PMCC PMCC

Search tips
Search criteria

Advanced
Results 1-25 (1161840)

Clipboard (0)
None

Related Articles

1.  Using Audit Information to Adjust Parameter Estimates for Data Errors in Clinical Trials 
Background
Audits are often performed to assess the quality of clinical trial data, but beyond detecting fraud or sloppiness, the audit data is generally ignored. In earlier work using data from a non-randomized study, Shepherd and Yu (2011) developed statistical methods to incorporate audit results into study estimates, and demonstrated that audit data could be used to eliminate bias.
Purpose
In this manuscript we examine the usefulness of audit-based error-correction methods in clinical trial settings where a continuous outcome is of primary interest.
Methods
We demonstrate the bias of multiple linear regression estimates in general settings with an outcome that may have errors and a set of covariates for which some may have errors and others, including treatment assignment, are recorded correctly for all subjects. We study this bias under different assumptions including independence between treatment assignment, covariates, and data errors (conceivable in a double-blinded randomized trial) and independence between treatment assignment and covariates but not data errors (possible in an unblinded randomized trial). We review moment-based estimators to incorporate the audit data and propose new multiple imputation estimators. The performance of estimators is studied in simulations.
Results
When treatment is randomized and unrelated to data errors, estimates of the treatment effect using the original error-prone data (i.e., ignoring the audit results) are unbiased. In this setting, both moment and multiple imputation estimators incorporating audit data are more variable than standard analyses using the original data. In contrast, in settings where treatment is randomized but correlated with data errors and in settings where treatment is not randomized, standard treatment effect estimates will be biased. And in all settings, parameter estimates for the original, error-prone covariates will be biased. Treatment and covariate effect estimates can be corrected by incorporating audit data using either the multiple imputation or moment-based approaches. Bias, precision, and coverage of confidence intervals improve as the audit size increases.
Limitations
The extent of bias and the performance of methods depend on the extent and nature of the error as well as the size of the audit. This work only considers methods for the linear model. Settings much different than those considered here need further study.
Conclusions
In randomized trials with continuous outcomes and treatment assignment independent of data errors, standard analyses of treatment effects will be unbiased and are recommended. However, if treatment assignment is correlated with data errors or other covariates, naive analyses may be biased. In these settings, and when covariate effects are of interest, approaches for incorporating audit results should be considered.
doi:10.1177/1740774512450100
PMCID: PMC3728661  PMID: 22848072
audit; bias; clinical trials; measurement error; multiple imputation
2.  Sensitivity Analyses Comparing Time-to-Event Outcomes Existing Only in a Subset Selected Postrandomization 
In some randomized studies, researchers are interested in determining the effect of treatment assignment on outcomes that may exist only in a subset chosen after randomization. For example, in preventative human immunodeficiency virus (HIV) vaccine efficacy trials, it is of interest to determine whether randomization to vaccine affects postinfection outcomes that may be right-censored. Such outcomes in these trials include time from infection diagnosis to initiation of antiretroviral therapy and time from infection diagnosis to acquired immune deficiency syndrome. Here we present sensitivity analysis methods for making causal comparisons on these postinfection outcomes. We focus on estimating the survival causal effect, defined as the difference between probabilities of not yet experiencing the event in the vaccine and placebo arms, conditional on being infected regardless of treatment assignment. This group is referred to as the always-infected principal stratum. Our key assumption is monotonicity—that subjects randomized to the vaccine arm who become infected would have been infected had they been randomized to placebo. We propose nonparametric, semiparametric, and parametric methods for estimating the survival causal effect. We apply these methods to the first Phase III preventative HIV vaccine trial, VaxGen’s trial of AIDSVAX B/B.
doi:10.1198/016214507000000130
PMCID: PMC2613336  PMID: 19122791
Acquired immune deficiency syndrome; Causal inference; Kaplan–Meier; Principal stratification
3.  Increasing the Efficiency of Prevention Trials by Incorporating Baseline Covariates 
Summary
Most randomized efficacy trials of interventions to prevent HIV or other infectious diseases have assessed intervention efficacy by a method that either does not incorporate baseline covariates, or that incorporates them in a non-robust or inefficient way. Yet, it has long been known that randomized treatment effects can be assessed with greater efficiency by incorporating baseline covariates that predict the response variable. Tsiatis et al. (2007) and Zhang et al. (2008) advocated a semiparametric efficient approach, based on the theory of Robins et al. (1994), for consistently estimating randomized treatment effects that optimally incorporates predictive baseline covariates, without any parametric assumptions. They stressed the objectivity of the approach, which is achieved by separating the modeling of baseline predictors from the estimation of the treatment effect. While their work adequately justifies implementation of the method for large Phase 3 trials (because its optimality is in terms of asymptotic properties), its performance for intermediate-sized screening Phase 2b efficacy trials, which are increasing in frequency, is unknown. Furthermore, the past work did not consider a right-censored time-to-event endpoint, which is the usual primary endpoint for a prevention trial. For Phase 2b HIV vaccine efficacy trials, we study finite-sample performance of Zhang et al.'s (2008) method for a dichotomous endpoint, and develop and study an adaptation of this method to a discrete right-censored time-to-event endpoint. We show that, given the predictive capacity of baseline covariates collected in real HIV prevention trials, the methods achieve 5-15% gains in efficiency compared to methods in current use. We apply the methods to the first HIV vaccine efficacy trial. This work supports implementation of the discrete failure time method for prevention trials.
doi:10.2202/1948-4690.1002
PMCID: PMC2997740  PMID: 21152074
Auxiliary; Covariate Adjustment; Intermediate-sized Phase 2b Efficacy Trial; Semiparametric Efficiency
4.  Intermittent Preventive Treatment for Malaria in Papua New Guinean Infants Exposed to Plasmodium falciparum and P. vivax: A Randomized Controlled Trial 
PLoS Medicine  2012;9(3):e1001195.
A three-arm randomized trial conducted among infants in Papua New Guinea estimates the preventive effect against malaria episodes of intermittent preventive treatment, in an area where children are exposed to both falciparum and vivax malaria.
Background
Intermittent preventive treatment in infants (IPTi) has been shown in randomized trials to reduce malaria-related morbidity in African infants living in areas of high Plasmodium falciparum (Pf) transmission. It remains unclear whether IPTi is an appropriate prevention strategy in non-African settings or those co-endemic for P. vivax (Pv).
Methods and Findings
In this study, 1,121 Papua New Guinean infants were enrolled into a three-arm placebo-controlled randomized trial and assigned to sulfadoxine-pyrimethamine (SP) (25 mg/kg and 1.25 mg/kg) plus amodiaquine (AQ) (10 mg/kg, 3 d, n = 374), SP plus artesunate (AS) (4 mg/kg, 3 d, n = 374), or placebo (n = 373), given at 3, 6, 9 and 12 mo. Both participants and study teams were blinded to treatment allocation. The primary end point was protective efficacy (PE) against all episodes of clinical malaria from 3 to 15 mo of age. Analysis was by modified intention to treat. The PE (compared to placebo) against clinical malaria episodes (caused by all species) was 29% (95% CI, 10–43, p≤0.001) in children receiving SP-AQ and 12% (95% CI, −11 to 30, p = 0.12) in those receiving SP-AS. Efficacy was higher against Pf than Pv. In the SP-AQ group, Pf incidence was 35% (95% CI, 9–54, p = 0.012) and Pv incidence was 23% (95% CI, 0–41, p = 0.048) lower than in the placebo group. IPTi with SP-AS protected only against Pf episodes (PE = 31%, 95% CI, 4–51, p = 0.027), not against Pv episodes (PE = 6%, 95% CI, −24 to 26, p = 0.759). Number of observed adverse events/serious adverse events did not differ between treatment arms (p>0.55). None of the serious adverse events were thought to be treatment-related, and the vomiting rate was low in both treatment groups (1.4%–2.0%). No rebound in malaria morbidity was observed for 6 mo following the intervention.
Conclusions
IPTi using a long half-life drug combination is efficacious for the prevention of malaria and anemia in infants living in a region highly endemic for both Pf and Pv.
Trial registration
ClinicalTrials.gov NCT00285662
Please see later in the article for the Editors' Summary
Editors' Summary
Background
Malaria is a major global public health problem. Half the world's population is at risk of this parasitic disease, which kills about one million people (mainly young children in sub-Saharan Africa) every year. Most of these deaths are caused by Plasmodium falciparum but P. vivax, the commonest and most widely distributed malaria parasite, is a major cause of malaria-related morbidity (illness and death) in many of the tropical and subtropical regions of the world where malaria is endemic (always present). Malaria is transmitted to people through the bites of night-flying mosquitoes. It can be prevented by controlling the mosquitoes that spread the parasite and by sleeping under insecticide-treated nets to avoid mosquito bites. Prompt treatment of malaria with antimalarial drugs can also reduce malaria transmission. In addition, intermittent preventative treatment (IPT)—the treatment of symptom-free individuals with full therapeutic courses of antimalarial drugs at fixed intervals regardless of their infection status—has been shown to reduce malaria-related morbidity among pregnant women in malaria-endemic areas and among African infants living in areas of high P. falciparum transmission.
Why Was This Study Done?
The World Health Organization recently recommended that, in Africa, IPT should be given during infancy (called IPTi) at the same time as routine immunizations. Because the studies on which this recommendation is based were all carried out in sub-Saharan Africa, in populations where P. falciparum is the predominant parasite and P. vivax is uncommon, it is not known whether IPTi would be an appropriate prevention strategy in non-African settings or in regions where both P. falciparum and P. vivax are endemic. In this randomized placebo-controlled trial, the researchers investigate the efficacy of IPTi in infants living in an area of Papua New Guinea where P. falciparum and P. vivax are both highly endemic. In a randomized placebo-controlled trial, the effects of an intervention and of a placebo (dummy) intervention are compared in groups of individuals chosen through the play of chance.
What Did the Researchers Do and Find?
The researchers assigned more than 1,000 infants to receive sulfadoxine/pyrimethamine (SP) plus amodiaquine (AQ) (SP and AQ are long-lasting antimalarial drugs), SP plus artesunate (AS) (AS is a short-lasting antimalarial), or placebo at 3, 6, 9, and 12 months old. They recorded the number of malaria episodes that occurred among the children between the ages of 3 and 15 months. Then, by comparing the number of episodes occurring among the children receiving SP-AS or SP-AQ with the number occurring among the children receiving placebo, the researchers calculated the protective efficacy of the two drug combinations over the study period. The protective efficacy of IPTi against all clinical malaria episodes (P. falciparum and P. vivax combined) was 29% for SP-AQ, but SP-AS was not associated with a statistically significant reduction in all malaria episodes as compared to placebo. For the two species of malaria separately, the incidence of P. falciparum malaria was 35% lower among the children receiving SP-AQ than among the children receiving placebo, whereas the incidence of P. vivax was reduced by 23%; IPTi with SP-AS provided protection only against P. falciparum malaria (protective efficacy 31%). Importantly, the number of adverse events (possible drug side effects) was similar in all the treatment arms, none of the severe adverse events were treatment-related, and there was no rebound in malaria-related morbidity for six months following the end of the intervention.
What Do These Findings Mean?
These findings show that IPTi using a combination of long-lasting antimalarial drugs (SP-AQ) can effectively and safely prevent malaria in a non-African population living in a region where P. falciparum and P. vivax are both highly endemic. Importantly, they also show that IPTi with SP-AQ can prevent both P. falciparum and P. vivax malaria. For Papua New Guinea, these findings suggest that SP-AQ is an appropriate drug choice for IPTi, particularly since the replacement of SP-AQ by artemether-lumefantrine as the national first line treatment for malaria will reduce the selection pressure for resistance against SP and AQ. However, although these finding provide proof-of-principle evidence for the efficacy and safety of IPTi, further studies are needed to identify the most effective combinations of long-lasting antimalarial drugs for use in IPTi in other malaria-endemic regions.
Additional Information
Please access these web sites via the online version of this summary at http://dx.doi.org/10.1371/journal.pmed.1001195.
Information is available from the World Health Organization on malaria (in several languages); the 2011 World Malaria Report provides details of the current global malaria situation; and the WHO policy recommendation on IPTi for P. falciparum malaria control in Africa is available
The US Centers for Disease Control and Prevention provide information on malaria (in English and Spanish), including a selection of personal stories about malaria
Information is available from the Roll Back Malaria Partnership on the global control of malaria, including a fact sheet about malaria in children and information on malaria in Papua New Guinea
The IPTi Consortium was established to evaluate IPTi and inform public health policy making
The Malaria Vaccine Initiative has a fact sheet on P. vivax malaria
Vivaxmalaria.com provides information about P. vivax
MedlinePlus provides links to additional information on malaria (in English and Spanish)
doi:10.1371/journal.pmed.1001195
PMCID: PMC3313928  PMID: 22479155
5.  Intermittent Intravaginal Antibiotic Treatment of Bacterial Vaginosis in HIV-Uninfected and -Infected Women: A Randomized Clinical Trial 
PLoS Clinical Trials  2007;2(2):e10.
Objective:
Assess efficacy of intermittent intravaginal metronidazole gel treatment in reducing frequency of bacterial vaginosis (BV).
Design:
Randomized, double-masked, placebo-controlled phase 3 trial.
Setting:
Postnatal and family planning clinics of the Queen Elizabeth Central Hospital and two health centers in Blantyre, Malawi.
Participants:
Nonpregnant HIV-uninfected and -infected women.
Intervention:
Intravaginal metronidazole treatment and placebo gels provided at baseline and every 3 mo for 1 y.
Outcome measures:
Primary: Cross-sectional and longitudinal comparisons of BV frequency at baseline, 1 mo after product dispensation (post-treatment evaluation [PTE]), and every quarterly visit. Secondary: Effect of treatment on BV clearance and recurrence.
Results:
Baseline: 842 HIV-uninfected and 844 HIV-infected women were enrolled. The frequency of BV at baseline in treatment and placebo arms, respectively, was 45.9% and 46.8% among HIV-uninfected women, and 60.5% and 56.9% among HIV-infected women. Primary outcomes: At the PTEs the prevalence of BV was consistently lower in treatment than placebo arms irrespective of HIV status. The differences were statistically significant mainly in HIV-uninfected women. Prevalence of BV was also reduced over time in both treatment and placebo arms. In a multivariable analysis that controlled for other covariates, the effect of intravaginal metronidazole treatment gel compared with placebo was not substantial: adjusted relative risk (RR) 0.90, 95% confidence interval (CI) 0.83–0.97 in HIV-uninfected women and adjusted RR 0.95, 95% CI 0.89–1.01 in HIV-infected women. Secondary outcomes: Intravaginal metronidazole treatment gel significantly increased BV clearance (adjusted hazard ratio [HR] 1.34, 95% CI 1.07–1.67 among HIV-uninfected women and adjusted HR 1.29, 95% CI 1.06–1.58 among HIV-infected women) but was not associated with decreased BV recurrence. Safety: No serious adverse events were related to use of intravaginal gels.
Conclusion:
Intermittent microbicide treatment with intravaginal gels is an innovative approach that can reduce the frequency of vaginal infections such as BV.
Editorial Commentary
Background: Bacterial vaginosis (BV) results from a change in the normal balance of bacteria in the vaginal tract, and is very common. In pregnant women, it is associated with poorer outcomes in pregnancy, and is also linked with HIV transmission (although it is not certain that BV actually increases the chance of getting HIV—just because these two occur together it does not necessarily follow that one causes the other). BV can be treated with metronidazole tablets, although these can cause gut symptoms and should not be taken repeatedly. The researchers wanted to carry out a multiclinic–based trial to find out whether a metronidazole gel applied intermittently to the vagina (for five nights every three months) would reduce the frequency of BV among women in Malawi. HIV-infected and HIV-uninfected women, recruited from postnatal and family planning clinics, were randomized to receive either metronidazole gels, or equivalent placebo gels, every three months and were then followed up for 12 months. The primary outcome for the trial was the proportion of women with BV at each quarterly follow-up visit, and the researchers intended to compare this outcome between treatment arms at each visit and also to look at the overall changes over time among women receiving either metronidazole or placebo, looking separately at HIV-infected and HIV-uninfected women.
What this trial shows: In total 1,686 women took part in the trial (842 not infected with HIV, and 844 infected with HIV). The proportion of HIV-uninfected women with BV dropped by around 20% over the course of the trial, both in women using metronidazole and in those using placebo. However, when comparing the proportion of HIV-uninfected women with BV between the two arms of the trial, there did not seem to be a consistent effect: differences were statistically significant at some time points and not others. Among HIV-infected women, there was also a drop over the course of the trial in the proportion of women with BV, irrespective of whether they used metronidazole or placebo. Again, when comparing the rate of BV among HIV-infected women between study arms (metronidazole versus placebo), the researchers did not see a consistent trend; differences were statistically significant at some time points but not others. Overall, when comparing metronidazole and placebo in an analysis that controlled for other factors, the metronidazole gel seemed to show a small effect in reduction of BV among HIV-uninfected women, but no obvious effect among HIV-infected women.
Strengths and limitations: Strengths in the design of this trial include the sample size, which was appropriate to detect an important effect of the metronidazole gel (versus placebo) had one existed, and the randomization and blinding procedures, which were designed to minimize the chance that the trialists or women being enrolled could anticipate to which arm of the trial they might be assigned. A key limitation of this study, as the researchers acknowledge, is the absence of a “no treatment” study arm. The frequency of BV dropped over the course of the trial in women using the placebo gel, raising the possibility that the placebo actually has some effect on bacteria in the vagina. However, a trial with a “no treatment” arm would pose its own problems, since trialists and participants would then not be fully blinded as to their treatment status.
Contribution to the evidence: This trial adds data on the efficacy of metronidazole gel when used intermittently, and among women in the community who may or may not actually have BV. Previous studies have evaluated treatment with metronidazole among women who already have symptoms or a diagnosis of BV. The findings of this trial rule out a substantial effect of metronidazole gel, as compared to placebo gel, in reducing the frequency of BV in this setting.
doi:10.1371/journal.pctr.0020010
PMCID: PMC1851729  PMID: 17318258
6.  Efficient Estimation of Semiparametric Transformation Models for Two-Phase Cohort Studies 
Under two-phase cohort designs, such as case-cohort and nested case-control sampling, information on observed event times, event indicators, and inexpensive covariates is collected in the first phase, and the first-phase information is used to select subjects for measurements of expensive covariates in the second phase; inexpensive covariates are also used in the data analysis to control for confounding and to evaluate interactions. This paper provides efficient estimation of semiparametric transformation models for such designs, accommodating both discrete and continuous covariates and allowing inexpensive and expensive covariates to be correlated. The estimation is based on the maximization of a modified nonparametric likelihood function through a generalization of the expectation-maximization algorithm. The resulting estimators are shown to be consistent, asymptotically normal and asymptotically efficient with easily estimated variances. Simulation studies demonstrate that the asymptotic approximations are accurate in practical situations. Empirical data from Wilms’ tumor studies and the Atherosclerosis Risk in Communities (ARIC) study are presented.
doi:10.1080/01621459.2013.842172
PMCID: PMC3960088  PMID: 24659837
Case-cohort design; EM algorithm; Kernel estimation; Nested case-control sampling; Nonparametric likelihood; Semiparametric efficiency
7.  Landmark Estimation of Survival and Treatment Effect in a Randomized Clinical Trial 
Summary
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.
doi:10.1080/01621459.2013.842488
PMCID: PMC3960087  PMID: 24659838
Efficiency Augmentation; Kaplan Meier; Landmark Prediction; Semi-competing Risks; Survival Analysis
8.  Methods of Blinding in Reports of Randomized Controlled Trials Assessing Pharmacologic Treatments: A Systematic Review 
PLoS Medicine  2006;3(10):e425.
Background
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.
Conclusions
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
Background.
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 http://dx.doi.org/10.1371/journal.pmed.0030425.
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
ClinicalTrials.gov, 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
doi:10.1371/journal.pmed.0030425
PMCID: PMC1626553  PMID: 17076559
9.  Semiparametric Estimation of Treatment Effect with Time-Lagged Response in the Presence of Informative Censoring 
Lifetime data analysis  2011;17(4):566-593.
In many randomized clinical trials, the primary response variable, for example, the survival time, is not observed directly after the patients enroll in the study but rather observed after some period of time (lag time). It is often the case that such a response variable is missing for some patients due to censoring that occurs when the study ends before the patient’s response is observed or when the patients drop out of the study. It is often assumed that censoring occurs at random which is referred to as noninformative censoring; however, in many cases such an assumption may not be reasonable. If the missing data are not analyzed properly, the estimator or test for the treatment effect may be biased. In this paper, we use semiparametric theory to derive a class of consistent and asymptotically normal estimators for the treatment effect parameter which are applicable when the response variable is right censored. The baseline auxiliary covariates and post-treatment auxiliary covariates, which may be time-dependent, are also considered in our semiparametric model. These auxiliary covariates are used to derive estimators that both account for informative censoring and are more efficient then the estimators which do not consider the auxiliary covariates.
doi:10.1007/s10985-011-9199-8
PMCID: PMC3217309  PMID: 21706378
Informative censoring; Influence function; Logrank test; Nuisance tangent space; Proportional hazards model; Regular and asymptotically linear estimators
10.  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.
doi:10.1002/sim.5433
PMCID: PMC3855673  PMID: 22733628
covariate adjustment; false selection rate control; oracle property; semiparametric treatment effect estimation; shrinkage methods; variable selection
11.  Switching HIV Treatment in Adults Based on CD4 Count Versus Viral Load Monitoring: A Randomized, Non-Inferiority Trial in Thailand 
PLoS Medicine  2013;10(8):e1001494.
Using a randomized controlled trial, Marc Lallemant and colleagues ask if a CD4-based monitoring and treatment switching strategy provides a similar clinical outcome compared to the standard viral load-based strategy for adults with HIV in Thailand.
Please see later in the article for the Editors' Summary
Background
Viral load (VL) is recommended for monitoring the response to highly active antiretroviral therapy (HAART) but is not routinely available in most low- and middle-income countries. The purpose of the study was to determine whether a CD4-based monitoring and switching strategy would provide a similar clinical outcome compared to the standard VL-based strategy in Thailand.
Methods and Findings
The Programs for HIV Prevention and Treatment (PHPT-3) non-inferiority randomized clinical trial compared a treatment switching strategy based on CD4-only (CD4) monitoring versus viral-load (VL). Consenting participants were antiretroviral-naïve HIV-infected adults (CD4 count 50–250/mm3) initiating non-nucleotide reverse transcriptase inhibitor (NNRTI)-based therapy. Randomization, stratified by site (21 public hospitals), was performed centrally after enrollment. Clinicians were unaware of the VL values of patients randomized to the CD4 arm. Participants switched to second-line combination with confirmed CD4 decline >30% from peak (within 200 cells from baseline) in the CD4 arm, or confirmed VL >400 copies/ml in the VL arm. Primary endpoint was clinical failure at 3 years, defined as death, new AIDS-defining event, or CD4 <50 cells/mm3. The 3-year Kaplan-Meier cumulative risks of clinical failure were compared for non-inferiority with a margin of 7.4%. In the intent to treat analysis, data were censored at the date of death or at last visit. The secondary endpoints were difference in future-drug-option (FDO) score, a measure of resistance profiles, virologic and immunologic responses, and the safety and tolerance of HAART. 716 participants were randomized, 356 to VL monitoring and 360 to CD4 monitoring. At 3 years, 319 participants (90%) in VL and 326 (91%) in CD4 were alive and on follow-up. The cumulative risk of clinical failure was 8.0% (95% CI 5.6–11.4) in VL versus 7.4% (5.1–10.7) in CD4, and the upper-limit of the one-sided 95% CI of the difference was 3.4%, meeting the pre-determined non-inferiority criterion. Probability of switch for study criteria was 5.2% (3.2–8.4) in VL versus 7.5% (5.0–11.1) in CD4 (p = 0.097). Median time from treatment initiation to switch was 11.7 months (7.7–19.4) in VL and 24.7 months (15.9–35.0) in CD4 (p = 0.001). The median duration of viremia >400 copies/ml at switch was 7.2 months (5.8–8.0) in VL versus 15.8 months (8.5–20.4) in CD4 (p = 0.002). FDO scores were not significantly different at time of switch. No adverse events related to the monitoring strategy were reported.
Conclusions
The 3-year rates of clinical failure and loss of treatment options did not differ between strategies although the longer-term consequences of CD4 monitoring would need to be investigated. These results provide reassurance to treatment programs currently based on CD4 monitoring as VL measurement becomes more affordable and feasible in resource-limited settings.
Trial registration
ClinicalTrials.gov NCT00162682
Please see later in the article for the Editors' Summary
Editors' Summary
Background
About 34 million people (most of them living in low-and middle-income countries) are currently infected with HIV, the virus that causes AIDS. HIV infection leads to the destruction of immune system cells (including CD4 cells, a type of white blood cell), leaving infected individuals susceptible to other infections. Early in the AIDS epidemic, most HIV-infected individuals died within 10 years of infection. Then, in 1996, highly active antiretroviral therapy (HAART)—combined drugs regimens that suppress viral replication and allow restoration of the immune system—became available. For people living in affluent countries, HIV/AIDS became a chronic condition but, because HAART was expensive, HIV/AIDS remained a fatal illness for people living in resource-limited countries. In 2003, the international community declared HIV/AIDS a global health emergency and, in 2006, it set the target of achieving universal global access to HAART by 2010. By the end of 2011, 8 million of the estimated 14.8 million people in need of HAART in low- and middle-income countries were receiving treatment.
Why Was This Study Done?
At the time this trial was conceived, national and international recommendations were that HIV-positive individuals should start HAART when their CD4 count fell below 200 cells/mm3 and should have their CD4 count regularly monitored to optimize HAART. In 2013, the World Health Organization (WHO) recommendations were updated to promote expanded eligibility for HAART with a CD4 of 500 cells/mm3 or less for adults, adolescents, and older children although priority is given to individuals with CD4 count of 350 cells/mm3 or less. Because HIV often becomes resistant to first-line antiretroviral drugs, WHO also recommends that viral load—the amount of virus in the blood—should be monitored so that suspected treatment failures can be confirmed and patients switched to second-line drugs in a timely manner. This monitoring and switching strategy is widely used in resource-rich settings, but is still very difficult to implement for low- and middle-income countries where resources for monitoring are limited and access to costly second-line drugs is restricted. In this randomized non-inferiority trial, the researchers compare the performance of a CD4-based treatment monitoring and switching strategy with the standard viral load-based strategy among HIV-positive adults in Thailand. In a randomized trial, individuals are assigned different interventions by the play of chance and followed up to compare the effects of these interventions; a non-inferiority trial investigates whether one treatment is not worse than another.
What Did the Researchers Do and Find?
The researchers assigned about 700 HIV-positive adults who were beginning HAART for the first time to have their CD4 count (CD4 arm) or their CD4 count and viral load (VL arm) determined every 3 months. Participants were switched to a second-line therapy if their CD4 count declined by more than 30% from their peak CD4 count (CD4 arm) or if a viral load of more than 400 copies/ml was recorded (VL arm). The 3-year cumulative risk of clinical failure (defined as death, a new AIDS-defining event, or a CD4 count of less than 50 cells/mm3) was 8% in the VL arm and 7.4% in the CD4 arm. This difference in clinical failure risk met the researchers' predefined criterion for non-inferiority. The probability of a treatment switch was similar in the two arms, but the average time from treatment initiation to treatment switch and the average duration of a high viral load after treatment switch were both longer in the CD4 arm than in the VL arm. Finally, the future-drug-option score, a measure of viral drug resistance profiles, was similar in the two arms at the time of treatment switch.
What Do These Findings Mean?
These findings suggest that, in Thailand, a CD4 switching strategy is non-inferior in terms of clinical outcomes among HIV-positive adults 3 years after beginning HAART when compared to the recommended viral load-based switching strategy and that there is no difference between the strategies in terms of viral suppression and immune restoration after 3-years follow-up. Importantly, however, even though patients in the CD4 arm spent longer with a high viral load than patients in the VL arm, the emergence of HIV mutants resistant to antiretroviral drugs was similar in the two arms. Although these findings provide no information about the long-term outcomes of the two monitoring strategies and may not be generalizable to routine care settings, they nevertheless provide reassurance that using CD4 counts alone to monitor HAART in HIV treatment programs in resource-limited settings is an appropriate strategy to use as viral load measurement becomes more affordable and feasible in these settings.
Additional Information
Please access these websites via the online version of this summary at http://dx.doi.org/10.1371/journal.pmed.1001494.
The World Health Organization provides information on all aspects of HIV/AIDS (in several languages); its 2010 recommendations for antiretroviral therapy for HIV infection in adults and adolescents are available as well as the June 2013 Consolidated guidelines on the use of antiretroviral drugs for treating and preventing HIV infection: recommendations for a public health approach
The 2012 UNAIDS World AIDS Day Report provides up-to-date information about the AIDS epidemic and efforts to halt it
Information is available from the US National Institute of Allergy and Infectious Diseases on HIV infection and AIDS
NAM/aidsmap provides basic information about HIV/AIDS and summaries of recent research findings on HIV care and treatment
Information is available from Avert, an international AIDS charity on many aspects of HIV/AIDS, including information on the global HIV/AIDS epidemic, on HIV and AIDS in Thailand, on universal access to AIDS treatment, and on starting, monitoring and switching HIV treatment (in English and Spanish)
The UK National Health Service Choices website provides information (including personal stories) about HIV and AIDS
More information about this trial (the PHPT-3 trial) is available
Patient stories about living with HIV/AIDS are available through Avert; the nonprofit website Healthtalkonline also provides personal stories about living with HIV, including stories about HIV treatment
doi:10.1371/journal.pmed.1001494
PMCID: PMC3735458  PMID: 23940461
12.  Short-Term Efficacy of Rofecoxib and Diclofenac in Acute Shoulder Pain: A Placebo-Controlled Randomized Trial 
PLoS Clinical Trials  2007;2(3):e9.
Objectives:
To evaluate the short-term symptomatic efficacy of rofecoxib and diclofenac versus placebo in acute episodes of shoulder pain.
Design:
Randomized controlled trial of 7 days.
Setting:
Rheumatologists and/or general practitioners totaling 47.
Participants:
Acute shoulder pain.
Interventions:
Rofecoxib 50 mg once daily, diclofenac 50 mg three times daily, and placebo.
Outcome measures:
Pain, functional impairment, patient's global assessment of his/her disease activity, and local steroid injection requirement for persistent pain. The primary variable was the Kaplan-Meier estimates of the percentage of patients at day 7 fulfilling the definition of success (improvement in pain intensity and a low pain level sustained to the end of the 7 days of the study; log-rank test).
Results:
There was no difference in the baseline characteristics between the three groups (rofecoxib n = 88, placebo n = 94, and diclofenac n = 89). At day 7, the Kaplan-Meier estimates of successful patients was higher in the treatment groups than in the placebo (54%, 56%, and 38% in the diclofenac, rofecoxib, and placebo groups respectively, p = 0.0070 and p = 0.0239 for placebo versus rofecoxib and diclofenac, respectively). During the 7 days of the study, there was a statistically significant difference between placebo and both active arms (rofecoxib and diclofenac) in all the evaluated outcome measures A local steroid injection had to be performed in 33 (35%) and 19 (22%) patients in the placebo and rofecoxib group respectively. Number needed to treat to avoid such rescue therapy was 7 patients (95% confidence interval 5–15).
Conclusion:
This study highlights the methodological aspects of clinical trials, e.g., eligibility criteria and outcome measures, in acute painful conditions. The data also establish that diclofenac and rofecoxib are effective therapies for the management of acute painful shoulder and that they reduce the requirement for local steroid injection.
Editorial Commentary
Background: Shoulder pain is a very common complaint that presents in primary care, and there are many different possible causes. Acute pain would normally be managed with nonsteroidal anti-inflammatory drugs (NSAIDs), supplemented with steroid injections (which are often reserved for the treatment of severe or persistent pain). One NSAID, diclofenac, is used frequently for this condition, but other NSAIDs might also be effective. A subgroup of NSAIDs called the Cox-2 selective inhibitors specifically inhibit one particular enzyme (cyclo-oxygenase, shortened to Cox-2) which is involved in inflammation and pain. These drugs are thought to be less likely to cause stomach irritation than other NSAIDs. Therefore the researchers in this study carried out a short-term, three-way clinical trial comparing diclofenac with one particular Cox-2 inhibitor, rofecoxib, and placebo in patients with acute shoulder pain. However, rofecoxib was withdrawn from the market in September 2004 because of evidence that use of the drug was associated with an increased risk of heart attacks and strokes, and controversy remains regarding the risk of such events among users of other Cox-2 inhibitors.
What this trial shows: The main aim of this trial was to compare the level of pain relief over seven days of treatment with either diclofenac or rofecoxib, as compared to placebo. The primary outcome measure used in the trial was the proportion of patients achieving a 50% or greater decrease in pain levels over the course of the study, measured using a numerical rating scale. A total of 273 participants were recruited into the trial and at day 7 the proportion achieving a 30% decrease in pain was 38% in the placebo arm, 54% in the diclofenac arm, and 56% in the rofecoxib arm. The differences in this outcome measure between diclofenac and placebo and between rofecoxib and placebo were statistically significant; however, the researchers did not carry out a direct comparison between diclofenac and rofecoxib. The rates of adverse events were roughly comparable between all three arms of the trial, although the study was not originally planned to be large enough to detect differences in the rates of such events, so it is not possible to conclude whether there was any true difference.
Strengths and limitations: The randomization procedures used in the study minimize the possibility of bias in assigning patients to treatment arms. Bias in assessment of outcomes was also minimized by ensuring that steps were taken to prevent investigators and patients from knowing which drugs a particular patient received until the end of the trial. A key limitation of the study is the short follow-up, only seven days, and it is therefore unclear whether efficacy and safety of these drugs would continue for the much longer periods of time (weeks or even months) for which these patients might need pain relief. Finally, patients randomized to the placebo arm received no treatment for the seven days of the study other than acetaminophen or steroid injections (which would result in withdrawal from the trial). This design does not limit interpretation of the data but could be criticized because of concern over whether the patients receiving placebo received adequate pain relief.
Contribution to the evidence: This study provides some data on the efficacy of diclofenac and rofecoxib, as compared to placebo in treatment of this condition. Given that rofecoxib is now withdrawn, the efficacy of this drug is no longer relevant. However, the information from this trial should help in designing future studies of NSAIDs in shoulder pain, for example to define appropriate trial outcomes, sample size, and other aspects of study design.
doi:10.1371/journal.pctr.0020009
PMCID: PMC1817652  PMID: 17347681
13.  Estimating the Effect of a Community-Based Intervention with Two Communities 
Journal of causal inference  2013;1(1):83-106.
Due to the need to evaluate the effectiveness of community-based programs in practice, there is substantial interest in methods to estimate the causal effects of community-level treatments or exposures on individual level outcomes. The challenge one is confronted with is that different communities have different environmental factors affecting the individual outcomes, and all individuals in a community share the same environment and intervention. In practice, data are often available from only a small number of communities, making it difficult if not impossible to adjust for these environmental confounders. In this paper we consider an extreme version of this dilemma, in which two communities each receives a different level of the intervention, and covariates and outcomes are measured on a random sample of independent individuals from each of the two populations; the results presented can be straightforwardly generalized to settings in which more than two communities are sampled. We address the question of what conditions are needed to estimate the causal effect of the intervention, defined in terms of an ideal experiment in which the exposed level of the intervention is assigned to both communities and individual outcomes are measured in the combined population, and then the clock is turned back and a control level of the intervention is assigned to both communities and individual outcomes are measured in the combined population. We refer to the difference in the expectation of these outcomes as the marginal (overall) treatment effect. We also discuss conditions needed for estimation of the treatment effect on the treated community. We apply a nonparametric structural equation model to define these causal effects and to establish conditions under which they are identified. These identifiability conditions provide guidance for the design of studies to investigate community level causal effects and for assessing the validity of causal interpretations when data are only available from a few communities. When the identifiability conditions fail to hold, the proposed statistical parameters still provide nonparametric treatment effect measures (albeit non-causal) whose statistical interpretations do not depend on model specifications. In addition, we study the use of a matched cohort sampling design in which the units of different communities are matched on individual factors. Finally, we provide semiparametric efficient and doubly robust targeted MLE estimators of the community level causal effect based on i.i.d. sampling and matched cohort sampling.
doi:10.1515/jci-2012-0011
PMCID: PMC4254657  PMID: 25485209
causal effect; causal effect among the treated; community-based intervention; efficient influence curve; environmental confounding
14.  Robust extraction of covariate information to improve estimation efficiency in randomized trials‡ 
Statistics in medicine  2011;30(19):2389-2408.
In randomized trials, investigators typically rely upon an unadjusted estimate of the mean outcome within each treatment arm to draw causal inferences. Statisticians have underscored the gain in efficiency that can be achieved from covariate adjustment in randomized trials with a focus on problems involving linear models. Despite recent theoretical advances, there has been a reluctance to adjust for covariates based on two primary reasons: (i) covariate-adjusted estimates based on conditional logistic regression models have been shown to be less precise and (ii) concern over the opportunity to manipulate the model selection process for covariate adjustments to obtain favorable results. In this paper, we address these two issues and summarize recent theoretical results on which is based a proposed general methodology for covariate adjustment under the framework of targeted maximum likelihood estimation in trials with two arms where the probability of treatment is 50%. The proposed methodology provides an estimate of the true causal parameter of interest representing the population-level treatment effect. It is compared with the estimates based on conditional logistic modeling, which only provide estimates of subgroup-level treatment effects rather than marginal (unconditional) treatment effects. We provide a clear criterion for determining whether a gain in efficiency can be achieved with covariate adjustment over the unadjusted method. We illustrate our strategy using a resampled clinical trial dataset from a placebo controlled phase 4 study. Results demonstrate that gains in efficiency can be achieved even with binary outcomes through covariate adjustment leading to increased statistical power.
doi:10.1002/sim.4301
PMCID: PMC4113477  PMID: 21751231
clinical trials; efficiency; covariate adjustment; variable selection
15.  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.
doi:10.1080/01621459.2011.646932
PMCID: PMC3545451  PMID: 23329858
Causal inference; Double robustness; Longitudinal data analysis; Missing data; Rubin causal model; Semiparametric efficient estimation
16.  ADAPTIVE MATCHING IN RANDOMIZED TRIALS AND OBSERVATIONAL STUDIES 
Journal of statistical research  2012;46(2):113-156.
SUMMARY
In many randomized and observational studies the allocation of treatment among a sample of n independent and identically distributed units is a function of the covariates of all sampled units. As a result, the treatment labels among the units are possibly dependent, complicating estimation and posing challenges for statistical inference. For example, cluster randomized trials frequently sample communities from some target population, construct matched pairs of communities from those included in the sample based on some metric of similarity in baseline community characteristics, and then randomly allocate a treatment and a control intervention within each matched pair. In this case, the observed data can neither be represented as the realization of n independent random variables, nor, contrary to current practice, as the realization of n/2 independent random variables (treating the matched pair as the independent sampling unit). In this paper we study estimation of the average causal effect of a treatment under experimental designs in which treatment allocation potentially depends on the pre-intervention covariates of all units included in the sample. We define efficient targeted minimum loss based estimators for this general design, present a theorem that establishes the desired asymptotic normality of these estimators and allows for asymptotically valid statistical inference, and discuss implementation of these estimators. We further investigate the relative asymptotic efficiency of this design compared with a design in which unit-specific treatment assignment depends only on the units’ covariates. Our findings have practical implications for the optimal design and analysis of pair matched cluster randomized trials, as well as for observational studies in which treatment decisions may depend on characteristics of the entire sample.
PMCID: PMC4119765  PMID: 25097298
Cluster randomized trials; matching; asymptotic linearity of an estimator; causal effect; efficient influence curve; empirical process; confounding; dependent treatment allocation; G-computation formula; influence curve; loss function; adaptive randomization; semiparametric statistical model; targeted maximum likelihood estimation; targeted minimum loss based estimation (TMLE)
17.  Endovascular Radiofrequency Ablation for Varicose Veins 
Executive Summary
Objective
The objective of the MAS evidence review was to conduct a systematic review of the available evidence on the safety, effectiveness, durability and cost–effectiveness of endovascular radiofrequency ablation (RFA) for the treatment of primary symptomatic varicose veins.
Background
The Ontario Health Technology Advisory Committee (OHTAC) met on August 26th, 2010 to review the safety, effectiveness, durability, and cost-effectiveness of RFA for the treatment of primary symptomatic varicose veins based on an evidence-based review by the Medical Advisory Secretariat (MAS).
Clinical Condition
Varicose veins (VV) are tortuous, twisted, or elongated veins. This can be due to existing (inherited) valve dysfunction or decreased vein elasticity (primary venous reflux) or valve damage from prior thrombotic events (secondary venous reflux). The end result is pooling of blood in the veins, increased venous pressure and subsequent vein enlargement. As a result of high venous pressure, branch vessels balloon out leading to varicosities (varicose veins).
Symptoms typically affect the lower extremities and include (but are not limited to): aching, swelling, throbbing, night cramps, restless legs, leg fatigue, itching and burning. Left untreated, venous reflux tends to be progressive, often leading to chronic venous insufficiency (CVI). A number of complications are associated with untreated venous reflux: including superficial thrombophlebitis as well as variceal rupture and haemorrhage. CVI often results in chronic skin changes referred to as stasis dermatitis. Stasis dermatitis is comprised of a spectrum of cutaneous abnormalities including edema, hyperpigmentation, eczema, lipodermatosclerosis and stasis ulceration. Ulceration represents the disease end point for severe CVI. CVI is associated with a reduced quality of life particularly in relation to pain, physical function and mobility. In severe cases, VV with ulcers, QOL has been rated to be as bad or worse as other chronic diseases such as back pain and arthritis.
Lower limb VV is a very common disease affecting adults – estimated to be the 7th most common reason for physician referral in the US. There is a very strong familial predisposition to VV. The risk in offspring is 90% if both parents affected, 20% when neither affected and 45% (25% boys, 62% girls) if one parent affected. The prevalence of VV worldwide ranges from 5% to 15% among men and 3% to 29% among women varying by the age, gender and ethnicity of the study population, survey methods and disease definition and measurement. The annual incidence of VV estimated from the Framingham Study was reported to be 2.6% among women and 1.9% among men and did not vary within the age range (40-89 years) studied.
Approximately 1% of the adult population has a stasis ulcer of venous origin at any one time with 4% at risk. The majority of leg ulcer patients are elderly with simple superficial vein reflux. Stasis ulcers are often lengthy medical problems and can last for several years and, despite effective compression therapy and multilayer bandaging are associated with high recurrence rates. Recent trials involving surgical treatment of superficial vein reflux have resulted in healing and significantly reduced recurrence rates.
Endovascular Radiofrequency Ablation for Varicose Veins
RFA is an image-guided minimally invasive treatment alternative to surgical stripping of superficial venous reflux. RFA does not require an operating room or general anaesthesia and has been performed in an outpatient setting by a variety of medical specialties including surgeons and interventional radiologists. Rather than surgically removing the vein, RFA works by destroying or ablating the refluxing vein segment using thermal energy delivered through a radiofrequency catheter.
Prior to performing RFA, color-flow Doppler ultrasonography is used to confirm and map all areas of venous reflux to devise a safe and effective treatment plan. The RFA procedure involves the introduction of a guide wire into the target vein under ultrasound guidance followed by the insertion of an introducer sheath through which the RFA catheter is advanced. Once satisfactory positioning has been confirmed with ultrasound, a tumescent anaesthetic solution is injected into the soft tissue surrounding the target vein along its entire length. This serves to anaesthetize the vein, insulate the heat from damaging adjacent structures, including nerves and skin and compresses the vein increasing optimal contact of the vessel wall with the electrodes or expanded prongs of the RF device. The RF generator is then activated and the catheter is slowly pulled along the length of the vein. At the end of the procedure, hemostasis is then achieved by applying pressure to the vein entry point.
Adequate and proper compression stockings and bandages are applied after the procedure to reduce the risk of venous thromboembolism and to reduce postoperative bruising and tenderness. Patients are encouraged to walk immediately after the procedure. Follow-up protocols vary, with most patients returning 1 to 3 weeks later for an initial follow-up visit. At this point, the initial clinical result is assessed and occlusion of the treated vessels is confirmed with ultrasound. Patients often have a second follow-up visit 1 to 3 months following RFA at which time clinical evaluation and ultrasound are repeated. If required, additional procedures such as phlebectomy or sclerotherapy may be performed during the RFA procedure or at any follow-up visits.
Regulatory Status
The Closure System® radiofrequency generator for endovascular thermal ablation of varicose veins was approved by Health Canada as a class 3 device in March 2005, registered under medical device license 67865. The RFA intravascular catheter was approved by Health Canada in November 2007 for the ClosureFast catheter, registered under medical device license 16574. The Closure System® also has regulatory approvals in Australia, Europe (CE Mark) and the United States (FDA clearance). In Ontario, RFA is not an insured service and is currently being introduced in private clinics.
Methods
Literature Search
The MAS evidence–based review was performed to support public financing decisions. The literature search was performed on March 9th, 2010 using standard bibliographic databases for studies published up until March, 2010.
Inclusion Criteria
English language full-reports and human studies Original reports with defined study methodologyReports including standardized measurements on outcome events such as technical success, safety, effectiveness, durability, quality of life or patient satisfaction Reports involving RFA for varicose veins (great or small saphenous veins)Randomized controlled trials (RCTs), systematic reviews and meta-analysesCohort and controlled clinical studies involving ≥ 1 month ultrasound imaging follow-up
Exclusion Criteria
Non systematic reviews, letters, comments and editorials Reports not involving outcome events such as safety, effectiveness, durability, or patient satisfaction following an intervention with RFAReports not involving interventions with RFA for varicose veinsPilot studies or studies with small samples (< 50 subjects)
Summary of Findings
The MAS evidence search on the safety and effectiveness of endovascular RFA ablation of VV identified the following evidence: three HTAs, nine systematic reviews, eight randomized controlled trials (five comparing RFA to surgery and three comparing RFA to ELT), five controlled clinical trials and fourteen cohort case series (four were multicenter registry studies).
The majority (12⁄14) of the cohort studies (3,664) evaluating RFA for VV involved treatment with first generation RFA catheters and the great saphenous vein (GSV) was the target vessel in all studies. Major adverse events were uncommonly reported and the overall pooled major adverse event rate extracted from the cohort studies was 2.9% (105⁄3,664). Imaging defined treatment effectiveness of vein closure rates were variable ranging from 68% to 96% at post-operative follow-up. Vein ablation rate at 6-month follow-up was reported in four studies with rates close to 90%. Only one study reported vein closure rates at 2 years but only for a minority of the eligible cases. The two studies reporting on RFA ablation with the more efficient second generation catheters involved better follow-up and reported higher ablation rates close to 100% at 6-month follow-up with no major adverse events. A large prospective registry trial that recruited over 1,000 patients at thirty-four largely European centers reported on treatment success in six overlapping reports on selected patient subgroups at various follow-up points up to 5 year. However, the follow-up for eligible recruited patients at all time points was low resulting in inadequate estimates of longer term treatment efficacy.
The overall level of evidence of randomized trials comparing RFA with surgical ligation and vein stripping (n = 5) was graded as low to moderate. In all trials RFA ablation was performed with first generation catheters in the setting of the operating theatre under general anaesthesia, usually without tumescent anaesthesia. Procedure times were significantly longer after RFA than surgery. Recovery after treatment was significantly quicker after RFA both with return to usual activity and return to work with on average a one week less of work loss. Major adverse events occurring after surgery were higher [(1.8% (n=4) vs. 0.4% (n = 1) than after RFA but not significantly. Treatment effectiveness measured by imaging defined vein absence or vein closure was comparable in the two treatment groups. Significant improvements in vein symptoms and quality of life over baseline were reported for both treatment groups. Improvements in these outcomes were significantly greater in the RFA group than the surgery group in the peri-operative period but not in later follow-up. Follow-up in these trials was inadequate to evaluate longer term recurrence for either treatment. Patient satisfaction was reported to be high for both treatments but was higher for RFA.
The studies comparing endovascular treatment approaches for VV (RFA and ELT) were more limited. Three RCT studies compared RFA (two with the second generation catheter) with ELT but mainly focused on peri-procedural outcomes such as pain, complications and recovery. Vein ablation rates were not evaluated in the trials, except for one small trial involving bilateral VV. Pain responses in patients undergoing ablation were extremely variable and up to 2 weeks, mean pain levels were significantly less with RFA than ELT ablation but differences were not significant at one month. Recovery, evaluated as return to usual activity or return to work, however, was similar in the treatment groups. Vein symptom and QOL improvements were improved in both groups but were significantly better in the RFA group than the ELT group at 2 weeks, but not at one month. Vein ablation rates were evaluated in several controlled clinical studies comparing the treatments between centers or within centers between individuals or over time. Comparisons in these studies were inconsistent with vein ablation rates for RFA reported to be similar to, higher than and lower than those with ELT.
Economic Analysis
RFA and surgical vein stripping, the main comparator reimbursed by the public system, are comparable in clinical benefits. Hence a cost-analysis was conducted to identify the differences in resources and costs between both procedures and a budgetary impact analysis (BIA) was conducted to project costs over a 5- year period in the province of Ontario. The target population of this economic analysis was patients with symptomatic varicose veins and the primary analytic perspective was that of the Ministry of Health and Long-Term Care.
The average case cost (based on Ontario hospital costs and medical resources) for surgical vein stripping was estimated to be $1,799. In order to calculate a procedural cost for RFA it was assumed that the hospital cost and physician labour fees, excluding anaesthesia and surgical assistance, were the same as vein stripping surgery. The manufacturer also provided details on the generator with a capital cost of $27,500 and a lifespan of 5 years and the disposables (catheter, sheath, guidewire) with a cost of $673 per case. The average case cost for RFA was therefore estimated to be $1,356. One-way sensitivity analysis was also conducted with hospital cost of RFA varied to 60% that of vein stripping surgery (average cost per case = $627.08) to calculate an impact to the province.
Historical volumes of vein stripping surgeries in Ontario were used to project surgeries in a linear fashion up to five years into the future. Volumes for RFA and ELT were calculated based on share capture from the surgery market based on discussion with clinical expert opinion and existing private data based on discussion with the manufacturer. RFA is expected to compete with ELT and capture some of the market. If ELT is reimbursed by the public sector then numbers will continue to increase from previous private data and share capture from the conventional surgical treatment market. Therefore, RFA cases will also increase since it will be capturing a share of the ELT market. A budget impact to the province was then calculated by multiplying volumes by the cost of the procedure.
RFA is comparable in clinical benefits to vein stripping surgery. It has the extra upfront cost of the generator and cost per case for disposables but does not require an operating theater, anaesthetist or surgical assistant fees. The impact to the province is expected to be 5 M by Year 5 with the introduction of new ELT and RFA image guided endovascular technologies and existing surgery for varicose veins.
Conclusion
The conclusions on the comparative outcomes between endovascular RFA and surgical ligation and saphenous vein stripping and between endovascular RFA and laser ablation for VV treatment are summarized in the table below (ES Table 1).
Outcome comparisons of RFA vs. surgery and RFA vs ELT for varicose veins
ELT refers to endovascular laser ablation; RFA, radiofrequency ablation
The outcomes of the evidence-based review on these treatments for VV based on different perspectives are summarized below:
RFA First versus Second Generation Catheters and Segmental Ablation
Ablation with second generation catheters and segmental ablation offered technical advantages with improved ease and significant decreases in procedure time. RFA ablation with second generation catheters is also no longer restricted to smaller (< 12 mm diameter) saphenous veins. The safety profile with the new device and method of energy delivery is as good as or improved over the first generation device. No major adverse events were reported in two multicenter prospective cohort studies in 6 month follow-up with over 500 patients. Post-operative complications such as bruising and pain were significantly less with RFA ablation with second generation catheters than ELT in two RCT trials.RFA treatment with second generation catheters has ablation rates that are higher than with first generation catheters and are more comparable with the consistently high rates of ELT.
Endovascular RFA versus Surgery
RFA has a quicker recovery attributable to decreased pain and lower minor complications.RFA, in the short term was comparable to surgery in treatment effectiveness as assessed by imaging defined anatomic outcomes such as vein closure, flow or reflux. Other treatment outcomes such as symptomatic relief and HRQOL were significantly improved in both groups and between group differences in the early peri-operative period were likely influenced by pain experiences. Longer term follow-up was inadequate to evaluate recurrence after either treatment.Patient satisfaction was high after both treatments but was higher for RFA than surgery.
Endovascular RFA versus ELT
RFA has significantly less post-operative pain than ELT but differences were not significant when pain was adjusted for analgesic use and pain differences between groups did not persist at 1 month follow-up.Treatment effectiveness, measured as symptom relief and QOL improvement were similar between the endovascular treatments in the short term (within 1 month) Treatment effectiveness measured as imaging defined vein ablation was not measured in any RCT trials (only for bilateral VV disease) and results were inconsistently reported in observational trials.Longer term follow-up was not available to assess recurrence after either treatment.
System Outcomes – RFA Replacing Surgery or Competing with ELT
RFA may offer system advantages in that the treatment can be offered by several medical specialties in outpatient settings and because it does not require an operating theatre or general anaesthesia. The treatment may result in decanting of patients from OR with decreased pre-surgical investigations, demand on anaesthetists’ time, hospital stay and wait time for VV treatment. It may also provide more reliable outpatient scheduling. Procedure costs may be less for endovascular approaches than surgery but the budget impact may be greater with insurance of RFA because of the transfer of cases from the private market to the public payer system.Competition between RFA and ELT endovascular approaches is likely to continue to stimulate innovation and technical changes to advance patient care and result in competitive pricing.
PMCID: PMC3377553  PMID: 23074413
18.  Phase 1 Study of Two Merozoite Surface Protein 1 (MSP142) Vaccines for Plasmodium falciparum Malaria 
PLoS Clinical Trials  2007;2(4):e12.
Objectives:
To assess the safety and immunogenicity of two vaccines, MSP142-FVO/Alhydrogel and MSP142-3D7/Alhydrogel, targeting blood-stage Plasmodium falciparum parasites.
Design:
A Phase 1 open-label, dose-escalating study.
Setting:
Quintiles Phase 1 Services, Lenexa, Kansas between July 2004 and November 2005.
Participants:
Sixty healthy malaria-naïve volunteers 18–48 y of age.
Interventions:
The C-terminal 42-kDa region of merozoite surface protein 1 (MSP142) corresponding to the two allelic forms present in FVO and 3D7 P. falciparum lines were expressed in Escherichia coli, refolded, purified, and formulated on Alhydrogel (aluminum hydroxide). For each vaccine, volunteers in each of three dose cohorts (5, 20, and 80 μg) were vaccinated at 0, 28, and 180 d. Volunteers were followed for 1 y.
Outcome Measures:
The safety of MSP142-FVO/Alhydrogel and MSP142-3D7/Alhydrogel was assessed. The antibody response to each vaccine was measured by reactivity to homologous and heterologous MSP142, MSP119, and MSP133 recombinant proteins and recognition of FVO and 3D7 parasites.
Results:
Anti-MSP142 antibodies were detected by ELISA in 20/27 (74%) and 22/27 (81%) volunteers receiving three vaccinations of MSP142-FVO/Alhydrogel or MSP142-3D7/Alhydrogel, respectively. Regardless of the vaccine, the antibodies were cross-reactive to both MSP142-FVO and MSP142-3D7 proteins. The majority of the antibody response targeted the C-terminal 19-kDa domain of MSP142, although low-level antibodies to the N-terminal 33-kDa domain of MSP142 were also detected. Immunofluorescence microscopy of sera from the volunteers demonstrated reactivity with both FVO and 3D7 P. falciparum schizonts and free merozoites. Minimal in vitro growth inhibition of FVO or 3D7 parasites by purified IgG from the sera of the vaccinees was observed.
Conclusions:
The MSP142/Alhydrogel vaccines were safe and well tolerated but not sufficiently immunogenic to generate a biologic effect in vitro. Addition of immunostimulants to the Alhydrogel formulation to elicit higher vaccine-induced responses in humans may be required for an effective vaccine.
Editorial Commentary
Background: Generally, adults living in parts of the world where malaria is common develop protective immunity against the parasite. This means they may get infected but not become ill as a result. However, there are individuals, such as pregnant women and children under the age of five, who are more likely to develop symptoms of malaria due to no (or reduced) natural immunity. A successful malaria vaccine would stimulate an individual's immune system to respond to the malaria parasite and prevent serious clinical disease. Many different groups are currently developing potential vaccines. Several candidates are based on a protein called MSP1 (merozoite surface protein 1) which is found on the surface of the blood-stage form of the malaria parasite. However, in nature parasites carry different versions of the MSP1 protein, and ideally a successful vaccine would bring about immune responses against these different versions. The researchers carrying out this trial wanted to compare the safety and immune responses against candidate vaccines representing two different MSP1 proteins, which covered many different parasite lines. As a phase 1 trial, the study was carried out in healthy adult volunteers. Sixty individuals were assigned to receive an injection of the vaccines, either containing a recombinant protein analogous to the FVO parasite line (termed MSP142-FVO) or the 3D7 parasite line (termed MSP142-3D7) at three different dose levels. The trial's primary objective was to assess safety, which was done by collecting data on any abnormal signs or symptoms up to 14 d after each of three vaccinations. These outcomes were graded and then defined as related to the vaccine or not. The researchers also looked at antibody levels in participants' blood against different variants of the MSP1 protein, as well as using in vitro tests to see whether antibodies from vaccinated individuals could prevent malaria parasites from growing in lab culture.
What the trial shows: The safety outcomes of the trial showed that the most common type of side effect experienced by the volunteers was pain at the injection site. The vast majority of such events were graded as mild, although there was one single case of a severe event (high levels of pain experienced by one volunteer at the injection site). There was no significant association between the chance of side effects and the vaccine dosage that an individual received. Following vaccination, antibody levels against the protein on which the vaccine was based were detected, although these levels dropped over time. The researchers did not see a strong association between the vaccine dosage that individuals received and the level of antibody response. However, the two vaccines when compared seemed to be equally good at raising an immune response and both caused antibodies to be raised corresponding to different variants of the MSP1 protein. However, the antibodies raised did not seem to be particularly effective at preventing malaria parasites from growing in lab culture.
Strengths and limitations: Strengths of this study include a comparison of three different dosage levels of the vaccines under study, as well as a comparison of two vaccines based on the same protein, representing different parasite lines. Limitations to the study include the small number of participants, which makes the trial underpowered to detect all but large differences in side effects between the groups being compared. A placebo arm was not included in the trial, so it is not possible to be sure that the numbers of side effects observed here can be attributed to the vaccines or not. Finally, the procedure for assigning individuals to the two different vaccines involved alternation, rather than true randomization, which could have minimized the risk of bias.
Contribution to the evidence: The trial reported here is an essential step in vaccine development. The results provide the first evidence relating to safety for these two vaccines, and do not raise any safety concerns at this stage. Although the vaccines raised an immune response, the antibodies raised did not seem to have much of an effect on malaria parasites in vitro. While these vaccines are safe, alternative MSP1 vaccine formulations anticipated to bring about a greater immune response will likely be studied before proceeding to field studies.
doi:10.1371/journal.pctr.0020012
PMCID: PMC1847697  PMID: 17415408
19.  Covariate adjustment for two-sample treatment comparisons in randomized clinical trials: A principled yet flexible approach 
Statistics in medicine  2008;27(23):4658-4677.
SUMMARY
There is considerable debate regarding whether and how covariate adjusted analyses should be used in the comparison of treatments in randomized clinical trials. Substantial baseline covariate information is routinely collected in such trials, and one goal of adjustment is to exploit covariates associated with outcome to increase precision of estimation of the treatment effect. However, concerns are routinely raised over the potential for bias when the covariates used are selected post hoc; and the potential for adjustment based on a model of the relationship between outcome, covariates, and treatment to invite a “fishing expedition” for that leading to the most dramatic effect estimate. By appealing to the theory of semiparametrics, we are led naturally to a characterization of all treatment effect estimators and to principled, practically-feasible methods for covariate adjustment that yield the desired gains in efficiency and that allow covariate relationships to be identified and exploited while circumventing the usual concerns. The methods and strategies for their implementation in practice are presented. Simulation studies and an application to data from an HIV clinical trial demonstrate the performance of the techniques relative to existing methods.
doi:10.1002/sim.3113
PMCID: PMC2562926  PMID: 17960577
baseline variables; clinical trials; covariate adjustment; efficiency; semiparametric theory; variable selection
20.  Improving efficiency of inferences in randomized clinical trials using auxiliary covariates 
Biometrics  2008;64(3):707-715.
Summary
The primary goal of a randomized clinical trial is to make comparisons among two or more treatments. For example, in a two-arm trial with continuous response, the focus may be on the difference in treatment means; with more than two treatments, the comparison may be based on pairwise differences. With binary outcomes, pairwise odds-ratios or log-odds ratios may be used. In general, comparisons may be based on meaningful parameters in a relevant statistical model. Standard analyses for estimation and testing in this context typically are based on the data collected on response and treatment assignment only. In many trials, auxiliary baseline covariate information may also be available, and it is of interest to exploit these data to improve the efficiency of inferences. Taking a semiparametric theory perspective, we propose a broadly-applicable approach to adjustment for auxiliary covariates to achieve more efficient estimators and tests for treatment parameters in the analysis of randomized clinical trials. Simulations and applications demonstrate the performance of the methods.
doi:10.1111/j.1541-0420.2007.00976.x
PMCID: PMC2574960  PMID: 18190618
Covariate adjustment; Hypothesis test; k-arm trial; Kruskal-Wallis test; Log-odds ratio; Longitudinal data; Semiparametric theory
21.  Statistical properties on semiparametric regression for evaluating pathway effects 
Most statistical methods for microarray data analysis consider one gene at a time, and they may miss subtle changes at the single gene level. This limitation may be overcome by considering a set of genes simultaneously where the gene sets are derived from prior biological knowledge. We call a pathway as a predefined set of genes that serve a particular cellular or physiological function. Limited work has been done in the regression settings to study the effects of clinical covariates and expression levels of genes in a pathway on a continuous clinical outcome. A semiparametric regression approach for identifying pathways related to a continuous outcome was proposed by Liu et al. (2007), who demonstrated the connection between a least squares kernel machine for nonparametric pathway effect and a restricted maximum likelihood (REML) for variance components. However, the asymptotic properties on a semiparametric regression for identifying pathway have never been studied. In this paper, we study the asymptotic properties of the parameter estimates on semiparametric regression and compare Liu et al.’s REML with our REML obtained from a profile likelihood. We prove that both approaches provide consistent estimators, have n convergence rate under regularity conditions, and have either an asymptotically normal distribution or a mixture of normal distributions. However, the estimators based on our REML obtained from a profile likelihood have a theoretically smaller mean squared error than those of Liu et al.’s REML. Simulation study supports this theoretical result. A profile restricted likelihood ratio test is also provided for the non-standard testing problem. We apply our approach to a type II diabetes data set (Mootha et al., 2003).
PMCID: PMC3763850  PMID: 24014933
Gaussian random process; Kernel machine; Mixed model; Pathway analysis; Profile likelihood; Restricted maximum likelihood
22.  Semiparametric estimation of the average causal effect of treatment on an outcome measured after a postrandomization event, with missing outcome data 
In the past decade, several principal stratification–based statistical methods have been developed for testing and estimation of a treatment effect on an outcome measured after a postrandomization event. Two examples are the evaluation of the effect of a cancer treatment on quality of life in subjects who remain alive and the evaluation of the effect of an HIV vaccine on viral load in subjects who acquire HIV infection. However, in general the developed methods have not addressed the issue of missing outcome data, and hence their validity relies on a missing completely at random (MCAR) assumption. Because in many applications the MCAR assumption is untenable, while a missing at random (MAR) assumption is defensible, we extend the semiparametric likelihood sensitivity analysis approach of Gilbert and others (2003) and Jemiai and Rotnitzky (2005) to allow the outcome to be MAR. We combine these methods with the robust likelihood–based method of Little and An (2004) for handling MAR data to provide semiparametric estimation of the average causal effect of treatment on the outcome. The new method, which does not require a monotonicity assumption, is evaluated in a simulation study and is applied to data from the first HIV vaccine efficacy trial.
doi:10.1093/biostatistics/kxp034
PMCID: PMC2800161  PMID: 19815692
Causal inference; HIV vaccine trial; Missing at random; Posttreatment selection bias; Principal stratification; Sensitivity analysis
23.  Intra-Articular Viscosupplementation With Hylan G-F 20 To Treat Osteoarthritis of the Knee 
Executive Summary
Objective
To assess the effectiveness and cost-effectiveness of hylan G-F 20 as a substitute for existing treatments for pain due to osteoarthritis (OA) of the knee, other viscosupplementation devices, and/or as an adjunct to conventional therapy.
Hylan G-F 20 (brand name Synvisc, which is manufactured by Genzyme) is a high molecular weight derivative of hyaluronan, a component of joint synovial fluid. It acts as a lubricant and shock absorber. It is administered by injection into the joint space to treat pain associated with OA of the knee. Although the injection procedure is an insured service in Ontario, the device, hylan G-F 20, is not.
Clinical Need
Osteoarthritis is prevalent in 10% to 12% of Ontario adults, and exceeds 40% in Ontario residents aged 65 years and older. About one-half of these people have mild, moderate, or severe OA of the knee. Conventional treatment involves a combination of nonpharmacological management (e.g., weight loss, exercise, social support, and patient education), drugs, (e.g., acetaminophen, COX-2 inhibitors, nonsteroidal anti-inflammatory drugs with/without misoprostol, intra-articular glucocorticoids, opioids, and topical analgesics) and surgical interventions, such as debridement and total knee replacement, when pharmacological management fails.
The growing burden of OA of the knee in the aging Ontario population combined with recent safety concerns about COX-2 inhibitors and long wait times for total joint replacement is placing pressure on the demand for new, effective technologies to manage the pain of OA.
The Technology
Hylan G-F 20 is derived from rooster comb hyaluronan (HA). At the time of writing, eight viscosupplement hyaluronic products are licensed in Canada. Hylan G-F 20 is distinguished from the other products by its chemical structure (i.e., cross-linked hyaluronan, hence hylan) and relatively higher molecular weight, which may bestow greater therapeutic viscoelastic properties. A complete treatment cycle of hylan G-F 20 involves an intra-articular injection of 2 ml of hylan G-F 20 once a week for 3 weeks. It is licensed for use for patients in all stages of joint pathology, but should not be used in infected or severely inflamed joints, in joints with large effusion, in patients that have skin diseases or infections in the area of the injection site, or in patients with venous stasis. It is also contraindicated in patients with hypersensitivities to avian proteins.
Review Strategy
The Medical Advisory Secretariat used its standard search protocol to review the literature for evidence on the effectiveness of intra-articular hylan G-F 20 compared with placebo, as a substitute for alternate active treatments, or as an adjunct to conventional care for treatment of the pain of OA of the knee. All English-language journal articles and reviews with clearly described designs and methods (i.e., those sufficient to assign a Jadad score to) published or released between 1966 and February 2005 were included. Two more recently published meta-analyses were also included. The databases searched were Ovid MEDLINE, EMBASE, the Cochrane database and leading international organizations for health technology assessments, including the International Network of Agencies for Health Technology Assessments. The search terms were as follows: hyaluronan, hyaluronate adj sodium, hylan, hylan G-F 20 (Synvisc), Synvisc, Hyalgan, Orthovisc, Supartz, Artz, Artzal, BioHY, NASHA, NRD101, viscosupplementation, osteoarthritis, knee, knee joint. The primary outcome of interest was a clinically significant difference, defined as greater than 10 mm on 100 mm visual analogue scale, or a change from baseline of more than 20% in the mean magnitude of pain relief experienced among patients treated with hylan G-F 20 compared with those treated with the control intervention.
One clinical epidemiologist reviewed the full-text reports and extracted data using an extraction form. Key variables included, but were not limited to, the characteristics of the patients, method of randomization, type of control intervention, outcome measures for effectiveness and safety, and length of follow-up. The quality of the studies and level of the evidence was initially scored by one clinical epidemiologist using the Jadad scale and GRADE approach. Level of quality depends on the amount of certainty about the magnitude of effect and is based on study designs, extent of methodological limitations, consistency of results and applicability (i.e. directness) to the Ontario clinical context. The GRADE approach also permits comment on the strength of recommendations resulting from the evidence, based on estimates of the magnitude of effect relative to the magnitude of risk and burden and the level of certainty around these estimates. The quality assessments were subsequently peer-reviewed.
Summary of Findings
The literature search revealed 2 previous health technology assessments, 3 meta-analyses of placebo-controlled trials, 1 Cochrane review and meta-analysis encompassing 18 randomized controlled trials (RCTs) that compared hylan G-F 20 to either placebo or active treatments, 11 RCTs of hylan G-F 20 (all included in the Cochrane review), and 10 observational studies. Given the preponderance of evidence, the Medical Advisory Secretariat’s analysis focused on studies with Level 1 evidence of effectiveness (i.e., the meta-analyses of RCTs and the RCTs). Only safety data from the observational studies were included.
The authors of the 2 health technology assessments concluded that the data were sparse and poor quality. There was some evidence that hylan G-F 20 delivered a small, clinical benefit at 3 to 6 months after treatment on a magnitude comparable to NSAIDs and intra-articular steroids. Hylan G-F 20 appeared to carry a risk of a local adverse reaction of in the range of 3% to 18% per 100 injections, but there was no apparent risk of a severe adverse event, although the data were limited.
Each of the 3 meta-analyses of placebo-controlled trials of intra-articular hyaluronans had only 3 trials involving hylan G-F 20. There results were inconsistent, with one study concluding that intra-articular hyaluronans were efficacious, whereas the 2 other analyses concluded the effect size was small (0.32) and probably not clinically significant. The risk of a minor adverse event ranged from 8% to 19% per 100 injections. Major adverse events were rare.
The authors of the Cochrane review concluded that a pooled analysis supported the efficacy of hyaluronans, including hylan G-F 20. The 5- to 13-week post-injection period showed an improvement from baseline of 11% to 54% for pain and 9% to 15% for function. Comparable efficacy was noted against NSAIDs, and longer-term benefits were noted in against steroids. Few adverse events were noted.
When the Medical Advisory Secretariat applied the criterion of clinical significance to the magnitude of pain relief reported in the RCTs on hylan G-F 20, the following was noted:
There was inconsistent evidence that hylan G-F 20 was clinically superior to placebo at 5 to 26 weeks after treatment.
There was consistent evidence that, in terms of delivering pain relief, hylan G-F 20 was no better or worse than NSAIDs or intra-articular steroids at 5 to 26 weeks after treatment.
There was consistent evidence that hylan G-F 20 was not clinically superior to other hyaluronic products.
There was consistent evidence that hylan G-F 20 delivered a small magnitude of clinical benefit at 12 to 52 weeks post-injection when administered as an adjunct to conventional care.
There were limitations to the methods in many of the RCTs involving hylan G-F 20. When only the results from the higher-quality studies were considered, there was level 2 evidence that hylan G-F 20 was not clinically superior to placebo (or another hyaluronan) at 1 to 26 weeks after treatment in older patients with advanced disease for whom total knee replacement was indicated. There was level 2 evidence that hylan G-F 2- was comparable to NSAIDs at 4 to 13 weeks after treatment, and level 2 evidence that hylan G-F 20 was superior to placebo as an adjunct to conventional care 4 to 26 weeks after treatment.
With respect to safety, overall, hylan G-F 20 carries a risk of a minor, local adverse event rate of about 8% to 19% per 100 injections. Incidents of moderate-severe post-injection inflammatory joint reactions have been reported, but the likelihood appears to be low (0.15% of patients).
Economic Analysis
Case-costing estimates suggest that the annual cost of 2 treatment cycles of hylan G-F 20 (plus analgesics for breakthrough pain) is almost equivalent to the annual cost of taking a NSAID (with a gastroprotective agent) and is more expensive that taking intra-articular corticosteroids (plus analgesics for breakthrough pain). The estimated cost of funding hylan G-F 20 as an adjunct to conventional therapy (i.e., any of analgesics, NSAIDs, intra-articular steroids, physiotherapy, and surgery) is $700 per patient per year. Given the huge burden of mild to moderate OA among adults who seek medical care for it in Ontario (about 300,000), funding hylan G-F 20 as an adjunct to existing treatment could be expensive, depending on its diffusion and uptake. If only 10% to 30% of patients choose this option, then the estimated budget impact would be $21 million to $63 million (Cdn) per year.
Conclusions
When the benefits relative to the risks and costs are considered, NSAIDs and hylan G-F 20 appear comparable, as the table shows. Consequently, there’s little evidence on which to recommend hylan G-F 20 over NSAIDs, except perhaps for patients who cannot tolerate NSAIDs, although this evidence is indirect, since no studies looked specifically at this population.
CC indicates conventional care; IA, intra-articular; NSAID, nonsteroidal anti-inflammatory drug.
Intra-articular steroids appear to deliver the same risks and clinical benefits as hylan G-F 20 at a lower cost; therefore, there’s evidence that intra-articular steroids are the preferred option. Hylan G-F 20 as an adjunct to conventional care appears to deliver some clinical benefit, although funding hylan G-F 20 as an adjunct would have considerable budget impact, so the benefits of this option do not clearly outweigh the costs. There’s some uncertainty about the effect of hylan G-F 20 relative to other hyaluronans, mostly because some of the trials of this comparison were not published.
Many of the studies of hylan G-F 20 have considerable methodological limitations that result in uncertainty about the magnitude of effect. An upcoming review of the evidence by the Osteoarthritis Advisory Panel of clinical experts will likely help to reduce some of this uncertainty.
There is moderate evidence that hylan G-F 20 is no more clinically effective than NSAIDs. The evidence that hylan G-F 20 might be an appropriate option for a person with OA of the knee who cannot tolerate NSAIDs is indirect. The possible benefit of fewer cases of NSAID-induced gastropathy in this population must be weighed against the uncertainty of a severe inflammatory adverse reaction to hylan G-F 20.
Similarly, there is moderate evidence that hylan G-F 20 is no more clinically effective than intra-articular corticosteroids. The lower cost of intra-articular corticosteroids makes them the preferred option.
There is moderate evidence that hylan G-F 20 is effective as an adjunct to conventional care, delivering a small magnitude of temporary relief at 4 to 26 weeks after treatment. The estimated additional cost to the system of providing hylan G-F 20 as an adjunct to conventional care is about $700 (Cdn) per patient annually. The magnitude and duration of clinical benefit of hylan G-F 20 must be weighed against the uncertainty and potential magnitude of the budget impact (about $35 million to $105 million (Cdn) per year) of funding this device given the high burden of OA in Ontario adults.
There is level 2 evidence that hylan G-F 20 is not effective in people with advanced OA for whom total knee replacement is indicated.
PMCID: PMC3382385  PMID: 23074461
24.  Optimizing parameters in clinical trials with a randomized start or withdrawal design 
Disease-modifying (DM) trials on chronic diseases such as Alzheimer’s disease (AD) require a randomized start or withdrawal design. The analysis and optimization of such trials remain poorly understood, even for the simplest scenario in which only three repeated efficacy assessments are planned for each subject: one at the baseline, one at the end of the trial, and the other at the time when the treatments are switched. Under the assumption that the repeated measures across subjects follow a trivariate distribution whose mean and covariance matrix exist, the DM efficacy hypothesis is formulated by comparing the change of efficacy outcome between treatment arms with and without a treatment switch. Using a minimax criterion, a methodology is developed to optimally determine the sample size allocations to individual treatment arms as well as the optimum time when treatments are switched. The sensitivity of the optimum designs with respect to various model parameters is further assessed. An intersection-union test (IUT) is proposed to test the DM hypothesis, and determine the asymptotic size and the power of the IUT. Finally, the proposed methodology is demonstrated by using reported statistics on the placebo arms from several recently published symptomatic trials on AD to estimate necessary parameters and then deriving the optimum sample sizes and the time of treatment switch for future DM trials on AD.
doi:10.1016/j.csda.2013.07.013
PMCID: PMC3804275  PMID: 24159249
Alzheimer’s disease; Disease-modifying trials; Intersection-union test; Minimax criterion; Random intercept and slope models; Randomized start design
25.  Semiparametric estimation exploiting covariate independence in two-phase randomized trials 
Biometrics  2008;65(1):178-187.
SUMMARY
Recent results for case-control sampling suggest when the covariate distribution is constrained by gene-environment independence, semiparametric estimation exploiting such independence yields a great deal of efficiency gain. We consider the efficient estimation of the treatment-biomarker interaction in two-phase sampling nested within randomized clinical trials, incorporating the independence between a randomized treatment and the baseline markers. We develop a Newton–Raphson algorithm based on the profile likelihood to compute the semiparametric maximum likelihood estimate (SPMLE). Our algorithm accommodates both continuous phase-one outcomes and continuous phase-two biomarkers. The profile information matrix is computed explicitly via numerical differentiation. In certain situations where computing the SPMLE is slow, we propose a maximum estimated likelihood estimator (MELE), which is also capable of incorporating the covariate independence. This estimated likelihood approach uses a one-step empirical covariate distribution, thus is straightforward to maximize. It offers a closed-form variance estimate with limited increase in variance relative to the fully efficient SPMLE. Our results suggest exploiting the covariate independence in two-phase sampling increases the efficiency substantially, particularly for estimating treatment-biomarker interactions.
doi:10.1111/j.1541-0420.2008.01046.x
PMCID: PMC2892338  PMID: 18479485
case-only estimator; estimated likelihood; gene-environment independence; Newton–Raphson algorithm; profile likelihood; treatment-biomarker interactions

Results 1-25 (1161840)