Eleanor J Murray1, James M Robins1,2, George R Seage1, Sara Lodi1, Emily P Hyle3, Krishna P Reddy4, Kenneth A Freedberg3,5,6, Miguel A Hernán1,2,7. 1. Department of Epidemiology, Harvard School of Public Health, Boston, MA, USA (EJM, JMR, GRS, SL, MAH). 2. Department of Biostatistics, Harvard School of Public Health, Boston, MA, USA (JMR, MAH). 3. Division of Infectious Disease, Massachusetts General Hospital, Boston, MA, USA (EPH, KAF). 4. Division of Pulmonary and Critical Care Medicine, Massachusetts General Hospital, Boston, MA, USA (KPR). 5. Department of Health Policy and Management, Harvard School of Public Health, Boston, MA, USA (KAF). 6. Center for AIDS Research, Harvard University, Boston, MA, USA (KAF). 7. Harvard-MIT Division of Health Sciences and Technology, Boston, MA, USA (MAH).
Abstract
BACKGROUND: Individual-level simulation models are valuable tools for comparing the impact of clinical or public health interventions on population health and cost outcomes over time. However, a key challenge is ensuring that outcome estimates correctly reflect real-world impacts. Calibration to targets obtained from randomized trials may be insufficient if trials do not exist for populations, time periods, or interventions of interest. Observational data can provide a wider range of calibration targets but requires methods to adjust for treatment-confounder feedback. We propose the use of the parametric g-formula to estimate calibration targets and present a case-study to demonstrate its application. METHODS: We used the parametric g-formula applied to data from the HIV-CAUSAL Collaboration to estimate calibration targets for 7-y risks of AIDS and/or death (AIDS/death), as defined by the Center for Disease Control and Prevention under 3 treatment initiation strategies. We compared these targets to projections from the Cost-effectiveness of Preventing AIDS Complications (CEPAC) model for treatment-naïve individuals presenting to care in the following year ranges: 1996 to 1999, 2000 to 2002, or 2003 onwards. RESULTS: The parametric g-formula estimated a decreased risk of AIDS/death over time and with earlier treatment. The uncalibrated CEPAC model successfully reproduced targets obtained via the g-formula for baseline 1996 to 1999, but over-estimated calibration targets in contemporary populations and failed to reproduce time trends in AIDS/death risk. Calibration to g-formula targets improved CEPAC model fit for contemporary populations. CONCLUSION: Individual-level simulation models are developed based on best available information about disease processes in one or more populations of interest, but these processes can change over time or between populations. The parametric g-formula provides a method for using observational data to obtain valid calibration targets and enables updating of simulation model inputs when randomized trials are not available.
BACKGROUND: Individual-level simulation models are valuable tools for comparing the impact of clinical or public health interventions on population health and cost outcomes over time. However, a key challenge is ensuring that outcome estimates correctly reflect real-world impacts. Calibration to targets obtained from randomized trials may be insufficient if trials do not exist for populations, time periods, or interventions of interest. Observational data can provide a wider range of calibration targets but requires methods to adjust for treatment-confounder feedback. We propose the use of the parametric g-formula to estimate calibration targets and present a case-study to demonstrate its application. METHODS: We used the parametric g-formula applied to data from the HIV-CAUSAL Collaboration to estimate calibration targets for 7-y risks of AIDS and/or death (AIDS/death), as defined by the Center for Disease Control and Prevention under 3 treatment initiation strategies. We compared these targets to projections from the Cost-effectiveness of Preventing AIDS Complications (CEPAC) model for treatment-naïve individuals presenting to care in the following year ranges: 1996 to 1999, 2000 to 2002, or 2003 onwards. RESULTS: The parametric g-formula estimated a decreased risk of AIDS/death over time and with earlier treatment. The uncalibrated CEPAC model successfully reproduced targets obtained via the g-formula for baseline 1996 to 1999, but over-estimated calibration targets in contemporary populations and failed to reproduce time trends in AIDS/death risk. Calibration to g-formula targets improved CEPAC model fit for contemporary populations. CONCLUSION: Individual-level simulation models are developed based on best available information about disease processes in one or more populations of interest, but these processes can change over time or between populations. The parametric g-formula provides a method for using observational data to obtain valid calibration targets and enables updating of simulation model inputs when randomized trials are not available.
Authors: Rochelle P Walensky; Eric L Ross; Nagalingeswaran Kumarasamy; Robin Wood; Farzad Noubary; A David Paltiel; Yoriko M Nakamura; Sheela V Godbole; Ravindre Panchia; Ian Sanne; Milton C Weinstein; Elena Losina; Kenneth H Mayer; Ying Q Chen; Lei Wang; Marybeth McCauley; Theresa Gamble; George R Seage; Myron S Cohen; Kenneth A Freedberg Journal: N Engl J Med Date: 2013-10-31 Impact factor: 91.245
Authors: K A Freedberg; E Losina; M C Weinstein; A D Paltiel; C J Cohen; G R Seage; D E Craven; H Zhang; A D Kimmel; S J Goldie Journal: N Engl J Med Date: 2001-03-15 Impact factor: 91.245
Authors: Eleanor J Murray; James M Robins; George R Seage; Kenneth A Freedberg; Miguel A Hernán Journal: Am J Epidemiol Date: 2017-07-15 Impact factor: 4.897
Authors: Julia del Amo; Santiago Moreno; Heiner C Bucher; Hansjakob Furrer; Roger Logan; Jonathan Sterne; Santiago Pérez-Hoyos; Inma Jarrín; Andrew Phillips; Sara Lodi; Ard van Sighem; Wolf de Wolf; Caroline Sabin; Loveleen Bansi; Amy Justice; Joseph Goulet; José M Miró; Elena Ferrer; Laurence Meyer; Rémonie Seng; Giota Toulomi; Panagiotis Gargalianos; Dominique Costagliola; Sophie Abgrall; Miguel A Hernán Journal: Clin Infect Dis Date: 2012-03-28 Impact factor: 9.079
Authors: Sara Lodi; Shweta Sharma; Jens D Lundgren; Andrew N Phillips; Stephen R Cole; Roger Logan; Brian K Agan; Abdel Babiker; Hartwig Klinker; Haitao Chu; Matthew Law; James D Neaton; Miguel A Hernán Journal: AIDS Date: 2016-11-13 Impact factor: 4.177
Authors: Jane J Kim; Karen M Kuntz; Natasha K Stout; Salaheddin Mahmud; Luisa L Villa; Eduardo L Franco; Sue J Goldie Journal: Am J Epidemiol Date: 2007-05-25 Impact factor: 4.897
Authors: Alethea W McCormick; Nadia N Abuelezam; Erin R Rhode; Taige Hou; Rochelle P Walensky; Pamela P Pei; Jessica E Becker; Madeline A DiLorenzo; Elena Losina; Kenneth A Freedberg; Marc Lipsitch; George R Seage Journal: PLoS One Date: 2014-05-27 Impact factor: 3.240
Authors: Lauren E Cain; Roger Logan; James M Robins; Jonathan A C Sterne; Caroline Sabin; Loveleen Bansi; Amy Justice; Joseph Goulet; Ard van Sighem; Frank de Wolf; Heiner C Bucher; Viktor von Wyl; Anna Esteve; Jordi Casabona; Julia del Amo; Santiago Moreno; Remonie Seng; Laurence Meyer; Santiago Perez-Hoyos; Roberto Muga; Sara Lodi; Emilie Lanoy; Dominique Costagliola; Miguel A Hernan Journal: Ann Intern Med Date: 2011-04-19 Impact factor: 25.391
Authors: Eleanor J Murray; James M Robins; George R Seage; Kenneth A Freedberg; Miguel A Hernán Journal: Med Decis Making Date: 2020-01 Impact factor: 2.583