Missing Patient Reported Outcome Data Research Articles

Ovarian cancer study dropouts had worse health-related quality of life and psychosocial symptoms at baseline and over time.

Participant drop out is a major barrier to high-quality patient-reported outcome (PRO) data analysis in cancer research as patients with worsening health are more likely to dropout. To test the hypothesis that ovarian cancer patients with worse PROs would drop out earlier, we examined how patients differed by time of dropout on health-related quality of life (HRQOL), anxiety, depression, optimism and insomnia. This analysis included 619 participants, stratified by time of dropout, from the Australian Ovarian Cancer Study - Quality of Life substudy, in which participants completed PRO questionnaires at three-monthly intervals for 21 months. Trends in PROs over time were examined. Pearson correlations examined the relationship between time of dropout and baseline PROs. Multiple linear regression models including age, disease stage and time since diagnosis examined relationships between baseline and final PRO scores, and final PRO scores and dropout group. Participants who dropped out earlier had significantly worse baseline HRQOL (p<0.0001) and higher depression (p<0.0001). For all five PROs, final scores were significantly associated with baseline scores (p<0.0001). Time of dropout was significantly associated with final HRQOL (p=0.003), anxiety (p=0.05), depression (p=0.02) and optimism (p=0.02) scores. Depression, HRQOL and anxiety worsened at a faster rate overtime in dropouts than study completers. Poorer HRQOL and higher depression at baseline, and final HRQOL, anxiety, depression and optimism scores were predictive of time of dropout. These results highlight the importance of collecting auxiliary data to inform careful and considered handling of missing PRO data during analysis, interpretation and reporting.

Design, implementation and reporting strategies to reduce the instance and impact of missing patient-reported outcome (PRO) data: a systematic review

ObjectivesPatient-reported outcomes (PROs) provide important information about the impact of treatment from the patients' perspective. However, missing PRO data may compromise the interpretability and value of the findings. We aimed...

Mode of administration does not cause bias in patient-reported outcome results: a meta-analysis.

Technological advances in recent decades have led to the availability of new modes to administer patient-reported outcomes (PROs). To aid selecting optimal modes of administration (MOA), we undertook a systematic review to determine whether differences in bias (both size and direction) exist among modes. We searched five electronic databases from 2004 (date of last comprehensive review on this topic) to April 2014, cross-referenced and searched reference lists. Studies that compared two or more MOA for a health-related PRO measure in adult samples were included. Two reviewers independently applied inclusion and quality criteria and extracted findings. Meta-analyses and meta-regressions were conducted using random-effects models. Of 5100 papers screened, 222 were considered potentially relevant and 56 met eligibility criteria. No evidence of bias was found for: (1) paper versus electronic self-complete; and (2) self-complete versus assisted MOA. Heterogeneity for paper versus electronic comparison was explained by type of construct (i.e. physical vs. psychological). Heterogeneity for self-completion versus assisted modes was in part explained by setting (clinic vs. home); the largest bias was introduced when assisted completion occurred in the clinic and follow-up was by self-completion (either electronic or paper) in the home. Self-complete paper and electronic MOA can be used interchangeably for research in clinic and home settings. Self-completion and assisted completion produce equivalent scores overall, although heterogeneity may be induced by setting. These results support the use of mixed MOAs within a research study, which may be a useful strategy for reducing missing PRO data.

Bias in Area Under the Curve for Longitudinal Clinical Trials With Missing Patient Reported Outcome Data

A common approach to the analysis of longitudinal patient reported outcomes (PROs) is the use of summary measures such as area under the time curve (AUC). However, it is not clear how missing data affects the validity of AUC analysis. This study aimed to compare the use of AUC summary measures (in individuals) with AUC summary statistics (on groups, calculated from the estimated parameters of a mixed model) when data are complete, missing at random, and missing not at random. A simulation experiment based on a two-armed randomized trial was carried out to investigate the precision and bias of AUC in longitudinal analysis where missingness, trajectory, and missingness allocation were varied. Summary measures AUC with ad hoc approaches to missing data were compared with mixed model AUC summary statistics. AUC summary statistics were consistently superior to AUC summary measures in terms of precision and bias. The bias of AUC summary statistic approach was very small, even when data were missing not at random and when differential attrition between groups existed. AUC summary measures on individuals should not be used to analyze longitudinal PRO data in the presence of missing data.

Patient‐reported outcomes in bariatric surgery: a systematic review of standards of reporting

Bariatric surgery is increasingly being used to treat severe obesity, but little is known about its impact on patient-reported outcomes (PROs). For PRO data to influence practice, well-designed and reported studies are required. A systematic review identified prospective bariatric surgery studies that used validated PRO measures. Risk of bias in randomized controlled trials (RCTs) was assessed, and papers were examined for reporting of (i) who completed PRO measures; (ii) missing PRO data and (iii) clinical interpretation of PRO data. Studies meeting all criteria were classified as robust. Eighty-six studies were identified. Of the eight RCTs, risk of bias was high in one and unclear in seven. Sixty-eight different PRO measures were identified, with the Short Form (SF)-36 questionnaire most commonly used. Forty-one (48%) studies explicitly stated measures were completed by patients, 63 (73%) documented missing PRO data and 50 (58%) interpreted PRO data clinically. Twenty-six (30%) met all criteria. Although many bariatric surgery studies assess PROs, study design and reporting is often poor, limiting data interpretation and synthesis. Well-designed studies that include agreed PRO measures are needed with reporting to include integration with clinical outcomes to inform practice.

Assessment of Patient Reported Outcomes (PROs) From the Enestnd Trial with Minimum Follow-up of 36 Cycles.

AbstractAbstract 3166 Background:In ENESTnd, nilotinib (NI) significantly reduced progression and demonstrated superior molecular response rates vs. imatinib (IM) in patients newly diagnosed with Philadelphia chromosome positive chronic phase chronic myelogenous leukemia (CML-CP). Additionally, fewer NI patients discontinued therapy vs. IM (Table 1). Previously reported 24-month analyses of PROs indicated similar health-related quality of life (HRQoL) and functioning mean scores between the treatment arms, and scores similar to general population norms. Aim:To evaluate the PROs for patients in ENESTnd with minimum follow-up of 36 cycles and to understand cohort-level (per treatment arm) HRQoL outcomes Methods:In ENESTnd, nNewly diagnosed CML-CP patients were randomized to NI 300 mg twice daily (BID), NI 400 mg BID, or IM 400 mg once daily (QD). HRQoL was assessed using the Functional Assessment of Cancer Therapy – Leukemia (FACT-Leu) and the Short Form 36 Health Survey (SF-36). The FACT-Leu has two components: a) the FACT-General (FACT-G) which measures physical, social/family, emotional, and functional well-being and b) a 17-item leukemia specific subscale. The SF-36 assesses eight domains (physical functioning, role-physical, bodily pain, general health, vitality, social functioning, role-emotional, and mental health) that enable a mental component score (MCS) and physical component score (PCS). Questionnaires were administered at enrollment and end of cycles 3, 12, 24, and 36. Mixed effects models for longitudinal data were used to compare trends over time between treatment arms. Study discontinuation prevented collection of PROs from patients with inadequate response or intolerance to treatment. Pattern-mixture models were fit to attempt to control for missing PRO data resulting from discontinuation or other reasons. Separately, missing data were imputed consistently across arms for cohort-level analysis of the FACT-Leu subscale so mean cohort scores could be reported (Table 1). Results:In both mixed effects and pattern-mixture models, FACT-Leu subscale, FACT-G, and SF-36 scores (PCS and MCS) were similar across treatment arms over time. Of the patients remaining on study and completing PRO questionnaires, SF-36 PCS and MCS scores at cycle 36 in all arms were comparable to the general US population; FACT-G scores were slightly better (Table 2). In the cohort-level analysis with imputation according to reason, IM FACT-Leu subscale scores began to trend increasingly lower vs. NI arms beginning at day 168 (Figure 1). By day 1008, the IM arm mean score is 10% and 13% lower than the NI 300 mg and NI 400 mg arms, respectively (Figure 1). Higher rates of discontinuation in the IM arm are the main factors that lead to the HRQoL deficit. Conclusions:In ENESTnd, patients who respond to and tolerate treatment have consistent HRQoL that is comparable to the general population. Cohort-level analysis indicates that discontinuation rates due to inadequate response and intolerance must be considered when determining the HRQoL across the entire cohort. These results suggest that in a population of newly diagnosed patients with CML-CP, NI results in higher HRQoL than IM. These findings may have particularly important implications for payers and policy makers when evaluating treatment options.Table 1IM 400mg QD (n=283) (%)NI 300mg BID (n=282) (%)NI 400mg BID (n=281) (%)Imputed FACT-LEUbStill on treatment61.870.973.7MeanDisease progression3.50.71.40Suboptimal response/treatment failure4.21.82.814Adverse events/lab abnormalities11.09.914.214Death0.41.40.40Other a6.77.86.414aIncludes abnormal test result(s), condition no longer requires study drug, consent withdrawn, lost to follow-up, administrative problems, protocol deviationbImputed values per time point, 14 represents the lowest recorded scoreTable 2Mean (SD)IM 400mg QD (n=152)NI 300mg BID (n=176)NI 400mg BID (n=185)General US Pop.SF-36 PCS49.1 (9.0)47.9 (9.5)49.1 (8.2)49.6 (10.0)cSF-36 MCS51.3 (9.6)50.2 (9.7)49.9 (9.0)50.5 (9.9)cFACT-G86.0 (15.7)84.5 (15.4)85.1 (15.0)80.1 (18.1)dFACT-Leu subscale54.7 (9.4)54.9 (8.9)55.4 (8.5)NAcFor ages 45–54 (http://www.sf-36.org/research/sf98norms.pdf)dEval Health Prof 2005; 28:192–211. [Display omitted] Disclosures:Beaumont:Novartis: Research Funding. Magestro:Novartis: Employment. Coombs:Novartis: Employment. Fan:Novartis Pharmaceuticals Corp: Employment. Kemp:Novartis Pharmaceuticals Corp: Employment. Waltzman:Novartis: Employment.