Stroke units, certification, and outcomes in German hospitals: a longitudinal study of patient-based 30-day mortality for 2006-2014
Journal
BMC Health Services Research
Type
journal article
Date Issued
2018-11
Author(s)
Research Team
https://www.alexandria.unisg.ch/persons/8635
Abstract
Background
Treatment of stroke patients in stroke units has increased and studies have shown improved outcomes. However, a large share of patients in Germany is still treated in hospitals without stroke unit. The effects of stroke unit service line, and total hospital quality certification on outcomes remain unclear.
Methods
We employ annual hospital panel data for 1100–1300 German hospitals from 2006 to 2014, which includes structural data and 30-day standardized mortality. We estimate hospital- and time-fixed effects regressions with three main independent variables: (1) stroke unit care, (2) stroke unit certification, and (3) total hospital quality certification.
Results
Our results confirm the trend of decreasing stroke mortality ratios, although to a much lesser degree than previous studies. Descriptive analysis illustrates better stroke outcomes for non-certified and certified stroke units and hospitals with total hospital quality certification. In a fixed effects model, having a stroke unit has a significant quality-enhancing effect, lowering stroke mortality by 5.6%, while there is no significant improvement effect for stroke unit certification or total hospital quality certification.
Conclusions
Patients and health systems may benefit substantially from stroke unit treatment expansion as installing a stroke unit appears more meaningful than getting it certified or obtaining a total hospital quality certification. Health systems should thus prioritize investment in stroke unit infrastructure and centralize stroke care in stroke units. They should also prioritize patient-based 30-day mortality data as it allows a more realistic representation of mortality than admission-based data.
Treatment of stroke patients in stroke units has increased and studies have shown improved outcomes. However, a large share of patients in Germany is still treated in hospitals without stroke unit. The effects of stroke unit service line, and total hospital quality certification on outcomes remain unclear.
Methods
We employ annual hospital panel data for 1100–1300 German hospitals from 2006 to 2014, which includes structural data and 30-day standardized mortality. We estimate hospital- and time-fixed effects regressions with three main independent variables: (1) stroke unit care, (2) stroke unit certification, and (3) total hospital quality certification.
Results
Our results confirm the trend of decreasing stroke mortality ratios, although to a much lesser degree than previous studies. Descriptive analysis illustrates better stroke outcomes for non-certified and certified stroke units and hospitals with total hospital quality certification. In a fixed effects model, having a stroke unit has a significant quality-enhancing effect, lowering stroke mortality by 5.6%, while there is no significant improvement effect for stroke unit certification or total hospital quality certification.
Conclusions
Patients and health systems may benefit substantially from stroke unit treatment expansion as installing a stroke unit appears more meaningful than getting it certified or obtaining a total hospital quality certification. Health systems should thus prioritize investment in stroke unit infrastructure and centralize stroke care in stroke units. They should also prioritize patient-based 30-day mortality data as it allows a more realistic representation of mortality than admission-based data.
Language
English
Keywords
Stroke
Stroke unit
Hospital specialization
Certificate
Accreditation
Stroke outcomes
HSG Classification
contribution to scientific community
Refereed
Yes
Volume
18
Number
880
Subject(s)
Eprints ID
259576
File(s)![Thumbnail Image]()
Name
0.1186s129130183664.pdf
Size
750.3 KB
Format
Adobe PDF
Checksum (MD5)
4c8859165f83a4bb69251066af0461ae