Stephen Senn | Luxembourg Institute of Health (original) (raw)
Papers by Stephen Senn
Statistics & Probability Letters, 2011
A series of macros that have been created to perform fixed and random effects meta-analysis in SA... more A series of macros that have been created to perform fixed and random effects meta-analysis in SAS® are described as is the motivation for their creation. These macros are being made freely available on the internet for others to use. The application of the macros is illustrated using an example of trials in pre-eclampsia.
Pharmaceutical statistics
A recent analysis of R&D productivity suggests that there are grounds for 'cautious optimism&... more A recent analysis of R&D productivity suggests that there are grounds for 'cautious optimism' that the industry 'turned the corner' in 2008 and is 'on the comeback trail'. We believe that this analysis is flawed and most probably wrong. We present an alternative analysis of these same data to suggest that the industry is not yet 'out of the woods' and suggest that many of the systemic issues affecting pharmaceutical R&D productivity are still being resolved.
Pharmaceutical statistics
The European Medicines Agency (EMA) website lists all diseases that officially exist in adults on... more The European Medicines Agency (EMA) website lists all diseases that officially exist in adults only. The class waiver for juvenile melanoma was revoked in 2008 referring to US SEER statistics. This statistical justification is misleading. Melanoma in adolescents is much rarer than claimed by EMA/Paediatric Committee; < 1 ∕ 4 of adolescents with melanoma need systemic treatment; separate efficacy studies are neither medically justified nor feasible. The scarce adolescent patients should be allowed to participate in adult trials. To force companies to investigate them separately turns them into paediatric hostages, to adapt the term therapeutic orphans coined in 1968 by Shirkey. There are now five melanoma Paediatric Investigation Plans (PIPs). Probably none of the PIP-triggered clinical studies will ever be completed; we propose to call them ghost studies. An oncology research network considering a reasonable trial in melanoma, including adolescents, will compete for recruitment w...
Nature reviews. Drug discovery, 2015
ABSTRACT Progression-seeking bias gives rise to optimism bias in drug discovery and high levels o... more ABSTRACT Progression-seeking bias gives rise to optimism bias in drug discovery and high levels of late-stage attrition. By choosing to develop a product that is unmarketable we deny potentially marketable products the chance to be evaluated. False positives thereby incur opportunity costs, representing lost revenues to the organization, denying patients access to novel treatments. In this note we show how reframing false positives in terms of opportunity costs helps project teams view termination decisions as prospective gains. Reframing decisions in this way makes termination decisions easier.
Drugs in R&D, 2011
Background and Aim: Acetylsalicylic acid (ASA [aspirin]) is a commonly used over-the-counter drug... more Background and Aim: Acetylsalicylic acid (ASA [aspirin]) is a commonly used over-the-counter drug for the treatment of pain, fever, or colds, but data on the safety of this use are very limited. The aim of this study was to provide data on the safety of this treatment pattern, which is of interest to clinicians, regulators, and the public. Methods: A meta-analysis of individual patient data from 67 studies sponsored by Bayer HealthCare was completed. The primary endpoints were patient-reported gastrointestinal (GI) adverse events (AEs); the secondary endpoints were the incidence of patient-reported non-GI AEs. Event incidence and odds ratios (ORs) based on Cochran-Mantel-Haenszel estimates are reported. In total, 6181 patients were treated with ASA, 3515 with placebo, 1145 with acetaminophen (paracetamol), and 754 with ibuprofen. Exposure to ASA was short term (82.5% of patients had a single dose). Results: GI AEs were more frequent with ASA (9.9%) than with placebo (9.0%) [OR 1.3; 95% CI 1.1, 1.5]. Dyspeptic symptoms were infrequent (4.6% in placebo subjects). The ORs for ASA were 1.3 (95% CI 1.1, 1.6) versus placebo; 1.55 (95% CI 0.7, 3.3) versus ibuprofen; and 1.04 (95% CI 0.8, 1.4) versus acetaminophen. There were very few serious GI AEs (one ASA case; three placebo cases). No differences were found for non-GI AEs and no cases of cerebral hemorrhage were reported.
Drugs in R&D, 2013
Background and Objectives Aspirin is widely used for short-term treatment of pain, fever or colds... more Background and Objectives Aspirin is widely used for short-term treatment of pain, fever or colds, but there are only limited data regarding the safety of this use. To summarize the available data on this topic, we conducted a meta-analysis of the published clinical trial literature regarding the gastrointestinal adverse effects of short-term use of aspirin in comparison with placebo and other medications commonly used for the same purpose. Data Sources and Methods An extensive literature search identified 119,310 articles regarding possible adverse effects of aspirin, among which 23,131 appeared to possibly include relevant data. An automated text-mining procedure was used to score the references for potential relevance for the meta-analysis. The 3,983 highest-scoring articles were reviewed individually to identify those with data that could be included in this analysis. Ultimately, 78 relevant articles were identified that contained gastrointestinal adverse event data from clinical trials of aspirin versus placebo or an active comparator. Odds ratios (ORs) computed using a Mantel-Haenszel estimator were used to summarize the comparative effects on dyspepsia, nausea/vomiting, and abdominal pain, considered separately and also aggregated as 'minor gastrointestinal events'. Gastrointestinal bleeds, ulcers, and perforations were also investigated. Results Data were obtained regarding 19,829 subjects (34 % treated with aspirin, 17 % placebo, and 49 % an active comparator). About half of the aspirin subjects took a single dose. Aspirin was associated with a higher risk of minor gastrointestinal events than placebo or active comparators: the summary ORs were 1.46 (95 % confidence interval [CI] 1.15-1.86) and 1.81 (95 % CI 1.61-2.04), respectively. Ulcers, perforation, and serious bleeding were not seen after use of aspirin or any of the other interventions.
Statistics in Medicine, 2014
British Medical Journal, 2008
... Sara Hughes, chair of PSI (professional UK body of statisticians in the pharmaceutical indust... more ... Sara Hughes, chair of PSI (professional UK body of statisticians in the pharmaceutical industry) and director of statistics, GlaxoSmithKline, Greenford ... Pasi Korhonen, SSL chair (Statisticians in the Finnish Pharmaceutical Industry) and chief executive officer of StatFinn Oy, Finland ...
Pain, 2014
Proof-of-concept (POC) clinical trials play an important role in developing novel treatments and ... more Proof-of-concept (POC) clinical trials play an important role in developing novel treatments and determining whether existing treatments may be efficacious in broader populations of patients. The goal of most POC trials is to determine whether a treatment is likely to be efficacious for a given indication and thus whether it is worth investing the financial resources and participant exposure necessary for a confirmatory trial of that intervention. A challenge in designing POC trials is obtaining sufficient information to make this important go/no-go decision in a cost-effective manner. An IMMPACT consensus meeting was convened to discuss design considerations for POC trials in analgesia, with a focus on maximizing power with limited resources and participants. We present general design aspects to consider including patient population, active comparators and placebos, study power, pharmacokinetic-pharmacodynamic relationships, and minimization of missing data. Efficiency of single-do...
Christie/Simplicity, Complexity and Modelling, 2011
Statistics in Medicine, 2009
Statistics in Medicine, 2010
Some general points regarding efficiency in clinical trials are made. Reasons as to why fitting m... more Some general points regarding efficiency in clinical trials are made. Reasons as to why fitting many covariates to adjust the estimate of the treatment effect may be less problematic than commonly supposed are given. Two methods of dynamic allocation of patients based on covariates, minimization and Atkinson's approach, are compared and contrasted for the particular case where all covariates are binary. The results of Monte Carlo simulations are also presented. It is concluded that in the cases considered, Atkinson's approach is slightly more efficient than minimization although the difference is unlikely to be very important in practice. Both are more efficient than simple randomization, although it is concluded that fitting covariates may make a more valuable and instructive contribution to inferences about treatment effects than only balancing them.
Statistics in Medicine, 2009
Biostatisticians have frequently uncritically accepted the measurements provided by their medical... more Biostatisticians have frequently uncritically accepted the measurements provided by their medical colleagues engaged in clinical research. Such measures often involve considerable loss of information. Particularly, unfortunate is the widespread use of the so-called &amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;#39;responder analysis&amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;#39;, which may involve not only a loss of information through dichotomization, but also extravagant and unjustified causal inference regarding individual treatment effects at the patient level, and, increasingly, the use of the so-called number needed to treat scale of measurement. Other problems involve inefficient use of baseline measurements, the use of covariates measured after the start of treatment, the interpretation of titrations and composite response measures. Many of these bad practices are becoming enshrined in the regulatory guidance to the pharmaceutical industry. We consider the losses involved in inappropriate measures and suggest that statisticians should pay more attention to this aspect of their work.
Statistical Methods in Medical Research, 2011
It is a common belief that individual variation in response to treatment is an important explanat... more It is a common belief that individual variation in response to treatment is an important explanation for the variation in observed outcomes in clinical trials. If such variation is large, it seems reasonable to suppose that progress in treating disease will be advanced by classifying patients according to their abilities or not to &amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;#39;respond&amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;#39; to particular treatments. We consider that there is currently a lost opportunity in drug development. There is a great deal of talk about individual response to treatment and tailor-made drugs. However, relatively little work is being done to formally investigate, using suitable designs, where individual response to treatment may be important. Through a case study from a replicate cross-over study we show how, given suitable replication, it is possible to isolate the component of variation corresponding to patient-by-treatment interaction and hence investigate the possibility of individual response to treatment.
PLoS ONE, 2012
Objectives: REGULATE trial was designed to compare the efficacy and safety of benfluorex versus p... more Objectives: REGULATE trial was designed to compare the efficacy and safety of benfluorex versus pioglitazone in type 2 diabetes mellitus (DM) patients.
Pharmaceutical Statistics, 2007
A common approach to analysing clinical trials with multiple outcomes is to control the probabili... more A common approach to analysing clinical trials with multiple outcomes is to control the probability for the trial as a whole of making at least one incorrect positive finding under any configuration of true and false null hypotheses. Popular approaches are to use Bonferroni corrections or structured approaches such as, for example, closed-test procedures. As is well known, such strategies, which control the family-wise error rate, typically reduce the type I error for some or all the tests of the various null hypotheses to below the nominal level. In consequence, there is generally a loss of power for individual tests. What is less well appreciated, perhaps, is that depending on approach and circumstances, the test-wise loss of power does not necessarily lead to a family wise loss of power. In fact, it may be possible to increase the overall power of a trial by carrying out tests on multiple outcomes without increasing the probability of making at least one type I error when all null hypotheses are true. We examine two types of problems to illustrate this. Unstructured testing problems arise typically (but not exclusively) when many outcomes are being measured. We consider the case of more than two hypotheses when a Bonferroni approach is being applied while for illustration we assume compound symmetry to hold for the correlation of all variables. Using the device of a latent variable it is easy to show that power is not reduced as the number of variables tested increases, provided that the common correlation coefficient is not too high (say less than 0.75). Afterwards, we will consider structured testing problems. Here, multiplicity problems arising from the comparison of more than two treatments, as opposed to more than one measurement, are typical. We conduct a numerical study and conclude again that power is not reduced as the number of tested variables increases.
Statistics & Probability Letters, 2011
A series of macros that have been created to perform fixed and random effects meta-analysis in SA... more A series of macros that have been created to perform fixed and random effects meta-analysis in SAS® are described as is the motivation for their creation. These macros are being made freely available on the internet for others to use. The application of the macros is illustrated using an example of trials in pre-eclampsia.
Pharmaceutical statistics
A recent analysis of R&D productivity suggests that there are grounds for 'cautious optimism&... more A recent analysis of R&D productivity suggests that there are grounds for 'cautious optimism' that the industry 'turned the corner' in 2008 and is 'on the comeback trail'. We believe that this analysis is flawed and most probably wrong. We present an alternative analysis of these same data to suggest that the industry is not yet 'out of the woods' and suggest that many of the systemic issues affecting pharmaceutical R&D productivity are still being resolved.
Pharmaceutical statistics
The European Medicines Agency (EMA) website lists all diseases that officially exist in adults on... more The European Medicines Agency (EMA) website lists all diseases that officially exist in adults only. The class waiver for juvenile melanoma was revoked in 2008 referring to US SEER statistics. This statistical justification is misleading. Melanoma in adolescents is much rarer than claimed by EMA/Paediatric Committee; < 1 ∕ 4 of adolescents with melanoma need systemic treatment; separate efficacy studies are neither medically justified nor feasible. The scarce adolescent patients should be allowed to participate in adult trials. To force companies to investigate them separately turns them into paediatric hostages, to adapt the term therapeutic orphans coined in 1968 by Shirkey. There are now five melanoma Paediatric Investigation Plans (PIPs). Probably none of the PIP-triggered clinical studies will ever be completed; we propose to call them ghost studies. An oncology research network considering a reasonable trial in melanoma, including adolescents, will compete for recruitment w...
Nature reviews. Drug discovery, 2015
ABSTRACT Progression-seeking bias gives rise to optimism bias in drug discovery and high levels o... more ABSTRACT Progression-seeking bias gives rise to optimism bias in drug discovery and high levels of late-stage attrition. By choosing to develop a product that is unmarketable we deny potentially marketable products the chance to be evaluated. False positives thereby incur opportunity costs, representing lost revenues to the organization, denying patients access to novel treatments. In this note we show how reframing false positives in terms of opportunity costs helps project teams view termination decisions as prospective gains. Reframing decisions in this way makes termination decisions easier.
Drugs in R&D, 2011
Background and Aim: Acetylsalicylic acid (ASA [aspirin]) is a commonly used over-the-counter drug... more Background and Aim: Acetylsalicylic acid (ASA [aspirin]) is a commonly used over-the-counter drug for the treatment of pain, fever, or colds, but data on the safety of this use are very limited. The aim of this study was to provide data on the safety of this treatment pattern, which is of interest to clinicians, regulators, and the public. Methods: A meta-analysis of individual patient data from 67 studies sponsored by Bayer HealthCare was completed. The primary endpoints were patient-reported gastrointestinal (GI) adverse events (AEs); the secondary endpoints were the incidence of patient-reported non-GI AEs. Event incidence and odds ratios (ORs) based on Cochran-Mantel-Haenszel estimates are reported. In total, 6181 patients were treated with ASA, 3515 with placebo, 1145 with acetaminophen (paracetamol), and 754 with ibuprofen. Exposure to ASA was short term (82.5% of patients had a single dose). Results: GI AEs were more frequent with ASA (9.9%) than with placebo (9.0%) [OR 1.3; 95% CI 1.1, 1.5]. Dyspeptic symptoms were infrequent (4.6% in placebo subjects). The ORs for ASA were 1.3 (95% CI 1.1, 1.6) versus placebo; 1.55 (95% CI 0.7, 3.3) versus ibuprofen; and 1.04 (95% CI 0.8, 1.4) versus acetaminophen. There were very few serious GI AEs (one ASA case; three placebo cases). No differences were found for non-GI AEs and no cases of cerebral hemorrhage were reported.
Drugs in R&D, 2013
Background and Objectives Aspirin is widely used for short-term treatment of pain, fever or colds... more Background and Objectives Aspirin is widely used for short-term treatment of pain, fever or colds, but there are only limited data regarding the safety of this use. To summarize the available data on this topic, we conducted a meta-analysis of the published clinical trial literature regarding the gastrointestinal adverse effects of short-term use of aspirin in comparison with placebo and other medications commonly used for the same purpose. Data Sources and Methods An extensive literature search identified 119,310 articles regarding possible adverse effects of aspirin, among which 23,131 appeared to possibly include relevant data. An automated text-mining procedure was used to score the references for potential relevance for the meta-analysis. The 3,983 highest-scoring articles were reviewed individually to identify those with data that could be included in this analysis. Ultimately, 78 relevant articles were identified that contained gastrointestinal adverse event data from clinical trials of aspirin versus placebo or an active comparator. Odds ratios (ORs) computed using a Mantel-Haenszel estimator were used to summarize the comparative effects on dyspepsia, nausea/vomiting, and abdominal pain, considered separately and also aggregated as 'minor gastrointestinal events'. Gastrointestinal bleeds, ulcers, and perforations were also investigated. Results Data were obtained regarding 19,829 subjects (34 % treated with aspirin, 17 % placebo, and 49 % an active comparator). About half of the aspirin subjects took a single dose. Aspirin was associated with a higher risk of minor gastrointestinal events than placebo or active comparators: the summary ORs were 1.46 (95 % confidence interval [CI] 1.15-1.86) and 1.81 (95 % CI 1.61-2.04), respectively. Ulcers, perforation, and serious bleeding were not seen after use of aspirin or any of the other interventions.
Statistics in Medicine, 2014
British Medical Journal, 2008
... Sara Hughes, chair of PSI (professional UK body of statisticians in the pharmaceutical indust... more ... Sara Hughes, chair of PSI (professional UK body of statisticians in the pharmaceutical industry) and director of statistics, GlaxoSmithKline, Greenford ... Pasi Korhonen, SSL chair (Statisticians in the Finnish Pharmaceutical Industry) and chief executive officer of StatFinn Oy, Finland ...
Pain, 2014
Proof-of-concept (POC) clinical trials play an important role in developing novel treatments and ... more Proof-of-concept (POC) clinical trials play an important role in developing novel treatments and determining whether existing treatments may be efficacious in broader populations of patients. The goal of most POC trials is to determine whether a treatment is likely to be efficacious for a given indication and thus whether it is worth investing the financial resources and participant exposure necessary for a confirmatory trial of that intervention. A challenge in designing POC trials is obtaining sufficient information to make this important go/no-go decision in a cost-effective manner. An IMMPACT consensus meeting was convened to discuss design considerations for POC trials in analgesia, with a focus on maximizing power with limited resources and participants. We present general design aspects to consider including patient population, active comparators and placebos, study power, pharmacokinetic-pharmacodynamic relationships, and minimization of missing data. Efficiency of single-do...
Christie/Simplicity, Complexity and Modelling, 2011
Statistics in Medicine, 2009
Statistics in Medicine, 2010
Some general points regarding efficiency in clinical trials are made. Reasons as to why fitting m... more Some general points regarding efficiency in clinical trials are made. Reasons as to why fitting many covariates to adjust the estimate of the treatment effect may be less problematic than commonly supposed are given. Two methods of dynamic allocation of patients based on covariates, minimization and Atkinson's approach, are compared and contrasted for the particular case where all covariates are binary. The results of Monte Carlo simulations are also presented. It is concluded that in the cases considered, Atkinson's approach is slightly more efficient than minimization although the difference is unlikely to be very important in practice. Both are more efficient than simple randomization, although it is concluded that fitting covariates may make a more valuable and instructive contribution to inferences about treatment effects than only balancing them.
Statistics in Medicine, 2009
Biostatisticians have frequently uncritically accepted the measurements provided by their medical... more Biostatisticians have frequently uncritically accepted the measurements provided by their medical colleagues engaged in clinical research. Such measures often involve considerable loss of information. Particularly, unfortunate is the widespread use of the so-called &amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;#39;responder analysis&amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;#39;, which may involve not only a loss of information through dichotomization, but also extravagant and unjustified causal inference regarding individual treatment effects at the patient level, and, increasingly, the use of the so-called number needed to treat scale of measurement. Other problems involve inefficient use of baseline measurements, the use of covariates measured after the start of treatment, the interpretation of titrations and composite response measures. Many of these bad practices are becoming enshrined in the regulatory guidance to the pharmaceutical industry. We consider the losses involved in inappropriate measures and suggest that statisticians should pay more attention to this aspect of their work.
Statistical Methods in Medical Research, 2011
It is a common belief that individual variation in response to treatment is an important explanat... more It is a common belief that individual variation in response to treatment is an important explanation for the variation in observed outcomes in clinical trials. If such variation is large, it seems reasonable to suppose that progress in treating disease will be advanced by classifying patients according to their abilities or not to &amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;#39;respond&amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;#39; to particular treatments. We consider that there is currently a lost opportunity in drug development. There is a great deal of talk about individual response to treatment and tailor-made drugs. However, relatively little work is being done to formally investigate, using suitable designs, where individual response to treatment may be important. Through a case study from a replicate cross-over study we show how, given suitable replication, it is possible to isolate the component of variation corresponding to patient-by-treatment interaction and hence investigate the possibility of individual response to treatment.
PLoS ONE, 2012
Objectives: REGULATE trial was designed to compare the efficacy and safety of benfluorex versus p... more Objectives: REGULATE trial was designed to compare the efficacy and safety of benfluorex versus pioglitazone in type 2 diabetes mellitus (DM) patients.
Pharmaceutical Statistics, 2007
A common approach to analysing clinical trials with multiple outcomes is to control the probabili... more A common approach to analysing clinical trials with multiple outcomes is to control the probability for the trial as a whole of making at least one incorrect positive finding under any configuration of true and false null hypotheses. Popular approaches are to use Bonferroni corrections or structured approaches such as, for example, closed-test procedures. As is well known, such strategies, which control the family-wise error rate, typically reduce the type I error for some or all the tests of the various null hypotheses to below the nominal level. In consequence, there is generally a loss of power for individual tests. What is less well appreciated, perhaps, is that depending on approach and circumstances, the test-wise loss of power does not necessarily lead to a family wise loss of power. In fact, it may be possible to increase the overall power of a trial by carrying out tests on multiple outcomes without increasing the probability of making at least one type I error when all null hypotheses are true. We examine two types of problems to illustrate this. Unstructured testing problems arise typically (but not exclusively) when many outcomes are being measured. We consider the case of more than two hypotheses when a Bonferroni approach is being applied while for illustration we assume compound symmetry to hold for the correlation of all variables. Using the device of a latent variable it is easy to show that power is not reduced as the number of variables tested increases, provided that the common correlation coefficient is not too high (say less than 0.75). Afterwards, we will consider structured testing problems. Here, multiplicity problems arising from the comparison of more than two treatments, as opposed to more than one measurement, are typical. We conduct a numerical study and conclude again that power is not reduced as the number of tested variables increases.