Table 3.

Percentage of studies in each risk category based on the quality of the sample [34]. U is the total population size, and N is the sample size in relation to U.

Origin of the sample	Sample	Risk level	Percentage of studies
(1) When the sample is determined by the total number of people or an age interval	(b) When N is less than 80% of the necessary value for its representation with a margin of error of up to 5%.	High	10
	(b) When N is less than 80% of the necessary value for its representation with a margin of error of up to 5%. (c) When there is no information about U or N.	High	2
	(c) When there is no information about U or N.	High	56
	(a) When U is equal to N.	Low	3
	(b) When N is representative of U with a randomized sample and a margin of error of up to 5%.	Low	3
	(a) When N is extracted from U with a randomized sample and a margin of error greater than 5% but less than 10%.	Moderate	3
(2) When the sample is based on heads of household (one or two per household)	(b) When N is less than 80% of the value necessary to represent the heads of household with a margin of error of up to 5%.	High	2
(3) When the sample is based on households	(b) When N is less than 80% of the value necessary to represent the households with a margin of error of up to 5%.	High	3
	(b) When N is less than 80% of the value necessary to represent the households with a margin of error of up to 5%. (c) When there is no information on the number of households or N.	High	2
	(c) When there is no information on the number of households or N.	High	7
	(b) When, in the representative number of homes, one of the household members is interviewed, with a randomized sample and a margin of error of up to 5%.	Low	2
(4) When the sample is intentionally focused on an interest group (e.g., midwives, herbalists, or local specialists)	(d) In cases of local specialists, when there is no indication of the total, but the snowball technique is used to select the principal people with knowledge.	Moderate	2
(5) When participatory methods are used	(b) When there is no information about the size of the population or group in question, but information about the number of participants is provided.	Moderate	2
(6) Diffuse selection criteria	(a) When there is no information on N or U.	High	5

A total of 80% of the complete (100%) sample is used with a margin of error of less than 5%.