Readmission in Norwegian hospitals
Report
|Updated
Readmission rates are increasingly being used as an indicator of quality of care in other countries.
Background
The Norwegian Health Directorate has commissioned The Norwegian Knowledge Centre for the health services (NOKC) to propose a suitable method for calculating and presenting readmission rates in Norway. This report describes the method for the indicator calculation and presents the indicator results on hospital and municipality level.
Methods
The analyses in this report are based on patient administrative data from 20 Norwegian health trusts and four private hospitals, covering all acute care hospitals. The methods and definitions are based on a similar study in Denmark from 2009, where a readmission indicator developed by Sundhedsstyrelsen was studied in an elderly population. In the present report we studied readmissions in patients 67 years and older, admitted to hospitals with a diagnosis belonging to one of eleven predetermined diagnosis groups: asthma/chronic obstructive pulmonary disease (COPD), urinary infection, fracture, dehydration, constipation, gastroenteritis, rheumatoid arthritis, heart failure, anemia, pneumonia and stroke. The data were linked to the National Registry in Norway. A readmission was defined as an acute admission between 8 hours and 30 days after a previous hospital discharge (primary admission). Data from the time period 2005-2009 was used in the present study.
The readmission rate was calculated for the eleven diagnosis groups combined, and presented stratified by both hospital and municipality. Moreover, readmission rates after a primary admission for asthma/COPD, fracture, heart failure and pneumonia are presented separately. The results are adjusted for gender and age, and for the combined readmission indicator, diagnosis group as well.
Results
The readmission rates differ substantially both between hospitals and between municipalities. For the eleven diagnosis groups combined, eight hospitals had lower and eight hospitals had higher, readmission rate than other hospitals. One municipality had significantly lower readmission rate than average, whereas 30 municipalities had higher readmission rates. For the majority of hospitals and municipalities, however, the differences were small and insignificant.
Conclusion
This report describes a method for calculating the readmission rates in Norwegian hospitals for patients 67 years and older. Eleven diagnosis groups were included in the analyses. The results are presented on hospital and municipality level, and show that readmissions are common in Norwegian hospitals. Although various definitions and methods for data capture and analyses have been used in other readmission studies, our results seem to be reasonably in accordance with studies from other countries. Significant geographical variations, as well as differences between hospitals, were observed. For most municipalities and hospitals, however, the differences are small.
Unplanned readmissions are costly and may reflect suboptimal patient outcomes. It is outside the scope for this report to investigate the causes of differences in readmission rates, however, previous studies have shown that general hospitalization rates, discharge routines, and factors related to the primary healthcare system may impact the readmission rates. Accordingly, readmission rates do not only reflect hospital quality of care, but also care given by the primary healthcare systems in local municipalities. Furthermore, the data used in the present study cannot be used to evaluate the proportion of readmissions that are avoidable.