The annual incidence of stroke in the community is about 2 per 1000 population,1 whereas among hospital inpatients it is 11 per 1000.2 However, a study that systematically and simultaneously identifies all inpatients experiencing stroke and all patients admitted with stroke does not exist. Previous work on stroke among inpatients has excluded some patients—for example, those with3 or without2 obvious iatrogenic predisposing factors. Similarly, although risk factors for stroke have been used as predictors of an event in the context of a study,3 these are often not documented clinically. Secondary prevention is dependent on identification and documentation of risk factors.
We compared outcomes and the identification and documentation of known risk factors in a cohort of patients admitted with a stroke or having had a stroke while in hospital (having been admitted to hospital with a primary diagnosis other than stroke).
Subjects, methods, and results
University Hospital Aintree serves a predominantly urban population of 250 000 and admits about 32 000 patients annually. Its stroke unit has 18 acute and 25 rehabilitation beds. Guidelines for the management of acute stroke are available throughout the hospital.
We identified all patients with a primary diagnosis of stroke (excluding transient ischaemic attacks and subarachnoid haemorrhages) on a stroke register. From October 1994 to March 1997, 100 inpatients with stroke and 1274 patients admitted with stroke were identified prospectively by a 24 hour, on-call stroke research team or retrospectively from the hospital discharge coding. Data collection was by retrospective review of case notes.
Median ages were 75 (interquartile range 67-82) years for inpatients and 74 (66-81) years for admitted patients. Fifty four (54%) inpatients and 647 (51%) admitted patients were female. Forty seven (47%) inpatients and 537 (42%) admitted patients were managed in the stroke unit.
The table shows the numbers of patients for whom known risk factors for stroke were clearly documented and the numbers for whom no documentation existed. According to documentation, cardiovascular risk factors were significantly higher in inpatients whereas previous strokes or transient ischaemic attacks were more common among admitted patients. Documentation was less complete for inpatients than for admitted patients.
Of the 80 (80%) inpatients and 1092 (86%) admitted patients who had computed tomography, 5 (6%) and 144 (13%) respectively had a primary intracerebral haemorrhage (χ2=2.64, P>0.05).
The inpatients remained in hospital longer after stroke (median 31 (interquartile range 13-59) days) than the admitted patients (16 (6-43) days). Twenty four (24%) inpatients returned to their previous residence, compared with 799 (63%) admitted patients (odds ratio 0.19, 95% confidence interval 0.11 to 0.31). Sixteen (16%) inpatients were newly discharged to an institution, compared with 124 (10%) admitted patients (1.77, 95% confidence interval 0.93 to 3.16), which may partly account for the longer stay for inpatients. Sixty (60%) inpatients died in hospital, compared with 351 (28%) admitted patients (3.94, 95% confidence interval 2.55 to 6.15); stroke was the primary or secondary cause of death for 51 (85%) inpatients and 301 (86%) admitted patients.
Comment
Although the inpatients and the admitted patients were similar in terms of age and sex, inpatients stayed in hospital longer, were more likely to die in hospital, and had less well documented risk factors. Improving staff awareness on medical and surgical wards regarding the importance of the early identification and documentation of known risk factors for stroke may improve outcome.
Table.
Identification and documentation of known stroke risk factors among patients who had a stroke while in hospital and among patients who were admitted with stroke. Values are numbers (percentages) of patients
| Identified
|
Not documented
|
||||||
|---|---|---|---|---|---|---|---|
| Inpatients (n=100) | Admitted patients (n=1274) | P value | Inpatients (n=100) | Admitted patients (n=1274) | P value | ||
| Cardiac failure | 35 (35) | 41 (3) | <0.01 | 60 (60) | 954 (75) | <0.01 | |
| Atrial fibrillation | 32 (32) | 253 (20) | <0.01 | 38 (38) | 223 (18) | <0.01 | |
| Myocardial infarction | 25 (25) | 181 (14) | <0.01 | 30 (30) | 207 (16) | <0.01 | |
| Angina | 19 (19) | 138 (11) | <0.03 | 68 (68) | 830 (65) | >0.61 | |
| Hypertension | 31 (31) | 480 (38) | >0.19 | 30 (30) | 196 (15) | <0.01 | |
| Diabetes mellitus | 11 (11) | 149 (12) | >0.88 | 33 (33) | 195 (15) | <0.01 | |
| Previous stroke | 16 (16) | 329 (26) | <0.04 | 38 (38) | 188 (15) | <0.01 | |
| Previous transient ischaemic attack | 9 (9) | 240 (19) | <0.02 | 56 (56) | 334 (26) | <0.01 | |
| Ever smoked | 39 (39) | 685 (54) | <0.01 | 20 (20) | 162 (13) | >0.05 | |
All tests were with Yates's corrected χ2.
Acknowledgments
We thank Liz Lightbody, Hazel Dickinson, and Dimitrios Theofanidis, who collected data from the European stroke database, and the BMJ reviewers (Gord Gubitz and M J Campbell) for their comments.
Footnotes
Funding: None.
Competing interests: None declared.
References
- 1.Bamford J, Sandercock P, Dennis M, Warlow C, Jones L, McPherson K, et al. A prospective study of acute cerebrovascular disease in the community: the Oxfordshire community stroke project 1981-86: 1. Methodology, demography and incident cases of first-ever stroke. J Neurol Neurosurg Psychiatry. 1988;51:1371–1380. doi: 10.1136/jnnp.51.11.1373. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Azzimondi G, Nonino F, Fiorani L, Vignatelli L, Stracciari A, Pazzaglia P, et al. Incidence of stroke among inpatients in a large Italian hospital. Stroke. 1994;25:1752–1754. doi: 10.1161/01.str.25.9.1752. [DOI] [PubMed] [Google Scholar]
- 3.Mahaffey KW, Granger CB, Sloan MA, Thompson TD, Gore JM, Weaver WD, et al. Risk factors for in-hospital nonhemorrhagic stroke in patients with acute myocardial infarction treated with thrombolysis: results from GUSTO-I. Circulation. 1998;97:757–764. doi: 10.1161/01.cir.97.8.757. [DOI] [PubMed] [Google Scholar]