Tying Community Engagement with Appropriate Technology at the Last Mile:

Ruchit Nagar w Preethi Venkat w Yale School of Public Health w New Haven, CT

Motivation

500,000 children under the age of five die from vaccine-preventable death in India every year. Low awareness around the importance of vaccines continues to be a driver for underimmunization. This awareness gap is coupled with a lack of data on patterns of health behavior, especially at the last mile. We sought out to develop a novel, culturally-tailored platform that could streamline the interface between patient and provider, report patient-centered, actionable data to health admins, and engage the community to elicit behavior change around maternal and child health.

Methods

•  A cluster randomized controlled trial was conducted in which 96 village health camps were randomized to three arms across 5 blocks in the Udaipur District, serviced by our partner NGO, Seva Mandir: •  Mothers given NFC sticker placed on the

immunization card (control group) •  Mothers given NFC pendant (P) •  Mothers given NFC pendant with voice call

reminder to come to camp (P+V) •  Completion and timeliness of administration of

DTP3 from birth was assessed using Kaplan-Meier curves and regression modeling.

•  A Kruskal-Wallis H test was used to assess discussion level around the system.

•  User satisfaction (ranked on a Likert scale) was analyzed using a chi-squared test.

•  Visibility of the data collection tool to key relatives (grandmothers and fathers) was modeled using logistic regression.

Results

•  Necklace is better retained than the paper record (2.1% lost or forgotten vs. 7.4% at first follow-up, 19.5% at second follow-up)

•  No significant difference in arms with respect to adherence from DTP1 through DTP3 within two monthly immunization camps, although effect estimates are positive

•  Mothers discussed the data collection tool with significantly more people in P compared to control, and in P+V compared to control (Figure 3)

•  The proportion of mothers who rated the necklace “very good” in P and in P+V was significantly higher than the proportion of control group mothers that rated the sticker “very good” (Figure 4).

•  Mother’s occupation, camp interruptions, and electricity access were significant predictors of adherence (Figure 5)

•  No significant change in overall immunization camp attendance rate, although effect estimates are positive

•  Dai ma identified as a highly networked member of the village (Figure 6)

•  P and P+V were significant predictors of the grandmother having seen the necklace, as per the mother (p < 0.001 for both). P was a significant predictor of the father having seen the data collection tool (p = 0.008). However, missing data limits the validity of the logistic regression models used to assess visibility.

Conclusion

Despite the null results, the fact that the necklace was well retained by the community suggests that it can be a valuable social symbol and community engagement tool. The use of a culturally appropriate wearable as the data collection tool is amplifying discussion around the KB system, but little impact of the voice calls is evident as call deployment was inconsistent. The significant levels of discussion and likeability of the necklace exhibited in this study, relative to the control group, warrant further research to assess what kinds of targeting strategies would be most effective in influencing health-seeking behaviors using the necklace. With more robust evidence that the KB system is at once well liked, visible, and can increase on-time vaccination adherence and health camp attendance, the system can be integrated at the local government level to improve maternal and child health on a much larger scale. Low power and short term follow-up may have masked true effects of the system, which will undergo refinement before a larger trial in 318 villages in the Udaipur District.

A Cluster Randomized Trial to Determine the Effectiveness of a Novel, Digital Necklace and Voice Reminder Platform on Increasing Infant Immunization Adherence among Mothers in Rural Udaipur, India

The main objective of this study was to test a digital NFC (Near Field Communication) pendant and voice call reminder system for effectiveness in improving DTP3 adherence within two monthly camps from DTP1 administration. An intermediate assessment of the randomized trial was also conducted to quantitatively examine secondary outcomes, indicative of whether the necklace promotes discussion, is well liked by mothers, and is visible to key community members.

Figure 1: Visual overview of Khushi Baby system (Khushi Baby, Inc.)

Figure 3: Boxplot of number of persons data collection tool was shown to and discussed with, by study arm

Figure 4: Mothers’ reactions to data collection tool, by study arm

11.1%

35.3% 36.4%

Mean: 1.00

Mean: 3.77 Mean: 2.82 p = 0.005

p = 0.008 p = 0.008

Figure 2: Workflow of the KB system, from the field to the cloud

Figure 6: Social Network of a representative village. Pink dots represent females, blue dots represent males. Green connections are health related ties; yellow connections are friendship ties; red connections are anti-health ties

Figure 5: Adjusted relative risks of finishing DTP3 within two successive camps of receiving DTP1

Methodology

Objectives