Sample Size and Power When Designing a Randomized Trial for the Estimation of Treatment, Selection, and Preference Effects

BACKGROUND

A 2-stage randomized trial design, incorporating participant choice, provides unbiased estimates of the effects of the treatment or intervention (treatment effect), the difference between outcomes for participants who prefer one treatment compared with another (selection effect), and the interaction between participants' preferences for treatment and the treatment actually received (preference effect). It is important to ensure that such trials are adequately powered to estimate these effects.

SAMPLE SIZE FORMULAS

This paper presents methods for determining the required sample sizes for estimating treatment, selection, and preference effects. We demonstrate the changes in sample size as various key parameters are changed. In general, approximately twice as many participants (in total) are needed to have equivalent power for detecting both treatment and selection/preference effects compared with a trial of the treatment effect alone.

PRIMARY SCREENING EXAMPLE

We illustrate their application for the design of a primary screening trial comparing human papillomavirus DNA testing versus cervical screening (by Pap smear). Our example would require 520 participants to have 80% power to detect moderate-sized preference and selection effects and a small to moderate treatment effect.

CONCLUSIONS

With the growing interest in understanding treatment choices and with the use of decision aids, well-designed and adequately powered 2-stage randomized trial designs offer the opportunity to determine the effects of participants' preferences. Our sample size formulas will help future studies ensure that they have adequate power to detect selection and preference effects.

Presence of other allergic disease modifies the effect of early childhood traffic-related air pollution exposure on asthma prevalence

Nitrogen dioxide (NO2), a surrogate measure of traffic-related air pollution (TRAP), has been associated with incident childhood asthma. Timing of exposure and atopic status may be important effect modifiers. We collected cross-sectional data on asthma outcomes from Toronto school children aged 5-9years in 2006. Lifetime home, school and daycare addresses were obtained to derive birth and cumulative NO2 exposures for a nested case-control subset of 1497 children. Presence of other allergic disease (a proxy for atopy) was defined as self-report of one or more of doctor-diagnosed rhinitis, eczema, or food allergy. Generalized estimating equations were used to adjust for potential confounders, and examine hypothesized effect modifiers while accounting for clustering by school. In children with other allergic disease, birth, cumulative and 2006 NO2 were associated with lifetime asthma (OR 1.46, 95% CI 1.08-1.98; 1.37, 95% CI 1.00-1.86; and 1.60, 95% CI 1.09-2.36 respectively per interquartile range increase) and wheeze (OR 1.44, 95% CI 1.10-1.89; 1.31, 95% CI 1.02-1.67; and 1.60, 95% CI 1.16-2.21). No or weaker effects were seen in those without allergic disease, and effect modification was amplified when a more restrictive algorithm was used to define other allergic disease (at least 2 of doctor diagnosed allergic rhinitis, eczema or food allergy). The effects of modest NO2 levels on childhood asthma were modified by the presence of other allergic disease, suggesting a probable role for allergic sensitization in the pathogenesis of TRAP initiated asthma.

Serum 25-hydroxyvitamin d level and influenza vaccine immunogenicity in children and adolescents

Background

Vaccination is an important strategy in the prevention of influenza, but immunologic response to vaccination can vary widely. Recent studies have shown an association between serum 25-hydroxyvitamin D (25[OH]D) levels and immune function. The purpose of this study was to determine if serum 25(OH)D level correlates with influenza vaccine immunogenicity in children and adolescents.

Methods

We conducted a prospective cohort study of children age 3 to 15 years of age vaccinated with trivalent influenza vaccine (A/Brisbane/59/2007[H1N1]-like virus, A/Brisbane/10/2007 [H3N2]-like virus and B/Florida/4/2006-like virus) in Hutterite communities in Alberta, Saskatchewan and Manitoba. Serum 25(OH)D levels were measured at baseline and immunogenicity was assessed using hemagluttination inhibition (HAI) titers done at baseline and 3–5 weeks post vaccination. Logistic regression was used to assess the relationship between serum 25(OH)D level as both a continuous and dichotomous variable and seroprotection, seroconversion, fold increase in geometric mean titer (GMT) and post vaccination titer.

Results

A total of 391 children and adolescents were included in the study and 221 (57% had post-vaccination HAI titers. The median serum 25(OH)D level was 61.0 nmol/L (Interquartile range [IQR] 50.0, 71.0). No relationship was found between serum 25(OH)D level and seroprotection (post-vaccination titer ≥40 and ≥320) or seroconversion (post-vaccination titer ≥40 for participants with pre-vaccine titer <10 or four-fold rise in post-vaccination titer for those with a pre-vaccine titer ≥10).

Conclusion

Serum 25(OH)D level was not associated with influenza vaccine immunogenicity in otherwise healthy children and adolescents. Other strategies to enhance influenza vaccine response should continue to be evaluated in this population.

The role of serum 25(OH)D level in vaccine responsiveness in other populations, especially those hyporesponsive to influenza vaccination, requires further study.

How to report and interpret screening test properties: guidelines for driving researchers

One important goal of driving research is the development of a short but valid office-based screening test for fitness to drive of aging drivers. Several candidate tests have been proposed already, and no doubt others will be proposed in the future. It might seem obvious that authors advocating for the adoption of a particular screening test or procedure should report sensitivity, specificity, and other common screening test properties. Unfortunately, driving researchers have frequently failed to report any screening test properties. Others have reported screening test properties but have made basic mistakes such as calculating predictive values of positive and negative tests but reporting them incorrectly as sensitivity and specificity. These omissions and errors suggest that some driving researchers may be unaware of the importance of accurately reporting test properties when proposing a screening procedure and that others may need a refresher on how to calculate and interpret the most common screening test properties. Many good learning resources for screening and diagnostic tests are available, but most of them are intended for students and researchers in medicine, epidemiology, or public health. We hope that this tutorial in a prominent transportation journal will help lead to improved reporting and interpretation of screening test properties in articles that assess the usefulness of potential screening tools for fitness to drive.

Factors explaining the heterogeneity of effects of patient decision aids on knowledge of outcome probabilities: a systematic review sub-analysis

BACKGROUND

There is considerable unexplained heterogeneity in previous meta-analyses of randomized controlled trials (RCTs) evaluating the effects of patient decision aids on the accuracy of knowledge of outcome probabilities. The purpose of this review was to explore possible effect modification by three covariates: the type of control intervention, decision aid quality and patients' baseline knowledge of probabilities.

METHODS

A sub-analysis of studies previously identified in the 2011 Cochrane review on decision aids for people facing treatment and screening decisions was conducted. Additional unpublished data were requested from relevant study authors to maximize the number of eligible studies. RCTs (to 2009) comparing decision aids with standardized probability information to control interventions (lacking such information) and assessing the accuracy of patient knowledge of outcome probabilities were included. The proportions of patients with accurate knowledge of outcome probabilities in each group were converted into relative effect measures. Intervention quality was assessed using the International Patient Decision Aid Standards instrument (IPDASi) probabilities domain.

RESULTS

A main effects analysis of 17 eligible studies confirmed that decision aids significantly improve the accuracy of patient knowledge of outcome probabilities (relative risk = 1.80 [1.51, 2.16]), with considerable heterogeneity (87%). The type of control did not modify effects. Meta-regression suggested that the IPDASi probabilities domain score (reflecting decision aid quality) is a potential effect modifier (P = 0.037), accounting for a quarter of the variability (R² = 0.28). Meta-regression indicated the control event rate (reflecting baseline knowledge) is a significant effect modifier (P = 0.001), with over half the variability in ln(OR) explained by the linear relationship with log-odds for the control group (R² = 0.52); this relationship was slightly strengthened after correcting for the statistical dependence of the effect measure on the control event rate.

CONCLUSIONS

Patients' baseline level of knowledge of outcome probabilities is an important variable that explains the heterogeneity of effects of decision aids on improving accuracy of this knowledge. Greater relative effects are observed when the baseline proportion of patients with accurate knowledge is lower. This may indicate that decision aids are more effective in populations with lower knowledge.

Initiation and continuation of randomized trials after the publication of a trial stopped early for benefit asking the same study question: STOPIT-3 study design

BACKGROUND

Randomized control trials (RCTs) stopped early for benefit (truncated RCTs) are increasingly common and, on average, overestimate the relative magnitude of benefit by approximately 30%. Investigators stop trials early when they consider it is no longer ethical to enroll patients in a control group. The goal of this systematic review is to determine how investigators of ongoing or planned RCTs respond to the publication of a truncated RCT addressing a similar question.

METHODS/DESIGN

We will conduct systematic reviews to update the searches of 210 truncated RCTs to identify similar trials ongoing at the time of publication, or started subsequently, to the truncated trials ('subsequent RCTs'). Reviewers will determine in duplicate the similarity between the truncated and subsequent trials. We will analyze the epidemiology, distribution, and predictors of subsequent RCTs. We will also contact authors of subsequent trials to determine reasons for beginning, continuing, or prematurely discontinuing their own trials, and the extent to which they rely on the estimates from truncated trials.

DISCUSSION

To the extent that investigators begin or continue subsequent trials they implicitly disagree with the decision to stop the truncated RCT because of an ethical mandate to administer the experimental treatment. The results of this study will help guide future decisions about when to stop RCTs early for benefit.

Predictors of mortality in patients with an implantable cardiac defibrillator: a systematic review and meta-analysis

BACKGROUND

Many current predictors of mortality in heart failure (HF) were evaluated before the use of implantable cardioverter defibrillators (ICDs). We conducted a meta-analysis to identify factors associated with mortality in ICD-HF patients.

METHODS

We searched in MedLine, EMBASE, and CINAHL in May 2012. Two reviewers selected citations that included ambulatory ICD patients and addressed the association between any predictor and mortality using multivariable regression. We meta-analyzed mortality using random-effects models.

RESULTS

Of 10,420 studies reviewed, 72 studies evaluating 63 predictors on 257,692 ICD patients proved eligible. High confidence in estimates was found for age (hazard ratio [HR], 1.45 for 10-year increase; 95% confidence interval [CI], 1.35-1.56), baseline glomerular filtration rate (HR, 1.25 for 15-mL/min decrease; 95% CI, 1.15-1.35), chronic obstructive pulmonary disease (HR, 1.54; 95% CI, 1.38-1.71), diabetes (HR, 1.56; 95% CI, 1.37-1.79), peripheral vascular disease (HR, 1.43; 95% CI, 1.2-1.72), left ventricular ejection fraction (HR, 0.77 for 10% increase; 95% CI, 0.73-0.83), and appropriate or inappropriate ICD shocks (HR, 2.34; 95% CI 1.59-3.44) New York Heart Association class, atrial fibrillation, and congestive HF were strongly associated with mortality but the confidence in estimates was low. Ischemic cardiomyopathy and male sex were not independent predictors of mortality.

CONCLUSIONS

This meta-analysis identified strong reliable mortality predictors in ICD-HF patients. Age, renal dysfunction, chronic obstructive pulmonary disease, diabetes, peripheral vascular disease, decreased left ventricular ejection fraction, and ICD shocks during follow-up were strong predictors of mortality; ischemic cardiomyopathy and male sex were not. Further research is needed to study other potential predictors, particularly biomarkers.

Changes in circulating progenitor cells are associated with outcome in heart failure patients: a longitudinal study

BACKGROUND

Circulating progenitor cells (CPCs) are involved in the process of endothelial repair and are a prognostic factor in cardiovascular diseases. We evaluated the association between serial measurements of CPCs and functional capacity and outcomes in heart failure (HF).

METHODS

We included 156 consecutive consenting ambulatory HF patients (left ventricular ejection fraction < 40%). We evaluated CPCs and functional capacity (peak VO2) every 6 months for up to 2 years. CPCs were measured as early-outgrowth colony-forming units (EO-CFUs) and circulating CD34+, VEGFR2+ and/or CD133+ cells. We recorded mortality, HF hospital admissions, transplant, and ventricular assist device.

RESULTS

The mean age was 55 ± 15 years. A decrease in CD34+VEGFR2+ cells was independently associated with increased functional capacity; a 10-cell decrease in CD34+VEGFR2+ cells was associated with an increase of 0.2 mL/kg/min in peak VO2 (P < 0.05). We found an interaction effect (P = 0.02) between EO-CFUs and diabetes: in patients without diabetes, a 10-EO-CFU increase was independently associated with increased peak VO2 of 0.28 mL/kg/min (P = 0.01), and in patients with diabetes, a decrease in EO-CFUs was associated with an increased peak VO2 (P < 0.05). Higher EO-CFUs were associated with reduced mortality (hazard ratio, 0.25; 95% confidence interval, 0.09-0.69).

CONCLUSIONS

We noted differential relations between CPCs and outcomes in patients with vs without diabetes. Higher EO-CFUs and lower CD34+VEGFR2+ cells were associated with improved functional capacity and reduced mortality in nondiabetic patients. In patients with diabetes, lower EO-CFUs were associated with improved functional capacity. The basis for these differences requires further examination.

Cognitive behavioral therapy age effects in child and adolescent anxiety: an individual patient data metaanalysis

BACKGROUND

Investigations of age effects on youth anxiety outcomes in randomized trials (RCTs) of cognitive behavior therapy (CBT) have failed to yield a clear result due to inadequate statistical power and methodologic weaknesses. We conducted an individual patient data metaanalysis to address this gap.

QUESTION

Does age moderate CBT effect size, measured by a clinically and statistically significant interaction between age and CBT exposure?

METHODS

All English language RCTs of CBT for anxiety in 6-19 year olds were identified using systematic review methods. Investigators of eligible trials were invited to submit their individual patient data. The anxiety disorder interview schedule (ADIS) primary diagnosis severity score was the primary outcome. Age effects were investigated using multilevel modeling to account for study level data clustering and random effects.

RESULTS

Data from 17 of 23 eligible trials were obtained (74%); 16 studies and 1,171 (78%) cases were available for the analysis. No interaction between age and CBT exposure was found in a model containing age, sex, ADIS baseline severity score, and comorbid depression diagnosis (power ≥ 80%). Sensitivity analyses, including modeling age as both a categorical and continuous variable, revealed this result was robust.

CONCLUSIONS

Adolescents who receive CBT in efficacy research studies show benefits comparable to younger children. However, CBT protocol modifications routinely carried out by expert trial therapists may explain these findings. Adolescent CBT protocols are needed to facilitate the transportability of efficacy research effects to usual care settings where therapists may have less opportunity for CBT training and expertise development.

Epidemiologic approaches to evaluating the potential for human papillomavirus type replacement postvaccination

Currently, 2 vaccines exist that prevent infection by the genotypes of human papillomavirus (HPV) responsible for approximately 70% of cervical cancer cases worldwide. Although vaccination is expected to reduce the prevalence of these HPV types, there is concern about the effect this could have on the distribution of other oncogenic types. According to basic ecological principles, if competition exists between ≥2 different HPV types for niche occupation during natural infection, elimination of 1 type may lead to an increase in other type(s). Here, we discuss this issue of "type replacement" and present different epidemiologic approaches for evaluation of HPV type competition. Briefly, these approaches involve: 1) calculation of the expected frequency of coinfection under independence between HPV types for comparison with observed frequency; 2) construction of hierarchical logistic regression models for each vaccine-targeted type; and 3) construction of Kaplan-Meier curves and Cox models to evaluate sequential acquisition and clearance of HPV types according to baseline HPV status. We also discuss a related issue concerning diagnostic artifacts arising when multiple HPV types are present in specific samples (due to the inability of broad-spectrum assays to detect certain types present in lower concentrations). This may result in an apparent increase in previously undetected types postvaccination.

Association of psychotherapy with disability benefit claim closure among patients disabled due to depression

Background

Depression is the most frequent reason for receiving disability benefits in North America, and treatment with psychotherapy is often funded by private insurers. No studies have explored the association between the provision of psychotherapy for depression and time to claim closure.

Methods

Using administrative data from a Canadian disability insurer, we evaluated the association between the provision of psychotherapy and short-term disability (STD) and long-term disability (LTD) claim closure by performing Cox proportional hazards regression.

Results

We analyzed 10,508 STD and 10,338 LTD claims for depression. In our adjusted analyses, receipt of psychotherapy was associated with longer time to STD closure (HR [99% CI] = 0.81 [0.68 to 0.97]) and faster LTD claim closure (1.42 [1.33 to 1.52]). In both STD and LTD, older age (0.90 [0.88 to 0.92] and 0.83 [0.80 to 0.85]), per decade), a primary diagnosis of recurrent depression versus non-recurrent major depression (0.78 [0.69 to 0.87] and 0.80 [0.72 to 0.89]), a psychological secondary diagnosis (0.90 [0.84 to 0.97] and 0.66 [0.61 to 0.71]), or a non-psychological secondary diagnosis (0.81 [0.73 to 0.90] and 0.77 [0.71 to 0.83]) versus no secondary diagnosis, and an administrative services only policy ([0.94 [0.88 to 1.00] and 0.87 [0.75 to 0.996]) or refund policy (0.86 [0.80 to 0.92] and 0.73 [0.68 to 0.78]) compared to non-refund policy claims were independently associated with longer time to STD claim closure.

Conclusions

We found, paradoxically, that receipt of psychotherapy was independently associated with longer time to STD claim closure and faster LTD claim closure in patients with depression. We also found multiple factors that were predictive of time to both STD and LTD claim closure. Our study has limitations, and well-designed prospective studies are needed to establish the effect of psychotherapy on disabling depression.

Addressing continuous data for participants excluded from trial analysis: a guide for systematic reviewers

BACKGROUND

No methods directly address the impact of missing participant data for continuous outcomes in systematic reviews on risk of bias.

METHODS

We conducted a consultative, iterative process to develop a framework for handling missing participant data for continuous outcomes. We considered sources reflecting real observed outcomes in participants followed-up in individual trials included in the systematic review, and developed a range of plausible strategies. We applied our approach to two systematic reviews.

RESULTS

We used five sources of data for imputing the means for participants with missing data. To impute standard deviation (SD), we used the median SD from the control arms of all included trials. Using these sources, we developed four progressively more stringent imputation strategies. In the first example review, effect estimates diminished and lost significance as strategies became more stringent, suggesting rating down confidence in estimates of effect for risk of bias. In the second, effect estimates maintained statistical significance using even the most stringent strategy, suggesting missing data does not undermine confidence in results.

CONCLUSIONS

Our approach provides a useful, reasonable, and relatively simple, quantitative guidance for judging the impact of risk of bias as a result of missing participant data in systematic reviews of continuous outcomes.

Low serum 25-hydroxyvitamin D level and risk of upper respiratory tract infection in children and adolescents

Background. Vitamin D may be important for immune function. Studies to date have shown an inconsistent association between vitamin D and infection with respiratory viruses. The purpose of this study was to determine if serum 25-hydroxyvitamin D (25(OH)D) was associated with laboratory-confirmed viral respiratory tract infections (RTIs) in children.

Methods. Serum 25(OH)D levels were measured at baseline and children from Canadian Hutterite communities were followed prospectively during the respiratory virus season. Nasopharyngeal specimens were obtained if symptoms developed and infections were confirmed using polymerase chain reaction. The association between serum 25(OH)D and time to laboratory-confirmed viral RTI was evaluated using a Cox proportional hazards model.

Results. Seven hundred forty-three children aged 3–15 years were followed between 22 December 2008 and 23 June 2009. The median serum 25(OH)D level was 62.0 nmol/L (interquartile range, 51.0–74.0). A total of 229 participants (31%) developed at least 1 laboratory-confirmed viral RTI. Younger age and lower serum 25(OH)D levels were associated with increased risk of viral RTI. Serum 25(OH)D levels <75 nmol/L increased the risk of viral RTI by 50% (hazard ratio [HR], 1.51; 95% confidence interval [CI], 1.10–2.07, P = .011) and levels <50 nmol/L increased the risk by 70% (HR, 1.67; 95% CI, 1.16–2.40, P = .006).

Conclusions. Lower serum 25(OH)D levels were associated with increased risk of laboratory-confirmed viral RTI in children from Canadian Hutterite communities. Interventional studies evaluating the role of vitamin D supplementation to reduce the burden of viral RTIs are warranted.

The GRADE approach is reproducible in assessing the quality of evidence of quantitative evidence syntheses

OBJECTIVE

We evaluated the inter-rater reliability (IRR) of assessing the quality of evidence (QoE) using the Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) approach.

STUDY DESIGN AND SETTING

On completing two training exercises, participants worked independently as individual raters to assess the QoE of 16 outcomes. After recording their initial impression using a global rating, raters graded the QoE following the GRADE approach. Subsequently, randomly paired raters submitted a consensus rating.

RESULTS

The IRR without using the GRADE approach for two individual raters was 0.31 (95% confidence interval [95% CI] = 0.21-0.42) among Health Research Methodology students (n = 10) and 0.27 (95% CI = 0.19-0.37) among the GRADE working group members (n = 15). The corresponding IRR of the GRADE approach in assessing the QoE was significantly higher, that is, 0.66 (95% CI = 0.56-0.75) and 0.72 (95% CI = 0.61-0.79), respectively. The IRR further increased for three (0.80 [95% CI = 0.73-0.86] and 0.74 [95% CI = 0.65-0.81]) or four raters (0.84 [95% CI = 0.78-0.89] and 0.79 [95% CI = 0.71-0.85]). The IRR did not improve when QoE was assessed through a consensus rating.

CONCLUSION

Our findings suggest that trained individuals using the GRADE approach improves reliability in comparison to intuitive judgments about the QoE and that two individual raters can reliably assess the QoE using the GRADE system.

Randomized trials published in higher vs. lower impact journals differ in design, conduct, and analysis

OBJECTIVE

To compare methodological characteristics of randomized controlled trials (RCTs) published in higher vs. lower impact Core Clinical Journals.

STUDY DESIGN AND SETTING

We searched MEDLINE for RCTs published in 2007 in Core Clinical Journals. We randomly sampled 1,140 study reports in a 1:1 ratio in higher (five general medicine journals with the highest total citations in 2007) and lower impact journals.

RESULTS

Four hundred sixty-nine RCTs proved eligible: 219 in higher and 250 in lower impact journals. RCTs in higher vs. lower impact journals had larger sample sizes (median, 285 vs. 39), were more likely to receive industry funding (53% vs. 28%), declare concealment of allocation (66% vs. 36%), declare blinding of health care providers (53% vs. 41%) and outcome adjudicators (72% vs. 54%), report a patient-important primary outcome (69% vs. 50%), report subgroup analyses (64% vs. 26%), prespecify subgroup hypotheses (42% vs. 20%), and report a test for interaction (54% vs. 27%); P < 0.05 for all differences.

CONCLUSION

RCTs published in higher impact journals were more likely to report methodological safeguards against bias and patient-important outcomes than those published in lower impact journals. However, sufficient limitations remain such that publication in a higher impact journal does not ensure low risk of bias.

High-frequency oscillation in early acute respiratory distress syndrome

BACKGROUND

Previous trials suggesting that high-frequency oscillatory ventilation (HFOV) reduced mortality among adults with the acute respiratory distress syndrome (ARDS) were limited by the use of outdated comparator ventilation strategies and small sample sizes.

METHODS

In a multicenter, randomized, controlled trial conducted at 39 intensive care units in five countries, we randomly assigned adults with new-onset, moderate-to-severe ARDS to HFOV targeting lung recruitment or to a control ventilation strategy targeting lung recruitment with the use of low tidal volumes and high positive end-expiratory pressure. The primary outcome was the rate of in-hospital death from any cause.

RESULTS

On the recommendation of the data monitoring committee, we stopped the trial after 548 of a planned 1200 patients had undergone randomization. The two study groups were well matched at baseline. The HFOV group underwent HFOV for a median of 3 days (interquartile range, 2 to 8); in addition, 34 of 273 patients (12%) in the control group received HFOV for refractory hypoxemia. In-hospital mortality was 47% in the HFOV group, as compared with 35% in the control group (relative risk of death with HFOV, 1.33; 95% confidence interval, 1.09 to 1.64; P=0.005). This finding was independent of baseline abnormalities in oxygenation or respiratory compliance. Patients in the HFOV group received higher doses of midazolam than did patients in the control group (199 mg per day [interquartile range, 100 to 382] vs. 141 mg per day [interquartile range, 68 to 240], P<0.001), and more patients in the HFOV group than in the control group received neuromuscular blockers (83% vs. 68%, P<0.001). In addition, more patients in the HFOV group received vasoactive drugs (91% vs. 84%, P=0.01) and received them for a longer period than did patients in the control group (5 days vs. 3 days, P=0.01).

CONCLUSIONS

In adults with moderate-to-severe ARDS, early application of HFOV, as compared with a ventilation strategy of low tidal volume and high positive end-expiratory pressure, does not reduce, and may increase, in-hospital mortality. (Funded by the Canadian Institutes of Health Research; Current Controlled Trials numbers, ISRCTN42992782 and ISRCTN87124254, and ClinicalTrials.gov numbers, NCT00474656 and NCT01506401.).

Accuracy of p53 codon 72 polymorphism status determined by multiple laboratory methods: a latent class model analysis

Introduction

Studies on the association of a polymorphism in codon 72 of the p53 tumour suppressor gene (rs1042522) with cervical neoplasia have inconsistent results. While several methods for genotyping p53 exist, they vary in accuracy and are often discrepant.

Methods

We used latent class models (LCM) to examine the accuracy of six methods for p53 determination, all conducted by the same laboratory. We also examined the association of p53 with cytological cervical abnormalities, recognising potential test inaccuracy.

Results

Pairwise disagreement between laboratory methods occurred approximately 10% of the time. Given the estimated true p53 status of each woman, we found that each laboratory method is most likely to classify a woman to her correct status. Arg/Arg women had the highest risk of squamous intraepithelial lesions (SIL). Test accuracy was independent of cytology. There was no strong evidence for correlations of test errors.

Discussion

Empirical analyses ignore possible laboratory errors, and so are inherently biased, but test accuracy estimated by the LCM approach is unbiased when model assumptions are met. LCM analysis avoids ambiguities arising from empirical test discrepancies, obviating the need to regard any of the methods as a “gold” standard measurement. The methods we presented here to analyse the p53 data can be applied in many other situations where multiple tests exist, but where none of them is a gold standard.

Predictors of seizure outcomes in children with tuberous sclerosis complex and intractable epilepsy undergoing resective epilepsy surgery: an individual participant data meta-analysis

Objective

To perform a systematic review and individual participant data meta-analysis to identify preoperative factors associated with a good seizure outcome in children with Tuberous Sclerosis Complex undergoing resective epilepsy surgery.

Data Sources

Electronic databases (MEDLINE, EMBASE, CINAHL and Web of Science), archives of major epilepsy and neurosurgery meetings, and bibliographies of relevant articles, with no language or date restrictions.

Study Selection

We included case-control or cohort studies of consecutive participants undergoing resective epilepsy surgery that reported seizure outcomes. We performed title and abstract and full text screening independently and in duplicate. We resolved disagreements through discussion.

Data Extraction

One author performed data extraction which was verified by a second author using predefined data fields including study quality assessment using a risk of bias instrument we developed. We recorded all preoperative factors that may plausibly predict seizure outcomes.

Data Synthesis

To identify predictors of a good seizure outcome (i.e. Engel Class I or II) we used logistic regression adjusting for length of follow-up for each preoperative variable.

Results

Of 9863 citations, 20 articles reporting on 181 participants were eligible. Good seizure outcomes were observed in 126 (69%) participants (Engel Class I: 102(56%); Engel class II: 24(13%)). In univariable analyses, absence of generalized seizure semiology (OR = 3.1, 95%CI = 1.2–8.2, p = 0.022), no or mild developmental delay (OR = 7.3, 95%CI = 2.1–24.7, p = 0.001), unifocal ictal scalp electroencephalographic (EEG) abnormality (OR = 3.2, 95%CI = 1.4–7.6, p = 0.008) and EEG/Magnetic resonance imaging concordance (OR = 4.9, 95%CI = 1.8–13.5, p = 0.002) were associated with a good postoperative seizure outcome.

Conclusions

Small retrospective cohort studies are inherently prone to bias, some of which are overcome using individual participant data. The best available evidence suggests four preoperative factors predictive of good seizure outcomes following resective epilepsy surgery. Large long-term prospective multicenter observational studies are required to further evaluate the risk factors identified in this review.

Vitamin D deficiency and depression in adults: systematic review and meta-analysis

BACKGROUND

There is conflicting evidence about the relationship between vitamin D deficiency and depression, and a systematic assessment of the literature has not been available.

AIMS

To determine the relationship, if any, between vitamin D deficiency and depression.

METHOD

A systematic review and meta-analysis of observational studies and randomised controlled trials was conducted.

RESULTS

One case-control study, ten cross-sectional studies and three cohort studies with a total of 31 424 participants were analysed. Lower vitamin D levels were found in people with depression compared with controls (SMD = 0.60, 95% CI 0.23-0.97) and there was an increased odds ratio of depression for the lowest v. highest vitamin D categories in the cross-sectional studies (OR = 1.31, 95% CI 1.0-1.71). The cohort studies showed a significantly increased hazard ratio of depression for the lowest v. highest vitamin D categories (HR = 2.21, 95% CI 1.40-3.49).

CONCLUSIONS

Our analyses are consistent with the hypothesis that low vitamin D concentration is associated with depression, and highlight the need for randomised controlled trials of vitamin D for the prevention and treatment of depression to determine whether this association is causal.

Adjudication of bleeding outcomes in an international thromboprophylaxis trial in critical illness

INTRODUCTION

Measuring bleeding in critical care trials is challenging. We determined the reliability of adjudicated bleeding assessments in a large thromboprophylaxis trial in the intensive care unit (ICU).

MATERIALS AND METHODS

PROphylaxis for ThromboEmbolism in Critical Care Trial (PROTECT) was an international randomized controlled trial that compared dalteparin to unfractionated heparin for the prevention of deep vein thrombosis in the ICU. Daily bleeding data were collected prospectively using a validated tool. Bleeds were adjudicated in duplicate by 2 of 4 members comprising a central adjudication committee. Bleeds were stratified by severity and study drug, then randomly assigned to adjudicator pairs. Adjudicators were blinded to treatment allocation, study centre and peer-assessments. We calculated agreement on bleeding severity and examined the effect of adjudication on overall trial results.

RESULTS

In PROTECT, 491 patients had bleeding events including 208 with major bleeding and 283 with minor bleeding only. Of 491 patients, 446 were adjudicated in duplicate: 182 with major, 250 with minor and 14 with no bleeding. After adjudication, 52 of 244 bleeds were downgraded to minor; whereas only 15 of 244 were upgraded to major. Overall agreement among adjudicators was excellent (crude agreement=86.3%; kappa=0.76). Hazard ratios for major or any bleeding with dalteparin or unfractionated heparin were similar when analyzed using non-adjudicated events.

CONCLUSIONS

Major bleeds were sometimes over-called by research coordinators in a large ICU thromboprohylaxis trial. Adjudicator agreement was excellent. Central adjudication allowed reliable bleeding assessment and enhanced the rigor and validity of this major safety outcome.

Longitudinal measurements of oxygen consumption in growing infants during the first weeks after birth: old data revisited

BACKGROUND

In a study conducted in 1966-1969, longitudinal measurements were made of the metabolic rate in growing infants. Statistical methods for analyzing longitudinal data weren't readily accessible at that time.

OBJECTIVES

To measure minimal rates of oxygen consumption (V·O2, ml/min) in growing infants during the first postnatal weeks and to determine the relationships between postnatal increases in V·O2, body size and postnatal age.

METHODS

We studied 61 infants of any birth weight or gestational age, including 19 of very low birth weight. The infants, nursed in incubators, were clinically well and without need of oxygen supplementation or respiratory assistance. Serial measures of V·O2 using a closed-circuit method were obtained at approximately weekly intervals. V·O2 was measured under thermoneutral conditions with the infant asleep or resting quietly. Data were analyzed using mixed-effects models.

RESULTS

During early postnatal growth, V·O2 rises as surface area (m(2))(1.94) (standard error, SE 0.054) or body weight (kg)(1.24) (SE 0.033). Multivariate analyses show statistically significant effects of both size and age. Reference intervals (RIs) for V·O2 for fixed values of body weight and postnatal age are presented. As V·O2 rises with increasing size and age, there is an increase in the skin-operative environmental temperature gradient (T skin-op) required for heat loss. Required T skin-op can be predicted from surface area and heat loss (heat production minus heat storage).

CONCLUSIONS

Generation of RIs for minimal rates of V·O2 in growing infants from the 1960s was enabled by application of mixed-effects statistical models for analyses of longitudinal data. Results apply to the precaffeine era of neonatal care.

Effectiveness of cognitive behavioral therapy for depression in patients receiving disability benefits: a systematic review and individual patient data meta-analysis

Objectives

To systematically summarize the randomized trial evidence regarding the relative effectiveness of cognitive behavioural therapy (CBT) in patients with depression in receipt of disability benefits in comparison to those not receiving disability benefits.

Data Sources

All relevant RCTs from a database of randomized controlled and comparative studies examining the effects of psychotherapy for adult depression (http://www.evidencebasedpsychotherapies.org), electronic databases (MEDLINE, EMBASE, PSYCINFO, AMED, CINAHL and CENTRAL) to June 2011, and bibliographies of all relevant articles.

Study Eligibility Criteria, Participants and Intervention

Adult patients with major depression, randomly assigned to CBT versus minimal/no treatment or care-as-usual.

Study Appraisal and Synthesis Methods

Three teams of reviewers, independently and in duplicate, completed title and abstract screening, full text review and data extraction. We performed an individual patient data meta-analysis to summarize data.

Results

Of 92 eligible trials, 70 provided author contact information; of these 56 (80%) were successfully contacted to establish if they captured receipt of benefits as a baseline characteristic; 8 recorded benefit status, and 3 enrolled some patients in receipt of benefits, of which 2 provided individual patient data. Including both patients receiving and not receiving disability benefits, 2 trials (227 patients) suggested a possible reduction in depression with CBT, as measured by the Beck Depression Inventory, mean difference [MD] (95% confidence interval [CI]) = −2.61 (−5.28, 0.07), p = 0.06; minimally important difference of 5. The effect appeared larger, though not significantly, in those in receipt of benefits (34 patients) versus not receiving benefits (193 patients); MD (95% CI) = −4.46 (−12.21, 3.30), p = 0.26.

Conclusions

Our data does not support the hypothesis that CBT has smaller effects in depressed patients receiving disability benefits versus other patients. Given that the confidence interval is wide, a decreased effect is still possible, though if the difference exists, it is likely to be small.

Biofeedback dialysis for hypotension and hypervolemia: a systematic review and meta-analysis

BACKGROUND

Intradialytic hypotension (IDH) is associated with morbidity and mortality. We conducted a systematic review to determine whether biofeedback hemodialysis (HD) can improve IDH and other outcomes, compared with HD without biofeedback.

METHODS

Data sources included the Cochrane Central Register of Controlled Trials, MEDLINE, EMBASE and ISI Web of Science. We included randomized trials that enrolled adult patients (>18 years) with IDH or extracellular fluid expansion and that used biofeedback to guide ultrafiltration and/or dialysate conductivity. Two authors assessed trial quality and independently extracted data in duplicate. We assessed heterogeneity using I(2). We applied the GRADE framework for rating the quality of evidence.

RESULTS

We found two parallel-arm randomized controlled clinical trials and six randomized crossover trials meeting inclusion criteria. All trials were open-label and at least four were industry-sponsored. Studies were small (median n = 27). No study evaluated hospitalization and the evidence for effect on mortality was of very low quality. Three studies assessed quality of life (QoL); none demonstrated benefit or harm, and quality of evidence was very low. Biofeedback significantly reduced IDH (risk ratio 0.61, 95% confidence interval 0.44-0.86; I(2)= 0%). Quality of evidence for this outcome was low due to risk of bias and potential publication bias.

CONCLUSIONS

Biofeedback dialysis significantly reduces the frequency of IDH. Large and well-designed randomized trials are needed to assess the effects on survival, hospitalization and QoL.