Objective To determine accuracy of the Rose Angina Questionnaire (RAQ) for diagnosing coronary heart disease (CHD) among Bangladeshi adults, by comparing the classification based on the questionnaire with cardiologists' diagnosis.

Methods A case-control study of non-smoking Bangladeshi adults aged 40–75 years, was conducted in 2010. Cases were incident cases of CHD from two cardiac hospitals if diagnosed as such by the cardiologists; controls were non-cardiac patients from cardiac outpatient departments. One control was matched to each case on age and gender. Full version of the original RAQ questionnaire was used for study participants.

Results The sample comprised 302 CHD cases and 302 controls (mean age 53±8.5 years). RAQ detected 194 cases (32%) and 409 controls (68%) from either hospital. RAQ categorised 17.5% of sample as having CHD who were considered not to have CHD by the cardiologists; RAQ categorised 34.5% of sample as not suffering from CHD, who were diagnosed as having CHD by the cardiologists. Among 301 CHD cases, 160 (53.2%) were diagnosed as CHD by both hospital cardiologists and RAQ. Among 302 controls, 268 (88.7%) were diagnosed as not having CHD by both hospital cardiologists and RAQ. CHD patients are five times more likely to have RAQ positive result compared to controls.

Conclusion The RAQ had sensitivity of 53%, specificity of 89% and likelihood ratio positive of 4.8 in diagnosing CHD among Bangladeshi adults compared with diagnosis by cardiologists. RAQ can be used as an alternate tool for diagnosing CHD at field sites where there are limited resources.