Bloom 2006.

Methods	Cluster‐randomised trial; 4 districts each were randomised to intervention and control arms
Participants	The study was conducted in rural districts in Cambodia, and served a population of about 1.26 million people ‐ adults and children (11% of Cambodia's population).
Interventions	NGPs were contracted to provide all preventive, promotional, and basic curative healthcare services mandated for a district by the Ministry of Health. These NGPs were contracted to deliver specific services and corresponding targets at health facilities. The programme was implemented over 4 years ‐ between 1999 and 2003.
Outcomes	We reported the following outcomes (all measured over a 12 month period): immunisation of children 12 to 24 months old, high‐dose vitamin A to children 6 to 59 months old, antenatal visits, birth deliveries by trained professionals, female use of contraceptives, use of district public healthcare facilities when sick, mortality in the past year of children younger than 1 year, incidence of diarrhoea in children younger than 5 years, government healthcare expenditures Health information: accuracy of facility registers Availability of selected essential medicines: availability of child vaccines at facilities over the previous 3 months Health financing: individual healthcare expenditures (ITT; important benefit) (measured over a 12 month period)
Notes	Contextual factors The intervention followed a period of political instability with a health system that was dysfunctional, mainly because of poor training, and had widespread corruption, seen as government‐employed health workers being absent during work and doctors running private practices on the side and diverting patients to their private practices. Contracts were based on competitive bidding, and only international non‐governmental organisations were successful.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	This is not described in the paper.
Allocation concealment (selection bias)	Unclear risk	This is not described in the paper.
Baseline outcome measurements	Low risk	The numbers reported in baseline measures appear to be similar across intervention and control arms; however no details of statistical differences are provided.
Baseline participant characteristics	High risk	Characteristics are not reported in text or tables.
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Blinding is not described, and it is unlikely that it was done. It is unclear whether this would affect the performance of participants or personnel.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	It is unclear whether the assessors ‐ those who conducted baseline and post‐intervention surveys ‐ were blinded to whether participants belonged to intervention or control arms.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	This is not described in the paper.
Protection against contamination	Low risk	Allocation to intervention and control arms was done at the district level, and it unlikely that the control group received the intervention.
Selective reporting (reporting bias)	Unclear risk	The study protocol could not be accessed.
Other bias	Unclear risk	Review authors did not identify other risks of bias.