Michael Lebenbaum1, Joyce Cheng2,3, Claire de Oliveira4,5,6, Paul Kurdyak4,6,7, Juveria Zaheer5,8, Rebecca Hancock-Howard4, Peter C Coyte4. 1. Institute of Health Policy, Management and Evaluation, University of Toronto, 4th Floor, 155 College St, Toronto, ON, M5T 3M6, Canada. Michael.Lebenbaum@hotmail.com. 2. Institute of Health Policy, Management and Evaluation, University of Toronto, 4th Floor, 155 College St, Toronto, ON, M5T 3M6, Canada. Joyce.Cheng.Mason@gmail.com. 3. Centre for Addiction and Mental Health (CAMH), 33 Russell Street, Toronto, ON, Canada. Joyce.Cheng.Mason@gmail.com. 4. Institute of Health Policy, Management and Evaluation, University of Toronto, 4th Floor, 155 College St, Toronto, ON, M5T 3M6, Canada. 5. Centre for Addiction and Mental Health (CAMH), 33 Russell Street, Toronto, ON, Canada. 6. ICES, Toronto, ON, Canada. 7. Health Outcomes and Performance Evaluation, Centre for Addiction and Mental Health, Toronto, ON, Canada. 8. Department of Psychiatry, University of Toronto, Toronto, ON, Canada.
Abstract
BACKGROUND: Although suicide-prevention campaigns have been implemented in numerous countries, Canada has yet to implement a strategy nationally. This is the first study to examine the cost utility of the implementation of a multidimensional suicide-prevention program that combines several interventions over a 50-year time horizon. METHODS: We used Markov modeling to capture the dynamic changes to health status and estimate the incremental cost per quality-adjusted life-year gained over a 50-year period for Ontario residents for a suicide-prevention strategy compared to no intervention. The strategy consisted of a package of interventions geared towards preventing suicide including a public health awareness campaign, increased identification of individuals at risk, increased training of primary-care physicians, and increased treatment post-suicide attempt. Four health states were captured by the Markov model: (1) alive and no recent suicide attempt; (2) suicide attempt; (3) death by suicide; (4) death (other than suicide). Analyses were from a societal perspective where all costs, irrespective of payer, were included. We used a probabilistic analysis to test the robustness of the model results to both variation and uncertainty in model parameters. RESULTS: Over the 50-year period, the suicide-prevention campaign had an incremental cost-effectiveness ratio (ICER) of $18,853 (values are in Canadian dollars) per QALY gained. In all one-way sensitivity analyses, the ICER remained under $50,000/QALY. In the probabilistic analysis, there was a probability of 94.8% that the campaign was cost effective at a willingness-to-pay of $50,000/QALY (95% confidence interval of ICER probabilistic distribution: 2650-62,375). Among the current population, the intervention was predicted to result in the prevention of 4454 suicides after 50 years (1033 by year 10; 2803 by year 25). A healthcare payer perspective sensitivity analysis showed an ICER of $21,096.14/QALY. INTERPRETATION: These findings demonstrate that a suicide-prevention campaign in Ontario is very likely a cost-effective intervention to reduce the incidence of suicide and suggest suicide-prevention campaigns are likely to be cost effective for some other Canadian provinces and potentially other countries.
BACKGROUND: Although suicide-prevention campaigns have been implemented in numerous countries, Canada has yet to implement a strategy nationally. This is the first study to examine the cost utility of the implementation of a multidimensional suicide-prevention program that combines several interventions over a 50-year time horizon. METHODS: We used Markov modeling to capture the dynamic changes to health status and estimate the incremental cost per quality-adjusted life-year gained over a 50-year period for Ontario residents for a suicide-prevention strategy compared to no intervention. The strategy consisted of a package of interventions geared towards preventing suicide including a public health awareness campaign, increased identification of individuals at risk, increased training of primary-care physicians, and increased treatment post-suicide attempt. Four health states were captured by the Markov model: (1) alive and no recent suicide attempt; (2) suicide attempt; (3) death by suicide; (4) death (other than suicide). Analyses were from a societal perspective where all costs, irrespective of payer, were included. We used a probabilistic analysis to test the robustness of the model results to both variation and uncertainty in model parameters. RESULTS: Over the 50-year period, the suicide-prevention campaign had an incremental cost-effectiveness ratio (ICER) of $18,853 (values are in Canadian dollars) per QALY gained. In all one-way sensitivity analyses, the ICER remained under $50,000/QALY. In the probabilistic analysis, there was a probability of 94.8% that the campaign was cost effective at a willingness-to-pay of $50,000/QALY (95% confidence interval of ICER probabilistic distribution: 2650-62,375). Among the current population, the intervention was predicted to result in the prevention of 4454 suicides after 50 years (1033 by year 10; 2803 by year 25). A healthcare payer perspective sensitivity analysis showed an ICER of $21,096.14/QALY. INTERPRETATION: These findings demonstrate that a suicide-prevention campaign in Ontario is very likely a cost-effective intervention to reduce the incidence of suicide and suggest suicide-prevention campaigns are likely to be cost effective for some other Canadian provinces and potentially other countries.