Closing the gap on institutional delivery in northern India: a case study of how integrated machine learning approaches can enable precision public health.

INTRODUCTION: Meeting ambitious global health goals with limited resources requires a precision public health (PxPH) approach. Here we describe how integrating data collection optimisation, traditional analytics and causal artificial intelligence/machine learning (ML) can be used in a use case for increasing hospital deliveries of newborns in Uttar Pradesh, India. METHODS: Using a systematic behavioural framework we designed a large-scale survey on perceptual, interpersonal and structural drivers of women's behaviour around childbirth (n=5613). Multivariate logistic regression identified factors associated with institutional delivery (ID). Causal ML determined the cause-and-effect ordering of these factors. Variance decomposition was used to parse sources of variation in delivery location, and a supervised learning algorithm was used to distinguish population subgroups. RESULTS: Among the factors found associated with ID, the causal model showed that having a delivery plan (OR=6.1, 95% CI 6.0 to 6.3), believing the hospital is safer than home (OR=5.4, 95% CI 5.1 to 5.6) and awareness of financial incentives were direct causes of ID (OR=3.4, 95% CI 3.3 to 3.5). Distance to the hospital, borrowing delivery money and the primary decision-maker were not causal. Individual-level factors contributed 69% of variance in delivery location. The segmentation analysis showed four distinct subgroups differentiated by ID risk perception, parity and planning. CONCLUSION: These findings generate a holistic picture of the drivers and barriers to ID in Uttar Pradesh and suggest distinct intervention points for different women. This demonstrates data optimised to identify key behavioural drivers, coupled with traditional and ML analytics, can help design a PxPH approach that maximise the impact of limited resources.

Investigating Voluntary Medical Male Circumcision Program Efficiency Gains through Subpopulation Prioritization: Insights from Application to Zambia.

BACKGROUND: Countries in sub-Saharan Africa are scaling-up voluntary male medical circumcision (VMMC) as an HIV intervention. Emerging challenges in these programs call for increased focus on program efficiency (optimizing program impact while minimizing cost). A novel analytic approach was developed to determine how subpopulation prioritization can increase program efficiency using an illustrative application for Zambia. METHODS AND FINDINGS: A population-level mathematical model was constructed describing the heterosexual HIV epidemic and impact of VMMC programs (age-structured mathematical (ASM) model). The model stratified the population according to sex, circumcision status, age group, sexual-risk behavior, HIV status, and stage of infection. A three-level conceptual framework was also developed to determine maximum epidemic impact and program efficiency through subpopulation prioritization, based on age, geography, and risk profile. In the baseline scenario, achieving 80% VMMC coverage by 2017 among males 15-49 year old, 12 VMMCs were needed per HIV infection averted (effectiveness). The cost per infection averted (cost-effectiveness) was USD $1,089 and 306,000 infections were averted. Through age-group prioritization, effectiveness ranged from 11 (20-24 age-group) to 36 (45-49 age-group); cost-effectiveness ranged from $888 (20-24 age-group) to $3,300 (45-49 age-group). Circumcising 10-14, 15-19, or 20-24 year old achieved the largest incidence rate reduction; prioritizing 15-24, 15-29, or 15-34 year old achieved the greatest program efficiency. Through geographic prioritization, effectiveness ranged from 9-12. Prioritizing Lusaka achieved the highest effectiveness. Through risk-group prioritization, prioritizing the highest risk group achieved the highest effectiveness, with only one VMMC needed per infection averted; the lowest risk group required 80 times more VMMCs. CONCLUSION: Epidemic impact and efficiency of VMMC programs can be improved by prioritizing young males (sexually active or just before sexual debut), geographic areas with higher HIV prevalence than the national, and high sexual-risk groups.

Toward a Systematic Approach to Generating Demand for Voluntary Medical Male Circumcision: Insights and Results From Field Studies.

By the end of 2014, an estimated 8.5 million men had undergone voluntary medical male circumcision (VMMC) for HIV prevention in 14 priority countries in eastern and southern Africa, representing more than 40% of the global target. However, demand, especially among men most at risk for HIV infection, remains a barrier to realizing the program's full scale and potential impact. We analyzed current demand generation interventions for VMMC by reviewing the available literature and reporting on field visits to programs in 7 priority countries. We present our findings and recommendations using a framework with 4 components: insight development; intervention design; implementation and coordination to achieve scale; and measurement, learning, and evaluation. Most program strategies lacked comprehensive insight development; formative research usually comprised general acceptability studies. Demand generation interventions varied across the countries, from advocacy with community leaders and community mobilization to use of interpersonal communication, mid- and mass media, and new technologies. Some shortcomings in intervention design included using general instead of tailored messaging, focusing solely on the HIV preventive benefits of VMMC, and rolling out individual interventions to address specific barriers rather than a holistic package. Interventions have often been scaled-up without first being evaluated for effectiveness and cost-effectiveness. We recommend national programs create coordinated demand generation interventions, based on insights from multiple disciplines, tailored to the needs and aspirations of defined subsets of the target population, rather than focused exclusively on HIV prevention goals. Programs should implement a comprehensive intervention package with multiple messages and channels, strengthened through continuous monitoring. These insights may be broadly applicable to other programs where voluntary behavior change is essential to achieving public health benefits.