Conducting Research at the Intersection of HCI and Health: Building and Supporting Teams with Diverse Expertise to Increase Public Health Impact.

Research at the intersection of human-computer interaction (HCI) and health is increasingly done by collaborative cross-disciplinary teams. The need for cross-disciplinary teams arises from the interdisciplinary nature of the work itself-with the need for expertise in a health discipline, experimental design, statistics, and computer science, in addition to HCI. This work can also increase innovation, transfer of knowledge across fields, and have a higher impact on communities. To succeed at a collaborative project, researchers must effectively form and maintain a team that has the right expertise, integrate research perspectives and work practices, align individual and team goals, and secure funding to support the research. However, successfully operating as a team has been challenging for HCI researchers, and can be limited due to a lack of training, shared vocabularies, lack of institutional incentives, support from funding agencies, and more; which significantly inhibits their impact. This workshop aims to draw on the wealth of individual experiences in health project team collaboration across the CHI community and beyond. By bringing together different stakeholders involved in HCI health research, together, we will identify needs experienced during interdisciplinary HCI and health collaborations. We will identify existing practices and success stories for supporting team collaboration and increasing HCI capacity in health research. We aim for participants to leave our workshop with a toolbox of methods to tackle future team challenges, a community of peers who can strive for more effective teamwork, and feeling positioned to make the health impact they wish to see through their work.

The Potential Contribution of the Health System to Reducing Stunting in SUN Countries.

BACKGROUND: The global nutrition community has called for a multisectoral approach to improve nutritional outcomes. While most essential nutrition interventions are delivered through the health system, nutrition-sensitive interventions from other sectors are critical. OBJECTIVE: We modeled the potential impact that Scaling Up Nutrition (SUN) interventions delivered by the health system would have on reaching World Health Assembly (WHA) stunting targets. We also included results for targets 2, 3, and 5. METHODS: Using all available countries enrolled in the SUN movement, we identified nutrition interventions that are delivered by the health system available in the Lives Saved Tool. We then scaled these interventions linearly from 2012 up to nearly universal coverage (90%) in 2025 and estimated the potential impact that this increase would have with regard to the WHA targets. RESULTS: Our results show that only 16 countries out of 56 would reach the 40% reduction in the number of stunted children by 2025, with a combined total reduction of 32% across all countries. Similarly, only 2 countries would achieve the 50% reduction in anemia for women of reproductive age, 41 countries would reach at least 50% exclusive breastfeeding in children under 6 months of age, and 0 countries would reach the 30% reduction in low birth weight. CONCLUSIONS: While the health system has an important role to play in the delivery of health interventions, focusing investments and efforts on the health system alone will not allow countries to reach the WHA targets by 2025. Concerted efforts across multiple sectors are necessary.

Nutrition data use and needs: Findings from an online survey of global nutrition stakeholders.

BACKGROUND: There is growing global demand for country-specific information to track nutritional status and its determinants, including intervention coverage. Periodic population-based surveys form the backbone of most national nutrition information systems. However, data on the coverage of many nutrition specific and sensitive interventions remain sparse. METHODS: An online survey was administered to the international nutrition community in 2018 through relevant listservs and professional networks to characterize their use of nutrition-related indicators and data sources. Respondents were asked about their professional background, access and use of specific indicators and data sources in the previous year, and unmet data needs. Results were tabulated by respondent characteristics and χ2 tests used for statistical testing. RESULTS: Complete survey responses were received from 235 respondents, the majority from non-governmental organizations and research communities, and few from governments. Demographic Health Surveys (DHS) were the most frequently accessed country-specific data source and the Global Nutrition Report (GNR) was the most accessed consolidated data source, each accessed by approximately 75% of respondents. Respondents with a multi-country focus were more likely to have accessed DHS than those with a single-country focus (85% vs 60%, P < 0.001). Similarly, respondents with a multi-country focus were more likely to have accessed the GNR compared to those with a single-country focus (82% vs 66%, P < 0.05). The most commonly accessed indicators overall were the prevalence of exclusive breastfeeding (69%), child minimum dietary diversity (66%), under-5 stunting (65%), and under-5 wasting (65%). Reported data gaps included adult and household diet quality indicators (n = 32), nutrition-sensitive intervention coverage (n = 25), and infant and young child feeding promotion coverage (n = 11). Lack of data availability for the desired geographic level (82%) or demographic group of interest (82%) and out-of-date data (77%) were common data challenges experienced by respondents. CONCLUSIONS: The survey results highlight the continued need for high-quality, actionable nutrition data to help facilitate progress towards national and global nutrition targets.

Understanding client and provider perspectives of antenatal care service quality: a qualitative multi-method study from Tanzania.

BACKGROUND: Measures of quality of care in low- and middle-income countries (LMICs) rarely include experience of care. This gap in service quality metrics may be driven by a lack of understanding of client and provider perspectives. Understanding these perspectives is a critical first step in not only improving metrics, but also in improving service delivery. This study identifies the items antenatal care (ANC) clients and health care providers in Tanzania associate with a quality ANC service and explores the experience of care domain from both client and provider perspectives. METHODS: We conducted semi-structured interviews with15 providers and 35 clients in Tanzania that included a free-listing activity to elicit items clients and providers associate with quality ANC services. We analyzed the free-listing for rank order and frequency to identify the most salient items, which were included in the second phase of data collection. We then conducted semi-structured interviews with a pile sort activity with the same 15 providers and 32 new clients to understand the importance of the items identified in the free-listing. We used a thematic analysis driven by the framework approach to analyze interview data. RESULTS: Both clients and providers perceived quality of ANC as being comprised of items related to experience of care, provision of care, and cross-cutting essential physical and human resources. The free-listing findings illuminated that the experience of care was equally important to clients and providers as the availability of physical and human resources and the content of the care delivered. In addition, clients and providers perceived that a positive patient care experience - marked by good communication, active listening, keeping confidentiality, and being spoken to politely - increased utilization of health services and improved health outcomes. CONCLUSIONS: The experience of care in LMICs is an overlooked, yet critically important topic. Understanding the experience of care from those who receive and deliver services is key to measuring and improving the quality of ANC. Our research highlights the importance of incorporating experience of care into future quality improvement activities and quality measures. By doing so, we identify barriers and facilitating factors of practical use to policy-makers and governments in LMICs.

Optimizing data visualization for reproductive, maternal, newborn, child health, and nutrition (RMNCH&N) policymaking: data visualization preferences and interpretation capacity among decision-makers in Tanzania.

BACKGROUND: Reproductive, maternal, newborn, child health, and nutrition (RMNCH&N) data is an indispensable tool for program and policy decisions in low- and middle-income countries. However, being equipped with evidence doesn't necessarily translate to program and policy changes. This study aimed to characterize data visualization interpretation capacity and preferences among RMNCH&N Tanzanian program implementers and policymakers ("decision-makers") to design more effective approaches towards promoting evidence-based RMNCH&N decisions in Tanzania. METHODS: We conducted 25 semi-structured interviews in Kiswahili with junior, mid-level, and senior RMNCH&N decision-makers working in Tanzanian government institutions. We used snowball sampling to recruit participants with different rank and roles in RMNCH&N decision-making. Using semi-structured interviews, we probed participants on their statistical skills and data use, and asked participants to identify key messages and rank prepared RMNCH&N visualizations. We used a grounded theory approach to organize themes and identify findings. RESULTS: The findings suggest that data literacy and statistical skills among RMNCH&N decision-makers in Tanzania varies. Most participants demonstrated awareness of many critical factors that should influence a visualization choice-audience, key message, simplicity-but assessments of data interpretation and preferences suggest that there may be weak knowledge of basic statistics. A majority of decision-makers have not had any statistical training since attending university. There appeared to be some discomfort with interpreting and using visualizations that are not bar charts, pie charts, and maps. CONCLUSIONS: Decision-makers must be able to understand and interpret RMNCH&N data they receive to be empowered to act. Addressing inadequate data literacy and presentation skills among decision-makers is vital to bridging gaps between evidence and policymaking. It would be beneficial to host basic data literacy and visualization training for RMNCH&N decision-makers at all levels in Tanzania, and to expand skills on developing key messages from visualizations.

Youth accessing reproductive health services in Malawi: drivers, barriers, and suggestions from the perspectives of youth and parents.

BACKGROUND: Malawi has made progress in increasing its overall modern contraceptive prevalence rate since 2000, resulting in a dramatic reduction in its total fertility rate. However, youth, 15-24 years, have not had the same successes. Teenage pregnancies are on the rise and little progress has been made in reducing unmet need for family planning among youth. With two-thirds of the population under the age of 25 and with Malawi's rapid population growth, reducing unmet need for family planning among youth remains a priority for the government's reproductive health agenda. To further explore this situation, we conducted a qualitative study to explore the perspectives of youth and adults about the drivers and barriers to youth accessing family planning in Malawi and their ideas to improve services. METHODS: We conducted 34 focus group discussions with youth aged 15-24 and parents or legal guardians of female youth in 3 districts in Malawi. Focus groups were translated and transcribed. Data was input into Dedoose and analyzed using a thematic framework to identify broader patterns and themes. RESULTS: Youth participants felt motivated to use family planning to protect themselves from sexually transmitted diseases and to prevent unwanted pregnancies. Females focused on the consequences of unplanned pregnancies and believed family planning services were targeted primarily at them, while males thought family planning services targeted males and females equally. Barriers to youth accessing family planning included contraception misconceptions, the costs of family planning services, and negative attitudes. Parents had mixed views on family planning. While many parents acknowledged they could play a role in supporting youth, most said they are reluctant to support youth using family planning. Participants said improving counseling services, integrating family planning services and education within school curricula, and utilizing youth clubs could improve family planning services for youth. CONCLUSIONS: Policy makers and program implementers should consider the diverse preferences among youth and parents and continue seeking their input when designing policies and programs. Youth clubs and school-based services were among the most common suggestions. However, the effectiveness of youth clubs and school-based initiatives to increase contraceptive use among youth in Malawi is not clear.

Explaining progress towards Millennium Development Goal 4 for child survival in Tanzania.

BACKGROUND: During the Millennium Development Goal (MDG) era (1990-2015) the government in Mainland Tanzania and partners launched numerous initiatives to advance child survival including the comprehensive One Plan for Maternal Newborn and Child Health in 2008-2015 and a "sharpened" One Plan strategy in early 2014. Moving into the Sustainable Development Goal era, the government needs to learn from successes and challenges of striving towards MDG 4. METHODS: We expand previous work by presenting data for the full MDG period and sub-national results. We used data from six nationally-representative household surveys conducted between 1999 and 2015 to examine trends in coverage of 22 lifesaving maternal, newborn, child health and nutrition (MNCH&N) interventions, nutritional status (stunting; wasting) and breastfeeding practice across Mainland Tanzania and sub-nationally in seven standardized geographic zones. We used the Lives Saved Tool (LiST) to model the relative contribution of included interventions which saved under 5 lives during the period from 2000-2015 compared to 1999 on a national level and within the seven zones. FINDINGS: Child survival and nutritional status improved across Mainland Tanzania and in each of the seven zones across the 15-year period. MNCH&N intervention coverage varied widely and across zones with several key interventions declining across Mainland Tanzania or in specific geographical zones during all or part the period. According to our national LiST model, scale-up of 22 MNCH&N interventions - together with improvements in breastfeeding practice, stunting and wasting - saved 838 460 child lives nationally between 2000 and 2015. CONCLUSIONS: Mainland Tanzania has made significant progress in child survival and nutritional outcomes but progress cannot be completely explained by changes in intervention coverage alone. Further examination of the implementation and contextual factors shaping these trends is important to accelerate progress in the SDG era.

Household economic and food security after the 2010 Pakistan floods.

BACKGROUND: The 2010 floods inundated one-fifth of Pakistan and affected more than 20 million people. OBJECTIVE: To characterize the impact of the floods and subsequent humanitarian response on household economy and food security. METHODS: A cross-sectional 80 x 20 cluster survey (n = 1,569 households) was conducted using probability proportional to size sampling in the four most flood-affected provinces 6 months after the floods. Analysis included both descriptive statistics and regression models, with receipt of food aid (in the first month), dietary quality, and household income at 6 months postflood as outcomes. RESULTS: Need for food aid was nearly ubiquitous (98.9%); however, only half of the study population ever received food aid. Displacement was not a significant predictor of food aid receipt (OR, 1.28; 95% CI, 0.83 to 1.98); however urban location (OR, 2.78; 95% CI, 2.00 to 3.86) and damage to the home (OR, 2.73; 95% CI, 1.34 to 5.60) were significantly associated. Some of the hardest-hit groups, including both farmers and day laborers, were significantly less likely to receive food aid (p < .05). Additionally, receipt of food aid was not necessarily associated with improved household economy or food security; although households in internally displaced people (IDP) camps were more likely to receive food aid (OR, 2.78; 95% CI, 2.00 to 3.86), they were less likely to report same or improved dietary quality (OR, 0.63; 95% CI, 0.46 to 0.88) or income status (OR, 0.55; 95% CI, 0.35 to 0.86). CONCLUSIONS: Food aid coverage following the 2010 floods was relatively low, and many of the most affected populations were less likely to receive aid, suggesting that targeting should be improved in future responses.