Table 1.

Summary of research sites and samples

	Cambodia	Sierra Leone	Uganda	Zimbabwe
Site selection	Six provinces (covering all four ecological regions)—one district from each, including urban, rural and those with more or less external support. The RinGS project was in one province, covering two operational districts	Four districts (covering all main regions)	Three districts in Acholi sub-region—most conflict-affected area	Two provinces—one well served and one under-served; three districts including urban, mixed and rural. The RinGs study was done in four districts in the Midlands Province
Sectors included	Public sector only	Public sector only	Public sector and private not-for-profit	Public (government, municipal and rural district council employees), mission and private sector
Timeframe	1979 onwards	2000 onwards (last phase of conflict; post-conflict since 2002)	2000 onwards (6 years during; 6 years after conflict)	1997 onward (economic crisis, and post-since 2009)
Research tools and participant numbers (by gender, where relevant)
1. Stakeholder mapping		23 (7 f/16 m)	17 (3f/14m)
2. Document review	59	57	59	76
3. Key informant interviews	19 (all male)	23 (10 f/13 m)	25 (12 f/13 m)	28 (13 f/15 m) RinGs N=11; 5f/6m
4. Life histories/in-depth interviews with health workers	19 (doctors, medical assistants, nurses, midwives) 14 f/5 m RinGS project: 20 (14f/6m)	23 (doctors, nurses, midwives, community health officers—CHOs) 12f/11m	26 (clinical officers, nurses, nursing assistants, midwives and others) 19f/7 m	34 (doctor, nurses, midwifes environmental health practitioners and clinical officers) 32 f/3 m RinGs study: N=19; 11f/8m
5. Quantitative analysis of routine data	√	√		√
6. Survey of health workers		310 (doctors, CHOs, nurses, environmental health officers. MCH Aides, Lab and pharmacy technicians) 178 f/132m		227 (doctors, clinical officers nurses, midwives, environmental health technicians) 127 f/100m; RinGs Study N=140 (nurses, midwives and environmental health technicians); 83f/57m