Have Dutch Hospitals Saved Lives and Reduced Costs? A longitudinal patient-level analysis over the years 2013–2017

Sandra Sülz*, Holger Wagenaar, Joris van de Klundert

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

1 Citation (Scopus)
32 Downloads (Pure)


The purpose of this paper is to shed light on the ongoing Dutch health system reforms and identify whether hospital costs and hospital outcomes have changed over time. We present an empirical analysis that is based on granular micro-costing data and focuses on conditions for which mortality is indicative of outcome quality, that is, acute myocardial infarction (AMI), chronic heart failure (CHF), and pneumonia (PNE). We deploy a dataset of more than 80,000 inpatient episodes over 5 years (2013–2017) to estimate regression models that control for variation between patients and hospitals. We have three main findings. First, our results do not indicate significant outcome improvements over the years; that is, there is no time trend for mortality. Second, there is heterogeneity in cost developments: for patients who survive their inpatient stay, our data indicate that costs increase significantly by 0.9% per year for AMI patients, while costs decrease significantly by 1.7% per year for CHF patients and by 1.9% per year for PNE patients. For patients who pass away during their inpatient stay, our data do not indicate significant time trends. Third and finally, our results suggest the existence of substantial cost variation between hospitals.

Original languageEnglish
Pages (from-to)2399-2408
Number of pages10
JournalHealth Economics (United Kingdom)
Issue number10
Early online date12 Jul 2021
Publication statusPublished - Sept 2021

Bibliographical note

Publisher Copyright:
© 2021 The Authors. Health Economics published by John Wiley & Sons Ltd.

Research programs

  • EMC NIHES-05-63-01 Management


Dive into the research topics of 'Have Dutch Hospitals Saved Lives and Reduced Costs? A longitudinal patient-level analysis over the years 2013–2017'. Together they form a unique fingerprint.

Cite this