Vortragssitzung

Healthcare Utilisation and Costs

Talks

Does increasing emergency department use really drive up costs? A decomposition analysis with accident claims data
Christoph A. Thommen, ZHAW
Michael Stucki, ZHAW

Einleitung / Introduction

In primary care, it is the role of general practitioners (GPs) that has undergone modifications during the last decades. Patients are increasingly being treated at hospital emergency departments (EDs) and GPs seem to lose ground as principal providers. However, systematic analyses of the extent of such a phenomenon are still scarce. Most importantly, whether and how these changes impact health care costs is not well understood - a gap that our study aims to address.

Methode / Method

We decompose the change in costs per patient between 2008 and 2018 along the full cost distribution into various explanatory variables. We apply a method based on estimating the conditional distribution (Chernozhukov et al. 2013). The conditional model includes five sets of covariates: 1. An indicator for initial care provider, 2. indicators for each involved main care provider, 3. the number of different types of supporting provider, 4. characteristics of the accident, 5. patient characteristics.

Ergebnisse / Results

We find that, contrary to our expectations, the shift in initial care provision in trauma care from GPs to EDs does not play a role in explaining the increase in accident care costs between 2008 and 2018. The number of different types of providers represents the most important factor in explaining the cost growth. The involved providers are the second most important explanatory factor. The cost growth appears to be driven largely by volume expansion in supporting services such as radiology examinations and physiotherapy treatments. Our indicative analyses show that the expansion in supporting providers is most accentuated among patients that receive initial care from GPs. These findings confirm the picture drawn by our main analyses, according to which the shift of the initial provider from the GP to the ED had no relevant consequences for the development of costs between 2008 and 2018.

Zusammenfassung / Conclusion

In our paper, we analyze how trauma care provision changed between 2008 and 2018 and to what extent these changes were associated with cost increases. We analyze how various factors contribute to changes in costs per case by decomposing the change in costs over time. Drawing on a large dataset of N=3,620,134 accident cases from 2008 and 2018 of the largest Swiss accident insurer, we find that the shift in initial care provision in trauma care from GPs to EDs does not contribute to the cost growth in accident care costs. This contradicts the hypothesis of the cost-driving effect of GPs losing ground as (initial) care providers at the cost of EDs. However, the cost growth seems to be mainly due to an expansion in the volume of supporting services such as radiology examinations, in particular among patients initially treated by a GP.


Authors
Marc Höglinger, ZHAW
Stefan Scholz-Odermatt, SUVA
Healthcare expenditure projections up to 2050 for Switzerland: ageing and the COVID-19 crisis
Carsten Colombier, Eidgenössische Finanzverwaltung

Einleitung / Introduction

A key element of a forward-looking health policy is the question of the sustainable financing of the healthcare system, which has been reinforced by the COVID-19 pandemic. The coming ageing of the population puts additional pressure on public finances. In terms of GDP, Switzerland had the second most expensive healthcare system after the United States in 2019 and one of the most dynamic healthcare-expenditure growth rates among OECD countries in the past 30 years. To evaluate the fiscal sustainability of the Swiss health system, we carry out healthcare expenditure projections from 2019 to 2050 under a "no-policy-change" assumption. To account for the uncertainty surrounding our demographic assumptions and the cost impact of non-demographic factors, we create different scenarios. We also produce a policy scenario that analyses the cost-containing effects of cost targets.

Methode / Method

We apply a cohort simulation model to project HCE decomposed by HCE without long-term care (LTC) and LTC from the age of 65.

Ergebnisse / Results

COVID-related HCE is a temporary burden on public budgets. Ageing puts more pressure on LTC expenditure than on HCE without LTC. While ageing dominates the increase of public HCE, non-demographic drivers exert the strongest pressure on the compulsory health insurance. Our policy scenario illustrates that cost targets can substantially relieve the burden on public budgets and the compulsory health insurance.

Zusammenfassung / Conclusion

Even before the COVID-19 crisis, rapidly growing healthcare expenditure was calling the sustainability of public finances into question. The pandemic has reinforced these concerns and also underlined the importance of resilient healthcare systems. To highlight the need for economic policy action in the healthcare sector, this paper provides expenditure projections for Switzerland up to 2050. The expenditure projections take into account the financial impact of the COVID-19 crisis and foreseeable ageing of the population. The projections show that while COVID-related healthcare expenditure is a burden on public budgets in the short term, the ageing of the population will put continued and growing pressure on public budgets and compulsory health insurance until 2050. In the medium to long term, however, healthcare expenditure is driven not only by demographic change, but also by non-demographic factors such as rising income, medical advances and Baumol's cost disease. The projections also suggest that long-term care will be affected by higher cost growth than the rest of the healthcare system. The sensitivity analyses show that the strongest cost pressure comes from alternative assumptions about the effect of the non-demographic cost drivers. In addition, a policy scenario discusses the cost-dampening effects of cost targets.


Authors
Carsten Colombier, Eidgenössische Finanzverwaltung
Thomas Brändle, Eidgenössische Finanzverwaltung
Discontinuity in Primary Care and its Effects on Healthcare Utilization - Evidence from resigning German General Practitioners
Daniel Monsees, RWI - Leibniz-Institut für Wirtschaftsforschung

Einleitung / Introduction

In healthcare markets, patients often cannot determine the quality of the services provided by their doctors and hence, they need to trust them. Additionally, to derive the right and patient-specific treatments from certain symptoms, doctors require much expertise and specific knowledge about, e.g., the medical history of the patients. These are two arguments why a long and close relationship between a doctor and a patient may be beneficial to the patient. On the other hand, a too close relationship may also be susceptible to moral hazard or doctors may become inaccessible for new and innovative treatments that would rather be optimal.

Methode / Method

In this paper, we evaluate the effects of a disruption of the patient-provider relationship on health care utilization using detailed administrative claims data from a German statutory health insurance comprising almost 9 million insurees. Specifically, we assess the health and health care utilization consequences for patients whose primary care provider resigns from his profession (e.g., due to retirement) in an event-study setting relative to the period of retirement. We thereby contribute to a recently emerging strand if the literature that investigates the effects of patient-provider disruptions using event study methods. Results of this literature suggest that albeit disruptions have significant effects on healthcare utilization, the magnitude and persistence of these effects is influenced by organizational factors of the healthcare regime.

Ergebnisse / Results

Our results show a persistent decline in the probability of consulting a general practitioner after patient's primary care physicians exit, which coincides with increased ambulatory care spending. We further explore effect heterogeneities between patients, in particular regarding morbidities and analysing shifts in healthcare utilization. We find that hospitalizations are only increased in the short run, while the increase in overall ambulatory care spending is mainly driven by patients with chronic conditions.


Authors
Daniel Monsees, RWI - Leibniz-Institut für Wirtschaftsforschung
Matthias Westphal, TU Dortmund
Christmas shopping in the prescription drug market
Linn Hjalmarsson, Universität Bern

Einleitung / Introduction

Since the introduction of health insurance, there has always been a tradeoff between risk protection and maintaining incentives. Following the seminal work of Arrow (1963), an extensive literature focused on moral hazard in health insurance. However, non-linear insurance contracts not only induce moral hazard, but also the incentive to strategically retime treatments in order to minimize out-of-pocket payments. Such behavior may have adverse consequences with respect to health outcomes if the temporally increased demand of some patients leads to supply constraints. Nevertheless, this so-called intertemporal substitution has only recently received attention in the literature. The present study aims at closing this gap in the literature by explicitly focusing on prescription drugs. Prescription drugs have the appealing feature that the time of purchase and consumption generally don’t overlap. This makes prescription drug purchases especially prone to strategic timing. In order to analyze whether such strategic timing takes place, I exploit a policy feature which many non-linear health insurance contracts share, namely the deductible reset at the start of each calendar year. This reset creates a sharp and salient increase in the out-of-pocket price for patients who exceeded the deductible in the previous year. I therefore analyze whether chronically ill patients anticipate this price increase leading them to stockpile prescription drugs in times where they face a relatively low price. Moreover, I assess the role of providers in prescribing and dispensing prescription drugs. Finally, I explore whether irrational stockpiling behavior of some patients may have adverse consequences with respect to health outcomes for other patients. I use rich claims level data containing individual level data of roughly 8 million prescription drug claims from the largest Swiss health insurer over a period of 5 years. Preliminary results show that chronically ill patients who exceed their deductible before October and consequently face a relatively low out-of-pocket price, purchase substantially more prescription drugs towards the end of the year.


Authors
Linn Hjalmarsson, Universität Bern