Bulletin of the World Health Organization

The impact of user fees on health service utilization in low- and middle-income countries: how strong is the evidence?

Mylene Lagarde, Natasha Palmer

Volume 86, Number 11, November 2008, 839-848

Table 2. Quality assessment criteria applied to studies included in review of the effects of user fees on health service utilization in low- and middle-income countries

C-RCTs	Random allocation is clearly described (unit and process).
	Outcomes are measured at baseline.
	Outcome measures/data are reliable.
	There is no risk of exclusion bias.
	There is no risk of detection bias.
	There is no risk of contamination.
	The sampling strategy takes clusters into account (for C-RCTs).^a
	Appropriate statistical analysis is (well) performed.^a

CBA studies	Control and intervention sites are comparable.
	Outcome measures/data are reliable.
	There is no difference in outcomes between control and intervention sites at baseline.
	There is no exclusion or selection bias.
	There is no risk of contamination.
	Appropriate statistical analysis is (well) performed.^a

ITS studies	There is no risk that concurrent changes/events might have affected outcomes.
	There is no risk of selection bias.
	There is no risk of detection bias.
	Outcome measures/data are reliable.
	Time of the intervention is clearly defined.
	Appropriate statistical analysis (ARIMA model or time series regression) is performed.
	Rationale for the number of points in the series collected is stated (and sufficient to control for the effects of potential seasonal variations in outcomes before and after).^a

Overall quality assessment	There is a low risk of bias: all criteria are clearly met.
	There is a moderate risk of bias: one or two criteria are not clear or not met.
	There is a high risk of bias: more than two criteria are not clear or not met.

ARIMA, auto-regressive integrated moving average; CBA, controlled “before and after”; C-RCTs, cluster randomized controlled trials; ITS, interrupted time series.^a Criterion added by the authors to the list of those suggested by the Cochrane Collaboration’s Effective Practice and Organisation of Care group.