Limits...
Assessing quality of life in a randomized clinical trial: correcting for missing data.

Gunnes N, Seierstad TG, Aamdal S, Brunsvig PF, Jacobsen AB, Sundstrøm S, Aalen OO - BMC Med Res Methodol (2009)

Bottom Line: Comparison of the treatment arms shows a significant difference in mean score at the end of treatment.Use of proper methodology developed for analysing data subject to missingness is necessary to reduce potential estimation bias.The conclusions are robust for the choice of statistical methods.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Biostatistics, University of Oslo, P,O, Box 1122 Blindern, N-0317 Oslo, Norway. nina.gunnes@medisin.uio.no

ABSTRACT

Background: Health-related quality of life is a topic of current interest. This paper considers a randomized phase III study of radiation therapy with concurrent chemotherapy (docetaxel) versus radiation therapy alone in non-small cell lung cancer, stage III A/B. Longitudinal data on quality of life have been obtained through repeated administration of a multi-item questionnaire (EORTC QLQ-C30) developed by the European Organisation for Research and Treatment of Cancer. Missingness in the data is owing to patients having failed to complete the questionnaire at some of the scheduled filling-in times.

Methods: We have analysed a monotone (in terms of missingness) subset of the data as regards estimation of the mean score of a summary measure of self-reported quality of life in a hypothetical drop-out-free population at different points in time. Missingness is a difficult issue of great importance. We have therefore chosen to compare three different methods that are relatively easy to implement: the linear-increments method, the inverse-probability-weighting method and the Markov-process method. Single imputation has been applied in a supplementary analysis to fill in for all the non-consecutive missing score values prior to the execution of the estimation procedure.

Results: For the response in focus, the observed mean score at a certain time is larger than the estimated mean scores, which implies that the true mean score is easily overestimated unless the missingness is appropriately adjusted for. Comparison of the treatment arms shows a significant difference in mean score at the end of treatment.

Conclusion: Use of proper methodology developed for analysing data subject to missingness is necessary to reduce potential estimation bias. The quality of life of patients receiving radiation therapy with concurrent chemotherapy (docetaxel) appears somewhat worse than that of patients receiving radiation therapy alone in the period during which treatment is given. The conclusions are robust for the choice of statistical methods.

Show MeSH

Related in: MedlinePlus

Observed and estimated mean scores for a mortal cohort (without single imputation). The figure displays the observed and estimated mean scores for arm A (upper panel) and arm B (lower panel) when considering a mortal cohort. Single imputation has not been applied. The black solid-line curve corresponds to the arithmetic average of the observed score values, the magenta dotted-line curve corresponds to the LI method, and the green dash-dotted-line curve corresponds to the LI method when considering an immortal cohort (for comparison).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC2698910&req=5

Figure 2: Observed and estimated mean scores for a mortal cohort (without single imputation). The figure displays the observed and estimated mean scores for arm A (upper panel) and arm B (lower panel) when considering a mortal cohort. Single imputation has not been applied. The black solid-line curve corresponds to the arithmetic average of the observed score values, the magenta dotted-line curve corresponds to the LI method, and the green dash-dotted-line curve corresponds to the LI method when considering an immortal cohort (for comparison).

Mentions: Table 1 presents the numbers of observed score values, with respect to monotone missingness, for both treatment arms at different control weeks. The corresponding numbers of missing score values are presented in Table 2. Obviously, the numbers of observed score values decrease over time as the patients fail to answer the current question. In the same way, the numbers of missing score values increase over time. Figure 1 displays the mean score estimates, plotted against time, for both treatment arms when considering an immortal cohort. In the plot corresponding to arm A, we notice a rapid decline in the curves right after start of treatment. At control week 6, they reach a low before increasing. This sudden dip at the end of treatment is most likely due to some of the adverse effects of chemotherapy, such as nausea and discomfort, which generally lead to low score values. The curves fluctuate somewhat after control week 24. In contrast, the curves in the plot corresponding to arm B fall gradually. They begin to rise again at control week 84. Figure 2 displays the LI estimates of the mean score, plotted against time, for both treatment arms when considering a mortal cohort. We observe no important differences between the immortal cohort analysis and the mortal cohort analysis as regards estimation of the mean score using the LI method.


Assessing quality of life in a randomized clinical trial: correcting for missing data.

Gunnes N, Seierstad TG, Aamdal S, Brunsvig PF, Jacobsen AB, Sundstrøm S, Aalen OO - BMC Med Res Methodol (2009)

Observed and estimated mean scores for a mortal cohort (without single imputation). The figure displays the observed and estimated mean scores for arm A (upper panel) and arm B (lower panel) when considering a mortal cohort. Single imputation has not been applied. The black solid-line curve corresponds to the arithmetic average of the observed score values, the magenta dotted-line curve corresponds to the LI method, and the green dash-dotted-line curve corresponds to the LI method when considering an immortal cohort (for comparison).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC2698910&req=5

Figure 2: Observed and estimated mean scores for a mortal cohort (without single imputation). The figure displays the observed and estimated mean scores for arm A (upper panel) and arm B (lower panel) when considering a mortal cohort. Single imputation has not been applied. The black solid-line curve corresponds to the arithmetic average of the observed score values, the magenta dotted-line curve corresponds to the LI method, and the green dash-dotted-line curve corresponds to the LI method when considering an immortal cohort (for comparison).
Mentions: Table 1 presents the numbers of observed score values, with respect to monotone missingness, for both treatment arms at different control weeks. The corresponding numbers of missing score values are presented in Table 2. Obviously, the numbers of observed score values decrease over time as the patients fail to answer the current question. In the same way, the numbers of missing score values increase over time. Figure 1 displays the mean score estimates, plotted against time, for both treatment arms when considering an immortal cohort. In the plot corresponding to arm A, we notice a rapid decline in the curves right after start of treatment. At control week 6, they reach a low before increasing. This sudden dip at the end of treatment is most likely due to some of the adverse effects of chemotherapy, such as nausea and discomfort, which generally lead to low score values. The curves fluctuate somewhat after control week 24. In contrast, the curves in the plot corresponding to arm B fall gradually. They begin to rise again at control week 84. Figure 2 displays the LI estimates of the mean score, plotted against time, for both treatment arms when considering a mortal cohort. We observe no important differences between the immortal cohort analysis and the mortal cohort analysis as regards estimation of the mean score using the LI method.

Bottom Line: Comparison of the treatment arms shows a significant difference in mean score at the end of treatment.Use of proper methodology developed for analysing data subject to missingness is necessary to reduce potential estimation bias.The conclusions are robust for the choice of statistical methods.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Biostatistics, University of Oslo, P,O, Box 1122 Blindern, N-0317 Oslo, Norway. nina.gunnes@medisin.uio.no

ABSTRACT

Background: Health-related quality of life is a topic of current interest. This paper considers a randomized phase III study of radiation therapy with concurrent chemotherapy (docetaxel) versus radiation therapy alone in non-small cell lung cancer, stage III A/B. Longitudinal data on quality of life have been obtained through repeated administration of a multi-item questionnaire (EORTC QLQ-C30) developed by the European Organisation for Research and Treatment of Cancer. Missingness in the data is owing to patients having failed to complete the questionnaire at some of the scheduled filling-in times.

Methods: We have analysed a monotone (in terms of missingness) subset of the data as regards estimation of the mean score of a summary measure of self-reported quality of life in a hypothetical drop-out-free population at different points in time. Missingness is a difficult issue of great importance. We have therefore chosen to compare three different methods that are relatively easy to implement: the linear-increments method, the inverse-probability-weighting method and the Markov-process method. Single imputation has been applied in a supplementary analysis to fill in for all the non-consecutive missing score values prior to the execution of the estimation procedure.

Results: For the response in focus, the observed mean score at a certain time is larger than the estimated mean scores, which implies that the true mean score is easily overestimated unless the missingness is appropriately adjusted for. Comparison of the treatment arms shows a significant difference in mean score at the end of treatment.

Conclusion: Use of proper methodology developed for analysing data subject to missingness is necessary to reduce potential estimation bias. The quality of life of patients receiving radiation therapy with concurrent chemotherapy (docetaxel) appears somewhat worse than that of patients receiving radiation therapy alone in the period during which treatment is given. The conclusions are robust for the choice of statistical methods.

Show MeSH
Related in: MedlinePlus