Abstract

The risk of suicide in cancer patients has been reported as elevated in several countries. These patients are exposed to many medicines that may confuse or provide a means for harm, potentially also increasing their risk from accidents. Ratios of observed/expected numbers of hospital admission and death events relating to deliberate self-harm (DSH) and accidents were calculated in the 5 years from a cancer diagnosis in Scotland 1981-1995, compared to the matched general population. The relative risk (RR) of suicide was 1.51 (95% confidence interval (CI): 1.29-1.76). The RR of hospital admissions for DSH was not significantly increased, suggesting a strong suicidal intent in DSH acts in cancer patients. Accidental poisonings and all other accidents were both increased (RR death=3.69, 95% CI: 2.10-6.00; and 1.58, 95% CI: 1.48-1.69, respectively) (RR hospital admissions=1.32, 95% CI: 1.19-1.47; and 1.55, 95% CI: 1.53-1.57, respectively). The association of only certain tumour types (e.g. respiratory) with suicide and accidental poisoning, and a broad range of tumour types with an elevated risk of all other accidents, suggests accidental poisoning categories may be a common destination for code shifting of some DSH events. A previous history of DSH or accidents, significantly increased the RR of suicide or fatal accidents, respectively (RR suicide=14.86 (95% CI: 4.69-34.97) vs 1.16 (95% CI: 0.84-1.55)) (RR accidental death=3.37 (95% CI: 2.53-4.41) vs 1.29 (95% CI: 1.12-1.49)). Within 5 years of a cancer diagnosis, Scottish patients are at increased RR of suicide and fatal accidents, and increased RR of hospital admissions for accidents. Some of these accidents, particularly accidental poisonings, may contain hidden deliberate acts. Previous DSH or accidents are potential markers for those most at risk, in whom to target interventional techniques.