We used data from the Tanzania DHS (2015-2016), a nationally representative, cross-sectional household survey with a two-stage cluster sampling design [19, 20]. Using this sampling strategy, 13,360 households were selected for the survey, of which 12,563 were occupied and completed the interview (98% response rate). In the interviewed households, 13,634 eligible women (age 15-49) were identified for individual interviews using the Woman’s Questionnaire, of which 13,266 women completed the interview (97% response rate). The Woman’s Questionnaire includes questions about the woman’s use of modern contraceptives, the source of the method, and the amount paid for it. The DHS report provides additional details about the sampling procedure, data collection, and specific interview questions [5].
We analyzed data for women who self-reported currently using a modern contraceptive [10]. Due to small sample sizes, we excluded those using male sterilization (n=3), female condom (n=2), emergency contraception (n=3), the DHS FP method category “other modern method” (n=1), and the Standard Days Method (n=2).
In Tanzania, FP is provided at a variety of facility types, which range in size from large hospitals, to medium-sized health centers, to small dispensaries, all of which are found in both the private and public sector. To capture this variation in facility types, we grouped women’s self-reported most recent source of modern FP method into seven categories: government hospital, government health center, government dispensary, private facility (including private hospital/clinic, private nursing/maternity home), non-governmental organization/faith-based facility, pharmacy/chemist, and other (including shop, mobile clinic, friend/relative, other, community health worker, community-based distributor, other private medical) [10]. We considered the first three categories to be public sector (government-provided services) and the latter four categories to be private sector [7, 10]. We excluded observations with missing data for most recent source of modern FP method (n=37).
The Tanzania 2015-2016 DHS did not collect cost data for all types of methods. Therefore, to analyze informal payments, we limited this analysis to women who self-reported currently using a modern contraceptive method for which data on price were collected; this included the IUD, injectable, implant, pill, and male condom. Observations with missing values for informal payments were removed (n=145). One woman responded with “do not know” for informal payments amount and was excluded. Self-reported payment values were inspected for improbable values. No values were found to be greater than 10 times the 95th percentile, the criterion specified for improbable values by Radovich and colleagues [10].
Several sociodemographic characteristics were included in our analysis. Quintile of household wealth was determined by the DHS based on measures of household assets appropriate for the Tanzanian context. Highest level of educational attainment was categorized as no education, primary, and secondary or higher. Due to sample size limitations, binary age was used, categorized as less than 30 years and 30 or more years.
Three main variables were of interest in our analysis: modern FP method, source of modern FP method, and payment for modern FP method. Payment for modern FP method was measured using two questions: 1) “Did you pay for (CURRENT METHOD)?”; 2) “How much did you pay for (CURRENT METHOD)?” [5]. We described modern contraceptive method mix by wealth quintile and source of FP. Injectable and implant use was reported by wealth quintile and source of FP. We then analyzed the proportion of modern contraceptive users reporting free FP planning by public sector facility type. Finally, we characterized the amount paid for current method among injectable and implant users who reported a non-zero payment for their current method at a public sector facility. Portions of our analysis (figure 3 and Table 2) were limited to injectables and implants, as these were the only methods dispensed as a single unit and for which there was a sample size of at least 30 participants with non-zero payment for their current method, once stratified by facility type. Self-reported informal payments for FP method among injectable and implant users was reported in Tanzanian Shillings (TSh), with one USD being equal to approximately TSh 2,300. Analyses were conducted in SAS version 9.4 (SAS Institute, Inc., Cary, North Carolina).