ACHS Clinical Indicator PDF
ACHS Clinical Indicator PDF
ACHS Clinical Indicator PDF
Determining the Potential to Improve Quality of Care 5th Edition. ACHS Clinical Indicator Results for Australia and New Zealand 1998 2003 The Australian Council on Healthcare Standards Health Services Research Group University of Newcastle Published by the ACHS November 2004 5 Macarthur Street Ultimo NSW 2007 Copies available from the ACHS Publications Service Telephone 02 9281 9955 Facsimile 02 9211 9633 E-mail [email protected] Electronic version available at www.achs.org.au The Australian Council on Healthcare Standards This work is copyright. Apart from any use as permitted under the Copyright Act 1968, no part may be reproduced by any process without prior written permission from the Australian Council on Healthcare Standards. Requests and enquiries concerning reproduction and rights should be addressed to the Chief Executive, The Australian Council on Healthcare Standards, 5 Macarthur Street, Ultimo NSW 2007. Previous volumes in this series: Hospital Wide Medical Indicator Data: Quantitative and Qualitative Results 1993 Measurement of Care In Australian Hospitals 1994: Hospital-Wide Medical Indicator results and comparisons to 1993 results Measurement of Care in Australian Hospitals 1996: Obstetrics and Gynaecology Indicators and Hospital-Wide Medical Indicators Measurement of Care in Australian Hospitals 1997: Anaesthesia, Day procedures, Emergency Medicine, Hospital-Wide, Internal Medicine, Obstetrics and Gynaecology and Psychiatry Indicators Determining the Potential to Improve the Quality of Care in Australian Health Care Organisations: Results from the ACHS Clinical Indicator Data 1998 and 1999. Determining the Potential to Improve the Quality of Care in Australian Health Care Organisations 2nd Edition: Trends in Quality of Care: Results of the ACHS Clinical Indicators 1998 2000 Version 1 Determining the Potential to Improve Quality of Care 3rd Edition. ACHS Clinical Indicator Results for Australia and New Zealand 1998 2001 Determining the Potential to Improve Quality of Care 4th Edition. ACHS Clinical Indicator Results for Australia and New Zealand 1998 2002 This edition: Determining the Potential to Improve Quality of Care 5th Edition ACHS Clinical Indicator Results for Australia and New Zealand 1998 2003 ISBN 1 876987 49 9 (Paperback) ISBN 1 876987 50 2 (Web)
Acknowledgments
Acknowledgments
The Australian Council on Healthcare Standards (ACHS) Performance and Outcomes Service wish to thank the key people involved in the development and review of the clinical indicators. Working Party membership includes representatives from the speciality Colleges, Associations and Societies, health care providers, consumers, the National Centre for Classification in Health (NCCH), Quality Health New Zealand, the Health Services Research Group at the University of Newcastle and staff from the ACHS. The ACHS would also like to thank the participating health care organisations for their ongoing commitment to the collection of data that forms the content of this report. A special thank you and acknowledgement to the Performance and Outcomes Service Clinical Director, Dr Chris Maxwell, who provides clinical guidance and advice to the Service. Also, thank you to Ms Heidi Edwards and Ms Cynthia Kiu for their continued support in the collection and aggregation of the clinical indicator data.
Authors:
Robert Gibberd is Associate Professor, School of Medical Practice and Population Health, Faculty of Health, University of Newcastle. His current research interest is the measurement of the quality of health care. Stephen Hancock has postgraduate qualifications in mathematics and statistics and has worked for twenty years in acute hospitals as a registered nurse. Kay Richards is the Team Leader of the ACHS Performance and Outcomes Service. Kay has extensive experience in management in both the public and private systems, holds a Masters degree in Health Services Management and is a registered nurse with a Certificate in Intensive Care. For further information contact: Ms Kay Richards Team Leader, ACHS Performance and Outcomes Service Telephone: 02 9281 9955, E-mail: [email protected]
Contents
Contents
Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Executive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Summary of Key Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Summary of Results for Each Set of Indicators. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Version 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Version 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Emergency Medicine Indicators Hospital in the Home Indicators Hospital-Wide Clinical Indicators Infection Control Indicators Intensive Care Indicators
Ophthalmology and Excimer Laser Indicators Oral Health Indicators Paediatric Indicators Pathology Indicators
Version 3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
Foreword
Foreword
This is the 5th edition of Determining the Potential to Improve Quality of Care which presents the findings of clinical indicator data reported to the Australian Council on Healthcare Standards from health care organisations across Australia and New Zealand for the year 2003. The report also identifies trends from 1998 to 2003, providing a unique six-year review of where improvements in clinical care have occurred and areas where further improvements are required. As well as assisting individual health care organisations to target their improvement activities, the report is potentially of value to governments and the health industry in determining where improvements could be made in the delivery of health care at a national level. For the first time the report is presented in two sections. The Introduction, Executive Summary and Summary of Key Findings are presented in a printed medium, with complete indicator results found on an attached CD. This report is provided to key health industry bodies and all ACHS members. It can also be downloaded from our website (www.achs.org.au). The annual publication of this information is an important method by which ACHS contributes to the continuing improvement of quality and safety in health care in Australia and New Zealand.
Introduction
Introduction
Objective of this report The Australian Council on Healthcare Standards is pleased to provide this fifth edition of the report on the quality of health care. The objective of the report is twofold. The first is to present the findings of the clinical indicator data reported to the ACHS from health care organisations (HCOs) across Australia and New Zealand for the year 2003. Secondly, the report identifies trends from 1998 to 2003, providing a six-year review of where improvements in clinical care have occurred and where areas for improvement are required. How does the collection of clinical indicators impact on clinical care?
to be a major barrier to change. Much of our information comes from reviewing medical records, which is costly and time consuming. Evidence suggests that we need the data from multiple sources, including incident monitoring systems, routine administrative data sources and the use of screening tools to pro-actively identify areas that may cause harm. B Barraclough Maximising national effectiveness to reduce harm and improve care Fifth report to the Australian Health Ministers Conference 2004 The above quotes span 90 years, all with a similar message: our health systems need data to monitor what they are doing well and to determine what they can improve. It may seem from the above statement by Professor Barraclough that the hospital data systems have not made improvements since the days of Doctor Codman in Boston (1914) or our own surgeon from Newcastle, Doctor Smyth (1959). However, this would be incorrect, as substantial progress has been made in many areas. Australia was one of the first countries to carry out a nationally representative study of adverse 1 events in hospitals , which was followed by a similar 2 study in New Zealand . Both studies found adverse events were associated with more than 12% of admissions. Governments have required all public and private hospitals to maintain computerised files on all admissions (the hospital in-patient data) and have encouraged hospitals to provide clinical indicator data to the ACHS. Hospitals are increasingly being asked to record critical incidents. Thus, the situation has improved and this has been made possible partially through the use of computers and better statistical methods. However, as with all data, it is not possible to always measure the aspects that we would like. Thus, these studies and developments represent a continuing attempt to fulfil the objectives of Doctors Codman and Smyth. Assessing quality of care has become increasingly important to health care providers, purchasers and regulators and in an era of assessment and accountability in health services it is important to be 3 able to assess quality. The media coverage of cases of injury or death as a result of the health care provided has increased the publics concern about safety in hospitals. This concern may or may not be warranted, but until measures of safety are monitored over time it will not be possible to identify the magnitude of the problem or where patient care can be improved. Clinical indicators are one tool that can assist in measuring the performance and outcomes of care and services provided. Indicators can describe the performance that should occur for a particular type
If the financial records of any hospital were audited as casually and as ineffectively as in the quality of its patient care, the Administrator and the Governing Board would probably land in gaol. Or barring such as unfortunate happening, the least they could hope for would be financial chaos, unpaid bills, and a richly deserved reputation for business incompetence. Actually there is much more reason to do a continuing and adequate medical audit in a hospital than there is to keep meticulous and informative financial records. The medical audit deals with the life and health of people: the financial audit is concerned only with money. John Smyth, Surgeon, Royal Newcastle Hospital NSW. The Medical Journal of Australia Vol.1 No.10.1959
Wilson R. McL., Runciman W.B., Gibberd R.W., et al The Quality in Australian Health Care Study. The Medical Journal of Australia (1995) 163(9). 458 471. Davis P, Lay-Yee R, et al. Adverse Events in New Zealand Public Hospitals: Principal Findings from a National Survey. Ministry of Health, Wellington, New Zealand. (2001) Clarke A, Rao M Developing Quality Indicators to assess quality of care Quality& Safety in Health Care Volume 13 Issue 4 August 2004 pp. 248-249 Determining the Potential to Improve Quality of Care 7
Introduction of patient or related health outcome and then evaluate whether the patient care is consistent with 4 the evidence-based standards of care . For example, the Australasian Triage Scale (ATS) developed by the Australasian College for Emergency Medicine indicates that patients who meet a category 1 on the ATS should be seen immediately and this should occur 100% of the time. Data provided to the ACHS suggests that on average this is achieved 98.6% of the time in Australian and New Zealand HCOs, whereas the better performing organisations achieve a rate of over 99.9%. Understanding where there are opportunities for improvement can assist clinicians, health policy planners, and governments to direct resources and research capabilities to identifying solutions and actions for improvements. Fortunately, during the last 15 years, the ACHS together with the Australian and NZ health care systems, colleges and HCOs have created an environment that has encouraged the measurement of indicators of quality. This has not been an easy task and this report reflects some of the lessons learnt which can be summarised as follows. covered by the 245 individual indicators for 2003. Each year specific sets are reviewed and new versions are introduced for the following collection period. Key stakeholder groups collaborate in a Working Party to conduct the review. Working Party membership includes representatives from the speciality Colleges, Associations and Societies, health care providers, consumers, the National Centre for Classification in Health (NCCH), Quality Health New Zealand, the Health Services Research Group at the University of Newcastle and staff from the ACHS. Based on statistical analysis of the indicator results, feedback from users and input from clinicians, an indicator set is reviewed to ensure:
There have been many attempts to define quality and many frameworks developed to measure performance. Tier 3 of the National Health Performance Framework, developed by the National Health Performance Committee, has nine dimensions to measure health system performance. The ACHS has linked the clinical indicators to those dimensions and in 2003, 55% of the indicators measured the safety dimension and 43% measured the effectiveness dimension. More than half of the indicators measured an association with an adverse event. How are the clinical indicators developed? There are 20 clinical speciality areas or sets that are
4 8
Mainz J., Defining and classifying clinical indicators for quality improvement International journal for Quality in Health Care 2003: Volume 15, Number 6: pp 523-530 Determining the Potential to Improve Quality of Care
Introduction How does improvement occur? Three components are required: the use of data, the QI tools to improve processes within the health system, and support from staff to change the culture. There is no one measure of quality, and the ACHS data complements the hospital in-patient data that can be used to detect variation in safety (deaths, unplanned return to operating theatre, the use of ICD codes that indicate an adverse event may have occurred), efficiency (average length of stay) and appropriateness (admission rates for elective surgery). The ACHS indicators focus on clinical areas involving processes that apply to the specialty involved. Although the health system is often seen to be in the midst of chaos, the ACHS indicators for the last six years have identified areas where improvements have been made, areas which have deteriorated and areas where there has been little or no improvement. Indicators that fall into these three categories are discussed in the Summary of Key Findings.
Executive Summary
Executive Summary
This is the 5th edition of the ACHS publication Determining the Potential to Improve Quality of Care. It contains the results for the ACHS indicators from 1998 2003, a six-year period. The information is relevant to policy makers, medical colleges, health care practitioners and providers, quality managers and surveyors. The ACHS is thankful to the staff and HCOs for providing time and resources to collect these data. Australia is one of the few countries that can report on trends in clinical indicator data, and this achievement is only possible with support from the staff in the HCOs. This report presents data for 245 clinical indicators. The results are presented in the main results section as Tables, Figures and a brief summary for each indicator. These individual reports on each indicator are then summarised by identifying 54 indicators of major interest. This is done by classifying the indicators into four categories: In this executive summary we cannot discuss all the above indicators, but would like to note the large proportion of process measures, which are in principle under the direct control of HCOs and staff. Those that have deteriorated or failed to improve are: Delays in emergency departments (triage times); Low rates of plans for catch-up immunisation; High variation in the use of episiotomy; Decline in vaginal births after a caesarean section birth; Post-operative review of patients by Anaesthetist; Waiting time to commence radiotherapy treatment; Access block to wards; Access block to the ICU; Turn-around-time for serum/plasma potassium reports; Surgical site infection rates; and Written asthma discharge plan. The process indicators that have shown an improvement are: pre-anaesthetic consultation; compliance of anaesthesia records with ANZCA standards; failure to arrive for a day only procedure; delays in emergency departments for the most urgent cases; timely thrombolysis in AMI patients in the emergency department and department of medicine; documented CT scan following stroke; poorer rates for timely haemoglobin reports have improved; functional assessment of rehabilitation patients; initial functional assessment of rehabilitation patients, functional assessment of rehabilitation patients on cessation; rehabilitation planning; and rehabilitation discharge planning. As can be seen from the diversity of processes that have been identified above, there are considerable opportunities to identify scientifically the factors that caused an improvement or deterioration. By identifying these factors, the health system can also learn from them. Ultimately, scientific studies that identify the important factors for change may be seen as an essential component for those who aim to improve the quality of health care. HCOs, by continuing to provide these data are helping to make these studies a reality.
Improved performance during 1998 2003 Deteriorating performance during 1998 2003 Failure to improve during 1998 2003 and Indicators that are of concern.
The remaining indicators, are of lesser concern or do not have enough data to determine whether they should be in this summary. Each indicator of interest is then categorised into whether the measure is a process or an outcome measure, and the proportion of indicators in each category are given in the Table below. Table 1. Proportion of indicators identified as falling into the above categories Category Process Outcome Improved 12 13 Deteriorating 6 2 Failure to improve 9 2 Area of concern 9 1 Total 36 18 Total 25 8 11 10 54
Each of the 54 clinical indicators has an aspect that suggests that it should be studied further, either by obtaining additional data and research or by using an in depth analysis by a panel of relevant experts. The research task is to explain the trends that have been identified, or the variation in rates between HCOs that have been documented. These data provide the motivation to ask what we can learn from these trends or differences between rates, and whether any identified causes can be used to improve the processes in HCOs.
10
Outcome Indicators There has been improvement in several outcome measures, and the causes of this may be less easy to determine. However, where possible, it would be appropriate to document changes in surgical or medical practice that could have brought about these improvements.
Outcome indicators The proportion of inpatients having surgery with a post-operative length of stay equal to or greater than seven days who develop post-operative pulmonary embolism has increased to 0.6% from a previous rate of 0.4%. Twenty percent of HCOs now have rates above 0.9%, an increase from 0.6% in 1998. The rate of injury to the ureter or bladder during an abdominal or vaginal hysterectomy has increased from 0.6% to 1.0% during the last four years.
Summary of Key Findings improved to 62%, but the better HCOs have rates of over 99% and hence there is the potential to significantly improve this process, which is required for auditing the morbidity associated with anaesthesia. The cancellation of booked procedures in day only clinics because of administrative or organisational reasons has not improved and remains high for the public sector at 1.5%. For the 110 reports from the public HCOs, this resulted in over 3,500 cancellations in 2002 and 2003. Providing an appropriate discharge plan to patients with acute asthma has been shown to improve the control of asthma. The mean rate for providing written plans has remained below 60%, with a large variation between the rates for HCOs. The better HCOs have rates above 80% while the poorer had values below 20%. Births are one of the more frequent types of hospital admission. However, the indicators for quality of obstetric care show large variations between HCOs that should be of concern given the quarter of a million births each year in Australia and NZ. The induction of labour for reasons that are not appropriate varies from 4% to 14% with a mean of 9%. When expressed as a rate for all inductions the variation is from 22% to 48%, a mean of 34%. These rates have not shown any trend, and given the high rates for this inappropriate procedure an evidence based medicine review of criteria to be used is recommended. The large difference between the public and private sectors (7% and 14%) is also of major concern. Catch-up immunisation in infants whose immunisations are not up-to-date when they are admitted to hospital is an important way of reaching children who have not been immunised through the primary care system. The rates for immunisation (or planned immunisation) were 40%, with the poorer rates being below 27% and the better rates above 55%. Given the importance placed on immunisation programs in Australia, these low rates reflect poor processes within the hospital systems. The proportion of serum/plasma potassium reports which have a turn-around-time less than one hour is low (60% to 82%, depending on urgency) but also varies considerably between HCOs. The same issues also were found with delays in carrying out haemoglobin tests. and the more serious cases continue to CABG.
Areas of Concern
Process indicators The proportion of patients who were unable to be admitted to the ICU because of inadequate resources has remained at about 5% for the last six years. The better HCOs had low rates of 0.2%, compared to the poorer HCOs with rates of over 8%. The public rate was approximately ten times the private rate in 2002 and 2003. The causes of the considerable variation between hospitals may be used to ensure that resources are more equitably distributed. Asthma is a common condition for hospital admittance. The rate of documented assessment of severity was 90% on initial presentation and 86% for a subsequent assessment. The formulation of an appropriate discharge plan had a low rate of 55%, with the lower and higher rates being 21% and 78%. Elective surgery for insulin-treated diabetic patients requires an alteration of the treatment schedule. In particular, to reduce the risk of hypoglycaemia, at least four blood glucose measurements should be taken on the first post-operative day. There appears to have been no improvement in the overall rate since 1998 and in 2003 the rate was 85.5%, with the better rates being above 95%. Psychotropic doses and concurrent medications can result in harm to the patient if they are not appropriate. The proportion of patients on two or more psychotropic medications from one subgroup category at discharge is 25%, or one in four. The variation in the treatment of mental health patients should warrant further investigation. Four process measures of particular concern are: E.C.T., seclusion, assault and time to prepare discharge summary. Evidence on the appropriate proportion of E.C.T. treatments given during a defined course is not well established, and the data indicate that clinical practice varies. The rate at which mental health patients have more than twelve treatments of E.C.T. during a course ranges from twenty percent of HCOs being less than 2% to twenty percent of HCOs being more than 12%. A review of the recommended maximum is required. A similar situation exists for the use of seclusion: the corresponding proportions for patients
Determining the Potential to Improve Quality of Care 13
Outcome indicators The proportion of patients who die in hospital after having a coronary artery bypass graft (CABG) operation has declined during the years 1998 2002, but showed an increase in 2003 from less than 2.0% to 2.1%. This increase was seen in elective cases and also in those over 70 years. The causes of this should be reviewed, as it could be related to casemix; as the increase in angioplasty may remove the less serious patients
Summary of Key Findings having seclusion are 4% and 16%. There are large differences between the States. The rate of patient assault is around 4% but the poorer rates were above 7%, whereas the better rates were below 0.5%. The proportion of patients with a final discharge summary recorded in the medical records within 2 weeks of discharge had rates of 67% or less for the poorer HCOs, while the better rates were 98% or more. Outcome indicators Hospital acquired infections were reported for the first time in 2003. Surgical site infection rates ranged from 0.2% to 4.4%, the later being for colectomy. Central line associated blood stream infections had rates for different locations that were less than 1%. The highest rate of blood steam infection was for haemodialysis, in centrally inserted (both cuffed and non-cuffed) dialysis lines.
14
Summary of Results for Each Set of Indicators from zero to 0.9%. There was no clear relationship between infection rate and insertion point (central or peripheral). While the daily rate is less than one percent, the rate for an individual may increase substantially as the proportion of central line days for that patient increases. The remaining four indicators measured the proportion of patient days in ICU where a central line was utilised. More than 60% of patient days were associated with central lines (centrally inserted 57% and peripherally inserted 6.75%). Indicators 3.1-3.5 related to blood stream infections in dialysis months for five different access types. The rates were less than 0.02% per patient month for AV fistula, synthetic graft and native vessel graft. Rates for the centrally inserted lines were 2.26% per patient month for non-cuffed dialysis line (temporary) and 3.45% for cuffed (semipermanent) dialysis lines. Intensive Care Indicators Version 2 Indicators 1.1 and 1.2 are measures related to the utilisation of patient assessment systems. The average rate for entering adult patients onto the National Patient Database has been around 90% since 2000. Indicator 1.3 gives the proportion of HCOs participating in the participating in ARCCCR survey. The proportions were 95% and 84% in 2002 and 2003 respectively. The proportion of patients who were not admitted to ICU because of inadequate resources has remained at a high rate of seven percent for the public sector. Determining the reasons why the demand and supply are not better matched may help to reduce this figure. The rate for an unplanned readmission to ICU is between 1.4% and 2% since 1998. There was no association between the rates of refusal of admission to ICU and the unplanned readmission to an ICU. Internal Medicine Indicators Version 3 There are 19 indicators for internal medicine, which cover eight specific aspects of care, ranging from cardiology to thoracic medicine. There are nine process indicators and these generally have the larger variation between the HCOs. In particular, for geriatric medicine the proportion of patients for whom there is documented assessment of mental function on admission, CI 4.1, has a high compliance rate of more than 99% for one fifth of HCOs, but the lower one fifth have rates of less than 66% for their patients. In principle, the current average of 84% could be increased to 99% but there has been no improvement in the rate during the years 1999-2003. It is accepted that patients with AMI should receive thrombolysis within one hour of presentation to the hospital. The average rate has remained between 75% and 80% during the last four years. The 20th and 80th centiles were 66% and 91% in 2003, suggesting that a more acceptable figure could be achieved. The proportion of patients who had a coronary bypass operation within twenty-four hours of having the procedure PTCA (with or without stenting) in the same admission has declined from 0.7% to 0.3%, indicating that angioplasty is now a safer procedure. The single process measure within the six indicators
16 Determining the Potential to Improve Quality of Care
for gastroenterology is CI 3.3, the proportion of patients admitted for haematemesis and/or melaena with a blood transfusion for whom a surgical staff member has been notified of the patients condition. There has been no sustained improvement in this indicator. In the area of neurology, the proportion of inpatients with a diagnosis of stroke that had a documented CT scan has improved from 76% to 83%. One fifth of HCOs are below 73% however, with the better performing fifth of HCOs above 90%. A documented discharge plan for asthma patients is believed to improve the management of that illness but the average rate remains at about 55%, with the centiles showing large differences of 20% to 78%. Rates of objective assessment have not increased since 2000. The rates for initial and subsequent assessment were both around 90%. Mental Health Inpatient Indicators Version 4 There are more indicators for this area than in the other sets: a total of 20, of which nine are process indicators. The following process measures all had large potential gains: aspects of care that relate to allocating a diagnosis within the admission and at discharge; carrying out a complete physical examination; providing a discharge summary and completing the medical record within two weeks of being discharged. This was because some hospitals were achieving rates close to 100%, while the poorer HCOs were often less than 80%, implying that a fifth of patients in these HCOs are not provided with the basic diagnostic and reporting requirements. The rates for the indicators relating to discharge summaries were less than 80%. Other process measures are concerned with the clinical aspects of care, and relate to the use of two or more psychotropic drugs at discharge, and the use of seclusion. In both these measures, there is large between hospital variation, often involving significant differences between the States. There has been an increase in the rate of seclusion for more than four hours, which is unexpected since the desirable rate is low. This seclusion indicator has increased from 20% to 27% and the 80th centiles have also increased to 57% in 2002 and 51% in 2003. In regards to the outcome indicators, the rate of unplanned readmissions within 28 days remain at around 4% or 5% and the mortality rate for inpatients has halved over the six year period to 0.13%. The indicators relating to suicide, assault and self harm have not changed substantially over the six year period. Mental Health Community Based Indicators Version 1 This is the first year these three outcome indicators were collected. More than one third of clients had five or more contacts with the outpatient service, 8% were admitted only once and 1.6% more than once. Fourteen of the 16 reporting HCOs were public. Obstetrics and Gynaecology Indicators Version 4 There are eleven indicators for obstetrics and two for gynaecology. The two aspects of obstetrics that show considerable clinical variations are the proportion of patients undergoing induction of labour for reasons
Summary of Results for Each Set of Indicators other than those defined as appropriate and the proportion of primiparous patients requiring surgical repair as a result of tearing and / or an episiotomy. The proportion of inappropriate or unnecessary inductions as a proportion of all deliveries was 8.7% percent, with the centiles ranging from 4% to 15%. While much of this variation is explained by the differences between the public and private sectors of 7% and 14%, within each sector there is unexplained variation. If the average could be shifted to the 20th centile of 5%, this would reduce the proportion of inductions per annum in Australia by approximately 10,000. Similarly, there is a large variation in the proportion of patients requiring surgical repair after their first birth, with an average of 71% and the centiles ranging from 60% to 80%. If the average could be reduced to the 61%, there would be about 10,000 more intact perineums per annum in Australia. There has been a major change in the proportion of patients delivering vaginally following a previous primary caesarean section in Australia. The mean rate has declined from 24% to 16%, with the centile rates declining by more than 5%. Although there are studies ongoing to determine the appropriate rate for vaginal birth after caesarean section, it would be an area where an appropriateness study could be carried out. The proportion of patients receiving a blood transfusion during or after a hysterectomy has doubled over the last three years, and the metropolitan HCOs had a high rate of 6% in 2002. The 20th centile has remained stable at about 2%, and hence there is reason to determine the causes of the increase and the higher rate in the metropolitan HCOs. The rate of injury to the ureter or bladder during an abdominal or vaginal hysterectomy has increased from 0.6% to 1.0% during the last four years. Ophthalmology and Excimer Laser Indicators Version 3 There are fourteen indicators for ophthalmology. The indicators are primarily outcome measures. The numerators are relatively small, since they focus on poor outcomes such as re-admission rates within 28 days, infections and re-treatment. The rate of unplanned readmission days after cataract surgery has declined from 0.51% in 1998 to 0.32% in 2003. The rates of unplanned overnight stay and anterior vitrectomy in cataract surgery, first collected in 2003, were 0.65% and 0.84%. In glaucoma surgery the rate of re-admissions within 28 days and the rate of patients having LOS greater than three days averaged 1.6% and 3.7% respectively over the six years to 2003 and both varied considerably from year to year. Readmission due to infection has been less than 0.2% since 1998. In retinal detachment surgery the rate of unplanned readmission in 28 days over the six year period was 4.5% and varied considerably from year to year. The readmission rate was usually less than 0.2%. The length of stay indicator for retinal detachment had a decline over the last six years, the proportion with length of stay greater than three days declined from 3.5 to 2.5%. The rate of revision within 28 days, first collected in 2003 was 3.63%. The re-treatment rates in surface and non-surface ablation were 5.45% and 8.54% respectively. Infection rates in excimer laser were low, 0.23% over the six year combined. Oral Health Indicators Version 1 There are 12 indicators and the results were obtained from between one and 10 HCOs in 2002 and in 2003. Paediatric Indicators Version 3 There are five indicators for paediatrics, two of which are measures of immunisation status. Whether catch-up immunisation was given or was planned for children without immunisation was only 38% in 2003, with the 20th and 80th centiles ranging from 27% to 55%. This indicates that there is an opportunity to increase immunisation rates for children who have not been previously immunised. The re-admission rate for children with asthma is a relatively high 4%, with the 20th centile being not much lower at 3%. The lack of any trend or a low centile suggests that it may be hard to reduce this rate. Pathology Indicators Version 2 The indicators for Pathology are primarily process measures, focussing on the turn-around time for the tests. There are no clear trends. There are large differences in the centiles for most of these indicators and the outlier gains are high for haematology and chemical pathology, due to the large sample sizes. As the better HCOs are able to achieve high rates of compliance with the indicators, there are opportunities for determining how these rates are achieved, and use the findings to improve the rates for all HCOs. Radiation Oncology Indicators Version 2 There are six process indicators for radiation oncology. The proportion of HCOs (all public and metropolitan) reporting each indicator ranged between four and 12 in 2003. The proportion of patients waiting more than 21 days for the commencement of radiotherapy treatment has increased from 10% to 36%, the 20th centile from 2% to 17% and the 80th centile from 23% to 51%. These three trends indicate a significant deterioration in the access to radiation therapy in Australia, with the decline seen in most of the HCOs. The rate of CT planning was 70% and the better rate was 87% indicating that the rate could be improved. The trial participation rate for 3,196 patients at five HCOs was 11.6%. The proportion of patients having complete follow up for glottic, prostate and breast cancer was around two thirds. Radiology Indicators Version 2 There is one process measure for radiology involving turn-around-time for non-procedural non-urgent plain radiographs. There were large differences between the rates for the HCOs, and large centile and outlier gains. The proportion of patients for whom there is documented evidence of pneumothorax and/or
Determining the Potential to Improve Quality of Care 17
Summary of Results for Each Set of Indicators haemothorax requiring intervention after undergoing percutaneous trans pleural biopsy of the lung or mediastinum has declined from 20% to 12%. The centiles have also declined by about 8%, indicating that the morbidity from this procedure has been significantly reduced. Rehabilitation Medicine Indicators Version 3 There are seven rehabilitation indicators of which five are process measures concerned with whether functional assessments were carried out, rehabilitation plans established and appropriate discharge plans prepared. These indicators had significant improvement in the mean values with the rates for not having these items of care more being reduced to one quarter of the 1998 rates over the last five years. The proportion of deaths within the rehabilitation unit/facility for rehabilitation patients has declined in 2003. Surgical Indicators Version 3 There are 19 indicators reported for surgery, and except for the reporting of the mean operating time and the weight of tissue removed for patients having a transurethral resection for benign prostatomegaly, the indicators are primarily outcome measures. Seven measures showed an improvement during the five-year period. The proportion of children who undergo appendicectomy with a normal histology has declined from 21% to 16%. There has been a gradual decline in the proportion of patients being administered a blood transfusion following transurethral prostatectomy, from 4.5% to 3.6%. The rates of a postoperative in-hospital infection in primary THJ replacement have remained close to half the 1998 rate, which was 2.6%, in subsequent years. The mean rates for bile duct injury while having a laparoscopic cholecystectomy procedure have shown a slight reduction from 0.51% to 0.29%. The centiles have also declined by a similar amount. The rate of patients with a new neurological deficit following a neurosurgery procedure more than halved in 1999 and have remained at less than 2% since then. The rate of stoke following carotid endarterectomy (within the same admission) declined from 2.3% in 1998 and has remained around 1.6% since then. Finally, the proportion of tonsillectomies that had a significant haemorrhage declined by half from 1.3% to 0.63%, and the centiles also improved. The other measures had no evidence of trend. The mortality rates for cardiothoracic coronary artery grafts, CAGS, have not declined since 1998. The overall mortality rate remains close to 2%, the rate for elective CAGS remains close to 1.5% and in patients aged 71 years or more the rate was 3.4% in 2003. Non-superficial neurosurgical infection in hospital rates averaged 1.4% since 2001. The mortality rate in abdominal aortic aneurism (AAA) was 4.8% in 2003, higher than in previous years however there was no difference between the better and poorer rates. With the exceptions of CAGS where age exceeds 70 years and in neurosurgical infection there were no major variations between the rates of the HCOs for these measures, but these outcome indicators do provide clinicians and patient information on the likelihood of a poorer outcome associated with these surgical procedures.
18
CI Field
Numerator
Denominator
Description Clinical Indicator number. Year of collection. Proportion of Health Care Organisations submitting the indicator each year. The proportion of events that occurred. The proportion of patients, admissions, procedures or patient days at risk of the event. The overall rate for all HCOs combined, expressed as a percentage. 20% of HCOs have a rate that is at or below Rate (20)%, the 20th centile rate. The rate is expressed as a percentage. 20% of HCOs have a rate that is at or above Rate (80)%, the 80th centile rate. The rate is expressed as a percentage. One of these two rates, the 20th or 80th centile rate, will usually be referred as the better rate depending on which is the more desirable (20th for a low rate, 80th for a high rate). The other will be referred to as the poorer rate. The centile gains are a measure of the potential gains that would be obtained if the overall rate were moved to the better rate. The stratum gains are a measure of the potential gains that could be obtained if overall rate were moved to the rate of the best stratum (Public/private, Metropolitan/rural or State). If any HCO has a poorer rate that is more than three standard deviations from the overall rate then that HCO is referred to as having statistically significantly high (or low) rate. The outlier gains measure the benefits of improving the rate of each of the outlier HCOs to the overall rate.
19
Explanation The results table contains three rates referred to as Rate %, Rate (20)% and Rate (80)%. They are the overall rate, the 20th centile rate and the 80th centile rate respectively. Twenty percent of HCOs have rates equal to or greater than the 80th centile rate. Because it is desirable that a high rate be achieved this rate is referred to as the better rate. The better rate can be considered as achievable because twenty percent of HCOs are performing at or better than this rate. Twenty percent of HCOs have rates equal to or less than the 20th centile rate. This rate is referred to as the poorer rate. The difference between the overall rate and the better and poorer rates provides a measure of the variation between HCOs, with 60% of rates lying between the 20th and 80th centiles. In 2003 the overall rate was 73% and hence approximately one in four patients were not seen within 10 minutes. The poorer rate (20th centile) was 65.3% and the better rate (80th centile) was 91.0%. The overall rate, the better and poorer rates are plotted against year in Figure 1.2. The plot demonstrates that:
During the period from 1998 to 2003 the better rate, the poorer rate and the overall rate have varied little. The difference between the better and poorer rates has remained approximately constant at 25% since 1998.
Figure 1.2 ATSC 2 attended wihin 10 minutes
100 90 80 70 60 Rate % 20 50 40 30 20 10 0 1998 1999 2000 Year 2001 2002 2003 Overall rate, average rate
th th
80
Centile rate
Centile rate
There is little evidence of improvement during the period 1998 to 2003. For the better performing hospitals, one in every 10 patients is not seen within the recommended time frame. The difference between the better and poorer rates in 2003 was 26%. The three columns on the right of the results table contain three measures of the potential for improvement which are calculated from the data. They are centile, stratum and outlier gains.
In this indicator:
22
Glossary and Examples Centile gains represent the proportion of extra patients that would be seen in the specified time if the overall rate (73%) could be increased to the better rate (91%). The calculation is as follows: Centile gains = Sum of Denominators x (80th centile rate Overall rate) Stratum gains represent the proportion of extra patients that would be seen in the specified time if the overall rate (73%) could be increased to the rate of the better stratum (State, public/private, metropolitan/rural 83.9% see Table 1.2.) Outlier gains represent the proportion of extra patients that would be seen in the specified time if those HCOs that had unusually low (referred to as statistically significantly low) rates were to achieve the overall rate (73%). These measures can assist in prioritising indicators for further investigation. Where variation between HCOs is large then centile gains tend to be large. If the average of all HCOs could be brought to the 80th centile, an additional 45,731 patients (see centile gains column) from the 254,122 presentations to the emergency department would have been seen within the specified time of 10 minutes. Most data are submitted from the public and metropolitan sectors. It can be seen from Table 1.2 that in 2003, 199,719 patients (approximately 80%) were from Metropolitan HCOs and almost 95% were from public HCOs. As might be expected, the emergency departments with the larger throughput had lower rates than the smaller units, (68% compared to 80%). There was significant variation between the States, with NSW and Victoria having the better rates. The differences between the public and private sectors and between metropolitan and rural HCOs were not statistically significant. The stratum gains in 2003 (Table 1.2) were largest for South Australia and New Zealand. If South Australia were to achieve the best state rate in Victoria of 78.7%, then 4,474 additional patients would be seen within 10 minutes. Note that the best state rate is more than 10% lower than the rate for the top 20% Table 1.2 ATSC 2 attended within 10 minutes No HCOs 66 74 27 29 13 12 6 28 31 18 18 7 10 8 81 80 87 99 148 158 20 21 Stratum rate % 77.6 77.3 70.4 76.2 68.2 64.2 55.9 80.3 78.7 80.2 73.3 55.3 66.7 66.1 75.2 72.4 71.6 74.9 74.1 72.4 78.9 83.9 Standard error 1.3 1.3 1.8 1.7 2.3 2.1 4.4 1.8 1.8 2.5 2.1 2.6 2.3 3.6 0.94 0.94 1.7 1.8 0.85 0.84 3.6 3.6 Stratum gains 4,008 2,961 4,474 1,712 1,650 4,894 3,349 1,294
Stratum NSW QLD SA TAS VIC WA NZ Other Metropolitan Rural Public Private
Year 2002 2003 2002 2003 2002 2003 2002 2002 2003 2002 2003 2003 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003
Numerator 58,652 58,210 28,652 34,668 16,773 19,841 3,924 34,755 31,609 16,744 22,676 11,596 16,530 6,813 135,065 144,653 40,965 40,760 166,084 174,476 9,946 10,937
Denominator 75,539 75,300 40,709 45,526 24,598 30,920 7,032 43,300 40,173 20,870 30,909 20,978 24,783 10,316 179,599 199,719 57,232 54,403 224,236 241,107 12,595 13,015
23
CI 2.4
he rates have decreased from 4.54% in 1998 to 3.64% in 2003. The rate of the poorer performing HCOs has decreased from 5.78% to 4.15% in 2003. There are small potential gains in 2003 of 34. There were no HCOs with statistically significantly low rates and there were no statistically significant differences between the strata. Hence no outlier or stratum gains appear in the table. The overall rate, the better and poorer rates are plotted against year in Figure 2.4. It is desirable that the rate of blood transfusion following prostatectomy be minimised. Hence the better rate is given in this case by the 20th centile rate, the bottom trend line in the plot. The plot demonstrates that the difference between the better and poorer rates has diminished since 1998 and both the poorer rate and the overall rate have decreased. (That is to say, the variation between HCOs has decreased over the period.) The better rate has remained close to 3% during the entire period. Figure 2.4 TUR for benign prostatomegaly blood transfusion.
8 7 80 6 5 Rate % Overall rate, average rate 4 3 2 1 0 1998 1999 2000 Year 2001 2002 2003 20
th th
Centile rate
Centile rate
24
Year Description
Mean
Std
Median
Clinical Indicator number. Year of collection. Proportion of Health Care Organisations submitting the indicator each year. Overall mean for all contributing Health Care Organisations. The standard deviation of the averages reported by all contributing Health Care Organisations. The 20th Centile of the averages reported by all contributing Health Care Organisations. The median of the averages reported by all contributing Health Care Organisations. The 80th Centile of the averages reported by all contributing Health Care Organisations.
25
Explanation The data provided by the HCOs is the average length of stay and the proportion of patients in each HCO is not known. The median as well as the mean of the average length of stay are provided in the table. If the mean is substantially greater than the median then the distribution of means is skewed. Twenty percent of HCOs had an average length of stay less than the 20th centile and twenty percent had length of stay exceeding the 80th centile length of stay. The average length of stay has decreased from 2.00 days in 1998 to 1.67 days in 2003. A similar trend in the median occurred. Twenty percent of HCOs had an average length of stay exceeding 2.50 days in 1998 and in 2003 twenty percent had an average length of stay above 1.92 days (see 80th centile in Table 2.1). Western Australia reported the highest average length of stay in 2003 (Table 2.1). Table 2.1 Average length of stay for children admitted for Asthma
State
NSW QLD SA TAS VIC WA ACT, NT, NZ
No HCOs
28 20 6 5 18 12 6
ALOS
1.71 1.59 1.70 1.63 1.52 2.05 1.50
26
CI. 1.1 The rate of adverse drug reactions among non same-day patients reported to ADRAC. For the year 2003, there were 1,301 adverse drug reactions reported for almost 750,000 patients, a rate of 0.18%. There were large differences in rates between the HCOs, the higher rates were ten times the lower rates. The mean rate for private HCOs of 0.12% was approximately half the public rate, 0.22%. There was no significant relationship between the proportion of admissions and the rates of the HCOs. The majority of the 11 outlier HCOs were public whereas almost two thirds were private. The State differences were similar for both 2002 and 2003, with NSW having lower rates (Table 1.1). Table 1.1 Adverse drug reactions reported to ADRAC. Stratum NSW QLD VIC WA Other Year 2002 2003 2002 2003 2002 2003 2003 2002 2003 No HCOs 29 53 19 20 33 36 7 7 8 Numerator 199 287 121 221 352 519 19 185 255 Denominator 190,411 303,241 59,610 91,483 206,166 226,103 43,349 40,683 71,879 Stratum rate % 0.11 0.10 0.20 0.24 0.17 0.23 0.05 0.45 0.35 Standard error 0.03 0.03 0.06 0.05 0.03 0.03 0.08 0.07 0.06 Stratum gains 135 297 140 182
CI. 1.2 The rate of adverse drug reactions among non same-day patients reported within the HCO. The overall rates have declined and the proportion of HCOs providing data has doubled since 2000. There were 5,535 adverse drug reactions reported for 766,063 non same day admissions (0.72%). The private rate (0.48%) was approximately half the public HCO rate (0.93%) although there was considerable variation in rates in both groups. There was no significant relationship between the numbers of patients represented and the rates of the HCOs. Twenty seven of the 36 outlier HCOs were public whereas almost two thirds of contributing HCOs were private. Differences in case-mix may account for some of the variation in rates. This is supported by the fact that the HCO reporting the highest rates in each of the four years 1999 to 2002 treats cancer patients.
Determining the Potential to Improve Quality of Care 27
2.2
2.3
2.4
2.5 2.6
CI. 2.1 The rate of abnormal bleeding among non same-day separations receiving Warfarin. The rate of abnormal bleeding has declined significantly from the 1999 figure of 1.9% to 0.7% in 2003. The rates for the better and poorer performing HCOs have also decreased. The proportion of HCOs reporting a rate of zero has increased from 60% in 1999 to 79% in 2003. The rates for the poorer performing HCOs are three times the rates for the better performing HCOs. The results suggest that the proportion of episodes of abnormal bleeding in patients taking Warfarin could be decreased by approximately two thirds. The two outlier HCOs (public, rural) had rates of 4% or more and reported four and seven episodes of abnormal bleeding and contributed greatly to the metropolitan/rural differences (Table 2.1).
28
Adverse Drug Reaction Indicators Version 2 Table 2.1 Abnormal bleeding. Stratum Metropolitan Rural Year 2003 2003 No HCOs 12 14 Numerator 1 14 Denominator 1,304 836 Stratum rate % 0.19 1.49 Standard error 0.21 0.26 Stratum gains 10
CI. 2.2 The rate of cerebral haemorrhage among non same-day separations receiving Warfarin. Ninety percent of HCOs had no cases of cerebral haemorrhage since 2000. Three HCOs reported a single case of cerebral haemorrhage for patients taking Warfarin in 2003. CI. 2.3 The rate of INR/prothrombin reading greater than five among non same-day separations receiving Warfarin. There has been an improvement in the rate since 2000 (Figure 2.3). Both the overall rates and the rates of the better performing HCOs have decreased. The rate of HCOs reporting poorer performance is twice that of the better performing HCOs. The data suggest that the proportion of elevated readings of INR could be reduced by approximately one third. There were four outlier HCOs in 2003. Two of these HCOs were outliers in previous years. Figure 2.3 Warfarin INR/prothrombin greater than 5.
10
Rate %
CI. 2.4 The mortality rate due to an adverse reaction to Warfarin among non same-day separations. This is a relatively rare outcome for which there is no evidence of a decline in deaths. CI. 2.5 The rate of providing written drug information about Warfarin for patients discharged on hospital initiated Warfarin. Combining the results for the two years, half of HCOs had rates of 100%. This is a process for which all HCOs should aim to achieve rates close to 100%. CI. 2.6 The rate of Warfarin dosage review before the next dose when the INR result is above the therapeutic range. In the last two years two thirds of HCOs have reported rates of 100%, but twenty percent had rates below94% in 2003. As with CI 2.5, all HCOs should be able to achieve rates of more than 95%.
29
CI. 1.1 The rate of documented evidence of a pre-anaesthetic consultation. The rates increased from 79% in 1998 to 95% in 2001 and have remained at this level. The proportion of HCOs having a rate of 100% has increased from 28% in 1998 to 43% in 2003 and the proportion of HCOs with statistically significantly low rates has decreased from 20% in 2000 to 15% in 2003. This represents a significant improvement although there is potential to improve further. In 2002 and 2003, the rates were lowest in Queensland (Table 1.1). Table 1.1 Documented evidence of a pre-anaesthetic consultation. No HCOs 64 57 35 40 18 17 42 33 10 12 12 9 Stratum rate % 95.8 98.9 88.1 88.6 94.6 91.0 93.0 93.0 97.6 96.1 97.2 97.6 Standard error 1.87 1.34 2.24 1.67 4.44 2.79 2.51 1.65 6.14 4.16 3.10 2.30 Stratum gains 4,686 1,279 2,754
Year 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003
Numerator 76,831 69,654 49,123 40,333 13,486 14,754 41,411 43,132 7,276 7,024 28,366 23,378
Denominator 80,174 70,397 55,781 45,505 14,248 16,206 44,523 46,381 7,452 7,309 29,188 23,943
30
CI. 2.1 The rate of substantial compliance with ANZCA requirements for anaesthetic records. This indicator improved in the year 2000, but has failed to make any improvement since then. The better performing HCOs have reported rates close to 100%. The poorer performing HCOs reported rates of compliance of 80% or less (Figure 2.1).
100 90 80 70
Figure 2.1 Anaesthetics record compliance with ANZCA requirements. The proportion of HCOs reporting rates of 100% has increased from 22% in 1998 to 36% in 2003. This indicator measures a process that is under the control of each HCO and hence further improvement could be achieved. The proportion of HCOs with significantly lower rates has decreased from 35% in 2000 to 26% in 2003. Rates were lowest in Queensland in 2002 and 2003 (Table 2.1). Private rates were lower than the public rates (89% and 94% respectively).
1998 1999 2000 2001 Year 2002 2003
Rate %
60 50 40 30 20 10 0
Table 2.1 Anaesthetics record compliance with ANZCA requirements. Stratum ACT NSW QLD SA VIC WA Other Year 2003 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003 No HCOs 5 58 50 33 37 15 16 38 31 8 11 9 5 Numerator 6,482 56,850 42,735 32,616 37,931 9,737 16,202 36,403 36,576 6,397 6,215 24,154 16,564 Denominator 6,552 60,859 46,925 47,062 46,225 11,090 16,902 42,797 38,640 6,706 6,394 25,402 16,963 Stratum rate % 98.9 93.4 91.1 69.3 82.1 87.8 95.8 85.1 94.7 95.4 97.2 95.1 97.6 Standard error 6.52 3.03 2.44 3.44 2.45 7.09 4.06 3.61 2.68 9.11 6.60 4.68 4.05 Stratum gains 12,127 4,288
31
respiratory distress cardiac arrest core temperature < 35C severe pain unplanned stay > 2 hours
3.2
3.3
3.4
3.5
In relation to indicators 3.1 to 3.4, it is likely that some of the differences between HCOs are related to factors such as type of surgery. This is consistent with the slightly higher rates in the metropolitan and public HCOs. The indicators provide evidence that the rates for intervention by an anaesthetist have remained low and are probably related to the complexity of the surgery. Thus, in practice, the reductions in the proportion of these adverse outcomes may not be achievable to the degree indicated in the tables.
32
Anaesthics Indicators Version 3 CI. 3.1 The rate of patients having documented evidence of respiratory distress in the recovery period requiring intervention by an anaesthetist. The rate has declined steadily since 1998 (Figure 3.1). The rates for the better performing HCOs have remained at approximately 0.05%, less than one third of the overall rate. The rate of the poorer performing HCOs has remained at five times the rate of the better performing HCOs. The public rate is approximately twice the rural rate (Table 3.1). Figure 3.1 Recovery respiratory distress.
0 .4 0 0 .3 5 0 .3 0 0 .2 5 0 .2 0 0 .1 5 0 .1 0 0 .0 5 0 .0 0 1 9 98 19 9 9 2000 2001 Year 2 00 2 2003
Table 3.1 Recovery respiratory distress. No Year 2002 2003 2002 2003
CI. 3.2 The rate of respiratory or cardiac arrest in recovery. The rates have remained stable and low at 0.02% since 1998. CI. 3.3 The rate of patients having low core temperature in the recovery period. The rates have steadily decreased since 1998 (Figure 3.3). The rate in 2003 is almost half the rate in 1998. The rates of the poorer performing HCOs have decreased as well, but remain at almost double the average rate. This is an avoidable outcome and the large variation between HCOs indicates that the rates could be reduced by at least half. Forty three HCOs had statistically significantly high rates for at least three of their reports during the six year period. CI. 3.4 The rate of occurrence of severe pain in recovery requiring the intervention of an anaesthetist. Figure 3.3 Recovery core temperature < 35C.
2 .0
1 .5
1 .0
0 .5
The rates have declined from 0.75% in 1998 to approximately 0.5% in 2003. The better performing HCOs had rates less than 0.1% since 1998. The public rates were approximately three times the private rates in both 2002 and 2003 (Table 3.4). This appears to be a potentially avoidable outcome and the large variation between HCOs indicates that the rates could be reduced by at least half. Thirty HCOs had statistically significantly high rates for at least three times during the six-year period. Table 3.4 Recovery severe pain. No Year 2002 2003 2002 2003 Stratum rate % 0.77 0.86 0.26 0.26 Standard error 0.05 0.06 0.04 0.04 Stratum gains 1,904 2,358
33
Anaesthics Indicators Version 3 CI. 3.5 The rate of unplanned recovery stay exceeding two hours. The rates have remained at about 1.7% since 1998 and there were significant differences between the HCOs. Public rates were three times higher than the private rates in 2002 and 2003 (Table 3.5). Twenty public metropolitan HCOs accounted for almost all of the 10,517 outliers. CI 3.5 Recovery unplanned stay > two hours. Stratum Public Private Year 2002 2003 2002 2003 No HCOs 107 112 179 182 Numerator 7,904 14,004 4,977 5,469 Denominator 344,988 411,030 657,942 697,219 Stratum rate % 2.28 3.39 0.77 0.79 Standard error 0.15 0.26 0.11 0.20 Stratum gains 5,210 10,686
34
CI. 4.1 The rate of unplanned admission to ICU within 24 hours of a procedure. The overall rates have remained at about 0.2% since 1999, while the rates of the better and the poorer performing HCOs have declined (Figure 4.1). Public metropolitan rates are the highest (0.4%) but public/private and metropolitan/rural differences do not explain the variation between HCOs. There were seven outlier public metropolitan HCOs which accounted for two thirds of the 544 outlier events. Eleven HCOs had statistically significantly high rates at least three times during the six year period. Figure 4.1 Anaesthetics - unplanned ICU admission.
0 .4 0 0 .3 5 0 .3 0 0 .2 5 0 .2 0 0 .1 5 0 .1 0 0 .0 5 0 .0 0 1 9 98 19 9 9 2000 2001 Year 2 00 2 2003
Rate %
35
CI. 5.1. The rate of documented evidence of post anaesthetic review within 48 hours. The rates have increased since 1998 (Figure 5.1). The proportion of HCOs reporting rates of 100% has increased from 6% in 1998 to 44% in 2003. However, these results suggest that there still remains an opportunity to significantly increase the rate of post anaesthetic review from 62% to over 90%. A single rural public HCO accounted for almost three quarters of the 7,370 outlier cases. When this outlier was removed, the public rate was lower than the private rate.
100 90 80 70
Rate %
Table 5.1 Anaesthetics patient review within 48 hours. No HCOs 12 15 8 9 7 6 5 9 6 24 17 18 23 Stratum rate % 82.9 81.5 46.6 36.3 85.8 90.8 74.6 55.7 57.3 81.6 37.8 46.6 89.9 Standard error 7.39 7.07 9.84 6.92 12.7 15.4 14.9 10.5 9.61 5.23 5.73 4.92 6.71 Stratum gains 3,528 8,463 2,344 2,698 8,928 12,660
Year 2002 2003 2002 2003 2002 2003 2003 2002 2003 2003 2003 2003 2003
Numerator 14,299 12,111 4,537 5,645 5,038 2,833 2,492 4,799 4,615 20,005 7,691 13,605 14,091
Denominator 17,238 14,856 9,724 15,531 5,868 3,118 3,339 8,614 8,052 24,521 20,375 29,223 15,673
36
1.2
1.3
1.4
CI. 1.1 The rate of patients failing to arrive after being booked into a day procedure facility. The rates for failing to arrive have decreased since 1998 (Figure 1.1). The decrease has occurred in both the public and private sectors. The rates for the public sector have been approximately three times the rates of the private sector since 1998 (Table 1.1). The low rates for many HCOs suggest that the rate could be reduced further.
1 .5
Rate %
1 0
0 .5
37
Day Surgery / Endoscopy Indicators Version 3 Table 1.1 Failure to arrive No HCOs 125 117 191 202 Stratum rate % 1.5 1.6 0.52 0.45 Standard error 0.084 0.087 0.057 0.058 Stratum gains 2,246 2,703
CI. 1.2 The rate of booked day procedure patients having their procedure cancelled because of a pre-exiting medical condition The rates have declined slightly since 2001. The rate for the public sector is three times the rate for the private sector. The rate of the better performing HCOs is less than a quarter of the average rate in each sector. CI. 1.3 The rate of booked day procedure patients having their procedure cancelled because of an acute medical condition The rates have changed little since 2001. The rate for the public sector was higher than the private sector. CI. 1.4 The rate of cancellation of booked procedures for administrative or organisational reasons. This is the more common cause of cancellation of a procedure and the rates have remained at about 0.53% since 2001. The public sector rate was approximately ten times the private sector rate (Table 1.4). In both sectors the rate of the better performing HCOs is less than one third of the average rate. There were 52 public HCOs and two private HCOs having significantly high rates. Table 1.4 Cancellation administrative/organisational reasons No HCOs 109 110 191 204 Stratum rate % 1.6 1.5 0.18 0.15 Standard error 0.12 0.06 0.078 0.037 Stratum gains 3,072 3,097
38
CI. 2.1 The rate of unplanned return to the operating/procedure room. The rates, while low, have remained at about 0.04% since 1998. The average rates were about twice the rates of the better performing HCOs.
39
CI. 3.1 The rate of intended same-day patients who have an unplanned overnight admission. There has been a decline in the rates since 1998 (Figure 3.1). The rates for the public sector were twice the rate of the private sector (Table 3.1). The public and private sectors contributed approximately equally to the 5,364 outlier admissions. Figure 3.1 Unplanned overnight admission.
6
Rate %
Table 3.1 Unplanned overnight admission. No HCOs 136 125 217 231 Stratum rate % 2.9 2.8 1.4 1.3 Standard error 0.11 0.12 0.074 0.076 Stratum gains 3,971 3,734
40
CI. 4.1 The rate of unplanned delay in discharge from a day procedure facility. The rates have halved since 1998 (Figure 4.1), primarily due to the decline in the poorer performing HCOs from 0.96% to 0.40%. The rates for the better performing HCOs have not changed. The public rates were approximately twice the private rates but there remains the large differences between the higher and lower rates between HCOs (0.40% and 0.06% in 2003). Figure 4.1 Unplanned delay in discharge
1 .0
0 .8
Rate %
0 6
0 .4
0 .2
41
Only CI. 1.1 was collected in 2003. Results CI 1.1 Year 1998 1999 2000 2001 2002 2003 1998 1999 2000 2001 1998 1999 2000 2001 1998 1999 2000 1998 1999 2000 2001 1998 1999 2000 1998 1999 2000 2001 1998 1999 2000 2001 No HCOs 2 4 3 2 1 1 2 4 3 1 3 4 4 1 3 4 4 3 4 4 1 3 4 4 3 4 4 1 3 3 4 1 Numerator 7 14 15 17 2 0 7 15 16 18 11 28 40 33 11 28 42 11 28 40 32 9 28 42 9 27 37 28 9 21 34 25 Denominator 7 15 19 37 19 2 7 15 19 19 11 28 42 33 11 28 42 11 28 42 33 9 28 42 9 28 42 33 9 23 42 33 Rate % 100 93.3 78.9 45.9 10.5 0.00 100 100 84.2 94.7 100 100 95.2 100 100 100 100 100 100 95.2 97.0 100 100 100 100 96.4 88.1 84.8 100 91.3 81.0 75.8 Rate (20)% 100 93.3 58.0 1.26 10.5 0.00 100 100 72.1 94.7 100 100 95.3 100 100 100 100 100 100 95.3 97.0 100 100 100 100 96.4 84.7 84.9 100 46.6 75.4 75.8 Rate (80)% 100 93.3 95.9 88.2 10.5 0.00 100 100 94.4 94.7 100 100 95.3 100 100 100 100 100 100 95.3 97.0 100 100 100 100 96.4 97.9 84.9 100 98.1 94.1 75.8 Centile gains 3 15 1 4 1 5 Stratum gains Outlier gains 1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
CI. 1.1 to 1.8 The results obtained from these indicators were based on four or less reports and few patients. Thus it is not possible to determine any trends or gains. These indicators have been discontinued.
42 Determining the Potential to Improve Quality of Care results
2.4
2.5
2.6
CI No. 2.1 to 2.6 The numbers were too small to comment on the results. These indicators have been revised for 2004.
43
3.2
3.3
3.4
3.5
In each of the following indicators there were at most seven HCOs that submitted data in the last three years. Hence there is insufficient data to determine trends or stratum differences. These indicators have been discontinued. CI. 3.1 SCC The rate histological confirmation of diagnosis, prior to treatment or after excision in patients managed for SCC. The HCOs reporting since 2001 have reported rates close to 100%. CI. 3.2 SCC The rate of excision margins that are clear based on histological examination. There was no clear trend with the rates being about 90%. CI. 3.3 SCC The rate of documented evidence of examination of nodal status in the primary drainage area in patients managed for SCC. The rates have declined to 11% and exhibited large variation between the HCOs.
44
CI. 3.4 SCC The rate of positive excision margins and documented plan of action in all SCC excised. This indicator was difficult to interpret in its present form and has been discontinued. CI. 3.5 SCC The rate of having a documented follow-up plan in patients managed for SCC. In 2002 and 2003 one HCO reported rates of 0%. The rates of the remaining HCOs were 59% and 95% respectively in those years. The better rates were above 92.6% for all years, but the lower rates ranged from 3.8% to 97.7%. This can be due to only four to seven HCOs reporting.
45
1 2 3 4 5
to to to to to
immediately within 10 minutes within 30 minutes within 60 minutes within 120 minutes
1.2
1.3
1.4
1.5
46
Emergency Medicine Indicators Version 3 CI. 1.1 The rate of patients allocated ATS Category 1 who are attended to immediately. The rates have now reached 99% showing improvement in each year. The differences between the better performing and the poorer performing HCOs have also been reduced (Figure 1.1). Two thirds of the 23 HCOs with rates less than 90% had fewer than one ATS Category 1 patient each week. Figure 1.1 ATS 1 attended timmediately.
100 90 80 70
Rate %
Rate % Rate %
CI 1.2 Th rate of patients allocated ATS catorgory 2 who are attended to within 10 minutes. For triage category 2, there was little evidence of improvement, with one in every four patients not being seen within 10 minutes. For the better performing hospitals, one in every 10 patients were not seen within the recommended time frame. If the average of all HCOs could be brought to this rate, an additional 50,000 patients would have received treatment in the recommended timeframe. Most data are submitted from the public and metropolitan sectors. As might be expected, the emergency departments with the larger throughput had poorer rates than the smaller units, (68% compared to 80%). There were significant variations between the States, with NSW and Victoria having better rates (Table 1.2).
47
Emergency Medicine Indicators Version 3 Table 1.2 ATSC 2 attended within 10 minutes No HCOs 66 74 27 29 13 12 6 28 31 18 18 7 10 8 81 80 87 99 148 158 20 21 Stratum rate % 77.6 77.3 70.4 76.2 68.2 64.2 55.9 80.3 78.7 80.2 73.3 55.3 66.7 66.1 75.2 72.4 71.6 74.9 74.1 72.4 78.9 83.9 Standard error 1.3 1.3 1.8 1.7 2.3 2.1 4.4 1.8 1.8 2.5 2.1 2.6 2.3 3.6 0.94 0.94 1.7 1.8 0.85 0.84 3.6 3.6 Stratum gains 4,008 2,961 4,474 1,712 1,650 4,894 3,349 1,294
Stratum NSW QLD SA TAS VIC WA NZ Other Metropolitan Rural Public Private
Year 2002 2003 2002 2003 2002 2003 2002 2002 2003 2002 2003 2003 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003
Numerator 58,652 58,210 28,652 34,668 16,773 19,841 3,924 34,755 31,609 16,744 22,676 11,596 16,530 6,813 135,065 144,653 40,965 40,760 166,084 174,476 9,946 10,937
Denominator 75,539 75,300 40,709 45,526 24,598 30,920 7,032 43,300 40,173 20,870 30,909 20,978 24,783 10,316 179,599 199,719 57,232 54,403 224,236 241,107 12,595 13,015
CI. 1.3 The rate of patients allocated to ATS Category 3 who are attended to within 30 minutes. Triage category 3 also showed no improvement, with in fact a decline from 67% to 60% for the overall rates. The rates for those with a smaller throughput was 80% and drops to 60% for those with a higher throughput. As with the triage category 2 patients, the better performing hospitals were able to achieve a rate of 90%, or one person in every 10 not being seen within 30 minutes. If the overall rate were increased to 90%, over a quarter of a million patients would be seen in the recommended timeframe. The rates vary between states, with Victoria having the better rates of over 70% (Table 1.3) Figure 1.3 ATSC three attended within 30 minutes
100 90 80 70
Rate %
42 48
Emergency Medicine Indicators Version 3 Table 1.3 ATSC 3 attended within 30 minutes No HCOs 67 73 27 29 12 12 6 28 31 18 18 7 10 8 81 80 87 98 148 157 20 21 Stratum rate % 56.7 59.2 61.6 60.1 48.6 44.7 62.0 71.8 73.1 71.4 67.3 48.0 55.1 63.9 58.3 58.0 65.7 66.5 59.0 58.5 80.2 82.3 Standard error 1.5 1.4 2.1 1.9 3.1 2.7 4.9 2.1 2.0 3.6 2.7 2.8 2.6 3.9 1.1 1.1 1.8 1.8 0.96 0.91 3.5 3.2 Stratum gains 54,470 47,250 18,017 25,310 19,188 27,779 5,520 22,242 19,769 4,281 199,647 227,942
Stratum NSW QLD SA TAS VIC WA NZ Other Metropolitan Rural Public Private
Year 2002 2003 2002 2003 2002 2003 2002 2002 2003 2002 2003 2003 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003
Numerator 204,419 202,579 109,105 117,073 40,211 43,662 20,533 129,406 123,779 43,545 64,306 42,585 65,154 29,712 417,214 441,947 195,159 181,749 555,918 560,953 56,455 62,743
Denominator 360,651 341,960 177,080 194,893 82,756 97,808 33,118 180,235 169,400 60,957 95,556 88,753 18,313 46,535 716,036 761,790 297,074 273,115 942,707 958,691 70,403 76,214
CI. 1.4 The rate of patients allocated to ATS Category 4 who are attended to within 60 minutes. The situation with triage category 4 is similar to categories 2 and 3. There has been no improvement in delays in emergency treatment for these patients, and there was considerable variation between HCOs. Those with less throughput have rates of about 85% while the larger units have rates of 55%. There was a considerable difference between the better and poorer rates, being 54% and 92% respectively, for 2003. Figure 1.4 ATSC 4 attended within 60 minutes.
100 90 80 70
Rate %
49
Emergency Medicine Indicators Version 3 Table 1.4 ATSC 4 attended within 60 minutes No HCOs 81 80 85 95 147 155 19 20 Stratum rate % 57.4 59.4 68.4 69.8 59.3 60.9 85.9 85.6 Standard error 1.3 1.2 1.7 1.6 1.0 0.95 3.4 3.1 Stratum gains 375,525 350,149
CI. 1.5 The rate of patients allocated to ATS Category 5 who are attended to within 120 minutes. For triage category 5, the rates were generally better than those for categories 2, 3 and 4. However, there was also no evidence of improvement, with the overall rate dropping to 85% from 87%. Figure 1.5 ATSC 5 attended within 120 minutes
100 90 80 70
Rate %
50
2.1 The rate of AMI patients receiving thrombolysis within 1 hour of presentation. As a result of the findings that thrombolysis should be provided as soon as possible, the proportion of AMI patients receiving treatment within this time frame has increased from 72% to 81% in 2001. However, there has been no improvement during the years 2001 to 2003 (Figure 2.1). The better performing HCOs have only managed to achieve rates of about 87%. The reasons for not being able to achieve higher rates are worth investigating. Figure 2.1 AMI - thrombolysis within one hour.
100 90 80 70
Rate %
51
CI. 3.1 The rate of access block for patients requiring in-patient care The proportion of patients who present at the emergency department and were delayed by more than eight hours before being admitted to a ward not only varies considerably between hospitals but also has rates that were excessively high. One in four patients were unable to be admitted within eight hours, and for the poorer performing hospitals the rate was 40% or two in every five patients. The better HCOs had rates of 5%, or one in every 20 patients was unable to be admitted within 8 hours. NSW and Victorian HCOs had higher rates of about 30%, with the other States having rates less than 15% (Table 3.1). Table 3.1 Access block No HCOs 23 24 6 7 5 5 8 5 5 11 6 32 35 18 20 Stratum rate % 30.9 29.5 11.1 16.0 10.4 12.6 29.1 11.6 12.4 28.0 25.7 30.4 27.8 12.2 14.0 Standard error 2.8 2.6 5.2 4.5 7.9 8.3 6.0 7.3 5.1 3.6 3.9 2.2 2.0 3.4 3.3 Stratum gains 28,410 31,519 5,948 14,732 11,162 39,786 44,064
Year 2002 2003 2002 2003 2002 2003 2003 2002 2003 2002 2003 2002 2003 2002 2003
Numerator 44,273 54,455 4,509 10,379 1,852 2,392 10,369 2,412 6,226 24,375 21,580 66,296 88,558 11,125 16,843
Denominator 143,073 184,845 40,695 64,786 17,827 19,010 35,661 20,795 50,228 86,993 83,984 217,857 318,037 91,526 120,477
52
1.2
1.3
1.4
CIs. 1.1 1.4 Unexpected telephone calls and staff callouts Two of the four indicators measuring patient selection (CI 1.1 and 1.3) have improved their rates (Figures 1.1 and 1.3). The other two indicators relating to patient selection (CI 1.2 and 1.4) have remained at fairly low levels. The reasons for the improvement may relate to better selection of patients or better patient education. The differences between the better and lower performing HCOs may reflect the different patient mix, although when analysed by the possible ways to stratify the HCOs, no significant differences were found for unexpected telephone calls. There were some differences between public and private rates for unexpected staff callouts (Tables 1.3 and 1.4). Victoria contributes to about two-thirds of these data, and is more involved in monitoring the success of the hospital in the home program.
53
Hospital in the Home Indicators Version 2 Figure 1.1. One unexpected telephone call
20
5 5
Rate %
1998 1999 2000 2001 Year 2002 2003
2 5 1
Table 1.3 One unscheduled staff callout No HCOs 25 23 5 6 Stratum rate % 1.7 1.3 3.7 3.2 Standard error 0.21 0.16 0.68 0.36 Stratum gains 28 52
Table 1.4 More than one unscheduled staff calloutgains Stratum Public Private Year 2003 2003 No HCOs 22 6 Numerator 21 29 Denominator 13,109 2,673 Stratum rate % 0.21 0.82 Standard error 0.05 0.12 Stratum gains 16
54
2.2
2.3
2.4
CIs. 2.1-2.4 The rate of unplanned return to hospital without return to the HITH program. There were four measures of unplanned return to hospital. The overall rate for 2003 was 5.4%, and of these 1.4% returned to hospital in the home within 24 hours, and a similar percent returned after 24 hours. About 2.6% of these patients were unable to return to the program. The percentage of patients unable to return has declined from 3.1% to 2.6% over the last four years (Figure 2.1). In 2003 the Victorian rate, 6.7%, was twice the rate of the seven other states (Table 2.4). Figure 2.1 Unplanned return to hospital no return to HITH
8 7 6
Rate %
Table 2.4 Unplanned return to hospital Stratum VIC Other Year 2003 2003 No HCOs 18 14 Numerator 712 224 Denominator 10,494 6,996 Stratum rate % 6.67 3.37 Standard error 0.51 0.62 Stratum gains 346
55
CI. 1.1 The rate of post-operative pulmonary embolism in patients whose length of stay is at least seven days. The proportions of patients having pulmonary embolism as a result of surgery increased from less than 0.4% for years 1998 2000 to the highest proportion in 2003 to 0.56% (Figure 1.1). The higher rates in 2001 and 2003 were also associated with more outlier HCOs. In 2003, six public and six private HCOs contributed one third of the cases of pulmonary embolism. The rates for these HCOs ranged from 1.9% to 10.9%. If they were excluded from the analysis, then the 2003 rate becomes 0.38%, the same as the rate in 2002. The outlier HCOs should examine their use of prophylaxis.
1 .0
0 .8
Rate %
0 .6
0 .4
0 .2
56
CI. 2.1 The rate of unplanned and unexpected re-admissions within twenty-eight days of separation. There has been a downward trend in the mean proportions as well as the centiles (Figure 2.1). Because of the large denominators, there were many statistically significantly high rates or outliers. Thirty HCOs reported statistically significantly high rates in five of the six years. The average rate of the outlier HCOs in 2003 was 4.5% and was 4.3% in the previous years combined. The public sector had higher rates than the private (Table 2.1). The extent to which this is the result of different case mix needs to be assessed.
Rate %
Table 2.1 Unplanned readmission No HCOs 110 119 50 45 28 25 8 7 90 84 19 28 8 9 139 136 174 181 Stratum rate % 1.8 2.2 1.7 1.9 1.5 2.0 1.3 0.94 1.8 1.9 1.6 1.4 3.1 3.8 2.7 3.2 1.1 1.1 Standard error 0.12 0.14 0.16 0.20 0.24 0.33 0.29 0.35 0.13 0.17 0.29 0.24 0.30 0.26 0.088 0.099 0.079 0.094 Stratum gains 10,744 4,287 1,630 5,124 6,737 16,257 27,035
Year 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003
Numerator 14,158 18,605 6,916 8,348 2,658 3,050 1,569 1,239 10,355 10,402 1,903 4,018 3,598 8,957 27,441 40,138 13,716 14,481
Denominator 779,834 829,599 404,226 429,914 179,055 150,315 120,261 131,810 576,034 560,182 120,747 286,880 116,912 234,014 1,018,819 1,233,503 1,278,250 1,389,211
57
CI. 3.1 The rate of unplanned return to the operating room during the same admission. There has been a slight decline in the unplanned return to the operating theatre (Figure 3.1). In 2003 the public and private sectors had rates of 0.59% and 0.34% respectively (Table 3.1).
0 .8
Rate %
0 .6
0 .4
0 .2
Table 3.1 Unplanned return to operating room No HCOs 118 121 162 170 Stratum rate % 0.58 0.59 0.39 0.34 Standard error 0.031 0.028 0.021 0.019 Stratum gains 1,140
58
CI. 4.1 The rate of pressure ulcers (one or more) in non day-only patients. These data were first collected in 2003 and the overall rate for the 79 HCOs was 0.45%. The spread of the rates was from 0.05% to 0.57%, with the best explanatory variable being public or private. The private proportion was 0.15% compared to the public rate of 0.56%. Differences were also found between the metropolitan and rural HCOs and between the States (Table 4.1). Table 4.1 The rate of pressure ulcers (one or more) in non day-only inpatients No HCOs 43 18 9 9 42 37 43 36 Stratum rate % 0.50 0.33 0.27 0.71 0.53 0.20 0.56 0.15 Standard error 0.07 0.09 0.17 0.13 0.05 0.10 0.05 0.09 Stratum gains 1,398
CI. 4.2 4.5 The proportion of pressure ulcers that are Stage 1, 2, 3 and 4. The ulcers were classified into four stages, with one being the least severe to stage 4 which involves full thickness skin loss or damage to structures such as muscle and bone. About 45% of ulcers were determined to be stage 1 with stage 2 also having 45%. The HCOs that presented data for all five indicators were used to estimate the proportion of admissions which developed stage 1 to 4 pressure ulcers. The results were 0.21%, 0.21%, 0.035% and 0.007%. The strata had similar proportions for the severity of the ulcers.
59
Year 2003 2003 2003 2003 2003 2003 2003 2003 2003 2003 2003 2003 2003 2003 2003 2003 2003 2003
Denominator 9,722 9,125 11,543 10,674 5,491 5,332 3,855 3,504 933 1,619 440 213 187 176 16,568 15,746 2,540 2,096
Rate % 1.38 0.53 1.02 0.49 1.22 0.96 1.48 0.26 4.39 1.54 3.64 1.41 2.14 0.57 1.24 0.20 1.10 0.38
Rate (20)% 1.07 0.40 0.65 0.27 0.33 0.44 0.91 0.23 2.29 1.22 2.51 1.41 2.14 0.50 0.67 0.12 0.71 0.37
Rate (80)% 1.93 0.54 1.21 0.50 1.57 1.35 2.07 0.30 5.94 2.27 7.73 1.41 2.14 0.68 2.05 0.20 1.33 0.38
Centile gains 30 11 42 23 49 27 21 19 5 4 93 12 9
Stratum gains 13 3 30 32 7 12 56 5
Outlier gains 10 3 19 6 4 1 2 22 4
Burke JP (2003) Infection Control a problem for patient safety. The New Eng. J. Med. 348: 651-656
60 Determining the Potential to Improve Quality of Care
Infection Control Indicators Version 2 CI. 1.1 Superficial SSI in hip prosthesis procedures. The overall rate was 1.38%. The rate of the poorer performing HCOs was approximately twice the rate of the better performing HCOs (1.9% compared to 1.1%). These results suggest that the proportion of infections could be reduced by about one quarter (approximately 34 infections). The private HCO rate was lower than the public rate (1.8% and 1.2% respectively). Table 1.1 Superficial SSI in hip prosthesis procedures. No HCOs 33 78 Stratum rate % 1.81 1.24 Standard error 0.09 0.05 Stratum gains 13
Numerator 59 75
If the above SSI rate is applied to the proportion of separations for hip prosthesis procedures in Australia in 2002/2003, (30,496 including 3,267 revisions) then the estimated proportion of superficial SSIs in Australia is 400 per year. CI. 1.2 Deep/organ space SSI in hip prosthesis procedures. Deep SSI in hip prosthesis occurred in 0.53% of patients. The results suggest the proportion of these infections could be reduced by approximately one quarter. There were small state differences. Table 1.2 Deep/organ space SSI in hip prosthesis procedures. No HCOs 30 30 9 21 6 9 Stratum rate % 0.48 0.51 0.48 0.47 0.77 0.59 Standard error 0.05 0.05 0.08 0.04 0.07 0.07 Stratum gains 3
Numerator 5 10 3 12 10 8
If the above SSI rate is applied to the Australia hip prosthesis procedures in 2002/2003, then the estimated proportion of deep SSIs nationally is 160 per year. CI. 1.3 Superficial SSI in knee prosthesis procedures. The overall rate was 1.02%. There is potential to reduce the proportion of these infections by approximately one third, from 118 to 80. The lowest state rate was reported by New South Wales (0.72%). There were two outlier reports. The rates were 5.2% (in 251 procedures) and 9.5% (in 81 procedures). These HCOs should review their procedures. Table 1.3 Superficial SSI in knee prosthesis procedures. No HCOs 36 30 9 21 5 8 Stratum rate % 0.72 1.05 0.94 1.36 0.92 1.29 Standard error 0.10 0.11 0.17 0.11 0.16 0.21 Stratum gains 9 17 4
Numerator 11 31 9 44 11 12
In 2002/2003 there were 26,368 separations for knee prosthesis procedures in Australia. If the average SSI rate is applied to these procedures then the estimated proportion of superficial SSIs nationally is 270 per year.
61
Infection Control Indicators Version 2 CI. 1.4 Deep/organ space SSI in knee prosthesis procedures. The overall rate was 0.49%. There is potential to reduce the proportion of these infections by approximately one third. There were two outlier reports, with rates 4.2% (in 71 procedures) and 7.7% (in 52 procedures) respectively. These HCOs should review their procedures. If the average SSI rate is applied to all such procedures in Australia, then the estimated proportion of deep SSIs nationally is 130 per year. CI. 1.5 Superficial SSI (in the chest incision site) in CABG. The overall rate was 1.22%. The better performing HCOs had rates of 0.3%. The rate for Victoria (4.4%) was twice the overall rate. The public rate was approximately twice the private rate (2.1% and 0.8% respectively). There were four outlier reports. Table 1.5 Superficial SSI (in the chest incision site) in CABG. No HCOs 8 6 5 5 7 17 Stratum rate % 0.94 0.48 2.43 0.65 2.06 0.83 Standard error 0.33 0.33 0.31 0.45 0.34 0.23 Stratum gains 32 21
Numerator 14 4 46 3 40 27
There were 15,791 procedures for coronary artery bypass grafts in Australia in 2002/2003. Applying the overall rate to these procedures indicates there were approximately 200 Superficial SSI (in the chest incision site) annually. CI. 1.6 Deep/organ space SSI (in the chest incision site) in CABG. The rate of deep SSI (in the chest incision site) in coronary artery bypass was 0.96%. The rural rate was twice metropolitan rate (2.1% and 0.8% respectively). There were two outlier reports from one rural HCO. This HCO should review its results and procedures. Table 1.6 Deep/organ space SSI (in the chest incision site) in CABG. No HCOs 19 5 Stratum rate % 0.81 2.09 Standard error 0.10 0.27 Stratum gains 7
Numerator 33 18
Applying the average rate to the proportion of Australian admissions indicates there were approximately 150 deep/organ space SSIs annually in Australia. CI. 1.7 Superficial SSI (in the leg donor incision site) in CABG. The overall rate was 1.48%, with the better HCOs having a high rate of 0.9%. The overall rate implies that there were approximately 230 superficial SSIs annually in Australia. CI. 1.8 Deep SSI (in the leg donor incision site) in CABG. The overall rate was 0.26%. There was no evidence of differences between HCOs. The overall rate indicates there were approximately 40 deep SSIs annually in Australia. CI. 1.9 Superficial SSI in elective partial or total colectomy procedures (with anastomosis and no stoma formed). The rate of superficial SSI in colectomy procedures was 4.39% and the higher and lower rates were 5.9% and 2.3%. CI. 1.10 Deep/organ space SSI in elective partial or total colectomy procedures (with anastomosis and no stoma formed). The overall rate was 1.54%. Victoria reported the lowest state rate (0.79%) and the greatest proportion of procedures.
62 Determining the Potential to Improve Quality of Care
Infection Control Indicators Version 2 Table 1.10 Deep/organ space SSI in elective colectomy procedures Stratum NSW QLD VIC Other Year 2003 2003 2003 2003 No HCOs 10 10 6 7 Numerator 5 13 3 4 Denominator 446 355 603 215 Stratum rate % 1.68 2.38 0.79 2.01 Standard error 0.25 0.28 0.21 0.36 Stratum gains 3 5 2
CI. 1.11 Superficial SSI in femoral-popliteal bypass procedures. The overall rate was 3.64% with the higher rate being 7.7%. CI. 1.12 Deep/organ space SSI in femoral-popliteal bypass procedures. The overall rate was 1.41%. There was no evidence of differences between HCOs. CI. 1.13 Superficial SSI in open abdominal aortic aneurysm (AAA) procedures. The overall rate was 2.14%. There was no evidence of differences between HCOs. CI. 1.14 Deep/organ space SSI in open abdominal aortic aneurysm (AAA) procedures. The overall rate was 0.57%. There was little variation between HCOs. CI. 1.15 Superficial SSI in lower segment caesarean section procedures. The overall rate was 1.24%. There is potential to almost halve the proportion of these infections, from 205 to approximately 100. New South Wales reported the lowest rate (0.8%) and the greatest proportion of procedures. The public rate (1.8%) was almost twice the private HCO rate. There were six outlier HCOs, all reporting rates of 4% or more. These HCOs should review their data and/or procedures. Table 1.15 Superficial SSI in lower segment caesarean section procedures. Stratum NSW QLD SA VIC Other Public Private Year 2003 2003 2003 2003 2003 2003 2003 No HCOs 23 19 5 17 13 32 45 Numerator 27 55 21 32 70 128 77 Denominator 4,426 4,386 2,377 3,044 2,335 6,263 10,305 Stratum rate % 0.81 1.26 1.02 1.11 2.38 1.78 0.91 Standard error 0.18 0.18 0.25 0.22 0.25 0.15 0.12 36 54 Stratum gains 20
There were 69,300 caesarean section procedures performed in Australia in 2002/2003. If the average rate is applied to these procedures then the estimated proportion of superficial SSIs nationally is 850. CI. 1.16 Deep/organ space SSI in lower segment caesarean procedures. The overall rate was 0.2%. Twelve infections could be avoided if the overall rate was reduced to the better rate. If the overall rate is applied to the proportion of procedures performed nationally, then the estimated proportion of deep SSIs is 140. CI. 1.17 Superficial SSI in abdominal hysterectomy procedures. The overall rate was 1.1%, with a difference between the public and private sectors (Table 1.17) . Table 1.17 Superficial SSI in abdominal hysterectomy procedures. Stratum Public Private Year 2003 2003 No HCOs 17 42 Numerator 19 9 Denominator 632 1,908 Stratum rate % 1.79 0.88 Standard error 0.15 0.09 Stratum gains 5
There were 32,600 hysterectomies performed in Australia in 2002/2003. If the overall rate is applied to these preceedures the estimated proportion of superficial SSIs that occur in Australia annually is 360. CI. 1.18 Deep/organ space SSI in abdominal hysterectomy procedures. The overall rate was 0.38%. There was little variation between HCOs. The estimated proportion of deep/organ space SSIs that occur in Australia annually is 125.
Determining the Potential to Improve Quality of Care 63
No data was submitted for CIs. 2.6 and 2.8 in 2003. Central line utilisation ratios CI 2.2 2.4 Year 2003 2003 No HCOs 18 9 Numerator 12,432 561 Denominator 21,793 8,305 Rate % 57.0 6.75
CI. 2.1 Adult ICU-related centrally-inserted CLAB infection rate. The overall rate was 0.49% per patient day, with the better rates being 0.2%.
64
Infection Control Indicators Version 2 CI. 2.2 Adult ICU-related centrally-inserted central line utilisation ratio. Of the reported patient-days in Adult ICU, 57% involved the utilisation of centrally-inserted central lines. The utilisation rate in New South Wales was lower than in the other states (Table 2.2). Table 2.2 Adult ICU-related centrally-inserted central line utilisation ratio No HCOs 10 8 Stratum rate % 45.4 67.7 Standard error % 0.49 0.44
CI. 2.3 Adult ICU related peripherally inserted CLAB infection rate. The overall rate was 0.45% per patient day. A small proportion of patient days were reported (672). CI. 2.4 Adult ICU-related peripherally inserted central line utilisation ratio. Of the reported patient-days in adult ICU, 6.75% involved the utilisation of peripherally-inserted central lines. CI. 2.5 Paediatric ICU-related centrally inserted CLAB infection rate. The paediatric ICU-related centrally-inserted CLAB infection rate was reported by a single HCO, as 0.88% per patient day. CI. 2.6 Paediatric ICU related centrally inserted utilisation ratio. There were no reports for this indicator. CI. 2.7 Paediatric ICU related peripherally inserted (PI) CLAB infection rate. The paediatric ICU-related peripherally-inserted CLAB infection rate was reported by a single HCO as 0.36% per patient day. CI. 2.8 Paediatric ICU related peripherally inserted utilisation ratio. There were no reports for this indicator. CI. 2.9 Haematology Unit centrally inserted CLAB infection rate. The Haematology Unit centrally-inserted CLAB infection rate was reported by five HCOs in 2002 and a single HCO in 2003. The annual rates were 0.15% and 0.33% per patient day respectively. CI. 2.10 Haematology Unit peripherally - inserted CLAB infection rate. The Haematology Unit peripherally-inserted CLAB infection rate was reported by three HCOs in 2002 and a single HCO in 2003. The annual rates were 0.70% and 0.32% per patient day respectively. CI. 2.11 Oncology Unit centrally-inserted CLAB infection rate. The oncology unit centrally-inserted CLAB infection rate was reported by only a few HCOs. The rate in 2003 was 0.49% per patient day. There was one outlier HCO. CI. 2.12 Oncology Unit PI CLAB infection rate. Five HCOs reported no Oncology Unit peripherally-inserted CLAB infections. The rate was 0.04% per patient day in 2003. CI. 2.13 Outpatient Intravenous Therapy Unit centrally inserted CLAB rate. The outpatient intravenous therapy unit centrally-inserted CLAB infection rate was reported by three HCOs in 2002 and a single HCO in 2003. The annual rates were 0.26% and 0.49% per patient day respectively. CI. 2.14 Outpatient Intravenous Therapy Unit peripherally inserted CLAB infection rate. Outpatient Intravenous Therapy Unit peripherally-inserted CLAB infection rate was reported by three HCOs in 2002 and two HCOs in 2003.
65
CI. 3.1 Haemodialysis AV fistula-associated blood stream infection rate. There was one blood stream infection in 4,657 patient months where haemodialysis was performed via AV fistula. CI. 3.2 Haemodialysis synthetic graft-associated blood stream infection rate. There were no infections in 377 patient dialysis months from four HCOs where haemodialysis was performed via AV synthetic graft. CI. 3.3 Haemodialysis native vessel graft - associated blood stream infection rate. There were no infections in 84 patient dialysis months from two HCOs where haemodialysis was performed via native vessel graft. CI. 3.4 Haemodialysis centrally inserted non-cuffed dialysis line (temporary) - associated blood stream infection rate. Three HCOs reported a total of two infections in 38 patient dialysis months (5.26%). CI. 3.5 Haemodialysis centrally inserted cuffed (semipermanent) dialysis line-associated blood stream infection rate. Four HCOs reported a total of 14 infections in 406 patient dialysis months (3.45%).
66
CI. 4.1 Early invasive infection rate (Blood and/or CSF within 48 hours) all neonates. There were a total of two of these infections reported in 2,923 neonates from 8 HCOs (0.07%). CI. 4.2 Early invasive infection rate (Blood and/or CSF within 48 hours) neonates admitted to intensive care. A single HCO reported one infection in 73 neonates (1.37%). CI. 4.3 Late intensive care blood stream infection rate (after 48 hours from birth) neonates of birth weight < 1000g or < 28 weeks gestation, admitted to intensive care. A single HCO reported three of these infections in nine neonates (33.3%). CI. 4.4 Late intensive care blood stream infection rate (after 48 hours from birth) neonates of birth weight > 1000g or > 28 weeks admitted to intensive care. A single HCO reported no infections in 48 neonates.
67
1.2
CI. 1.1 The rate of entry into ANZICS Adult Patient Database. The proportion of admissions that were forwarded to this database was relatively high, with 80% of HCOs submitting more than 87% of their admissions. The better performing HCOs have rates near 100%. This would suggest that compliance is relatively high and appropriate to allow the data to be used to assess mortality rates. CI. 1.2 The rate of entry into the ANZPIC Registry. The proportion of HCOs providing data to this database increased from nine in 2002 to 15 in 2003. The ANZPIC data are used to describe paediatric intensive care practices and outcomes in Australia and New Zealand and was established in 1997. New South Wales had a rate of 99% in 2003 (Table 1.2). Table 1.2 Australian and NZ Paediatric Intensive Care Registry No HCOs 6 9 Stratum rate % 99.2 58.1 Standard error 5.71 11.6 Stratum gains 99
CI. 1.3 Participation in ARCCCR survey In 2002, 95% of HCOs responded to the most recent ARCCCR survey. In 2003, 84% of HCOs responded to the survey. Table 1.3 Proportion of HCOs participating in ARCCCR survey. CI 1.3 Year 2002 2003 No. Reports 25 45 No. HCOs 20 32 HCOs (%) 95 84
68
CI. 2.1 The rate of ICU access block The proportion of patients who were unable to be admitted to the ICU because of inadequate resources has remained at about 5% for the last 6 years (Figure 2.1). The better HCOs had rates of 0.2%, compared to the poorer HCOs with rates of 8%. The public rate was approximately ten times the private rate in 2002 and 2003 (Table 2.1). Figure 2.1 ICU - access block
10
Rate %
Table 2.1 ICU access block No HCOs 46 43 28 29 Stratum rate % 7.02 6.74 0.76 0.56 Standard error 0.68 0.63 1.04 0.96 Stratum gains 1,727 1,752
69
CI. 3.1 The rate of unplanned readmission to ICU within 72 hours. There appears to have been neither an improvement in the overall rates or in the better and poorer rates since 1998 (Figure 3.1). The poorer rates have been approximately twice the better rates. This variation suggests that there is potential to reduce the proportion of unplanned readmissions to ICU within 72 hours by one third. Figure 3.1 ICU - unplanned readmission
3 .0
2 .5
2 .0
Rate %
1 .5
1 .0
0 .5
70
CI. 1.1 The rate of thrombolysis within one hour of presentation in AMI patients who had thrombolysis. The rates for timely thrombolysis increased from 68% in 1998 to 79% in 2003 (Figure 1.1). In 2003, one in five patients who received thrombolysis did not receive it within one hour. The better rates have been about ten percentage points higher than the overall rates during that period. This variation indicates that the rates could be improved by ten percentage points. Victoria reported the highest rate in 2003 (Table 1.1). Figure 1.1 AMI thrombolysis within one hour
100 90 80 70
Rate %
No HCOs 17 7 12 10
Stratum gains 34
71
Area 1 Cardiology II
Rationale PTCA should be performed with a high primary success rate and a low rate of immediate complications, such as Acute Myocardial Infarction (AMI) or complications requiring Coronary Artery Bypass Grafts (CABG). The indicators are: CI. 1.2 The rate of primary success in vessels in which PTCA (with or without stenting) is undertaken CI. 1.3 The rate of inpatients undergoing CABG within twenty-four hours of a PTCA (with or without stenting) in the same admission Results CI 1.2 Year 1998 1999 2000 2001 2002 2003 1998 1999 2000 2001 2002 2003 No HCOs 20 21 24 23 24 27 25 25 28 26 25 28 Numerator 5,047 6,471 8,460 8,540 8,634 13,049 41 31 32 37 26 34 Denominator 5,350 6,847 8,933 9,220 8,992 13,505 5,562 7,807 9,117 8,984 9,535 12,628 Rate % 94.3 94.5 94.7 92.6 96.0 96.6 0.74 0.40 0.35 0.41 0.27 0.27 Rate (20)% 93.4 93.8 94.1 93.4 94.3 95.0 0.71 0.35 0.35 0.21 0.09 0.10 Rate (80)% 95.8 97.0 95.7 97.2 97.2 97.7 0.81 0.44 0.35 0.62 0.43 0.51 Centile gains 80 170 90 419 106 145 1 3 18 17 21 Stratum gains 64 160 59 2 Outlier gains 47 15 228 23 5 5 3
1.3
CI. 1.2 The rate of primary success in vessels in which PTCA (with or without stenting) is undertaken. The rates have increased from 94% 1998 to 97% in 2003. Over this period there were only small differences between the HCOs with the better and the poorer rates. In 2001 one large HCO reported a rate of 51%. When this HCO is removed the rate was 95% in 2001. CI. 1.3 The rate of inpatients undergoing CABG within twenty-four hours of a PTCA (with or without stenting) in the same admission. The rates have decreased from 0.7% in 1998 to 0.3% in 2003, that is less than three per 1000. Both the better and poorer rates have decreased over that period.
72
Area 2 Endocrinology
Rationale Elective surgery for insulin-treated diabetic patients requires an alteration of the treatment schedule. It is estimated that between 0.5 and 2.0% of patients undergoing elective surgery are insulin-treated diabetics. Inappropriate management creates the risk of hypoglycaemia. The indicators are: CI. 2.1 The rate of insulin treated diabetic inpatients having an elective operation, and a length of stay equal to or greater than 48 hours, whose medical record shows at least four blood glucose measurements on the first post-operative day CI. 2.2 The rate of insulin treated diabetic inpatients having an elective operation, and a length of stay equal to or greater than forty-eight hours with a recorded blood glucose level, BSL, less than 4mmol/l in the post operative period Results CI 2.1 Year 1998 1999 2000 2001 2002 2003 1998 1999 2000 2001 2002 2003 No HCOs 24 19 20 18 14 19 19 18 17 16 13 18 Numerator 145 127 195 151 137 230 43 46 28 31 28 54 Denominator 275 135 213 175 152 269 188 151 184 188 145 266 Rate % 52.7 94.1 91.5 86.3 90.1 85.5 22.9 30.5 15.2 16.5 19.3 20.3 Rate (20)% 52.8 94.1 90.8 79.5 90.4 81.7 11.2 23.0 12.7 13.8 15.1 18.5 Rate (80)% 88.8 94.1 95.1 94.3 94.9 95.3 33.3 33.2 17.7 20.3 22.3 21.9 Centile gains 99 7 14 7 6 22 11 4 5 6 4 Stratum gains 65 9 26 6 4 Outlier gains 44 4 2 19 5
2.2
CI. 2.1 The rate of insulin treated diabetic inpatients undergoing elective surgery having at least 4 BSL on the first post-operative day. There appears to have been no improvement in the overall rates since 1998. In 2003 the rate was 85.5%. The proportion of patients represented by this indicator is small and consequently the rates varied considerably. Over the entire six-year period the overall rate, the better rates and the poorer rates were 81%, 94% and 79% respectively. Since 1998, more than half of HCOs have had rates of 100% and hence it should be possible for more HCOs to have rates close to 95%. The New South Wales rate and the public rate were highest in 2003, being 95% and 93% respectively (Table 2.1). Table 2.1 Insulin treated diabetics at least 4 BSLs No HCOs 7 12 12 7 Stratum rate % 95.3 72.9 93.4 71.4 Standard error 5.34 6.04 5.08 6.76 Stratum gains 26 21
CI. 2.2 The rate of insulin treated diabetics undergoing elective surgery having post operative BSL < 4mmol/l There appears to have been no improvement in the overall rates since 1998, however the difference between the better and poorer rates has been less than ten percentage points since 2000 (Figure 2.2). The proportion of patients represented by this indicator was small and consequently the overall rates varied considerably, ranging from 15% to 31% over the six year period. Over the entire six-year period the overall rate, the better rate and the poorer rate were 21%, 16% and 24% respectively.
73
Internal Medicine Indicators Version 3 Figure 2.2 Insulin treated diabetics - post operative BSL < 4mmol/l
40 35 30 25
Rate %
74
Area 3 Gastroenterology
Rationale Haematemesis and melaena (H&M) are common symptoms necessitating admission to hospital and often indicate significant potential morbidity. These indicators relate to the management of patients admitted with haematemesis and/or melaena who also received a blood transfusion. The indicators are: CI. 3.1 The rate of gastroscopy within 24 hours in H&M patients who had a blood transfusion CI. 3.2 The rate of having a specific discharge diagnosis explaining the cause of bleeding in H&M patients who had a blood transfusion CI. 3.3 The rate of surgical staff notification in H&M patients who had a blood transfusion CI. 3.4 The rate of operation in H&M patients who had a blood transfusion CI. 3.5 The rate of operation in H&M patients who had a blood transfusion and endoscopic therapy CI. 3.6 The rate of death in H&M who had a blood transfusion Results CI 3.1 Year 1998 1999 2000 2001 2002 2003 1998 1999 2000 2001 2002 2003 1998 1999 2000 2001 2002 2003 1998 1999 2000 2001 2002 2003 1998 1999 2000 2001 2002 2003 1998 1999 2000 2001 2002 2003 No HCOs 31 29 29 24 17 25 30 26 24 20 14 20 26 22 20 15 12 16 26 22 26 23 16 21 21 23 22 17 15 20 27 25 25 27 19 25 Numerator 492 561 329 393 274 397 531 641 324 37 260 333 268 398 207 225 167 155 52 48 58 91 77 72 89 77 34 24 19 20 59 60 39 86 85 82 Denominator 696 808 540 670 457 767 675 782 427 552 327 454 626 734 382 328 293 359 651 734 501 694 450 682 351 760 461 456 433 672 686 777 497 825 775 793 Rate % 70.7 69.4 60.9 58.7 60.0 51.8 78.7 82.0 75.9 79.2 79.5 73.3 42.8 54.2 54.2 68.6 57.0 43.2 7.99 6.54 11.6 13.1 17.1 10.6 25.4 10.1 7.38 5.26 4.39 2.98 8.60 7.72 7.85 10.4 11.0 10.3 Rate (20)% 45.2 44.2 40.3 33.4 44.9 35.7 68.2 79.9 65.8 62.0 58.1 54.3 36.4 40.2 42.8 48.4 64.2 42.1 6.30 6.12 2.84 7.69 6.03 3.61 23.3 2.38 1.76 4.72 3.03 1.10 8.60 6.13 7.85 10.4 8.68 9.32 Rate (80)% 82.9 81.3 68.8 61.8 60.9 60.6 84.2 84.1 85.1 88.7 88.1 72.1 1.4 70.6 75.8 89.0 89.3 77.7 11.2 7.56 23.7 16.4 30.9 22.6 24.1 4.87 12.5 5.59 3.91 3.92 8.60 8.52 7.85 10.4 11.0 11.5 Centile gains 85 95 42 20 4 67 37 16 39 52 27 178 120 82 66 94 123 21 7 60 23 62 47 7 58 25 2 5 12 12 17 8 Stratum gains 98 95 91 20 223 40 10 13 21 11 153 25 63 90 72 17 6 51 8 10 7 Outlier gains 33 19 14 60 11 6 10 9 44 31 32 40 11 2 6
3.2
3.3
3.4
3.5
3.6
75
Internal Medicine Indicators Version 3 CI. 3.1 The rate of gastroscopy within 24 hours in H&M patients who had a blood transfusion. The overall rates have decreased from 70.7% in 1998 to 51.8% in 2003 (Figure 3.1). Both the better rates and the poorer rates have decreased over the period and the differences between these have decreased suggesting more uniform practice among reporting HCOs. In 2003 half of the patients who received blood transfusions had a gastroscopy. The better rates were about two thirds and the poorer rates about one third. This variation indicates that the rate of gastroscopy in these patients could be increased, if it is desirable that a gastroscopy be performed. New South Wales reported the lowest rate, 31% in 2003 (Table 3.1). Figure 3.1 H&M with blood transfusion gastroscopy within 24 hour
100 90 80 70
Rate %
Table 3.1 H&M with blood transfusion - gastroscopy within 24 hours No HCOs 5 9 5 6 Stratum rate % 30.6 49.2 61.0 80.9 Standard error 5.24 6.74 8.93 6.41 Stratum gains 147 56 20
Numerator 87 84 65 161
CI. 3.2 The rate of having a specific discharge diagnosis explaining the cause of bleeding in H&M patients who had a blood transfusion. The overall rates do not appear to have improved since 1998. The rate was lowest, 73%, in 2003. In years prior to 2003 the better rate was in excess of 80%. CI. 3.3 The rate of surgical staff notification in H&M patients who had a blood transfusion. The rates of notification of surgical staff in these patients varied between 43% and 69% over the period 1998 to 2003. The better rate was at least 25 percentage points above the average over that period indicating that the rate could be increased. Figure 3.4 H&M with blood transfusion 0peration
40 35 30 25
CI. 3.4 The rate of operation in H&M patients who had a blood transfusion The proportion of patients having surgery varied from 6.5% to 17.1% between 1998 and 2003. The better rates ranged from 2.84% to 7.69% over that period. The rural rate was 28.1%, which may be of concern.
Rate %
76
Internal Medicine Indicators Version 3 Table 3.4 H&M with blood transfusion operation Stratum Metropolitan Rural Year 2003 2003 No HCOs 12 9 Numerator 46 26 Denominator 597 85 Stratum rate % 8.11 28.1 Standard error 3.41 9.03 Stratum gains 17
CI. 3.5 The rate of operation in H&M patients who had a blood transfusion and endoscopic therapy. The rates have decreased from 25.4% in 1988 to 2.98% in 2003. The difference between the better and poorer rates since 2001 has been a few percentage points. Since 2000, the overall rates and the better and poorer rates were lower in this group of patients than in those who did not have an endoscopy (CI. 3.4).New South Wales had the lowest rate of 1.2% in 2003 (Table 3.5). Table 3.5 H&M with blood transfusion, endoscopic therapy and operation Stratum NSW QLD Other Year 2003 2003 2003 No HCOs 5 6 9 Numerator 2 12 6 Denominator 293 101 278 Stratum rate % 1.20 9.19 2.64 Standard error 1.02 1.74 1.05 Stratum gains 8
CI. 3.6 The rate of death in H&M patients having blood transfusion. The death rates varied between 8% and 11% between 1998 and 2003. During that period there were differences of one or two percentage points between the better and poorer rates. The stratum differences were small. Thus, although the mortality rate is high, there are no indications that the outcome varies between HCOs.
77
CI. 4.1 The rate of documented assessment of mental function in patients admitted to ageriatric medicine or geriatric rehabilitation unit. The rates have increased from 72.9% in 1998 to 84.1% in 2003 (Figure 4.1). The poorer rates increased from 42.2% to 70.6%. The better rates have exceeded 97% for the last six years indicating that all HCOs could achieve higher rates. Figure 4.1 Geriatrics documented mental function
100 90 80 70
Rate %
78
CI. 4.2 The rate of documented assessment of physical function in patients admitted to a geriatric medicine or geriatric rehabilitation unit. The rates have increased marginally from 87% in 1998 to above 90% in the subsequent years (Figure 4.2). The better rates exceeded 99.5% in all years indicating that rates close to 99% could be attained by all HCOs. Figure 4.2 Geriatrics docomented physical assessment
100 90 80 70
Rate %
50 40 30 20 10
Rate %
60
79
Area 5 Nephrology
Rationale Renal biopsy is a procedure commonly performed as part of the assessment of patients with renal disease. It is an invasive technique, which may put the patient at risk. The indicator is: CI. 5.1 Adequacy and safety of renal biopsy as determined by the presence of macroscopic haematuria Results CI 5.1 Year 1998 1999 2000 2001 2002 2003 No HCOs 13 14 19 21 23 16 Numerator 42 69 38 58 47 29 Denominator 653 892 1,114 1,819 1,531 571 Rate % 6.43 7.74 3.41 3.19 3.07 5.08 Rate (20)% 6.43 2.99 2.36 1.96 2.75 2.96 Rate (80)% 6.43 13.1 4.66 4.62 5.27 7.60 Centile gains 42 11 22 4 12 Stratum gains 5 4 Outlier gains 7 13 2
CI. 5.1 The rate of developing macroscopic haematuria within 24 hours in inpatients having a renal biopsy. The rates from 2000 have been lower than those in 1998 and 1999. The better and poorer rates have been variable over the period, with the better rates being less than half the poorer rates from 1999 onwards. This variation indicates some potential to reduce the rate that patients develop macroscopic haematuria.
80
Area 6 Neurology
Rationale Stroke is a common cause of admission to hospital that is resource intensive. Computerised tomography (CT) scanning makes an important contribution to patient management. The indicator is: CI. 6.1 Stroke investigation using a CT Scan Results CI 6.1 Year 1998 1999 2000 2001 2002 2003 No HCOs 82 72 76 74 65 70 Numerator 3,709 3,782 4,741 5,396 4,232 5,294 Denominator 4,902 4,594 5,853 6,746 5,165 6,341 Rate % 75.7 82.3 81.0 80.0 81.9 83.5 Rate (20)% 57.5 63.6 66.8 68.0 72.3 73.1 Rate (80)% 88.3 92.0 92.1 92.4 92.5 90.0 Centile gains 618 446 649 836 547 410 Stratum gains 293 112 531 218 223 Outlier gains 194 145 213 254 120 177
CI. 6.1 The rate of documented CT scan in patients with a discharge diagnosis of stroke. The overall rates of documented CT scan in stroke patients increased from 76% in 1998 to 80% or more in subsequent years (Figure 6.1). The better rates have been approximately ten percentage points higher than the overall rates suggesting that there is potential to increase the rate in all HCOs to 90% or more. New South Wales and Victoria reported the highest rates in both 2002 and 2003 (Table 6.1). Figure 6.1 Stroke - documented CT scan
100 90 80 70
Rate %
Table 6.1 Stroke documented CT scan Stratum NSW QLD SA TAS VIC WA Other Year 2002 2003 2002 2003 2002 2002 2003 2002 2003 2003 2002 2003 No HCOs 18 21 14 15 7 5 5 14 16 5 7 8 Numerator 1,280 1,435 1,013 1,127 204 279 287 1,045 1,664 299 411 482 Denominator 1,461 1,629 1,319 1,478 297 347 413 1,222 1,905 355 519 561 Stratum rate % 87.1 87.5 77.1 76.7 70.9 80.3 71.8 85.1 87.3 83.9 79.8 85.2 Standard error 2.11 1.75 2.22 1.84 4.67 4.32 3.48 2.30 1.62 3.75 3.53 2.99 Stratum gains 132 158 48 64 37
81
Area 7 Oncology
Rationale Stage II carcinoma of the breast with nodal involvement is a common malignancy in pre-menopausal women. The standard management usually includes systemic adjuvant therapy. The indicator is: CI. 7.1 Use of systemic adjuvant treatment for Stage II breast cancer Results CI 7.1 Year 1998 1999 2000 2001 2002 2003 No HCOs 18 12 14 8 9 14 Numerator 238 220 110 106 152 166 Denominator 247 227 123 116 169 174 Rate % 96.4 96.9 89.4 91.4 89.9 95.4 Rate (20)% 95.9 97.4 80.9 83.7 84.5 91.2 Rate (80)% 97.5 98.0 99.1 97.2 96.0 97.7 Centile gains 2 2 11 6 10 4 Stratum gains 3 Outlier gains 8 3 2 1
CI. 7.1 The rate of documented evidence of treatment, or intention to treat, with poly-chemotherapy in pre-menopausal patients with stage II carcinoma of the breast. The overall rates do not appear to have improved over the six years 1998 to 2003. In 2003 all but eight of the 174 patients covered by this indicator were offered the use of poly-chemotherapy. The better rates have remained at 96% or more indicating that all HCOs could achieve rates close to this.
82
8.2
8.3
CI. 8.1 The rate of documented objective assessment of severity on initial presentation in patients admitted to hospital with acute asthma. The rates increased from 83% in 1998 to approximately 90% from 2001 onwards. The poorer rates have been above 80% since then (Figure 8.1). The better rates have been above 97% since 1999 indicating that all HCOs could achieve rates close to this. Figure 8.1 Acute asthma documented severity on presentation
100 90 80 70
Rate %
83
Internal Medicine Indicators Version 3 CI. 8.2 The rate of documented subsequent objective assessment of severity in patients admittedto hospital with a acute asthma. The rates of subsequent objective assessment were a couple of percentage points lower than the corresponding rates on admission (CI. 8.1). The better rates have been 95% or above during the six year period indicating that an all HCOs could achieve rates close to this. There is evidence that the rates have declined slightly in 2003 (Figure 8.2). Figure 8.2 Acute asthma documented severity in addition to presentation
100 90 80 70
Rate %
CI. 8.3 The rate of documented evidence of an appropriate discharge plan in patients admittedto hospital with a acute asthma. The rates have ranged from 50% to 60% between 1999 and 2003 (Figure 8.3). Approximately 50% of asthma patients have documented evidence of an appropriate discharge plan. The rates for the better performing HCOs have been three or four times the rates for the poorer HCOs and twenty percentage points above the overall rates. These differences indicate that substantial improvements in the overall rate could be made. In 1998/1999 there were 42,336 admissions to the Australian Refined Diagnosis Related Groups (AR-DRG) E69C, Bronchitis and Asthma Age less than 50 without complications or co-morbidities. The number declined to 28,759 in 2002/2003, a reduction by one third. If the 2003 rate is applied to these 28,759 admissions then the estimated proportion of patients not receiving an appropriate discharge plan is approximately 13,000 (0.45%). Figure 8.3 Acute asthma documented discharge plan
100 90 80 70
Rate %
84
1.2
CI. 1.1 The rate of allocation of a diagnosis within twenty-four hours of admission. The better rates have exceeded 99% since 1999 and the difference between the better and poorer performing HCOs has been around ten percentage points during that period. Nine HCOs have had statistically significantly low rates during three or more years since 1998. Since this is a process indicator there is potential for the HCOs with the poorer rates to improve. Figure 1.1 Diagnosis within 24 hours of admission
100 90 80 70
Rate %
85
Mental Health Inpatient Indicators Version 4 CI. 1.2 The rate of having a diagnosis on separation recorded in the medical record. There were considerable differences between the HCOs, with the better rates exceeding 99.4% in all years. However, the poorer rates were around twenty percentage points lower than these rates (Figure 1.2). Thirteen HCOs have had statistically significantly low rates during the three or more years since 1998. There is potential for the HCOs with the poorer rates to improve. Figure 1.2 Diagnosis in medical record on discharge
100 90 80 70
Rate %
86
CI. 2.1 The rate of a documented complete physical examination within forty-eight hours of admission. There has been a decline in the overall rates since 1999, from 90% to 85%. However, the HCOs with the better rates have exceeded 99% since 1999. The HCOs with the poorer rates provide a complete physical examination to less than 75% of their patients. Fifteen HCOs have had statistically significantly low rates during three or more years since 1998. Since this is a process indicator the causes of the decline should be investigated. Table 2.1 Documented complete physical exam within 48 hours
100 90 80 70
Rate %
87
CI. 3.1 The rate of inpatients on two or more psychotropic medications from one sub-group category at discharge. The overall rates have remained close to 26% since 1998. The poorer rates have been ten or fifteen percentage points higher than the overall rates and the lower rates were fifteen percent below (Figure 3.1). The causes of the large variation in rates should be determined as they may indicate a difference in casemix or a difference in clinical practice. Ten HCOs had statistically significantly high rates during three or more years. Victoria and South Australia had the highest rates in 2003 (Table 3.1). Figure 3.1 Two or more psychotropic medications at discharge
50
40
Rate %
30
20
10
Year 2002 2003 2002 2003 2003 2002 2003 2002 2003
No HCOs 20 19 20 16 5 11 10 11 9
Numerator 1,268 1,899 1,231 902 774 792 1,056 1,434 785
Denominator 7,563 8,816 4,240 4,206 2,223 2,474 2,193 4,473 2,859
Stratum rate % 17.2 21.7 28.9 21.4 34.8 31.9 48.0 32.0 27.3
Standard error 2.4 2.7 3.2 3.9 5.3 4.1 5.4 3.1 4.7
88
4.2
Rate %
CI. 4.1 The rate of inpatients undergoing more than twelve treatments of E.C.T. during a course. The rates do not appear to have decreased during the period 1998 to 2003 (Figure 4.1). The better rates have been approximately half the overall rates during the period. The variation between HCOs exists in both the public and private sectors, and indicates that there is a capacity to reduce rates.
15
10
Rate %
Figure 4.2 Major medical complications with E.C.T. CI. 4.2 The rate of inpatients experiencing major medical complications while undergoing E.C.T. The rates have not decreased during the period 1998 to 2 .0 2003 (Figure 4.2). There was relatively small variation between HCOs, the better rates having been approximately one and a half times the poorer rates over the period. 1 .5 The variation between HCOs exists in both the public and private sectors.
1 .0
0 .5
89
5.2
5.3
5.4 5.5
Except for one or two private HCOs, these data are for public HCOs.
90
Mental Health Inpatient Indicators Version 4 CI. 5.1 The rate of inpatients having at least one episode of seclusion in an admission. The overall rates have remained close to 10%, the better and poorer rates being 5% and 15% approximately (Figure 5.1). The rates of the better performing HCOs have been one third of the rates of the poorer performing HCOs. These differences suggest that there could be an opportunity to decrease the rates of seclusion. Queensland had the lowest rates in 2002 and 2003 (Table 5.1). Figure 5.1 At least one episode of seclusion in an admission
20
15
Rate %
10
Table 5.1 At least one episode of seclusion in an admission Stratum NSW QLD VIC Other Year 2002 2003 2002 2003 2002 2003 2002 2003 No HCOs 17 19 14 14 18 19 10 13 Numerator 1,466 1,517 497 670 1,398 1,401 600 929 Denominator 16,020 16,306 8,775 10,290 9,517 9,441 5,881 8,156 Stratum rate % 9.1 9.3 5.9 6.7 14.6 14.6 10.2 11.4 Standard error 0.88 0.82 1.2 1.0 1.1 1.1 1.5 1.2 Stratum gains 513 423 824 751 254 386
CI. 5.2 The proportion of secluded patients having at least two episodes of seclusion. For those patients who have had at least one episode of seclusion, nearly 40% have two or more episodes. The rates for secluded patients being secluded more than once does not appear to have decreased over the period however the differences between the better and poorer performing HCOs has diminished. The difference between the better and poorer performing HCOs has been around ten percentage points in the last few years (Figure 5.2). Figure 5.2 At least two episodes of seclusion in an admission
50
40
Rate %
30
20
10
91
Mental Health Inpatient Version 4 CI. 5.3 The proportion of secluded patients having seclusion for more than four hours in one episode. The rates appear to have increased from 21% in 1998 to 27% in 2003. There has been an apparent increase in both the lower and the higher rates as well (Figure 5.3). The higher rates have been approximately twice the overall rates indicating that there is the capacity to determine the causes of these differences. The New South Wales rates were the lowest in 2002 and 2003 (Table 5.3). Figure 5.3 Seclusion for more than four hours in one episode
60
50
40
Rate %
30
20
10
Table 5.3 Seclusion for more than four hours in one episode No HCOs 16 16 13 12 17 20 8 12 Stratum rate % 8.4 16.3 45.2 51.5 43.9 34.9 28.1 17.2 Standard error 2.5 3.4 5.4 5.7 3.2 3.2 5.0 4.3 Stratum gains 177 177 497 292 108
CI. 5.4 The proportion of secluded patients who were not reviewed by sight at least half-hourly. All but six HCOs had rates less than 10% in 2003 and all but three HCOs had rates less than 5% in 2003. In 2003 the better rate of 0.55% was approximately half the overall rate and one third of the poorer rate. CI. 5.5 The proportion of secluded patients who experience major complications while in seclusion. The rates have decreased from more than 1% to 0.3% in 2003. The rates have been low in recent years.
92
of of of of of
inpatient attempted or actual suicide in an admission inpatients assaulting in an admission assaulting inpatients having assaulted twice or more in an admission inpatients undertaking significant self-mutilation in an admission inpatients suffering significant other injuries in an admission
6.2
6.3 6.4
6.5
CI. 6.1 The rate of attempted or actual suicide in an admission. The overall rates have remained around 0.7% since 1998 however there was a decline in the poorer rates from 1.5% to 1% in the 2002 and 2003 (Figure 6.1).
93
Mental Health Inpatient Indicators Version 4 CI. 6.1 Attempted or actual suicide.
2 .0
1 .5
Rate %
1 .0
0 .5
CI. 6.2 The rate of inpatients assaulting in an admission. The overall rates have remained around 4% during the period 1998 to 2003. The variation between HCOs is large. The poorer rates have remained close to 7% since 2000, whereas the better rates have been around 0.5% 2003 (Figure 6.2). The public rates were approximately seven times the private rates in 2002 and 2003 (Table 6.2). All the private HCOs had rates less than 5% and approximately half of the public HCOs had rates less than 5% in 2002 and 2003. Ten public HCOs have had statistically significantly high rates, (usually between 10% and 20%) for at least three of the six years. This degree of variation indicates that the causes should be determined. Figure 6.2 Assault in an admission
8 7 6
Rate %
Table 6.2 Assault in an admission Stratum Public Private Year 2002 2003 2002 2003 No HCOs 48 61 30 31 Numerator 2,006 2,049 100 123 Denominator 33,032 39,740 15,973 18,580 Stratum rate % 6.01 5.10 0.76 0.79 Standard error 0.49 0.44 0.71 0.64 Stratum gains 1,736 1,712
CI. 6.3 The proportion of assaulting inpatients having assaulted twice or more in an admission. Almost one third of patients that had assaulted once assaulted at least once more during the same admission. The variation between HCOs within the public and the private sectors were similar, both had rates of approximately 30% and similar better and poorer rates (20% and 35%). CI. 6.4 The rate of inpatients undertaking significant self-mutilation in an admission. These rates have not decreased since 1998, varying around 1% (Figure 6.4). The rates for both the public and the private sectors were similar. Seven HCOs had statistically significantly high rates between 2% and 10%.
94
Mental Health Inpatient Indicators Version 4 Figure 6.4 Significant self mutilation
3 .0
2 .5
2 .0
Rate %
1 .5
1 .0
0 .5
CI. 6.5 The rate of inpatients suffering significant other injuries in an admission. The overall rates did not decrease since 1998, ranging from between 0.52% and 1.1%, but the higher rates have declined from 1.87% in 1998 to less than 1% in 2003 (Figure 6.5). Despite this reduction in differences between the rates, the lower rates of 0.2% were significantly below the overall rates, which suggests that there is capacity to improve. Figure 6.5 Significant other injuries
2 .0
1 .5
Rate %
1 .0
0 .5
95
CI. 7.1 The rate of unplanned re-admissions within twenty-eight days. The rates have not decreased and have ranged from 7.73% to 9.27%. The overall rates were approximately twice the better rates during the six years (Figure 7.1). The public rates, 9.9%, were three percentage points above the private rates in 2002 and 2003 (Table 7.1). There was considerable variation within each sector, a two fold difference between the better and poorer rates, which indicates that there is capacity to improve. Figure 7.1 Unplanned readmissions within 28 days
20
15
Rate %
10
Table 7.1 Unplanned readmissions within 28 days No HCOs 54 66 34 32 Stratum rate % 9.85 9.85 6.90 6.01 Standard error 0.43 0.43 0.58 0.61 Stratum gains 843 1,280
96
Area 8 Mortality
Rationale Whilst not all deaths may be preventable, this indicator measures the inpatient death rates as a measure of the safety and quality of care for psychiatric patients. The data refer to inpatient psychiatric services only. The indicator is: CI. 8.1 The rate of inpatient deaths Results CI 8.1 Year 1998 1999 2000 2001 2002 2003 No HCOs 67 74 77 82 93 102 Numerator 102 65 65 36 52 76 Denominator 38,777 40,150 51,085 42,433 50,936 58,595 Rate % 0.26 0.16 0.13 0.08 0.10 0.13 Rate (20)% 0.11 0.05 0.05 0.05 0.05 0.07 Rate (80)% 0.43 0.19 0.23 0.13 0.11 0.15 Centile gains 61 44 39 15 26 35 Stratum gains 27 28 Outlier gains 31 11 11 1 11 9
CI. 8.1 The rate of inpatient deaths. The mortality rates have decreased from 0.26% in 1998 to around 0.1% from 2000 onwards (Figure 8.1). The difference between the better and poorer rates has decreased over the period, from 0.3% in 1998 to 0.075% in 2001. Figure 8.1 Inpatient deaths
0 .5 0
0 .4 0
Rate %
0 .3 0
0 .2 0
0 .1 0
97
CI. 9.1 The rate of inpatients having a discharge summary or letter at discharge. Almost three quarters of inpatients were discharged with a discharge summary or letter in 2002 and 2003. The rates for the poorer performing HCOs were less than 50% whereas the better rates were in excess of 97%. In 2003 the median rates were approximately 80% in both the public and prive sectors and both had similar better and poorer rates (approximately 95% and 50%). These differences suggest that the process of providing a discharge summary could be improved in both sectors. Figure 9.2 Final discharge summary in medical record within two weeks
100 90 80 70
Rate %
CI. 9.2 The rate of inpatients having a final discharge summary recorded in the medical record within two weeks of discharge. The overall rates have not increased since 1998 although the poorer rates have improved from 57% to 67% over the period (Figure 9.2). In 2002 and 2003 both the public and private sectors had similar rates. The New South Wales rates were significantly poorer in 2002 and 2003 (Table 9.2).
Table 9.2 Final discharge summary in medical record within two weeks. Stratum NSW QLD SA VIC WA Other
98
Year 2002 2003 2002 2003 2003 2002 2003 2002 2003 2002 2003
No HCOs 25 23 17 14 5 19 17 6 5 10 8
Numerator 7,939 8,932 3,542 3,051 2,101 6,418 7,322 2,035 2,081 3,835 3,138
Denominator 11,420 13,688 4,628 3,836 2,292 7,537 8,168 2,152 2,509 4,526 3,483
Stratum rate % 69.6 65.3 76.6 79.5 91.5 85.1 89.6 94.4 83.0 84.6 90.0
Standard error 3.1 3.0 4.9 5.6 7.3 3.8 3.9 7.2 7.0 4.9 5.9
Two private HCOs and 14 public HCOs submitted data for the indicators in this set. CI. 1.1 The rate of clients having five or more contacts with the outpatient service. In 2003, 37% of patients had five or more contacts. Ten of the 16 HCOs had rates between 28% and 51%. The New South Wales rate was the highest at 51% (Table 1.1). Table 1.1 Outpatients Clients having five or more outpatient contacts. No HCOs 6 10 Stratum rate % 50.7 32.0 Standard error 4.28 2.69 Stratum gains 2,859
CI. 1.2 The rate of outpatient clients being admitted to hospital for psychiatric reasons only once. In 2003, 8.2% of patients had one admission for psychiatric reasons only once. The better rate, 6.5%, was similar to the overall rate. CI. 1.3 The rate of outpatient clients being admitted to hospital for psychiatric reasons more than once. In 2003, 1.6% of outpatients were admitted to hospital more than once. The overall rate was twice the lower rate of 0.83%.
99
1.2
CI. 1.1 The rate of induction of labour for indications other than those listed above (excluding augmentation of labour) in patients undergoing induction. The proportions have remained between 33% and 34% for the last five years, with no evidence of a trend. Also, the HCOs with lower rates were significantly less than those with higher rates (Figure 1.1). This significant variation between HCOs, with 20% of the rates being less than 22% and another 20% being greater than 48% indicates that there is a capacity to reduce inductions by more than 4,000. The public and private HCOs had significantly different proportions of 27% and 46% respectively in 2002 and 2003 (Table 1.1). Figure 1.1 Induction for other than defined reasons (all inductions)
50
There are opportunities to determine the causes of the significantly high proportions in some individual HCOs. In all years, more than 1,600 inductions would not have been carried out if the HCOs that were outliers were to have the average rate of 34%.
40
Rate %
30
20
10
Obstretrics and Gynaecology Indicators Version 4 Table 1.1 Induction for other than defined reasons (all inductions) No HCOs 120 126 62 61 Stratum rate % 27.0 27.7 45.7 48.8 Standard error 0.69 0.72 0.93 1.07 Stratum gains 2,476 2,494
CI. 1.2 The rate of induction of labour for indications other than those listed above (including augmentation of labour) in all patients delivering. When the proportion of inductions is expressed as a proportion of all deliveries, the percentages also showed no trends, remaining at a little less than 9% over the last four years.The variation between HCOs has also remained large and relatively stable. The rate could be halved if the average were to be reduced to 4.4%, which is the rate for the better HCOs. This would reduce the proportion of inductions by about 6,000. The difference in proportions between the public and private sectors also provided evidence that lower rates could be achieved. The public HCOs had mean rates that were one half of the private rates. As with CI 1.1 there were HCOs with significantly high rates, which if their rates were reduced to the overall average would results in 2,000 less inductions. Both of these indicators show a consistent pattern: large clinical practice variations in the public and private sectors, and a large proportion of individual HCOs with high rates. Research to determine the appropriate criteria for inductions is recommended. Figure 1.2 Induction for other than defined reasons (all deliveries)
20
15
Rate %
10
Table 1.2 Induction for other than defined reasons (all deliveries) No HCOs 118 124 61 60 Stratum rate % 6.48 6.64 13.4 14.2 Standard error 0.25 0.27 0.37 0.43 Stratum gains 2,956 3,041
Area 2 Obstetrics The rate of vaginal delivery following primary caesarean section Rationale
This indicator is used to monitor the conduct of labour in those patients having a caesarean section for their first delivery. Evidence supports the finding that vaginal delivery after a previous caesarean birth is a reasonable option for many women. The recommended rate is not known, but rates of over 30% have been suggested. Thus, we have taken the desired rate to be high, and the potential gains are given in terms of increasing vaginal deliveries . The indicator is: CI. 2.1 Rate of vaginal delivery following primary caesarean section Results CI 2.1 Year 1998 1999 2000 2001 2002 2003 No HCOs 133 118 139 142 148 151 Numerator 1,879 1,790 2,118 1,876 2,143 2,174 Denominator 7,714 7,314 9,449 9,369 12,856 13,574 Rate % 24.4 24.5 22.4 20.0 16.7 16.0 Rate (20)% 17.2 18.1 15.7 14.0 11.5 11.0 Rate (80)% 30.1 34.1 30.4 26.6 25.0 21.3 Centile gains 440 707 756 612 1,072 721 Stratum gains 314 681 541 505 1,544 842 Outlier gains 68 74 115 56 173 39
CI. 2.1 The rate of vaginal delivery following a previous primary caesarean section. The rates for this indicator have declined, especially over the period 2000 2002. The proportion in 2003 was 16% compared to 24% in 1998 and 1999. Further there has been a decline in the centiles (Figure 2.1). This dramatic decline is probably due to a change in clinical attitudes to vaginal birth after a caesarean and research or studies on what would be a desirable rate could be conducted. The differences in rates remains high, with the HCOs with higher rates being twice that of the HCOs with lower rates. The difference between the States and the public and private sectors also provided evidence of significant variation in clinical practice (Table 2.1). Queensland (13%) and Victoria (13% and 15% in 2002 and 2003 respectively) had lower rates than the other States. The public rates are 6% higher than the private rates (18% and 12% respectively). Figure 2.1 Vaginal delivery following caesarean
40
30
Rate %
20 10 0 1998 1999 2000 2001 Year 2002 2003
Obstretrics and Gynaecology indicators Version 4 Table 2.1 Vaginal delivery following caesarean No HCOs 53 53 26 22 12 12 34 35 14 18 9 11 91 92 57 59 Stratum rate % 19.0 16.4 13.4 12.7 18.9 22.2 13.0 15.3 15.4 14.3 28.7 22.2 18.8 18.3 12.8 12.4 Standard error 0.83 0.67 1.23 0.89 1.95 1.35 0.98 0.83 1.89 1.09 2.25 1.56 0.66 0.48 0.89 0.61 Stratum gains 457 262 334 241 85 543 203 122 134 273 306
Year 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003
Numerator 905 743 287 307 173 255 434 442 138 219 206 208 1,591 1,575 552 599
Denominator 4,752 4,482 2,188 2,532 874 1,097 3,467 2,935 923 1,701 652 827 8,284 8,351 4,572 5,223
3.2
CI. 3.1 and 3.2 The rate of primary caesarean section for failure to progress after a period of labour with cervical dilatation of 3cm or less and cervical dilatation of more than 3cm. These indicators were reported for all patients undergoing a primary non-elective caesarean section, and estimated the proportion that failed to progress with dilatation of 3cm or less and the proportion with dilatation of 3cm or more. The rates for CI 3.1 were about 10% while for 3.2 they increased from 25% to 31% in 1998 2003. It is interesting to note that the differences in rates between the HCOs were larger in 1998 than in 2003: the rates for the HCOs with a higher proportion have declined by 7% for CI 3.1 and by 5.5% for CI 3.2. That is, the HCOs with the more extreme rates for caesarean have reduced their rates, while those with lower rates have remained stable (Figures 3.1 and 3.2). The reasons for these changes may be important to determine. The private sector had a higher mean rate of 15%, compared to the public rate of 8% for CI 3.1 (Table 3.1). For CI 3.2 the differences were small, the private sector rate was 29%, compared to the public rate of 31%. Figure 3.1 Primary caesarean section for failure to progress, < 3cm dilation
30
25
20
Rate %
15
10
Obstretrics and Gynaecology Indicators Version 4 Table 3.1 Primary caesarean section for failure to progress, < 3cm dilation. No HCOs 105 103 57 60 Stratum rate % 7.94 8.11 14.6 14.6 Standard error 0.50 0.41 0.76 0.65 Stratum gains 312 284
Figure 3.2 Primary caesarean section for failure to progress, > 3cm dilation
ea 50
40
Rate %
30
20
10
4.2
CI. 4.1 The rate of primary caesarean section for foetal distress in all deliveries. All three rates (mean and centiles) have shown an increase of about 0.6% over the six-year period. This increase may be related to changes in measuring foetal distress, the increase in caesarean section rates and/or to the increasing age of mothers who are having their first child. There have not been any significant differences between the strata, and the variation between the high and low rates is not large. Figure 4.1 Primary caesarean section for distress (all deliveries) CI. 4.2 The rate of primary caesarean section for foetal distress in all patients delivering by primary caesarean section only. 6 One of the reasons for a caesarean is foetal distress, and 5 a high rate may indicate that this procedure was carried out more appropriately. However, the rate was about 4 20% and the reasons for the other 80% of caesareans were not provided, some of which may be appropriate, 3 elective or unnecessary. Assuming that higher rates indicate better care, the rates have not shown a marked 2 increase. However, there are large differences between the 1 public and private sectors and the States (Table 4.2). These may reflect some known differences. For exam0 ple, the private sector has higher caesarean section 1998 1999 2000 2001 2002 2003 rates and hence it would be expected to have lower Year rates for foetal distress if the additional cases were due to higher rates of elective caesareans. The public and private rates differ by 8% (25% and 17% respectively). The differences between NSW and SA with rates of 24% and 28% were significantly different to Victoria with rates of 17% and 20% (Table 4.2). WA showed an unusual decline from 26% to 20%.
Rate %
Obstretrics and Gynaecology Indicators Version 4 Figure 4.2 Primary caesarean section for foetal distress (all primary caesareans)
30
25
20
Rate %
15
10
Table 4.2 Primary caesarean section for foetal distress (all primary caesareans) No HCOs 57 57 28 27 14 17 5 5 41 38 14 20 5 7 82 85 82 86 105 110 59 61 Stratum rate % 23.3 24.4 16.4 18.1 27.3 28.0 21.4 21.2 17.1 20.4 26.1 19.8 23.1 20.8 21.0 22.3 19.3 19.9 22.1 24.5 17.9 16.7 Standard error 0.78 0.78 1.1 0.92 2.0 1.6 3.5 2.9 0.94 0.99 2.3 1.4 2.8 2.0 0.61 0.54 1.0 0.96 0.64 0.52 0.84 0.69 Stratum gains 372 295 552 586 39 652 390 218 94 729
Stratum NSW QLD SA TAS VIC WA Other Metropolitan Rural Public Private
Year 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003
Numerator 2,185 2,093 796 1,042 403 539 99 121 1,060 1,021 291 534 164 276 3,814 4,424 1,184 1,202 3,460 4,132 1,538 1,494
Denominator 9,225 8,380 5,044 5,952 1,427 1,879 449 583 6,393 5,133 1,093 2,681 694 1,326 17,957 19,732 6,368 6,202 15,422 16,543 8,903 9,391
Area 5 Obstetrics Incidence of an intact lower genital tract in primiparous patients delivering vaginally
Rationale A high incidence of an intact perineum is considered to be a desirable outcome. Lower genital tract is defined as those structures below and not including the cervix. Surgical repair is defined as suture of the lower genital tract following delivery. This indicator relates to those patients who are having their first delivery. Factors leading to a low rate are a higher use of episiotomy (rates lower than 10% have been recommended) and tears while delivering. The indicator is: CI. 5.1 Rate of primiparous patients not requiring surgical repair of the lower genital tract Results CI 5.1 Year 1998 1999 2000 2001 2002 2003 No HCOs 180 165 181 177 173 172 Numerator 10,269 9,984 11,283 11,303 11,755 11,665 Denominator 35,526 33,792 39,157 37,030 40,512 40,439 Rate % 28.9 29.5 28.8 30.5 29.0 28.8 Rate (20)% 18.6 19.0 18.3 19.3 19.8 18.1 Rate (80)% 39.4 41.4 38.6 43.7 39.0 40.0 Centile gains 3,739 4,004 3,815 4,896 4,062 4,512 Stratum gains 6,345 4,360 5,958 5,645 6,314 6,817 Outlier gains 1,413 1,543 1,403 1,449 1,435 1,446
CI. 5.1 The rate of primiparous patients not requiring surgical repair of the lower genital tract. As shown in Figure 5.1, there has been no major change in the proportions, which have remained comparatively stable for the six years. However, there is substantial variation in the rates between HCOs, with 20% of rates being less than 18% and 20% being greater than 40%. The potential to improve this outcome is large in all years with about 4,000 more primiparous patients not requiring surgical repair of the lower genital tract if the 80th centile could be achieved. The State differences (Table 5.1) reflect an interesting pattern, with the smaller States (SA, Tasmania and the remainder) having higher rates of about 40%. In contrast, Victoria and NSW had lower rates of 25%. The large within strata variations suggest that there is an unusually large clinical variation in the use of episiotomy in Australia. Publications have shown that reducing the rate of episiotomy can increase the rate of intact perineum after delivery and hence the rate for this indicator6. These data suggest that there is a need to review the use of episiotomy in all HCOs, and to compare the practices in the States. Figure 5.1 Intact lower genital tracts in primiparous vaginal birth
50
40
Rate %
30
20
10
Rockner G. and Fianu-Jonasson A. Changed pattern in the use of episiotomy in Sweden. British J. Obstet. and Gynaecology (1999), 106, 95 101.
Obstretrics and Gynaecology Indicators Version 4 Table 5.1 Intact lower genital tracts in primiparous vaginal birth No HCOs 56 55 26 25 19 18 5 5 46 42 16 21 5 6 Stratum rate % 26.4 24.1 29.1 32.3 44.9 35.5 40.0 45.8 24.6 24.7 33.6 32.3 40.8 43.7 Standard error 1.11 1.32 2.04 2.00 2.68 2.71 4.63 4.63 1.44 1.58 2.68 2.41 3.65 3.48 Stratum gains 3,162 3,039 794 826 339 2,027 2,047 329 565
Year 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003
Numerator 4,468 3,336 1,466 1,982 1,326 1,185 404 537 2,451 2,371 986 1,367 654 887
Denominator 17,069 14,011 5,049 6,099 2,912 3,308 977 1,136 10,021 9,687 2,917 4,189 1,567 2,009
6.2
CI. 6.1 The rate of babies born with an Apgar score of four or below at five minutes post delivery. The rates showed a consistent although small decline from 0.8% to 0.6% for a sample of about half of the births in Australia. The higher and lower proportions also show a similar decline. The 20th centile is now 0.3% and the 80th is 0.7%, a rate that is now below the mean rate for the year 1998. For this indicator there were no significant differences between the public and private sectors or the States and rural/urban HCOs . Figure 6.1 Apgar score four or below at five minutes.
1 .0
0 .8
Rate % % Rate
0 .6
0 .4
0 .2
Obstretrics and Gynaecology Indicators Version 4 CI. 6.2 The rate of babies born with an Apgar score of six or below at ten minutes post delivery. The mean rates declined by more than half in 1999, from 0.8% to 0.3% but has remained at this level during the next five years. Figure 6.2 Apgar score six or below at ten minutes
1 .0
0 .8
Rate %
0 .6
0 .4
0 .2
For this indicator there were differences between the States with NSW and SA having lower rates. Table 6.2 Apgar score - six or below at ten minutes. No HCOs 60 19 13 36 21 10 Stratum rate % 0.23 0.38 0.17 0.47 0.38 0.46 Standard error 0.03 0.06 0.10 0.04 0.06 0.07 Stratum gains 18 71 20 20
Numerator 95 51 3 151 48 49
Area 7 Obstetrics Term babies transferred or admitted to a Neonatal Intensive Care Unit for reasons other than congenital abnormality
Rationale This indicator provides an index of the overall management of labour in terms of the outcome for the baby. A term baby is one born at 37 weeks gestation or later. Babies born after a prolonged pregnancy of 41 completed weeks or more, are included. Babies having routine or non-intensive care observations carried out in a neonatal intensive care unit are excluded. Transfers/admissions due to congenital abnormality are excluded. A neonatal intensive care unit is defined as a separate and self-contained facility in the hospital capable of providing complex multi system life support for an indefinite period. The indicator is: CI. 7.1 The rate of term babies transferred/admitted to a NICU for reasons other than congenital abnormality Results CI 7.1 Year 1998 1999 2000 2001 2002 2003 No HCOs 141 139 155 155 157 162 Numerator 1,024 1,283 1,355 1,093 1,305 1,414 Denominator 93,700 93,922 113,747 105,336 118,479 128,319 Rate % 1.09 1.37 1.19 1.04 1.10 1.10 Rate (20)% 0.36 0.39 0.27 0.36 0.22 0.21 Rate (80)% 1.32 1.55 1.17 1.28 1.32 1.21 Centile gains 688 915 1,044 718 1,039 1,149 Stratum gains 258 251 739 Outlier gains 327 526 501 392 477 614
CI. 7.1 The rate of term babies transferred/admitted to a neonatal intensive care unit for reasons other than congenital abnormality. The proportions of transfers have remained slightly over 1% during the six years. The higher and lower proportions show no clear trend, but there is a marked difference in the rates (1.2% and 0.2% in 2003). The variation in rates would be related to the availability of neonatal ICU facilities and the complexity of the births in the hospital. The differences between States in 2002 and 2003 is due to Victoria having lower rates of 0.7% and 0.4%, while Tasmania and other States had higher rates of over 4% (Table 7.1). The reasons for these higher rates should be determined by HCOs in these regions. Table CI. 7.1 Term babies admitted to NICU. No HCOs 57 60 21 20 14 14 5 5 40 39 15 18 5 6 Stratum rate % 0.98 1.00 0.94 0.99 1.00 1.07 4.29 5.39 0.72 0.43 0.92 0.93 4.76 3.00 Standard error 0.14 0.15 0.27 0.25 0.38 0.34 0.55 0.52 0.18 0.20 0.37 0.30 0.56 0.44 Stratum gains 273 99 120 213 130 152
Year 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003
Numerator 511 465 121 172 73 110 154 246 211 117 67 116 168 188
Denominator 53,878 48,428 13,455 17,778 7,123 9,888 3,382 4,296 30,204 29,404 7,212 12,566 3,225 5,959
CI. 8.1 The rate of blood transfusion during/post abdominal or vaginal hysterectomy (excluding laparoscopic hysterectomy). The decline over the three years 1998 to 2000 has been reversed in 2001 to 2003. The higher rates have shown a 1.2% increase from 1998 to 2003. There were significant differences between the States, where NSW and Tasmania had lower rates of 3%. However, there were also differences between the rural and metropolitan HCOs, with the metropolitan HCOs having a rate of 6% and rural HCOs having a rate of 3% (Table 8.1). Figure 8.1 Hysterectomy - blood transfusion
8 7 6
Rate %
Table 8.1 Hysterectomy blood transfusion Stratum NSW QLD SA TAS VIC WA Other Metropolitan Rural Year 2003 2003 2003 2003 2003 2003 2003 2002 2002 No HCOs 38 32 9 8 35 13 7 84 59 Numerator 64 93 48 25 117 47 51 443 77 Denominator 2,592 2,398 646 974 2,466 1,000 764 7,466 3,350 Stratum rate % 2.98 4.11 6.58 2.97 4.33 4.64 5.79 5.58 3.06 Standard error 0.40 0.41 0.80 0.65 0.41 0.64 0.73 0.25 0.37 Stratum gains 26 23 33 16 21 188
CI. 9.1 The rate of injury (with or without repair) to ureter/s or bladder during an abdominal or vaginal hysterectomy (excluding laparoscopic hysterectomy). The proportion of procedures with injury while low has increased to about 1% over the last two years. In 2002, the higher rate was due to one HCO with 54 injuries from 217 procedures, and excluding this HCO the mean rate was 0.8%. This then suggests a general trend for an increase from 0.7% to 1%. The rate for the poorer performing HCOs in 2003 was 1.4%.
of of of of
unplanned readmissions within 28 days after cataract surgery treatment for infection within 28 days after cataract surgery unplanned overnight stay after cataract surgery anterior vitrectomy in cataract surgery
1.2
1.3 1.4
CI. 1.1 The rate of unplanned readmissions within 28 days after cataract surgery. The rates have decreased from more than 0.5% in 1998 to 0.32% in 2003. The decrease also occurred in the poorer (higher) rates which halved over that period. The difference between the better rates and the poorer rates has diminished over the period (Figure 1.1). This reduction in variation between HCOs is indicative of improvement having been made across the majority of HCOs. The public rate, 0.66%, was more than double the private rate in 2002 (Table 1.1). Figure 1.1 Cataract surgery unplanned readmissions within 28 days.
1 .0
0 .8
Rate %
0 .6
0 .4
0 .2
Ophthamology and Excimer Laser Indicators Version 3 Table 1.1 Cataract surgery - unplanned readmissions within 28 days. No HCOs 26 93 Stratum rate % 0.66 0.29 Standard error 0.04 0.02 Stratum gains 49
CI. 1.2 The rate of treatment for infection within 28 days after cataract surgery. The rates were low and have ranged between 0.07% and 0.18% from 1998 to 2003. For 2003, the rates for the HCOs were very similar, except for one outlier. CI. 1.3 The rate of unplanned overnight stay after cataract surgery. The rate was 0.65% in 2003. The higher rate was six times the lower rate indicating that the rates could be reduced. Those HCOs with statistically significantly high rates were from both the public and private sectors and had rates in excess of 2%. One HCO reported a rate of 47% on more than 100 patients in a six-month period. CI. 1.4 The rate of anterior vitrectomy in cataract surgery. In 2003, the overall rate was 0.84%, with rates for the centiles being 0.11% and 0.66%. The public and metropolitan rates were higher (Table 1.4). The rate for the public metropolitan HCOs was 2.2%. This rate may be consistent with casemix differences. If this is not the case then the causes should be determined. Table 1.4 Cataract surgery anterior vitrectomy No HCOs 38 30 15 53 Stratum rate % 1.03 0.41 1.83 0.48 Standard error 0.10 0.14 0.12 0.07 Stratum gains 148 121
2.2
2.3
CI. 2.1 The rate of re-admissions (related to the operated eye) within 28 days following glaucoma surgery. The overall rates were between 1.2% and 3.7%. In the last two years the overall rates were similar to the lower rates. Two large specialist hospitals reported statistically significantly higher rates of between 4% and 10% several times during the six year period. These high rates may be consistent with differences in casemix. These two HCOs accounted for more than two thirds of the unplanned readmissions following this surgery and more than half of the patients and hence contributed substantially to the differences in rates between public and private HCOs (Table 2.1). Table 2.1 Glaucoma surgery unplanned readmissions within 28 days No HCOs 12 14 43 31 Stratum rate % 3.71 6.28 0.66 0.35 Standard error 0.36 0.62 0.20 0.25 Stratum gains 15 29
Numerator 19 34 10 8
CI. 2.2 The rate of re-admission within 28 days following glaucoma surgery due to infection in the operated eye. The overall rates have been low, less than 0.2% in all years since 1998 with the exception of 2001. In 2001 there were 10 re-admissions and nine of these were to one major metropolitan specialist hospital. When that HCO was excluded for 2001 there was no variation between HCOs and all rates were less than 0.2% in all years. CI. 2.3 The rate of patients having LOS greater than three days following glaucoma surgery. The rates were between 1.7% and 7.1% during the period 1998 to 2003. In 2002 and 2003 the differences between the overall rates and the better rates were relatively small. The public private differences in 2003 (Table 2.3) were primarily due to a single specialist hospital, which had rates between 22% and 41% from 1999 onwards.
Ophthamology and Excimer Laser Indicators Version 3 Table 2.3 Glaucoma surgery LOS greater than three days. No HCOs 15 28 Stratum rate % 11.7 1.14 Standard error 1.17 0.86 Stratum gains 52
Numerator 63 3
3.2
3.3
3.4
CI. 3.1 The rate of unplanned re-admissions within 28 days of discharge following retinal detachment surgery. The overall rates were between 2.6% and 5.1%. The public rate was twice the private rate in 2003 (Table 3.1). The better rate was less than half the overall rate and there was considerable variation within both the public and private sectors indicating that the rates could be improved. Table 3.1 Retinal detachment surgery unplanned readmissions within 28 days. No HCOs 9 13 Stratum rate % 4.48 2.24 Standard error 0.66 0.76 Stratum gains 31
Numerator 63 23
CI. 3.2 The rate of unplanned re-admissions within 28 days of discharge following retinal detachment surgery, due to infection in the operated eye. The overall rates were low, less than 0.2% since 1999. The data provide no evidence of differences between HCOs.
Ophthamology and Excimer Laser Indicators Version 3 CI. 3.3 The rate of patients with a LOS greater than four days, following retinal detachment surgery. The overall rates have decreased from 3.6% in 1998 to 2.5% in 2003. The rates for the poorer performing HCOs decreased from 19% in 1998 to around 7% in subsequent years (Figure 3.3). The public HCO rate was three times the private rate in 2003 (Table 3.3). Figure 3.3 Retinal detachment surgery LOS greater than four days.
20
15
Rate %
10
Table 3.3 Retinal detachment surgery LOS greater than four days. No HCOs 10 12 Stratum rate % 3.39 0.97 Standard error 0.46 0.58 Stratum gains 34
Numerator 51 6
CI. 3.4 The rate for a revision of the operated eye within 28 days, following retinal detachment surgery. The overall rate was 3.6% in 2003. The public HCO was ten times the private rate in 2003 (Table 3.4), largely due to a major specialist HCO, which had more than nine tenths of the revisions and more than two thirds of the patients. Table 3.4 Retinal detachment surgery revision within 28 days. No HCOs 6 12 Stratum rate % 6.17 0.59 Standard error 0.50 0.55 Stratum gains 56
Numerator 65 3
4.2
CI. 4.1 The rate of surface ablation re-treatment. A small proportion of HCOs reported this indicator on a small proportion of patients in both 2002 and 2003. In 2003 the overall rate was 5.5% and there was no evidence of differences between HCOs. In 2002 one private HCO contributed eight of the re-treatments in less than one third of the patients. Consequently the remaining HCOs had one re-treatment in the remaining 79 patients, a rate of less than 1% for 2002. CI. 4.2 The rate of non-surface ablation re-treatment. A small proportion of HCOs reported this indicator on few patients in 2003. Since 1998 the overall rates have ranged between 5% and 10%. In 2002, one private HCO had more than half the re-treatments in approximately one quarter of the patients. The rate of the remaining HCOs in 2002 was 4%.
CI. 5.1 The rate for which patients are treated for infection in the operated eye within 28 days of excimer laser surgery. For all years between 1999 and 2003 except 2000 the rates were low, being less than 0.2%.
CI 1.1 The rate of unplanned return following acute care attendance. The rates were close to 3% in both 2002 and 2003. Six HCOs submitted data in both years.
CI. 1.2 The rate for re-treatment of teeth within 12 months of restorative treatment. The rate in 2003 was 5.1%. Eight public HCOs submitted data in 2003 and there was considerable differences in the rates: a lower rate of 0.7% compared to 6.5% for the higher rate.
Oral Health Indicators Version 1 The indicator is: CI. 1.3 Rate of attendances for complications within seven days of routine extraction Results CI 1.3 Year 2002 2003 No HCOs 4 10 Numerator 58 248 Denominator 2,818 14,629 Rate % 2.06 1.70 Rate (20)% 2.06 0.22 Rate (80)% 2.06 2.36 Centile gains 215 Stratum gains Outlier gains 48
CI. 1.3 The rate of return for complications within seven days of routine extraction. The overall rates in both years were close to 2%. The ten HCOs that submitted in 2003 were public and had significant differences in their rates.
CI. 1.4 The rate of return with complications within seven days of surgical extraction. Only two HCOs submitted data in 2002. The four public HCOs that submitted in 2003 had a combined rate of 5%.
CI. 1.5 The rate of extraction of teeth within 12 months of endodontic treatment. The five public HCOs that submitted data in 2003 had a combined rate of 1.4%.
CI. 1.6 The rate of unplanned return within 12 months after having crown and bridge treatment. One private HCO submitted data in 2002 and had a rate equal to 3.4%. No HCOs reported in 2003.
Denture remakes
Rationale The purpose of this indicator is to identify cases where the patient is unable to wear the current denture due to functional problems or aesthetic failure and the denture is replaced. It is important to exclude from this indicator cases where the remake is planned, e.g. after immediate denture or after recent full clearance. The indicator is: CI. 1.7 The rate of denture remakes within 12 months Results CI 1.7 Year 2002 2003 No HCOs 3 5 Numerator 14 60 Denominator 839 3,143 Rate % 1.67 1.91 Rate (20)% 1.35 1.25 Rate (80)% 2.27 2.21 Centile gains 2 20 Stratum gains Outlier gains
CI. 1.7 The rate of denture remakes within 12 months Public HCOs alone submitted data for this indicator. The rates were similar, 1.7% in 2002 and 1.9% in 2003.
CI. 2.1 2.3 Modes of treatment. In 2002 a single private HCO submitted data for 205 patients.100 had anaesthesia alone, and 105 were supplemented with sedation. In 2003, a public HCO provided data on 41 patients who all had general anaesthesia.
CI. 3.1 The rate of completed courses of endodontic treatment on the same tooth within six months of initial treatment. Two HCOs, one private, submitted data in 2002 and their combined rate was 1.1%. The private HCO had no re-treatments within six months. The public HCO rate in 2002 was 27%. In 2003, a different HCO had a rate of 0%. CI. 3.2 The rate of re-treatment of teeth within 12 months of completing a course of endodontic treatment. The two public and one private HCO in 2002 had no returns. There was a single return from one of the three public HCOs in 2003.
1.2
CI. 1.1 The rate of infants admitted as inpatients having documented current immunisation status. The overall rates have not improved since 1998 and range from 80% to 91%. However, the better rates have been consistently greater than 95% since 1998 which indicates that there is potential to improve the rate of documenting immunisation status. New South Wales and Queensland reported the highest rates in 2002 and 2003. Table 1.1 Paediatric inpatients documented immunization status. No HCOs 19 17 10 11 5 7 6 Stratum rate % 92.1 91.2 90.7 92.9 86.3 84.5 51.3 Standard error 1.09 2.98 1.94 6.49 11.1 2.59 4.23 Stratum gains 109 1,392
CI. 1.2 The rate of catch-up or planning for immunisation in inpatient infants whose immunisations are not up-to-date. The rates have increased from 21% in 1998 to approximately 40% in 2002 and 2003. During the period from 1998 the poorer rates have increased from 6% to more than 20% whereas the better rates have varied around 60% since 2000. These results are extremely low given that immunisation is one of the health priorities in Australia, and rates of 90% are obtained by GPs. The differences in the higher and lower rates indicate that the rates could be improved. In 2003, the metropolitan HCOs and Queensland had low rates (Table 1.2).
Paediatric Indicators Version 3 Table 1.2 Paediatric inpatients - catch up immunisation. No HCOs 13 13 9 10 5 6 9 20 Stratum rate % 34.6 49.2 58.0 24.6 43.3 46.6 32.2 45.7 Standard error 5.25 4.17 7.13 4.08 8.49 6.46 4.13 4.53 Stratum gains 108 117 82
Area 2 Asthma
Rationale Asthma is a common childhood condition, the severity and frequency of which may be decreased by careful management. Clinical indicator 2.2 refers to those children re-admitted to the same organisation within 28 days. The indicators are: CI. 2.1 The average length of stay (ALOS) in all children admitted with a primary diagnosis of asthma CI. 2.2 The average length of stay in children admitted with a primary diagnosis of asthma, excluding same day admissions CI. 2.3 The rate of children with a primary diagnosis of asthma, having a readmission to hospital for asthma within 28 days Results CI 2.1 Year 1998 1999 2000 2001 2002 2003 2001 2002 2003 No HCOs 65 61 63 69 66 65 64 64 63 Mean 2.00 1.83 1.80 1.77 1.74 1.67 1.85 1.89 1.83 Std 0.56 0.44 0.47 0.41 0.43 0.42 0.40 0.51 0.54 20th centile 1.50 1.50 1.46 1.41 1.40 1.32 1.41 1.48 1.40 Median 1.91 1.78 1.73 1.70 1.70 1.62 1.86 1.81 1.73 80th centile 2.50 2.11 2.11 2.10 2.00 1.92 2.15 2.17 2.28
2.2
CI 2.3
No HCOs 61 58 64 66 62 64
Centile gains 88 81 82 89 82 93
Stratum gains 98 61 36 64 31 82
Outlier gains 39 31 15 24 16 11
CI. 2.1 The average length of stay in all children admitted with a primary diagnosis of asthma. The average length of stay, ALOS, has decreased from 2.0 days in 1998 to 1.7 days in 2003. A similar trend in the median occurred. Twenty percent of HCOs had ALOS exceeding 2.5 days in 1998 and in 2003 twenty percent had ALOS above 1.9 days. Western Australia reported the highest average length of stay in 2003 (Table 2.1). Table 2.1 Average length of stay for children admitted for asthma No HCOs 28 20 6 5 18 12 6
Paediatric Indicators Version 3 CI. 2.2 The average length of stay in children admitted with a primary diagnosis of asthma, excluding same day admissions. The average length of stay when same day admissions are excluded has remained about 1.85 days since 2001. Twenty percent of HCOs had ALOS exceeding 2.2 days between 2001 and 2003. Western Australia reported the highest average length of stay in 2003 (Table 2.2). Table 2.2 Average length of stay for children admitted for asthma: excluding same day admissions No HCOs 27 17 6 5 17 9 6
CI. 2.3 The rate of children with a primary diagnosis of asthma, having a readmission to hospital for asthma within 28 days. The rates were 4.0% in 2002 and 2003 and the poorer rates were 4.2% and 4.6% in those years. Since 1998 the better rates have been approximately 1% lower than the average rates. The lack of trend or a low centile suggests that it may be hard to reduce this rate. Western Australia had the highest readmission rates in 2002 and 2003. No HCOs 23 21 12 17 6 9 10 6 5 6 11 Stratum rate % 3.49 3.98 3.77 2.99 4.48 3.50 4.27 6.20 4.91 3.44 4.23 Standard error 0.22 0.28 3.37 0.38 0.35 0.37 0.40 0.44 0.47 0.54 0.34 Stratum gains 26 11 16 19 17 22
Year 2002 2003 2002 2003 2002 2002 2003 2002 2003 2002 2003
Numerator 98 115 37 31 58 34 56 49 48 13 73
Denominator 2,883 2,664 1,003 1,442 1,179 1,032 1,267 724 921 486 1,763
1.4
CI. 1.1 The rate of urgent serum/plasma potassium validated report results with a turn-around-time less than 60 minutes, during normal working hours. The rate was 77.6% in 2003, with the better rates being 86.5%. Twenty-seven of the thirty HCOs were public and they represented 99% of the urgent validated reports during normal working hours. The better rate was 25 percentage points above the poorer rate. The variation between HCOs suggests that there is potential to increase the rates to close to 90%. Victoria had the lowest rates in 2003 (Table 1.2). For years 1998 to 2002 this indicator was defined in terms of turn around time less than 45 minutes. The rates obtained were about 65%. Table 1.1 Urgent serum/plasma potassium within 60 minutes, in working hours No HCOs 12 9 9 Stratum rate % 80.8 62.4 87.8 Standard error 2.50 2.87 2.75 Stratum gains 2,695 7,396
Pathology Indicators Version 2 CI. 1.2 The rate of urgent serum/plasma potassium validated report results with a turn-around-time less than 60 minutes, outof-hours. The rate was 83.1% in 2003, with the better rates being 92.0%. Twenty six of the twenty eight HCOs were public and they represented 99% of the urgent validated results out of hours. The better rate was 20 percentage points above the poorer rate. The variation between HCOs suggests that there is potential to increase the rates to over 90%. Victoria had the lowest rates in 2003 (Table 1.2). For years 1998 to 2002 this indicator was defined in terms of turn around time less than 45 minutes. The rates obtained were about 73%. Table 1.2 Urgent serum/plasma potassium within 60 minutes, out of hours No HCOs 11 8 9 Stratum rate % 87.7 74.1 92.3 Standard error 2.67 2.12 2.59 Stratum gains 7,192
CI. 1.3 The rate of non-urgent serum/plasma potassium validated report results with a turn-around-time less than 60 minutes, during normal working hours. Since 1999 the rates have varied around 55% while the better rates have varied around 80% (Figure 1.3). The poorer rates have been very low at about 30%. Twenty five of the twenty eight HCOs were public and they represented 99% of the non-urgent validated reports. A proportion of HCOs had statistically significantly low rates in three or more of the years since 1998. Victoria had the lowest rates in 2002 and 2003 (Table 1.3). Figure 1.3 Non-urgent serum/plasma potassium within 60 minutes, in working hours.
100 90 80 70
Rate %
Table 1.3 Non-urgent serum/plasma potassium within 60 minutes, in working No HCOs 9 11 10 9 5 8 Stratum rate % 69.8 71.6 43.5 45.6 39.7 72.4 Standard error 4.27 4.87 6.75 5.14 5.22 6.80 Stratum gains _ 18,139 32,350 34,683
Pathology Indicators Version 2 CI. 1.4 The rate of non-urgent serum/plasma potassium results validated with a turn-around-time less than 60 minutes, out-of-hours. The rates have varied around 70% while the better rates have varied around 85% (Figure 1.4). The poorer HCOs have validated the results for only 50% requests. Twenty four of the twenty five HCOs were public and they represented 99% of the non-urgent validated reports. A proportion of HCOs have had statistically significantly low rates in three or more of the years since 1998. Victoria had the lowest rates in 2002 and 2003 (Table 1.4). The large differences in rates between HCOs for all four indicators suggests that turn around times could be significantly improved if the methods used in the better HCOs were applied to all organisations. Figure 1.4 Non-urgent serum/plasma potassium within 60 minutes, out of hours
100 90 80 70
Rate %
Table 1.4 Non-urgent serum/plasma potassium within 60 minutes, out of hours No HCOs 8 9 10 8 5 8 Stratum rate % 85.0 84.3 59.1 61.1 61.2 62.2 Standard error 3.47 4.29 5.62 3.72 3.84 6.11 Stratum gains 6,390 12,066 12,589 4,258
Area 2 Haematology
Rationale Timely knowledge of haemoglobin provides valuable information on the causation and management of certain disorders. The indicators are: CI. 2.1 The rate of urgent haemoglobin validated report results with a turn-around-time less than 60 minutes, during normal working hours CI. 2.2 The rate of urgent haemoglobin validated report results with a turn-around-time less than 60 minutes, out-of-hours CI. 2.3 The rate of non-urgent haemoglobin validated report results with a turn-around-time less than 60 minutes, during working hours CI. 2.4 The rate of non-urgent haemoglobin results validated with a turn-around-time less than 60 minutes, out-of-hours No HCOs 30 28 13 14 19 21 23 27 12 14 17 20 22 25 Rate % 87.7 92.1 69.9 76.2 77.7 76.9 70.1 74.2 83.0 84.1 84.5 86.3 79.6 84.5 Rate (20)% 81.3 86.2 44.8 54.9 47.3 56.7 65.5 65.9 44.6 61.5 71.4 63.5 67.0 80.3 Rate (80)% 95.8 95.4 90.1 90.7 89.7 85.8 85.3 89.0 91.2 90.9 92.8 95.1 94.3 92.9 Centile gains 10,300 3,112 37,313 30,996 39,402 32,352 56,064 58,480 5,394 4,695 10,302 11,039 19,244 10,526 Stratum gains 27,535 46,077 7,575 17,122 Outlier gains 5,028 1,481 13,796 9,472 18,801 14,983 28,682 23,505 2,934 1,589 4,678 3,559 7,398 5,203
Year 2003 2003 1998 1999 2000 2001 2002 2003 1998 1999 2000 2001 2002 2003
Numerator 110,893 85,942 129,046 163,248 255,796 280,051 256,984 294,323 55,009 57,584 105,718 108,790 104,816 106,915
Denominator 126,496 93,323 184,615 214,202 329,089 364,204 366,844 396,571 66,244 68,510 125,049 126,022 131,605 126,474
2.4
CI. 2.1 The rate of urgent haemoglobin validated report results with a turn-around-time less than 60 minutes, during normal working hours. The rate was 88% in 2003 and the better rate was 96%. Twenty-eight of the thirty HCOs were public and they represented 99% of the urgent validated results during normal working hours. The variation between HCOs is relatively small, with the lower rate being 81% or seven percent below the overall rate. For years 1998 to 2002 this indicator was defined in terms of turn around time less than 45 minutes. The rates obtained were about 86%. CI. 2.2 The rate of urgent haemoglobin validated report results with a turn-around-time less than 60 minutes, out-of-hours. The rate was 92% in 2003, with the better rates being 95%. The lower rate was relatively high being 81%. Twentyseven of the twenty eight HCOs were public and they represented 99% of the urgent validated results out of hours. CI. 2.3 The rate of non-urgent haemoglobin validated report results with a turn-around-time less than 60 minutes, during working hours. Since 1998 the rates have varied around 74% while the better rates have varied around 87% (Figure 2.3). The lower rates were less than 66%. Twenty five of the twenty seven HCOs were public and they represented 99% of the nonurgent validated reports. A proportion of HCOs have had statistically significantly low rates in three or more of the years since 1998. There is potential to increase the overall rate.
Pathology Indicators Version 2 Figure 2.3 Non-urgent haemoglobin within 60 minutes, in working hours
100 90 80 70
Rate %
CI. 2.4 The rate of non-urgent haemoglobin results validated with a turn-around-time less than 60 minutes, out-of-hours. The rates have varied around 83% while the better rates have varied around 92%. The poorer rates have made significant improvements reaching 85% in 2003 (Figure 2.4). Twenty four of the twenty five HCOs were public and they represented 99% of the non-urgent validated reports. A proportion of HCOs have had statistically significantly low rates in three or more of the years since 1998. Figure 2.4 Non-urgent haemoglobin within 60 minutes, out of hours
100 90 80 70
Rate %
3.2
CI. 3.1 The rate of validated small biopsy results from the gastro-intestinal tract with a turn-around-time less than 48 hours. The rates have varied around 82% since 2000. The better rates were in excess 96% between 1998 and 2002 indicating that there is potential to increase the overall rate by approximately 10%. Seventeen of the eighteen HCOs were public. CI. 3.2 The rate of validated large biopsy results from the gastro-intestinal tract with a turn-around-time less than 96 hours. The rates do not appear to have improved since 1998. The better rates were in excess 93% between 1998 and 2002 indicating that there is potential to increase the overall rate by approximately 10%. Seventeen of the eighteen HCOs were public. Figure 3.2 GIT large biopsy in less than 96 hours
100 90 80 70
Rate %
CI. 1.1 The rate of patients waiting more than 21 days, from the date ready for care, to the date of commencing radiotherapy treatment. The overall rates have increased from 10% in 1998 to 36% in 2003. The better rates have also deteriorated from 2% to 17%, while the poorer rates are now over 50%. There was no relationship between the numbers of patients and the rates. All reports were from metropolitan public HCOs. In 1999 one in ten ready for care patients and in 2003 one in three waited for more than 21 days to the commencement of radiotherapy. This worsening rate also reflects a shortage of supply. In was estimated that in Australia in 2002, there were approximately 15,000 patients who could have benefited from radiotherapy but were unable to access it.7 Figure 1.1 Wait more than 21 days after ready for care.
60
50
40
30
20
10
7 Kenny, Liz & Lehman, Margot (2004) Sequential audits of unacceptable delays in radiation therapy in Australia and New Zealand. Australasian Radiology 48 (1), 29-34. 138 Determining the Potential to Improve Quality of Care
CI. 2.1 The rate of treatment courses provided, where CT planning was utilised during the period under study. The rate of CT planning was 70% in 2003. One HCO reported a rate of 100%, another 81%. There is potential to increase the rate. All HCOs were public metropolitan.
CI. 3.1 The rate of patients entered on prospective clinical trials. The rate was rather low, being 11.8%. All HCOs were public metropolitan.
who survived three years and those who died within the three years; whether or not a total laryngectomy was performed.
The indicator is: CI. 4.1 The rate of patients who had radiotherapy treatment for glottic cancer (T1-2 N0 M0), having complete follow up (T1-2 N0 M0 signifies Tumour stage 1 or 2, no metastasis to regional lymph nodes and no distant metastases) Results CI 4.1 Year 1999 2000 2001 2002 2003 No HCOs 7 7 6 7 6 Numerator 55 60 74 47 20 Denominator 71 64 84 70 32 Rate % 77.5 93.8 88.1 67.1 62.5 Rate (20)% 75.9 93.7 86.4 65.9 63.7 Rate (80)% 93.0 93.7 92.2 75.7 69.1 Centile gains 11 3 6 2 Stratum gains Outlier gains 4
CI. 4.1 The rate of patients who had radiotherapy treatment for glottic cancer (T1-2 N0 M0), having complete follow up. The rates have varied between 63% and 94% but the numbers of patients were small, varying between 32 and 84. All HCOs were public metropolitan. There was little variation between HCOs in 2003.
who survived three years and those who died within the three years; the presence or absence of proctitis occurring at least three months after treatment.
The indicator is: CI. 5.1 The rate of patients who had radiotherapy treatment for prostate cancer (T1-4 N0/Nx) having complete follow up (T1-4 N0/Nx signifies tumour stage 1 to 4, no metastasis to regional lymph nodes or unable to assess) Results CI 5.1 Year 1999 2000 2001 2002 2003 No HCOs 6 6 5 7 7 Numerator 458 502 220 144 236 Denominator 639 618 266 207 314 Rate % 71.7 81.2 82.7 69.6 75.2 Rate (20)% 75.5 77.6 81.9 52.0 55.4 Rate (80)% 97.4 92.4 97.0 87.4 95.7 Centile gains 164 69 37 36 64 Stratum gains Outlier gains 90 28 11 9
CI. 5.1 The rate of patients who had radiotherapy treatment for prostate cancer (T1-4 N0/Nx) having complete follow up. The rates for complete follow up in these patients did not change since 1998, varying around 75%. The rates of the better performing HCOs remained close to 90%. There is considerable variation between the better and poorer performing HCOs and consequently, the rate could be improved.
who survived five years and those who died within the five years; whether or not a total mastectomy as performed.
The indicator is: CI. 6.1 The rate of patients who had radiotherapy treatment for breast conservation (T1-3 N0-1 M0), having complete follow up (T1-3 N0-1 M0 signifies tumour stage 1 to 3, no metastasis regional lymph nodes or metastasis to movable ipsilateral axillary node(s) and no distant metastases) Results CI 6.1 Year 2003 No HCOs 4 Numerator 220 Denominator 349 Rate % 63.0 Rate (20)% 39.6 Rate (80)% 76.0 Centile gains 45 Stratum gains Outlier gains 14
CI. 6.1 The rate of patients who had radiotherapy treatment for breast conservation (T1-3 N0-1 M0), having complete follow up. The rate of complete follow up in these patients was 63%. The better rate was 76%. All HCOs were public metropolitan.
CI. 1.1 The rate of reports on non-procedural non-urgent plain radiographs not available in the form defined above within 24 hours. The rates have improved, decreasing from 47% in 1998 to around 30% in subsequent years (Figure 1.1). Public HCOs accounted for 95% of radiographs covered by this indicator. There was considerable variation between HCOs and the better rates were 7% or less. Hence there is potential to increase the proportion of non-urgent plain radiographs that are available within 24 hours from 75% to about 95%. Figure 1.1 Nonprocedural non-urgent plain radiographs not available within 24 hours
60
50
Rate %
40
30
20
10
2.2
2.3
2.4
CI. 2.1 The rate for patients undergoing cerebral angiography to have documented evidence of a temporary neurological deficit following the procedure. The overall rates have declined from about 1.6% to 1.2%. There was no major variation between the rates for HCO. In 2003, two thirds of the angiographies were from public HCOs. In 2003 half of the HCOs reported no cases. CI. 2.2 The rate of patients undergoing cerebral angiography having documented evidence of stroke within 24 hours of the procedure. The overall rates were small and did not change substantially since 1998. In 2003, two thirds of the angiographies were from public HCOs and three quarters of the HCOs reported no cases. For all years, there was little variation in rates between the HCOs. CI. 2.3 The rate of patients undergoing cerebral angiography who die within 24 hours of the procedure. This is also a rare event, the rates being no more than 0.5% since 1998. There was no evidence of differences between HCOs in 2002 and 2003.
Radiology Indicators Version 2 CI. 2.4 The rate of patients undergoing percutaneous trans pleural biopsy of the lung or mediastinum, having documented evidence of pneumothorax and/or haemothorax requiring intervention following the procedure. The rates have decreased from 20% in 1998 to around 12% in subsequent years (Figure 2.4). The better rates have decreased from 14% to 4% during the period. There is considerable variation between HCOs that is not accounted for by stratum differences. This suggests that the rates could continue to improve. Figure 2.4 Percutaneous trans pleural biopsy lung/mediastinum pneumothorax/haemothorax
30
25
20
15
10
1.2
CI. 1.1 The rate for patients admitted to a rehabilitation unit/facility having no documented evidence of a functional assessment within seven working days of patient admission. The rates have decreased from 14% in 1998 to 3% in 2003 (Figure 1.1). The proportion of HCOs having no patients without documented evidence of functional assessment has increased from 21% in 1998 to 46% in 2003. This represents a significant improvement and since the better HCOs have rates of 0% there is potential to improve further. Figure 1.1 No functional assessment documented within seven working days
30
25
20
Rate %
15
10
Rehabilitation Medicine Indicators Version 2 CI. 1.2 The rate of patients having no documented evidence of a functional assessment prior to cessation of an active inpatient rehabilitation program. The rates have decreased from 21% in 1998 to 4% in 2003 (Figure 1.2). The proportion of HCOs providing a functional assessment on all their patients has increased from 16% in 1998 to 44% in 2003. As with CI 1.1 above, this represents a significant improvement and the rates should continue to improve. NSW had the lowest rate in 2003. Figure 1.2 No functional assessment documented prior to cessation of program
30
25
20
Rate %
15
10
Table 1.2 No functional assessment documented prior to cessation of program No HCOs 38 12 5 22 8 Stratum rate % 1.97 6.07 6.08 4.81 11.9 Standard error 0.86 2.25 2.00 0.91 2.17 Stratum gains 100 330 205
CI. 2.1 The rate of patients admitted to a rehabilitation unit/facility having no documented established multi-disciplinary rehabilitation plan within seven working days of patient admission. The rates have decreased from 14% in 1998 to 3% in 2003 (Figure 2.1). The proportion of HCOs having documented rehabilitation plan on all their patients has increased from 40% in 1998 to 58% in 2003. This represents a significant improvement and Figure 2.1 suggests that further improvement is likely. The public rate of 5% was significantly higher than the private rate in 2003 (Table 2.1). Figure 2.1 No documented multi-disciplinary rehabilitation plan within seven working days.
20
15
Rate %
10
Table 2.1 No documented multi-disciplinary rehabilitation plan within seven working days. No HCOs 46 39 Stratum rate % 4.94 0.82 Standard error 0.70 0.69 Stratum gains 613
CI. 2.2 The rate of patients having no appropriate discharge plan on separation. The rates have decreased from 18.2% in 1998 to 2.9% in 2003 (Figure 2.2). The proportion of HCOs having a documented rehabilitation plan on all their patients has increased from 40% in 1998 to 58% in 2003. This represents a significant improvement and there is potential to improve further. NSW reported the lowest rate in 2002 and 2003 (Table 2.2). Figure 2.2 No discharge plan on separation
20
15
Rate %
10
Table 2.2 No discharge plan on separation No HCOs 39 38 12 12 23 23 11 11 Stratum rate % 1.60 1.49 2.64 2.16 6.04 3.12 7.41 7.81 Standard error 0.77 0.72 2.03 1.67 0.92 0.78 1.53 1.35 Stratum gains 395 190 188 244
CI. 3.1 The rate of patients having an unplanned interruption to their rehabilitation program. The rates have varied around 7.5% since 1998 and there was no trend. The public rate, 7.8%, was higher than the private rate of 5.3%. Figure 3.1 Unplanned interruption to program
20
15
Rate %
10
Table 3.1 Unplanned interruption to program No HCOs 44 42 Stratum rate % 7.78 5.32 Standard error 0.44 0.39 Stratum gains 372
4.2
1 .5
1 .0
0 .5
CI. 4.1 The rate of rehabilitation patient deaths within the rehabilitation unit/facility. The rates have decreased from more than 1% to approximately 0.7% in 2003. Both the better and the poorer rates have also decreased during that period (Figure 4.1). There were significant differences between the higher rates of 1.3% and the lower rates of 0.2%. Also, some organisations had rates that were statistically high and the proportion of outlier deaths in 2003 was 47. The rates were lowest in South Australia and Victoria in 2003.
Rate %
Table 4.1 Rehabilitation patient deaths No HCOs 42 13 5 24 6 Stratum rate % 0.43 1.07 0.23 0.69 2.74 Standard error 0.08 0.20 0.22 0.09 0.25 Stratum gains 31 24 60 44
Numerator 55 37 3 88 58
CI. 4.2 The rate of addressing rehabilitation patient deaths within an audit process/quality improvement study. The rates have increased from 62% in 1998 to 95% in 2003. The proportion of HCOs reporting rates of 100% has increased from 69% to 85% over the period. In 2003, ten out of 208 cases were not formally reviewed.
152 Determining the Potential to Improve Quality of Care
CI. 1.1 The rate of patients having a pyloromyotomy in which mucosal perforation occurs and is detected at the time of operation or later. The proportion of pyloromyotomy patients reported has more than doubled since 1998 and the rates of mucosal perforation have declined from more than 4% in 1998 to less than 2% in 2003. Since 1998 there has been no statistically significant difference between HCOs in any year.
Appendicectomy
Rationale Appendicectomy is a commonly performed operation in childhood. Good management should achieve a low rate of negative (normal) histology. Acute appendicitis excludes interval and incidental appendicectomy. Significant other intra abdominal pathology includes findings such as perforated Meckels diverticulum, torsion of an ovarian cyst, but excludes mesenteric adenitis. The indicators are: CI. 1.2 The rate of children with a pre-operative diagnosis of acute appendicitis, who undergo appendicectomy having normal histology CI. 1.3 The rate of children with a pre-operative diagnosis of acute appendicitis who undergo appendicectomy with normal histology, but have significant other intra abdominal pathology Results CI 1.2 Year 1999 2000 2001 2002 2003 1998 1999 2000 2001 2002 2003 No HCOs 49 61 78 78 62 43 46 54 69 72 56 Numerator 183 199 236 210 232 38 44 42 86 69 59 Denominator 854 953 1,583 1,428 1,440 753 767 797 1,332 1,284 1,322 Rate % 21.4 20.9 14.9 14.7 16.1 5.05 5.74 5.27 6.46 5.37 4.46 Rate (20)% 18.4 15.7 13.9 11.4 11.7 3.44 3.08 2.69 2.69 3.99 2.83 Rate (80)% 26.5 26.0 21.1 19.1 22.0 6.79 7.24 6.37 6.80 7.81 7.50 Centile gains 26 49 16 47 64 12 20 20 50 17 21 Stratum gains 13 47 11 8 43 21 Outlier gains 11 4 4 2 15 1 2
1.3
CI. 1.2 The rate of children with a pre-operative diagnosis of acute appendicitis, who undergo appendicectomy having normal histology. The rates have decreased from 21% in 1998 to 16% in 2003 (Figure 1.2). The rates for both the better and the poorer performing HCOs have decreased. There is considerable variation between HCOs; the rates of the poorer performing HCOs were almost twice the rates of the better performing HCOs. The lower rate of 12% suggests that further improvement of approximately 4% could be obtained. Western Australia had the highest rate in 2003 of 22% (Table 1.2). Figure 1.2 Paediatrics pre-operative diagnosis of acute appendicitis
30
25
Rate %
20
15
10
Surgical Indicators Version 3 Table 1.2 Paediatrics pre-operative diagnosis of acute appendicitis No HCOs 19 17 6 9 6 5 Stratum rate % 15.5 17.3 16.5 13.4 22.3 12.3 Standard error 1.36 1.69 2.64 1.70 1.84 1.90 Stratum gains 12 12 21
Numerator 58 44 18 29 59 24
CI. 1.3 The rate of children with a pre-operative diagnosis of acute appendicitis who undergo appendicectomy with normal histology, but have significant other intra abdominal pathology. The average rates have varied around 5% and the rates of the better performing HCOs have remained below 4%. Queensland has a higher rate of 7.1% (Table 1.3) Figure 1.3 Paediatrics appendicectomy normal histology but pathology
8 7 6 5 4 3 2 1 0 1998 1999 2000 2001 Year 2002 2003
Table 1.3 Paediatrics appendicectomy normal histology but pathology No HCOs 16 16 6 8 5 5 Stratum rate % 4.76 7.09 6.20 2.99 3.84 2.48 Standard error 0.58 0.66 0.99 0.64 0.69 0.72 Stratum gains 7 11 3
Rate %
Numerator 18 22 7 3 7 2
2.2
2.3
CI 2.4
Year 1998 1999 2000 2001 2002 2003 1998 1999 2000 2001 2002 2003
Numerator 172 135 180 220 160 182 172 187 206 275 198 232
Denominator 3,787 3,450 4,172 5,596 5,241 5,000 3,929 3,908 4,539 5,452 4,917 4,897
Rate % 4.54 3.91 4.31 3.93 3.05 3.64 4.38 4.79 4.54 5.04 4.03 4.74
Rate (20)% 3.00 2.67 2.63 2.97 2.47 2.94 3.51 4.46 3.57 4.04 3.66 3.65
Rate (80)% 5.78 5.22 5.44 4.64 3.70 4.15 4.99 5.46 5.31 6.35 4.42 6.02
Centile gains 58 42 70 54 30 34 33 12 43 54 17 53
Stratum gains 47 40 16
Outlier gains 3 2 10 2
2.5
CI. 2.1 The average operating time (minutes) in patients undergoing TUR for benign prostatomegaly. The average and median operating time for TUR have remained close to 49 minutes since 1998. The variation between HCOs has not changed with 60% of HCOs patients having average operation times between 40 and 60 minutes.
Surgical Indicators Version 3 CI. 2.2 The average length of stay (days) following TUR for benign prostatomegaly. The average length of stay has declined from 3.8 days in 1998 to 3.3 days in 2003. One in five HCOs had average length of stay less than 2.2 days in 2003 whereas in 1998 one in five HCOs had average length of stay less than 2.8 days. Four HCOs reported average length of stay in excess of 5 days for a total of 112 patients. CI. 2.3 The average weight of tissue (grams) removed from patients undergoing TUR for benign prostatomegaly. The average weight of tissue removed has increased from 19 grams in 1998 to 20.5 grams in 2003. In 1998 four in five HCOs had average weights greater than 13 grams whereas in 2003 four in five HCOs had weights greater than 16 grams. CI. 2.4 The rate of patients having a blood transfusion following TUR for benign prostatomegaly, (intra-operatively or post-operatively). The rates have decreased from 4.5% in 1998 to 3.6% in 2003 (Figure 2.4). The higher rates have decreased from 5.8% to 4.2%. Since 1998 40% of HCOs, reported no transfusions. There are small potential gains. There were 8,144 public and 12,079 private TUR prostatectomy procedures in 2003. If the average rate is applied to these procedures then the estimated number blood transfusions annually is 736. If the overall rate could be reduced to the better rate the number would decrease by 141. Figure 2.4 TUR for benign prostatomegaly blood transfusion
8 7 6
Rate %
CI. 2.5 The rate of patients having an unplanned readmission within 28 days of discharge following TUR for benign prostatomegaly. The rates have remained between 4% and 5% since 1998. During that period more than one third of HCOs reported no readmissions within 28 days. The lower rates have remained at or below 4% since 2000.
CI. 3.1 The rate of a post-operative in-hospital infection in primary THJ replacement. The rate in 1998 was 2.6% and has remained around 1.4% since then (Figure 3.1). The rates for the better performing HCOs have remained at round 1%, half the rate of the poorer performing HCOs. This variation between HCOs indicates that a reduction in the proportion of infections may be achievable. If the above rate is applied to the 30,496 separations (including 3,267 revisions) in Australia for hip prosthesis procedures, then the estimated proportion of infections in Australia is 430 per year. Figure 3.1 THJR post-operative in-hospital infection
4 .0 3 .5 3 .0
Rate %
CI. 4.1 The rate of completely excised malignant skin tumours. The rates have remained close to 90% (Figure 4.1). The rates for the poorer performing HCOs have remained about 10% lower than the rates of the better performing HCOs. In 2003 there were 9 HCOs with statistically significantly low rates, all of which were non-metropolitan. This may reflect differences in surgical practice or severity of patient skin tumour; however rural rates have been less than five percentage points below the metropolitan rates since 2000. One non-metropolitan HCO had statistically significantly high rates in both 2002 and 2003 and should review its procedures. Figure 4.1 Plastic surgery completely excised malignant skin tumours
100 90 80 70
Rate %
5.2
5.3
CI. 5.1 The rate of patients who die in the same admission as having CAGS. The rates have remained at around 2% and the rates of the better performing HCOs at around 1.5%. The extent to which differences in case-mix have contributed to the differences between HCOs cannot be determined. In 2003 there was an outlier HCO accounting for nine deaths. For most years there have been no outlier HCOs. Figure 5.1 CAGS death
3 .0
2 .5
2 .0
Rate %
1 .5
1 .0
0 .5
Surgical Indicators Version 3 CI. 5.2 The rate of elective patients who die in the same admission as having CAGS. Since 1998 the rates have remained around 1.5% and the rates of the better performing HCOs have remained around 1.3%. The rates of the poorer performing HCOs has been more variable. There is a small difference between the pubic and private rates for this indicator (Table 5.2). Table 5.2 CAGS death (elective) No HCOs 12 15 Stratum rate % 1.44 1.84 Standard error 0.11 0.12 Stratum gains 9
Numerator 35 47
CI. 5.3 The rate of patients aged 71 years or greater who die in the same admission as CAGS. The rates have changed little since 1998 and remain at about 3.4% and were higher than the rates for all CAGS. In 2003 there was considerable variation between HCOs, the rates for the poorer performing HCOs being twice the rates of the better performing HCOs. One in five HCOs has a rate exceeding 5%.
Area 6 Neurosurgery
Neurosurgical infection Rationale The occurrence of infection following neurosurgery may have undesirable effects. Neurosurgical infection is defined as a wound infection, meningitis or shunt infection. For the purpose of these indicators (CI No: 6.1 6.2) only intracranial and spinal operations should be included as neurosurgery procedures. The indicator is: C.I. 6.1 The rate of patients having a neurosurgical infection in hospital excluding superficial wound infections Results CI 6.1 Year 2001 2002 2003 No HCOs 13 16 19 Numerator 39 56 78 Denominator 3,108 3,495 5,651 Rate % 1.25 1.60 1.38 Rate (20)% 0.97 0.96 0.57 Rate (80)% 2.24 2.36 2.31 Centile gains 8 22 46 Stratum gains 49 Outlier gains 5 12
CI. 6.1 The rate of patients having a neurosurgical infection in hospital excluding superficial wound infections. The rate of infection was 1.38% in 2003, little different from the previous two years. There was considerable variation between HCOs, which suggests that the proportion of infections could be reduced. The Queensland rate was the highest in 2003 (2.4%) and the public rate as statistically significantly higher than the private rate (Table 6.1). Table 6.1 Neurosurgery neurosurgical infection No HCOs 9 5 5 14 5 Stratum rate % 1.53 2.42 0.73 1.63 0.51 Standard error 0.31 0.56 0.37 0.25 0.48 Stratum gains 22 14 49
Numerator 43 22 13 75 3
Surgical Indicators Version 3 Neurological deficit following a neurological procedure Rationale Neurological deficit after surgery may imply a less than optimal technique. The indicator is: C.I. 6.2 The rate of patients with a new neurological deficit following a neurosurgery procedure Results CI 6.2 Year 1998 1999 2000 2001 2002 2003 No HCOs 14 15 11 8 13 15 Numerator 189 99 97 34 66 92 Denominator 3,985 4,941 4,916 2,841 3,621 5,545 Rate % 4.74 2.00 1.97 1.20 1.82 1.66 Rate (20)% 1.06 0.86 0.86 0.48 1.30 1.05 Rate (80)% 9.22 2.29 2.28 2.41 2.26 2.39 Centile gains 146 56 54 20 18 33 Stratum gains 26 Outlier gains 95 23 21 6 7
CI. 6.2 The rate of patients with a new neurological deficit following a neurosurgery procedure. The rates have decreased from 4.7% in 1998 and appears to have stabilised to about 1.6%. The NSW rate was the highest in 2003. A single HCO had a significantly high rate of 4.2%. Figure 6.2 Neurosurgery neurological deficit following procedure
10
Rate %
Table 6.2 Neurosurgery neurological deficit following procedure No HCOs 8 7 Stratum rate % 1.92 1.18 Standard error 0.20 0.27 Stratum gains 26
Numerator 74 18
CI. 7.1 The rate of patients having bile duct injury requiring operative intervention during laparoscopic cholecystectomy. The rates have decreased from 0.5% in 1998 to 0.3% in 2003 (Figure 7.1). The proportion of HCOs having no cases of bile duct injury requiring operative intervention has increased from 70% in 1998 to 82% in 2003. In the last three years the difference between the better and the poorer performing HCOs has been reduced. CI. 7.1 Laparoscopic cholecystectomy bile duct injury requiring operative intervention.
0 .8
0 .6
Rate %
0 .4
0 .2
CI. 8.1 The rate of death following elective AAA repair (within the same admission). The rates were highest in 2003, although there was no evidence of any differences in outcomes between HCOs. Since 1999 more than two thirds of HCOs had no deaths and the differences between the better performing and the poorer performing HCOs is negligible.
Carotid endarterectomy
Rationale Carotid endarterectomy is undertaken as a prophylactic measure to reduce the occurrence of stroke. A minimal rate of post-operative stroke is desirable. The indicator is: C.I. 8.2 The rate of stoke following carotid endarterectomy (within the same admission) Results CI 8.2 Year 1998 1999 2000 2001 2002 2003 No HCOs 39 43 46 65 68 52 Numerator 28 17 18 30 25 18 Denominator 1,202 1,319 1,112 1,759 1,661 1,309 Rate % 2.33 1.29 1.62 1.71 1.51 1.38 Rate (20)% 2.33 1.05 1.62 1.45 1.51 0.93 Rate (80)% 2.33 1.23 1.62 2.05 1.51 1.40 Centile gains 3 4 5 Stratum gains 2 Outlier gains
CI. 8.2 The rate of stoke following carotid endarterectomy (within the same admission). The rates have declined from 2.3% in 1998 and then remained around 1.5%. Since 1999 more than two thirds of HCOs had no strokes following carotic endarterectomy. The difference between the better and poorer performing HCOs remains small.
CI. 9.1 The rate of significant reactionary haemorrhage following tonsillectomy. The rate was 1.3% in 1998 but has been around 0.7% for the subsequent years. Since 1998, two thirds of HCOs had no cases of reactionary haemorrhage. In 2003, the difference between the better and poorer performing HCOs was small. The proportion of tonsillectomy / adenectomies in Australian was 61,000, and applying the above rate to these suggests that they were about 420 cases of significant reactionary haemorrhage.
8 9
Determining to Improve the Quality of Care in Australian Health Care Organisations: Results from the ACHS Clinical Indicator Data 1998 and 1999. The Australian Council of Health Care Standards (ACHS). Volume 4: Measurement of Health Care in Australian Hospitals, 1997.
Appendix Statistical Methods Finally, the observed numerator is compared to the expected numerator, based on the shrunken estimates for each HCO. The expected numerator is the number of events that would have occurred if the HCO has the same rate as the overall rate. The observed minus expected provides a measure of how different each HCO is from the overall mean. A value of zero implies that the HCO has the same mean as the mean of all the HCOs. If the observed minus the expected is more than three standard deviations from zero, then there is evidence of a systematic difference between this HCO and the other HCOs. If this difference is clinically important, a quality improvement team should determine the possible causes. If the causes can be removed, we can quantify the gains thereby obtained as the sum of the observed minus expected above the three standard deviations (assuming that a low rate represents better quality). The outlier observations are summed across all outlier HCOs. The outlier gains are calculated as Outlier gains = Sum of (observed expected) for outlier HCOs Effect of shrinkage The distribution of results for indicator CI 2.1, AMI thrombolysis within one hour are shown in the Figure 2.1a below. The crude rates for the 184 reports range from 0% to 100%, and the shrunken rates range from 39% to 96%. This indicator reveals the problem associated with reporting the actual HCO rates they are based on small numbers. The HCOs with rates of 0% and 100% had denominators that were less than or equal to 7 and 41 respectively. The shrinkage estimators adjust the HCOs observed rate using the HCOs numerator and denominator together with the summary results from the other HCOs, (the mean and standard deviation), and provide a more precise estimate of the individual HCOs true rates. Figure 2.1a. Distribution of rates for CI 2.1 AMI thrombolysis within one hour Crude Rates (%) Shrunken Rates (%)
50
60
40 30
40
20 10
20
10
20
30
40
50
60
70
80
90 100
10
20
30
40
50
60
70
80
90 100