AB1362
VALIDATION OF ICD-10 CODES FOR STROKE IN PATIENTS WITH RHEUMATOID ARTHRITIS
M. Holmqvist 1,*E. Gränsmark 1L. Alfredsson 2J. Askling 1
1Clinical epidemiology unit, Department of Medicine, Solna, Karolinska University Hospital, 2Institute of Environmental Medicine, Karolinska Institutet, Stockholm, Sweden
Background: It is well known that patients with rheumatoid arthritis (RA) are at increased risk of ischemic heart disease. The results pertaining to risk of stroke in RA are conflicting; several studies have indicated no increased risk. When using administrative claims data to detect outcome, the validity of the registered codes are of pivotal importance. Misclassification of outcome could dilute risks that actually are present. In Sweden, hospitalization data are registered using international classification of diseases (ICD) codes. A validation study of stroke codes in the ICD-8 and ICD-9 based on the general population published in 1993 indicated that the codes for stroke in ICD-8 or ICD-9 classifications in the general unselected population had a high validity [1]; the positive predictive value of an ICD-8 or ICD-9 code indicating stroke was found to be 94% using medical charts and current clinical criteria to validate register information. There have been no studies of the validity of ICD-10 codes for stroke specifically among patients with RA.
Objectives: To assess the validity of the ICD-10 codes indicating ischemic and hemorrhagic stroke.
Methods: The Epidemiological Investigation of Rheumatoid Arthritis (EIRA) case-control study of incident RA was linked to the Swedish Hospital Discharge register and the charts of all individuals who after inclusion in EIRA were hospitalized with an ICD-10 code indicating IS (I63) or HS (I61) (n= 36 events among 26 cases and 40 events among 30 controls) were reviewed. The gold standard was stroke verified via computed tomography (CT). In situations where the CT was normal but there was a clinically strong suspicion of stroke, the MONICA criteria [2] were used to classify IS. The positive predictive value (PPV) was calculated for both ICD-codes, overall, and stratified by RA-status.
Results: Out of the 68 listings with I63 (IS), 62 were verified as IS either via CT or the MONICA criteria. Out of the 7 listings with I61 (HS), 7 were verified as HS via CT. The PPV for IS given a listing of I63 was 91% and for HS given a listing of I61 was 100%. Overall, the PPV of IS and HS combined was 92% among RA patients and 90% among the controls.
Conclusions: The validity of ICD-10 listings of stroke in the Swedish Hospital Discharge register is high, and can safely be used in register linkage studies. This also applies for patients with RA.
References: 1. Lindblad, U., et al., Validity of register data on acute myocardial infarction and acute stroke: the Skaraborg Hypertension Project. Scand J Soc Med, 1993. 21(1): p. 3-9.
2. WHO, MONICA manual, Part IV: Event registration, Section 2: Stroke event registration data component. 1990, World Health Organization: Geneva.
Disclosure of Interest: None Declared