This indicator is calculated for each service delivery point (SDP) type, so it can be disaggregated by SDP sector (public or private), type (e.g., hospital, family planning [FP] clinic, pharmacy, or community health worker) and location (i.e., poor/not poor, rural/urban, and geographic region). It can also be disaggregated by method.

In most countries, the source of supply will vary substantially by type of method. Permanent methods, IUDs, and implants require a clinic-based facility (including mobile clinics). Pills are available through clinics in addition to commercial and community-based distributor (CBD) outlets. DepoProvera, once a clinic-based method, has been introduced into CBD programs and is available in pharmacies in some countries. Condoms and spermicides can be dispensed from any type of facility. Thus, data on source of supply are particularly useful when classified by method.

“Source of supply” yields two types of information: type of facility and type of sector (public/private). Type of facility generally includes hospital, health center, FP clinic, mobile clinic, pharmacy, field worker, private doctor, and shop, among others. Sector distinguishes between governmental programs and those in the private sector (including the local FP association, commercial retailers, private physicians, and other private providers). Ideally, data on source of supply should yield the percentage of contraceptive use attributable to the government program, the private FP association, the private sector (pharmacies, private doctors), and other relevant sources.

The distinction between public and private is often difficult to make, especially in countries with multiple sources of contraception. The respondent may incorrectly identify a given clinic as a government clinic, when in fact it is private (or she simply may not know if it is public or private). A private physician may in fact be participating in a subsidized program to offer low cost services to specific groups. In response to this problem, the DHS questionnaire provides a line for entering the actual name of the facility. Subsequent to the interview, a member of the research team codes the place mentioned according to the correct classification, based on master lists of service delivery points. To classify those not on the list, researchers can later contact key informants from the area.