HEALTH INFORMATION SYSTEM FOR MEDICAL LABORATORY SCIENCE

Lesson 8: HMIS Data Quality

Data Quality Advantages of LQAS for Local Program Management

→ The overall utility of a dataset(s) as a function of its • Can be used at a local level as it requires only modest
ability to be processes easily and analyzed for database, supervision.
data warehouse, or data analytics system. • Identifies where the successes and challenges are located
→ Signifies the data’s appropriateness to serve its purpose using rigorous sampling theory.
in each context. • Introduces information that can be rapidly interpreted by
→ High quality data means that the data is: local managers.
o Useful • Paper/pencil analyses rather than requiring computer
o Consistent analyses.
→ Poor data quality generates costs, affects customer • Data can be used for local management as well as for
satisfaction, company reputation, and even strategic national reporting.
decisions of the management.
Routine Data Quality Assessment (RDQA)
Aspects of data quality
→ A simplified version of the Data Quality Audit (DQA) tool
1) Accuracy 2) Reliability which allows programs and projects to verify and assess
3) Competence 4) Presentability the quality of their reported data.
5) Relevance 6) Accessibility
7) Consistency
Steps in applying RDQA

Lot Quality Assurance Sampling (LQAS) 1. Determine the purpose.

2. select levels and sites to be included.
→ A tool that allows the use of small random samples to 3. Identify indicators, data sources, and reporting
distinguish between different groups of data elements period.
(or lots) with high and low data quality. 4. conduct site visits.
˃ Most frequently used size = <20 per one admin. 5. review outputs and findings
- it provides the most acceptable level of error 6. develop a system strengthening plan, including
→ Widely applied in the health care industry for decades follow up actions.
and has been used for quality assurance of products.
→ Adopted in the context of District Health Information
System (DHIS) data quality assurance. RDQA Objectives
→ Designating health facilities, monthly reports, sections 1) the quality of reported data for key
of monthly reports, sections of monthly reports, and 1. Verify Rapidly indicators at selected sites; and
group of data elements as “lots” to provide 2) the ability of data‐ management
representative samples. systems to collect, manage and
report quality data.
Steps in applying LQAS

1. Define the service to be assessed (e.g... DQA of DHIS) corrective measures with action plans
2. Implement for strengthening the data
2. Identify the unit of interest (e.g., a supervisory area,
management and reporting system
facility, hospital, a district)
and improving data quality.
3. Define the higher and lower thresholds of
performance based on prior information about the Capacity improvements and
expected performance of the region of interest. 3. Monitor performance of the data management
4. Determine the level of acceptable error. (0%) and reporting system to produce
5. Determine the sample size and decision rule for quality data.
acceptable errors to declare an area as performing
"below expectations."
6. Identify the number of errors observed (mismatched
data elements will be reliably determined if the
facility is performing above or below expectations).
 Uses of RDQA Tool Key steps of implementation plan

• The RQDA is a multipurpose tool that is most effective Define “What do you want to
when routinely used. Following are the uses for the goals/objectives accomplish.”
RDQA tool:
Schedule milestones Outline the deadlines and
RDQA Use Care Example
timelines.
Routine data quality Routine data quality checks
checks as part of on-going can be included in already
Allocate resources Determine whether you have
supervision planned supervision visits at
sufficient resources and decide
the service delivery sites
how you will procure those
Initial and follow-up Repeated assessments (e.g.,
missing.
assessments of data biannually, or annually) of a
management and system’s ability to collect and Designate team Create a general team plan with
reporting systems report quality data at all member overall roles that each team
levels can be used to identify responsibilities member will play
gaps and monitor necessary
improvements
Strengthening of the Monitoring and evaluation Define metrics for measurement for success
program staff’s capacity in (M&E) staff can be trained on success
data management and the RDQA and sensitized to
reporting the need to strengthen the
key functional areas linked to Data Quality Tools
data management and
reporting in order to produce → Analyzes information and identifies incomplete or
quality data. incorrect data.
Preparation for a formal The RDQA tool can help → The maintaining the process enhances the reliability and
data quality audit identify data quality issues integrity of the info being used by the organization.
and areas of weaknesses in
the data management and
reporting system that would Data Cleansing
need to be strengthened to → the process in which could range from removing
increase readiness for a abnormalities to merge repeated information.
formal data quality audit.
→ Enhances the reliability of the information.
External assessment by Such use of the RDQA for
partners of the quality of external assessments could
data be more frequent, more Addressing Problems in Data Quality
streamlined, and less
resource intensive than Decomposition/breakdown of fields
comprehensive data quality Parsing and into component parts and formatting
audits that use the DQA standardization the values into consistent layouts
version for auditing. based on industry standards and
patterns and user- defined business
Development Implementation Plan rules.

→ Tool that illustrates how a project is expected to Modification of data values to meet
Generalized domain restrictions, constraints on
progress at a high level.
“Cleansing” integrity, or other rules that define
→ Ensure that the development team is working to deliver
data quality.
and complete task on time.
→ Ensuring the efficient flow of communication between Identification and merging of related
those who are involved in the project to minimize issues Matching entries.
that would delay delivery of the project.
→ Validates the estimation and schedule of the project Capture statistics or metadata to
plan. Profiling determine the data quality and
identify data quality issues.
 Deployment of controls to ensure → Lays down the potential causes in a bar graph and tracks
Monitoring conformity of data to business rules. the collective percentage in a line graph to the top of the
table.
Enhancement of the values of the → Uses the Pareto principle (20% of the work creates 80%
Enrichment data by using related attributes from of the results)
external sources.
USED WHEN: there are multiple potential causes to a
problem.
Application and Scope of Data Quality Tools

• First generation of data quality tools – was characterized

by dedicated data cleansing tools designed to address
normalization and reduplication.
• Last 10 years – It was observed that there is a
generalization of:

Extract, Transform, Load (ETL) tools

→ Which allow the optimization of the alimentation
process. Figure 1. Pareto Exercise Chart Food Serving Survey
• Recently – these tools started to focus on: (3) Fault Tree Analysis (FTA)
Data Quality Management (DQM) → It is a top-down, deductive analysis which visually depicts
→ Which generally integrates, profiling, parsing, a failure path or failure chain.
standardization, cleansing, and matching processes. → Uses Boolean logic to determine the root cause of an
undesirable event.
Root Cause Analysis → Undesired results at the top and potential causes are
→ A systematic problem-solving method that identifies the listed down which forms an upside-down tree.
root causes of problems or events instead of simply USED IN: risk and safety analysis.
addressing the obvious symptoms
(4) Current Reality Tree (CRT)
Goal: to improve the quality of products and services by
→ It is a down-top, inductive analysis to determine the
using systematic ways to address problems to effective.
effects/results of the problem.
→ Problems are listed down followed by the potential
Techniques in Root Cause Analysis cause/s of the problem and the results/effects are listed
at the top.
(1) Failure Mode and Effects Analysis (FMEA) ˃ By doing so, a cause common to all problems
→ Aims to find various of failure within a system and will appear.
addresses the following questions:
USED WHEN: The root causes of multiple problems need to
a. How many times does a cause of failure occur?
be analyzed all at once.
b. What actions are implemented to prevent this
cause from occurring again?
c. Are these actions effective and efficient?

Failure mode – ways (/modes) in which something might fail.

Effects analysis – studying the consequence of those failures.

USED WHEN: There is a new product/process or when there

are changes/updates in a product when a problem is reported
through customer feedback (but not necessarily the
customers in other cases)

(2) Pareto Analysis

→ statistical technique in decision-making used for the
selection of a limited number of tasks that produce Figure 2. Current Reality Tree Chart of Overcrowded Bus
significant overall effect.
(5) Fishbone Diagram (Ishikawa or cause-and-effect Sustaining a Culture of Information Use
diagram)
• Information culture affects outcomes of information use.
→ Looks like a fishbone as it shows the categorized causes
and sub-causes of a problem. Information culture
→ is determined by the ff variables:
USED WHEN: in identifying and grouping possible causes into
o mission
categories.
o history
6 Major categories of causes o leadership
Methods People o employee traits
o industry
Materials Environment
o national culture
Equipment Measurement → It can also be shaped by cognitive and epistemic
expectations which are influenced by the way tasks
are performed and decisions are made.

• In order to have sense of information attitudes and

values, managers should consider taking the
pulse/rhythm of information of their own organizations.

• Management play an important role in sustaining a

culture of information and should continuously work on
maintaining and improving the quality of data
information used in daily operation.
Figure 3. Fishbone diagram presenting the main causes for
patients' dissatisfaction.

(6) Kepner-Tregoe Technique

→ Breaks down a problem to its root cause by assessing a
situation using priorities and orders of concern for
specific issues.
→ The idea is not to find a perfect solution but rather the
best possible choice, based on achieving the outcome
with minimal negative consequences.

(7) Rapid Problem Resolution (RPR)

→ Deals with failures, incorrect output, and performance
issues.

3 RPR Phases

Discover Data gathering and analysis of the

findings

Investigate Creation of a diagnostic plan and

identification of the root cause through
careful analysis of the diagnostic data

Fix fixing the problem and monitoring to

confirm and validate that the correct root
cause was identified

USED WHEN: diagnosing of ongoing and recurring problems.