Table 1. Characteristics of studies included in the meta-analysis of the association of diarrhea morbidity with neighbourhood sanitation.
| Reference | Study design | Study location, period | Subjects (#, age) | Diarrhea outcome | Exposure | Risk of bias assessmenta,b | ||||
|---|---|---|---|---|---|---|---|---|---|---|
| E | O | S | A | P | ||||||
| Neighbourhood sanitation | ||||||||||
| Al-Ghamdi 2009[37] | Cross-sectional | Urban, Saudi Arabia, 2004–2005 | 1064 (7–12 yrs) | Incidence 1-month recall (yes/no) | No sewage spillage around house (yes/no*) | n | n | n | n | n |
| Anteneh & Kumie 2010[34]c | Cross-sectional | Rural, Ethiopia, 2006 | 447 (<5 yrs) | Incidence 2-week recall (yes/no) | No observable feces in the neighborhood yard (yes/no*) | n | y | y | n | - |
| Ferrer 2008[38] | Case-control | Urban, Brazil, 2002–2004 | 3364 (<10 yrs) | Clinic admission for diarrhea | No open sewage ditch nearby(yes/no*) | n | y | y | y | - |
| Graf 2008[42] | Cross-sectional | Urban, Nairobi, 2006 | 717 (<5 yrs) | Incidence 2-week recall (yes/no) | No rubbish and fecal material lying around, blocked open drains around home and nearby streets(1 scale increment/4 scales) | n | y | y | n | - |
| Heller 2003[40] | Case-cohort | Urban, Brazil, 1993–1994 | 1996 (<5 yrs) | Physician diagnosis of diarrhea | No wastewater in street (yes/no*) | n | y | n | y | n |
| Moraes 2003[43] | Cohort | Urban, Brazil, 1980–1990 | 683 (<5yrs) | Incidence 2-week recall, more than twice expected number of episodes | Communities with simplified sewerage and surface drainage vs. surface drainage only* | y | y | y | y | y |
| Household sanitation | ||||||||||
| Agustina 2013[35]c | Cross-sectional | Urban, Indonesia, 2004–2005 | 274 (12-59mths) | Defecation description 1-week record, categorized as diarrhea by field worker | Child feces disposal in latrine (yes/no*),. Presence of household latrine (yes/no*) | y | y | n | n | - |
| Anteneh & Kumie 2010[34]c | Cross-sectional | Rural, Ethiopia, 2006 | 447 (<5 yrs) | Incidence 2-week recall (yes/no) | Child defecation in latrine (yes/no*), Presence of functional household latrine (with sub & superstructure, providing service) (yes/no*) | y | y | y | n | - |
| Aziz 1990[19] | NRCTd | Rural, Bangladesh, 1984–1987 | 1359 (<5yrs) | Incidence 1-week recall (yes/no) | Child defecation or feces disposal in latrine | y | y | y | y | - |
| Baltazar 1989[44] | Case-control | Urban, Philippines, 1985 | 665 (< 2yrs) | Clinic admission for diarrhea | Child defecation or feces disposal in latrine (yes/no*) | y | y | n | n | y |
| Clasen 2014[41]c | CRCTd | Rural, India, 2010–2013 | 3835 (<5 yrs) | Prevalence 1-week record | Presence of functional household latrine (with roof; pan not broken; no hindrance for usage) (yes/no*) | y | y | y | y | y |
| Daniels 1990[45] | Case-control | Rural, Lesotho, 1987–1988 | 1613 (<5 yrs) | Clinic admission for diarrhea | Presence of household latrine (yes/no*) | y | y | y | y | y |
| Dessalegn 2011[46] | Cross-sectional | Rural/urban, Ethiopia, 2009 | 768 (<5 yrs) | Incidence 2-week recall (yes/no) | Presence of household latrine (yes/no*) | n | y | y | y | y |
| Dikassa 1993[47] | Case-control | Urban, Zaire, 1988 | 214 (<3 yrs) | Clinic admission for diarrhea | Child feces disposal in latrine (yes/no*) | y | y | y | y | y |
| Garrett 2008[20] | NRCTd | Rural, Kenya, 2001 | 960 (<5 yrs) | Incidence 1-week recall (yes/no) | Presence of household toilet (yes/no*) | y | y | y | y | n |
| Godana 2013[36]b | Case-control | Rural, Ethiopia, 2013 | 593 (<5 yrs) | Diarrhea incidence 2-week recall (yes/no) | Disposal of infant feces in latrine (yes/no*), Presence of household latrine (yes/no*) | y | y | y | n | y |
| Knight 1992[48] | Case-control | Rural, Malaysia, 1989 | 196 (4–59 months) | Clinic admission for diarrhea | Presence of household toilet (yes/no*) | y | y | n | y | y |
| Mbonye 2004[49] | Cross-sectional | Rural, Uganda, 2001 | 323 (<2 yrs) | Incidence 2-week recall (yes/no) | Presence of household pit latrine (yes/no*) | n | y | y | y | - |
| Mertens 1992[50] | Case-control | Rural, Sri Lanka, 1987–1988 | 3694 (<5 yrs) | Clinic admission for diarrhea | Child defecation in latrine or covered pit vs. open defecation* | y | y | n | y | n |
| Mihrete 2014[51] | Cross-sectional | Rural/urban, Ethiopia, 2012 | 925 (<5 yrs) | Diarrhea incidence 2-week recall (yes/ no) | Presence of household toilet (yes/no*) | y | y | y | y | - |
| Oketcho 2012[52] | Case-control | Rural, Tanzania, 2011 | 303 (<5 yrs) | Clinic admission for diarrhea | Child defecation in toilet or latrine (yes/no*) | y | y | n | y | - |
| Traore 1994[53] | Case-control | Urban, Burkina Faso, 1991 | 2793 (<3 yrs) | Clinic admission for diarrhea | Child feces disposal in latrine vs. elsewhere (yard, etc.)* | y | y | y | y | n |
| Tumwine 2001[39] | Cross-sectional | Rural/urban, East Africa, 1997 | 1015 (All age) | Incidence 1-week recall (yes/no) | Presence of household latrine (yes/no*) | y | y | y | n | - |
a. E: Adequate measure of exposure O: Adequate measure of outcome; S: Appropriate sample and/or study group selection; A: Adjustment for child age, household socioeconomic status, and water or hygiene; P: Adequate participation or follow-up rate
b. y:yes; n:no;-:uncertain
c. Multiple datasets are extracted
d. NRCT: Cluster non-randomized controlled trials, CRCT: Cluster randomized controlled trials.
*Reference state.