Background: Cluster randomized trials (CRTs) are a design used to test interventions where individual randomization is not appropriate. The mixed model for repeated measures (MMRM) is a popular choice for individually randomized trials with longitudinal continuous outcomes. This model's appeal is due to avoidance of model misspecification and its unbiasedness for data missing completely at random or at random. Methods: We extended the MMRM to cluster randomized trials by adding a random intercept for the cluster and undertook a simulation experiment to investigate statistical properties when data are missing at random. We simulated cluster randomized trial data where the outcome was continuous and measured at baseline and three post-intervention time points. We varied the number of clusters, the cluster size, the intra-cluster correlation, missingness and the datageneration models. We demonstrate the MMRM-CRT with an example of a cluster randomized trial on cardiovascular disease prevention among diabetics. Results: When simulating a treatment effect at the final time point we found that estimates were unbiased when data were complete and when data were missing at random. Variance components were also largely unbiased. When simulating under the null, we found that type I error was largely nominal, although for a few specific cases it was as high as 0.081. Conclusions:Although there have been assertions that this model is inappropriate when there are more than two repeated measures on subjects, we found evidence to the contrary. We conclude that the MMRM for CRTs is a good analytic choice for cluster randomized trials with a continuous outcome measured longitudinally.Trial registration: ClinicalTrials.gov, ID: NCT02804698.

ObjectiveTo assess the feasibility and acceptability of a beverage intervention in Hispanic adults.DesignEligible individuals identified as Hispanic, were 18–64 years old and had BMI 30·0–50·0 kg/m2. Participants were randomized 2:2:1 to one of three beverages: Mediterranean lemonade (ML), green tea (GT) or flavoured water control (FW). After a 2-week washout period, participants were asked to consume 32 oz (946 ml) of study beverage daily for 6 weeks and avoid other sources of tea, citrus, juice and sweetened beverages; water was permissible. Fasting blood samples were collected at baseline and 8 weeks to assess primary and secondary efficacy outcomes.SettingTucson, AZ, USA.ParticipantsFifty-two participants were recruited over 6 months; fifty were randomized (twenty-one ML, nineteen GT, ten FW). Study population mean (sd) age 44·6 (sd 10·2) years, BMI 35·9 (4·6) kg/m2; 78 % female.ResultsForty-four (88 %) completed the 8-week assessment. Self-reported adherence was high. No significant change (95 % CI) in total cholesterol (mg/dl) from baseline was shown −1·7 (−14·2, 10·9), −3·9 (−17·2, 9·4) and −13·2 (−30·2, 3·8) for ML, GT and FW, respectively. Mean change in HDL-cholesterol (mg/dl) −2·3 (−5·3, 0·7; ML), −1·0 (−4·2, 2·2; GT), −3·9 (−8·0, 0·2; FW) and LDL-cholesterol (mg/dl) 0·2 (−11·3, 11·8; ML), 0·5 (−11·4, 12·4; GT), −9·8 (−25·0, 5·4; FW) were also non-significant. Fasting glucose (mg/dl) increased significantly by 5·2 (2·6, 7·9; ML) and 3·3 (0·58, 6·4; GT). No significant change in HbA1c was demonstrated. Due to the small sample size, potential confounders and effect modifiers were not investigated.ConclusionsRecruitment and retention figures indicate that a larger-scale trial is feasible; however, favourable changes in cardiometabolic biomarkers were not demonstrated.

Missing data are common in NI and equivalence trials, and they are often handled by methods which may bias estimates and lead to incorrect conclusions.

Patient-reported outcomes, such as quality of life, functioning, and symptoms, are used widely in therapeutic and behavioral trials and are increasingly used in drug development to represent the patient voice. Missing patient reported data is common and can undermine the validity of results reporting by reducing power, biasing estimates, and ultimately reducing confidence in the results. In this paper, we review statistically principled approaches for handling missing patient-reported outcome data and introduce the idea of estimands in the context of behavioral trials. Specifically, we outline a plan that considers missing data at each stage of research: design, data collection, analysis, and reporting. The design stage includes processes to prevent missing data, define the estimand, and specify primary and sensitivity analyses. The analytic strategy considering missing data depends on the estimand. Reviewed approaches include maximum likelihood-based models, multiple imputation, generalized estimating equations, and responder analysis. We outline sensitivity analyses to assess the robustness of the primary analysis results when data are missing. We also describe ad-hoc methods, including approaches to avoid. Last, we demonstrate methods using data from a behavioral intervention, where the primary outcome was self-reported cognition.

Objectives Cognitive symptoms are commonly reported among cancer patients and survivors, yet guidance on when self‐reported cognitive symptoms warrant follow‐up is lacking. We sought to establish cut‐off scores for identifying patients with perceived low cognitive functioning on widely used self‐report measures of cognition and a novel single item Cognitive Change Score. Methods Adult patients diagnosed with invasive cancer who had completed at least one cycle of chemotherapy completed a questionnaire containing the EORTC‐Cognitive Function (CF) subscale, Functional Assessment of Cancer Therapy‐Cognitive Function (FACT‐COG) Perceived Cognitive Impairment (PCI) and our Cognitive Change Score (CCS). We used receiver operating characteristic analyses to establish the discriminative ability of these measures against the Patient's Assessment of Own Functioning Inventory (PAOFI) as our reference standard. We chose cut‐off scores on each measure that maximised both sensitivity and specificity for identifying patients with self‐reported low CF. Results We recruited 294 participants (55.8% women, mean age 56.6 years) with mixed cancer diagnoses (25.5 months since diagnosis). On the CCS, 77.6% reported some cognitive change since starting chemotherapy. On the PAOFI 36% had low CF. The following cut‐off scores identified cases of low CF: ≥28.5 on the CCS (75.5% sensitivity, 67.6% specificity); ≤75.0 on the European Organisation for Research and Treatment of Cancer, QLQ‐C30 Cognitive Functioning scale (90.9% sensitivity, 57.1% specificity); ≤55.1 on the FACT‐COG PCI‐18 (84.8% sensitivity, 76.2% specificity), and ≤59.5 on the FACT‐COG PCI‐20 (78.8% sensitivity, 84.1% specificity). Conclusions We found a single item question asking about cognitive change has acceptable discrimination between patients with self‐reported normal and low CF when compared to other more comprehensive self‐report measures of cognitive symptoms. Further validation work is required.