Is the intention to vaccinate enough? Systematic variation in the value of timely vaccinations and preferences for monetary vs non-monetary incentives among pregnant women in southern Tanzania

Abstract

Background: Globally, approximately 19.7 million children remain under-vaccinated; many more receive delayed vaccinations. Sustained progress towards global vaccination targets requires overcoming, or compensating for, incrementally greater barriers to vaccinating hard-to-reach and hard-to-vaccinate children. We prospectively assessed pregnant women's valuations of routine childhood vaccinations and preferences for alternative incentives to inform interventions aiming to increase vaccination coverage and timeliness in southern Tanzania.

Methods: Between August and December 2017, 406 women in their last trimester of pregnancy were enrolled from health facilities and communities in the Mtwara region of Tanzania and asked contingent valuation questions about their willingness to vaccinate their child if they were (a) given an incentive, or (b) facing a cost for each vaccination. Interval censored regressions assessed correlates of women's willingness to pay (WTP) for timely vaccinations. Participants were asked to rank monetary and non-monetary incentive options for the timely vaccination of their children.

Findings: All women expected to get their children vaccinated according to the recommended schedule, even without incentives. Nearly all women (393; 96.8 %) were willing to pay for vaccinations. The average WTP was Tanzania Shilling (Tsh) 3,066 (95 % confidence interval Tsh 2,523-3,610; 1 USD ∼ Tsh 2,200) for each vaccination. Women's valuations of timely vaccinations varied significantly with vaccine-related knowledge and attitudes, economic status, and rural vs urban residence. Women tended to prefer non-monetary over monetary incentives for the timely vaccination of their children.

Interpretation: Women placed a high value on timely childhood vaccinations, suggesting that unexpected system-level barriers rather than individual-level demand factors are likely to be the primary drivers of missed vaccinations. Systematic variation in the value of vaccinations across women reflects variation in perceived benefits and opportunity costs. In this setting, nonmonetary incentives and other interventions to increase demand and compensate for system-level barriers hold significant potential for improving vaccination coverage and timeliness.

Clinicaltrialsgov protocol: NCT03252288.

Keywords: Childhood vaccinations; Contingent valuation; Incentives; Tanzania; Vaccination timeliness; Willingness to pay.

Fig. 1

Question format and potential response scenarios. Abbreviations: TSH – Tanzania shilling. Notes: $d^{c1,c2,c3}$ denotes response scenarios, with c_t describing the response (y = yes, n = no) to choice t ($t=1\cdots 3$). Depending on the response scenario, a woman’s valuation of timely vaccinations could be captured exactly (solid line; point data) or be known to fall within some interval (dashed line; interval data). At the time of the study 1 US Dollar was worth approximately TSH 2,200.

Fig. 2

Nearly universally positive willingness to pay for routine childhood vaccinations (southern Tanzania, 2017, N = 406). Notes: Abbreviations: Tsh. – Tanzania shilling. Groupings of responses in the ranges of 1 ≤ WTP ≤ 499 Tsh and ≥ 2000 Tsh are for visualization only. Exact values were used in the model estimation. At the time of the study 1 US Dollar was worth approximately Tsh 2,200. Two women answered yes to all 3 dichotomous choice questions and stated Tsh 1,000 as their maximum willingness to pay in the open-ended follow-up question; these two women are excluded from the bar chart.

Fig. 3

Estimated marginal effects on women’s WTP for timely routine childhood vaccinations (southern Tanzania, 2017, N = 406). Notes: Abbreviations: WTP–willingness to pay; Tsh–Tanzania shilling; CI–confidence interval. Sample average marginal effects were calculated based on estimates from a multivariable interval censored regression model predicting log(WTP) as a function of the covariates shown in Table 2, (see Methods for details). Circles represent point estimates; error bars represent 95 % confidence intervals. Estimates in 2017 Tanzanian shilling (Tsh). At the time of the study 1 US Dollar was worth approximately Tsh 2,200. For binary variables, the estimated marginal effects show the sample average expected change in WTP for a discrete change in the variable from 0 to 1. For continuous variables, the estimated marginal effects show the change in WTP when the variable increases from the lower to the upper bound of the respective interval (see also Fig. 4).

Fig. 4

Estimated marginal effects of women’s age, vaccination attitudes, household assets, and distance to care on their WTP for timely routine childhood vaccinations (southern Tanzania, 2017, N = 406). Notes: Abbreviations: WTP – willingness to pay; Tsh – Tanzania shilling. Sample average marginal effects were calculated based on estimates from a multivariable interval censored regression model predicting log(WTP) as a function of the covariates shown in Table 2, (see Methods for details). Circles represent point estimates; error bars represent 95 % confidence intervals. Estimates in 2017 Tanzanian shilling (Tsh). At the time of the study 1 US Dollar was worth approximately Tsh 2,200.

Fig. 5

Women's rankings of alternative incentive options for timely vaccinations (southern Tanzania, 2017, N = 406). Notes: Abbreviations: Tsh – Tanzania shilling. Women were asked to rank six potential incentives for each timely vaccination of their children, including (a) mobile phone credit of Tsh 2,000, (b) a pharmacy voucher valued at Tsh 2,000, (c) a lottery ticket with the chance of winning Tsh 20,000, (d) a free health check for the mother, (e) a birth certificate, and (f) a mobile money payment of Tsh 2,000. Women’s odds of winning the lottery or the costs of health checks (estimated at Tsh 8,000 in 2019) or birth certificates (Tsh 3,500 in 2020) were not specified.