Productivity and Quality Analysis of Norwegian Hospital Mergers

Link to article:



Emanuelsson, Natalie




Andre skrifter


Background and research question: Hospital mergers have been a hot topic in the Norwegian healthcare sector the past two decades. This study is using a sample of 22 Norwegian hospitals that have merged from 1999 until 2014. A major ownership reform have influenced this period and contributed to both hospital mergers and reorganization. I have examined whether hospital mergers have resulted in increased productivity and quality in Norwegian hospitals.

Data: To perform this study, the Norwegian Patient Register has provided patient-level data. These data are necessary to develop case-mix adjustments for all hospitals included in the analysis and to calculate each patient’s travel time from their municipality to the specific hospital. Moreover, hospital data from 1999 until 2014 have been aggregated and merged with the patient-level data from the same time. A sample of approximately 83,000,000 patient records from 1999 until 2014 were used in the analysis.

Methodology: Case-mix adjustment variables such as DRG, patient characteristics, and treatment variables were used to adjust for differences in patient composition between hospitals and to estimate the performance indicators in three models. Productivity scores for both merged and non-merged hospitals were estimated through a Simar & Wilson two-stage approach using Data Envelopment Analysis (DEA-SW), where the Difference-in-Differences estimation (DiD) for productivity were calculated simultaneously. In addition, DiD estimation on quality-scores were calculated to detect the potential differences of quality between merged and non-merged hospitals.

Most important findings: The DEA-SW results showed that merged hospital does not perform with better productivity than non-merged hospitals. However, there is a positive trend over time, which indicate that the productivity off all hospitals in the analysis are increasing. Furthermore, the DiD estimation did not show any evidence of increased quality for merged hospitals compared to non-merged hospitals. The merged hospitals had higher and statistically significant DID-coefficient at a 5% level in all five tables, which means that they have lower quality compared to the non-merged hospitals.


Oppdragsgiver: NFR via NTNU
Oppdragsgivers prosjektnr.: 214338
Frisch prosjekt: 4115 - The effect of DRG-based financing on hospital