The association of organizational contextual factors and HIV-Tuberculosis service integration following exposure to quality improvement interventions in primary healthcare clinics in rural KwaZulu-Natal.
MetadataShow full item record
A key strategy to reduce Tuberculosis (TB)-related mortality among people living with HIV is integrating HIV and TB diagnostic and treatment services. In South Africa, integrated HIV-TB service provision is standard of care, however, there is evidence that patients accessing primary healthcare clinics (PHC) are missed for HIV and TB testing and screening, diagnosis, linkage to treatment, and preventive services. Gaps in the HIV-TB care cascade are indicative of weaknesses in healthcare systems at the frontline. Quality Improvement (QI) collaboratives are a widely adopted approach to facilitating improvement among multiple clinics and scaling up best practices to improve on a given health topic. Little is known of the effectiveness of QI collaboratives and less is known of the role of organizational contextual factors (OCFs) in influencing the success of QI collaboratives to improve integrated HIV-TB services. Scaling up TB/HIV Integration (SUTHI) was a cluster-randomised trial designed to test the effectiveness of a QI intervention to enhance integrated HIV-TB services on mortality in HIV, TB, and HIV-TB patients. The study was from 01 December 2016-31 December 2018. Sixteen nurse supervisors (clusters) overseeing 40 PHC clinics were randomized (1:1) to receive either a structured QI intervention (QI group), which comprised, clinical training, three QI workshops timed at 6-month intervals, and in-person mentorship visits; or standard of care (SOC group) supervision and support for HIV-TB service delivery. This PhD project was a nested sub-study embedded in the SUTHI trial which aimed to describe and assess the influence of OCFs on the QI intervention to improve process indicators of HIV-TB services. A description of the QI intervention, including change ideas generated and lessons learned from practical application of the intervention in 20 QI clinics are presented in Paper I. Baseline performance of indicators was highlighted as important in influencing the size of improvements. OCFs that undermined the QI process were poor data quality, data capturing backlogs, lack of data analytic skills among clinic staff, poor transfer of training knowledge to peers, low clinic staff motivation to consistently track performance and limited involvement of the clinic management team in QI activities due to heavy workloads. A comparison between the QI and SOC group clinics showed that the QI intervention was only effective in improving two of five HIV-TB indicators, HIV testing services (HTS) andIsoniazid Preventive Therapy (IPT) initiation rates in new antiretroviral therapy patients. HTS was 19% higher (94.5% versus (vs) 79.6%; Relative Risk (RR)=1.19; 95% CI:1.02% - 1.38%; p=0.029) and IPT initiation was 66% higher (61.2% vs 36.8%; RR=1.66; 95% CI:1.02% -2.72%; p=0.044), in the QI group compared to the SOC group. Small clusters showed larger improvements in IPT initiation rates compared to big clusters, likely due to better coordination of efforts (Paper II). Several OCFs were quantitatively assessed and inserted into a linear mixed model to determine which factors likely influenced the improvement observed in the IPT initiation rates (Paper III). The practice of monitoring data for improvement was significantly associated with higher IPT initiation rates (Beta coefficient (β)=0.004; p=0.004). The main recommendations made from the PhD project are to encourage the practice of monitoring data for improvement among clinic teams; provision of widespread QI training for all levels of staff, different staff categories and leadership; to ensure good quality of routine data, and provision of regular performance feedback from upper management to the clinics.