Table 2 GEHIP PHIT implementation progress: success, challenges, adaptations.

Successes

Development of a toolkit for budgeting and finances. With technical assistance from the University of Dar es Salaam Computing Centre and the Ifakara Health Institute, GEHIP re-engineered Tanzania’s PLANREP budgeting toolkit for Ghana. This involved utilizing Navrongo Health Research System generated burden of disease data and Government of Ghana accounting procedures to develop a model for translating budget plans into visualized data on their burden of disease implications: The District Health Planning and Reporting Toolkit (DiHPART).

Development of simplified HMIS data capture and data utilization procedures GEHIP provided direct support to a reform of the national “District Health Information Management System Version-2” by providing community-level components of the reformed system. This involved streamlining registers, testing their application, revising content, and developing feedback tools, data visualization, and supervisory leadership mechanisms for improving the use of HMIS for decision-making.

Mortality auditing. All frontline workers have been trained to produce simple-to-interpret narrative reports on all known maternal and neonatal deaths. A medical review panel has been constituted to conduct weekly reviews of these audits.

Training of frontline workers in emergency management All frontline workers were trained to refer deliveries and care for newborn needs. Missing elements of emergency management were identified by GEHIP mortality audit scheme with training instituted in response to problems.

Improved collaboration and co-financing of community based primary health care by district and local government. All district managers were trained in CHPS implementation and equipped to use DiHPART for orienting district political leaders to the benefits of CHPS investment. Improved understanding of the health benefits of CHPS has catalysed incremental funding for implementing community-based care. CHPS coverage accelerated as a consequence.

Challenges

DiHPART. Trial of the system demonstrated that categories used for data visualization are inconsistent with decision-making options. Also, changes in the national accounting system are not yet reflected in the DiHPART tool. DiHPART requires re-engineering based on lessons learned.

Cash flow and planning. DiHPART assumes that the district level common fund is available for managers to allocate according to plans. Long delays in the allocation of Common Fund revenue challenge that assumption. Actual expenditure patterns differ from budget parameters because of unpredictable flow of essential revenue.

Fidelity to proven operational models. Implementation of the National Health Insurance System (NHIS) has dysfunctionally shifted the focus of care to clinic-based services, detracting from outreach and operational strategies that have been proven to work.

Excess mortality Audits have revealed excess neonatal mortality from asphyxia, and maternal mortality from convulsions, and haemorrhaging.

Effective supervision at the community level. Research shows that supervisory field encounters are the main factor affecting community health worker performance. However, NHIS reimbursement policies reimburse supervisory staff for clinical services rendered. Supervisory field work has diminished.

Timing of Systems Changes. As Figure 3 shows, GEHIP has taken time to implement: changing the leadership system, accelerating the flow of resources from the development sector, and the implementing community health services has taken 18 months of project time that impacted on CHPS coverage in project Years 2 and 3. The full child survival impact of GEHIP will be realized well after the project is completed.

Adaptations

DiHPART. While DiHPART was conceived as a resource allocation tool, its value to GEHIP has shifted somewhat to resource development. Visualizing the health benefits of investment in CHPS has facilitated district dialogue about ways for development revenue to bridge critical resource gaps. The absence of budget lines for CHPS start-up costs prevents the expansion of community-based primary health care, leadership training, field demonstration, and DiHPART have been combined into a paradigm for multi-sectoral investment in CHPS expansion.

Excess mortality. High neonatal and maternal mortality from preventable causes has fostered strategic planning about training needs, referral systems development, and information systems reform. Research findings and plans have been translated into fund-raising initiatives that are successfully augmenting GEHIP with resources for addressing excess mortality.

Fidelity to proven CHPS strategies. GEHIP has been a mechanism for systems diagnosis, documentation, and national dialogue about implementation lapses in the national CHPS program. Simple to implement corrective measures that are instituted by GEHIP are developing treatment districts into a national learning platform for health systems development. The UER, in turn, is becoming a learning region for guiding national strategies for achieving universal health coverage, accessible care, and comprehensive community health services.

Timing of Systems Changes. The national CHPS monitoring system lapsed in 2008 with the conclusion of external funding arrangements. In response, GEHIP has developed reformed CHPS coverage monitoring tools that enable the project to conduct longitudinal observation of the scale-up of CHPS implementation (as illustrated by Figure 3). New monitoring tools, integrated into DHIMS-2, will enable the GHS to have real-time access to information about systems development progress and lapses.