Table 2.

Minimum methodological criteria for comparing quality of care between public and private sector providers

Criterion	Rationale
Does the study have a viable comparison group?	Documenting good or poor quality of care without knowing how good the care is amongst equivalent providers in the community isn’t that informative.
Does the study control for patient mix?	Different types of providers are frequented by different types of patients. Alternately, the same patient may select provider sector based on what s/he perceives to be the problem. This can bias comparisons because the compared providers aren’t treating patients with the same type or severity of problems.
Does the study control for differences in provider mix?	Public sector providers generally have a narrow range of certified medical qualifications. Private sector includes people with those same qualifications plus pharmacists, traditional healers, informal providers with no medical training, and public providers moonlighting in private sector. Comparing providers without taking into account differences in qualifications confounds the effect of the sector with the effect of medical training.
Does the study control for differences in resources?	Different types of providers have different resources available to them – both financially and in terms of equipment and trainings, etc. Cost analyses should be included in the analysis of quality of care and should consider subsidies that providers receive as well as charges and costs to patients.