Table 2. Example of factors affecting decisions about strength of recommendations—Lay or community health workers to reduce childhood mortality.

Population: Children in high mortality settingsIntervention: Lay health workers (LHWs) delivering health promotion, treatment, and referral interventionsComparison: No LHW intervention / usual careOutcome: Childhood mortality
Key factors—is there uncertainty regarding:	Decision regarding whether there is uncertainty (yes / no)	Explanation of the decision made
Quality of evidence	Yes	The use of LHWs in maternal and child health programmes may lead to fewer deaths among children under five (low quality evidence—GRADE). In addition, the use of LHWs probably leads to an increase in the number of women who breastfeed and to the number of children who have their immunisation schedule up to date (moderate quality evidence for both outcomes— GRADE). These additional outcomes are also related to mortality reduction
Balance of benefits versus harms and burdens	Yes	Potentially important benefits (mortality reduction) but confidence interval also includes harm. Additional evidence on LHWs suggests effectiveness, e.g., LHWs associated with increased uptake of interventions of proven cost-effectiveness (immunisation, breastfeeding)
Acceptability	Yes	• Some evidence that LHWs acceptable to service users and used widely• Varied acceptability to other services providers in different settings (e.g., [43],[44])
Resource use	Yes	Potentially large investment needed over long period but alternatives likely to be more costly
Feasibility (or local factors that influence the translation of evidence into practice)	Yes	There may be constraints to scaling up trained LHWs and supporting them, but it is even less feasible to scale up professional cadres. There are a number of well-documented examples of LHW programmes that have been taken to scale for which monitoring has suggested some positive outcomes, e.g., in Ethiopia and Pakistan [45],[46]
Recommended options for consideration This assessment of evidence within a wider health system context might result in the following recommended options for consideration:• Option 1: Where child mortality is high; an infrastructure for LHWs can be developed rapidly; it is unlikely that the numbers of other cadres could be expanded; and this cadre is acceptable to other providers and to service users and has strong political support:○ Strong recommendation to implement LHWs to reduce childhood mortality (i.e., there is confidence that the desirable effects of LHWs delivering interventions to reduce childhood mortality outweigh the undesirable effects).• Option 2: Where child mortality is high; LHWs are acceptable to other providers and to service users; but governance and financing mechanisms for LHWs will need to be established, and there is little experience of running such programmes and uncertainty among policy makers:○ Conditional recommendation to implement LHWs to reduce childhood mortality (i.e., the desirable effects of LHWs delivering interventions to reduce childhood mortality probably outweigh the undesirable effects, but there is uncertainty).• Option 3: Where child mortality is moderate; an infrastructure for LHWs can be developed rapidly; but there is evidence that the scaling up of LHWs may be challenged by health care professionals:○ Conditional recommendation to implement LHWs to reduce childhood mortality, dependent on the ability to overcome professional opposition (i.e., the desirable effects of LHWs delivering interventions to reduce childhood mortality probably outweigh the undesirable effects, but there is uncertainty).

Source: This table draws on evidence from [47].