AIMS: To design, pilot, and validate a questionnaire to test contraceptive knowledge in combined oral contraceptive pill users. METHOD: To ensure face and content validity the questionnaire was developed using existing unvalidated instruments and in consultation with GPs, local and national family planning experts, and with pill users. The questionnaire was then piloted with 15 current oral contraceptive pill users and 10 local GPs, modified, and a contraceptive knowledge 'score' developed. Construct validity-that the degree of family planning training should predict contraceptive knowledge-was tested in four groups: family planning trainers (n=28), GPs (n=40), current pill users (n=53), and male medical students (n=59). Thirty current pill users were sent the questionnaire after two weeks to determine test-retest reliability. RESULTS: The questionnaire showed construct validity: there was a gradient of scores across the four knowledge groups. Family planning trainers had the highest scores, followed by GPs, and current pill users, with male medical students having the lowest scores (Kruskal-Wallis test p<0.001). Women had good knowledge of situations when pill efficacy is reduced but poorer knowledge of what action to take subsequently. Predictors of knowledge in pill users were educational level, age, and the importance attached to not falling pregnant-thus providing further evidence for the construct validity of the questionnaire. Test-retest reliability was good (rank correlation 0. 73). CONCLUSION: A contraceptive knowledge questionnaire suitable for use in audit or research has been developed which is reliable and has face, content and construct validity. Pill users have poor knowledge of what to do in situations of pill failure and about the details of emergency contraception. Trials are needed to assess the effectiveness of different strategies to improve contraceptive knowledge in women.
AIMS: To design, pilot, and validate a questionnaire to test contraceptive knowledge in combined oral contraceptive pill users. METHOD: To ensure face and content validity the questionnaire was developed using existing unvalidated instruments and in consultation with GPs, local and national family planning experts, and with pill users. The questionnaire was then piloted with 15 current oral contraceptive pill users and 10 local GPs, modified, and a contraceptive knowledge 'score' developed. Construct validity-that the degree of family planning training should predict contraceptive knowledge-was tested in four groups: family planning trainers (n=28), GPs (n=40), current pill users (n=53), and male medical students (n=59). Thirty current pill users were sent the questionnaire after two weeks to determine test-retest reliability. RESULTS: The questionnaire showed construct validity: there was a gradient of scores across the four knowledge groups. Family planning trainers had the highest scores, followed by GPs, and current pill users, with male medical students having the lowest scores (Kruskal-Wallis test p<0.001). Women had good knowledge of situations when pill efficacy is reduced but poorer knowledge of what action to take subsequently. Predictors of knowledge in pill users were educational level, age, and the importance attached to not falling pregnant-thus providing further evidence for the construct validity of the questionnaire. Test-retest reliability was good (rank correlation 0. 73). CONCLUSION: A contraceptive knowledge questionnaire suitable for use in audit or research has been developed which is reliable and has face, content and construct validity. Pill users have poor knowledge of what to do in situations of pill failure and about the details of emergency contraception. Trials are needed to assess the effectiveness of different strategies to improve contraceptive knowledge in women.