Statistics Collaborative Inc.

Washington, DC, United States

Statistics Collaborative Inc.

Washington, DC, United States
SEARCH FILTERS
Time filter
Source Type

Dugel P.U.,Retinal Consultants of Arizona | Bebchuk J.D.,Statistics Collaborative Inc | Nau J.,NeoVista Inc | Reichel E.,Tufts Medical Center | And 4 more authors.
Ophthalmology | Year: 2013

Purpose: To evaluate the safety and efficacy of epimacular brachytherapy (EMBT) for the treatment of neovascular age-related macular degeneration (AMD). Design: Multicenter, randomized, active-controlled, phase III clinical trial. Participants: Four hundred ninety-four participants with treatment-naïve neovascular AMD. Methods: Participants with classic, minimally classic, and occult lesions were randomized in a 2:1 ratio to EMBT or a ranibizumab monotherapy control arm. The EMBT arm received 2 mandated, monthly loading injections of 0.5 mg ranibizumab. The control arm received 3 mandated, monthly loading injections of ranibizumab then quarterly injections. Both arms also received monthly as needed (pro re nata) retreatment. Main Outcome Measures: The proportion of participants losing fewer than 15 Early Treatment Diabetic Retinopathy Study (ETDRS) letters from baseline visual acuity (VA) and the proportion gaining more than 15 ETDRS letters from baseline VA. Results: At 24 months, 77% of the EMBT group and 90% of the control group lost fewer than 15 letters. This difference did not meet the prespecified 10% noninferiority margin. This end point was noninferior using a 20% margin and a 95% confidence interval for the group as a whole and for classic and minimally classic lesions, but not for occult lesions. The EMBT did not meet the superiority end point for the proportion of participants gaining more than 15 letters (16% for the EMBT group vs. 26% for the control group): this difference was statistically significant (favoring controls) for occult lesions, but not for predominantly classic and minimally classic lesions. Mean VA change was -2.5 letters in the EMBT arm and +4.4 letters in the control arm. Participants in the EMBT arm received a mean of 6.2 ranibizumab injections versus 10.4 in the control arm. At least 1 serious adverse event occurred in 54% of the EMBT arm, most commonly postvitrectomy cataract, versus 18% in the control arm. Mild, nonproliferative radiation retinopathy occurred in 3% of the EMBT participants, but no case was vision threatening. Conclusions: The 2-year efficacy data do not support the routine use of EMBT for treatment-naïve wet AMD, despite an acceptable safety profile. Further safety review is required. Financial Disclosure(s): Proprietary or commercial disclosure may be found after the references. © 2013 American Academy of Ophthalmology.


Jackson T.L.,King's College London | Dugel P.U.,Retinal Consultants of Arizona | Bebchuk J.D.,Statistics Collaborative Inc. | Smith K.R.,Statistics Collaborative Inc. | And 4 more authors.
Ophthalmology | Year: 2013

Purpose: To report the fluorescein angiography (FA) and optical coherence tomography (OCT) results of a clinical trial of epimacular brachytherapy (EMBT) used for the treatment of neovascular age-related macular degeneration (AMD). Design: Pivotal multicenter, active-controlled, randomized clinical trial. Participants: A total of 494 participants with treatment-naïve, neovascular AMD. Methods: Participants with classic, minimally classic, and occult lesions were randomized to receive (a) EMBT and 2 mandated monthly ranibizumab injections followed by pro re nata (PRN) ranibizumab or (b) 3 mandated monthly ranibizumab injections followed by mandated quarterly plus PRN ranibizumab. Participants underwent FA at screening and at months 1, 6, 12, 18, and 24. Optical coherence tomography scans were undertaken monthly for 24 months. The FA and OCT images were analyzed at respective independent reading centers. Main Outcome Measures: Change at 24 months in mean FA total lesion size and choroidal neovascularization (CNV) size and change in mean OCT centerpoint thickness. Results: The mean (standard deviation) changes in FA total lesion size in the EMBT and control arms were +1.9 (8.7) and -3.0 (7.2) mm2, respectively, with a mean change in total CNV size of +0.4 (8.4) and -4.7 (6.5) mm2, respectively. Mean (standard deviation) changes in OCT centerpoint thickness were -144 (246) and -221 (185) μm, respectively. Retrospective subgroup analyses showed no significant difference between treatment arms in mean centerpoint thickness in some subgroups, including eyes with classic lesions. The control arm showed a significantly larger reduction in mean total lesion size and mean CNV size than the EMBT arm in all subgroups analyzed. Nine eyes in the EMBT arm showed features consistent with mild, nonproliferative radiation retinopathy, but with a mean gain of 5.0 Early Treatment Diabetic Retinopathy Study letters. Conclusions: Both FA and OCT suggest that EMBT with PRN ranibizumab results in an inferior structural outcome than quarterly plus PRN ranibizumab. Some subgroup analyses suggest that classic lesions may be more responsive than occult lesions, although generally both subgroups are inferior to ranibizumab. A non-vision-threatening radiation retinopathy occurs in 2.9% of eyes over 24 months, but longer follow-up is needed. © 2013 by the American Academy of Ophthalmology Published by Elsevier Inc.


Pitt B.,University of Michigan | Bakris G.L.,University of Chicago | Bushinsky D.A.,University of Rochester | Garza D.,Relypsa | And 5 more authors.
European Journal of Heart Failure | Year: 2015

Aims We evaluated the effects of patiromer, a potassium (K+)-binding polymer, in a pre-specified analysis of hyperkalaemic patients with heart failure (HF) in the OPAL-HK trial. Methods and results Chronic kidney disease (CKD) patients on renin-angiotensin-aldosterone system inhibitors (RAASi) with serum K+ levels ≥5.1 mEq/L to <6.5 mEq/L (n = 243) received patiromer (4.2 g or 8.4 g BID initially) for 4 weeks (initial treatment phase); the primary efficacy endpoint was mean change in serum K+ from baseline to week 4. Eligible patients (those with baseline K+ ≥5.5 mEq/L to <6.5 mEq/L and levels ≥3.8 mEq/L to <5.1 mEq/L at the end of week 4) entered an 8-week randomized withdrawal phase and were randomly assigned to continue patiromer or switch to placebo; the primary efficacy endpoint was the between-group difference in median change in the serum K+ over the first 4 weeks of that phase. One hundred and two patients (42%) had heart failure (HF). The mean [± standard error (SE)] change in serum K+ from baseline to week 4 was -1.06 ± 0.05 mEq/L [95% confidence interval (CI), -1.16,-0.95; P < 0.001]; 76% (95% CI, 69,84) achieved serum K+, 3.8 mEq/L to <5.1 mEq/L. In the randomized withdrawal phase, the median increase in serum K+ from baseline of that phase was greater with placebo (n = 22) than patiromer (n = 27) (P < 0.001); recurrent hyperkalaemia (serum K+, ≥5.5 mEq/L) occurred in 52% on placebo and 8% on patiromer (P < 0.001). Mild-to-moderate constipation was the most common adverse event (11%); hypokalaemia occurred in 3%. Conclusion In patients with CKD and HF who were hyperkalaemic on RAASi, patiromer was well tolerated, decreased serum K+, and, compared with placebo, reduced recurrent hyperkalaemia. © 2015 The Authors. European Journal of Heart Failure published by John Wiley & Sons Ltd on behalf of European Society of Cardiology.


Dickstein K.,University of Bergen | Dickstein K.,University of Stavanger | Bebchuk J.,Statistics Collaborative Inc | Wittes J.,Statistics Collaborative Inc
Progress in Cardiovascular Diseases | Year: 2012

Coronary artery disease and myocardial infarction represent a major cause of morbidity and mortality. Four randomized, controlled, double-blind clinical trials -VALIANT, EPHESUS, OPTIMAAL, and CAPRICORN evaluated pharmacologic intervention in a total of 28,771 high-risk patients following acute MI complicated with signs of heart failure or evidence of left ventricular dysfunction. The demographic profiles of the 4 study cohorts were similar. The High-Risk MI Database Initiative constructed a common database by merging the data captured by these 4 large trials. The merged data set did not contain the randomized study treatment, so no comparisons could be made between the agents investigated. A total of more than 17,600 subjects experienced a cardiovascular end point. Approximately 5100 deaths occurred, and more than 15,700 subjects experienced a hospitalization. The primary objectives of this initiative were to use this large database to define more precisely the prognostic profile of this high-risk population, to perform rigorous, adequately-sized, subset analyses, to provide epidemiologic information and event rate estimation based on baseline demographics. The methodological challenges and limitations of such an analyses are discussed. It is proposed that some thoughtful foresight and planning could enable us to use the large number of clinical events that accrue during randomized clinical trials to address questions of scientific and clinical interest. © 2012 Elsevier Inc.


Fonseca V.,Tulane University | McDuffie R.,Tulane University | Calles J.,Case Western Reserve University | Cohen R.M.,University of Cincinnati | And 7 more authors.
Diabetes Care | Year: 2013

OBJECTIVE-Identify determinants of weight gain in people with type 2 diabetes mellitus (T2DM) allocated to intensive versus standard glycemic control in the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial. RESEARCH DESIGNANDMETHODS-We studied determinants of weight gain over 2 years in 8,929 participants (4,425 intensive arm and 4,504 standard arm) with T2DM in the ACCORD trial.We used general linear models to examine the association between each baseline characteristic and weight change at the 2-year visit.We fit a linear regression of change in weight and A1C and used general linear models to examine the association between each medication at baseline and weight change at the 2-year visit, stratified by glycemia allocation. RESULTS-There was significantly more weight gain in the intensive glycemia arm of the trial compared with the standard arm (3.0±6 7.0 vs. 0.36±6.3 kg). On multivariate analysis, younger age, male sex, Asian race, no smoking history, high A1C, baseline BMI of 25-35, high waist circumference, baseline insulin use, and baseline metformin use were independently associated with weight gain over 2 years. Reduction of A1C from baseline was consistently associated with weight gain only when baseline A1C was elevated. Medication usage accounted for <15% of the variability of weight change, with initiation of thiazolidinedione (TZD) use the most prominent factor. Intensive participants who never took insulin or a TZD had an average weight loss of 2.9 kg during the first 2 years of the trial. In contrast, intensive participants who had never previously used insulin or TZD but began this combination after enrolling in the ACCORD trial had a weight gain of 4.6-5.3 kg at 2 years. CONCLUSIONS-Weight gain in ACCORD was greater with intensive than with standard treatment and generally associated with reduction of A1C from elevated baseline values. Initiation of TZD and/or insulin therapy was the most important medication-related factor associated with weight gain. © 2013 by the American Diabetes Association.


Lan K.K.G.,Janssen Pharmaceutical | Wittes J.T.,Statistics Collaborative Inc.
Clinical Trials | Year: 2012

Background Traditional calculations of sample size do not formally incorporate uncertainty about the likely effect size. Use of a normal prior to express that uncertainty, as recently recommended, can lead to power that does not approach 1 as the sample size approaches infinity. Purpose To provide approaches for calculating sample size and power that formally incorporate uncertainty about effect size. The relevant formulas should ensure that power approaches one as sample size increases indefinitely and should be easy to calculate. Methods We examine normal, truncated normal, and gamma priors for effect size computationally and demonstrate analytically an approach to approximating the power for a truncated normal prior. We also propose a simple compromise method that requires a moderately larger sample size than the one derived from the fixed effect method. Results Use of a realistic prior distribution instead of a fixed treatment effect is likely to increase the sample size required for a Phase 3 trial. The standard fixed effect method for moving from estimates of effect size obtained in a Phase 2 trial to the sample size of a Phase 3 trial ignores the variability inherent in the estimate from Phase 2. Truncated normal priors appear to require unrealistically large sample sizes while gamma priors appear to place too much probability on large effect sizes and therefore produce unrealistically high power. Limitations The article deals with a few examples and a limited range of parameters. It does not deal explicitly with binary or time-to-failure data. Conclusions Use of the standard fixed approach to sample size calculation often yields a sample size leading to lower power than desired. Other natural parametric priors lead either to unacceptably large sample sizes or to unrealistically high power. We recommend an approach that is a compromise between assuming a fixed effect size and assigning a normal prior to the effect size. © 2012 The Author(s).


Cole E.B.,Medical University of South Carolina | Toledano A.Y.,Statistics Collaborative Inc | Lundqvist M.,Philips | Pisano E.D.,Medical University of South Carolina
Academic Radiology | Year: 2012

Rationale and Objectives: The purpose of this study was to assess the performance of a MicroDose photon-counting full-field digital mammography (PCM) system in comparison to full-field digital mammography (FFDM) for area under the receiver-operating characteristic (ROC) curve (AUC), sensitivity, specificity, and feature analysis of standard-view mammography for women presenting for screening mammography, diagnostic mammography, or breast biopsy. Materials and Methods: A total of 133 women were enrolled in this study at two European medical centers, with 67 women who had a pre-existing 10-36 months FFDM enrolled prospectively into the study and 66 women who underwent breast biopsy and had screening PCM and diagnostic FFDM, including standard craniocaudal and mediolateral oblique views of the breast with the lesion, enrolled retrospectively. The case mix consisted of 49 cancers, 17 biopsy-benign cases, and 67 normal cases. Sixteen radiologists participated in the reader study and interpreted all 133 cases in both conditions, separated by washout period of ≥4 weeks. ROC curve and free-response ROC curve analyses were performed for noninferiority of PCM compared to FFDM using a noninferiority margin Δ value of 0.10. Feature analysis of the 66 cases with lesions was conducted with all 16 readers at the conclusion of the blinded reads. Mean glandular dose was recorded for all cases. Results: The AUC for PCM was 0.947 (95% confidence interval [CI], 0.920-0.974) and for FFDM was 0.931 (95% CI, 0.898-0.964). Sensitivity per case for PCM was 0.936 (95% CI, 0.897-0.976) and for FFDM was 0.908 (95% CI, 0.856-0.960). Specificity per case for PCM was 0.764 (95% CI, 0.688-0.841) and for FFDM was 0.749 (95% CI, 0.668-0.830). Free-response ROC curve figures of merit were 0.920 (95% CI, 0.881-0.959) and 0.903 (95% CI, 0.858-0.948) for PCM and FFDM, respectively. Sensitivity per lesion was 0.903 (95% CI, 0.846-0.960) and 0.883 (95% CI, 0.823-0.944) for PCM and FFDM, respectively. The average false-positive marks per image of noncancer cases were 0.265 (95% CI, 0.171-0.359) and 0.281 (95% CI, 0.188-0.374) for PCM and FFDM, respectively. Noninferiority P values for AUC, sensitivity (per case and per lesion), specificity, and average false-positive marks per image were all statistically significant (P <.001). The noninferiority P value for free-response ROC was <.025, from the 95% CI for the difference. Feature analysis resulted in PCM being preferred to FFDM by the readers for ≥70% of the cases. The average mean glandular dose for PCM was 0.74 mGy (95% CI, 0.722-0.759 mGy) and for FFDM was 1.23 mGy (95% CI, 1.199-1.262 mGy). Conclusions: In this study, radiologist performance with PCM was not inferior to that with conventional FFDM at an average 40% lower mean glandular dose. © 2012 AUR.


Schumi J.,Statistics Collaborative Inc. | Wittes J.T.,Statistics Collaborative Inc.
Trials | Year: 2011

Non-inferiority trials test whether a new product is not unacceptably worse than a product already in use. This paper introduces concepts related to non-inferiority, and discusses the regulatory views of both the European Medicines Agency and the United States Food and Drug Administration. © 2011 Schumi and Wittes; licensee BioMed Central Ltd.


This paper briefly describes Jerome Cornfield's approach to Bayesian statistics, his discomfort with frequentist inference, and his contribution to two major clinical trials, the University Group Diabetes Program and the Coronary Drug Project. I mention the role of Bayesian statistics in current randomized clinical trials and conjecture why Cornfield's contributions to Bayesian methods are not more widely cited today. I then provide some personal recollections of Jerry as a role model and mentor and conclude with a recommendation that biostatisticians read his seminal papers because of their thoughtfulness, insight, wit, and clarity. © 2012 John Wiley & Sons, Ltd.


Wittes J.,Statistics Collaborative Inc.
Clinical Trials | Year: 2012

Background This article addresses a problem arising when a trial shows such strong evidence of benefit of the tested intervention that it stops early with an observed effect size for the experimental treatment that is statistically significantly better than the control. Within the classical frequentist framework of group sequential trials, the observed estimated effect size, the associated nave confidence interval, and the p-value are all biased estimates of the true values. The bias is in the direction of the overestimation of the treatment effect, creation of narrower confidence intervals than appropriate, and a p-value that is too small. Purpose To discuss methods for correcting the bias in observed effect sizes, confidence intervals, and p-values for trials stopped early and to show the extent to which such correction would have modified the conclusions of the Randomized Aldactone Evaluation Study (RALES). Results In RALES, the effect of not correcting for bias is negligible. Limitations This article does not show general results; it only explores a few examples that use conservative methods for early stopping. It does not consider sequential methods that allow a relatively high probability of stopping early. Conclusions This article points out that there is no unique solution to the correction of the p-value, but it recommends stagewise ordering, which states that earlier stopping of a trial is ipso facto stronger evidence of effect than later stopping so long as the stopping is governed by a monitoring boundary that preserves the Type I error rate. Associated with stagewise ordering is a method for calculating the estimated effect size and its confidence interval. In the RALES trial, which stopped at 50% information time, the corrections to the estimated values are small. © The Author(s), 2012.

Loading Statistics Collaborative Inc. collaborators
Loading Statistics Collaborative Inc. collaborators