Linked data and inclusion health: Harmonised international data linkage to identify determinants of health inequalities

Abstract

A recent article in The Lancet establishing the principles of inclusion health, highlighted substantial gaps in our understanding of the drivers of health inequalities in socially excluded groups such as people with a history of incarceration, people who experience homelessness, sex workers, people with mental illness, and people who inject drugs¹. Cross-sectoral data linkage of electronic health records with services working with socially excluded groups was one of the key recommendations of this article. The magnitude of health disparities observed in people that experience social exclusion necessitates an international public health response and addressing the determinants of social exclusion has been identified as a key component of closing the gap of Indigenous disadvantage². This symposium will establish data linkage as a key component of the inclusion health and will complement the efforts of the Pan American Health Oranization's (PAHO) Commission on Equity and Health Inequalities in the Americas.

Traditional survey methodology is costly and often results in studies that are highly parochial in nature. Due to difficulties recruiting and retaining marginalized groups, these studies are commonly forced to adopt methodological concessions, often selecting the most convenient participants (i.e., selection bias) or incurring increased rates of loss-to-follow-up (i.e., attrition bias). Conversely, global studies aimed at modelling the burden of disease are often not sufficiently nuanced to answer specific inferential research questions. Data-linkage has the potential to overcome these common biases and limitations. Thus, harmonised international data-linkage studies are an important component of the inclusion health response to identify the determinants of health inequalities in socially excluded groups and inform the global inclusion health agenda.

This symposium will bring together facilitators from three countries with extensive experience conducting data linkage studies that generate evidence on health and social inequality in socially excluded groups. Using a current multinational study as an example, barriers to international data-linkage studies, methodological solutions, and distributed approaches to generating international comparative evidence will be presented. Innovative examples of cross-sectoral approaches to linkage with social service, correctional and national survey data will be discussed. The development of a novel framework for identifying social exclusion exposures and determinants of health inequalities typically not captured in administrative health data will also be discussed. The session will conclude with a discussion aimed at forming the foundation of an international data linkage project to address these current gaps identified in the inclusion health series and best practice for translation to policy and practice to address health disparities in socially excluded groups.

References

1. Aldridge et al. Morbidity and mortality in homeless individuals, prisoners, sex workers, and individuals with substance use disorders in high-income countries: a systematic review and meta-analysis. The Lancet. 2017;391(10117):241-250. https://doi.org/10.1016/S0140-6736(17)31869-X


2. Greenwood M et al. Challenges in health equity for Indigenous peoples in Canada. The Lancet. 2018;Epub ahead of print. https://doi.org/10.1016/S0140-6736(18)30177-6

Introduction

High rates of self-harm resulting in acute health service contact have been observed in adults released from prison. Contact with health services due to self-harm is a key intervention opportunity to prevent deleterious health outcomes. Little is known about subsequent mental health service contact after discharge from acute health services.

Objectives and Approach

We aimed to describe mental health service contact after discharge from acute health service contacts following self-harm in a representative sample of adults released from prison. Ambulance, emergency department (ED), hospital inpatient and ambulatory mental health service records were probabilistically linked to pre-release interview data. Self-harm events after release were identified from ICD codes and coded from case notes in ambulance, ED, and hospital records. We calculated the time between discharge from ambulance, ED, or hospital after self-harm and subsequent contact with mental health services. Factors predicting the likelihood of mental health service contact were examined using multivariate logistic regression.

Results

Of 1307 adults released from prison, 108 (8.3%) experienced 218 self-harm events resulting in acute health service use in the community. Of these presentations, 0%, 59%, and 50% of discharges from ambulance attendances, ED and hospital, respectively, had subsequent contact with a specialist mental health service within 7 days of that acute service contact. Mental health service contact within 7 days of acute service contact was positively associated with being female (adjusted odds ratio [AOR]: 3.27; 95%CI: 1.26-8.47) and being identified by prison staff as at risk for self-harm (AOR: 3.34; 95%CI: 1.29-8.62), and was negatively associated with dual diagnosis (AOR: 0.19: 95%CI: 0.06-0.61), substance use disorder only (AOR: 0.13; 95%CI: 0.04-0.48) and physical health functioning (AOR: 0.96; 95%CI: 0.92-0.99).

Conclusion/Implications

Almost half of adults with a recent history of incarceration discharged from acute health service after self-harm did not receive timely specialist mental health care. Improved integration of acute health services and ambulatory mental health services could improve outcomes for adults who present with self-harm.