Predictors of Homicide-Suicide Affirmation in Acute and Chronic Pain Patients

Objectives. 1) Determine and compare prevalence for homicide-suicide (H-S) affirmation in community non-patients (N = 478), community patients (N = 158), acute pain patients (APPs; N = 326), and chronic pain patients (CPPs; N = 341); and 2) Develop H-S predictor models in APPs and CPPs. Design. A large set of items containing the H-S item was administered to the above groups, who were compared statistically for H-S endorsement. APPs and CPPs affirming the H-S item were compared with those not affirming on all available variables including the Battery for Health Improvement (BHI 2) with significant variables (P < 0.001) utilized in predictor models for H-S in APPs and CPPs. Setting. Community plus rehabilitation facilities. Results. The above population groups affirmed the H-S item according to the following percentages: healthy community 1.88%, community patients 3.16%, rehabilitation patients without pain 3.64%, rehabilitation AAPs 3.99%, and rehabilitation CPPs 4.40%. For both APPs and CPPs, the H-S item was significantly correlated with some suicidality items and some homicide items. The model for APPs identified "having a suicide plan" as being predictive of H-S affirmation. For CPPs, the items of having thoughts of revenge killing, being motivated to seek revenge without any verbal warning, and the Doctor Dissatisfaction Scale of the BHI 2 predicted H-S affirmation. The APPs model classified 96% of the APPs correctly, while the CPPs model classified 97% of the CPPs correctly. These predictor rates, however, were no better than the base rate. Conclusion. The prevalence of H-S affirmation within APPs and CPPs is not insignificant. The APPs predictor model points to a close association between H-S affirmation and suicidality. The CPPs model indicates that there is a close association between H-S affirmation, and anger/hostility and anger directed at physicians. These results, however, should not lead to the belief that CPPs are at greater risk for actual H-S completion for the following reasons: 1) H-S is an extremely rare event; and 2) predictive validity of the H-S item for actual H-S completion has not been determined.