Marian R. Fisher

A Randomized Trial Comparing the Efficacy and Safety of Intravitreal Triamcinolone With Observation to Treat Vision Loss Associated With Macular Edema Secondary to Central Retinal Vein Occlusion: The Standard Care vs Corticosteroid for Retinal Vein Occlusion (SCORE) Study Report 5

OBJECTIVE To compare the efficacy and safety of 1-mg and 4-mg doses of preservative-free intravitreal triamcinolone with observation for eyes with vision loss associated with macular edema secondary to perfused central retinal vein occlusion (CRVO). METHODS Multicenter, randomized, clinical trial of 271 participants. MAIN OUTCOME MEASURE Gain in visual acuity letter score of 15 or more from baseline to month 12. RESULTS Seven percent, 27%, and 26% of participants achieved the primary outcome in the observation, 1-mg, and 4-mg groups, respectively. The odds of achieving the primary outcome were 5.0 times greater in the 1-mg group than the observation group (odds ratio [OR], 5.0; 95% confidence interval [CI], 1.8-14.1; P = .001) and 5.0 times greater in 4-mg group than the observation group (OR, 5.0; 95% CI, 1.8-14.4; P = .001); there was no difference identified between the 1-mg and 4-mg groups (OR, 1.0; 95% CI, 0.5-2.1; P = .97). The rates of elevated intraocular pressure and cataract were similar for the observation and 1-mg groups, but higher in the 4-mg group. CONCLUSIONS Intravitreal triamcinolone is superior to observation for treating vision loss associated with macular edema secondary to CRVO in patients who have characteristics similar to those in the SCORE-CRVO trial. The 1-mg dose has a safety profile superior to that of the 4-mg dose. Application to Clinical Practice Intravitreal triamcinolone in a 1-mg dose, following the retreatment criteria applied in the SCORE Study, should be considered for up to 1 year, and possibly 2 years, for patients with characteristics similar to those in the SCORE-CRVO trial. Trial Registration clinicaltrials.gov Identifier: NCT00105027.

Detection of diabetic macular edema. Ophthalmoscopy versus photography - early treatment diabetic retinopathy study report number 5

Clinical and photographic methods were used to assess retinopathy during the examinations of diabetic patients enrolled in the Early Treatment Diabetic Retinopathy Study (ETDRS). In analyzing available data from eyes randomly selected for deferral of treatment, the authors compare the clinical detection (including contact lens biomicroscopy) with photographic detection (30 degrees stereoscopic color fundus photographs) of diabetic macular edema. Based on clinical detection, 53% (1778 patients) had hard exudates within 1 disc diameter (DD) of the center of macula, 56% (1868 patients) had retinal thickening within this region, and 31% (1027 patients) had thickening at the center of macula. These analyses show agreements of 83, 78, and 83% between retinal specialists and photographic graders when assessing these three characteristics, respectively. Agreement was 81% in the detection of macular edema for which treatment is indicated (clinically significant macular edema). Each method has its advantages but in general there was close agreement between these methods, particularly for clinically significant macular edema, which supports the reliability of each method.

Sample sizes for long-term medical trial with time-dependent dropout and event rates

A general model is formulated that allows for time-dependent dropout and event rates in the determination of sample sizes for long-term medical trials when a therapy group and a control group are to be compared. The need for time-dependent event (dropout) rate is illustrated by using the Framingham Heart Study Mortality data to estimate sample size for the NHLBI (National Heart, Lung, and Blood Institute) Multiple Risk Factor Intervention Trial (MRFIT). A further generalization of the model allows for participants who drop out of the control group (e.g., because of therapeutic measures prescribed by their own physicians) to return (or not return) to the control group at a later date. The question of unequal sample sizes for the therapy and the control groups is also discussed.

A Severity Scale for Diabetic Macular Edema Developed from ETDRS Data

PURPOSE To develop a severity scale for diabetic macular edema (DME) and to assess relationships between severity and duration of DME and visual acuity (VA). METHODS From the Early Treatment Diabetic Retinopathy Study (ETDRS), mean baseline VA scores were tabulated for 7422 eyes cross-classified by (1) location of retinal thickening (RT) and its area within 1 disc diameter of the macular center, and (2) degree of RT at the center. Adjacent (row, column, and off-diagonal) cells with the greatest similarity in baseline VA (mean and SD) based on a Gaussian (normal) likelihood were merged. An initial eight-step scale was chosen using the Schwarz criterion (Bayesian information criterion; BIC) and was revised based on clinical judgment to nine steps. Relationships between baseline VA and other photographic and fluorescein angiographic characteristics were examined singly and in combination with the scale. RESULTS Modeling baseline VA as a function of the nine-step scale yielded an R(2) of 38.0%, compared with 38.4% using the full cross-classification of these variables. Addition of each of the other baseline characteristics changed the adjusted R(2) for the combination very little. Between scale levels 1A and 5B mean (SD) VA decreased from 86.8 (5.8) letters to 59.8 (13.6) letters. In a model of change in VA as a function of time spent at each DME severity level, VA loss increased progressively from 1 letter per year at level 2 to 17 letters per year at level 5B. CONCLUSIONS The scale facilitates documentation of the relationship of severity and duration of DME with VA.

The Role of an Independent Statistical Analysis Center in the Industry-Modified National Institutes of Health Model

During the past four decades, the randomized clinical trial has evolved as a major research methodology for the rigorous evaluation of new medical interventions and therapies. The role of statistics in clinical trial methodology and the organizational structure of clinical trials have also evolved. More recently, there has been a shift in the funding of clinical trials, from the public (eg, National Institutes of Health, or NIH) to the private (pharmaceutical and device industries) sector. This paper describes an Industry-Modified NIH Model for the conduct of industry-sponsored clinical trials that involves a distinct statistical analysis center (SAC) and data management center, as opposed to the single unit, or coordinating center, utilized in most NIH-sponsored trials. The role of the SAC in support of the monitoring activities of an independent data monitoring committee (IDMC) is described in detail. Since the Greenberg Report was published in 1967, the number of multicenter confirmatory clinical trials that have an IDMC has steadily increased, and the recent International Conference on Harmonisation (ICH) E9 regulatory guideline on clinical trials supports this trend. This manuscript provides explicit guidance regarding the activities of an independent SAC for industry-sponsored clinical trials based on our experiences as a SAC for a number of such trials. Activities discussed include protocol development, drafting of written operating procedures for an IDMC and summary notes of its meetings, and preparation of an interim analysis plan and interim analysis reports.

The WIZARD Trial as a Case Study of Flexible Clinical Trial Design

The design of a randomized, controlled, confirmatory efficacy clinical trial is influenced by multiple competing factors, including ethical, scientific, economic, and regulatory interests. Design elements, including choice of end points, study duration, and interim monitoring plans, need to be adapted to specific circumstances. The Weekly Intervention With Zithromax for Atherosclerosis and Its Related Disorders (WIZARD) study design was initially intended to meet the sponsors scientific goals while being subject to practical constraints. In response to changing circumstances, modifications to the study design were required. The WIZARD trial demonstrates the substantial flexibility in the design and conduct of group-sequential randomized trials not only during the design stage but also after the trial is under way, effectively satisfying evolving and competing demands while preserving the statistical and scientific validity of the study.

Relationship between rate of patient enrollment and quality of clinical center performance in two multicenter trials in ophthalmology

Those responsible for managing and overseeing multicenter clinical trials often express concern that differential rates of patients enrollment among participating clinical centers may be reflected in poorer compliance with study protocols or poorer patient care in the centers enrolling fewer patients. The purposes of the present investigation were (1) to determine whether rate of patient enrollment was related to other measures of clinical center performance in two multicenter clinical trials in ophthalmology, the Macular Photocoagulation Study (MPS) and the Early Treatment Diabetic Retinopathy Study (ETDRS), and (2) to compare findings from these two studies with findings reported from cooperative oncology groups. Percent of major data forms ever failing computerized edit was the only measure of performance examined that demonstrated a strong positive relationship to rate of patient enrollment in both studies under both analytic approaches employed (F test for linear trend in means and Spearmans rank correlation). In the MPS, but not in the ETDRS, percent of patients who had missed all visits scheduled during the past 12 months (inactive patients) also was related to rate of patient enrollment. However, the major finding from this investigation has been that it is not possible to predict overall performance of an individual clinical center based on rate of patient enrollment. Clinical centers enrolling fewer patients in these multicenter clinical trials performed in some areas as well as or better than centers enrolling larger numbers of patients.

Axial Length Estimation in Strabismic Patients

BACKGROUND Previous studies have indicated that axial length determination is important in strabismic patients for defining the limit for a safe maximum recession of the medial rectus. Also, the response to strabismus surgery may be, in part, a function of axial length. We previously published a formula for predicting axial length based on age and refractive error; however, its accuracy has not been tested in a patient population that is different from the one used to generate the formula. The purpose of this study is to test a formula for estimating axial length, given age and refractive error, in a population that is different from that from which it was generated. METHOD We measured axial length using A-scan ultrasonography in 163 consecutive patients undergoing strabismus surgery. Twenty-nine patients were younger than 18 months of age; 134 patients were between 18 months and 10 years of age. We compared the measured axial length determination with the axial length value estimated by a formula generated from our previous published series. RESULTS For patients younger than 18 months of age, the equation estimated axial length within 0.5 mm in 41.4% of patients, within 1.0 mm in 79.3% of patients, and within 1.5 mm in 93.1% of patients. For patients between 18 months and 18 years of age, the formula estimated axial length within 0.5 mm in 37.3% of patients, within 1.0 mm in 73.1% of patients, and within 1.5 mm in 87.3% of patients. CONCLUSIONS The formula may be useful for the strabismus surgeon in estimating axial length when A-scan ultrasonography is not available in an operating room setting, particularly in congenital esotropes who require larger recessions in small eyes. If, however, A-scan ultrasonography is available, it is preferable to using the formula. The formula is not sufficiently accurate for use for calculating intraocular lens power.

When is “final”?

(We don’t really need to divide by T here and use Slutsky in the first equation. Since the X’s are fixed, we can treat them as constants, and just move them outside the plim. Nonetheless, for the sake of propriety, we use plim... Also, in the second equation, one of the X’s is actually defined as the dependent variable in the first equation, and there is clearly a random element in it, so we really should not assume fixed X’s.)