Failure to account for time‐dependent treatment use when developing a prognostic model can result in biased future predictions. We reviewed currently available methods to account for treatment use when developing a prognostic model. First, we defined the estimands targeted by each method and examined their mechanisms of action with directed acyclic graphs (DAGs). Next, methods were implemented in data from 1,906 patients; 325 received selective β‐blockers (SBBs) during follow‐up. We demonstrated seven Cox regression modeling strategies: (a) ignoring SBB treatment; (b) excluding SBB users or (c) censoring them when treated; (d) inverse probability of treatment weighting after censoring (IPCW), including SBB treatment as (e) a binary or (f) a time‐dependent covariate; and (g) marginal structural modeling (MSM). Using DAGs, we demonstrated IPCW and MSM have the best properties and target a similar estimand. In the case study, compared to (a), approaches (b) and (e) provided predictions that were 1% and 2% higher on average. Performance (c‐statistic, Brier score, calibration slope) varied minimally between approaches. Our review of methods confirmed that ignoring treatment is theoretically inferior, but differences between the prediction models obtained using different methods can be modest in practice. Future simulation studies and applications are needed to assess the value of applying IPCW or MSM to adjust for treatments in different treatment and disease settings.
Background: Little evidence on the validity of simple and widely applicable tools to predict mortality in patients with chronic obstructive pulmonary disease (COPD) exists. Objective: To conduct a large international study to validate the ADO index that uses age, dyspnoea and FEV1 to predict 3-year mortality and to update it in order to make prediction of mortality in COPD patients as generalisable as possible. Design: Individual subject data analysis of 10 European and American cohorts (n=13 914). Setting: Population-based, primary, secondary and tertiary care. Patients: COPD GOLD stages I–IV. Measurements: We validated the original ADO index. We then obtained an updated ADO index in half of our cohorts to improve its predictive accuracy, which in turn was validated comprehensively in the remaining cohorts using discrimination, calibration and decision curve analysis and a number of sensitivity analyses. Results: 1350 (9.7%) of all subjects with COPD (60% male, mean age 61 years, mean FEV1 66% predicted) had died at 3 years. The original ADO index showed high discrimination but poor calibration (p<0.001 for difference between predicted and observed risk). The updated ADO index (scores from 0 to 14) preserved excellent discrimination (area under curve 0.81, 95% CI 0.80 to 0.82) but showed much improved calibration with predicted 3-year risks from 0.7% (95% CI 0.6% to 0.9%, score of 0) to 64.5% (61.2% to 67.7%, score of 14). The ADO index showed higher net benefit in subjects at low-to-moderate risk of 3-year mortality than FEV1 alone. Interpretation: The updated 15-point ADO index accurately predicts 3-year mortality across the COPD severity spectrum and can be used to inform patients about their prognosis, clinical trial study design or benefit harm assessment of medical interventions. ; The Barmelweid cohort (Switzerland) was funded by the Swiss National Science Foundation (grant number 3233B0-115216) and by the Klinik Barmelweid. Basque study (Spain): No external funding. The Cardiovascular Health Study is supported by NHLBI Grant/Contract numbers N01-HC-85239, N01-HC-85079 through N01-HC-85086; N01-HC-35129, N01 HC-15103, N01 HC-55222, N01-HC-75150, N01-HC-45133; HL080295, HL-075366; NIA Grant/Contract numbers AG-023269, AG-15928, AG-20098, and AG-027058; University of Pittsburgh Claude. D. Pepper Older Americans Independence Center grant number P30-AG-024827; with additional contribution from NINDS. See also http://www.chs-nhlbi.org/pi.htm The Copenhagen City Heart City study (Denmark) was supported by grants from The Danish Heart Foundation, The Danish Lung Association and Danish Medical Research Council. The Jackson Heart Study (JHS) is a collaborative study supported by the National Institutes of Health and the National Center on Minority Health and Health Disparities (study ID numbers: 5001; N01 HC95170; N01 HC95171; N01 HC95172) in partnership with Jackson State University, Tougaloo College, and University of Mississippi Medical Center. The Lung Health Study (USA) was supported by contract NIH/N01-HR-46002 from the National Heart, Lung and Blood Institute (NHLBI). The National Emphysema Treatment Trial (USA) is funded by the National Heart, Lung and Blood Institute, the Centers for Medicare and Medicaid Services, and the Agency for Healthcare Research and Quality. The PAC-COPD Study is funded by grants from Fondo de Investigación Sanitaria (FIS PI020541), Ministry of Health, Spain; Agència d'Avaluació de Tecnologia i Recerca Mèdiques (AATRM 035/20/02), Catalonia Government; Spanish Society of Pneumology and Thoracic Surgery (SEPAR 2002/137); Catalan Foundation of Pneumology (FUCAP 2003 Beca Marià Ravà); Red RESPIRA (RTIC C03/11); Red RCESP (RTIC C03/09), Fondo de Investigación Sanitaria (PI052486); Fondo de Investigación Sanitaria (PI052302); Fundació La Marató de TV3 (num. 041110); DURSI (2005SGR00392); and an unrestricted educational grant from Novartis Farmacèutica, Spain. CIBERESP and CIBERES are funded by the Instituto de Salud Carlos III, Ministry of Health, Spain. PLATINO study was funded by ALAT (Associacion Latino Americana del Tórax); Boehringer Ingelheim GmbH (BI), and GlaxoSmithKline (GSK). The SEPOC study (Spain) was supported by grants from Fondo de Investigación Sanitaria 99/0690 and CIRIT 1999SGR00240. Judith Garcia-Aymerich has a researcher contract from the Instituto de Salud Carlos III (CP05/00118), Ministry of Health, Spain. Karel G.M. Moons receives funding from the Netherlands Organisation for Scientific Research (project 9120.8004 and 918.10.615).
Readers' note This article is a living systematic review that will be updated to reflect emerging evidence. Updates may occur for up to two years from the date of original publication. This version is update 3 of the original article published on 7 April 2020 (BMJ 2020;369:m1328). Previous updates can be found as data supplements (https://www.bmj.com/content/369/bmj.m1328/related#datasupp). When citing this paper please consider adding the update number and date of access for clarity. Funding: LW, BVC, LH, and MDV acknowledge specific funding for this work from Internal Funds KU Leuven, KOOR, and the COVID-19 Fund. LW is a postdoctoral fellow of Research Foundation-Flanders (FWO) and receives support from ZonMw (grant 10430012010001). BVC received support from FWO (grant G0B4716N) and Internal Funds KU Leuven (grant C24/15/037). TPAD acknowledges financial support from the Netherlands Organisation for Health Research and Development (grant 91617050). VMTdJ was supported by the European Union Horizon 2020 Research and Innovation Programme under ReCoDID grant agreement 825746. KGMM and JAAD acknowledge financial support from Cochrane Collaboration (SMF 2018). KIES is funded by the National Institute for Health Research (NIHR) School for Primary Care Research. The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR, or the Department of Health and Social Care. GSC was supported by the NIHR Biomedical Research Centre, Oxford, and Cancer Research UK (programme grant C49297/A27294). JM was supported by the Cancer Research UK (programme grant C49297/A27294). PD was supported by the NIHR Biomedical Research Centre, Oxford. MOH is supported by the National Heart, Lung, and Blood Institute of the United States National Institutes of Health (grant R00 HL141678). ICCvDH and BCTvB received funding from Euregio Meuse-Rhine (grant Covid Data Platform (coDaP) interref EMR187). The funders played no role in study design, data collection, data analysis, data interpretation, or reporting. ; Peer reviewed ; Publisher PDF
