Alexander Thompson1, Bruce Guthrie2, Katherine Payne1. 1. Manchester Centre for Health Economics, University of Manchester, Manchester, UK (AT, KP). 2. Population Health Sciences Division, The University of Dundee, UK (BG).
Abstract
BACKGROUND: The payoff time represents an estimate of when the benefits of an intervention outweigh the costs. It is particularly useful for benefit-harm decision making for interventions that have deferred benefits but upfront harms. The aim of this study was to expand the application of the payoff time and provide an example of its use within a decision-analytic model. METHODS: Three clinically relevant patient vignettes based on varying levels of estimated 10-year cardiovascular risk (10%, 15%, 20%) were developed. An existing state-transition Markov model taking a health service perspective and a life-time horizon was adapted to include 3 levels of direct treatment disutility (DTD) associated with ongoing statin use: 0.005, 0.01, and 0.015. For each vignette and DTD we calculated a range of outputs including the payoff time inclusive and exclusive of healthcare costs. RESULTS: For a 10% 10-year cardiovascular risk (vignette 1) with low-levels of DTD (0.005), the payoff time was 8.5 years when costs were excluded and 16 years when costs were included. As the baseline risk of cardiovascular increased, the payoff time shortened. For a 15% cardiovascular risk (vignette 2) and for a low-level of DTD, the payoff time was 5.5 years and 9.5 years, respectively. For a 20% cardiovascular risk (vignette 3), the payoff time was 4.2 and 7.2 years, respectively. For higher levels of DTDs for each vignette, the payoff time lengthened, and in some instances the intervention never paid off, leading to an expected net harm for patients. CONCLUSIONS: This study has shown how the payoff time can be readily applied to an existing decision-analytic model and be used to complement existing measures to guide healthcare decision making.
BACKGROUND: The payoff time represents an estimate of when the benefits of an intervention outweigh the costs. It is particularly useful for benefit-harm decision making for interventions that have deferred benefits but upfront harms. The aim of this study was to expand the application of the payoff time and provide an example of its use within a decision-analytic model. METHODS: Three clinically relevant patient vignettes based on varying levels of estimated 10-year cardiovascular risk (10%, 15%, 20%) were developed. An existing state-transition Markov model taking a health service perspective and a life-time horizon was adapted to include 3 levels of direct treatment disutility (DTD) associated with ongoing statin use: 0.005, 0.01, and 0.015. For each vignette and DTD we calculated a range of outputs including the payoff time inclusive and exclusive of healthcare costs. RESULTS: For a 10% 10-year cardiovascular risk (vignette 1) with low-levels of DTD (0.005), the payoff time was 8.5 years when costs were excluded and 16 years when costs were included. As the baseline risk of cardiovascular increased, the payoff time shortened. For a 15% cardiovascular risk (vignette 2) and for a low-level of DTD, the payoff time was 5.5 years and 9.5 years, respectively. For a 20% cardiovascular risk (vignette 3), the payoff time was 4.2 and 7.2 years, respectively. For higher levels of DTDs for each vignette, the payoff time lengthened, and in some instances the intervention never paid off, leading to an expected net harm for patients. CONCLUSIONS: This study has shown how the payoff time can be readily applied to an existing decision-analytic model and be used to complement existing measures to guide healthcare decision making.
Entities:
Keywords:
cardiovascular; cost-effectiveness; decision-analytic modelling; direct treatment disutility; payoff time
Authors: Karen Barnett; Stewart W Mercer; Michael Norbury; Graham Watt; Sally Wyke; Bruce Guthrie Journal: Lancet Date: 2012-05-10 Impact factor: 79.321
Authors: Ghizelda R Lagerweij; G Ardine de Wit; Karel Gm Moons; Yvonne T van der Schouw; Wm Monique Verschuren; Jannick An Dorresteijn; Hendrik Koffijberg Journal: Eur J Prev Cardiol Date: 2018-02-07 Impact factor: 7.804