Suicide risk among cancer survivors: Head and neck versus other cancers.

Abstract

146 Background: While head and neck cancer (HNC) only accounts for 4% of new cancer cases in the United States, incidence of HNC-associated suicide may be significantly higher due to quality of life issues associated with HNC survivorship. This study estimated incidence of HNC-associated suicide versus other common cancers, and quantified suicide rate among HNC survivors compared with non-HNC. Methods: We queried the Surveillance, Epidemiology and End Results 18 database from 2000-2014 for all cancer deaths confirmed as suicide. Mortality rates from suicide was estimated for the 21 most common cancers in the United States, including HNC. Negative binomial regression estimated mortality rate ratios (MRRs) and 95% confidence intervals (CIs), between HNC and the other 20 other cancers combined, and as individual sites. Models were stratified by sex (when applicable), controlling for race, marital status, age, year and stage of diagnosis. Results: There were 4,769 suicides observed among 4,613,123 cancer survivors from 2000-2014, yielding an incidence rate of 23.6 suicides per 100,000 person-years. Combined, all other cancers had a 45% decreased suicide ratio compared to HNC for both males (MRR = 0.55, 95% CI 0.48, 0.64) and females (MRR = 0.55, 95% CI 0.37, 0.81). For specific cancer sites, only pancreatic cancer had a higher mortality rate due to suicide than HNC (86.4 suicides per 100,000 person-years, versus 63.4 suicides per 100,000 person-years). Stratified by gender, only male pancreatic cancer survivors had a significantly higher suicide MRR compared with HNC (MRR = 1.54, 95% CI 1.23, 1.90). For females, all cancer sites (including pancreas) had either significantly or insignificantly lower suicide MRR compared with HNC. Conclusions: Risk of suicide is significantly higher among HNC survivors compared to all other cancers, except pancreatic cancer. Our findings confirm that HNC survivors are prime candidates for lifelong psychosocial surveillance.