Date of Award

January 2024

Document Type

Thesis

Degree Name

Master of Public Health (MPH)

Department

School of Public Health

First Advisor

Michaela A. Dinan

Abstract

Objective: Patients with cancer are at elevated suicide risks compared to the general population, and patients diagnosed with advanced cancers have higher suicide risks compared to those diagnosed with early-staged cancers. Immune Checkpoint Inhibitors (ICIs) were initially approved to treat metastatic melanoma, non-small cell lung cancer (NSCLC), and renal cell carcinoma (RCC), and an unpublished study observed a decreased suicide risk in patients with these cancers at all stages. This study evaluates suicide risks in patients with breast, colorectal, liver, pancreatic, and prostate cancer, of which ICIs have limited effects on, to assess if the lower suicide rates observed previously could be attributed to cancer-specific availability of ICIs. We hypothesize that potentially curative therapy will provide increased hope for patients with previously incurable disease and therefore be associated with decreased rates of death by suicide only in cancers which ICIs are effective, especially metastatic or advanced cancer.

Methods: Data was obtained from the Surveillance, Epidemiology and End Results (SEER) database for patients diagnosed with breast, colorectal, liver, pancreatic, and prostate cancer between 2004 to 2010 (pre-period) and 2016 to 2019 (post-period). Difference-in-difference models were used to compare the difference in suicide rates during the pre- and post-periods. Multiple chi-square tests were performed to describe the baseline characteristics. In the aggregated level analysis, standardized mortality ratios of suicide were compared across periods (post vs. pre) and cancer stages (regional/advanced vs. localized). In the individual-level analysis, a Poisson regression was fitted to identify the association between suicide risk and period.

Results: A total of 1,394,567 cases were identified from SEER, of which 0.2% died from suicide (N=2,097) and 99.8% died from other causes or were still alive (N=1,392,470). Among all patients, 61.7% were diagnosed with localized cancers (N=860,159), 26.0% were diagnosed with regional cancers (N=363,010), and 12.3% were diagnosed with advanced cancers (N=171,398). Pre-period cases accounted for 76.5% of all patients (N=1,066,346) and post-period cases accounted for 23.5% (N=360,621). Our aggregated-level model suggested that SMR of suicide for patients with regional/advanced cancers during the post-period of ICI approval was significantly higher. The individual-level model showed that suicide risk among patients with regional/advanced cancers was higher after the approval of ICIs at a borderline significance.

Conclusions: Our findings suggest that suicide rates increased overall among cancer patients. We did not observe evidence that ICI adoption was associated with reduced suicide risk in patients with advanced-stage disease for which ICI treatments are not as effective. Future research may link SEER with claims data to examine suicide attempts instead of suicidal deaths and focus on sex-specific cancers to clarify whether the increased risk of suicide is associated with the sex difference.

Comments

This thesis is restricted to Yale network users only. It will be made publicly available on 05/07/2026

Share

COinS