PubMed Health⌕ Search

Biomedical subjects

Stephen Senn

Publications and source records attributed to Stephen Senn.

14 recordsLinked to original sources
PubMed2007

Trying to be precise about vagueness.

A previous investigation by Lambert et al., which used computer simulation to examine the influence of choice of prior distribution on inferences from Bayesian random effects meta-analysis, is critically examined from a number of viewpoints. The practical example used is shown to be problematic. The various prior distributions are shown to be unreasonable in terms of what they imply about the joint distribution of the overall treatment effect and the random effects variance. An alternative form of prior distribution is tentatively proposed. Finally, some practical recommendations are made that stress the value both of fixed effect analyses and of frequentist approaches as well as various diagnostic investigations.

Bayes Theorem↗
PubMed2006

Change from baseline and analysis of covariance revisited.

The case for preferring analysis of covariance (ANCOVA) to the simple analysis of change scores (SACS) has often been made. Nevertheless, claims continue to be made that analysis of covariance is biased if the groups are not equal at baseline. If the required equality were in expectation only, this would permit the use of ANCOVA in randomized clinical trials but not in observational studies. The discussion is related to Lord's paradox. In this note, it is shown, however that it is not a necessary condition for groups to be equal at baseline, not even in expectation, for ANCOVA to provide unbiased estimates of treatment effects. It is also shown that although many situations can be envisaged where ANCOVA is biased it is very difficult to imagine circumstances under which SACS would then be unbiased and a causal interpretation could be made.

Analysis of Variance↗
PubMed2006

Cross-over trials in Statistics in Medicine: the first '25' years.

Papers on cross-over trials that have appeared in the first 25 years of Statistics in Medicine are reviewed. Papers on bioequivalence are also considered. After a brief statistical summary, individual papers are discussed under seven headings: 1. The two-stage analysis of AB/BA trials, 2. Baselines, 3. Binary and categorical data, 4. Survival data, 5. Modelling carry-over, 6. Bioequivalence and 7. Components of variation. Finally, a brief assessment of the importance in this field of Statistics in Medicine is given.

Cross-Over Studies↗
PubMed2006

Early piribedil monotherapy of Parkinson's disease: A planned seven-month report of the REGAIN study.

Piribedil is a D2 dopamine agonist, which has been shown to improve symptoms of Parkinson's disease (PD) when combined with L-dopa. The objective of this study was to compare the efficacy of piribedil monotherapy to placebo in patients with early PD over a 7-month period. Four hundred and five early PD patients were randomized (double-blind) to piribedil (150-300 mg/day) or placebo. L-dopa open-label supplementation was permitted. Unified Parkinson Disease Rating Scale part III (UPDRS III) score as the last observation on monotherapy over 7 months was the primary outcome measure. Secondary outcomes were proportion of responders (UPDRS III improvement > 30%), patients remaining on monotherapy after 7 months, UPDRS III subscores, and UPDRS II. UPDRS III improved on piribedil (-4.9 points) versus a worsening on placebo (2.6 points; estimated effect = 7.26 points; 95% CI = 5.38-9.14; P < 0.0001). The proportion of responders was significantly higher for piribedil (42%) than for placebo (14%) (OR = 4.69; 95% CI = 2.82-7.80; P < 0.001). Piribedil significantly improved several UPDRS III subscores. UPDRS II improved on piribedil by -1.2 points, while it deteriorated by 1.5 points on placebo (estimated effect = 2.71; 95% CI = 1.8-3.62; P < 0.0001). The proportion of patients remaining on monotherapy after 7 months was greater in the piribedil group (OR = 3.72; 95% CI = 2.26-6.11; P < 0.001). Safety was consistent with that reported for other dopamine agonists, gastrointestinal side effects being the most common (22% of patients in piribedil group vs. 14% on placebo). Piribedil is effective and safe as early PD therapy.

Adult↗
PubMed2004

Controversies concerning randomization and additivity in clinical trials.

'As ye randomise so shall ye analyse', is one way of describing Fisher's defence of randomization. Yet, when it comes to clinical trials we nearly always randomize but we rarely analyse the way we randomize and Fisher himself was no exception. Two controversies involving Fisher in the 1930s are discussed: one with Neyman concerning additivity and the other with Student concerning randomization. Their relevance today is considered, as is whether randomization inference in clinical trials is dead and whether modelling rules the day, whether minimization is an acceptable procedure and to what extent trialists confuse experiments with surveys. It will be maintained that a number of different possible purposes of clinical trials have been confused because in the case of the general linear model, under strong additivity, they can all be satisfied by a single analysis. More generally, however, this is not the case.

Data Interpretation, Statistical↗
PubMed2003

A note on non-parametric ANCOVA for covariate adjustment in randomized clinical trials.

Koch et al. recently (1998) proposed two covariate-adjusted approaches for the comparison of continuous, ordinal and binary responses in a randomized clinical trial (Statist. Med. 1998; 17: 1863-1892). The first is a randomization approach while the second assumes that the study is a sample of a population. Here, we study the second approach and consider the simplest cases of two treatments with a continuous response and with a binary response. Koch's second approach will be compared with the classical ANCOVA for a continuous response. From this relationship we demonstrate that Koch's method cannot preserve the probability of the type I error. Simulations with continuous responses as well as with binary outcomes confirm the aforementioned theoretical result on the performance of Koch's method under the null hypothesis of no treatment effect. However, this poses only a problem for relatively small to moderate sample sizes. Further, as specified in the original paper of Koch et al., the first approach does preserve the type I error for any sample size, as the P-values can be reported in an exact manner (Statist. Med. 1998; 17: 1863-1892). Finally, we propose a correction factor for Koch's test statistic that better preserves the type I error.

Computer Simulation↗