Fig. 3: Total burden and risk of self-harm after diagnosis of psychiatric disorders in patients with cancer.

Case (with psychiatric disorder) and control (no psychiatric disorder) groups were obtained via propensity score matching (Extended Data Fig. 1). Controls were matched by age at cancer diagnosis, cancer type, sex, IMD and primary care practice ID. a, Cumulative burden of all self-harm events in cases and controls. b, Cumulative incidence of the first self-harm event in cases and controls. Gray’s test was used to assess statistical significance. c, HRs for risk of self-harm for each psychiatric disorder were further adjusted for noncancer comorbidities, cancer treatment and presence of other psychiatric disorders. The numbers of patients with psychiatric disorder were as follows: depression, 21,609; anxiety disorder, 20,070; schizophrenia, 7,679; bipolar disorder, 557; personality disorder, 194; substance abuse, 115,868. The numbers of matched controls for each psychiatric disorder were as follows: depression, 75,087; anxiety disorder, 67,887; schizophrenia, 23,130; bipolar disorder, 1,636; personality disorder, 628; and substance abuse, 126,057. Self-harm risks during the first 12 months and subsequent years of follow-up are shown. Strata with low event numbers (n < 10) were not analyzed. Data are presented as HRs, and error bars represent 95% CIs. Numbers in the graphs in panel c represent P values. The likelihood ratio test was used. Full data and 95% CIs are provided in the supplementary tables.