BACKGROUND: Missing data resulting from premature study withdrawal are a common problem in the analysis of longitudinal data in clinical trials. To date, this subject has received little attention in the context of economic evaluations and with regard to the analysis of cost data. OBJECTIVES: To (i) demonstrate the impact of patients who drop out during the study on the outcomes of an economic evaluation, and (ii) to compare the mean and variation in costs after applying five different methods to deal with incomplete data: multiple imputation, complete cases analysis, linear extrapolation, predicted mean and hot decking. STUDY DESIGN: The study was performed using cost data collected in two randomised clinical trials comparing patients with chronic obstructive pulmonary disease receiving eithertiotropium bromide or ipratropium bromide. The overall dropout rate was 17%, with the daily costs of the dropouts approximately 4 times higher than the costs of the completers. METHODS: Multiple imputation is a principled method that deals with missing observations by replacing each missing observation with a set of multiple plausible values. The variance between the resulting multiple datasets is combined with the variance between the datasets to take account of the extra uncertainty that results from missing data. The outcomes after multiple imputation were compared with the results of four naive methods to deal with missing observations: complete cases analysis, linear extrapolation, predicted mean and hot decking. All costs were expressed in 2001 euros. RESULTS: In the tiotropium bromide group, mean (standard error) costs varied from Euro 955 (137) after complete cases analysis to Euro 1298 (198) after linear extrapolation. The corresponding estimates in the ipratropium bromide group were Euro 970 (125) and Euro 1561 (244), respectively. The difference in costs between treatment groups varied from -Euro 15 (95% CI: -379 to 349) after complete cases analysis to -Euro 402 (95% CI: -883 to 79) after predicted mean, in favour of the tiotropium bromide group. The difference in costs according to the other methods varied from -Euro 263 (95% CI: -878 to 353) after linear extrapolation to -Euro 265 (95% CI: -709 to 180) after multiple imputation to -Euro 359 (95% CI: -771 to 54) after hot decking. CONCLUSION: This study showed that the method of dealing with the data of the dropouts had a large impact on the outcomes of an economic evaluation. Information about the rate of patient withdrawal and the way data of dropouts are treated is of vital importance in assessing the results of economic evaluations and should always be reported. Multiple imputation is a principled method that can be used to deal with the data of these patients.
RCT Entities:
BACKGROUND: Missing data resulting from premature study withdrawal are a common problem in the analysis of longitudinal data in clinical trials. To date, this subject has received little attention in the context of economic evaluations and with regard to the analysis of cost data. OBJECTIVES: To (i) demonstrate the impact of patients who drop out during the study on the outcomes of an economic evaluation, and (ii) to compare the mean and variation in costs after applying five different methods to deal with incomplete data: multiple imputation, complete cases analysis, linear extrapolation, predicted mean and hot decking. STUDY DESIGN: The study was performed using cost data collected in two randomised clinical trials comparing patients with chronic obstructive pulmonary disease receiving either tiotropium bromide or ipratropium bromide. The overall dropout rate was 17%, with the daily costs of the dropouts approximately 4 times higher than the costs of the completers. METHODS: Multiple imputation is a principled method that deals with missing observations by replacing each missing observation with a set of multiple plausible values. The variance between the resulting multiple datasets is combined with the variance between the datasets to take account of the extra uncertainty that results from missing data. The outcomes after multiple imputation were compared with the results of four naive methods to deal with missing observations: complete cases analysis, linear extrapolation, predicted mean and hot decking. All costs were expressed in 2001 euros. RESULTS: In the tiotropium bromide group, mean (standard error) costs varied from Euro 955 (137) after complete cases analysis to Euro 1298 (198) after linear extrapolation. The corresponding estimates in the ipratropium bromide group were Euro 970 (125) and Euro 1561 (244), respectively. The difference in costs between treatment groups varied from -Euro 15 (95% CI: -379 to 349) after complete cases analysis to -Euro 402 (95% CI: -883 to 79) after predicted mean, in favour of the tiotropium bromide group. The difference in costs according to the other methods varied from -Euro 263 (95% CI: -878 to 353) after linear extrapolation to -Euro 265 (95% CI: -709 to 180) after multiple imputation to -Euro 359 (95% CI: -771 to 54) after hot decking. CONCLUSION: This study showed that the method of dealing with the data of the dropouts had a large impact on the outcomes of an economic evaluation. Information about the rate of patient withdrawal and the way data of dropouts are treated is of vital importance in assessing the results of economic evaluations and should always be reported. Multiple imputation is a principled method that can be used to deal with the data of these patients.
Authors: W Vincken; J A van Noord; A P M Greefhorst; Th A Bantje; S Kesten; L Korducki; P J G Cornelissen Journal: Eur Respir J Date: 2002-02 Impact factor: 16.671
Authors: M Decramer; R Gosselink; M Rutten-Van Mölken; J Buffels; O Van Schayck; P-A Gevenois; R Pellegrino; E Derom; W De Backer Journal: Thorax Date: 2005-04 Impact factor: 9.139
Authors: J C Davis; C A Marra; M C Robertson; K M Khan; M Najafzadeh; M C Ashe; T Liu-Ambrose Journal: Osteoporos Int Date: 2010-08-04 Impact factor: 4.507
Authors: Jennifer C Davis; Carlo A Marra; M Clare Robertson; Mehdi Najafzadeh; Teresa Liu-Ambrose Journal: J Am Geriatr Soc Date: 2011-06-30 Impact factor: 5.562
Authors: Katherine W Reeves; Roslyn A Stone; Francesmary Modugno; Roberta B Ness; Victor G Vogel; Joel L Weissfeld; Laurel A Habel; Marike Vuga; Jane A Cauley Journal: Ann Epidemiol Date: 2011-04 Impact factor: 3.797
Authors: Anna D T Muntingh; Christina M van der Feltz-Cornelis; Harm W J van Marwijk; Philip Spinhoven; Willem J J Assendelft; Margot W M de Waal; Leona Hakkaart-van Roijen; Herman J Adèr; Anton J L M van Balkom Journal: BMC Health Serv Res Date: 2009-09-08 Impact factor: 2.655
Authors: L Hakkaart-van Roijen; B W C Zwirs; C Bouwmans; S S Tan; T W J Schulpen; L Vlasveld; J K Buitelaar Journal: Eur Child Adolesc Psychiatry Date: 2007-05-04 Impact factor: 4.785
Authors: Jennifer C Davis; Stirling Bryan; Carlo A Marra; Ging-Yuek R Hsiung; Teresa Liu-Ambrose Journal: Br J Sports Med Date: 2013-11-06 Impact factor: 13.800