Table 4 Generic lessons learnt from operational research with cotrimoxazole preventive therapy in Malawi

There were high case fatality rates of TB patients on anti-TB treatment alone, and thus a need for HIV-specific interventions

There were high early death rates of people living with HIV starting antiretroviral treatment

Research questions must be relevant to programme needs.

Operational research leadership and coordination must be placed within the programme.

Research on cotrimoxazole was endorsed by MoH, and district studies were designed and implemented in conjunction with national programme staff

Research should be endorsed and designed with programme MoH staff in order to increase the probability of findings and recommendations from the study being accepted and implemented

Research was carried out at district or facility level using routine systems; data were collected using registers and treatment cards; all patients were included with no special inclusion and exclusion criteria

Research can and should be effectively carried out within programme settings and routine health services

Key actors or "policy entrepreneurs" in the programmes helped to move forward the process of policy making

National meetings were held to engage all stakeholders, to obtain "buy-in" of the results and to get advice and direction as to how to move forward

Publication of results in international-peer reviewed journals brought credibility to findings as a result of the peer-review process, and allowed dissemination of results internationally

Key actors or "policy entrepreneurs" must be identified and given the task of moving forward the policy process

When research is completed, dissemination must occur nationally, and if judged of wider importance then internationally as well

Publication of operational research in peer-reviewed journals adds credibility to the study findings

Clear policy decisions were obtained from MoH about the study findings, and directives given about how to implement the new interventions

Research should influence national policy and practice

Policy documents were prepared and widely distributed through circulars around the country

National Guidelines were updated with new evidence and new policy

Monitoring tools were prepared and disseminated; drug forecasting was integrated into established processes; training materials were developed and used at different levels; uptake of new interventions were reported in national quarterly reports

Programmes need to implement the new policy and practices

Key actors and "policy entrepreneurs within programmes play an important role in this process

International guidelines or a road-map need to be developed to better direct the national steps that logically help move research to policy and practice

There was a clear demonstration of impact in reducing case fatality and increasing treatment success in TB patients, and in reducing early death rates in people with HIV starting ART

The ultimate benefit is an impact on programme performance and treatment outcomes