Urban violence and pacification in Rio's favelas: innovative approaches to measurement and quantifying health system impacts

Many cities, especially in low- and middle-income countries in Latin America, are plagued by high levels of urban violence. Urban violence affects the everyday life of citizens - with increased homicides, injuries, and crime, poorer mental health and higher levels of stress, and disrupted community culture and day-to-day activities. Much of this violence is located in urban slums where the provision of public services, including policing and healthcare, is a major challenge. Currently, one billion currently live in urban slums, which will increase to two billion by 2030.

High-quality health systems are crucial for ensuring everyone can access high quality healthcare without financial hardship - particularly for the urban poor and populations affected by urban violence. However, how urban violence affects the provision and performance of health systems is not well documented or researched. More research is needed on the best ways to measure urban violence in local areas to better characterise health system impacts. Many cities globally have introduced interventions to address urban violence, but there a few robust evaluations of these interventions on health systems performance. Better understanding how health systems perform when exposed to urban violence is vital for improving services, introducing healthcare policies to respond to the challenge of urban violence, and designing interventions to address weaknesses in the system.

This research project will explore the role of urban violence on health system performance in the city of Rio de Janeiro. Rio de Janeiro is a valuable setting for this type of research as there are high-levels of urban violence driven by drug-gangs and large inequalities which are concentrated in large urban slums (called favelas) where one-third of the population live. There have been investments in the city in recent years aiming to address urban violence and expand healthcare provision. This includes a pacification programme for gang-controlled slums and the opening of new primary care clinics. Understanding how urban violence influences and is influenced by these investments is the key aim of this research.

The research will cover four key areas. The first will examine innovative approaches to measuring urban violence in Rio de Janeiro. It will use data form multiple sources, including locations of homicides, app-based reports of violence, police data, or nearby school closures. The second will understand to what extent urban violence deterred the opening of healthcare clinics during a large expansion in primary health care in the city between 2008 and 2016. The third will explore the relationship between urban violence and health system performance and will test whether local episodes of violence affected the access or quality of healthcare. This will be measured through metrics such as clinic closures, home visits, opening hours, and staff turnover. The fourth area of research will understand if the city's pacification programme affected health system performance.

These research aims will deliver evidence that can be used by communities and policymakers both in Brazil and internationally in other cities burdened by urban violence. It will also form preparatory work for future research to better understand how the relationships between urban violence and health systems can impact population health and health inequalities.

Evidence gaps relating to health system functioning in areas of high urban violence are inhibiting progress towards universal health coverage (UHC) and the Sustainable Development Goals. This is particularly the case in cities in low- and middle-income countries in Latin America with high homicide rates and large urban slums. This study aims to address important research gaps by exploring innovative approaches to measuring urban violence and quantifying health system impacts. The research will be based in the city of Rio de Janeiro where there are large challenges from urban violence, and recent interventions aiming to address urban violence (a pacification programme) and expand primary healthcare. It will build on strong collaborations between Imperial College and Brazilian-based investigators and advance existing research on this globally important topic.

Four specific areas will addressed in the research:
i) multiple data sources, including policing data, geocoding homicides, app-based reports, and school closures, will be used to explore different approaches to measuring local exposure to urban violence;
ii) the impact of urban violence on clinic roll-out as part of a major investment in PHC between 2008-16 will be examined. Regression modelling will be used to adjust for other local factors;
iii) quasi-experimental fixed effects panel regression will be used to understand how episodes of local violence impact health service functioning, with metrics such as clinic closures, opening times, home visits used a proxies for healthcare access and quality;
iv) the pacification programme's impact on health system functioning will be evaluated using quasi-experimental method such as interrupted time series and event studies.