Assessment of the introduction of DRG-based reimbursement in Switzerland: Evidence on the short-term effects on length of stay compliance in university hospitals
Introduction
In 2012, Switzerland implemented a nationwide diagnosis-related groups (DRG) hospital reimbursement system. The main goals of the SwissDRG system were to increase transparency and efficiency in the hospital sector. Policy-makers also aimed at slowing down the growth of health care expenditures, fostering competition between hospitals, increasing the quality of inpatient care, and improving access [1], [2].
The Swiss health system with its decentralized structure offers an interesting setup to study the impacts of a national DRG system. First, Swiss cantons play a multifaceted role, as planners, regulators, payers, and owners of public hospitals, which inevitably creates conflicts of interest and may be at odds with a national payment scheme [3] Second, there is ample evidence for regional disparities in health care provision (e.g., [4]), which, among other things, translates into specific requirements for a payment system. Third, already since 2001 some hospitals used prospective payments based on the all-patient DRG (APDRG) classification. The APDRG system has been identified as a potential driving force behind the reduction in the length of stay (LOS), although limited evidence exists that would establish a causal link (e.g., [5], [6], [7]).
In this study, we explore the coexistence of hospitals with an APDRG payment system and hospitals with a retrospective per diem reimbursement scheme before 2012 to evaluate the introduction of SwissDRG in a before-after comparison using the APDRG hospitals as a comparison group. This quasi-experimental approach allows us to control for factors that potentially confound the regional variation in inpatient services and are difficult to observe otherwise, e.g., differences in patient characteristics or other demand-side factors, the heterogeneity in the medical and non-medical staff, or differences in treatment styles.
We use the length of stay compliance (LOSC) as a main outcome to evaluate the impact of SwissDRG. LOSC relates the actual LOS to a yearly reference value per DRG and has been suggested as a meaningful indicator to compare hospital performance and the efficiency of services [8], [9], [10], [11]. Using administrative data from all Swiss university hospitals and a difference-in-differences methodology, we find a significant reduction in LOSC for hospitals that experienced a change in reimbursement scheme from per diem to DRG, compared to those that had a DRG system in place already before 2012. A reduction in LOSC implies that on average actual LOS values increased relative to the benchmark value, which is indicative of a lower hospital performance or service efficiency in response to SwissDRG. We interpret this as a short-term impact of SwissDRG and provide suggestive evidence for experience effects that may be a potential driving force behind this result, i.e., hospitals that worked with prospective DRG payments already before more quickly and efficiently adapted to the new system.
Our research contributes to an extensive literature on the impacts of DRG (e.g., [12], [13], [14], [15], [16], [17], [18], [19], [20], [21]). This literature studies the consequences of DRG payment systems on hospital operations, with no conclusive evidence yet on the changes in quantity, quality and efficiency of services. In Switzerland, Busato and von Below [5] present a small area analysis and find generally fewer hospitalizations and a higher share of resources in the outpatient sector in areas with APDRG reimbursement. Using a similar approach as in our study, Widmer and Weaver [6] do not find evidence for higher re-hospitalization rates or more early discharges after the introduction of APDRG (see also [22]). Weissenberger et al. [23] combine LOS, patient satisfaction, and quality of life measures and find a positive effect on the quality of hospital services of the APDRG compared to the per diem system (see also [24]). Felder et al. [1] find evidence for a shift from inpatient to outpatient care, but it remains unclear to what extent this result can be attributed to DRG or to general technological progress in the hospital sector. Pellegrini and Roth (2016, 2017) report cost increases from 2011 to 2012 in the inpatient sector, which could be the result of SwissDRG, but they find similar increases in psychiatric hospitalizations, which were not reimbursed by SwissDRG. Moreover, when looking over a longer time horizon, they present descriptive evidence that the cost increases in inpatient acute care after 2012 were mainly driven by a rather constant relative increase in case numbers [25], [26]. Widmer et al. [27] do not find much evidence for an expansion in the quantity of hospital services after the introduction of SwissDRG, but their results indicate that LOS in hospitals reduced due to shifts between the acute and post-acute phase. With our research we aim at supplementing the existing evidence on the impacts of DRG with (i) the use of LOSC as outcome to compare hospital performance, (ii) a quasi-experimental estimation of the short-term effects of SwissDRG on LOSC, and (iii) an empirical evaluation of different channels that might explain these impacts of SwissDRG on LOSC.
Section snippets
Institutional context
The hospital sector in Switzerland can be characterized by a regulated competition market form. There are currently approximately 300 hospitals, 40% of which are classified as general hospitals, the remaining 60% are specialized clinics [44]. The regulatory framework is given by the national health insurance law (KVG/LaMal), in particular articles 49/49a, but the hospital sector in general is rather decentralized and cantons (similar to US states) have the legal authority over the hospital
Data description
Our analysis is based on the medical statistics of the hospitals for the years 2011–2013, provided by the Swiss Federal Statistical Office. Since cantons did not apply payment systems uniformly to all their hospitals before 2012, we accessed the data release with the hospital type and focus on the university hospitals (type K111). Using hospital identifiers and the case numbers per hospital, we were able to uniquely identify these hospitals and assign them to two groups based on their payment
Descriptive findings
Fig. 2 shows the trends in LOSC by type of hospital (APDRG vs. non-APDRG) over the available time frame. The solid lines connect the quarterly means of LOSC, the dotted lines indicate 95% confidence intervals. It can be seen that the mean LOSC values are of similar magnitude for the treatment and the control group in 2011, prior to the introduction of SwissDRG, which confirms our argument of similar pre-treatment trends of both groups in Section 3.3. In 2012, we observe a decline in LOSC for
Discussion
Our results contribute to the literature on the impacts of DRG in at least three dimensions: First, we consider LOSC (as opposed to LOS) as outcome variable in our evaluation, following a recent literature that proposes this measure as suitable indicator for performance comparisons. We think that considering a benchmarked LOS in the evaluation of DRG systems is important because LOS has several limitations. For example, there is a general decline in LOS observed over time that makes it
Conclusion
This study aimed at evaluating the short-term impacts of the introduction of SwissDRG on LOSC, a relative measure that compares the actual LOS of a case to a benchmark value. We find a relative reduction in LOSC for hospitals that changed from per diem to SwissDRG reimbursement, compared to hospitals that already used a DRG system before (based on APDRGs). If interpreted as a proxy for efficiency, our results point to a reduction in service efficiency, at least in the short-run. We explain the
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Conflict of interest
Both authors Prof. Stefan Boes and Christoph Napierala do not have any conflict of interest regarding the research presented in “DRG hospital reimbursement and length of stay compliance: Evidence from a system change in Switzerland”. There has not been any financing nor any other incentive been provided to pursue this research. Funding has been solely covered through University of Lucerne.
References (48)
- et al.
A new method of classifying prognostic comorbidity in longitudinal studies: development and validation
J Chronic Dis
(1987) - et al.
New ICD-10 version of the Charlson comorbidity index predicted in-hospital mortality
J Clin Epidemiol
(2004) - et al.
Analyzing health insurance claims on different timescales to predict days in hospital
J Biomed Informatics
(2016) - et al.
Mortality and life expectancy in persons with severe unipolar depression
J Affect Disord
(2016) - et al.
Survival after in-hospital cardiac arrest is highly associated with the Age-combined Charlson Co-morbidity Index in a cohort study from a two-site Swedish University hospital
Resuscitation
(2016) - et al.
Hospital ownership and cost and quality of care: is there a dime?.s worth of difference?
J Health Econ
(2001) - et al.
Medicare upcoding and hospital ownership
J Health Econ
(2004) - et al.
Leistungs- und Kostenverschiebungen zwischen dem akutstatioäen und dem spital- und praxisambulanten Sektor
(2014) Kennzahlen der Schweizer Spitäler 2008/2009. Federal Office of Public Health (FOPH)
(2015)- et al.
Die Spitalversorgung im Spannungsfeld der kantonalen Spitalpolitik: Aktualisierung 2015
(2016)
Explaining regional variations in health care utilization between swiss cantons using panel econometric models
BMC Health Serv Res
The implementation of DRG-based hospital reimbursement in Switzerland: a population-based perspective
Health Res Policy Syst
Der Einfluss von APDRG auf Aufenthaltsdauer und Rehospitalisierungen: Auswirkungen von Fallpauschalen in Schweizer Spitälern zwischen 2001 und 2008
Obsan Bericht 49
Qualität der stationären Leistungen unter der neuen Spitalfinanzierung. Monitoring der Qualitätsindikatoren 2009-2016
Obsan Dossier 65
Diagnosis-related group analysis and strategic healthcare budgeting
Capsula Eburnea
Management Sanitario: principi, esperienze locali e internazionali, prospettive
Compliance of length of stay a measure of choice in a DRG system. Working paper
Is the timing of radiological intervention and treatment day associated with economic outcomes in DRG-financed health care systems: a case study
BMC Health Serv Res
Hospital management by diagnosis related groups: United States and European experience
Third international conference on system science in health care
Comparing outcomes of care before and after implementation of the DRG-based prospective payment system
JAMA
DRGs: their design and development
First results of the introduction of DRGs in Germany and overview of experience from other DRG countries
J Public Health
Hospital case payment systems in Europe
Health Care Manag Sci
Impact of DRGs. Introducing a DRG reimbursement system: a literature review
Cited by (6)
Extended Length of Hospital Stay for Surgical and Medical Patients - Insights from Hospital and Psychosocial Predictors
2022, Proceedings of the Annual Hawaii International Conference on System Sciences