RT Journal Article
SR Electronic
T1 Quality of care for tuberculosis and HIV in the private health sector: a cross-sectional, standardised patient study in South Africa
JF BMJ Global Health
JO BMJ Global Health
FD BMJ Publishing Group Ltd
SP e005250
DO 10.1136/bmjgh-2021-005250
VO 6
IS 5
A1 Jody Boffa
A1 Sizulu Moyo
A1 Jeremiah Chikovore
A1 Angela Salomon
A1 Benjamin Daniels
A1 Ada T Kwan
A1 Madhukar Pai
A1 Amrita Daftary
YR 2021
UL http://gh.bmj.com/content/6/5/e005250.abstract
AB Background South Africa has high burdens of tuberculosis (TB) and TB-HIV, yet the quality of patient care in the private sector is unknown. We describe quality of TB and TB-HIV care among private general practitioners (GPs) in two South African cities using standardised patients (SPs).Methods Sixteen SPs presented one of three cases during unannounced visits to private GPs in selected high-TB burden communities in Durban and Cape Town: case 1, typical TB symptoms, HIV-positive; case 2, TB-specified laboratory report, HIV-negative and case 3, history of incomplete TB treatment, HIV-positive. Clinical practices were recorded in standardised exit interviews. Ideal management was defined as relevant testing or public sector referral for any reason. The difference between knowledge and practice (know-do gap) was assessed through case 1 vignettes among 25% of GPs. Factors associated with ideal management were assessed using bivariate logistic regression.Results 511 SP visits were completed with 212 GPs. Respectively, TB and HIV were ideally managed in 43% (95% CI 36% to 50%) and 41% (95% CI 34% to 48%) of case 1, 85% (95% CI 78% to 90%) and 61% (95% CI 73% to 86%) of case 2 and 69% (95% CI 61% to 76%) and 80% (95% CI 52% to 68%) of case 3 presentations. HIV status was queried in 35% (95% CI 31% to 39%) of visits, least with case 1 (24%, 95% CI 18% to 30%). The difference between knowledge and practice was 80% versus 43% for TB and 55% versus 37% for HIV, resulting in know-do gaps of 37% (95% CI 19% to 55%) and 18% (95% CI −1% to 38%), respectively. Ideal TB management was associated with longer visit time (OR=1.1, 95% CI 1.1 to 1.2), female GPs (3.2, 95% CI 2.0 to 5.1), basic symptom inquiry (2.0, 95% CI 1.7 to 2.3), HIV-status inquiry (OR=11.2, 95% CI 6.4 to 19.6), fewer medications dispensed (OR=0.6, 95% CI 0.5 to 0.7) and Cape Town (OR=2.2, 95% CI 1.5 to 3.1). Similar associations were observed for HIV.Conclusions Private providers ideally managed TB more often when a diagnosis or history of TB was implied or provided. Management of HIV in the context of TB was less than optimal.All data relevant to the study are de-identified and included in the article or uploaded as supplementary information.