Group comparisons involving missing data in clinical trials: a comparison of estimates and power (size) for some simple approaches

A Review of Completed and Ongoing Complement Inhibitor Trials for Geographic Atrophy Secondary to Age-Related Macular Degeneration

Age-related macular degeneration (AMD) is a leading cause of irreversible blindness among older adults in the Western world. While therapies exist for patients with exudative AMD, there are currently no approved therapies for non-exudative AMD and its advanced form of geographic atrophy (GA). The discovery of genetic variants in complement protein loci with increased susceptibility to AMD has led to the investigation of the role of complement inhibition in AMD with a focus on GA. Here, we review completed and ongoing clinical trials evaluating the safety and efficacy of these studies. Overall, complement inhibition in GA has yielded mixed results. The inhibition of complement factor D has failed pivotal phase 3 trials. Studies of C3 and C5 inhibition meeting their primary endpoint are limited by high rates of discontinuation and withdrawal in the treatment arm and higher risks of conversion to exudative AMD. Studies evaluating other complement members (CFB, CFH, CFI and inhibitors of membrane attack complex—CD59) are ongoing and could offer other viable strategies.

C5 Inhibitor Avacincaptad Pegol for Geographic Atrophy Due to Age-Related Macular Degeneration: A Randomized Pivotal Phase 2/3 Trial

PURPOSE

The complement pathway may play a key role in the pathogenesis of age-related macular degeneration (AMD). The safety and efficacy of avacincaptad pegol (Zimura, IVERIC bio Inc, New York, NY), a C5 inhibitor, were assessed in participants with geographic atrophy (GA) secondary to AMD (GATHER1 Study).

DESIGN

International, prospective, randomized, double-masked, sham-controlled, pivotal phase 2/3 clinical trial.

PARTICIPANTS

A total of 286 participants with GA secondary to AMD.

MAIN OUTCOME MEASURES

The primary efficacy endpoint was the mean rate of change in GA over 12 months measured by fundus autofluorescence (FAF) at 3 timepoints: baseline, month 6, and month 12.

RESULTS

The reduction in the mean rate of GA growth (square root transformation) over 12 months was 27.4% (P = 0.0072) for the avacincaptad pegol 2 mg cohort and 27.8% (P = 0.0051) for the avacincaptad pegol 4 mg cohort compared with their corresponding sham cohorts. The results for both dose groups were statistically significant. Avacincaptad pegol was generally well tolerated after monthly administration over 12 months. There were no avacincaptad pegol-related adverse events (AEs) or inflammation. Further, there were no ocular serious AEs (SAEs) and no cases of endophthalmitis. The most frequent ocular AEs were related to the injection procedure.

CONCLUSIONS

Intravitreal administration of avacincaptad pegol 2 mg and 4 mg led to a significant reduction of GA growth in eyes with AMD over a 12-month period. Because C5 inhibition theoretically preserves C3 activity, it may offer additional safety advantages. A second confirmatory pivotal clinical trial is underway to confirm the efficacy and safety of avacincaptad pegol in slowing the GA growth (GATHER2 Study).

The short and long term effects of exercise training in non-cystic fibrosis bronchiectasis--a randomised controlled trial

BACKGROUND

Exercise training is recommended for non-cystic fibrosis (CF) bronchiectasis, but the long-term effects are unclear. This randomised controlled trial aimed to determine the effects of exercise training and review of airway clearance therapy (ACT) on exercise capacity, health related quality of life (HRQOL) and the incidence of acute exacerbations in people with non-CF bronchiectasis.

METHODS

Participants were randomly allocated to 8 weeks of supervised exercise training and review of ACT, or control. Primary outcomes of exercise capacity and HRQOL (Chronic respiratory disease questionnaire) and secondary outcomes of cough-related QOL (Leicester cough questionnaire) and psychological symptoms (Hospital anxiety and depression scale) were measured at baseline, following completion of the intervention period and at 6 and 12 months follow up. Secondary outcomes of the exacerbation rate and time to first exacerbation were analysed over 12 months.

RESULTS

Eighty-five participants (mean FEV1 74% predicted; median Modified Medical Research Council Dyspnoea grade of 1 (IQR [1-3]) were included. Exercise training increased the incremental shuttle walk distance (mean difference to control 62 m, 95% CI 24 to 101 m) and the 6-minute walking distance (mean difference to control 41 m, 95% CI 19 to 63 m), but these improvements were not sustained at 6 or 12 months. Exercise training reduced dyspnoea (p = 0.009) and fatigue (p = 0.01) but did not impact on cough-related QOL or mood. Exercise training reduced the frequency of acute exacerbations (median 1[IQR 1-3]) compared to the control group (2[1-3]) over 12 months follow up (p = 0.012), with a longer time to first exacerbation with exercise training of 8 months (95% CI 7 to 9 months) compared to the control group (6 months [95% CI 5 to 7 months], p = 0.047).

CONCLUSIONS

Exercise training in bronchiectasis is associated with short term improvement in exercise capacity, dyspnoea and fatigue and fewer exacerbations over 12 months.

TRIAL REGISTRY

ClinicalTrials.gov (NCT00885521).

Sample size determination for constrained longitudinal data analysis

The longitudinal data analysis model proposed by Liang and Zeger (Sankhyā: Indian J. Stat. Ser. B 2000; 62:134-148) uses the baseline as well as postbaseline values as the dependent variables, and the baseline mean responses are constrained to be the same across treatment groups due to randomization. Compared with the conventional longitudinal analysis of covariance, this approach can correctly estimate the variance of within-group mean changes and achieve the specified coverage probabilities. General results on the sample size and power calculations for this model in the presence of missing data are obtained. The sample size relationship between the constrained and unconstrained longitudinal data analysis is established. Simple expressions for sample size calculation are obtained for the compound symmetry and first-order autoregressive correlation structures. The sensitivity of the sample size requirement to the configuration of correlation structure and retention pattern is assessed. The performance of several ad hoc approximations for longitudinal data analysis sample size calculation is evaluated. Simulation studies are conducted to assess the validity of the proposed sample size formulas with deviation from normality. The sample size formulas are also illustrated in detail using real clinical trial data.

Imputation methods for missing outcome data in meta-analysis of clinical trials

Background

Missing outcome data from randomized trials lead to greater uncertainty and possible bias in estimating the effect of an experimental treatment. An intention-to-treat analysis should take account of all randomized participants even if they have missing observations.

Purpose

To review and develop imputation methods for missing outcome data in meta-analysis of clinical trials with binary outcomes.

Methods

We review some common strategies, such as simple imputation of positive or negative outcomes, and develop a general approach involving ‘informative missingness odds ratios’ (IMORs). We describe several choices for weighting studies in the meta-analysis, and illustrate methods using a meta-analysis of trials of haloperidol for schizophrenia.

Results

IMORs describe the relationship between the unknown risk among missing participants and the known risk among observed participants. They are allowed to differ between treatment groups and across trials. Application of IMORs and other methods to the haloperidol trials reveals the overall conclusion to be robust to different assumptions about the missing data.

Limitations

The methods are based on summary data from each trial (number of observed positive outcomes, number of observed negative outcomes and number of missing outcomes) for each intervention group. This limits the options for analysis, and greater flexibility would be available with individual participant data.

Conclusions

We propose that available reasons for missingness be used to determine appropriate IMORs. We also recommend a strategy for undertaking sensitivity analyses, in which the IMORs are varied over plausible ranges.

Discordance between reported intention-to-treat and per protocol analyses

OBJECTIVE

To quantify the degree of disagreement between the two most popular methods for dealing with missing data: intention to treat (ITT) and per protocol (PP).

STUDY DESIGN AND SETTING

We performed a systematic review of randomized two-armed clinical trials (CTs) published between 2001 and 2003, abstracted in PubMed and reporting both the ITT and PP analyses on a primary binary endpoint, out of which 74 papers were finally selected. The treatment effect of each CT was measured by the odds ratio, and the disagreement between them was quantified by the Bland-Altman method.

RESULTS

On average, the PP estimator provides greater values Log(e)ORPP=1 x 25.Log(e)ORITT, (95% CI: 1.15, 1.35) than the corresponding ITT estimator, although the limits of concordance showed that the ratio between the two estimators varies greatly from 0.39 up to 2.53.

CONCLUSION

These results confirm that missing values may cause both systematic and unpredictable bias in CTs. Further efforts should be made to minimize protocol deviations and to use better statistical methods to highlight the drawbacks of missing information. In the presence of protocol deviations, the conclusion of a CT cannot rest on the single reporting of either the ITT or the PP approach alone.

The Action to Control Cardiovascular Risk in Diabetes Memory in Diabetes Study (ACCORD-MIND): rationale, design, and methods

Type 2 diabetes mellitus and cognitive impairment are 2 of the most common chronic conditions found in persons aged > or = 60 years. Clinical studies have shown a greater prevalence of global cognitive impairment, incidence of cognitive decline, and incidence of Alzheimer disease in patients with type 2 diabetes. To date, there have been no randomized trials of the effects of long-term glycemic control on cognitive function and structural brain changes in patients with type 2 diabetes. The primary aim of the Action to Control Cardiovascular Risk in Diabetes (ACCORD) Memory in Diabetes Study (ACCORD-MIND) is to test whether there is a difference in the rate of cognitive decline and structural brain change in patients with diabetes treated with standard-care guidelines compared with those treated with intensive-care guidelines. This comparison will be made in a subsample of 2,977 patients with diabetes participating in the ongoing ACCORD trial, a clinical trial sponsored by the National Heart, Lung, and Blood Institute (NHLBI) with support from the National Institute on Aging (NIA). Data from this ACCORD substudy on the possible beneficial or adverse effects of intensive treatment on cognitive function will be obtained from a 30-minute test battery, administered at baseline and 20-month and 40-month visits. In addition, full-brain magnetic resonance imaging will be performed on 630 participants at baseline and at 40 months to assess the relation between the ACCORD treatments and structural brain changes. The general aim of ACCORD-MIND is to determine whether the intensive treatment of diabetes, a major risk factor for Alzheimer disease and vascular dementia, can reduce the early decline in cognitive function that could later evolve into more cognitively disabling conditions. This report presents the design, rationale, and methods of the ACCORD-MIND substudy.

Clinical trial of exercise for shoulder pain in chronic spinal injury

BACKGROUND AND PURPOSE

The high prevalence of shoulder pain in wheelchair users may be related to the repetitive use of the upper limbs during self-care and wheelchair-related activities. The purpose of this study was to determine the effects of a controlled 8-week, scapula-focused exercise intervention on pain and functional disability in people with spinal cord injury (SCI) and shoulder impingement symptoms.

SUBJECTS

Forty-one manual wheelchair users (with SCI and spina bifida), both with (n=21) and without (n=20) shoulder impingement symptoms, participated.

METHODS

The study design was a clinical trial with an asymptomatic control group. Subjects completed the Wheelchair User's Shoulder Pain Index (WUSPI) and the Shoulder Rating Questionnaire (SRQ) and provided patient satisfaction scores at initial and 8-week visits. Subjects in the intervention group were instructed in a home exercise program consisting of stretching and strengthening exercises. Subjects in the asymptomatic control group received no intervention. An analysis of variance model was used to test for group and time effects for the WUSPI, SRQ, and satisfaction scores.

RESULTS

Subjects in the intervention group showed significant improvements in all measures as a result of the intervention, whereas asymptomatic control group subjects remained stable.

DISCUSSION AND CONCLUSION

A selective 8-week home exercise program is effective in reducing pain and improving function and satisfaction in this population of wheelchair users.

Uncertainty method improved on best–worst case analysis in a binary meta-analysis

BACKGROUND

Most systematic reviewers aim to perform an intention-to-treat meta-analysis, including all randomized participants from each trial. This is not straightforward in practice: reviewers must decide how to handle missing outcome data in the contributing trials.

OBJECTIVE

To investigate methods of allowing for uncertainty due to missing data in a meta-analysis.

STUDY DESIGN AND SETTING

The Cochrane Library was surveyed to assess current use of imputation methods. We developed a methodology for incorporating uncertainty, with weights assigned to trials based on uncertainty interval widths. The uncertainty interval for a trial incorporates both sampling error and the potential impact of missing data. We evaluated the performance of this method using simulated data.

RESULTS

The survey showed that complete-case analysis is commonly considered alongside best-worst case analysis. Best-worst case analysis gives an interval for the treatment effect that includes all of the uncertainty due to missing data. Unless there are few missing data, this interval is very wide. Simulations show that the uncertainty method consistently has better power and narrower interval widths than best-worst case analysis.

CONCLUSION

The uncertainty method performs consistently better than best-worst case imputation and should be considered along with complete-case analysis whenever missing data are a concern.

Intention-to-treat approach to data from randomized controlled trials: a sensitivity analysis

The intention-to-treat (ITT) approach to randomized controlled trials analyzes data on the basis of treatment assignment, not treatment receipt. Alternative approaches make comparisons according to the treatment received at the end of the trial (as-treated analysis) or using only subjects who did not deviate from the assigned treatment (adherers-only analysis). Using a sensitivity analysis on data for a hypothetical trial, we compare these different analytical approaches in the context of two common protocol deviations: loss to follow-up and switching across treatments. In each case, two rates of deviation are considered: 10% and 30%. The analysis shows that biased estimates of effect may occur when deviation is nonrandom, when a large percentage of participants switch treatments or are lost to follow-up, and when the method of estimating missing values accounts inadequately for the process causing loss to follow-up. In general, ITT analysis attenuates between-group effects. Trialists should use sensitivity analyses on their data and should compare the characteristics of participants who do and those who do not deviate from the trial protocol. The ITT approach is not a remedy for unsound design, and imputation of missing values is not a substitute for complete, good quality data.

Effects of a home exercise programme on shoulder pain and functional status in construction workers

BACKGROUND

Repetitive or sustained elevated shoulder postures have been identified as a significant risk factor for occupationally related shoulder musculoskeletal disorders. Construction workers exposed to routine overhead work have high rates of shoulder pain that frequently progresses to functional loss and disability. Exercise interventions have potential for slowing this progression.

AIMS

To evaluate a therapeutic exercise programme intended to reduce pain and improve shoulder function.

METHODS

Construction worker volunteers were screened by history and clinical examination to test for inclusion/exclusion criteria consistent with shoulder pain and impingement syndrome. Sixty seven male symptomatic workers (mean age 49) were randomised into a treatment intervention group (n = 34) and a control group (n = 33); asymptomatic subjects (n = 25) participated as an additional control group. Subjects in the intervention group were instructed in a standardised eight week home exercise programme of five shoulder stretching and strengthening exercises. Subjects in the control groups received no intervention. Subjects returned after 8-12 weeks for follow up testing.

RESULTS

The intervention group showed significantly greater improvements in the Shoulder Rating Questionnaire (SRQ) score and shoulder satisfaction score than the control groups. Average post-test SRQ scores for the exercise group remained below levels for asymptomatic workers. Intervention subjects also reported significantly greater reductions in pain and disability than controls.

CONCLUSIONS

Results suggest a home exercise programme can be effective in reducing symptoms and improving function in construction workers with shoulder pain.