Comparing cancer stage at diagnosis between migrants and non-migrants: a meta-analysis

Talk Code: 
3C.2
Presenter: 
Adam Harvey-Sullivan
Twitter: 
Co-authors: 
Sana Ali, Parveen Dhesi, Joseph Hart, Helena Painter, Garth Funston, Dominik Zenner
Author institutions: 
Queen Mary University of London, University College London

Problem

Migrants worldwide face significant barriers in accessing healthcare services; resulting delays in provision of healthcare can lead to poor health outcomes. Stage at diagnosis is a key factor in determining cancer survival. Delays in cancer diagnosis can result in detection at more advanced stage that is harder to treat. Factors such as linguistic, cultural, systemic, and socio-economic barriers risk delays in cancer diagnosis for migrants. This study aims to compare the differences in cancer stage at diagnosis between migrant and non-migrant populations, hypothesizing that migrants are less likely to be diagnosed at an early stage.

Approach

We conducted a systematic review with meta-analysis (PROSPERO CRD42023385332) following the PRISMA guidelines. Three databases (MEDLINE, Embase, and Web of Science) and grey literature were searched. Eligible studies were published from January 2000 to January 2023, conducted in OECD countries, focused on symptomatic primary cancer diagnoses, and compared stage at diagnosis among adult migrants with non-migrants. Quality assessment was performed using the ROBINS-I tool. Meta-analyses were performed using a random-effects model to calculate pooled odds ratios (ORs) comparing early-stage diagnosis in these populations.

Findings

Our search strategy identified 11,549 articles. 41 studies met the inclusion criteria, and 34 were suitable for meta-analysis. Studies encompassed eight cancer types, eight countries and a variety of migrant populations. Overall, we found that migrants were significantly less likely to be diagnosed with early stage cancer compared to non-migrants (OR 0.85; 95% confidence interval (CI) 0.78-0.91). Stratified pooled estimates across seven cancer types consistently showed this association but it was statistically significant only for breast (OR 0.78; 95% CI 0.70-0.87) and prostate cancer (OR 0.92; 95% CI 0.85-0.99). Subgroup analysis by study location and migrant region or origin demonstrated variation but the overall trend of migrants being less likely to have early stage at diagnosis was maintained. Sensitivity analysis for imputed missing data and by cancer stage classification scheme did not change the interpretation of the outcomes.

Consequences

Our results demonstrate that migrants are less likely to be diagnosed with early stage cancer compared with non-migrants. This inequality can be attributed to multiple barriers to healthcare access. Study heterogeneity is a significant limitation that we mitigated through stratification and sensitivity analyses. Variations by cancer type highlight that different cancer-, patient-, clinician- and health system factors can influence the diagnostic pathway for each malignancy. Delays in cancer diagnosis for migrants have significant implications for clinicians and policymakers as late-stage diagnosis is not only associated with significant suffering, morbidity and mortality but also incurs greater healthcare costs. This underscores the urgent need for targeted interventions to improve early cancer diagnosis among migrant populations, such as culturally sensitive healthcare services, community outreach programs, and policy reforms to ensure equitable healthcare access.

Submitted by: 
Adam Harvey-Sullivan
Funding acknowledgement: 
Doctors Harvey-Sullivan, Painter, Dhesi and Ali are NIHR-funded Academic Clinical Fellows in Primary Care.