Format

Send to

Choose Destination
BMC Health Serv Res. 2015 Dec 30;15:583. doi: 10.1186/s12913-015-1244-y.

Agreement between routine electronic hospital discharge and Scottish Stroke Care Audit (SSCA) data in identifying stroke in the Scottish population.

Author information

1
Division of Applied Medicine, Department of Medicine and Therapeutics, Polwarth Building, Foresterhill, University of Aberdeen, Aberdeen, UK. m.e.turner@abdn.ac.uk.
2
NHS Lanarkshire Stroke MCN, Stroke Unit, Monklands Hospital, Monkscourt Avenue, Airdrie, UK. mark.barber@nhs.net.
3
Information Services Division, NHS National Services Scotland, Edinburgh, UK. hazeldodds@nhs.net.
4
Centre for Clinical Brain Sciences, University of Edinburgh, Royal Infirmary of Edinburgh, Edinburgh, UK. martin.dennis@ed.ac.uk.
5
Institute of Cardiovascular and Medical Sciences, University of Glasgow, Royal Infirmary, Glasgow, UK. Peter.Langhorne@glasgow.ac.uk.
6
Division of Applied Medicine, Department of Medicine and Therapeutics, Polwarth Building, Foresterhill, University of Aberdeen, Aberdeen, UK. m.j.macleod@abdn.ac.uk.

Abstract

BACKGROUND:

In Scotland all non-obstetric, non-psychiatric acute inpatient and day case stays are recorded by an administrative hospital discharge database, the Scottish Morbidity Record (SMR01). The Scottish Stroke Care Audit (SSCA) collects data from all hospitals managing acute stroke in Scotland to support and improve quality of stroke care. The aim was to assess whether there were discrepancies between these data sources for admissions from 2010 to 2011.

METHODS:

Records were matched when admission dates from the two data sources were within two days of each other and if an International Classification of Diseases (ICD) code of I61, I63, I64, or G45 was in the primary or secondary diagnosis field on SMR01. We also carried out a linkage analysis followed by a case-note review within one hospital in Scotland.

RESULTS:

There were a total of 22 416 entries on SSCA and 22 200 entries on SMR01. The concordance between SSCA and SMR01 was 16 823. SSCA contained 5593 strokes that were not present in SMR01, whereas SMR01 contained 185 strokes that were not present in SSCA. In the case-note review the concordance was 531, with SSCA containing 157 strokes that were not present in SMR01 and SMR01 containing 32 strokes that were not present in SSCA.

CONCLUSIONS:

When identifying strokes, hospital administrative discharge databases should be used with caution. Our results demonstrate that SSCA most accurately represents the number of strokes occurring in Scotland. This resource is useful for determining the provision of adequate patient care, stroke services and resources, and as a tool for research.

PMID:
26719156
PMCID:
PMC4697331
DOI:
10.1186/s12913-015-1244-y
[Indexed for MEDLINE]
Free PMC Article

Supplemental Content

Full text links

Icon for BioMed Central Icon for PubMed Central
Loading ...
Support Center