Werner Brannath

A graphical approach to sequentially rejective multiple test procedures

For clinical trials with multiple treatment arms or endpoints a variety of sequentially rejective, weighted Bonferroni-type tests have been proposed, such as gatekeeping procedures, fixed sequence tests, and fallback procedures. They allow to map the difference in importance as well as the relationship between the various research questions onto an adequate multiple test procedure. Since these procedures rely on the closed test principle, they usually require the explicit specification of a large number of intersection hypotheses tests. The underlying test strategy may therefore be difficult to communicate. We propose a simple iterative graphical approach to construct and perform such Bonferroni-type tests. The resulting multiple test procedures are represented by directed, weighted graphs, where each node corresponds to an elementary hypothesis, together with a simple algorithm to generate such graphs while sequentially testing the individual hypotheses. The approach is illustrated with the visualization of several common gatekeeping strategies. A case study is used to illustrate how the methods from this article can be used to tailor a multiple test procedure to given study objectives.

Recursive Combination Tests

We present a method that extends the flexibility of adaptive designs to the number of interim analyses and to the choice of decision boundaries. At each stage of the trial, the design of the next stage can be determined using all of the information gathered so far. Additionally, one can specify the next stage as the final stage or can plan a further interim analysis. The method is based on a recursive application of the two-stage combination tests for p values. The crucial point is the appropriate definition of a p value function that combines p values from two separate stages to a single p value. Formally, we start with a two-stage combination test. However, the p value of the second stage can be replaced by the p value of a further combination test. This applies if the experimenter decides in the first interim analysis to perform another interim analysis, thereby extending the design to at least three stages. Obviously, this can be also done in a recursive way in the following interim analyses. The test decision is based on an overall p value, which can be easily calculated by concatenation of the p value functions for the combination tests used. Under very general conditions, this procedure controls the level α. Recursive combination tests cover the classical group sequential and adaptive two-stage tests as special cases. By construction, they also allow for simple computation of confidence intervals and median unbiased point estimates.

25-gauge vs 20-gauge system for pars plana vitrectomy: a prospective randomised clinical trial.

Aim: To compare 25-gauge vs 20-gauge system for pars plana vitrectomy in a prospective, randomised, controlled clinical trial. Methods: Three-port pars plana vitrectomy was performed in 60 patients belonging to 2 groups. Evaluations were performed preoperatively, intraoperatively, during the first three postoperative days, at 1 week, and at 1 and 3 months. The main outcome measure was time for surgery, divided into duration of wound opening, vitrectomy, retinal manipulation and wound closure. Results: The total duration of surgery showed no significant difference between the groups (p = 0.67). The 25-gauge group showed significantly shorter duration of wound opening (p<0.001) and wound closure (p<0.001). In contrast, the vitrectomy duration was significantly longer in the 25-gauge group (p<0.001). Conjunctival injection and subjective postoperative pain showed significantly lower irritation in the 25-gauge group (p<0.001 for both). Conclusion: The 25-gauge vitrectomy system offered significantly improved patient comfort during the first postoperative week. The smaller surgical openings facilitated wound healing and minimised pain. Duration of surgery was comparable between the two systems—the shorter time needed for wound opening and closure in the 25-gauge group being equalised by the longer vitrectomy duration. Intraoperative as well as retinal manipulation and illumination caused more surgical difficulties using the 25-gauge system.

Adaptive designs for confirmatory clinical trials

Adaptive designs play an increasingly important role in clinical drug development. Such designs use accumulating data of an ongoing trial to decide how to modify design aspects without undermining the validity and integrity of the trial. Adaptive designs thus allow for a number of possible adaptations at midterm: Early stopping either for futility or success, sample size reassessment, change of population, etc. A particularly appealing application is the use of adaptive designs in combined phase II/III studies with treatment selection at interim. The expectation has arisen that carefully planned and conducted studies based on adaptive designs increase the efficiency of the drug development process by making better use of the observed data, thus leading to a higher information value per patient.In this paper we focus on adaptive designs for confirmatory clinical trials. We review the adaptive design methodology for a single null hypothesis and how to perform adaptive designs with multiple hypotheses using closed test procedures. We report the results of an extensive simulation study to evaluate the operational characteristics of the various methods. A case study and related numerical examples are used to illustrate the key results. In addition we provide a detailed discussion of current methods to calculate point estimates and confidence intervals for relevant parameters.

Confirmatory adaptive designs with Bayesian decision tools for a targeted therapy in oncology

The ability to select a sensitive patient population may be crucial for the development of a targeted therapy. Identifying such a population with an acceptable level of confidence may lead to an inflation in development time and cost. We present an approach that allows to decrease these costs and to increase the reliability of the population selection. It is based on an actual adaptive phase II/III design and uses Bayesian decision tools to select the population of interest at an interim analysis. The primary endpoint is assumed to be the time to some event like e.g. progression. It is shown that the use of appropriately stratified logrank tests in the adaptive test procedure guarantees overall type I error control also when using information on patients that are censored at the adaptive interim analysis. The use of Bayesian decision tools for the population selection decision making is discussed. Simulations are presented to illustrate the operating characteristics of the study design relative to a more traditional development approach. Estimation of treatment effects is considered as well.

Prospective analysis of factors related to migraine attacks: the PAMINA study.

Migraine is related to numerous factors such as hormones, stress or nutrition, but information about their actual importance is limited. Therefore, we analysed prospectively a wide spectrum of factors related to headache in migraineurs. We examined 327 migraineurs recruited via newspapers who kept a comprehensive diary for 3 months. Statistical analysis comprising 28 325 patient days and 116 dichotomous variables was based on the interval between two successive headache attacks. We calculated univariate Cox regression analyses and included covariables with a P-value of <0.05 in two stepwise multivariate Cox regression analyses, the first accounting for a correlation of the event times within a subject, the second stratified by the number of headache-free intervals. We performed similar analyses for the occurrence of migraine attacks and for the persistence of headache and migraine. Menstruation had the most prominent effect, increasing the hazard of occurrence or persistence of headache and migraine by up to 96%. All other factors changed the hazard by <35%. The two days before menstruation and muscle tension in the neck, psychic tension, tiredness, noise and odours on days before headache onset increased the hazard of headache or migraine, whereas days off, a divorced marriage, relaxation after stress, and consumption of beer decreased the hazard. In addition, three meteorological factors increased and two others decreased the hazard. In conclusion, menstruation is most important in increasing the risk of occurrence and persistence of headache and migraine. Other factors increase the risk less markedly or decrease the risk.

Prospective analysis of factors related to migraine aura--the PAMINA study.

Objectives.— The aim of this study was to examine factors increasing and decreasing the risk of occurrence of migraine aura and of headache and migraine not associated with aura (HoA, MoA) prospectively by means of a daily diary.

Choroidal blood flow and progression of age-related macular degeneration in the fellow eye in patients with unilateral choroidal neovascularization.

PURPOSE Cardiovascular risk factors such as smoking, hypertension, and atherosclerosis seem to play an important role in the development of choroidal neovascularization (CNV). Recent studies have also provided evidence suggesting that choroidal and retinal blood flow is decreased in patients with AMD. On the basis of these results, the hypothesis for this study was that lower choroidal blood flow is associated with an increased risk of CNV in patients with AMD. METHODS Forty-one patients with unilateral choroidal neovascular AMD were included in this observational longitudinal study. The fellow eyes of the patients served as study eyes. Subfoveal choroidal blood flow (FLOW) and fundus pulsation amplitude (FPA) were assessed with laser Doppler flowmetry and laser interferometry, respectively. A multivariate COX-regression model was used to test the hypothesis that low choroidal perfusion parameters are associated with the development of CNV. RESULTS Of the 37 patients that were followed up until the end of the study, 17 developed CNV and 20 did not. The univariate COX-regression analysis shows that lower FLOW, systolic blood pressure, intraocular pressure, and FPA are risk factors for development of CNV. Moreover, the more advanced the AMD in the study eye, the higher the risk for CNV to develop in the fellow eye. Multivariate COX regression analysis indicated that only FLOW (P = 0.0071), FPA (P = 0.0068), and staging (P = 0.031) had statistically significant influences on the progression to CNV. CONCLUSIONS The present study indicates that lower choroidal perfusion is a risk factor for the development of CNV in the fellow eye of patients with unilateral CNV.

The Effect of Remifentanil on the Heat Pain Threshold in Volunteers

Remifentanil offers a wide range of clinical uses and has been successfully combined with general anesthetics. However, there are few human experimental studies demonstrating the analgesic property of remifentanil. It was our aim to determine the analgesic effect of remifentanil with regard to dose-dependent increments in a human model of heat pain threshold assessment. Twenty healthy volunteers were randomized in a double-blinded cross-over design to receive an infusion of remifentanil or saline. The stepped infusion was increased every 5 min by 0.01 &mgr;g · kg−1 · min−1 up to 0.17 &mgr;g · kg−1 · min−1and terminated in case of defined safety limits. Thermal sensory testing of the heat pain threshold was performed every 5 min at the left forearm. The dose-response relationship and the effective dose for at least 50% of the subjects (ED50) were determined. Remifentanil led to a clear dose-dependent increase of the heat pain threshold differing significantly from placebo (P < 0.0007). The ED50 of remifentanil equals 0.05 &mgr;g · kg−1 · min−1 (first quartile 0.025 &mgr;g · kg−1 · min−1 and third quartile 0.06 &mgr;g · kg−1 · min−1) in this experimental setting. In conclusion, an opioid-mediated analgesic effect of remifentanil was determined in a human heat pain threshold model. The dose of 0.05 &mgr;g · kg−1 · min−1 is an effective and safe increment in healthy volunteers. Implications This study investigated remifentanil as a single infused drug to determine its analgetic property. A dose-dependent effect was found in a human heat pain threshold assessment.

Predictors of mortality at initiation of peritoneal dialysis in children after cardiac surgery

BACKGROUND The development of renal dysfunction in the postoperative course of cardiac surgery is still associated with high mortality in pediatric patients. In particular for small infants peritoneal dialysis offers a secure and useful treatment option. The aim of the present study was to investigate if routinely used laboratory and clinical variables could help predict mortality at initiation of peritoneal dialysis. METHODS We performed a retrospective chart analysis of pediatric intensive care unit patients with renal dysfunction who were treated with peritoneal dialysis after cardiac surgery between 1993 and 2001 and analyzed variables obtained 3 hours or less before starting peritoneal dialysis. RESULTS Results are documented as means and standard errors. A total of 1141 children underwent a cardiac operation on cardiopulmonary bypass. Sixty-two children (5.4%) were treated with peritoneal dialysis. Mortality was 40.3% (37 survivors, 25 nonsurvivors). The pH in survivors was 7.35 (0.01); in nonsurvivors it was 7.23 (0.03; p = 0.0037). Base excess in survivors was -1.37 mmol/L (0.61); in nonsurvivors it was -7.17 mmol/L (1.49; p = 0.0026). Lactate in survivors was 4.5 mmol/L (0.60); in nonsurvivors it was 10.5 mmol/L (1.78; p = 0.0089). Positive inspiratory pressure in survivors was 24.6 cm H(2)O (0.78); in nonsurvivors it was 28.9 cm H(2)O (1.08; p = 0.0274). Tidal volume per kilogram bodyweight in survivors was 11.0 mL/kg (0.48); in nonsurvivors it was 8.7 mL/kg (0.50; p = 0.0493). CONCLUSIONS We conclude from our data that the consideration of pH, base excess, lactate, positive inspiratory pressure, and tidal volume per kilogram bodyweight help predict mortality at initiation of peritoneal dialysis. We were able to observe significant differences between survivors and nonsurvivors using these variables.