Secondary Logo

Journal Logo

Brief Reports

Individual and Area-level Factors Contributing to the Geographic Variation in Ambulatory Care Sensitive Conditions in Finland

A Register-based Study

Satokangas, Markku MD*,†; Arffman, Martti MSc; Antikainen, Harri PhD; Leyland, Alastair H. PhD§; Keskimäki, Ilmo MD, PhD†,∥

Author Information
doi: 10.1097/MLR.0000000000001454

Abstract

Hospitalizations for ambulatory care sensitive conditions (ACSC) is among the most commonly used proxy indicators to measure primary health care (PHC) performance.1 ACSCs have been suggested as potentially avoidable by well-functioning PHC.2 As equity in delivery of health care can be assessed by measuring geographic variation in different medical practices,3 delivery of PHC can be assessed through geographic variation in ACSCs. Variation in ACSC is a common phenomenon,4 but it is driven more by individual socioeconomic position (SEP) and health status than general practitioner (GP) workload.5 Thus, the link between ACSCs and PHC performance remains controversial (Table 1).

TABLE 1 - Glossary
ACSC Ambulatory care sensitive conditions
COPD Chronic obstructive pulmonary disease
GP General practitioner
ICD-10 International Classification of Diseases 10th revision
IRR Incidence rate ratio
MRR Median rate ratio
PHC Primary health care
PCV Proportional change in variance
SEP Socioeconomic position

ACSCs seem to associate with increased health needs of individuals.6,7 From the patients’ perspective ACSCs result from nonadherence to treatment; for example, due to combined lack of support and mental health issues.8 This helps to understand why ACSCs are reduced by better patient support through specified family physicians,9 payment models that reward comprehensive care10 and continuity of care.11,12 However, ACSC rates do not mirror the results of PHC clinical quality indicators13—and their connection with the number of GPs remains inconsistent.9,14

Further, some individual and population characteristics contributing to variation in ACSCs might be partially addressable by PHC policies. Although the individual socioeconomic gradient in ACSC rates disfavors the poorest, these rates still differ depending on the incomes of residential areas.15 And while individual comorbidities are a major predictor of ACSCs,16,17 even area-level disease prevalence explains variation in ACSC rates.18 Shorter travel time to PHC reduces ACSCs rates,9,19 while shorter travel time to hospitals,19 high rurality,20 and high hospital bed supply12,21,22 increase them.

Few studies have assessed over time the development of geographic distribution23,24 or variation in ACSC rates,25,26 but describe these only with area-level factors.23,24 It is still unclear how a comprehensive array of area-level factors over time contributes to geographic variation in ACSCs, when individual SEP and health status are adjusted for.

FINNISH CONTEXT

The mainly tax-funded Finnish health care system offers a good framework to assess variation in ACSCs, comprising universal access, hierarchical structure and a long tradition of collecting individual hospitalization data. Finland has a strong public PHC where GPs operate as gatekeepers to specialist care provided by public hospitals.27 PHC is provided through ∼150 health centers,28 which act also as community hospitals by providing inpatient care in GP led wards. Hospital care is provided through 20 hospital districts—each having mainly a single 24/7 emergency hospital. An individual receives PHC services from a single health center, which receives its specialist care through a single hospital district. However, occupational and private health care offer alternative routes to GP and specialist care outpatient consultations.

METHODS

Outcome Variable

The Finnish Care Register for Health Care provided individual hospitalizations for the total Finnish population aged 20 years or more in 2011–2017. Of these we identified ACSCs using the UK definition29—with an addition of unspecified pneumonia (ICD-10 diagnosis code J18.9) as used previously in Finland.23 Further, we divided ACSCs into subgroups of acute, chronic, and vaccine-preventable (Supplemental Digital Content 1, https://links.lww.com/MLR/C130). Competent authorities linked hospitalization data into individual sociodemographic data. We formed annual cohorts and applied municipality of residence to allocate individuals into health center areas according to arrangement of PHC; and into hospital districts according to arrangement of hospital care. To account for hospital transfers we combined hospitalizations that occurred within 1 day of each other.

With these cohorts we asked: (1) how geographic variance in ACSCs was distributed between PHC and hospital care; (2) which factors predicted ACSCs; (3) which area-level factors explained variance in ACSCs when analyzed separately; (4) what were the relative proportions of variance explained by area-level factors after adjustment for individual factors; and (5) how the heterogeneity in risk of ACSCs developed with each added factor?

Individual Factors

The annual FOLK databases provided data for sex, age, municipality of residence, and household income for each individual with ACSCs. For these individuals we selected 5 comorbidities [chronic heart failure, chronic obstructive pulmonary disease (COPD), diabetes, hypertension, and dementia] as suggested by Saver et al16 from inpatient hospitalizations and specialist outpatient visits during the previous 5 years.

Area-level Factors

We chose area-level factors previously suggested to affect ACSCs, such as disease prevalence18 and hospital supply.12 These factors were allocated to health center areas and further categorized for the purposes of reporting—as summarized in Table 2. We categorized the proportion of the population aged 65 and above receiving pensioner’s care allowance as disease burden because it captures both disease prevalence and functional limitations. It is granted when a chronic disease or disability limits daily living activities.

TABLE 2 - The Categories of Individual and Area-level Factors Added Into the Models Estimating the Geographic Variance of ACSCs
Added to Category Factors Hypothesized Pathway for Risk of ACSC Hospitalization Data Obtained From
Model 1 Individual demographics Age and sex Null model FOLK database maintained by Statistics Finland
Model 2 Individual SEP and health status Household incomes Lower SEP predisposes to material deprivation and possibly to poorer care,30 for example through communication mismatch with GPs31 FOLK database maintained by Statistics Finland
Number of comorbidities Multimorbidity complicates treatments and associates with acute diseases leading to hospitalizations32 The Care Register for Health Care maintained by Finnish Institute of Health and Welfare
Model 3 Area-level disease burden Proportion of population aged ≥65 receiving pensioner’s care allowance (%) Higher proportion of multimorbid elderly (with limitations of activities in daily living) within an area strain health care, possibly lowering the threshold of hospital utilization Statistical database Kelasto maintained by Social Insurance Institution of Finland33
Model 4 Area-level arrangement and usage of hospital care Proportion of ACSCs occurring in GP led wards of all ACSCs (%) GP led wards are likely to have less constrictive intake criteria than specialist care Aggregated from the Care Register for Health Care maintained by Finnish Institute of Health and Welfare
Rate of hospital bed utilization in specialist health care Available hospital beds promote utilization of hospital care34
Model 5 Area-level distance to health services Populations’ average distance to health center (km) Long distance to health services might cause delay and lead to disease exacerbation Road and street network data provided by both Esri Finland and the Finnish Transport Agency (Digiroad database)
Populations’ average distance to emergency hospital (km)
Model 6 Other area-level factors related to health services Number of GPs per 1000 inhabitants Fewer GPs represent poorer availability of care General Practitioner survey provided by Finnish Medical Association
Income median (€) Area’s wealth might affect the care of multimorbid patients,35 differences in provision of primary care15 Aggregated from the FOLK database maintained by Statistics Finland
The subsequent models include all factors from the previous ones. All area-level factors were allocated to health center areas.
ACSC indicates ambulatory care sensitive conditions; GP, general practitioner; SEP, socioeconomic position.

Statistical Methods

Age-standardized rates were calculated using the direct standardization.36 We built on the analysis strategy presented by Falster et al5 and allocated the annual cohorts to 3-level Poisson multilevel models: individuals were nested within 131 health center areas, which were in turn nested within 20 hospital districts. We analyzed separate models for total ACSCs and each of the 3 ACSC subgroups in 3 consecutive time periods: 2011–2012, 2013–2014, and 2015–2017. The total Finnish population of comparable age was applied as the population at risk and added into the models as an offset (range, 4020–519,853 in health centers and 22,410–1,288,747 in hospital districts).

By first analyzing an age-adjusted and sex-adjusted model (model 1) we estimated the random parameters (σ2) representing the variance at health center and hospital district levels to which we compared all subsequent models. We added the factors stepwise into the models; and measured the effects of each addition with the proportional change in variance (PCV).37 To analyze which factors predicted ACSCs we calculated incidence rate ratios (IRR). Finally, we calculated the areas’ heterogeneity in risk of ACSCs with median rate ratios.38 The statistical analyses were performed with R, release versions 3.5.139—using the Laplace approximation method.40

RESULTS

We observed 729,008 ACSCs in Finland in 2011–2017 (Table 3). The age-standardized ACSC rate in the adult population decreased from 2.67 (95% confidence interval, 2.66–2.68) per 100 person-years in 2011–2012 to 2.57 (2.56–2.58) in 2015–2017. In the age-adjusted and sex-adjusted models—model 1—of total ACSCs the hospital district level variance was approximately twice that of the health center level. These variances decreased slightly over time at both area levels, although the decrease was more pronounced at hospital district level.

TABLE 3 - Cohort Characteristics for All Hospitalizations for ACSCs and Average ACSC Rates per 100 Person-Years in 3 Studied Time Periods
2011–2012 2013–2014 2015–2017
Variables Range of Area-level Factors Persons ACSC Rate Range of Area-level Factors Persons ACSC Rate Range of Area-level Factors Persons ACSC Rate
Total study population 4,203,024 2.5 4,254,995 2.5 4,388,766 2.6
Age (y)
 20–54 1,449,327 0.6 1,474,922 0.6 1,573,093 0.6
 55–64 1,463,115 1.1 1,428,219 1.0 1,403,760 1.0
 65–74 672,900 2.6 708,863 2.5 739,912 2.5
 75–84 392,232 6.1 406,370 5.8 426,180 5.8
 85+ 225,450 15.9 236,621 15.6 245,821 16.2
Sex
 Males 2,043,104 2.6 2,071,126 2.5 2,143,146 2.7
 Females 2,159,920 2.4 2,183,869 2.4 2,245,620 2.5
Income
 Quintile 1 855,827 4.7 865,091 4.5 904,007 4.4
 Quintile 2 812,472 3.9 824,835 3.8 845,825 4.1
 Quintile 3 803,963 2.1 815,350 2.0 842,794 2.3
 Quintile 4 836,117 1.3 845,743 1.3 870,886 1.4
 Quintile 5 894,479 0.9 903,808 0.9 925,084 0.9
No. comorbidities
 0 3,829,145 1.3 3,853,135 1.2 3,961,902 1.3
 1 285,115 10.5 301,763 9.8 316,921 9.8
 2 74,391 23.8 84,120 22.5 92,130 22.4
 3+ 14,373 54.7 15,977 52.6 17,813 55.4
Proportion of population aged ≥65 receiving pensioner’s care allowance (%)
 Tercile 1 8.3–16.7 1,763,672 1.9 9.0–15.8 1,827,338 1.9 8.1–14.6 1,877,242 2.0
 Tercile 2 16.8–19.5 1,562,636 2.7 15.9–18.5 1,550,841 2.6 14.7–17.5 1,664,597 2.7
 Tercile 3 19.6–25.6 876,716 3.5 18.6–25.0 876,816 3.3 17.6–24.0 846,927 3.5
Proportion of ACSCs occurring in GP led wards of all ACSCs (%)
 Tercile 1 32.9–25.4 1,842,453 2.3 1.1–29.7 1,953,356 2.3 0.5–26.8 2,454,779 2.2
 Tercile 2 25.5–41.2 1,162,415 2.6 29.8–42.5 1,581,758 2.2 26.9–45.3 1,016,471 2.9
 Tercile 3 41.3–72.9 1,198,156 2.8 42.6–73.7 719,881 3.4 45.4–73.2 917,516 3.3
Rate of hospital bed utilization in specialist health care
 Tercile 1 0.13–0.23 2,395,000 2.0 0.14–0.22 2,362,852 1.9 0.14–0.21 2,400,369 2.0
 Tercile 2 0.24–0.28 1,204,534 3.0 0.23–0.27 1,329,090 2.8 0.22–0.26 1,285,578 2.9
 Tercile 3 0.29–0.39 603,490 3.9 0.28–0.41 563,053 3.9 0.27–0.48 702,819 3.9
Populations’ average distance to health center (km)
 Tercile 1 1.2–4.7 2,374,959 2.1 1.2–4.6 2,410,214 2.0 1.2–4.8 2,517,459 2.1
 Tercile 2 4.8–6.6 1,070,273 3.0 4.7–6.6 1,039,132 2.9 4.8–6.6 1,026,532 3.0
 Tercile 3 6.7–26.4 757,792 3.4 6.7–26.8 805,649 3.2 6.7–26.9 844,775 3.3
Populations’ average distance to emergency hospital (km)
 Tercile 1 2.8–20.8 2,673,218 2.2 2.8–21.1 2,741,170 2.1 2.8–21.8 2,900,462 2.3
 Tercile 2 20.9–45.8 957,014 2.7 21.2–45.8 948,449 2.6 21.9–51.3 941,334 2.6
 Tercile 3 45.9–331.1 572,792 3.9 45.9–331.1 565,376 3.8 51.4–331.1 546,970 3.9
No. GPs per 1000 inhabitants
 Tercile 1 (median)* 0.58 2,281,069 2.3 0.61 2,500,773 2.2 0.59 2,803,283 2.3
 Tercile 2 (median)* 0.71 1,265,989 2.7 0.74 932,620 2.8 0.74 877,856 2.8
 Tercile 3 (median)* 0.82 655,966 3.2 0.84 821,602 3.0 0.85 707,627 3.4
Income median (€)
 Tercile 1 17,900–20,600 772,433 3.7 19,000–21,900 722,894 3.6 20,000–22,800 868,618 3.4
 Tercile 2 20,700–22,200 1,695,028 2.6 22,000–23,700 1,767,878 2.6 22,800–24,400 1,695,119 2.8
 Tercile 3 22,300–36,800 1,735,563 1.9 23,800–38,600 1,764,223 1.8 24,500–40,600 1,825,029 2.0
*We had no permission to report the range for number of GPs acquired from the General Practitioner survey.
ACSC indicates ambulatory care sensitive conditions; GP, general practitioner.

In the final models (model 6) male sex, higher age, lower income, and higher number of comorbidities associated with higher IRRs in total ACSCs and all ACSC subgroups (Table 4). Area-level higher disease burden, hospital bed utilization rate, and median income were associated with higher IRRs in total ACSCs in all time periods—but a higher proportion of ACSCs in GP led wards only in 2013–2014 (Supplemental Digital Content 1, https://links.lww.com/MLR/C130).

TABLE 4 - IRRs of Individual and Area-level Factors Significantly Associated With Total ACSCs and ACSC Subgroups in Finland in 2015–2017; From the Multilevel Poisson Models Adjusted Simultaneously for All Individual and Area-level Factors (Model 6)
Total Acute Chronic Vaccine-preventable
Variables IRR (95% CI) P IRR (95% CI) P IRR (95% CI) P IRR (95% CI) P
Sex
 Male 1.00 1.00 1.00 1.00
 Female 0.74 (0.73–0.74) <0.001 0.95 (0.93–0.96) <0.001 0.75 (0.74–0.76) <0.001 0.60 (0.59–0.61) <0.001
Age (y)
 20–54 1.00 1.00 1.00 1.00
 55–64 1.77 (1.74–1.80) <0.001 1.11 (1.08–1.14) <0.001 3.73 (3.59–3.88) <0.001 2.16 (2.09–2.22) <0.001
 65–74 3.58 (3.52–3.63) <0.001 1.66 (1.62–1.70) <0.001 8.95 (8.62–9.29) <0.001 4.88 (4.74–5.02) <0.001
 75–84 6.28 (6.18–6.37) <0.001 2.47 (2.41–2.53) <0.001 16.12 (15.54–16.73) <0.001 9.20 (8.95–9.46) <0.001
 85+ 12.06 (11.88–12.25) <0.001 4.71 (4.60–4.83) <0.001 28.69 (27.66–29.77) <0.001 19.66 (19.13–20.21) <0.001
Income quintile
 Lowest 1.00 1.00 1.00 1.00
 2 0.78 (0.77–0.78) <0.001 0.74 (0.72–0.75) <0.001 0.77 (0.76–0.78) <0.001 0.80 (0.79–0.82) <0.001
 3 0.65 (0.64–0.65) <0.001 0.63 (0.61–0.64) <0.001 0.62 (0.61–0.63) <0.001 0.69 (0.67–0.70) <0.001
 4 0.54 (0.53–0.55) <0.001 0.53 (0.51–0.54) <0.001 0.50 (0.49–0.51) <0.001 0.59 (0.58–0.60) <0.001
 Highest 0.43 (0.43–0.44) <0.001 0.44 (0.43–0.45) <0.001 0.38 (0.37–0.39) <0.001 0.48 (0.47–0.49) <0.001
No. comorbidities: 0–5 (+1 comorbidity) 2.33 (2.32–2.33) <0.001 1.89 (1.88–1.91) <0.001 2.80 (2.79–2.81) <0.001 2.01 (2.00–2.02) <0.001
Proportion of population aged 65+ receiving pensioner’s care allowance (+1 SD) 1.06 (1.02–1.11) 0.008 1.12 (1.05–1.19) <0.001 1.00 (0.95–1.06) 0.867 1.09 (1.04–1.15) 0.001
Proportion of ACSCs occurring in GP led wards of all ACSC hospitalizations (+1 SD) 1.03 (1.00–1.07) 0.024 1.01 (0.97–1.05) 0.622 1.04 (1.00–1.07) 0.060 1.06 (1.02–1.10) 0.002
Rate of hospital bed utilization in specialist health care (+1 SD) 1.08 (1.04–1.12) <0.001 1.08 (1.03–1.14) 0.002 1.12 (1.07–1.17) <0.001 1.02 (0.98–1.07) 0.325
Income median (+1 SD) 1.07 (1.03–1.11) 0.001 1.10 (1.05–1.16) <0.001 1.05 (1.01–1.10) 0.029 1.06 (1.01–1.11) 0.020
The IRR of area-level distances to health services and number of GPs did not significantly associate with ACSCs. Over time all these associations were stable as shown in Supplementary Digital Content 1 (https://links.lww.com/MLR/C130). A bold font indicate statistically significant P values.
ACSC indicates ambulatory care sensitive conditions; CI, confidence interval; GP, general practitioner; IRR, incidence rate ratios.

Each area-level factor explained variation in ACSCs when analyzed separately (Supplemental Digital Content 1, https://links.lww.com/MLR/C130). For total ACSCs our final model explained a little less than half (PCV=39.8%–52.1%) the health center level variance and almost two thirds (PCV=61.1%–74.2%) the hospital district level variance. The highest PCV occurred in 2011–2012 from when it decreased over time along with the variances in age-adjusted and sex-adjusted models. Although individual SEP and health status explained 18.7%–29.9% of health center level variance and 24.6%–35.7% of hospital district level variance, a combination of area-level disease burden and both arrangement and usage of hospital care accounted for an additional 14.1%–16.2% and 32.4%–33.0% of the respective variances. And while the distance to health services and number of GPs did not add much to the models, the areas’ median income accounted for an additional 2.7%–6.5% and 4.2%–7.6% of these variances. Areas had some heterogeneity in risk of ACSCs which decreased with subsequent models (Table 5).

TABLE 5 - Variance in Age-adjusted and Sex-adjusted Model (Model 1) in Total ACSCs and ACSC Subgroups Between HC Areas and HD in Finland at 3 Consecutive Time Periods
2011–2012 2013–2014 2015–2017
HC HD HC HD HC HD
Models σ2 PCV (%) MRR σ2 PCV (%) MRR σ2 PCV (%) MRR σ2 PCV (%) MRR σ2 PCV (%) MRR σ2 PCV (%) MRR
Total ACSCs
 Model 1 0.023 1.16 0.050 1.24 0.020 1.14 0.034 1.19 0.019 1.14 0.037 1.20
 Model 2 0.016 29.9 1.13 0.032 35.7 1.19 0.014 27.9 1.12 0.026 24.6 1.16 0.016 18.7 1.13 0.025 31.0 1.16
 Model 3 0.014 37.3 1.12 0.018 63.3 1.14 0.013 35.2 1.11 0.018 46.5 1.14 0.015 21.7 1.12 0.019 48.3 1.14
 Model 4 0.013 44.0 1.11 0.016 68.1 1.13 0.011 44.1 1.10 0.014 57.3 1.12 0.013 34.3 1.11 0.013 64.0 1.12
 Model 5 0.013 45.6 1.11 0.017 66.6 1.13 0.011 44.7 1.10 0.015 54.8 1.13 0.013 34.5 1.11 0.013 63.3 1.12
 Model 6 0.011 52.1 1.11 0.013 74.2 1.11 0.010 47.5 1.10 0.013 61.1 1.12 0.012 39.8 1.11 0.012 67.4 1.11
Acute ACSCs
 Model 1 0.027 1.17 0.038 1.21 0.025 1.16 0.035 1.20 0.026 1.17 0.055 1.25
 Model 2 0.027 0.2 1.17 0.029 24.2 1.18 0.022 9.6 1.15 0.026 25.9 1.17 0.024 8.1 1.16 0.041 26.2 1.21
 Model 3 0.027 0.8 1.17 0.018 52.9 1.14 0.022 11.8 1.15 0.016 53.6 1.13 0.023 10.8 1.16 0.029 46.8 1.18
 Model 4 0.026 3.8 1.17 0.017 56.8 1.13 0.021 16.0 1.15 0.013 62.4 1.12 0.022 12.7 1.15 0.023 59.0 1.15
 Model 5 0.025 6.7 1.16 0.017 57.0 1.13 0.020 17.8 1.15 0.012 65.6 1.12 0.022 13.4 1.15 0.022 59.5 1.15
 Model 6 0.021 21.5 1.15 0.013 66.1 1.11 0.019 21.8 1.14 0.010 71.4 1.10 0.020 22.7 1.14 0.019 65.7 1.14
Chronic ACSCs
 Model 1 0.034 1.19 0.072 1.29 0.029 1.18 0.055 1.25 0.027 1.17 0.035 1.20
 Model 2 0.024 28.9 1.16 0.049 32.6 1.23 0.025 13.4 1.16 0.044 21.5 1.22 0.023 13.9 1.16 0.026 27.6 1.16
 Model 3 0.023 31.8 1.16 0.033 55.0 1.19 0.025 14.9 1.16 0.035 36.7 1.20 0.023 13.3 1.16 0.022 38.5 1.15
 Model 4 0.020 41.9 1.14 0.027 63.1 1.17 0.021 29.7 1.15 0.026 53.3 1.17 0.018 34.7 1.13 0.016 55.3 1.13
 Model 5 0.019 44.3 1.14 0.030 57.9 1.18 0.020 32.0 1.14 0.030 46.7 1.18 0.017 36.2 1.13 0.017 51.4 1.13
 Model 6 0.018 45.6 1.14 0.028 61.4 1.17 0.020 32.3 1.14 0.027 51.2 1.17 0.017 38.5 1.13 0.016 53.5 1.13
Vaccine-preventable ACSCs
 Model 1 0.024 1.16 0.033 1.19 0.026 1.17 0.021 1.15 0.024 1.16 0.037 1.20
 Model 2 0.018 25.9 1.14 0.026 20.1 1.17 0.020 26.1 1.14 0.021 1.4 1.15 0.021 13.1 1.15 0.032 14.8 1.18
 Model 3 0.014 43.0 1.12 0.019 42.2 1.14 0.016 40.1 1.13 0.019 9.6 1.14 0.020 19.9 1.14 0.026 31.0 1.16
 Model 4 0.013 46.4 1.11 0.017 47.4 1.13 0.015 44.0 1.12 0.018 15.0 1.14 0.018 28.4 1.13 0.020 45.9 1.14
 Model 5 0.013 46.6 1.11 0.017 48.7 1.13 0.015 44.1 1.12 0.018 15.2 1.14 0.018 28.4 1.13 0.020 46.0 1.14
 Model 6 0.011 52.9 1.11 0.015 56.1 1.12 0.014 48.9 1.12 0.016 21.1 1.13 0.017 31.8 1.13 0.019 48.8 1.14
Each subsequent model builds on the previous one and adds a category of explanatory factors: individual socioeconomic position and health status (model 2), area-level disease burden (model 3), area-level arrangement and usage of hospital care (model 4), area-level distance to health services (model 5), and other area-level factors related to health services (model 6).
ACSCs indicates ambulatory care sensitive conditions; HC, health center; HD, hospital districts; MRR, median rate ratio; PCV, proportional change in variance, calculated as percentual decrease in variance between each model and model 1.

DISCUSSION

This study analyzed how individual and area-level factors over time contributed to geographic variance in ACSCs between 2 nested levels of health service providers in Finland. Among total ACSCs our final models explained less than half (PCV=39.8%–52.1%) the variance between health center areas and almost two thirds (PCV=61.1%–74.2%) between hospital districts. Even after adjusting for individual SEP and health status, area-level disease burden and both arrangement and usage of hospital care still explained 14.1%–16.2% and 32.4%–33.0% of these variances. The proportions explained were consistent over time. In age-adjusted and sex-adjusted models hospital districts showed more variation than health center areas—a disparity which evened out after adjusting for the studied factors. This suggest that variation in age-standardized and sex-standardized ACSC rates could be driven more by factors related to hospital services rather than PHC.

Our findings support the previous studies stating that variation in ACSCs reflects health status,5,16–18,41 SEP,5,12,15,21 and factors related to hospital care.12,19,21,42–44 However, this study emphasized that these factors explained more of the variance occurring in hospital district level than in health center level.

The finding for area-level disease burden—a combination of disease prevalence and functional limitations—adds to earlier knowledge.18 It explained additional variance independent of individual SEP and comorbidities. We suggest caution when interpreting ACSC rates adjusted with only disease prevalence—especially if this factor is used as a proxy for individual health status. We had a few possible hypotheses for this independent effect. At the hospital district level, it might reflect either different admission criteria between hospitals or systematically insufficient capacity of PHC to answer the high morbidity and disabilities. For health centers, it might reflect inadequate response of PHC in some areas; a possible link between ACSCs and PHC performance.

The finding that local GP led wards maintained variation in ACSCs was consistent with previous studies.42,43 Countries applying ACSC rates as PHC performance indicators should consider if their arrangement of hospital care affects these rates. ACSCs also reflected areas’ overall tendency for hospital utilization, suggesting unnecessary use of available hospital supply34 rather than different GP referral practices.45 The effects of distances to health services and the number of GPs were captured by other factors—such as the arrangement of hospital care. Our finding that higher area-level income predicted higher ACSCs contradicts the previously reported low ACSC rates in wealthy areas.15 These lower rates might not reflect better performing PHC, but rather other factors such as population’s favorable health status.

Strengths and Limitations

The main strength of this study was that we were able to apply nested multilevel Poisson model to distinguish the variance in ACSCs between PHC and hospital care. Further, the individual hospitalization data used are of good quality46—and their comprehensive usage ensured the generalizability of our results in Finnish context. The relative proportions will differ when analyzed elsewhere, but the phenomena behind the analyzed factors are transferrable internationally.

Our study was limited by the lack of individual-level survey data to enrich the applied registers. Moreover, we did not analyze the effect of alternative routes to physician consultations in Finland: occupational and private health care—but believe that including individual and area-level incomes account for them. As the geographic diagnosis coverage in Finnish Register of PHC visits was partial,47 we had to collect individual comorbidities from both specialist care outpatient visits and hospital discharges.

CONCLUSIONS

This study observed that administrative areas’ disease burden and hospital care utilization patterns contributed to geographic variation in ACSCs—potential links to PHC performance. This followed adjusting ACSCs for individual SEP and health status, as well as for the country-specific arrangement of hospital care. Countries measuring PHC performance with ACSC rates should (1) consider that these rates might be driven more by hospital care than PHC; (2) interpret prevalence adjusted rates with caution; and (3) consider that the arrangement of their hospital care might affect not only ACSC rates, but also other factors thought to reflect only the provision of PHC services.

ACKNOWLEDGMENTS

The authors thank the Finnish Medical Association for the permission to use their General Practitioner survey data. Results were presented at the European Public Health Conference 2019 in Marseilles, France on November 23, 2019.

REFERENCES

1. Billings J, Anderson GM, Newman LS. Recent findings on preventable hospitalizations. Health Aff. 1996;15:239–249.
2. Billings J, Zeitel L, Lukomnik J, et al. Impact of socioeconomic status on hospital use in New York City. Health Aff. 1993;12:162–173.
3. Westert GP, Groenewegen PP, Boshuizen HC, et al. Medical practice variations in hospital care; time trends of a spatial phenomenon. Health Place. 2004;10:215–220.
4. Busby J, Purdy S, Hollingworth W. A systematic review of the magnitude and cause of geographic variation in unplanned hospital admission rates and length of stay for ambulatory care sensitive conditions. BMC Health Serv Res. 2015;15:324.
5. Falster MO, Jorm LR, Douglas KA, et al. Sociodemographic and health characteristics, rather than primary care supply, are major drivers of geographic variation in preventable hospitalizations in Australia. Med Care. 2015;53:436–445.
6. Falster MO, Jorm LR, Leyland AH. Visualising linked health data to explore health events around preventable hospitalisations in NSW Australia. BMJ Open. 2016;6:e012031.
7. Vuik SI, Fontana G, Mayer E, et al. Do hospitalisations for ambulatory care sensitive conditions reflect low access to primary care? An observational cohort study of primary care usage prior to hospitalisation. BMJ Open. 2017;7:e015704.
8. Longman JM, Rix E, Johnston JJ, et al. Ambulatory care sensitive chronic conditions: what can we learn from patients about the role of primary health care in preventing admissions? Aust J Prim Health. 2018;24:304–310.
9. Carneiro CS. Hospitalisation of ambulatory care sensitive conditions and access to primary care in Portugal. Public Health. 2018;165:117–124.
10. Laberge M, Wodchis WP, Barnsley J, et al. Hospitalizations for ambulatory care sensitive conditions across primary care models in Ontario, Canada. Social Sci Med. 2017;181:24–33.
11. Barker I, Steventon A, Deeny SR. Association between continuity of care in general practice and hospital admissions for ambulatory care sensitive conditions: cross sectional study of routinely collected, person level data. BMJ. 2017;356:j84.
12. Busby J, Purdy S, Hollingworth W. How do population, general practice and hospital factors influence ambulatory care sensitive admissions: a cross sectional study. BMC Fam Pract. 2017;18:67.
13. van der Pol M, Olajide D, Dusheiko M, et al. The impact of quality and accessibility of primary care on emergency admissions for a range of chronic ambulatory care sensitive conditions (ACSCs) in Scotland: longitudinal analysis. BMC Fam Pract. 2019;20:32.
14. Chang C-H, O’Malley AJ, Goodman DC. Association between temporal changes in primary care workforce and patient outcomes. Health Serv Res. 2017;52:634–655.
15. Lofqvist T, Burstrom B, Walander A, et al. Inequalities in avoidable hospitalisation by area income and the role of individual characteristics: a population-based register study in Stockholm County, Sweden. BMJ Qual Saf. 2014;23:206–214.
16. Saver BG, Wang CY, Dobie SA, et al. The central role of comorbidity in predicting ambulatory care sensitive hospitalizations. Eur J Public Health. 2014;24:66–72.
17. Eggli Y, Desquins B, Seker E, et al. Comparing potentially avoidable hospitalization rates related to ambulatory care sensitive conditions in Switzerland: the need to refine the definition of health conditions and to adjust for population health status. BMC Health Serv Res. 2014;14:25.
18. Pollmanns J, Romano PS, Weyermann M, et al. Impact of disease prevalence adjustment on hospitalization rates for chronic ambulatory care-sensitive conditions in Germany. Health Serv Res. 2018;53:1180–1202.
19. Huang Y, Meyer P, Jin L. Spatial access to health care and elderly ambulatory care sensitive hospitalizations. Public Health. 2019;169:76–83.
20. Burgdorf F, Sundmacher L. Potentially avoidable hospital admissions in Germany: an analysis of factors influencing rates of ambulatory care sensitive hospitalizations. Dtsch Arztebl Int. 2014;111:215–223.
21. Weeks WB, Ventelou B, Paraponaris A. Rates of admission for ambulatory care sensitive conditions in France in 2009–2010: trends, geographic variation, costs, and an international comparison. Eur J Health Econ. 2016;17:453–470.
22. Van Loenen T, Faber MJ, Westert GP, et al. The impact of primary care organization on avoidable hospital admissions for diabetes in 23 countries. Scand J Primary Health Care. 2016;34:5–12.
23. Satokangas M, Lumme S, Arffman M, et al. Trajectory modelling of ambulatory care sensitive conditions in Finland in 1996–2013: assessing the development of equity in primary health care through clustering of geographic areas—an observational retrospective study. BMC Health Serv Res. 2019;19:629.
24. Sheringham J, Asaria M, Barratt H, et al. Are some areas more equal than others? Socioeconomic inequality in potentially avoidable emergency hospital admissions within English local authority areas. J Health Serv Res Policy. 2017;22:83–90.
25. Manderbacka K, Arffman M, Satokangas M, et al. Regional variation of avoidable hospitalisations in a universal health care system: a register-based cohort study from Finland 1996−2013. BMJ Open. 2019;9:e029592.
26. Librero J, Ibañez-Beroiz B, Peiró S, et al. Trends and area variations in potentially preventable admissions for COPD in Spain (2002–2013): a significant decline and convergence between areas. BMC Health Serv Res. 2016;16:367.
27. Keskimäki I, Tynkkynen L-K, Reissell E, et al. Finland: health system review. Health Syst Transit. 2019;21:1–166.
28. Parhiala K. Avainlukuja perusterveydenhuollon järjestämisestä Suomessa 2013-2019 [in Finnish]. Helsinki: Finnish Institute of Health and Welfare; 2019:1–4.
29. NHS Institute for Innovation and Improvement. Indicator construction: managing variation in emergency admissions. 2016. Available at: https://digital.nhs.uk/binaries/content/assets/website-assets/data-and-information/data-tools-and-services/data-services/innovative-uses-of-data/acsc-appendix-a.pdf. Accessed August 11, 2017.
30. Blustein J. Preventable hospitalizations and socioeconomic status. Health Affairs. 1998;17:177–189.
31. Willems S, De Maesschalck S, Deveugele M, et al. Socio-economic status of the patient and doctor-patient communication: does it make a difference? Patient Educ Counsel. 2005;56:139–146.
32. Buurman BM, Frenkel WJ, Abu-Hanna A, et al. Acute and chronic diseases as part of multimorbidity in acutely hospitalized older patients. Eur J Intern Med. 2016;27:68–75.
33. Official Statistics of Finland: disability benefits [e-publication]. 2018. Available at: https://www.kela.fi/web/en/496. Accessed January 23, 2019.
    34. Delamater PL, Messina JP, Grady SC, et al. Do more hospital beds lead to higher hospitalization rates? A spatial examination of Roemer’s law. PLoS One. 2013;8:e54900.
    35. Mercer SW, Zhou Y, Humphris GM, et al. Multimorbidity and socioeconomic deprivation in primary care consultations. Ann Family Med. 2018;16:127–131.
      36. Pace M, Lanzieri G, Glickman M, et al. Revision of the European Standard Population: Report of Eurostat’s Task Force. Luxembourg: Publications Office of the European Union; 2013.
      37. Merlo J, Yang M, Chaix B, et al. A brief conceptual tutorial on multilevel analysis in social epidemiology: investigating contextual phenomena in different groups of people. J Epidemiol Community Health. 2005;59:729–736.
      38. Austin PC, Stryhn H, Leckie G, et al. Measures of clustering and heterogeneity in multilevel Poisson regression analyses of rates/count data. Stat Med. 2018;37:572–589.
      39. R Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing; 2019.
      40. Bates D, Mächler M, Bolker B, et al. Fitting linear mixed-effects models using lme4. J Stat Softw. 2015;67:1–48.
      41. Paul MC, Dik J-WH, Hoekstra T, et al. Admissions for ambulatory care sensitive conditions: a national observational study in the general and COPD population. Eur J Public Health. 2019;29:213–219.
      42. Falster MO, Leyland AH, Jorm LR. Do hospitals influence geographic variation in admission for preventable hospitalisation? A data linkage study in New South Wales, Australia. BMJ Open. 2019;9:e027639.
      43. Berlin C, Busato A, Rosemann T, et al. Avoidable hospitalizations in Switzerland: a small area analysis on regional variation, density of physicians, hospital supply and rurality. BMC Health Serv Res. 2014;14:289.
      44. Kim AM, Park JH, Yoon TH, et al. Hospitalizations for ambulatory care sensitive conditions as an indicator of access to primary care and excess of bed supply. BMC Health Serv Res. 2019;19:259.
      45. Franks P, Mooney C, Sorbero M. Physician referral rates: style without much substance? Med Care. 2000;38:836–846.
      46. Sund R. Quality of the Finnish Hospital Discharge Register: a systematic review. Scand J Public Health. 2012;40:505–515.
      47. Ketola E, Pitkänen V, Huvinen S, et al. The national primary health care case-mix described [in Finnish]. Finnish Med J. 2019;74:2027–2034.
      Keywords:

      health services research; multilevel modelling; preventable hospitalizations; primary care

      Supplemental Digital Content

      Copyright © 2020 The Author(s). Published by Wolters Kluwer Health, Inc.