Statistical Monitoring in Clinical Trials: Best Practices for Detecting Data Anomalies Suggestive of Fabrication or Misconduct.

BACKGROUND: Traditional site-monitoring techniques are not optimal in finding data fabrication and other nonrandom data distributions with the greatest potential for jeopardizing the validity of study results. TransCelerate BioPharma conducted an experiment testing the utility of statistical methods for detecting implanted fabricated data and other signals of noncompliance. METHODS: TransCelerate tested statistical monitoring on a data set from a chronic obstructive pulmonary disease (COPD) clinical study with 178 sites and 1554 subjects. Fabricated data were selectively implanted in 7 sites and 43 subjects by expert clinicians in COPD. The data set was partitioned to simulate studies of different sizes. Analyses of vital signs, spirometry, visit dates, and adverse events included distributions of standard deviations, correlations, repeated values, digit preference, and outlier/inlier detection. An interpretation team, including clinicians, statisticians, site monitoring, and data management, reviewed the results and created an algorithm to flag sites for fabricated data. RESULTS: The algorithm identified 11 sites (19%), 19 sites (31%), 28 sites (16%), and 45 sites (25%) as having potentially fabricated data for studies 2A, 2, 1A, and 1, respectively. For study 2A, 3 of 7 sites with fabricated data were detected, 5 of 7 were detected for studies 2 and 1A, and 6 of 7 for study 1. Except for study 2A, the algorithm had good sensitivity and specificity (>70%) for identifying sites with fabricated data. CONCLUSIONS: We recommend a cross-functional, collaborative approach to statistical monitoring that can adapt to study design and data source and use a combination of statistical screening techniques and confirmatory graphics.

Change in area of geographic atrophy in the Age-Related Eye Disease Study: AREDS report number 26.

OBJECTIVE: To characterize progression of geographic atrophy (GA) associated with age-related macular degeneration in AREDS as measured by digitized fundus photographs. METHODS: Fundus photographs from 181 of 4757 AREDS participants with a GA area of at least 0.5 disc areas at baseline or from participants who developed bilateral GA during follow-up were scanned, digitized, and evaluated longitudinally. Geographic atrophy area was determined using planimetry. Rates of progression from noncentral to central GA and of vision loss following development of central GA included the entire AREDS cohort. RESULTS: Median initial lesion size was 4.3 mm(2). Average change in digital area of GA from baseline was 2.03 mm(2) (standard error of the mean, 0.24 mm(2)) at 1 year, 3.78 mm(2) (0.24 mm(2)) at 2 years, 5.93 mm(2) (0.34 mm(2)) at 3 years, and 1.78 mm(2) (0.086 mm(2)) per year overall. Median time to developing central GA after any GA diagnosis was 2.5 years (95% confidence interval, 2.0-3.0). Average visual acuity decreased by 3.7 letters at first documentation of central GA, and by 22 letters at year 5. CONCLUSIONS: Growth of GA area can be reliably measured using standard fundus photographs that are digitized and subsequently graded at a reading center. Development of GA is associated with subsequent further growth of GA, development of central GA, and loss in central vision.

Risk of advanced age-related macular degeneration after cataract surgery in the Age-Related Eye Disease Study: AREDS report 25.

PURPOSE: To assess the risk of advanced age-related macular degeneration (AMD) developing after cataract surgery. DESIGN: Cohort study. PARTICIPANTS: Four thousand five hundred seventy-seven participants (8050 eyes) from a multicenter, controlled, randomized clinical trial, the Age-Related Eye Disease Study (AREDS). METHODS: Development of advanced AMD, either neovascular (NV) AMD or geographic atrophy (GA), was evaluated with annual fundus photographs, and history of cataract surgery was assessed every 6 months. Cox proportional hazard models with time-dependent covariates were conducted for NV AMD and GA separately. MAIN OUTCOME MEASURES: Neovascular AMD, GA, and central GA (CGA; involving the center of the macula). RESULTS: The Cox proportional hazards model of right eyes showed nonsignificant hazard ratios of 1.20 (95% confidence interval [CI], 0.82-1.75) for NV AMD, 0.80 (95% CI, 0.61-1.06) for GA, and 0.87 (95% CI, 0.64-1.18) for CGA. Similar results were obtained for left eyes: 1.07 (95% CI, 0.72-1.58) for NV AMD, 0.94 (95% CI, 0.71-1.25) for GA, and 0.86 (95% CI, 0.63-1.19) for CGA. For participants with advanced AMD in 1 eye (AREDS category 4), the hazard ratios for fellow eyes were 1.08 (95% CI, 0.65-1.72) for NV AMD and 0.98 (95% CI, 0.64-1.49) for CGA. CONCLUSIONS: The AREDS results showed no clear effect of cataract surgery on the risk of progression to advanced AMD. FINANCIAL DISCLOSURE(S): The author(s) have no proprietary or commercial interest in any materials discussed in this article.

The relationship of dietary lipid intake and age-related macular degeneration in a case-control study: AREDS Report No. 20.

OBJECTIVE: To evaluate the association of lipid intake with baseline severity of age-related macular degeneration (AMD) in the Age-Related Eye Disease Study (AREDS). METHODS: Age-Related Eye Disease Study participants aged 60 to 80 years at enrollment (N = 4519) provided estimates of habitual nutrient intake through a self-administered semiquantitative food frequency questionnaire. Stereoscopic color fundus photographs were used to categorize participants into 4 AMD severity groups and a control group (participants with <15 small drusen). RESULTS: Dietary total omega-3 long-chain polyunsaturated fatty acid (LCPUFA) intake was inversely associated with neovascular (NV) AMD (odds ratio [OR], 0.61; 95% confidence interval [CI], 0.41-0.90), as was docosahexaenoic acid, a retinal omega-3 LCPUFA (OR, 0.54; 95% CI, 0.36-0.80), comparing highest vs lowest quintile of intake, after adjustment for total energy intake and covariates. Higher fish consumption, both total and broiled/baked, was also inversely associated with NV AMD (OR, 0.61; 95% CI, 0.37-1.00 and OR, 0.65; 95% CI, 0.45-0.93, respectively). Dietary arachidonic acid was directly associated with NV AMD prevalence (OR, 1.54; 95% CI, 1.04-2.29). No statistically significant relationships existed for the other lipids or AMD groups. CONCLUSION: Higher intake of omega-3 LCPUFAs and fish was associated with decreased likelihood of having NV AMD.