Search tips
Search criteria

Results 1-25 (1614216)

Clipboard (0)

Related Articles

When comparing a new treatment with a control in a randomized clinical study, the treatment effect is generally assessed by evaluating a summary measure over a specific study population. The success of the trial heavily depends on the choice of such a population. In this paper, we show a systematic, effective way to identify a promising population, for which the new treatment is expected to have a desired benefit, utilizing the data from a current study involving similar comparator treatments. Specifically, using the existing data, we first create a parametric scoring system as a function of multiple multiple baseline covariates to estimate subject-specific treatment differences. Based on this scoring system, we specify a desired level of treatment difference and obtain a subgroup of patients, defined as those whose estimated scores exceed this threshold. An empirically calibrated threshold-specific treatment difference curve across a range of score values is constructed. The subpopulation of patients satisfying any given level of treatment benefit can then be identified accordingly. To avoid bias due to overoptimism, we utilize a cross-training-evaluation method for implementing the above two-step procedure. We then show how to select the best scoring system among all competing models. Furthermore, for cases in which only a single pre-specified working model is involved, inference procedures are proposed for the average treatment difference over a range of score values using the entire data set, and are justified theoretically and numerically. Lastly, the proposals are illustrated with the data from two clinical trials in treating HIV and cardiovascular diseases. Note that if we are not interested in designing a new study for comparing similar treatments, the new procedure can also be quite useful for the management of future patients, so that treatment may be targeted towards those who would receive nontrivial benefits to compensate for the risk or cost of the new treatment.
PMCID: PMC3775385  PMID: 24058223
Cross-training-evaluation; Lasso procedure; Personalized medicine; Prediction; Ridge regression; Stratified medicine; Subgroup analysis; Variable selection
Biometrics  2010;67(2):427-435.
In a longitudinal study, suppose that the primary endpoint is the time to a specific event. This response variable, however, may be censored by an independent censoring variable or by the occurrence of one of several dependent competing events. For each study subject, a set of baseline covariates is collected. The question is how to construct a reliable prediction rule for the future subject’s profile of all competing risks of interest at a specific time point for risk-benefit decision makings. In this paper, we propose a two-stage procedure to make inferences about such subject-specific profiles. For the first step, we use a parametric model to obtain a univariate risk index score system. We then estimate consistently the average competing risks for subjects which have the same parametric index score via a nonparametric function estimation procedure. We illustrate this new proposal with the data from a randomized clinical trial for evaluating the efficacy of a treatment for prostate cancer. The primary endpoint for this study was the time to prostate cancer death, but had two types of dependent competing events, one from cardiovascular death and the other from death of other causes.
PMCID: PMC2970653  PMID: 20618311
Local likelihood function; Nonparametric function estimation; Perturbation-resampling method; Risk index score
3.  Methods of Blinding in Reports of Randomized Controlled Trials Assessing Pharmacologic Treatments: A Systematic Review 
PLoS Medicine  2006;3(10):e425.
Blinding is a cornerstone of therapeutic evaluation because lack of blinding can bias treatment effect estimates. An inventory of the blinding methods would help trialists conduct high-quality clinical trials and readers appraise the quality of results of published trials. We aimed to systematically classify and describe methods to establish and maintain blinding of patients and health care providers and methods to obtain blinding of outcome assessors in randomized controlled trials of pharmacologic treatments.
Methods and Findings
We undertook a systematic review of all reports of randomized controlled trials assessing pharmacologic treatments with blinding published in 2004 in high impact-factor journals from Medline and the Cochrane Methodology Register. We used a standardized data collection form to extract data. The blinding methods were classified according to whether they primarily (1) established blinding of patients or health care providers, (2) maintained the blinding of patients or health care providers, and (3) obtained blinding of assessors of the main outcomes. We identified 819 articles, with 472 (58%) describing the method of blinding. Methods to establish blinding of patients and/or health care providers concerned mainly treatments provided in identical form, specific methods to mask some characteristics of the treatments (e.g., added flavor or opaque coverage), or use of double dummy procedures or simulation of an injection. Methods to avoid unblinding of patients and/or health care providers involved use of active placebo, centralized assessment of side effects, patients informed only in part about the potential side effects of each treatment, centralized adapted dosage, or provision of sham results of complementary investigations. The methods reported for blinding outcome assessors mainly relied on a centralized assessment of complementary investigations, clinical examination (i.e., use of video, audiotape, or photography), or adjudication of clinical events.
This review classifies blinding methods and provides a detailed description of methods that could help trialists overcome some barriers to blinding in clinical trials and readers interpret the quality of pharmalogic trials.
Following a systematic review of all reports of randomized controlled trials assessing pharmacologic treatments involving blinding, a classification of blinding methods is proposed.
Editors' Summary
In evidence-based medicine, good-quality randomized controlled trials are generally considered to be the most reliable source of information about the effects of different treatments, such as drugs. In a randomized trial, patients are assigned to receive one treatment or another by the play of chance. This technique helps makes sure that the two groups of patients receiving the different treatments are equivalent at the start of the trial. Proper randomization also prevents doctors from controlling or affecting which treatment patients get, which could distort the results. An additional tool that is also used to make trials more precise is “blinding.” Blinding involves taking steps to prevent patients, doctors, or other people involved in the trial (e.g., those people recording measurements) from finding out which patients got what treatment. Properly done, blinding should make sure the results of a trial are more accurate. This is because in an unblinded study, participants may respond better if they know they have received a promising new treatment (or worse if they only got placebo or an old drug); doctors may “want” a particular treatment to do better in the trial, and unthinking bias could creep into their measurements or actions; the same applies for practitioners and researchers who record patients' outcomes in the trial. However, blinding is not a simple, single step; the people carrying out the trial often have to set up a variety of different procedures that depend on the type of trial that is being done.
Why Was This Study Done?
The researchers here wanted to thoroughly examine different methods that have been used to achieve blinding in randomized trials of drug treatments, and to describe and classify them. They hoped that a better understanding of the different blinding methods would help people doing trials to design better trials in the future, and also help readers to interpret the quality of trials that had been done.
What Did the Researchers Do and Find?
This group of researchers conducted what is called a “systematic review.” They systematically searched the published medical literature to find all randomized, blinded drug trials published in 2004 in a number of different “high-impact” journals (journals whose articles are often mentioned in other articles). Then, the researchers classified information from the published trial reports. The researchers ended up with 819 trial reports, and nearly 60% of them described how blinding was done. Their classification of blinding was divided up into three main areas. First, they detailed methods used to hide which drugs are given to particular patients, such as preparing identically appearing treatments; using strong flavors to mask taste; matching the colors of pills; using saline injections and so on. Second, they described a number of methods that could be used to reduce the risk of unblinding (of doctors or patients), such as using an “active placebo” (a sugar pill that mimics some of the expected side effects of the drug treatment). Finally, they defined methods for blinded measurement of outcomes (such as using a central committee to collect data).
What Do These Findings Mean?
The researchers' classification will help people to work out how different techniques can be used to achieve, and keep, blinding in a trial. This will assist others to understand whether any particular trial was likely to have been blinded properly, and therefore work out whether the results are reliable. The researchers also suggest that, generally, blinding methods are not described in enough detail in published scientific papers, and recommend that guidelines for describing results of randomized trials be improved.
Additional Information.
Please access these Web sites via the online version of this summary at
James Lind Library has been created to help patients and researchers understand fair tests of treatments in health care by illustrating how fair tests have developed over the centuries, a trial registry created by the US National Institutes of Health, has an introduction to understanding clinical trials
National Electronic Library for Health introduction to controlled clinical trials
PMCID: PMC1626553  PMID: 17076559
4.  The Project Data Sphere Initiative: Accelerating Cancer Research by Sharing Data 
The Oncologist  2015;20(5):464-e20.
By providing access to large, late-phase, cancer-trial data sets, the Project Data Sphere initiative has the potential to transform cancer research by optimizing research efficiency and accelerating progress toward meaningful improvements in cancer care. This type of platform provides opportunities for unique research projects that can examine relatively neglected areas and that can construct models necessitating large amounts of detailed data.
In this paper, we provide background and context regarding the potential for a new data-sharing platform, the Project Data Sphere (PDS) initiative, funded by financial and in-kind contributions from the CEO Roundtable on Cancer, to transform cancer research and improve patient outcomes. Given the relatively modest decline in cancer death rates over the past several years, a new research paradigm is needed to accelerate therapeutic approaches for oncologic diseases. Phase III clinical trials generate large volumes of potentially usable information, often on hundreds of patients, including patients treated with standard of care therapies (i.e., controls). Both nationally and internationally, a variety of stakeholders have pursued data-sharing efforts to make individual patient-level clinical trial data available to the scientific research community.
Potential Benefits and Risks of Data Sharing.
For researchers, shared data have the potential to foster a more collaborative environment, to answer research questions in a shorter time frame than traditional randomized control trials, to reduce duplication of effort, and to improve efficiency. For industry participants, use of trial data to answer additional clinical questions could increase research and development efficiency and guide future projects through validation of surrogate end points, development of prognostic or predictive models, selection of patients for phase II trials, stratification in phase III studies, and identification of patient subgroups for development of novel therapies. Data transparency also helps promote a public image of collaboration and altruism among industry participants. For patient participants, data sharing maximizes their contribution to public health and increases access to information that may be used to develop better treatments. Concerns about data-sharing efforts include protection of patient privacy and confidentiality. To alleviate these concerns, data sets are deidentified to maintain anonymity. To address industry concerns about protection of intellectual property and competitiveness, we illustrate several models for data sharing with varying levels of access to the data and varying relationships between trial sponsors and data access sponsors.
The Project Data Sphere Initiative.
PDS is an independent initiative of the CEO Roundtable on Cancer Life Sciences Consortium, built to voluntarily share, integrate, and analyze comparator arms of historical cancer clinical trial data sets to advance future cancer research. The aim is to provide a neutral, broad-access platform for industry and academia to share raw, deidentified data from late-phase oncology clinical trials using comparator-arm data sets. These data are likely to be hypothesis generating or hypothesis confirming but, notably, do not take the place of performing a well-designed trial to address a specific hypothesis. Prospective providers of data to PDS complete and sign a data sharing agreement that includes a description of the data they propose to upload, and then they follow easy instructions on the website for uploading their deidentified data. The SAS Institute has also collaborated with the initiative to provide intrinsic analytic tools accessible within the website itself.
As of October 2014, the PDS website has available data from 14 cancer clinical trials covering 9,000 subjects, with hopes to further expand the database to include more than 25,000 subject accruals within the next year. PDS differentiates itself from other data-sharing initiatives by its degree of openness, requiring submission of only a brief application with background information of the individual requesting access and agreement to terms of use. Data from several different sponsors may be pooled to develop a comprehensive cohort for analysis. In order to protect patient privacy, data providers in the U.S. are responsible for deidentifying data according to standards set forth by the Privacy Rule of the U.S. Health Insurance Portability and Accountability Act of 1996.
Using Data Sharing to Improve Outcomes in Cancer: The “Prostate Cancer Challenge.”
Control-arm data of several studies among patients with metastatic castration-resistant prostate cancer (mCRPC) are currently available through PDS. These data sets have multiple potential uses. The “Prostate Cancer Challenge” will ask the cancer research community to use clinical trial data deposited in the PDS website to address key research questions regarding mCRPC.
General themes that could be explored by the cancer community are described in this article: prognostic models evaluating the influence of pretreatment factors on survival and patient-reported outcomes; comparative effectiveness research evaluating the efficacy of standard of care therapies, as illustrated in our companion article comparing mitoxantrone plus prednisone with prednisone alone; effects of practice variation in dose, frequency, and duration of therapy; level of patient adherence to elements of trial protocols to inform the design of future clinical trials; and age of subjects, regional differences in health care, and other confounding factors that might affect outcomes.
Potential Limitations and Methodological Challenges.
The number of data sets available and the lack of experimental-arm data limit the potential scope of research using the current PDS. The number of trials is expected to grow exponentially over the next year and may include multiple cancer settings, such as breast, colorectal, lung, hematologic malignancy, and bone marrow transplantation. Other potential limitations include the retrospective nature of the data analyses performed using PDS and its generalizability, given that clinical trials are often conducted among younger, healthier, and less racially diverse patient populations. Methodological challenges exist when combining individual patient data from multiple clinical trials; however, advancements in statistical methods for secondary database analysis offer many tools for reanalyzing data arising from disparate trials, such as propensity score matching. Despite these concerns, few if any comparable data sets include this level of detail across multiple clinical trials and populations.
Access to large, late-phase, cancer-trial data sets has the potential to transform cancer research by optimizing research efficiency and accelerating progress toward meaningful improvements in cancer care. This type of platform provides opportunities for unique research projects that can examine relatively neglected areas and that can construct models necessitating large amounts of detailed data. The full potential of PDS will be realized only when multiple tumor types and larger numbers of data sets are available through the website.
PMCID: PMC4425388  PMID: 25876994
Project Data Sphere; Data sharing; Prostate cancer; Comparative effectiveness research
5.  Profile local linear estimation of generalized semiparametric regression model for longitudinal data 
Lifetime data analysis  2013;19(3):317-349.
This paper studies the generalized semiparametric regression model for longitudinal data where the covariate effects are constant for some and time-varying for others. Different link functions can be used to allow more flexible modelling of longitudinal data. The nonparametric components of the model are estimated using a local linear estimating equation and the parametric components are estimated through a profile estimating function. The method automatically adjusts for heterogeneity of sampling times, allowing the sampling strategy to depend on the past sampling history as well as possibly time-dependent covariates without specifically model such dependence. A K -fold cross-validation bandwidth selection is proposed as a working tool for locating an appropriate bandwidth. A criteria for selecting the link function is proposed to provide better fit of the data. Large sample properties of the proposed estimators are investigated. Large sample pointwise and simultaneous confidence intervals for the regression coefficients are constructed. Formal hypothesis testing procedures are proposed to check for the covariate effects and whether the effects are time-varying. A simulation study is conducted to examine the finite sample performances of the proposed estimation and hypothesis testing procedures. The methods are illustrated with a data example.
PMCID: PMC3710313  PMID: 23471814
Asymptotics; Kernel smoothing; Link function; Sampling adjusted estimation; Testing time-varying effects; Weighted least squares
6.  Evaluating the Effect of Early Versus Late ARV Regimen Change if Failure on an Initial Regimen: Results From the AIDS Clinical Trials Group Study A5095 
The current goal of initial antiretroviral (ARV) therapy is suppression of plasma human immunodeficiency virus (HIV)-1 RNA levels to below 200 copies per milliliter. A proportion of HIV-infected patients who initiate antiretroviral therapy in clinical practice or antiretroviral clinical trials either fail to suppress HIV-1 RNA or have HIV-1 RNA levels rebound on therapy. Frequently, these patients have sustained CD4 cell counts responses and limited or no clinical symptoms and, therefore, have potentially limited indications for altering therapy which they may be tolerating well despite increased viral replication. On the other hand, increased viral replication on therapy leads to selection of resistance mutations to the antiretroviral agents comprising their therapy and potentially cross-resistance to other agents in the same class decreasing the likelihood of response to subsequent antiretroviral therapy. The optimal time to switch antiretroviral therapy to ensure sustained virologic suppression and prevent clinical events in patients who have rebound in their HIV-1 RNA, yet are stable, is not known. Randomized clinical trials to compare early versus delayed switching have been difficult to design and more difficult to enroll. In some clinical trials, such as the AIDS Clinical Trials Group (ACTG) Study A5095, patients randomized to initial antiretroviral treatment combinations, who fail to suppress HIV-1 RNA or have a rebound of HIV-1 RNA on therapy are allowed to switch from the initial ARV regimen to a new regimen, based on clinician and patient decisions. We delineate a statistical framework to estimate the effect of early versus late regimen change using data from ACTG A5095 in the context of two-stage designs.
In causal inference, a large class of doubly robust estimators are derived through semiparametric theory with applications to missing data problems. This class of estimators is motivated through geometric arguments and relies on large samples for good performance. By now, several authors have noted that a doubly robust estimator may be suboptimal when the outcome model is misspecified even if it is semiparametric efficient when the outcome regression model is correctly specified. Through auxiliary variables, two-stage designs, and within the contextual backdrop of our scientific problem and clinical study, we propose improved doubly robust, locally efficient estimators of a population mean and average causal effect for early versus delayed switching to second-line ARV treatment regimens. Our analysis of the ACTG A5095 data further demonstrates how methods that use auxiliary variables can improve over methods that ignore them. Using the methods developed here, we conclude that patients who switch within 8 weeks of virologic failure have better clinical outcomes, on average, than patients who delay switching to a new second-line ARV regimen after failing on the initial regimen. Ordinary statistical methods fail to find such differences. This article has online supplementary material.
PMCID: PMC3545451  PMID: 23329858
Causal inference; Double robustness; Longitudinal data analysis; Missing data; Rubin causal model; Semiparametric efficient estimation
7.  Evidence for the Selective Reporting of Analyses and Discrepancies in Clinical Trials: A Systematic Review of Cohort Studies of Clinical Trials 
PLoS Medicine  2014;11(6):e1001666.
In a systematic review of cohort studies, Kerry Dwan and colleagues examine the evidence for selective reporting and discrepancies in analyses between journal publications and other documents for clinical trials.
Please see later in the article for the Editors' Summary
Most publications about selective reporting in clinical trials have focussed on outcomes. However, selective reporting of analyses for a given outcome may also affect the validity of findings. If analyses are selected on the basis of the results, reporting bias may occur. The aims of this study were to review and summarise the evidence from empirical cohort studies that assessed discrepant or selective reporting of analyses in randomised controlled trials (RCTs).
Methods and Findings
A systematic review was conducted and included cohort studies that assessed any aspect of the reporting of analyses of RCTs by comparing different trial documents, e.g., protocol compared to trial report, or different sections within a trial publication. The Cochrane Methodology Register, Medline (Ovid), PsycInfo (Ovid), and PubMed were searched on 5 February 2014. Two authors independently selected studies, performed data extraction, and assessed the methodological quality of the eligible studies. Twenty-two studies (containing 3,140 RCTs) published between 2000 and 2013 were included. Twenty-two studies reported on discrepancies between information given in different sources. Discrepancies were found in statistical analyses (eight studies), composite outcomes (one study), the handling of missing data (three studies), unadjusted versus adjusted analyses (three studies), handling of continuous data (three studies), and subgroup analyses (12 studies). Discrepancy rates varied, ranging from 7% (3/42) to 88% (7/8) in statistical analyses, 46% (36/79) to 82% (23/28) in adjusted versus unadjusted analyses, and 61% (11/18) to 100% (25/25) in subgroup analyses. This review is limited in that none of the included studies investigated the evidence for bias resulting from selective reporting of analyses. It was not possible to combine studies to provide overall summary estimates, and so the results of studies are discussed narratively.
Discrepancies in analyses between publications and other study documentation were common, but reasons for these discrepancies were not discussed in the trial reports. To ensure transparency, protocols and statistical analysis plans need to be published, and investigators should adhere to these or explain discrepancies.
Please see later in the article for the Editors' Summary
Editors' Summary
In the past, clinicians relied on their own experience when choosing the best treatment for their patients. Nowadays, they turn to evidence-based medicine—the systematic review and appraisal of trials, studies that investigate the benefits and harms of medical treatments in patients. However, evidence-based medicine can guide clinicians only if all the results from clinical trials are published in an unbiased and timely manner. Unfortunately, the results of trials in which a new drug performs better than existing drugs are more likely to be published than those in which the new drug performs badly or has unwanted side effects (publication bias). Moreover, trial outcomes that support the use of a new treatment are more likely to be published than those that do not support its use (outcome reporting bias). Recent initiatives—such as making registration of clinical trials in a trial registry (for example, a prerequisite for publication in medical journals—aim to prevent these biases, which pose a threat to informed medical decision-making.
Why Was This Study Done?
Selective reporting of analyses of outcomes may also affect the validity of clinical trial findings. Sometimes, for example, a trial publication will include a per protocol analysis (which considers only the outcomes of patients who received their assigned treatment) rather than a pre-planned intention-to-treat analysis (which considers the outcomes of all the patients regardless of whether they received their assigned treatment). If the decision to publish the per protocol analysis is based on the results of this analysis being more favorable than those of the intention-to-treat analysis (which more closely resembles “real” life), then “analysis reporting bias” has occurred. In this systematic review, the researchers investigate the selective reporting of analyses and discrepancies in randomized controlled trials (RCTs) by reviewing published studies that assessed selective reporting of analyses in groups (cohorts) of RCTs and discrepancies in analyses of RCTs between different sources (for example, between the protocol in a trial registry and the journal publication) or different sections of a source. A systematic review uses predefined criteria to identify all the research on a given topic.
What Did the Researchers Do and Find?
The researchers identified 22 cohort studies (containing 3,140 RCTs) that were eligible for inclusion in their systematic review. All of these studies reported on discrepancies between the information provided by the RCTs in different places, but none investigated the evidence for analysis reporting bias. Several of the cohort studies reported, for example, that there were discrepancies in the statistical analyses included in the different documents associated with the RCTs included in their analysis. Other types of discrepancies reported by the cohort studies included discrepancies in the reporting of composite outcomes (an outcome in which multiple end points are combined) and in the reporting of subgroup analyses (investigations of outcomes in subgroups of patients that should be predefined in the trial protocol to avoid bias). Discrepancy rates varied among the RCTs according to the types of analyses and cohort studies considered. Thus, whereas in one cohort study discrepancies were present in the statistical test used for the analysis of the primary outcome in only 7% of the included studies, they were present in the subgroup analyses of all the included studies.
What Do These Findings Mean?
These findings indicate that discrepancies in analyses between publications and other study documents such as protocols in trial registries are common. The reasons for these discrepancies in analyses were not discussed in trial reports but may be the result of reporting bias, errors, or legitimate departures from a pre-specified protocol. For example, a statistical analysis that is not specified in the trial protocol may sometimes appear in a publication because the journal requested its inclusion as a condition of publication. The researchers suggest that it may be impossible for systematic reviewers to distinguish between these possibilities simply by looking at the source documentation. Instead, they suggest, it may be necessary for reviewers to contact the trial authors. However, to make selective reporting of analyses more easily detectable, they suggest that protocols and analysis plans should be published and that investigators should be required to stick to these plans or explain any discrepancies when they publish their trial results. Together with other initiatives, this approach should help improve the quality of evidence-based medicine and, as a result, the treatment of patients.
Additional Information
Please access these websites via the online version of this summary at
Wikipedia has pages on evidence-based medicine, on systematic reviews, and on publication bias (note: Wikipedia is a free online encyclopedia that anyone can edit; available in several languages) provides information about the US National Institutes of Health clinical trial registry, including background information about clinical trials
The Cochrane Collaboration is a global independent network of health practitioners, researchers, patient advocates, and others that aims to promote evidence-informed health decision-making by producing high-quality, relevant, accessible systematic reviews and other synthesized research evidence; the Cochrane Handbook for Systematic Reviews of Interventions describes the preparation of systematic reviews in detail
PLOS Medicine recently launched a Reporting Guidelines Collection, an open-access collection of reporting guidelines, commentary, and related research on guidelines from across PLOS journals that aims to help advance the efficiency, effectiveness, and equitability of the dissemination of biomedical information
PMCID: PMC4068996  PMID: 24959719
8.  Subgroup identification from randomized clinical trial data 
Statistics in medicine  2011;30(24):10.1002/sim.4322.
We consider the problem of identifying a subgroup of patients who may have an enhanced treatment effect in a randomized clinical trial, and it is desirable that the subgroup be defined by a limited number of covariates. For this problem, the development of a standard, pre-determined strategy may help to avoid the well-known dangers of subgroup analysis. We present a method developed to find subgroups of enhanced treatment effect. This method, referred to as “Virtual Twins”, involves predicting response probabilities for treatment and control “twins” for each subject. The difference in these probabilities is then used as the outcome in a classification or regression tree, which can potentially include any set of the covariates. We define a measure Q(Â) to be the difference between the treatment effect in estimated subgroup  and the marginal treatment effect. We present several methods developed to obtain an estimate of Q(Â), including estimation of Q(Â) using estimated probabilities in the original data, using estimated probabilities in newly simulated data, two cross-validation-based approaches and a bootstrap-based bias corrected approach. Results of a simulation study indicate that the Virtual Twins method noticeably outperforms logistic regression with forward selection when a true subgroup of enhanced treatment effect exists. Generally, large sample sizes or strong enhanced treatment effects are needed for subgroup estimation. As an illustration, we apply the proposed methods to data from a randomized clinical trial.
PMCID: PMC3880775  PMID: 21815180
randomized clinical trials; subgroups; random forests; regression trees; tailored therapeutics
9.  Evaluating Marker-Guided Treatment Selection Strategies 
Biometrics  2014;70(3):489-499.
A potential venue to improve healthcare efficiency is to effectively tailor individualized treatment strategies by incorporating patient level predictor information such as environmental exposure, biological, and genetic marker measurements. Many useful statistical methods for deriving individualized treatment rules (ITR) have become available in recent years. Prior to adopting any ITR in clinical practice, it is crucial to evaluate its value in improving patient outcomes. Existing methods for quantifying such values mainly consider either a single marker or semi-parametric methods that are subject to bias under model misspecification. In this paper, we consider a general setting with multiple markers and propose a two-step robust method to derive ITRs and evaluate their values. We also propose procedures for comparing different ITRs, which can be used to quantify the incremental value of new markers in improving treatment selection. While working models are used in step I to approximate optimal ITRs, we add a layer of calibration to guard against model misspecification and further assess the value of the ITR non-parametrically, which ensures the validity of the inference. To account for the sampling variability of the estimated rules and their corresponding values, we propose a resampling procedure to provide valid confidence intervals for the value functions as well as for the incremental value of new markers for treatment selection. Our proposals are examined through extensive simulation studies and illustrated with the data from a clinical trial that studies the effects of two drug combinations on HIV-1 infected patients.
PMCID: PMC4213325  PMID: 24779731
Biomarker-analysis Design; Counterfactual Outcome; Personalized Medicine; Perturbation-resampling; Predictive Biomarkers; Subgroup Analysis
10.  On the covariate-adjusted estimation for an overall treatment difference with data from a randomized comparative clinical trial 
Biostatistics (Oxford, England)  2012;13(2):256-273.
To estimate an overall treatment difference with data from a randomized comparative clinical study, baseline covariates are often utilized to increase the estimation precision. Using the standard analysis of covariance technique for making inferences about such an average treatment difference may not be appropriate, especially when the fitted model is nonlinear. On the other hand, the novel augmentation procedure recently studied, for example, by Zhang and others (2008. Improving efficiency of inferences in randomized clinical trials using auxiliary covariates. Biometrics 64, 707–715) is quite flexible. However, in general, it is not clear how to select covariates for augmentation effectively. An overly adjusted estimator may inflate the variance and in some cases be biased. Furthermore, the results from the standard inference procedure by ignoring the sampling variation from the variable selection process may not be valid. In this paper, we first propose an estimation procedure, which augments the simple treatment contrast estimator directly with covariates. The new proposal is asymptotically equivalent to the aforementioned augmentation method. To select covariates, we utilize the standard lasso procedure. Furthermore, to make valid inference from the resulting lasso-type estimator, a cross validation method is used. The validity of the new proposal is justified theoretically and empirically. We illustrate the procedure extensively with a well-known primary biliary cirrhosis clinical trial data set.
PMCID: PMC3297822  PMID: 22294672
ANCOVA; Cross validation; Efficiency augmentation; Mayo PBC data; Semi-parametric efficiency
Annals of statistics  2011;39(1):305-332.
The complexity of semiparametric models poses new challenges to statistical inference and model selection that frequently arise from real applications. In this work, we propose new estimation and variable selection procedures for the semiparametric varying-coefficient partially linear model. We first study quantile regression estimates for the nonparametric varying-coefficient functions and the parametric regression coefficients. To achieve nice efficiency properties, we further develop a semiparametric composite quantile regression procedure. We establish the asymptotic normality of proposed estimators for both the parametric and nonparametric parts and show that the estimators achieve the best convergence rate. Moreover, we show that the proposed method is much more efficient than the least-squares-based method for many non-normal errors and that it only loses a small amount of efficiency for normal errors. In addition, it is shown that the loss in efficiency is at most 11.1% for estimating varying coefficient functions and is no greater than 13.6% for estimating parametric components. To achieve sparsity with high-dimensional covariates, we propose adaptive penalization methods for variable selection in the semiparametric varying-coefficient partially linear model and prove that the methods possess the oracle property. Extensive Monte Carlo simulation studies are conducted to examine the finite-sample performance of the proposed procedures. Finally, we apply the new methods to analyze the plasma beta-carotene level data.
PMCID: PMC3109949  PMID: 21666869
Asymptotic relative efficiency; composite quantile regression; semiparametric varying-coefficient partially linear model; oracle properties; variable selection
12.  Collaborative Double Robust Targeted Maximum Likelihood Estimation* 
Collaborative double robust targeted maximum likelihood estimators represent a fundamental further advance over standard targeted maximum likelihood estimators of a pathwise differentiable parameter of a data generating distribution in a semiparametric model, introduced in van der Laan, Rubin (2006). The targeted maximum likelihood approach involves fluctuating an initial estimate of a relevant factor (Q) of the density of the observed data, in order to make a bias/variance tradeoff targeted towards the parameter of interest. The fluctuation involves estimation of a nuisance parameter portion of the likelihood, g. TMLE has been shown to be consistent and asymptotically normally distributed (CAN) under regularity conditions, when either one of these two factors of the likelihood of the data is correctly specified, and it is semiparametric efficient if both are correctly specified.
In this article we provide a template for applying collaborative targeted maximum likelihood estimation (C-TMLE) to the estimation of pathwise differentiable parameters in semi-parametric models. The procedure creates a sequence of candidate targeted maximum likelihood estimators based on an initial estimate for Q coupled with a succession of increasingly non-parametric estimates for g. In a departure from current state of the art nuisance parameter estimation, C-TMLE estimates of g are constructed based on a loss function for the targeted maximum likelihood estimator of the relevant factor Q that uses the nuisance parameter to carry out the fluctuation, instead of a loss function for the nuisance parameter itself. Likelihood-based cross-validation is used to select the best estimator among all candidate TMLE estimators of Q0 in this sequence. A penalized-likelihood loss function for Q is suggested when the parameter of interest is borderline-identifiable.
We present theoretical results for “collaborative double robustness,” demonstrating that the collaborative targeted maximum likelihood estimator is CAN even when Q and g are both mis-specified, providing that g solves a specified score equation implied by the difference between the Q and the true Q0. This marks an improvement over the current definition of double robustness in the estimating equation literature.
We also establish an asymptotic linearity theorem for the C-DR-TMLE of the target parameter, showing that the C-DR-TMLE is more adaptive to the truth, and, as a consequence, can even be super efficient if the first stage density estimator does an excellent job itself with respect to the target parameter.
This research provides a template for targeted efficient and robust loss-based learning of a particular target feature of the probability distribution of the data within large (infinite dimensional) semi-parametric models, while still providing statistical inference in terms of confidence intervals and p-values. This research also breaks with a taboo (e.g., in the propensity score literature in the field of causal inference) on using the relevant part of likelihood to fine-tune the fitting of the nuisance parameter/censoring mechanism/treatment mechanism.
PMCID: PMC2898626  PMID: 20628637
asymptotic linearity; coarsening at random; causal effect; censored data; crossvalidation; collaborative double robust; double robust; efficient influence curve; estimating function; estimator selection; influence curve; G-computation; locally efficient; loss-function; marginal structural model; maximum likelihood estimation; model selection; pathwise derivative; semiparametric model; sieve; super efficiency; super-learning; targeted maximum likelihood estimation; targeted nuisance parameter estimator selection; variable importance
13.  On Sparse Estimation for Semiparametric Linear Transformation Models 
Journal of multivariate analysis  2010;101(7):1594-1606.
Semiparametric linear transformation models have received much attention due to its high flexibility in modeling survival data. A useful estimating equation procedure was recently proposed by Chen et al. (2002) for linear transformation models to jointly estimate parametric and nonparametric terms. They showed that this procedure can yield a consistent and robust estimator. However, the problem of variable selection for linear transformation models is less studied, partially because a convenient loss function is not readily available under this context. In this paper, we propose a simple yet powerful approach to achieve both sparse and consistent estimation for linear transformation models. The main idea is to derive a profiled score from the estimating equation of Chen et al. (2002), construct a loss function based on the profile scored and its variance, and then minimize the loss subject to some shrinkage penalty. Under regularity conditions, we have shown that the resulting estimator is consistent for both model estimation and variable selection. Furthermore, the estimated parametric terms are asymptotically normal and can achieve higher efficiency than that yielded from the estimation equations. For computation, we suggest a one-step approximation algorithm which can take advantage of the LARS and build the entire solution path efficiently. Performance of the new procedure is illustrated through numerous simulations and real examples including one microarray data.
PMCID: PMC2869045  PMID: 20473356
Censored survival data; Linear transformation models; LARS; Shrinkage; Variable selection
14.  Landmark Estimation of Survival and Treatment Effect in a Randomized Clinical Trial 
In many studies with a survival outcome, it is often not feasible to fully observe the primary event of interest. This often leads to heavy censoring and thus, difficulty in efficiently estimating survival or comparing survival rates between two groups. In certain diseases, baseline covariates and the event time of non-fatal intermediate events may be associated with overall survival. In these settings, incorporating such additional information may lead to gains in efficiency in estimation of survival and testing for a difference in survival between two treatment groups. If gains in efficiency can be achieved, it may then be possible to decrease the sample size of patients required for a study to achieve a particular power level or decrease the duration of the study. Most existing methods for incorporating intermediate events and covariates to predict survival focus on estimation of relative risk parameters and/or the joint distribution of events under semiparametric models. However, in practice, these model assumptions may not hold and hence may lead to biased estimates of the marginal survival. In this paper, we propose a semi-nonparametric two-stage procedure to estimate and compare t-year survival rates by incorporating intermediate event information observed before some landmark time, which serves as a useful approach to overcome semi-competing risks issues. In a randomized clinical trial setting, we further improve efficiency through an additional calibration step. Simulation studies demonstrate substantial potential gains in efficiency in terms of estimation and power. We illustrate our proposed procedures using an AIDS Clinical Trial Protocol 175 dataset by estimating survival and examining the difference in survival between two treatment groups: zidovudine and zidovudine plus zalcitabine.
PMCID: PMC3960087  PMID: 24659838
Efficiency Augmentation; Kaplan Meier; Landmark Prediction; Semi-competing Risks; Survival Analysis
15.  Reporting Methods of Blinding in Randomized Trials Assessing Nonpharmacological Treatments  
PLoS Medicine  2007;4(2):e61.
Blinding is a cornerstone of treatment evaluation. Blinding is more difficult to obtain in trials assessing nonpharmacological treatment and frequently relies on “creative” (nonstandard) methods. The purpose of this study was to systematically describe the strategies used to obtain blinding in a sample of randomized controlled trials of nonpharmacological treatment.
Methods and Findings
We systematically searched in Medline and the Cochrane Methodology Register for randomized controlled trials (RCTs) assessing nonpharmacological treatment with blinding, published during 2004 in high-impact-factor journals. Data were extracted using a standardized extraction form. We identified 145 articles, with the method of blinding described in 123 of the reports. Methods of blinding of participants and/or health care providers and/or other caregivers concerned mainly use of sham procedures such as simulation of surgical procedures, similar attention-control interventions, or a placebo with a different mode of administration for rehabilitation or psychotherapy. Trials assessing devices reported various placebo interventions such as use of sham prosthesis, identical apparatus (e.g., identical but inactivated machine or use of activated machine with a barrier to block the treatment), or simulation of using a device. Blinding participants to the study hypothesis was also an important method of blinding. The methods reported for blinding outcome assessors relied mainly on centralized assessment of paraclinical examinations, clinical examinations (i.e., use of video, audiotape, photography), or adjudications of clinical events.
This study classifies blinding methods and provides a detailed description of methods that could overcome some barriers of blinding in clinical trials assessing nonpharmacological treatment, and provides information for readers assessing the quality of results of such trials.
An assessment of blinding methods used in nonpharmacological trials published in one year in high-impact factor journals classifies methods used and describes methods that could overcome some barriers of blinding.
Editors' Summary
Well-conducted “randomized controlled trials” are generally considered to be the most reliable source of information about the effects of medical treatments. In a randomized trial, the play of chance is used to decide whether each patient receives the treatment under investigation, or whether he/she is assigned to a “control” group receiving the standard treatment for their condition. This helps makes sure that the two groups of patients receiving the different treatments are equivalent at the start of the trial. Proper randomization also prevents doctors from deciding which treatment individual patients are given, which could distort the results. An additional technique used is “blinding,” which involves taking steps to prevent patients, doctors, or other people involved in the trial (e.g., those recording measurements) from finding out which patients have received which treatment. Properly done, blinding should make sure the results of a trial are more accurate. This is because in an unblinded study, participants may respond better if they know they have received a promising new treatment (or worse if they only got a placebo or an old drug). In addition, doctors and others in the research team may “want” a particular treatment to perform better in the trial, and unthinking bias could creep into their measurements or actions. However, blinding is not a simple, single step; the people carrying out the trial often have to set up a variety of different procedures.
Why Was This Study Done?
The authors of this study had already conducted research into the way in which blinding is done in trials involving drug (“pharmacological”) treatment. Their work was published in October 2006 in PLoS Medicine. However, concealing from patients the type of pill that they are being given is much easier than, for example, concealing whether or not they are having surgery or whether or not they are having psychotherapy. The authors therefore set out to look at the methods that are in use for blinding in nonpharmacological trials. They hoped that a better understanding of the different blinding methods would help people doing trials to design better trials in the future, and also help readers to interpret the quality of completed trials.
What Did the Researchers Do and Find?
The authors systematically searched the published medical literature to find all randomized, blinded drug trials published in just one year (2004) in a number of different “high-impact” journals (well-regarded journals whose articles are often mentioned in other articles). Then, they classified information from the published trial reports. They ended up with 145 trial reports, of which 123 described how blinding was done. The trials covered a wide range of medical conditions and types of treatment. The blinding methods used mainly involved the use of “sham” procedures. Thus, in 80% of the studies in which the treatment involved a medical device, a pretend device had been used to make patients in the control group think they were receiving treatment. In many of the treatments involving surgery, researchers had devised elaborate ways of making patients think they had had an operation. When the treatment involved manipulation (e.g. physiotherapy or chiropractic), fake “hands-on” techniques were given to the control patients. The authors of this systematic review classify all the other techniques that were used to blind both the patients and members of the research teams. They found that some highly innovative ideas have been successfully put into practice.
What Do These Findings Mean?
The authors have provided a detailed description of methods that could overcome some barriers of blinding in clinical trials assessing nonpharmacological treatment. The classification of the techniques used will be useful for other researchers considering what sort of blinding they will use in their own research.
Additional Information.
Please access these Web sites via the online version of this summary at
The James Lind Library has been created to help patients and researchers understand fair tests of treatments in health care by illustrating how fair tests have developed over the centuries, a trial registry created by the US National Institutes of Health, has an introduction to understanding clinical trials
The UK National Health Service National Electronic Library for Health has an introduction to controlled clinical trials
The CONSORT statement is intended to strengthen evidence-based reporting of clinical trials
PMCID: PMC1800311  PMID: 17311468
16.  A method to estimate treatment efficacy among latent subgroups of a randomized clinical trial 
Statistics in medicine  2010;30(7):709-717.
Subgroup analysis arises in clinical trials research when we wish to estimate a treatment effect on a specific subgroup of the population distinguished by baseline characteristics. Many trial designs induce latent subgroups such that subgroup membership is observable in one arm of the trial and unidentified in the other. This occurs, for example, in oncology trials when a biopsy or dissection is performed only on subjects randomized to active treatment. We discuss a general framework to estimate a biological treatment effect on the latent subgroup of interest when the survival outcome is right-censored and can be appropriately modelled as a parametric function of covariate effects. Our framework builds on the application of instrumental variables methods to all-or-none treatment noncompliance. We derive a computational method to estimate model parameters via the EM algorithm and provide guidance on its implementation in standard software packages. The research is illustrated through an analysis of a seminal melanoma trial that proposed a new standard of care for the disease and involved a biopsy that is available only on patients in the treatment arm.
PMCID: PMC3161831  PMID: 21394747
survival analysis; accelerated failure time model; treatment noncompliance; mixture model; EM algorithm
17.  Efficient design and inference for multistage randomized trials of individualized treatment policies 
Biostatistics (Oxford, England)  2011;13(1):142-152.
Clinical demand for individualized “adaptive” treatment policies in diverse fields has spawned development of clinical trial methodology for their experimental evaluation via multistage designs, building upon methods intended for the analysis of naturalistically observed strategies. Because often there is no need to parametrically smooth multistage trial data (in contrast to observational data for adaptive strategies), it is possible to establish direct connections among different methodological approaches. We show by algebraic proof that the maximum likelihood (ML) and optimal semiparametric (SP) estimators of the population mean of the outcome of a treatment policy and its standard error are equal under certain experimental conditions. This result is used to develop a unified and efficient approach to design and inference for multistage trials of policies that adapt treatment according to discrete responses. We derive a sample size formula expressed in terms of a parametric version of the optimal SP population variance. Nonparametric (sample-based) ML estimation performed well in simulation studies, in terms of achieved power, for scenarios most likely to occur in real studies, even though sample sizes were based on the parametric formula. ML outperformed the SP estimator; differences in achieved power predominately reflected differences in their estimates of the population mean (rather than estimated standard errors). Neither methodology could mitigate the potential for overestimated sample sizes when strong nonlinearity was purposely simulated for certain discrete outcomes; however, such departures from linearity may not be an issue for many clinical contexts that make evaluation of competitive treatment policies meaningful.
PMCID: PMC3276275  PMID: 21765180
Adaptive treatment strategy; Efficient SP estimation; Maximum likelihood; Multi-stage design; Sample size formula
18.  Utilizing the integrated difference of two survival functions to quantify the treatment contrast for designing, monitoring and analyzing a comparative clinical study 
Consider a comparative, randomized clinical study with a specific event time as the primary endpoint. In the presence of censoring, standard methods of summarizing the treatment difference are based on Kaplan-Meier curves, the logrank test and the point and interval estimates via Cox’s procedure. Moreover, for designing and monitoring the study, one usually utilizes an event-driven scheme to determine the sample sizes and interim analysis time points.
When the proportional hazards assumption is violated, the logrank test may not have sufficient power to detect the difference between two event time distributions. The resulting hazard ratio estimate is difficult, if not impossible, to interpret as a treatment contrast. When the event rates are low, the corresponding interval estimate for the “hazard ratio” can be quite large due to the fact that the interval length depends on the observed numbers of events. This may indicate that there is not enough information for making inferences about the treatment comparison even when there is no difference between two groups. This situation is quite common for a post marketing safety study. We need an alternative way to quantify the group difference.
Instead of quantifying the treatment group difference using the hazard ratio, we consider an easily interpretable and model-free parameter, the integrated survival rate difference over a pre-specified time interval, as an alternative. We present the inference procedures for such a treatment contrast. This approach is purely nonparametric and does not need any model assumption such as the proportional hazards. Moreover, when we deal with equivalence or non-inferiority studies and the event rates are low, our procedure would provide more information about the treatment difference. We used a cardiovascular trial data set to illustrate our approach.
The results using the integrated event rate differences have a heuristic interpretation for the treatment difference even when the proportional hazards assumption is not valid. When the event rates are low, for example, for the cardiovascular study discussed in the paper, the procedure for the integrated event rate difference provides tight interval estimates in contrast to those based on the event-driven inference method.
The design of a trial with the integrated event rate difference may be more complicated than that using the event-driven procedure. One may use simulation to determine the sample size and the estimated duration of the study.
The procedure discussed in the paper can be a useful alternative to the standard proportional hazards method in survival analysis.
PMCID: PMC3705645  PMID: 22914867
Equivalence study; Event-driven study; Kaplan-Meier curve; Non-inferiority trial; Post-market study; Proportional hazards estimate
19.  Comparing Biomarkers as Principal Surrogate Endpoints 
Biometrics  2011;67(4):1442-1451.
Recently a new definition of surrogate endpoint, the ‘principal surrogate’, was proposed based on causal associations between treatment effects on the biomarker and on the clinical endpoint. Despite its appealing interpretation, limited research has been conducted to evaluate principal surrogates, and existing methods focus on risk models that consider a single biomarker. How to compare principal surrogate value of biomarkers or general risk models that consider multiple biomarkers remains an open research question. We propose to characterize a marker or risk model’s principal surrogate value based on the distribution of risk difference between interventions. In addition, we propose a novel summary measure (the standardized total gain) that can be used to compare markers and to assess the incremental value of a new marker. We develop a semiparametric estimated-likelihood method to estimate the joint surrogate value of multiple biomarkers. This method accommodates two-phase sampling of biomarkers and ismore widely applicable than existing nonparametric methods by incorporating continuous baseline covariates to predict the biomarker(s), and is more robust than existing parametric methods by leaving the error distribution of markers unspecified. The methodology is illustrated using a simulated example set and a real data set in the context of HIV vaccine trials.
PMCID: PMC3163011  PMID: 21517791
Estimated likelihood; Predictiveness curve; Principal stratification; Semiparametric; Surrogate marker; Total gain
20.  KRAS Testing for Anti-EGFR Therapy in Advanced Colorectal Cancer 
Executive Summary
In February 2010, the Medical Advisory Secretariat (MAS) began work on evidence-based reviews of the literature surrounding three pharmacogenomic tests. This project came about when Cancer Care Ontario (CCO) asked MAS to provide evidence-based analyses on the effectiveness and cost-effectiveness of three oncology pharmacogenomic tests currently in use in Ontario.
Evidence-based analyses have been prepared for each of these technologies. These have been completed in conjunction with internal and external stakeholders, including a Provincial Expert Panel on Pharmacogenomics (PEPP). Within the PEPP, subgroup committees were developed for each disease area. For each technology, an economic analysis was also completed by the Toronto Health Economics and Technology Assessment Collaborative (THETA) and is summarized within the reports.
The following reports can be publicly accessed at the MAS website at: or at
Gene Expression Profiling for Guiding Adjuvant Chemotherapy Decisions in Women with Early Breast Cancer: An Evidence-Based and Economic Analysis
Epidermal Growth Factor Receptor Mutation (EGFR) Testing for Prediction of Response to EGFR-Targeting Tyrosine Kinase Inhibitor (TKI) Drugs in Patients with Advanced Non-Small-Cell Lung Cancer: an Evidence-Based and Economic Analysis
K-RAS testing in Treatment Decisions for Advanced Colorectal Cancer: an Evidence-Based and Economic Analysis.
The objective of this systematic review is to determine the predictive value of KRAS testing in the treatment of metastatic colorectal cancer (mCRC) with two anti-EGFR agents, cetuximab and panitumumab. Economic analyses are also being conducted to evaluate the cost-effectiveness of KRAS testing.
Clinical Need: Condition and Target Population
Metastatic colorectal cancer (mCRC) is usually defined as stage IV disease according to the American Joint Committee on Cancer tumour node metastasis (TNM) system or stage D in the Duke’s classification system. Patients with advanced colorectal cancer (mCRC) either present with metastatic disease or develop it through disease progression.
KRAS (Kristen-RAS, a member of the rat sarcoma virus (ras) gene family of oncogenes) is frequently mutated in epithelial cancers such as colorectal cancer, with mutations occurring in mutational hotspots (codons 12 and 13) of the KRAS protein. Involved in EGFR-mediated signalling of cellular processes such as cell proliferation, resistance to apoptosis, enhanced cell motility and neoangiogenesis, a mutation in the KRAS gene is believed to be involved in cancer pathogenesis. Such a mutation is also hypothesized to be involved in resistance to targeted anti-EGFR (epidermal growth factor receptor with tyrosine kinase activity) treatments such as cetuximab and panitumumab, hence, the important in evaluating the evidence on the predictive value of KRAS testing in this context.
KRAS Mutation Testing in Advanced Colorectal Cancer
Both cetuximab and panitumumab are indicated by Health Canada in the treatment of patients with metastatic colorectal cancer whose tumours are WT for the KRAS gene. Cetuximab may be offered as monotherapy in patients intolerant to irinotecan-based chemotherapy or in patients who have failed both irinotecan and oxaliplatin-based regimens and who received a fluoropyrimidine. It can also be administered in combination with irinotecan in patients refractory to other irinotecan-based chemotherapy regimens. Panitumumab is only indicated as a single agent after failure of fluoropyrimidine-, oxaliplatin-, and irinotecan-containing chemotherapy regimens.
In Ontario, patients with advanced colorectal cancer who are refractory to chemotherapy may be offered the targeted anti-EGFR treatments cetuximab or panitumumab. Eligibility for these treatments is based on the KRAS status of their tumour, derived from tissue collected from surgical or biopsy specimens. It is believed that KRAS status is not affected by treatments, therefore, for patients for whom surgical tissue is available for KRAS testing, additional biopsies prior to treatment with these targeted agents is not necessary. For patients that have not undergone surgery or for whom surgical tissue is not available, a biopsy of either the primary or metastatic site is required to determine their KRAS status. This is possible as status at the metastatic and primary tumour sites is considered to be similar.
Research Question
To determine if there is predictive value of KRAS testing in guiding treatment decisions with anti-EGFR targeted therapies in advanced colorectal cancer patients refractory to chemotherapy.
Research Methods
Literature Search
The Medical Advisory Secretariat followed its standard procedures and on May 18, 2010, searched the following electronic databases: Ovid MEDLINE, EMBASE, Ovid MEDLINE In-Process & Other Non-Indexed Citations, Cochrane Central Register of Controlled Trials, Cochrane Database of Systematic Reviews and The International Network of Agencies for Health Technology Assessment database.
The subject headings and keywords searched included colorectal cancer, cetuximab, panitumumab, and KRAS testing. The search was further restricted to English-language articles published between January 1, 2009 and May 18, 2010 resulting in 1335 articles for review. Excluded were case reports, comments, editorials, nonsystematic reviews, and letters. Studies published from January 1, 2005 to December 31, 2008 were identified in a health technology assessment conducted by the Agency for Healthcare Research and Quality (AHRQ), published in 2010. In total, 14 observational studies were identified for inclusion in this EBA: 4 for cetuximab monotherapy, 7 for the cetuximab-irinotecan combination therapy, and 3 to be included in the review for panitumumab monotherapy
Inclusion Criteria
English-language articles, and English or French-language HTAs published from January 2005 to May 2010, inclusive.
Randomized controlled trials (RCTs) or observational studies, including single arm treatment studies that include KRAS testing.
Studies with data on main outcomes of interest, overall and progression-free survival.
Studies of third line treatment with cetuximab or panitumumab in patients with advanced colorectal cancer refractory to chemotherapy.
For the cetuximab-irinotecan evaluation, studies in which at least 70% of patients in the study received this combination therapy.
Exclusion Criteria
Studies whose entire sample was included in subsequent publications which have been included in this EBA.
Studies in pediatric populations.
Case reports, comments, editorials, or letters.
Outcomes of Interest
Overall survival (OS), median
Progression-free-survival (PFS), median.
Response rates.
Adverse event rates.
Quality of life (QOL).
Summary of Findings of Systematic Review
Cetuximab or Panitumumab Monotherapy
Based on moderate GRADE observational evidence, there is improvement in PFS and OS favouring patients without the KRAS mutation (KRAS wildtype, or KRAS WT) compared to those with the mutation.
Cetuximab-Irinotecan Combination Therapy
There is low GRADE evidence that testing for KRAS may optimize survival benefits in patients without the KRAS mutation (KRAS wildtype, or KRAS WT) compared to those with the mutation.
However, cetuximab-irinotecan combination treatments based on KRAS status discount any effect of cetuximab in possibly reversing resistance to irinotecan in patients with the mutation, as observed effects were lower than for patients without the mutation. Clinical experts have raised concerns about the biological plausibility of this observation and this conclusion would, therefore, be regarded as hypothesis generating.
Economic Analysis
Cost-effectiveness and budget impact analyses were conducted incorporating estimates of effectiveness from this systematic review. Evaluation of relative cost-effectiveness, based on a decision-analytic cost-utility analysis, assessed testing for KRAS genetic mutations versus no testing in the context of treatment with cetuximab monotherapy, panitumumab monotherapy, cetuximab in combination with irinotecan, and best supportive care.
Of importance to note is that the cost-effectiveness analysis focused on the impact of testing for KRAS mutations compared to no testing in the context of different treatment options, and does not assess the cost-effectiveness of the drug treatments alone.
KRAS status is predictive of outcomes in cetuximab and panitumumab monotherapy, and in cetuximab-irinotecan combination therapy.
While KRAS testing is cost-effective for all strategies considered, it is not equally cost-effective for all treatment options.
PMCID: PMC3377508  PMID: 23074403
21.  Mixtures of Varying Coefficient Models for Longitudinal Data with Discrete or Continuous Nonignorable Dropout 
Biometrics  2004;60(4):854-864.
The analysis of longitudinal repeated measures data is frequently complicated by missing data due to informative dropout. We describe a mixture model for joint distribution for longitudinal repeated measures, where the dropout distribution may be continuous and the dependence between response and dropout is semiparametric. Specifically, we assume that responses follow a varying coefficient random effects model conditional on dropout time, where the regression coefficients depend on dropout time through unspecified nonparametric functions that are estimated using step functions when dropout time is discrete (e.g., for panel data) and using smoothing splines when dropout time is continuous. Inference under the proposed semiparametric model is hence more robust than the parametric conditional linear model. The unconditional distribution of the repeated measures is a mixture over the dropout distribution. We show that estimation in the semiparametric varying coefficient mixture model can proceed by fitting a parametric mixed effects model and can be carried out on standard software platforms such as SAS. The model is used to analyze data from a recent AIDS clinical trial and its performance is evaluated using simulations.
PMCID: PMC2677904  PMID: 15606405
Clinical trials; Equivalence trial; Linear mixed model; Missing data; Nonignorable dropout; Pattern-mixture model; Pediatric AIDS; Selection bias; Smoothing splines
22.  Bayesian Inference in Semiparametric Mixed Models for Longitudinal Data 
Biometrics  2009;66(1):70-78.
We consider Bayesian inference in semiparametric mixed models (SPMMs) for longitudinal data. SPMMs are a class of models that use a nonparametric function to model a time effect, a parametric function to model other covariate effects, and parametric or nonparametric random effects to account for the within-subject correlation. We model the nonparametric function using a Bayesian formulation of a cubic smoothing spline, and the random effect distribution using a normal distribution and alternatively a nonparametric Dirichlet process (DP) prior. When the random effect distribution is assumed to be normal, we propose a uniform shrinkage prior (USP) for the variance components and the smoothing parameter. When the random effect distribution is modeled nonparametrically, we use a DP prior with a normal base measure and propose a USP for the hyperparameters of the DP base measure. We argue that the commonly assumed DP prior implies a nonzero mean of the random effect distribution, even when a base measure with mean zero is specified. This implies weak identifiability for the fixed effects, and can therefore lead to biased estimators and poor inference for the regression coefficients and the spline estimator of the nonparametric function. We propose an adjustment using a postprocessing technique. We show that under mild conditions the posterior is proper under the proposed USP, a flat prior for the fixed effect parameters, and an improper prior for the residual variance. We illustrate the proposed approach using a longitudinal hormone dataset, and carry out extensive simulation studies to compare its finite sample performance with existing methods.
PMCID: PMC3081790  PMID: 19432777
Dirichlet process prior; Identifiability; Postprocessing; Random effects; Smoothing spline; Uniform shrinkage prior; Variance components
23.  Limbal Stem Cell Transplantation 
Executive Summary
The objective of this analysis is to systematically review limbal stem cell transplantation (LSCT) for the treatment of patients with limbal stem cell deficiency (LSCD). This evidence-based analysis reviews LSCT as a primary treatment for nonpterygium LSCD conditions, and LSCT as an adjuvant therapy to excision for the treatment of pterygium.
Clinical Need: Condition and Target Population
The outer surface of the eye is covered by 2 distinct cell layers: the corneal epithelial layer that overlies the cornea, and the conjunctival epithelial layer that overlies the sclera. These cell types are separated by a transitional zone known as the limbus. The corneal epithelial cells are renewed every 3 to 10 days by a population of stem cells located in the limbus.
Nonpterygium Limbal Stem Cell Deficiency
When the limbal stem cells are depleted or destroyed, LSCD develops. In LSCD, the conjunctival epithelium migrates onto the cornea (a process called conjunctivalization), resulting in a thickened, irregular, unstable corneal surface that is prone to defects, ulceration, corneal scarring, vascularization, and opacity. Patients experience symptoms including severe irritation, discomfort, photophobia, tearing, blepharospasm, chronic inflammation and redness, and severely decreased vision.
Depending on the degree of limbal stem cell loss, LSCD may be total (diffuse) or partial (local). In total LSCD, the limbal stem cell population is completed destroyed and conjunctival epithelium covers the entire cornea. In partial LSCD, some areas of the limbus are unharmed, and the corresponding areas on the cornea maintain phenotypically normal corneal epithelium.
Confirmation of the presence of conjunctivalization is necessary for LSCD diagnosis as the other characteristics and symptoms are nonspecific and indicate a variety of diseases. The definitive test for LSCD is impression cytology, which detects the presence of conjunctival epithelium and its goblet cells on the cornea. However, in the opinion of a corneal expert, diagnosis is often based on clinical assessment, and in the expert’s opinion, it is unclear whether impression cytology is more accurate and reliable than clinical assessment, especially for patients with severe LSCD.
The incidence of LSCD is not well understood. A variety of underlying disorders are associated with LSCD including chemical or thermal injuries, ultraviolet and ionizing radiation, Stevens-Johnson syndrome, multiple surgeries or cryotherapies, contact lens wear, extensive microbial infection, advanced ocular cicatricial pemphigoid, and aniridia. In addition, some LSCD cases are idiopathic. These conditions are uncommon (e.g., the prevalence of aniridia ranges from 1 in 40,000 to 1 in 100,000 people).
Pterygium is a wing-shaped fibrovascular tissue growth from the conjunctiva onto the cornea. Pterygium is the result of partial LSCD caused by localized ultraviolet damage to limbal stem cells. As the pterygium invades the cornea, it may cause irregular astigmatism, loss of visual acuity, chronic irritation, recurrent inflammation, double vision, and impaired ocular motility.
Pterygium occurs worldwide. Incidence and prevalence rates are highest in the “pterygium belt,” which ranges from 30 degrees north to 30 degrees south of the equator, and lower prevalence rates are found at latitudes greater than 40 degrees. The prevalence of pterygium for Caucasians residing in urban, temperate climates is estimated at 1.2%.
Existing Treatments Other Than Technology Being Reviewed
Nonpterygium Limbal Stem Cell Deficiency
In total LSCD, a patient’s limbal stem cells are completely depleted, so any successful treatment must include new stem cells. Autologous oral mucosal epithelium transplantation has been proposed as an alternative to LSCT. However, this procedure is investigational, and there is very limited level 4c evidence1 to support this technique (fewer than 20 eyes examined in 4 case series and 1 case report).
For patients with partial LSCD, treatment may not be necessary if their visual axis is not affected. However, if the visual axis is conjunctivalized, several disease management options exist including repeated mechanical debridement of the abnormal epithelium; intensive, nonpreserved lubrication; bandage contact lenses; autologous serum eye drops; other investigational medical treatments; and transplantation of an amniotic membrane inlay. However, these are all disease management treatments; LSCT is the only curative option.
The primary treatment for pterygium is surgical excision. However, recurrence is a common problem after excision using the bare sclera technique: reported recurrence rates range from 24% to 89%. Thus, a variety of adjuvant therapies have been used to reduce the risk of pterygium recurrence including LSCT, amniotic membrane transplantation (AMT), conjunctival autologous (CAU) transplantation, and mitomycin C (MMC, an antimetabolite drug).
New Technology Being Reviewed
To successfully treat LSCD, the limbal stem cell population must be repopulated. To achieve this, 4 LSCT procedures have been developed: conjunctival-limbal autologous (CLAU) transplantation; living-related conjunctival-limbal allogeneic (lr-CLAL) transplantation; keratolimbal allogeneic (KLAL) transplantation; and ex vivo expansion of limbal stem cells transplantation. Since the ex vivo expansion of limbal stem cells transplantation procedure is considered experimental, it has been excluded from the systematic review. These procedures vary by the source of donor cells and the amount of limbal tissue used. For CLAU transplants, limbal stem cells are obtained from the patient’s healthy eye. For lr-CLAL and KLAL transplants, stem cells are obtained from living-related and cadaveric donor eyes, respectively.
In CLAU and lr-CLAL transplants, 2 to 4 limbal grafts are removed from the superior and inferior limbus of the donor eye. In KLAL transplants, the entire limbus from the donor eye is used.
The recipient eye is prepared by removing the abnormal conjunctival and scar tissue. An incision is made into the conjunctival tissue into which the graft is placed, and the graft is then secured to the neighbouring limbal and scleral tissue with sutures. Some LSCT protocols include concurrent transplantation of an amniotic membrane onto the cornea.
Regulatory Status
Health Canada does not require premarket licensure for stem cells. However, they are subject to Health Canada’s clinical trial regulations until the procedure is considered accepted transplantation practice, at which time it will be covered by the Safety of Human Cells, Tissues and Organs for Transplantation Regulations (CTO Regulations).
Review Strategy
The Medical Advisory Secretariat systematically reviewed the literature to assess the effectiveness and safety of LSCT for the treatment of patients with nonpterygium LSCD and pterygium. A comprehensive search method was used to retrieve English-language journal articles from selected databases.
The GRADE approach was used to systematically and explicitly evaluate the quality of evidence and strength of recommendations.
Summary of Findings
Nonpterygium Limbal Stem Cell Deficiency
The search identified 873 citations published between January 1, 2000, and March 31, 2008. Nine studies met the inclusion criteria, and 1 additional citation was identified through a bibliography review. The review included 10 case series (3 prospective and 7 retrospective).
Patients who received autologous transplants (i.e., CLAU) achieved significantly better long-term corneal surface results compared with patients who received allogeneic transplants (lr-CLAL, P< .001; KLAL, P< .001). There was no significant difference in corneal surface outcomes between the allogeneic transplant options, lr-CLAL and KLAL (P = .328). However, human leukocyte antigen matching and systemic immunosuppression may improve the outcome of lr-CLAL compared with KLAL. Regardless of graft type, patients with Stevens-Johnson syndrome had poorer long-term corneal surface outcomes.
Concurrent AMT was associated with poorer long-term corneal surface improvements. When the effect of the AMT was removed, the difference between autologous and allogeneic transplants was much smaller.
Patients who received CLAU transplants had a significantly higher rate of visual acuity improvements compared with those who received lr-CLAL transplants (P = .002). However, to achieve adequate improvements in vision, patients with deep corneal scarring will require a corneal transplant several months after the LSCT.
No donor eye complications were observed.
Epithelial rejection and microbial keratitis were the most common long-term complications associated with LSCT (complications occurred in 6%–15% of transplantations). These complications can result in graft failure, so patients should be monitored regularly following LSCT.
The search yielded 152 citations published between January 1, 2000 and May 16, 2008. Six randomized controlled trials (RCTs) that evaluated LSCT as an adjuvant therapy for the treatment of pterygium met the inclusion criteria and were included in the review.
Limbal stem cell transplantation was compared with CAU, AMT, and MMC. The results showed that CLAU significantly reduced the risk of pterygium recurrence compared with CAU (relative risk [RR], 0.09; 95% confidence interval [CI], 0.01–0.69; P = .02). CLAU reduced the risk of pterygium recurrence for primary pterygium compared with MMC, but this comparison did not reach statistical significance (RR, 0.48; 95% CI, 0.21–1.10; P = .08). Both AMT and CLAU had similar low rates of recurrence (2 recurrences in 43 patients and 4 in 46, respectively), and the RR was not significant (RR, 1.88; 95% CI, 0.37–9.5; P = .45). Since sample sizes in the included studies were small, failure to detect a significant difference between LSCT and AMT or MMC could be the result of type II error. Limbal stem cell transplantation as an adjuvant to excision is a relatively safe procedure as long-term complications were rare (< 2%).
GRADE Quality of Evidence
Nonpterygium Limbal Stem Cell Deficiency
The evidence for the analyses related to nonpterygium LSCD was based on 3 prospective and 7 retrospective case series. Thus, the GRADE quality of evidence is very low, and any estimate of effect is very uncertain.
The analyses examining LSCT as an adjuvant treatment option for pterygium were based on 6 RCTs. The quality of evidence for the overall body of evidence for each treatment option comparison was assessed using the GRADE approach. In each of the comparisons, the quality of evidence was downgraded due to serious or very serious limitations in study quality (individual study quality was assessed using the Jadad scale, and an assessment of allocation concealment and the degree of loss to follow-up), which resulted in low- to moderate-quality GRADE evidence ratings (low-quality evidence for the CLAU and AMT and CLAU and MMC comparisons, and moderate-quality evidence for the CLAU and CAU comparison).
Ontario Health System Impact Analysis
Nonpterygium Limbal Stem Cell Deficiency
Since 1999, Ontario’s out-of-country (OOC) program has approved and reimbursed 8 patients for LSCTs and 1 patient for LSCT consultations. Similarly, most Canadian provinces have covered OOC or out-of-province LSCTs. Several corneal experts in Ontario have the expertise to perform LSCTs.
As there are no standard guidelines for LSCT, patients who receive transplants OOC may not receive care aligned with the best evidence. To date, many of the patients from Ontario who received OOC LSCTs received concurrent AMTs, and the evidence from this analysis questions the use of this procedure. In addition, 1 patient received a cultured LSCT, a procedure that is considered investigational. Many patients with LSCD have bilateral disease and therefore require allogeneic transplants. These patients will require systemic and topical immunosuppression for several years after the transplant, perhaps indefinitely. Thus, systemic side effects associated with immunosuppression are a potential concern, and patients must be monitored regularly.
Amniotic membrane transplantation is a common addition to many ocular surface reconstruction procedures, including LSCT. Amniotic membranes are recovered from human placentas from planned, uneventful caesarean sections. Before use, serological screening of the donor’s blood should be conducted. However, there is still a theoretical risk of disease transmission associated with this procedure.
Financial Impact
For the patients who were reimbursed for OOC LSCTs, the average cost of LSCT per eye was $18,735.20 Cdn (range, $8,219.54–$33,933.32). However, the actual cost per patient is much higher as these costs do not include consultations and follow-up visits, multiple LSCTs, and any additional procedures (e.g., corneal transplants) received during the course of treatment OOC. When these additional costs were considered, the average cost per patient was $57,583 Cdn (range, $8,219.54–$130,628.20).
The estimated average total cost per patient for performing LSCT in Ontario is $2,291.48 Cdn (range, $951.48–$4,538.48) including hospital and physician fees. This cost is based on the assumption that LSCT is technically similar to a corneal transplant, an assumption which needs to be verified. The cost does not include corneal transplantations, which some proportion of patients receiving a LSCT will require within several months of the limbal transplant.
Pterygium recurrence rates after surgical excision are high, ranging from 24% to 89%. However, according to clinical experts, the rate of recurrence is low in Ontario. While there is evidence that the prevalence of pterygium is higher in the “pterygium belt,” there was no evidence to suggest different recurrence rates or disease severity by location or climate.
Nonpterygium Limbal Stem Cell Deficiency
Successful LSCTs result in corneal re-epithelialization and improved vision in patients with LSCD. However, patients who received concurrent AMT had poorer long-term corneal surface improvements. Conjunctival-limbal autologous transplantation is the treatment option of choice, but if it is not possible, living-related or cadaveric allogeneic transplants can be used. The benefits of LSCT outweigh the risks and burdens, as shown in Executive Summary Table 1. According to GRADE, these recommendations are strong with low- to very low-quality evidence.
Benefits, Risks, and Burdens – Nonpterygium Limbal Stem Cell Deficiency
Short- and long-term improvement in corneal surface (stable, normal corneal epithelium and decreased vascularization and opacity)
Improvement in vision (visual acuity and functional vision)
Long-term complications are experienced by 8% to 16% of patients
Risks associated with long-term immunosuppression for recipients of allogeneic grafts
Potential risk of induced LSCD in donor eyes
High cost of treatment (average cost per patient via OOC program is $57,583; estimated cost of procedure in Ontario is $2,291.48)
Costs are expressed in Canadian dollars.
GRADE of recommendation: Strong recommendation, low-quality or very low-quality evidence
benefits clearly outweigh risks and burdens
case series studies
strong, but may change if higher-quality evidence becomes available
Conjunctival-limbal autologous transplantations significantly reduced the risk of pterygium recurrence compared with CAU. No other comparison yielded statistically significant results, but CLAU reduced the risk of recurrence compared with MMC. However, the benefit of LSCT in Ontario is uncertain as the severity and recurrence of pterygium in Ontario is unknown. The complication rates suggest that CLAU is a safe treatment option to prevent the recurrence of pterygium. According to GRADE, given the balance of the benefits, risks, and burdens, the recommendations are very weak with moderate quality evidence, as shown in Executive Summary Table 2.
Benefits, Risks, and Burdens – Pterygium
Reduced recurrence; however, if recurrence is low in Ontario, this benefit might be minimal
Long-term complications rare
Increased cost
GRADE of recommendation: Very weak recommendations, moderate quality evidence.
uncertainty in the estimates of benefits, risks, and burden; benefits, risks, and burden may be closely balanced
very weak, other alternatives may be equally reasonable
PMCID: PMC3377549  PMID: 23074512
24.  Meta-analyses of Adverse Effects Data Derived from Randomised Controlled Trials as Compared to Observational Studies: Methodological Overview 
PLoS Medicine  2011;8(5):e1001026.
Su Golder and colleagues carry out an overview of meta-analyses to assess whether estimates of the risk of harm outcomes differ between randomized trials and observational studies. They find that, on average, there is no difference in the estimates of risk between overviews of observational studies and overviews of randomized trials.
There is considerable debate as to the relative merits of using randomised controlled trial (RCT) data as opposed to observational data in systematic reviews of adverse effects. This meta-analysis of meta-analyses aimed to assess the level of agreement or disagreement in the estimates of harm derived from meta-analysis of RCTs as compared to meta-analysis of observational studies.
Methods and Findings
Searches were carried out in ten databases in addition to reference checking, contacting experts, citation searches, and hand-searching key journals, conference proceedings, and Web sites. Studies were included where a pooled relative measure of an adverse effect (odds ratio or risk ratio) from RCTs could be directly compared, using the ratio of odds ratios, with the pooled estimate for the same adverse effect arising from observational studies. Nineteen studies, yielding 58 meta-analyses, were identified for inclusion. The pooled ratio of odds ratios of RCTs compared to observational studies was estimated to be 1.03 (95% confidence interval 0.93–1.15). There was less discrepancy with larger studies. The symmetric funnel plot suggests that there is no consistent difference between risk estimates from meta-analysis of RCT data and those from meta-analysis of observational studies. In almost all instances, the estimates of harm from meta-analyses of the different study designs had 95% confidence intervals that overlapped (54/58, 93%). In terms of statistical significance, in nearly two-thirds (37/58, 64%), the results agreed (both studies showing a significant increase or significant decrease or both showing no significant difference). In only one meta-analysis about one adverse effect was there opposing statistical significance.
Empirical evidence from this overview indicates that there is no difference on average in the risk estimate of adverse effects of an intervention derived from meta-analyses of RCTs and meta-analyses of observational studies. This suggests that systematic reviews of adverse effects should not be restricted to specific study types.
Please see later in the article for the Editors' Summary
Editors' Summary
Whenever patients consult a doctor, they expect the treatments they receive to be effective and to have minimal adverse effects (side effects). To ensure that this is the case, all treatments now undergo exhaustive clinical research—carefully designed investigations that test new treatments and therapies in people. Clinical investigations fall into two main groups—randomized controlled trials (RCTs) and observational, or non-randomized, studies. In RCTs, groups of patients with a specific disease or condition are randomly assigned to receive the new treatment or a control treatment, and the outcomes (for example, improvements in health and the occurrence of specific adverse effects) of the two groups of patients are compared. Because the patients are randomly chosen, differences in outcomes between the two groups are likely to be treatment-related. In observational studies, patients who are receiving a specific treatment are enrolled and outcomes in this group are compared to those in a similar group of untreated patients. Because the patient groups are not randomly chosen, differences in outcomes between cases and controls may be the result of a hidden shared characteristic among the cases rather than treatment-related (so-called confounding variables).
Why Was This Study Done?
Although data from individual trials and studies are valuable, much more information about a potential new treatment can be obtained by systematically reviewing all the evidence and then doing a meta-analysis (so-called evidence-based medicine). A systematic review uses predefined criteria to identify all the research on a treatment; meta-analysis is a statistical method for combining the results of several studies to yield “pooled estimates” of the treatment effect (the efficacy of a treatment) and the risk of harm. Treatment effect estimates can differ between RCTs and observational studies, but what about adverse effect estimates? Can different study designs provide a consistent picture of the risk of harm, or are the results from different study designs so disparate that it would be meaningless to combine them in a single review? In this methodological overview, which comprises a systematic review and meta-analyses, the researchers assess the level of agreement in the estimates of harm derived from meta-analysis of RCTs with estimates derived from meta-analysis of observational studies.
What Did the Researchers Do and Find?
The researchers searched literature databases and reference lists, consulted experts, and hand-searched various other sources for studies in which the pooled estimate of an adverse effect from RCTs could be directly compared to the pooled estimate for the same adverse effect from observational studies. They identified 19 studies that together covered 58 separate adverse effects. In almost all instances, the estimates of harm obtained from meta-analyses of RCTs and observational studies had overlapping 95% confidence intervals. That is, in statistical terms, the estimates of harm were similar. Moreover, in nearly two-thirds of cases, there was agreement between RCTs and observational studies about whether a treatment caused a significant increase in adverse effects, a significant decrease, or no significant change (a significant change is one unlikely to have occurred by chance). Finally, the researchers used meta-analysis to calculate that the pooled ratio of the odds ratios (a statistical measurement of risk) of RCTs compared to observational studies was 1.03. This figure suggests that there was no consistent difference between risk estimates obtained from meta-analysis of RCT data and those obtained from meta-analysis of observational study data.
What Do These Findings Mean?
The findings of this methodological overview suggest that there is no difference on average in the risk estimate of an intervention's adverse effects obtained from meta-analyses of RCTs and from meta-analyses of observational studies. Although limited by some aspects of its design, this overview has several important implications for the conduct of systematic reviews of adverse effects. In particular, it suggests that, rather than limiting systematic reviews to certain study designs, it might be better to evaluate a broad range of studies. In this way, it might be possible to build a more complete, more generalizable picture of potential harms associated with an intervention, without any loss of validity, than by evaluating a single type of study. Such a picture, in combination with estimates of treatment effects also obtained from systematic reviews and meta-analyses, would help clinicians decide the best treatment for their patients.
Additional Information
Please access these Web sites via the online version of this summary at
The US National Institutes of Health provide information on clinical research; the UK National Health Service Choices Web site also has a page on clinical trials and medical research
The Cochrane Collaboration produces and disseminates systematic reviews of health-care interventions
Medline Plus provides links to further information about clinical trials (in English and Spanish)
PMCID: PMC3086872  PMID: 21559325
25.  Variable selection for covariate-adjusted semiparametric inference in randomized clinical trials 
Statistics in medicine  2012;31(29):10.1002/sim.5433.
Extensive baseline covariate information is routinely collected on participants in randomized clinical trials, and it is well-recognized that a proper covariate-adjusted analysis can improve the efficiency of inference on the treatment effect. However, such covariate adjustment has engendered considerable controversy, as post hoc selection of covariates may involve subjectivity and lead to biased inference, while prior specification of the adjustment may exclude important variables from consideration. Accordingly, how to select covariates objectively to gain maximal efficiency is of broad interest. We propose and study the use of modern variable selection methods for this purpose in the context of a semiparametric framework, under which variable selection in modeling the relationship between outcome and covariates is separated from estimation of the treatment effect, circumventing the potential for selection bias associated with standard analysis of covariance methods. We demonstrate that such objective variable selection techniques combined with this framework can identify key variables and lead to unbiased and efficient inference on the treatment effect. A critical issue in finite samples is validity of estimators of uncertainty, such as standard errors and confidence intervals for the treatment effect. We propose an approach to estimation of sampling variation of estimated treatment effect and show its superior performance relative to that of existing methods.
PMCID: PMC3855673  PMID: 22733628
covariate adjustment; false selection rate control; oracle property; semiparametric treatment effect estimation; shrinkage methods; variable selection

Results 1-25 (1614216)