BMC Health Services Research

BioMed Central

Open Access

Research article

Preventable hospitalization and access to primary health care in an

area of Southern Italy
Paolo Rizza1, Aida Bianco1, Maria Pavia*1 and Italo F Angelillo2
Address: 1Chair of Hygiene, Medical School, University of Catanzaro "Magna Grcia", Catanzaro, Italy and 2Department of Public, Clinical and
Preventive Medicine, Second University of Naples, Naples, Italy
Email: Paolo Rizza - [email protected]; Aida Bianco - [email protected]; Maria Pavia* - [email protected];
Italo F Angelillo - [email protected]
* Corresponding author

Published: 30 August 2007

BMC Health Services Research 2007, 7:134

doi:10.1186/1472-6963-7-134

Received: 20 November 2006

Accepted: 30 August 2007

This article is available from: http://www.biomedcentral.com/1472-6963/7/134

2007 Rizza et al; licensee BioMed Central Ltd.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract
Background: Ambulatory care-sensitive conditions (ACSC), such as hypertension, diabetes,
chronic heart failure, chronic obstructive pulmonary disease and asthma, are conditions that can
be managed with timely and effective outpatient care reducing the need of hospitalization.
Avoidable hospitalizations for ACSC have been used to assess access, quality and performance of
the primary care delivery system. The aims of this study were to quantify the proportion of
avoidable hospital admissions for ACSCs, to identify the related patient's socio-demographic profile
and health conditions, to assess the relationship between the primary care access characteristics
and preventable hospitalizations, and the usefulness of avoidable hospitalizations for ACSCs to
monitor the effectiveness of primary health care.
Methods: A random sample of 520 medical records of patients admitted to medical wards
(Cardiology, Internal Medicine, Pneumology, Geriatrics) of a non-teaching acute care 717-bed
hospital located in Catanzaro (Italy) were reviewed.
Results: A total of 31.5% of the hospitalizations in the sample were judged to be preventable. Of
these, 40% were for congestive heart failure, 23.2% for chronic obstructive pulmonary disease,
13.5% for angina without procedure, 8.4% for hypertension, and 7.1% for bacterial pneumonia.
Preventable hospitalizations were significantly associated to age and sex since they were higher in
older patients and in males. The proportion of patients who had a preventable hospitalization
significantly increased with regard to the number of hospital admissions in the previous year and to
the number of patients for each primary care physician (PCP), with lower number of PCP accesses
and PCP medical visits in the previous year, with less satisfaction about PCP health services, and,
finally, with worse self-reported health status and shorter length of hospital stay.
Conclusion: The findings from this study add to the evidence and the urgency of developing and
implementing effective interventions to improve delivery of health care at the community level and
provided support to the usefulness of avoidable hospitalizations for ACSCs to monitor this
process.

Page 1 of 8
(page number not for citation purposes)

BMC Health Services Research 2007, 7:134

http://www.biomedcentral.com/1472-6963/7/134

Background

Methods

Ambulatory care-sensitive conditions (ACSC), such as

hypertension, diabetes, chronic heart failure, chronic
obstructive pulmonary disease and asthma, are conditions that can be managed with timely and effective outpatient care reducing the need of hospitalization. Rates of
avoidable hospitalizations, also called preventable hospitalizations, for ACSC have been used to assess access,
quality and performance of the primary care delivery system. This includes primary prevention, early detection
and monitoring of acute episodes and follow-up and
monitoring of chronic conditions.

Study population
This cross-sectional study was conducted from April to
July 2005 by reviewing a random sample of 520 medical
records of patients aged 18 and over admitted to medical
wards (Cardiology, Internal Medicine, Pneumology, Geriatrics) of a non-teaching acute care 717-bed hospital
located in Catanzaro (Italy).

The Agency for Healthcare Research and Quality (AHRQ)

developed the Preventive Quality Indicators (PQIs), a set
of measures to identify ACSC as rates of admission to the
hospital based on the assumption that high hospitalization rates for ACSC may result from poor access to primary care and can be prevented [1]. ACSC have been
evaluated in many countries most in the US but also in
Australia, Canada and European countries [2-5] for determining the socioeconomic and medical conditions associated with hospitalized ACSC [6-12]; however, little is
known about role of access to primary care, measured as
number of accesses, number of patients per primary care
physician (PCP), health care seeking and physician practice style [13-17].
In Italy, where universal and free access to primary health
care is provided through the National Health Service,
these indicators can serve as a convenient and effective
evaluation tool to assess effectiveness of and barriers to
access to primary care. Community health care is provided by health districts, geographical units responsible
for coordinating and providing primary care, pharmaceuticals, home care, specialist and residential and semi-residential care. Primary care is provided by physicians and
pediatricians working under government contract, who
are paid a capitation fee based on the number of patients
(adults or children) on their list with a maximum number
of patients allowed (15001800 for general practitioners
and 8001000 for pediatricians). The PCPs act as "gatekeepers" for access to secondary services and this role has
a great impact both on quality of outcomes and cost of
healthcare, but for several reasons patients prefer to attend
specialists or hospitals [18].
The aims of this study were to quantify the proportion of
avoidable hospital admissions for ACSCs, to identify the
related patient's socio-demographic profile and health
conditions, to assess whether barriers to access to primary
care were related to preventable hospitalizations in Italy
and to assess the usefulness of avoidable hospitalizations
for ACSCs to monitor the effectiveness of primary health
care.

Two physicians, who were not involved in care and who

had been previously trained, collected the data by reviewing charts and by interviewing at bedside all the patients
who agreed to participate, independently. Actual data collection and extraction did not start until the performance
of the reviewers had been judged satisfactory and showed
satisfactory inter-rater reliability. The analysis on medical
records was based on data reviewed by both physicians in
which discrepancies were solved through discussion
among reviewers.
A total of 492 patients agreed to participate and were
enrolled for a response rate of 94.6%, thus minimizing
the role of non-respondent bias.
Review instrument
The following data were collected for each patient by
reviewing charts: socio-demographics (age, gender, marital status, education level), distance of patient's home
from hospital, Charlson et al. comorbidity index [19],
ward and type of admission, and who referred the patient
to the hospital. The following data were collected by interviewing at bedside all the patients who agreed to participate: socio-demographics (working activity, number of
persons in the household), the self-reported health status,
self-reported utilization of health services during the previous year (propensity to seek care, number and main reasons of visit and for not having been to a PCP medical
visit, difficulty to the access to and satisfaction with PCP
health services, number and main reasons for specialist
visit, emergency access, hospital admission), and the
name of their own PCP. The questionnaire focused particularly on frequency of utilization of health services during
the previous year with most questions in "yes/no" format.
If the answer was "yes" then the participants were asked
the number of accesses. The questions on satisfaction and
difficulty with PCP health services were scored on a fourpoint Likert scale with options for no, few, rather, and
much. The questionnaire was pretested on a sample to
ensure clarity of interpretation to improve the validity of
responses. The accuracy of patient self-report of health
services utilization is of paramount concern, however
such data is often used to estimate health care utilization
and have demonstrated to be a reasonably accurate
method to obtain information on most types of medical
utilization in the general population [20-22].

Page 2 of 8
(page number not for citation purposes)

BMC Health Services Research 2007, 7:134

We selected 9 out of 16 PQIs from AHRQ, defined by

diagnosis and procedure codes of International Classification of Diseases (ICD-9-CM) [23], to identify the preventable hospitalizations for ACS conditions: diabetes shortterm and long-term complication, uncontrolled diabetes,
chronic obstructive pulmonary disease, hypertension,
congestive heart failure, bacterial pneumonia, angina
without procedure and adult asthma. We excluded pediatric indicators (pediatric asthma, pediatric gastroenteritis,
low birth weight) and indicators for adult conditions that
were rarely encountered, in our setting, as discharge diagnosis, as measured by reviewing discharge records in the
previous two years (perforated appendix, dehydration,
urinary tract infection, lower-extremity amputation
among patients with diabetes). Assessment of avoidable
hospitalizations was made by using the discharge data
according to the Guide to Prevention Quality Indicators
from AHRQ [1].
The Ethics Committee of the "Mater Domini" Hospital of
Catanzaro (Italy) approved the protocol of the study
(Prot. E.C. n16/2005).
Statistical analysis
A multivariate logistic regression analysis was performed
to identify baseline characteristics independently associated with the following outcomes of interest: preventable
hospitalization for all investigated conditions, preventable hospitalization for cardiovascular diseases (hypertension, congestive heart failure, angina without procedure),
preventable hospitalization for respiratory diseases
(chronic obstructive pulmonary disease, bacterial pneumonia, adult asthma), preventable hospitalization for
diabetes (diabetes short-term complication, diabetes
long-term complication, uncontrolled diabetes). Model
building strategy and particularly ways to include independent variables in the model (continuous, ordinal or
categorical) took into account how each of these ways better fitted the data at the univariate analysis and we chose
that way in the multivariate analysis. Moreover, as regards
to the cut-off points for less or more than 12 accesses or
visits per year, our idea was that in many chronic patients,
one access per month is generally useful for monitoring
chronic conditions. In all models the explanatory variables included were the following: patient's age (continuous), patient's sex (male = 0, female = 1), distance in
kilometers between patient's home and hospital (continuous), educational level (no formal education = 0, primary school = 1, secondary school or higher = 2), living
condition (with family = 1, other = 2), additional persons
in the household (none = 0, 1 = 1, > 1 = 2), working activity (retired = 0, other = 1), type of admission (emergency
physician = 1, other = 2), length of hospital stay (continuous), age-adjusted Charlson et al. comorbidity index
(continuous), self-reported health status on a 10 points
scale ( 4 = 0, 5 = 1), number of PCP accesses in the pre-

http://www.biomedcentral.com/1472-6963/7/134

vious year ( 12 = 1, > 12 = 2), number of PCP medical

visits in the previous year ( 12 = 1, > 12 = 2), satisfaction
with PCP health services (no/few = 0, rather/much = 1),
number of patients for each PCP (< 1000 = 1, 10001300
= 2, > 1300 = 3), number of specialist visits in community
health services (none = 0, 1 = 1), number of emergency
accesses in the previous year (none = 0, 1 = 1), and
number of hospital admissions in the previous year (none
= 0, 1 = 1). The significance level for variables entering
the logistic regression models was set at 0.2 and for
removing from the model at 0.4. Adjusted odds ratio
(ORs) and 95% confidence intervals (CIs) were calculated. The data were analyzed using the Stata software program [24].

Results
The main characteristics of the study population are presented in Table 1. Fifty-two per cent were females, the
median age was 75 years (range 2395), and three quarters lived with their family, more than half were in general
medical wards (68.3%), the median length of stay was 9
days (range 183) and the median of Charlson et al.
comorbidity index was 4 (range 014). Almost all
(99.2%) had at least one PCP access in the previous year,
more than 40% had at least one hospital admission in the
previous year, and at least one district health services
access. More than 70% were satisfied with PCP health
services, and the main reasons for dissatisfaction were
long waiting times for access (18.2%), opening hours
(11.7%), and trust in hospital physicians (7.4%). Overall,
17.2% reported difficulty of access to PCP health services.
In the study period, a total of 31.5% of the hospitalizations in the sample were judged to be preventable. Of
these, 40% were for congestive heart failure, 23.2% for
chronic obstructive pulmonary disease, 13.5% for angina
without procedure, 8.4% for hypertension, 7.1% for bacterial pneumonia, 3.2% for diabetes short-term complication, 2.6% for adult asthma, 1.3% for diabetes long-term
complication and 0.6% for uncontrolled diabetes.
Patients admitted for a preventable hospitalization were
more frequently males (58.1% vs. 43.3%), older (median
age 76 vs. 74), with a higher unsatisfactory self-reported
health status (56.1% vs. 36.8%), reported more frequently less than 13 PCP medical visits (93.5% vs. 47.9%)
and less than 13 PCP accesses (41.6% vs. 19.5%) in the
previous year, attended a PCP with a higher number of
patients (65.6% vs. 31.1%), were more frequently unsatisfied by PCP health services (54% vs. 15.6%), and were
more likely to have had at least one emergency access
(52.6% vs. 45.2%) and hospital admission in the previous
year (54% vs. 40.1%).
Preventable hospitalizations were significantly associated
with age and sex since they were higher in older patients
Page 3 of 8
(page number not for citation purposes)

BMC Health Services Research 2007, 7:134

http://www.biomedcentral.com/1472-6963/7/134

Table 1: Selected characteristics of the study population

Characteristic

N*

236
256

48
52

106
139
196
51
75

21.5
28.3
39.8
10.4

177
186
129

36
37.8
26.2

286
206

58.1
41.9

368
124

74.8
25.2

110
205
177

22.4
41.6
36

411
81

83.5
16.5

227
153
112
20

46.1
31.1
22.8

432
60

87.8
12.2

Sex
Male
Female
Age group, years
< 65
6574
7584
85
Median
Education level
No formal education
Primary school
Secondary school or higher
Marital status
Married
Others
Living condition
With family
Other
Additional persons in the household
None
1
>1
Working activity
Retired
Other
Distance from home to hospital, km
5
635
> 35
Median
Type of admission
Emergency physician
Other
Length of hospital stay, days
Median
Age-adjusted Charlson et al. comorbidity index
Median
Self-reported health status on a 10 points scale
4
5
PCP accesses in the previous year
None
112
> 12
PCP medical visits in the previous year
12
> 12
Satisfaction with PCP health services
No/few
Rather/much
Difficulty of access to PCP health services
No/few
Rather/much
Patients for each PCP
< 1000
10001300
> 1300
District health services accesses in the previous year
None
1
Emergency accesses in the previous year
None
1
Hospital admissions in the previous year
None
1

(OR = 1.03, 95% CI = 1.011.05, p = 0.027) and in males

(OR = 0.52, 95% CI = 0.310.87, p = 0.013). The proportion of patients who had a preventable hospitalization
significantly increased with regard to the number of hospital admissions in the previous year (OR = 1.76, 95% CI
= 1.062.93, p = 0.03) and to the number of patients for
each PCP (OR = 2.25, 95% CI = 1.623.13, p < 0.001),
with lower number of PCP accesses (OR = 0.52, 95% CI =
0.30.93, p = 0.027) and PCP medical visits in the previous year (OR = 0.1, 95% CI = 0.050.23, p < 0.001), with
less satisfaction about PCP health services (OR = 0.34,
95% CI = 0.20.58, p < 0.001), and, finally, with worse
self-reported health status (OR = 0.53, 95% CI = 0.31
0.89, p = 0.017) and shorter length of hospital stay (OR =
0.95, 95% CI = 0.910.99, p = 0.011) (Model 1 in Table
2).
Preventable hospitalizations for cardiovascular diseases
were significantly more common for higher number of
patients for each PCP (OR = 2.2, 95% CI = 1.53.22, p <
0.001) and for lower number of PCP accesses (OR = 0.52,
95% CI = 0.270.98, p = 0.044) and PCP medical visits
(OR = 0.12, 95% CI = 0.050.3, p < 0.001) in the previous
year. Moreover patients who had a preventable hospitalization for cardiovascular diseases were significantly more
likely to be less satisfied for PCP health services (OR =
0.31, 95% CI = 0.170.57, p < 0.001), retired (OR = 0.37,
95% CI = 0.130.97, p = 0.044), with worse self-reported
health status (OR = 0.5, 95% CI = 0.270.93, p = 0.027)
and shorter length of hospital stay (OR = 0.94, 95% CI =
0.90.99, p = 0.013) (Model 2 in Table 2).

4
211
281

42.9
57.1

4
129
359

0.8
26.2
73

304
184

62.3
37.7

135
353

27.7
72.3

404
84

82.8
17.2

146
137
205

29.9
28.1
42

310
178

63.5
36.5

260
232

52.8
47.2

275
217

55.9
44.1

*The numbers that do not add to 492 are due to not applicable data for the
variable.
PCP, primary care physician.

Preventable hospitalizations for respiratory diseases were

significantly associated with higher number of patients for
each PCP (OR = 2.86, 95% CI = 1.664.94, p < 0.001) and
emergency accesses in the previous year (OR = 2.71, 95%
CI = 1.26.11, p = 0.016), with males (OR = 0.27, 95% CI
= 0.110.64, p = 0.003), worse satisfaction for PCP health
services (OR = 0.27, 95% CI = 0.130.58, p = 0.001) and
lower number of PCP accesses (OR = 0.34, 95% CI =
0.140.82, p = 0.017) and PCP medical visits (OR = 0.11,
95% CI = 0.030.4, p = 0.001) in the previous year
(Model 3 in Table 2).
Finally, preventable hospitalizations for diabetes were significantly higher with the number of hospital admissions
in the previous year (OR = 9.74, 95% CI = 1.1483.39, p
= 0.038) and the number of patients for each PCP (OR =
4.75, 95% CI = 1.2218.48, p = 0.025) (Model 4 in Table
2).

Discussion
One of the main findings of our study was that more than
30% of the hospitalizations were considered preventable
according to the PQI indicators. Although most of the

Page 4 of 8
(page number not for citation purposes)

BMC Health Services Research 2007, 7:134

http://www.biomedcentral.com/1472-6963/7/134

Table 2: Logistic regression models results 1,2

Variable

OR

SE

95% CI

Model 1. Outcome: Overall preventable hospitalization

Log-likelihood = -194.19, 2 = 220.14, p < 0.001, Number of observations = 488
Number of PCP medical visits in the previous year
0.1
0.04
Number of patients for each PCP
2.25
0.38
Satisfaction with PCP health services
0.34
0.09
Length of hospital stay
0.95
0.02
Sex
0.52
0.14
Self-reported health status
0.53
0.14
Number of PCP accesses in the previous year
0.52
0.15
Age
1.03
0.01
Number of hospital admissions in the previous year
1.76
0.46
Additional persons in the household
0.75
0.13
Type of admission
0.57
0.24
Distance between patient's home and hospital
0.99
0.01

0.050.23
1.623.13
0.20.58
0.910.99
0.310.87
0.310.89
0.30.93
1.011.05
1.062.93
0.531.06
0.251.3
0.981.01

< 0.001
< 0.001
< 0.001
0.011
0.013
0.017
0.027
0.027
0.03
0.105
0.18
0.399

Model 2. Outcome: Preventable hospitalization for cardiovascular diseases

Log-likelihood = -150.13, 2 = 153.38, p < 0.001, Number of observations = 429
Number of patients for each PCP
2.2
0.43
Satisfaction with PCP health services
0.31
0.09
Number of PCP medical visits in the previous year
0.12
0.06
Length of hospital stay
0.94
0.02
Self-reported health status
0.5
0.16
Number of PCP accesses in the previous year
0.52
0.17
Working activity
0.37
0.18
Sex
0.62
0.18
Number of hospital admissions in the previous year
1.53
0.46
Additional persons in the household
0.76
0.15
Age-adjusted Charlson et al. comorbidity index
1.08
0.07

1.53.22
0.170.57
0.050.3
0.90.99
0.270.93
0.270.98
0.130.97
0.351.1
0.852.77
0.511.13
0.961.22

< 0.001
< 0.001
< 0.001
0.013
0.027
0.044
0.044
0.106
0.156
0.172
0.19

Model 3. Outcome: Preventable hospitalization for respiratory diseases

Log-likelihood = -92.36, 2 = 116.38, p < 0.001, Number of observations = 385
Number of patients for each PCP
2.86
0.8
Satisfaction with PCP health services
0.27
0.11
Number of PCP medical visits in the previous year
0.11
0.07
Sex
0.27
0.12
Number of emergency accesses in the previous year
2.71
1.12
Number of PCP accesses in the previous year
0.34
0.15
Type of admission
0.2
0.17
Living condition
2.35
1.08
Self-reported health status
0.49
0.21
Age
1.03
0.02
Age-adjusted Charlson et al. comorbidity index
0.88
0.08
Length of hospital stay
0.97
0.03

1.664.94
0.130.58
0.030.4
0.110.64
1.26.11
0.140.82
0.041.05
0.955.81
0.211.12
0.991.07
0.731.05
0.921.02

< 0.001
0.001
0.001
0.003
0.016
0.017
0.058
0.063
0.092
0.131
0.159
0.232

Model 4. Outcome: Preventable hospitalization for diabetes

Log-likelihood = -21.54, 2 = 18.77, p = 0.0277, Number of observations = 144
Number of patients for each PCP
4.75
3.29
Number of hospital admissions in the previous year
9.74
10.67
Number of PCP accesses in the previous year
0.15
0.17
Education level
0.27
0.23
Length of hospital stay
0.88
0.09
Number of specialist visits in community health services
0.36
0.37
Self-reported health status
2.61
2.57
Sex
2.47
2.33
Age
1.04
0.05

1.2218.48
1.1483.39
0.021.34
0.051.46
0.721.08
0.052.79
0.3818
0.3915.65
0.951.14

0.025
0.038
0.090
0.129
0.216
0.326
0.332
0.338
0.340

level is set at p 0.05

variables are presented in order of decreasing significance.
PCP, primary care physician.
1Significance
2The

Page 5 of 8
(page number not for citation purposes)

BMC Health Services Research 2007, 7:134

studies conducted on this topic used hospitalization rates

[13,25-30], investigations on prevalence of avoidable hospitalizations showed variable results ranging from 5.9%
in the USA considering 12 ACS conditions [31] to 34% in
the same country [30]. Therefore, our prevalence is one of
the highest encountered in the literature and deserves
detailed comments. It has been argued that hospital
admissions rates for ACSC alone are not sufficient proof
that the provision of ambulatory care is inadequate, since
some hospitalizations for ACSC will occur only because
some ACSC are less manageable than others. Moreover,
higher proportions of ACSCs may reflect higher prevalence for those particular diseases in the population.
However, prevalence of the chronic diseases evaluated in
our study is not substantially different from that encountered in most European countries, and we strongly believe
that the high prevalence of admissions for ACSCs is
related to unsatisfactory access and delivery of primary
care, which leads to deterioration of health conditions
and to overutilization of hospitals. Indeed, this conclusion has already been adduced in a survey conducted by
some of us to assess satisfaction with PCPs which showed
that many patients prefer to seek assistance at a more
sophisticated level of care [18]. Moreover, we investigated
emergency department use and hospitalization in the
prior year as very interesting indicators of PCP access
problems. Indeed, findings of a previous study conducted
by some of us in the same area, regarding the utilization
of the emergency department as a source of non-urgent
care indicated high overutilization of the department and
that patients were self-referred or referred by relatives to
the emergency department [32]. This observation emphasizes that, although all patients in our national health care
system have a general practitioner who should provide
primary care, the emergency department is used as a primary health care facility and, therefore, this prevents general practitioners from playing an effective role in the
patient management and as an interface between hospital
and community based services. As a consequence, we
agree that the health of patients with certain chronic medical conditions deteriorates without access to regular medical care and that this decline contributes to increased use
of emergency departments and ultimately to increased
hospitalizations [33,34]. Programs to improve the access
and quality of primary care have notably reduced emergency department visits. Results of such programs suggest
that areas with high levels of emergency department visits
may have less access to primary care, and therefore poorer
population health [35].
We found that avoidable hospitalizations are more likely
to have a shorter length of stay. We have tested this association, since in previous studies performed in our area,
high inappropriateness of hospital use was found and we
believe that for chronic conditions in many cases this

http://www.biomedcentral.com/1472-6963/7/134

could happen also for avoidable hospitalizations and

indeed inappropriate admissions have generally shorter
length of stay [36].
Almost all prior studies have computed rates from databases, and factors not contained such as variations in disease prevalence, health care seeking behavior, and
physician practice style could also affect preventable hospitalization rates [13], and it has been reported that the
study of predictors of ACSC hospitalizations should have
been applied to individual-level data [17]. We examined
preventable hospitalizations in individual patients, comparing subjects admitted for ACSC conditions to those
admitted for other reasons, and this strategy allowed us to
identify the independent effects of various factors, pertaining to subjects' characteristics and primary care supply, on the probability of ACSC admissions.
Regarding subjects' demographic and social characteristics
we found that older males were more likely to be hospitalized for ACSCs, whereas no other socio-demographic
feature tested predicted ACSCs. Preventable hospitalizations have been associated with lower socioeconomic status, advanced age, poor health and higher education [37],
since people with higher levels of education are likely to
have greater awareness and knowledge of the local health
care system and a better knowledge of disease processes,
and may therefore seek medical treatment earlier. In our
study education did not influence admissions for ACSCs,
whereas, as in a previous study [38], perceived poor health
status was indeed a significant predictor of hospitalization
for ACSCs.
Primary care access measures, such as satisfaction with
PCP health services and number of patients per PCP, as
well as proxies of propensity to seek care, such as number
of accesses and visits to PCPs in the previous year, were all
related to hospitalization for ACSCs. Our results indicate
that poor access to primary health care increases the likelihood of hospitalization for ACSCs, after controlling for
most of the other factors that may affect hospital admission, such as socio-demographics and propensity to seek
care. Some of the variables included in the model, such as
number of PCP accesses and visits in the previous year,
number of patients per PCP and satisfaction with PCP
services, may be highly correlated; therefore we calculated
some selected correlation coefficients and we found that,
as expected, number of PCP accesses and PCP visits were
significantly correlated (r = 0.46, p < 0.001) and satisfaction with PCP services was significantly correlated to
number of PCP visits (r = 0.30, p < 0.001), whereas
number of patients per PCP was inversely correlated to
number of PCP visits (r = -0.17, p < 0.001) and satisfaction with PCP health services (r = -0.19, p < 0.001). To
assess the role of each of these variables independently

Page 6 of 8
(page number not for citation purposes)

BMC Health Services Research 2007, 7:134

from others, we included all of them into the logistic

regression model. Therefore, we believe that our results
strongly support the validity of ACSCs as an indicator of
quality of primary care and confirm the crucial role of
PCPs in reducing unnecessary hospitalizations. Similar
findings have been reported by Zhan et al., who showed
that living in a primary care shortage area or in a higher
supply of hospital beds area represented independent risk
factor for a preventable hospitalization, even after controlling for other socio-demographic characteristics associated with health care use [37]. Moreover, adults
identified as having a PCP were four times more likely to
be discharged with a non-ACSC diagnosis compared to
those without a PCP [17].
We collected data in one hospital and concern about generalizability of our results may arise. However, although
very scarce data is available in Italy on ACSCs, it seems, as
for many other health services indicators, that there are
differences between Northern and Southern Italy. In a
recent report from the Observatory of the Health of the
Italian Regions, avoidable hospitalization rates were similar in Southern regions and consistently higher as compared to Northern regions [39]. For example, for adult
asthma the preventable hospitalization rates in Southern
Italy ranged from 0.27 discharges/1.000 inhabitants in
Campania and Sicily to 0.45 discharges/1.000 inhabitants
in Sardinia (Calabria region has a rate of 0.35 discharges/
1.000 inhabitants). In the centre and in the north of Italy
rates ranged from 0.13 discharges/1.000 inhabitants in
Valle d'Aosta to 0.33 discharges/1.000 inhabitants in
Bozen's autonomous administration, whereas the overall
national rate is 0.24 discharges/1.000 inhabitants [39].
Therefore, although we cannot exclude that our results
pertain only to our area, it is reasonable to suppose that
an analogous context may be referred to the Southern part
of our country. To have more insight into provision of primary care in other areas as measured by avoidable hospitalization, we strongly suggest replication of the study in
other regions of the country.
Most Western countries are facing tremendous transformations in their healthcare systems, and one main element of change concerns the transition from an inpatient
hospital-centered health care system to a multi level community-spread healthcare system that allows selection of
health conditions that can be treated only in hospital.
However, this transition is far from being completely
implemented in Italy and our findings confirm that
improvement of access to primary care may reduce avoidable hospitalizations. Therefore, our results have significant implications for health care policymakers, since
ACSCs may be useful to monitor effectiveness and to
reduce barriers to access to primary health care.

http://www.biomedcentral.com/1472-6963/7/134

Conclusion
In conclusion, the findings from this study add to the evidence and the urgency of developing and implementing
effective interventions to improve delivery of health care
at the community level and provided support to the usefulness of avoidable hospitalizations for ACSCs to monitor this process.

Competing interests
The author(s) declare that they have no competing interests.

Authors' contributions
PR and AB participated in the design of the study, collected the data, and contributed to the data analysis and
interpretation. MP designed the study, the data analysis
and interpretation, and wrote the article. IFA designed the
study, the data analysis and interpretation, and gave final
approval of the version to be published.
All authors read and approved the final manuscript.

Acknowledgements
At the time of this study Italo F. Angelillo was with the Chair of Hygiene,
Medical School, University of Catanzaro "Magna Grcia", Catanzaro (Italy)

References
1.

2.
3.
4.

5.

6.
7.

8.
9.

10.
11.

Agency for Healthcare Research and Quality: AHRQ Quality Indicators Guide to Prevention Quality Indicators: Hospital
Admission for Ambulatory Care Sensitive Conditions. Revision 4. Rockville, MD, Agency for Healthcare Research and Quality;
2004.
Ansari Z, Barbetti T, Carson NJ, Auckland MJ, Cicuttini F: The Victorian ambulatory care sensitive conditions study: rural and
urban perspectives. Soz Praventivmed 2003, 48:33-43.
Sanderson C, Dixon J: Conditions for which onset or hospital
admission is potentially preventable by timely and effective
ambulatory care. J Health Serv Res Policy 2000, 5:222-230.
Roos LL, Walld R, Uhanova J, Bond R: Physician visits, hospitalizations, and socioeconomic status: ambulatory care sensitive conditions in a Canadian setting. Health Serv Res 2005,
40:1167-1185.
Bermudez-Tamayo C, Marquez-Calderon S, Rodriguez del Aguila
MM, Perea-Milla Lopez E, Ortiz Espinosa J: Organizational characteristics of primary care and hospitalization for to the main
ambulatory care sensitive conditions. Aten Primaria 2004,
15:305-311.
Caminal J, Starfield B, Sanchez E, Casanova C, Morales M: The role
of primary care in preventing ambulatory care sensitive conditions. Eur J Public Health 2004, 14:246-251.
Zeng F, O'Leary JF, Sloss EM, Lopez MS, Dhanani N, Melnick G: The
effect of Medicare Health Maintenance Organizations on
hospitalization rates for Ambulatory Care-Sensitive Conditions. Med Care 2006, 44:900-907.
Gaskin DJ, Hoffman C: Racial and ethnic differences in preventable hospitalizations across 10 states. Med Care Res Rev 2000,
57(suppl 1):85-107.
Laditka JN: Hazards of hospitalization for ambulatory care
sensitive conditions among older women: evidence of
greater risks for African Americans and Hispanics. Med Care
Res Rev 2003, 60:468-495.
Laditka JN, Laditka SB: Race, ethnicity and hospitalization for
six chronic ambulatory care sensitive conditions in the USA.
Ethn Health 2006, 11:247-263.
Parchman ML, Culler S: Preventable hospitalisations in primary
care shortage areas. Arch Fam Med 1999, 8:487-491.

Page 7 of 8
(page number not for citation purposes)

BMC Health Services Research 2007, 7:134

12.
13.
14.
15.
16.
17.

18.
19.
20.
21.
22.
23.
24.
25.
26.
27.

28.
29.
30.
31.
32.
33.
34.
35.

36.
37.

Ricketts TC, Randolph R, Howard HA, Pathman D, Carey T: Hospitalization rates as indicators of access to primary care. Health
Place 2001, 7:27-38.
Bindman AB, Grumbach K, Osmond D, Komaromy M, Vranizan K,
Lurie N, Billings J, Stewart A: Preventable hospitalisations and
access to health care. JAMA 1995, 274:305-311.
Epstein AJ: The role of public clinics in preventable hospitalizations among vulnerable populations. Health Serv Res 2001,
36:405-420.
Gill JM, Mainous AG III: The role of provider continuity in preventing hospitalizations. Arch Fam Med 1998, 7:352-357.
Billings J, Teicholz N: Uninsured patients in District of Columbia hospitals. Health Aff 1990, 9:158-165.
Shi L, Samuels ME, Pease M, Bailey WP, Corley EH: Patient characteristics associated with hospitalizations for ambulatory care
sensitive conditions in South Carolina. South Med J 1999,
92:989-998.
Pavia M, Carbone V, Pileggi C, Angelillo IF: Patients' perceptions
and related behaviours on role of primary care physician in
Italy. Eur J Public Health 2004, 14:258-260.
Charlson ME, Pompei P, Ales KL, MacKenzie CR: A new method of
classifying prognostic comorbidity in longitudinal studies:
development and validation. J Chronic Dis 1987, 40:373-383.
Bhandari A, Wagner T: Self-reported utilization of health care
services: improving measurement and accuracy. Med Care Res
Rev 2006, 63:217-235.
Reijneveld SA, Stronks K: The validity of self-reported use of
health care across socioeconomic strata: a comparison of
survey and registration data. Int J Epidemiol 2001, 30:1407-1414.
Lubeck DP, Hubert HB: Self-report was a viable method for
obtaining health care utilization data in community-dwelling
seniors. J Clin Epidemiol 2005, 58:286-290.
World Health Organization: International classification of disease 9th
Revision edition. Geneva, Switzerland; 1997.
Stata Corporation: Stata Statistical Software: Release 8.1. College Station, TX. Stata Corp; 2003.
Billings J, Anderson GM, Newman LS: Recent findings on preventable hospitalizations. Health Aff 1996, 15:239-249.
Billings J, Zeitel L, Lukomnik J, Carey TS, Blank AE, Newman L:
Impact of socioeconomic status on hospital use in New York
City. Health Aff 1993, 12:162-173.
Davis SK, Liu Y, Gibbons GH: Disparities in trends of hospitalization for potentially preventable chronic conditions among
African Americans during the 1990s: implications and benchmarks. Am J Public Health 2003, 93:447-455.
Niti M, Ng TP: Avoidable hospitalisation rates in Singapore,
19911998: assessing trends and inequities of quality in primary care. J Epidemiol Community Health 2003, 57:17-22.
Pappas G, Hadden WC, Kozak LJ, Fisher GF: Potentially avoidable
hospitalizations: inequalities in rates between US socioeconomic groups. Am J Public Health 1997, 87:811-816.
Backus L, Moron M, Bacchetti P, Baker LC, Bindman AB: Effect of
managed care on preventable hospitalization rates in California. Med Care 2002, 40:315-324.
Kozak LJ, Hall MJ, Owings MF: Trends in avoidable hospitalizations, 19801998. Health Aff 2001, 20:225-232.
Bianco A, Pileggi C, Angelillo IF: Non-urgent visits to a hospital
emergency department in Italy.
Public Health 2003,
117:250-255.
Grossman LK, Rich LN, Johnson C: Decreasing nonurgent emergency department utilization by medicaid children. Pediatrics
1998, 102:20-24.
Zuckerman S, Brennan N, Yemane A: Has medicaid managed
care affected beneficiary access and use? Inquiry 2002,
39:221-242.
Laditka JN, Laditka SB, Probst JC: More may be better: evidence
of a negative relationship between physician supply and hospitalization for ambulatory care sensitive conditions. Health
Serv Res 2005, 40:1148-1166.
Angelillo IF, Ricciardi G, Nante N, Boccia A, Collaborative Group:
Appropriateness of hospital utilisation in Italy. Public Health
2000, 114:9-14.
Zhan C, Miller MR, Wong H, Meyer GS: The effects of HMO penetration on preventable hospitalizations. Health Serv Res 2004,
39:345-361.

http://www.biomedcentral.com/1472-6963/7/134

38.
39.

Blustein J, Hanson K, Shea S: Preventable hospitalizations and

socioeconomic status. Health Aff 1998, 17:177-189.
Osservatorio Nazionale sulla Salute nelle Regioni Italiane: Rapporto
Osservasalute 2006. Milano 2006.

Pre-publication history
The pre-publication history for this paper can be accessed
here:
http://www.biomedcentral.com/1472-6963/7/134/pre
pub

Publish with Bio Med Central and every

scientist can read your work free of charge
"BioMed Central will be the most significant development for
disseminating the results of biomedical researc h in our lifetime."
Sir Paul Nurse, Cancer Research UK

Your research papers will be:

available free of charge to the entire biomedical community
peer reviewed and published immediately upon acceptance
cited in PubMed and archived on PubMed Central
yours you keep the copyright

BioMedcentral

Submit your manuscript here:

http://www.biomedcentral.com/info/publishing_adv.asp

Page 8 of 8
(page number not for citation purposes)