Study Setting, sample Knowledge measure Outcome
Dwamena[119] Accra, Ghana, 2012. Systematic sample of outpatients at anticoagulation clinic of a teaching hospital. Own tool, adapted from Taylor[121], pass rate set at 70% 112/175 (64%) passed. Better knowledge was associated with better INR control.
Assefa[82] Addis Ababa, Ethiopia, 2014. Outpatients on warfarin at teaching hospital. Own tool, adapted from Anticoagulation Knowledge Assessment (AKA)[122], pass rate set at 75% 18/130 (14%) passed. Mean score was 59%. Association between knowledge and INR control not reported.
Mariita[10] Nairobi, Kenya, 2016. Consecutive sample at cardiac, cardiothoracic, and hemato-oncology clinics of a teaching hospital. Own tool, adapted from Oral Anticoagulation Knowledge test (OAK)[123], pass rate set at 75% 15/147 (10%) passed. Mean score was 57%. Knowledge was not associated with INR control.
Iqbal[116] Nairobi, Kenya, 2017. Convenience sample at cardiac, cardiothoracic, hemato-oncology, and DVT clinics of a teaching hospital. Own tool, adapted from OAK[123], “satisfactory knowledge” set at >70% 12/45 (27%) had satisfactory knowledge pre-intervention. Association between knowledge and INR control not reported.
Hutheram[124] Gauteng, South Africa, 2016. Convenience sample at ten private sector INR clinics attached to a private pathology company. Own tool, adapted from OAK[123], pass rate set at 50% 31/34 (91%) passed. Association between knowledge and INR control not reported.
Samadoulougou[125] Ouagadougou, Burkina Faso, 2014. Convenience sample of patients in the cardiology clinic of a university hospital. Own tool, adapted from Janoly-Duménil[126] Participants scored low in questions relating to their ability to anticipate and make decisions in risky situations. Association between knowledge and INR control not reported.
Maramba[32] Harare, Zimbabwe, 2018. Convenience sample of outpatients with thrombophilia on long-term warfarin. Not clear 29/47 (62%) were not aware of the need for regular check-ups. Association between knowledge and INR control not reported.
Gregersen[120] Johannesburg, South Africa, 2006. Convenience sample of women of childbearing age with valvular heart disease, who had at least one pregnancy while on warfarin. Own questionnaire 38/124 (31%) were not using contraception; misperceptions about “the contraceptive effect of warfarin” were not uncommon. Knowledge about effects of warfarin on the fetus was often inaccurate, and not specific or detailed. Association between knowledge and INR control not reported.