Moderate agreement between self-reported stroke and hospital-recorded stroke in two cohorts of Australian women: a validation study

Table 3 Scenarios of agreement in the validation of self-reported stroke

Scenario	Prevalence (self-report)	Prevalence (hospital-recorded)	Sensitivity	Positive predictive value	Kappa (95% CI)	Prevalence index
1921-26 cohort
A*	6.6%	2.1%	78.1%	24.5%	0.35 (0.25 to 0.46)	0.91
B ^†	6.6%	3.1%	86.0%	41.2%	0.54 (0.44 to 0.63)	0.90
C ^‡	6.6%	3.4%	86.8%	45.1%	0.57 (0.48 to 0.67)	0.90
D	6.6%	4.0%	88.7%	53.9%	0.65 (0.57 to 0.74)	0.89
1946-52 cohort
A*	1.2%	0.7%	78.6%	44.0%	0.56 (0.37 to 0.75)	0.98
B ^†	1.2%	0.7%	80.0%	48.0%	0.60 (0.42 to 0.78)	0.98

*Scenario A: Agreement between self-reported and hospital-recorded stroke.
^†Scenario B: Hospital comparison group includes TIA and sequelae of cerebrovascular disease diagnoses, and hospital–recorded strokes occurring before the period of interest.
^‡Scenario C: Compares self-reported stroke with any hospital-recorded cerebrovascular disease.
^§Scenario D: Scenario C, plus includes women as having had a possible stroke if they provided comments on the stroke event in the questionnaire.

ISSN: 1471-2288