|Author(s)||Mary Kawonga1, Sharon Fonn2, Duane Blaauw3.
|Affiliation(s)||1Department of Community Health, Wits School of Public Health, Johannesburg, South Africa, 2School of Public Health, Wits School of Public Health, Johannesburg, South Africa, 3Centre for Health Policy, Wits School of Public Health, Johannesburg, South Africa.|
|Country - ies of focus||South Africa|
|Relevant to the conference tracks||Health Systems|
|Summary||With increasing global focus on the integration of vertical programmes within health systems, methods are needed to analyse whether general health service (horizontal) managers at district level exercise administrative authority over disease programmes (administrative integration). This study adapts "decision space" analysis to measure administrative integration of HIV programme monitoring and evaluation (M&E). The study shows that horizontal managers exercise high degrees of authority in producing HIV information but not in using it for decisions, while vertical managers use HIV information but in silos. The lack of M&E integration may undermine district health system strengthening aims.|
|Background||In South Africa, integration is a health sector reform priority, while several vertical programmes exist, notably for HIV, tuberculosis (TB), and maternal and child health (MCH). Historically a national HIV/AIDS directorate and specialist HIV managers have vertically managed the HIV programme and HIV programme managers account for ear-marked HIV programme funding through dedicated parallel reporting mechanisms. This is at odds with current health sector decentralisation reforms that envisage integrated management of health services under the control of generalist (horizontal) managers at a decentralised district level. National health policy envisions health districts as the foundation of the national health system.|
|Objectives||If health districts are to be the foundation of the health system as envisaged, then horizontal district managers would need to be allocated and to exercise authority over district health services, including disease-specific interventions (i.e. administrative integration). This study examines whether this is happening in the South African health system. We use the HIV programme as a case study given its traditionally vertical approach, and focus on the M&E (information) function as a tracer for analysing administrative integration. The research aims are to:
1. Describe the extent to which horizontal managers exercise authority over HIV M&E coordination.
2. Explore factors associated with exercised authority.
|Methodology||The research explores two hypotheses: a) vertical managers exercise higher degrees of authority than horizontal managers in administering HIV M&E; and higher management capacity and HIV M&E knowledge are associated with higher degrees of exercised authority. This cross-sectional study was conducted in two of South Africa’s nine provinces. Fifty one participants were interviewed including: a) managers primarily responsible for general health services or general health information (horizontal manager) and b) those responsible for vertical services or information (vertical manager). HIV M&E was defined as the production of HIV information (HIV data collection, collation, analysis) and use of HIV information for decisions.In the absence of existing methods for measuring 'exercised authority' over vertical programmes, Bossert's decision-space’ approach provided a useful frame. Since decision-space analysis has not been applied to either the M&E function or in the context of programme integration, it was adapted to measure ‘exercised (administrative) authority’ in this study. We defined exercised authority as a manager undertakes tasks to oversee HIV data collection, collation and analysis, and uses HIV data to review the programme and take action. To measure this, we first identified the M&E tasks that managers were expected to perform within each M&E domain (collection, collation, analysis, use) and then administered a semi-structured questionnaire to collect data on participants’ performance of these tasks. We developed four sub-scales to measure the degree of exercised authority for each M&E domain. Sub-scales comprised several items (M&E tasks), which we coded ‘no’ if a respondent did not perform the task (score zero) or ‘yes’ if s/he did. We computed an ordinal dependent variable for each HIV M&E domain and coded observed scores as ‘low’, ‘medium’, and ‘high’. We performed ordinal logistic regression to explore whether explanatory variables (actor type [horizontal or vertical], health system level, highest, qualification, duration of management experience, management capacity score, and M&E knowledge score) were predictive of higher degrees of exercised authority.|
|Results||More than 75% of participants were female, with an undergraduate degree or higher, and had some management
training. Participant characteristics were similar between vertical and horizontal managers. Horizontal managers attained higher mean scores for HIV data collection.
|Conclusion||In light of the increasing focus on health system strengthening and integration, our research makes a contribution
by providing a method and scales for measuring and monitoring administrative integration. We anticipate that
these scales will be strengthened further by empirical testing on larger samples and varied settings. In applying this
method to South Africa’s public sector HIV programme, we find that HIV M&E coordination is generally not administratively integrated, characterised by horizontal managers exercising little authority in using HIV data, and vertical managers using HIV data in sub-programme silos. We argue that this programme model potentially undermines South Africa's policy aims of integrated management of district health services under the authority of horizontal general health service managers. The research also indicates that plans for integrating the HIV programme within the health system at decentralised district level should include investments in strategies to equip horizontal managers with the knowledge and skills to use programme data for decision-making.