Excess healthcare utilization and costs linked to chronic conditions: a comparative study of nine European countries

European Journal of Public Health, Vol. 35, No. 2, 216–227 © The Author(s) 2025. Published by Oxford University Press on behalf of the European Public Health Association. This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial License (https://creativecommons. org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact [email protected] for reprints and translation rights for reprints. All other permissions can be obtained through our RightsLink service via the Permissions link on the article page on our site—for further information please contact journals.permis- [email protected]. https://doi.org/10.1093/eurpub/ckaf012 Advance Access published on 11 February 2025 ............................................................................................................... Excess healthcare utilization and costs linked to chronic conditions: a comparative study of nine European countries Boris Polanco 1,2 , Ana O~ na 2 , Armin Gemperli 1,3 , Diana Pacheco Barzallo 1,2,4, � 1 Faculty of Health Sciences and Medicine, University of Lucerne, Lucerne, Switzerland 2 Health Economics Group, Swiss Paraplegic Research, Nottwil, Switzerland 3 Center of Primary and Community Care, University of Lucerne, Lucerne, Switzerland 4 Center for Rehabilitation in Global Health Systems, University of Lucerne, Lucerne, Switzerland �Corresponding author. Faculty of Health Sciences and Medicine, University of Lucerne, Frohburgstrasse 3, 6002 Luzern, Switzerland. E-mail: [email protected] Abstract The increasing prevalence of chronic conditions is a significant challenge for healthcare systems worldwide, not only from a public health perspective but also for the aggregate cost that these represent. This paper estimates the additional use of healthcare services due to chronic health conditions and their associated costs in nine European countries. We analyzed inpatient and outpatient healthcare utilization using longitudinal data (Survey of Health, Ageing and Retirement in Europe [SHARE]). We implemented a difference-in-differences approach across multiple time periods. Monetary estimates were derived using WHO-CHOICE healthcare service costs. To compare countries, we calculated the healthcare cost burden of chronic conditions as a percentage of total health expenditure. People with chronic conditions require significantly more healthcare services than those without such conditions, averaging three additional outpatient visits and one extra overnight inpatient stay annually. These patterns vary across countries. In Germany, outpatient care usage is particularly high, with an average of four additional visits, while Switzerland leads in inpatient care with two extra overnight stays. The associated costs also differ widely, influenced by variations in healthcare demand, service pricing, and the prevalence of chronic conditions in each country. Chronic conditions significantly increase healthcare utilization, and demographic trends suggest this demand will continue to grow steadily. This rising pressure poses serious challenges for healthcare systems, necessitating a shift toward more efficient service delivery models. ............................................................................................................... Introduction The prevalence of chronic conditions is a challenge for healthcare systems. Since 1990, the burden of non-communicable diseases has risen by 60% [1], meaning more people will face long-term health conditions. Healthcare systems must expand services to meet these needs, adding pressure to a system already consuming about 10% of the gross domestic product (GDP) in industrialized economies [2]. Without proper planning, rising healthcare costs could become a significant economic burden for individuals, families, and the healthcare system [3]. Understanding healthcare expenditure is key to managing budgets and designing cost-effective interventions. Total healthcare expenditures are driven by the price and utilization of services. While the price depends on a country’s wealth [4], new technologies [5], and the supply of health services, the utilization of healthcare services is influenced by demographic dynamics, service providers, disease burden, and care preferences [6]. The latter is being pressured by the rise in non-communicable diseases, especially among older populations, which implies more frequent care for extended periods [7]. Related studies have examined the rising cost of specific conditions like chronic pain [8], cancer [9], and diabetes[10], but there are no estimates on the overall impact of chronic conditions on healthcare utilization. This paper estimated how much healthcare utilization changed after a person was diagnosed with a chronic condition and the corresponding costs. Using longitudinal data from nine European countries, we applied a difference-in-differences (DiD) approach to estimate the additional utilization due to new diagnoses. We focused on inpatient and outpatient care, as these two consume the largest portions of healthcare expenditure [11–13]. Our findings offer insights into population healthcare needs and the cost implications for decision-making on service provision to manage the economic impact of chronic conditions. Methods Data sources This study used three data sources: (i) the Survey of Health, Ageing and Retirement in Europe (SHARE) [14] to estimate the healthcare service utilization of people with chronic health conditions compared to those with similar characteristics in the general population, (ii) the WHO-CHOICE [15] to estimate the cost associated with the excess use of health services, and (iii) the World Bank and WHO Global Health Expenditure Database [16] to estimate the healthcare costs. SHARE includes data on individuals aged 50þfrom 28 European countries [14], covering health, socioeconomic status, and wellbeing. The survey collected regular data in multiple years since 2004, 2006/2007, 2008/2009, 2011/2012, 2013, 2015, 2017, 2019/ 2020, and 2021/2022. Data was collected face-to-face with physical and biomarker tests, switching to phone interviews for end-of-life cases and during COVID-19. The survey targets people living in Downloaded from https://academic.oup.com/eurpub/article/35/2/216/8009071 by ZHB Zentralund Hochschulbibliothek Luzern user on 25 November 2025 private households and excludes those incarcerated, hospitalized, out of the country, unable to speak a local language, or moved to an unknown address. All respondents are included in the longitudinal sample. SHARE aims to represent each country’s population, though representativeness may vary for some cohorts, especially among older groups, due to response rates and sampling differences [17]. We analyzed information on healthcare service utilization (outpatient and inpatient) from the survey across eight waves. Since not all countries participated in every wave, we focused on the countries with consistent data across waves, enabling panel data analysis: Austria, Belgium, Denmark, France, Germany, Italy, Spain, Sweden, and Switzerland. The WHO-CHOICE program, developed by the World Health Organization, provides evidence on the unit costs of inpatient and outpatient care at the country level at purchase power parity (PPP) [15]. For comparison purposes, we adjusted the cost estimates for inflation in our calculations [18]. Data analysis Part 1: Excess frequency in healthcare service utilization To estimate the excess in healthcare services utilization, we considered Population A, or the treated group, diagnosed with a chronic condition in any of the T periods, where T¼1;2;3;...;8 and Population B, or the control group, which has all people who never reported having a chronic condition. Once diagnosed, we expected the treated group to increase their utilization of healthcare services compared to the control group, which we estimated using a DiD with multiple time periods, applying the potential outcomes framework [19]. In formal terms, the DiD is defined as follows: •T¼8, the number of time periods that individuals in the sample were observed. •D it is the treatment indicator equals 1 if individual i was diagnosed with a chronic condition in time period t, and 0 otherwise. •G i is the first time period where individual i reported being diagnosed with a chronic condition, and 0 otherwise. •Y it is the utilization of healthcare services of individual i at time period t, where Y is measured by: 1. Outpatient care: reported number of visits to a doctor in the last 12 months. 2. Inpatient care: number of overnight stays in a hospital in the last 12 months. •Y it (g) is the treated potential outcome, which is the healthcare service utilization that individual i would experience at time period t if they were diagnosed with a chronic condition in period g. •Y it ¼Y it (0) is the untreated potential outcome, which is the healthcare service utilization that individual i would experience in time period t if i has never been diagnosed with a chronic condition. •For all individuals iand time periods t<Gi;we have that Yit ¼Yit 0 ð Þ ðno anticipationÞand Yit ¼Yit Gi ð Þ when t≥Gi. Following this approach, the unit-level treatment effect is given by the difference in healthcare service utilization between treated and control groups: τit ¼Yit g ð ÞYit 0 ð Þ The unit-level average treatment effect on the treated (ATT) is given by: τit (g ð Þ ¼1 Tgþ1X T t¼Gi Yit g ð ÞYit 0 ð Þ � � The sample ATT is given by: ATT g;t ð Þ ¼E Ytg ð ÞYt0 ð ÞjG¼g h i ATT is the average treatment effect on the treated for the group g in the time period t, which serves as the building block for other aggregated treatment effects with varying weight functions w g;t ð Þ. These weights depend on the group size among all groups that ever participated in the treatment and the number of post-treatment time periods for a particular group. These parameters are of the following form: θ¼X g2GX T t¼2 w g;t ð ÞATT g;t ð Þ The event study estimator can be constructed by the group-time average treatment effects as follows: ATTES e ð Þ ¼X g2G wES g;e ð ÞATT g;gþe ð Þ We considered three parameters: the overall average treatment effect, group-time average treatment effects, and the event study estimator. The DiD approach relies on the parallel trend assumption, matching treated and control groups based on the following characteristics: age, gender, country, household size, civil status, selfperceived health, financial status, and education level before treatment. To ensure exchangeability, we used inverse probability weighting and calculated standard errors with the multiplier bootstrap method (1000 iterations). The analysis was conducted in R (version 4.2.2) using the did package (version 2.1.2). The Supplementary Appendix includes a sensitivity analysis to assess the robustness of our estimates. Part 2: Cost of the excess frequency We used WHO-CHOICE data to estimate healthcare service costs, adjusted for inflation using the Harmonized Index of Consumer Prices (HICP) for healthcare services [18, 20]. The WHOCHOICE reports the estimated average cost per visit, excluding medication, and is indicated in PPP to allow comparison across countries. To account for the variation and compute the standard deviation of the cost estimates, we implemented a two-part model for frequency [21]. As healthcare costs vary by provider, we analyzed different cost scenarios: •Health centers exclusively for outpatient services (no beds). •Health centers with beds. •Primary-level hospitals that treat simple cases and have few specialties. •Secondary-level hospitals that treat referral cases (specialized hospitals). •Tertiary-level hospitals with highly specialized staff and technical equipment. For inpatient care, we used only the last three categories. Cross-country comparison: healthcare cost burden To compare our results across countries, we calculated the healthcare cost burden (HcB). HcB was calculated using the excess service utilization (ATT), estimated in Part I, and the unit cost estimates from Part II. To get aggregate estimates, we used the potential total demand for healthcare services by people with chronic conditions using the prevalence data from the Rehabilitation Needs Estimator project [22, 23]. Formally, HcB is the share of the total healthcare expenditure a country incurs for services utilization of persons with chronic conditions: HcBcountry ¼#People with chronic conditionscountry ×ATTcountry ×Pricecountry Total health expenditurecountry In the supplement, we incorporated the HcB results using incidence estimates of chronic conditions in each country. However, as incidence estimates are inexistent for all chronic conditions, we Excess healthcare utilization and chronic conditions costs 217 Downloaded from https://academic.oup.com/eurpub/article/35/2/216/8009071 by ZHB Zentralund Hochschulbibliothek Luzern user on 25 November 2025 estimated incidence rates in our sample, which serve only as a reference, as we lack mortality data. Results Descriptive statistics Table 1 displays the characteristics of the sample (eight waves) across the nine countries. Belgium has the largest sample with 34 722 observations (14.5%), followed by Spain with 29 253 (12.2%), Germany with 29 084 (12.2%), Italy with 29 031 (12.2%), and Switzerland contributing 18 116 observations (7.6%) to the overall dataset. The proportion of male individuals in the sample ranged from 41.9% in Austria to 47.1% in Germany, with an average of 45% across all countries. The average age ranged from 66.4 in Denmark and Sweden to 69.8 years. Self-perceived health varied across the sample. Spain had the highest proportion of people reporting “poor” health (15%), while Denmark and Sweden had the most reporting “excellent” health (19.4% and 15.9%, respectively). In contrast, Spain had the least people reporting “excellent” health (3.7%). On average, households had two members across all countries. Partnership status also varied, with Italy having the highest percentage of individuals living with a partner (77.2%) and Austria the lowest (64%). Education levels differed significantly, with Denmark having the highest proportion of individuals with tertiary education (42.5%), while Italy had the lowest (7.7%). Part 1: Excess frequency of health care services utilization The event study plots (Fig. 1A and B) display the estimated ATT of chronic conditions on the utilization of healthcare services. A table with the estimates is reported in Supplementary Table S1. Before diagnosis, treated and control groups show no significant differences in healthcare utilization, confirming their comparability. In contrast, post-diagnosis, outpatient, and inpatient care rises in all countries with varying effect sizes. For outpatient care (Fig. 1A), Germany exhibited the highest excess utilization, with an ATT of 3.96 more doctor visits (95% CI: 1.49–6.43). Belgium and Austria followed closely, with ATT values of 3.93 (95% CI: 2.62–5.24) and 3.88 (95% CI: 2.24–5.52) more visits, respectively. Italy and Spain also showed notable effects, with 3.37 (95% CI: 2.53–4.21) and 3.32 (95% CI: 2.39–4.25) more visits, respectively. Switzerland and France followed, with ATT of 3.31 (95% CI: 1.59–5.02) and 3.09 (95% CI: 2.35–3.82) more visits, respectively. Denmark and Sweden had the lowest estimates, with ATT values of 2.56 (95% CI: 1.83–3.29) and 1.46 (95% CI: 0.90–2.03) more visits, respectively. Notably, the excess utilization of outpatient care was statistically significant for all countries. For inpatient care (Fig. 1B), Switzerland showed the highest excess utilization, with an ATT of 2.29 more overnight stays for individuals diagnosed with a chronic condition (95% CI: 0.70–3.88). Austria followed with an ATT of 1.94 more stays (95% CI: 0.42– 3.46). Belgium and Germany reported moderate effects, with ATT of 0.93 (95% CI: 0.48–1.38) and 0.96 (95% CI: −0.03 to 1.95), respectively. France showed a similar ATT of 0.89 more stays (95% CI: 0.39–1.40). Smaller effects were observed for Spain (0.58, 95% CI: 0.22–0.93) and Italy (0.52, 95% CI: 0.19–0.85). Sweden and Denmark presented the lowest effects, with ATT of 0.42 (95% CI: 0.08–0.76) and 0.33 (95% CI: −0.17 to 0.83) more visits, respectively. Nevertheless, Denmark and Germany’s confidence interval included zero, indicating non-significant effects. Part 2: Cost of services Tables 2 and 3 present the estimated unit costs of healthcare services and their variation (standard deviation) across service providers in different countries. As expected, outpatient care costs are substantially lower than inpatient care costs. On average, the highest unit costs for outpatient and inpatient services are observed in Austria, Switzerland, and Denmark. In contrast, Italy, Spain, and France consistently exhibit lower costs across all providers. The other countries demonstrate comparable cost levels. Part 3: Healthcare cost burden Figure 2A (outpatient) and B (inpatient) display the estimated HcB in the analyzed countries. The x-axis displays the excess in healthcare utilization (ATT), and the y-axis shows the cost per visit. Interestingly, even when prices are in PPP terms, some variation remains, especially in inpatient care, which suggests that services are not homogeneous across countries. The size of the bubbles reflects the prevalence of chronic conditions. For outpatient care, the results indicate that Italy had the highest HcB, ranging from 2% in health centers to 2.9% in secondary/tertiary hospitals. This is primarily attributed to the country’s high prevalence of chronic conditions. Spain follows, with a burden ranging from 1.6% to 2.3%. Despite having similar costs and utilization levels to Italy, Spain’s burden is lower due to a smaller prevalence of chronic conditions. Belgium (1.3%–1.9%) and Germany (1.2%– 1.8%) exhibit burdens driven by higher utilization of outpatient services combined with a high prevalence of chronic conditions. Austria shows a similar burden (1.3%–1.8%), explained by elevated healthcare utilization and high service prices. France reports a relatively lower burden (0.9%–1.2%), primarily due to lower healthcare prices and less excessive utilization of outpatient services. At the lower end of the spectrum are Sweden (0.4%–0.6%), Switzerland (0.6%–0.9%), and Denmark (0.7%–1%). Despite having relatively similar healthcare prices, these countries reported a smaller prevalence of chronic conditions. Sweden, in particular, stands out with the smallest burden in the sample. For inpatient care, the highest HcBs were reported in Austria (9.1%– 12.8%) and Switzerland (6.4%–8.6%), driven by relatively higher prices and greater healthcare utilization. Belgium (4.5%–5.9%) and Germany (4.2%–5.5%) followed, with lower prices and less utilization overall. In Germany, however, the high prevalence of chronic conditions accounts for a significant share of the estimated HcB. Italy (4.2%–5.6%), Spain (3.7%–4.9%), and France (3.3%–4.5%) showed a lower burden, mainly due to reduced prices and utilization. Notably, Italy stands out for its higher prevalence of chronic conditions despite its lower overall costs. Finally, despite high healthcare prices, Sweden (1.5%–2.03%) and Denmark (1.2%– 1.61%) had the smallest HcB in the sample. In both countries, this is attributed to reduced utilization of healthcare services and a lower prevalence of chronic conditions, particularly in Sweden. Discussion This study estimated the increased utilization of healthcare services and the associated costs resulting from chronic conditions across nine European countries. For outpatient care, individuals diagnosed with a chronic condition experience an average of four additional visits per year in Germany, Belgium, and Austria. In Italy, Spain, Switzerland, France, and Denmark, this figure drops to 3, while Sweden sees just one additional visit annually. For inpatient care, Switzerland and Austria report an average of two extra overnight hospital stays per year, followed by Germany, Belgium, and France with one additional stay. In contrast, Spain, Italy, and Sweden record less than 1 extra overnight stay annually, while Denmark shows no significant effect. These findings highlight a substantial financial burden on healthcare systems. Italy, Belgium, Spain, Austria, and Germany face the highest costs for outpatient services, while Switzerland and Austria incur the most significant expenses for inpatient care. Our findings align with related literature highlighting the steady rise in healthcare spending driven by non-communicable diseases [24]. While much of the existing research focuses on total healthcare 218 Polanco et al. Downloaded from https://academic.oup.com/eurpub/article/35/2/216/8009071 by ZHB Zentralund Hochschulbibliothek Luzern user on 25 November 2025 Table 1. Descriptive statistics of the sample Variable Austria Belgium Denmark France Germany Italy Spain Sweden Switzerland Sample size (N) 23 623 (9.9%) 34 722 (14.5%) 21 962 (9.2%) 28 917 (12.1%) 29 084 (12.2%) 29 031 (12.2%) 29 253 (12.2%) 24 201 (10.1%) 18 116 (7.6%) Gender Male (%) 9891 (41.9%) 15 690 (45.2%) 10 160 (46.3%) 12 538 (43.4%) 13 712 (47.1%) 13 108 (45.2%) 13 092 (44.8%) 11 210 (46.3%) 8260 (45.6%) Female (%) 13 732 (58.1%) 19 032 (54.8%) 11 802 (53.7%) 16 379 (56.6%) 15 372 (52.9%) 15 923 (54.8%) 16 161 (55.2%) 12 991 (53.7%) 9856 (54.4%) Age (mean) 68.5 66.8 66.4 67.8 66.8 68.1 69.7 69.8 68.1 Self-perceived health status Excellent (%) 1814 (7.7%) 2528 (7.3%) 4254 (19.4%) 1878 (6.5%) 1 461 (5%) 1 809 (6.2%) 1066 (3.7%) 3841 (15.9%) 1996 (11%) Very good (%) 5421 (23%) 7085 (20.5%) 7197 (32.8%) 4097 (14.3%) 4346 (15%) 4054 (14%) 4277 (14.7%) 5651 (23.4%) 5167 (28.6%) Good (%) 8433 (35.8%) 14 887 (43%) 5243 (23.9%) 12 306 (42.9%) 11 639 (40.1%) 10 530 (36.3%) 10 875 (37.3%) 8298 (34.4%) 7498 (41.5%) Fair (%) 5999 (25.5%) 7923 (22.9%) 3957 (18.1%) 7324 (25.5%) 8785 (30.3%) 9345 (32.3%) 8565 (29.4%) 4929 (20.4%) 2760 (15.3%) Poor (%) 1879 (8%) 2217 (6.4%) 1263 (5.8%) 3089 (10.8%) 2786 (9.6%) 3237 (11.2%) 4369 (15%) 1426 (5.9%) 647 (3.6%) Income group Low (%) 6345 (30.3%) 9510 (30.2%) 6135 (30.6%) 8133 (30.4%) 7887 (30.2%) 7901 (30.3%) 7825 (30.3%) 6670 (30.2%) 4986 (30.3%) Medium (%) 8389 (40%) 12 600 (40.1%) 8036 (40.1%) 10 720 (40.1%) 10 487 (40.1%) 10 406 (39.9%) 10 370 (40.1%) 8834 (40%) 6580 (39.9%) High (%) 6215 (29.7%) 9332 (29.7%) 5863 (29.3%) 7912 (29.6%) 7759 (29.7%) 7784 (29.8%) 7671 (29.7%) 6583 (29.8%) 4912 (29.8%) Household size Small (%) 20 154 (85.3%) 28 294 (81.5%) 19 533 (88.9%) 24 555 (84.9%) 24 793 (85.2%) 18 719 (64.5%) 19 590 (67%) 22 447 (92.8%) 15 562 (85.9%) Medium (%) 2244 (9.5%) 3944 (11.4%) 1540 (7%) 2629 (9.1%) 2937 (10.1%) 6089 (21%) 5647 (19.3%) 1222 (5%) 1476 (8.1%) Large (%) 208 (0.9%) 186 (0.5%) 47 (0.2%) 188 (0.7%) 94 (0.3%) 310 (1.1%) 395 (1.4%) 24 (0.1%) 67 (0.4%) In partnership (%) 15 118 (64%) 23 855 (68.7%) 16 108 (73.3%) 19 246 (66.6%) 22 234 (76.4%) 22 414 (77.2%) 21 774 (74.4%) 17 606 (72.7%) 13 020 (71.9%) Education Primary (%) 5675 (24.3%) 13 847 (40.3%) 4003 (18.3%) 12 076 (42.6%) 3486 (12.1%) 20 269 (70.4%) 22 874 (80.1%) 9179 (38.7%) 3954 (22.2%) Secondary (%) 11 594 (49.7%) 9331 (27.2%) 8543 (39.2%) 9915 (34.9%) 16 266 (56.3%) 6315 (21.9%) 2829 (9.9%) 7470 (31.5%) 10 822 (60.7%) Tertiary (%) 6048 (25.9%) 11 157 (32.5%) 9271 (42.5%) 6383 (22.5%) 9121 (31.6%) 2207 (7.7%) 2839 (9.9%) 7059 (29.8%) 3042 (17.1%) Numbers in frequencies and percentages in parentheses. Excess healthcare utilization and chronic conditions costs 219 Downloaded from https://academic.oup.com/eurpub/article/35/2/216/8009071 by ZHB Zentralund Hochschulbibliothek Luzern user on 25 November 2025 expenditures—including costs for medications, prevention, assistive devices, infrastructure, and administration [10, 25, 26]—our study provides causal estimates specifically related to increased healthcare service utilization following chronic condition diagnoses. Consequently, our cost estimates are smaller than those in the broader literature, as they represent only a subset of total expenditures for non-communicable diseases. A comparable study estimated that inpatient service utilization for some chronic conditions in Switzerland contributed between 1.3% and 6.2% of the total variation of costs, which is closer to our findings [12]. Given that healthcare represents one of the largest categories of public expenditure, the growing demand for care poses significant Figure 1. (A) Outpatient care: estimated effect of being diagnosed with a chronic condition on service utilization. (B) Inpatient care: estimated effect of being diagnosed with a chronic condition on service utilization. ATT is the average estimated excess utilization of healthcare services by persons diagnosed with a chronic condition compared to a control group. 220 Polanco et al. Downloaded from https://academic.oup.com/eurpub/article/35/2/216/8009071 by ZHB Zentralund Hochschulbibliothek Luzern user on 25 November 2025 challenges to public finances [8]. Thus, to address the increasing care needs of the population effectively, it is essential to understand the implications for healthcare system organization. These are shaped by several key factors: (1) the cost of healthcare services, (2) the patterns of service utilization, and (3) the overall health status of the population. In outpatient care, our findings suggest that the primary driver of the HcB is the prevalence of chronic conditions. While some variations in prices and utilization were observed, there are no marked differences across countries except for Sweden. Sweden stands out with the smallest burden in the sample, attributed to its comparatively low service utilization and prevalence of chronic conditions. In contrast, inpatient care revealed more pronounced variations across countries. Even after adjusting for price differences, notable disparities persist, suggesting marked differences in how inpatient services are organized and delivered across nations. Switzerland and Austria exhibited the highest burdens, driven by relatively higher Figure 1. Continued. Excess healthcare utilization and chronic conditions costs 221 Downloaded from https://academic.oup.com/eurpub/article/35/2/216/8009071 by ZHB Zentralund Hochschulbibliothek Luzern user on 25 November 2025 prices for inpatient care and elevated levels of healthcare utilization. Belgium, Germany, Italy, Spain, and France showed a mid-range burden, primarily explained by the high prevalence of chronic conditions despite lower prices and healthcare utilization levels. Like outpatient care, Sweden and Denmark experienced the lowest burden due to their reduced healthcare utilization and lower prevalence of chronic conditions. Our findings highlight critical areas for improving the organization of healthcare services to address the growing care needs of the population. First, healthcare prices are highly complex. Our results demonstrate that even after price adjustments, differences persist, particularly in inpatient care. This reflects the heterogeneity of healthcare services across countries [27]. These differences suggest that factors beyond inflation, such as technological advancements, population preferences, or structural composition effects, may contribute to higher costs. For instance, Switzerland and Austria exhibit relatively more expensive services, likely influenced by these exogenous factors [27]. To mitigate financial pressures, alternative models that minimize reliance on costly and potentially unnecessary services should be explored. Many chronic conditions, particularly in their Table 2. Outpatient care: estimated costs by type of facility [numbers in USD (PPP)] WHO-CHOICE Estimates Cost estimates due to excess utilization of healthcare services A B C B × C D Country Type of facility Mean estimated unit costs (USD per visit) SD UB estimate LB estimate Mean individual costs (USD per person over a year) UB estimate LB estimate Number of people with chronic conditions 50+ yo. Total estimated costs (in USD millions) Healthcare cost burden (HcB) Austria Health centre (no beds) 58.32 42.87 174.63 12.36 226.28 963.96 27.69 3 338 346 755.41 1.3% Secondary hospital 82.83 64.15 257.83 16.71 321.38 1423.22 37.43 1072.88 1.8% Tertiary hospital 82.42 64.36 238.98 16.22 319.79 1319.17 36.33 1067.57 1.8% Primary hospital 80.12 58.99 242.18 15.84 310.87 1336.83 35.48 1037.78 1.8% Health centre (with beds) 70.89 55.33 219.45 14.55 275.05 1211.36 32.59 918.22 1.6% Belgium Tertiary hospital 70.55 51.74 206.02 15.56 277.26 1079.54 40.77 4 507 303 1249.70 1.9% Secondary hospital 68.42 48.57 205.44 14.96 268.89 1076.51 39.20 1211.97 1.8% Health centre (with beds) 61.98 48.65 192.66 12.95 243.58 1009.54 33.93 1097.90 1.7% Primary hospital 66.91 51.82 202.92 13.64 262.96 1063.30 35.74 1185.22 1.8% Health centre (no beds) 49.20 37.68 150.00 9.84 193.36 786.00 25.78 871.51 1.3% Denmark Tertiary hospital 76.45 60.41 227.76 16.33 195.71 749.33 29.88 2 167 188 424.14 1.0% Health centre (with beds) 63.73 51.08 202.33 14.38 163.15 665.67 26.32 353.57 0.8% Secondary hospital 76.65 59.04 240.83 14.99 196.22 792.33 27.43 425.25 1.0% Primary hospital 52.48 57.96 225.49 14.76 134.35 741.86 27.01 291.16 0.7% Health centre (no beds) 52.48 38.15 153.63 10.86 134.35 505.44 19.87 291.16 0.7% France Secondary hospital 61.58 49.16 182.38 12.63 190.28 696.69 29.68 24 406 554 4644.13 1.3% Health centre (no beds) 42.80 34.93 147.43 8.72 132.25 563.18 20.49 3227.82 0.9% Primary hospital 42.80 41.63 171.99 11.87 132.25 657.00 27.89 3227.82 0.9% Tertiary hospital 58.64 46.38 179.62 11.86 181.20 686.15 27.87 4422.41 1.2% Health centre (with beds) 52.12 42.92 164.40 9.98 161.05 628.01 23.45 3930.70 1.1% Germany Secondary hospital 70.68 53.72 212.84 14.72 279.89 1368.56 21.93 35 636 697 9974.45 1.8% Tertiary hospital 70.11 53.55 212.73 13.48 277.64 1367.85 20.09 9894.02 1.8% Primary hospital 48.31 50.18 207.39 13.56 191.31 1333.52 20.20 6817.57 1.2% Health centre (with beds) 57.40 41.17 171.71 12.15 227.30 1104.10 18.10 8100.36 1.5% Health centre (no beds) 48.31 36.88 148.35 9.92 191.31 953.89 14.78 6817.57 1.2% Italy Tertiary hospital 65.58 50.87 204.46 12.71 221.00 860.78 32.16 26 297 584 5811.89 2.9% Secondary hospital 63.27 49.47 185.65 13.56 213.22 781.59 34.31 5607.17 2.8% Primary hospital 44.23 47.68 193.96 13.13 149.06 816.57 33.22 3919.79 2.0% Health centre (no beds) 44.23 32.95 132.48 8.83 149.06 557.74 22.34 3919.79 2.0% Health centre (with beds) 57.44 48.04 179.74 11.03 193.57 756.71 27.91 5090.50 2.6% Spain Tertiary hospital 60.66 44.59 182.80 12.84 201.39 776.90 30.69 17 497 926 3523.93 2.3% Primary hospital 42.60 47.57 180.72 12.05 141.43 768.06 28.80 2474.77 1.6% Health centre (with beds) 52.20 39.01 150.26 11.29 173.30 638.61 26.98 3032.46 2.0% Secondary hospital 62.62 48.80 192.26 13.35 207.90 817.11 31.91 3637.79 2.4% Health centre (no beds) 42.60 33.91 129.78 8.19 141.43 551.57 19.57 2474.77 1.6% Sweden Tertiary hospital 75.23 59.57 236.49 14.77 109.84 480.07 13.29 3 680 911 404.30 0.6% Primary hospital 49.49 57.24 217.77 14.80 72.26 442.07 13.32 265.97 0.4% Secondary hospital 74.80 57.14 234.67 14.52 109.21 476.38 13.07 401.98 0.6% Health centre (with beds) 65.93 52.18 197.76 12.15 96.26 401.45 10.94 354.32 0.5% Health centre (no beds) 49.49 37.58 146.40 10.51 72.26 297.19 9.46 265.97 0.4% Switzerland Secondary hospital 74.94 58.52 228.71 16.21 248.05 1148.12 25.77 3 277 658 813.03 0.9% Primary hospital 51.69 56.34 231.47 13.92 171.09 1161.98 22.13 560.79 0.6% Health centre (no beds) 51.69 38.21 154.99 10.83 171.09 778.05 17.22 560.79 0.6% Health centre (with beds) 64.97 52.27 199.97 13.57 215.05 1003.85 21.58 704.86 0.7% Tertiary hospital 76.53 60.55 237.37 15.55 253.31 1191.60 24.72 830.28 0.9% The values are in 2021 USD, adjusted for healthcare inflation. UB and LB are the 95% confidence intervals around the estimated costs. The UB and LB indicate the range of uncertainty in the estimated costs. SD ¼standard deviation; yo. ¼years old; UB ¼upper bound estimate; LB ¼lower bound estimate. A: Unit cost estimates (per visit to a healthcare provider). Data comes from the WHO-CHOICE project. B: Mean individual costs are the unit cost of a visit to inpatient care times the estimated excess utilization (ATT) of healthcare services. C: Numbers calculated from the prevalence of chronic conditions estimated by the Rehabilitation Needs Estimator from the Institute of Health Metrics (IHME). Total estimated costs are the associated costs due to the excess utilization of healthcare services among people with chronic conditions. D: HcB is the proportion of excess healthcare utilization costs relative to a country’s total healthcare expenditure. 222 Polanco et al. Downloaded from https://academic.oup.com/eurpub/article/35/2/216/8009071 by ZHB Zentralund Hochschulbibliothek Luzern user on 25 November 2025 early stages, can be managed effectively and at a significantly lower cost within primary care settings, rather than through direct inpatient treatment [28, 29]. However, expanding primary care services presents challenges, particularly in Europe, where healthcare systems are already strained by workforce shortages [30]. Second, the unnecessary utilization of healthcare services is more prevalent in systems lacking gatekeeping structures or those that do not require copayments. Some healthcare systems provide broad access to services while incorporating referral systems for specialist care as a mechanism to control overuse [31]. This approach may account for the findings in Spain and France, where excess utilization is relatively low. However, potential trade-offs must be carefully evaluated, particularly when considering increased out-of-pocket payments. Such measures can disproportionately affect lowerincome households, potentially exacerbating health disparities and leading to pervasive negative effects on overall health outcomes [32]. Third—and perhaps most critical—is addressing the rising prevalence of chronic conditions. Effective public health interventions must prioritize two key strategies: prevention and rehabilitation. Enhancing preventive measures, such as promoting healthy nutrition and regular physical activity from an early age, represents one of the most cost-effective approaches to managing non-communicable diseases [33]. These measures can significantly reduce the incidence of chronic conditions and alleviate long-term healthcare burdens. Equally important is integrating rehabilitation services into standard care for individuals already living with chronic conditions. Rehabilitation not only aids patients in regaining independence and improving their quality of life but also helps prevent complications that could lead to costly hospital admissions [34]. Finally, it is important to acknowledge the limitations of this study. A key assumption in our model is that once an individual is diagnosed with a chronic condition, their healthcare utilization increases because of need. However, some of the observed increase may also be supply-driven, resulting from overdiagnosis—for instance, identifying asymptomatic conditions during routine screenings [35]. Future research should focus on accurately identifying overdiagnosis and assessing its impact on healthcare costs. Advanced risk models could play a pivotal role in this area by analyzing cost data to identify cost drivers and facilitate simulations for health economic evaluations [36]. Another limitation is that our estimates reflect changes in healthcare utilization following a diagnosis, meaning the total effect pertains only to new cases (i.e. the incidence of chronic conditions). Since incidence data is currently unavailable, we calculated total effects using prevalence data, thereby capturing the impact on all individuals with chronic conditions. As a reference, we re-estimated our results in the Supplementary section by computing incidence within our sample, with some simplifications. Table 3. Inpatient care: estimated costs by type of facility [numbers in USD (PPP)] WHO-CHOICE Estimates Cost estimates due to excess utilization of healthcare services A B C B × C D Country Type of facility Mean estimated unit costs (USD per overnight stay) SD UB estimate LB estimate Mean individual costs (USD per person over a year) UB estimate LB estimate Number of people with chronic conditions 50+ yo. Total estimated costs (in USD millions) Healthcare cost burden (HcB) Austria Primary hospital 820.4 396.4 1854.6 300.3 1591.6 6417.0 126.1 33 383 46 5313.4 9.1% Secondary hospital 863.1 395.0 1857.5 339.1 1674.5 6427.0 142.4 5590.0 9.6% Tertiary hospital 1144.4 537.7 2487.9 410.5 2220.2 8608.2 172.4 7411.8 12.8% Belgium Primary hospital 926.9 417.3 2042.7 352.9 862.0 2818.9 169.4 45 073 03 3885.2 5.9% Secondary hospital 698.0 314.6 1446.0 274.3 649.1 1995.5 131.7 2925.9 4.5% Tertiary hospital 707.8 307.1 1466.5 280.4 658.3 2023.8 134.6 2967.0 4.5% Denmark Primary hospital 1006.3 460.0 2141.0 393.4 332.1 1777.1 . 21 671 88 719.7 1.7% Secondary hospital 788.9 375.5 1731.8 311.1 260.3 1437.4 . 564.2 1.3% Tertiary hospital 754.2 346.0 1677.8 284.8 248.9 1392.5 . 539.4 1.3% France Primary hospital 546.7 267.5 1196.6 211.5 486.6 1675.3 82.5 244 065 54 11 876.0 3.3% Secondary hospital 590.8 271.6 1292.8 227.8 525.8 1809.9 88.8 12 832.6 3.5% Tertiary hospital 758.5 347.8 1693.4 302.5 675.1 2370.8 118.0 16 476.7 4.5% Germany Primary hospital 709.9 327.7 1545.0 268.0 681.5 3012.7 . 35 6366 97 24 287.9 4.4% Secondary hospital 890.9 422.1 1979.1 329.5 855.2 3859.2 . 30 477.8 5.5% Tertiary hospital 671.4 308.4 1475.1 265.8 644.5 2876.4 . 22 968.1 4.2% Italy Primary hospital 819.4 375.5 1730.1 306.7 426.1 1470.6 58.3 26 2975 84 11 204.5 5.6% Secondary hospital 611.9 281.2 1342.0 220.9 318.2 1140.7 42.0 8368.1 4.2% Tertiary hospital 621.5 293.4 1369.5 227.0 323.2 1164.1 43.1 8498.2 4.3% Spain Primary hospital 738.5 325.3 1511.5 295.3 428.3 1405.7 65.0 17 4979 26 7494.5 4.9% Secondary hospital 603.0 278.9 1317.0 235.4 349.7 1224.8 51.8 6119.7 4.0% Tertiary hospital 551.7 254.4 1182.9 206.7 320.0 1100.1 45.5 5599.0 3.7% Sweden Primary hospital 744.9 331.5 1542.5 291.5 312.9 1172.3 23.3 36 809 11 1151.6 1.6% Secondary hospital 991.8 474.1 2184.3 373.6 416.6 1660.0 29.9 1533.3 2.1% Tertiary hospital 732.2 330.3 1622.0 288.9 307.5 1232.7 23.1 1131.9 1.6% Switzerland Primary hospital 799.5 370.5 1793.8 294.5 1830.7 6960.1 206.1 32 776 58 6000.5 6.4% Secondary hospital 835.4 398.6 1871.5 311.7 1913.1 7261.5 218.2 6270.4 6.6% Tertiary hospital 1080.7 503.0 2370.0 408.3 2474.8 9195.6 285.8 8111.7 8.6% The values are in 2021 USD, adjusted for healthcare inflation. UB and LB are the 95% confidence intervals around the estimated costs. The UB and LB indicate the range of uncertainty in the estimated costs. SD ¼standard deviation; yo. ¼years old; UB ¼upper bound estimate; LB ¼lower bound estimate. A: Unit cost estimates (per visit to a healthcare provider). Data comes from the WHO-CHOICE project. B: Mean individual costs are the unit cost of a visit to inpatient care times the estimated excess utilization (ATT) of healthcare services. C: Numbers calculated from the prevalence of chronic conditions estimated by Rehabilitation Needs Estimator from the Institute of Health Metrics (IHME). Total estimated costs are the associated costs due to the excess utilization of healthcare services among people with chronic conditions. D: HcB is the proportion of excess healthcare utilization costs relative to a country’s total healthcare expenditure. Excess healthcare utilization and chronic conditions costs 223 Downloaded from https://academic.oup.com/eurpub/article/35/2/216/8009071 by ZHB Zentralund Hochschulbibliothek Luzern user on 25 November 2025 Conclusion This study estimates how much people with chronic conditions increase their utilization of inpatient and outpatient healthcare services. This excess utilization imposes a substantial economic burden on healthcare systems, with important variations across countries. These differences can be attributed to each country’s distinct organization of healthcare systems, social structures, Figure 2. (A) Outpatient care: estimated costs by type of facility [numbers in USD (PPP)]. (B) Inpatient care: estimated costs by type of facility [numbers in USD (PPP)]. The numbers reflect the healthcare cost burden (HcB), defined as the proportion of excess healthcare utilization costs relative to a country’s total healthcare expenditure. The size of the bubbles indicates the prevalence of chronic conditions. 224 Polanco et al. Downloaded from https://academic.oup.com/eurpub/article/35/2/216/8009071 by ZHB Zentralund Hochschulbibliothek Luzern user on 25 November 2025