Table 1.

Factors thought to be associated with skilled delivery service use in the literature

Determinant*	Rationale	Findings
Sociocultural factors
Maternal age +++	Older women: more experienced in using services, more confident, more say in household. Young women: more modern.	No difference, or older women more likely to use services in all multivariate studies examined.
Marital status ++	Single mothers more autonomous: more use. But maybe poorer and stigmatized: less use.	No association or either direction.
Ethnicity, religion, traditional beliefs +++	Certain cultural backgrounds, beliefs, norms and values as well as discrimination may decrease care-seeking.	Mixed results. Large differences in some studies, none in others.
Family composition +	Small children at home and no extended family to help should decrease use.	Some found less skilled care if higher number of births in previous five years.
Mother's education +++	Knowledge, access to written information, modern culture, more confident, higher earnings, control over resources, better communication with husband and providers, etc. should all increase use.	Consistently strong and dose-dependent positive effect on delivery service use.
Husband's education ++	Knowledge, modern attitudes, better communication between spouses, higher autonomy for wife, higher earnings, etc. should increase service use.	Higher husband's education consistently increases skilled attendance; effect often smaller than effect of mother's own education.
Women's autonomy ++	Decision-making power, mobility, control over resources, access to transport should increase use.	Most found some aspects to increase skilled attendance, but others found no effect.
Perceived need
Information availability +	Information about risks of childbirth and about service availability in radio or TV should increase use.	Information access associated with more skilled attendance in some studies but not in others.
Health knowledge +	Knowledge about risks of childbirth and the benefits of skilled care should increase wish to use services.	Expected association in some but not in other studies.
Pregnancy wanted +	Higher value attached to desired child justifies expenses for skilled attendance.	Expected association in some but not in other studies.
Perceived quality of care +	Perceived poor personal and medical quality of care, clash with culture and fear of procedures may decrease use.	Qualitative studies generally find that perceived low quality decreases use, some describe interaction with distance and cost. Very few quantitative studies.
ANC use ++	Familiarity with services, encouragement by health workers increases delivery service use.	Usually those attending ANC much more likely to receive skilled delivery care.
Previous facility delivery ++	Familiarity with services increases their use.	Nearly always very strongly associated with index facility delivery.
Birth order +++	First birth: more difficult, help from natal family, high value on pregnancy, or unplanned/unwanted. High order births: previous experience, confidence if no problems previously, difficulty to leave home with several small children, poorer families.	No difference or first births and lower order births more likely to have skilled attendance than high order births in the vast majority of studies examined.
Complications +	Pregnancy complications (→ ANC advice), complications during delivery, previous complications (→ women aware, medical risk) should increase use of skilled attendance.	Qualitative studies: important factor, decreases importance of other barriers. Few quantitative studies, several found that women with complications are more likely to seek skilled care.
Economic accessibility
Mother's occupation +	Own earnings, range of movement, information should increase use. Decreased use expected if work is poverty-induced.	No effect in several studies, association in either direction. Often less use of skilled attendance among women farmers.
Husband's occupation ++	Higher financial resources and health insurance with some occupations should increase service use.	In several but not in all studies increased skilled attendance if higher status occupations.
Ability to pay +++	Costs for transport, care, opportunity costs decrease use by the poor.	Poorer women less likely to have skilled attendance, in some studies no effect.
Physical accessibility
Region, urban/rural +++	Social and service environment differences between regions. In rural areas generally worse services and infrastructure, more poverty, more traditional beliefs, which all decrease use.	Nearly always moderate to large differentials with less service use in rural areas.
Distance, transport, roads ++	Distance as disincentive and actual obstacle to reach facilities, enhanced by lack of transport and poor roads.	Less service use when further away or no difference.

* Frequency of inclusion in quantitative studies: + rarely, ++ sometimes, +++ usually