A basic introduction to fixed-effect and random-effects models for meta-analysis

Meta-analysis and subgroups

Subgroup analysis is the process of comparing a treatment effect for two or more variants of an intervention-to ask, for example, if an intervention's impact is affected by the setting (school versus community), by the delivery agent (outside facilitator versus regular classroom teacher), by the quality of delivery, or if the long-term effect differs from the short-term effect. While large-scale studies often employ subgroup analyses, these analyses cannot generally be performed for small-scale studies, since these typically include a homogeneous population and only one variant of the intervention. This limitation can be bypassed by using meta-analysis. Meta-analysis allows the researcher to compare the treatment effect in different subgroups, even if these subgroups appear in separate studies. We discuss several statistical issues related to this procedure, including the selection of a statistical model and statistical power for the comparison. To illustrate these points, we use the example of a meta-analysis of obesity prevention.

Meta-analysis of alcohol price and income elasticities - with corrections for publication bias

BACKGROUND

This paper contributes to the evidence-base on prices and alcohol use by presenting meta-analytic summaries of price and income elasticities for alcohol beverages. The analysis improves on previous meta-analyses by correcting for outliers and publication bias.

METHODS

Adjusting for outliers is important to avoid assigning too much weight to studies with very small standard errors or large effect sizes. Trimmed samples are used for this purpose. Correcting for publication bias is important to avoid giving too much weight to studies that reflect selection by investigators or others involved with publication processes. Cumulative meta-analysis is proposed as a method to avoid or reduce publication bias, resulting in more robust estimates. The literature search obtained 182 primary studies for aggregate alcohol consumption, which exceeds the database used in previous reviews and meta-analyses.

RESULTS

For individual beverages, corrected price elasticities are smaller (less elastic) by 28-29 percent compared with consensus averages frequently used for alcohol beverages. The average price and income elasticities are: beer, -0.30 and 0.50; wine, -0.45 and 1.00; and spirits, -0.55 and 1.00. For total alcohol, the price elasticity is -0.50 and the income elasticity is 0.60.

CONCLUSIONS

These new results imply that attempts to reduce alcohol consumption through price or tax increases will be less effective or more costly than previously claimed.

Physical activity interventions and depression in children and adolescents : a systematic review and meta-analysis

CONTEXT

Evidence suggests chronic physical activity (PA) participation may be both protective against the onset of and beneficial for reducing depressive symptoms.

OBJECTIVE

The aim of this article is to assess the impact of PA interventions on depression in children and adolescents using meta-analysis.

DATA SOURCES

Published English language studies were located from manual and computerized searches of the following databases: PsycInfo, The Cochrane Database of Systematic Reviews and The Cochrane Central Register of Controlled Trials, Trials Register of Promoting Health Interventions (TRoPHI; EPPI Centre), Web of Science and MEDLINE.

STUDY SELECTION

Studies meeting inclusion criteria (1) reported on interventions to promote or increase PA; (2) included children aged 5-11 years and/or adolescents aged 12-19 years; (3) reported on results using a quantitative measure of depression; (4) included a non-physical control or comparison group; and (5) were published in peer-reviewed journals written in English, up to and including May 2011 (when the search was conducted).

DATA EXTRACTION

Studies were coded for methodological, participant and study characteristics. Comprehensive Meta-Analysis version-2 software was used to compute effect sizes, with subgroup analyses to identify moderating characteristics. Study quality was assessed using the Delphi technique.

RESULTS

Nine studies were included (n = 581); most were school-based randomized controlled trials, randomized by individual. Studies used a variety of measurement tools to assess depressive symptoms. The summary treatment effect was small but significant (Hedges' g = -0.26, standard error = 0.09, 95% confidence intervals = -0.43, -0.08, p = 0.004). Subgroup analyses showed that methodological (e.g. studies with both education and PA intervention; those with a higher quality score; and less than 3 months in duration) and participant characteristics (e.g. single-gender studies; those targeting overweight or obese groups) contributed most to the reduction in depression.

CONCLUSIONS

There was a small significant overall effect for PA on depression. More outcome-focused, high-quality trials are required to effectively inform the implementation of programmes to reduce depressive symptoms in children and adolescents.

Progesterone luteal support after ovulation induction and intrauterine insemination: a systematic review and meta-analysis

OBJECTIVE

To evaluate the effect of luteal phase P support after ovulation induction IUI.

DESIGN

A systematic review and meta-analysis.

SETTING

Not applicable.

PATIENT(S)

Undergoing ovulation induction IUI.

INTERVENTION(S)

Any form of exogenous P in ovulation induction IUI cycles.

MAIN OUTCOME MEASURE(S)

Clinical pregnancy and live birth.

RESULT(S)

Five trials were identified that met inclusion criteria and comprised 1,298 patients undergoing 1,938 cycles. Clinical pregnancy (odds ratio [OR] 1.47, 95% confidence interval [CI] 1.15-1.98) and live birth (OR 2.11, 95% CI 1.21-3.67) were more likely in P-supplemented patients. These findings persisted in analyses evaluating per IUI cycle, per patient, and first cycle only data. In subgroup analysis, patients receiving gonadotropins for ovulation induction had the most increase in clinical pregnancy with P support (OR 1.77, 95% CI 1.20-2.6). Conversely, patients receiving clomiphene citrate (CC) for ovulation induction showed no difference in clinical pregnancy with P support (OR 0.89, 95% CI 0.47-1.67).

CONCLUSION(S)

Progesterone luteal phase support may be of benefit to patients undergoing ovulation induction with gonadotropins in IUI cycles. Progesterone support did not benefit patients undergoing ovulation induction with CC, suggesting a potential difference in endogenous luteal phase function depending on the method of ovulation induction.

Systematic review and meta-analysis of nonpharmacological interventions for fatigue in children and adolescents with cancer

BACKGROUND

Fatigue is one of the most distressing and prevalent symptoms reported by pediatric oncology patients. With the increase in cancer survival rates, medical teams have focused on methods that control cancer-related fatigue in children during the disease and its treatment in order to increase the quality of life for these patients.

AIM

The objective of this systematic review was to synthesize the best available evidence concerning the effectiveness of nonpharmacological interventions for fatigue in children and adolescents with cancer.

METHODS

The search strategy was designed to retrieve studies published between 1960 and 2010 in either English or Chinese. This review included randomized controlled trials and quasi-experimental studies. The studies that were selected for retrieval were assessed by two independent reviewers for methodological validity prior to inclusion in the review using standardized critical-appraisal instruments.

RESULTS

The review included six studies, and the meta-analysis revealed a statistically significant effect of exercise interventions in reducing general fatigue (effect size = -0.76; 95% CI [-1.35, -0.17]) in children and adolescents with cancer.

CONCLUSIONS AND IMPLICATIONS

The review provides an evidence-based guide to future priorities for clinical practice. Exercise interventions could reduce the levels of general fatigue in children aged 6-18 years. In particular, exercise interventions for fatigue are feasible and safe.

Meta-analysis of diagnostic test accuracy in neurosurgical practice

Comparative effectiveness research (CER) allows evidence to be evaluated on the effectiveness, benefits, and detriments of management options, diagnostic tests, or ways to deliver health care. This process can be achieved in different ways, such as with well-designed randomized controlled trials or by meta-analyses. Several medical subspecialties are increasingly using CER, but CER remains underused by the neurosurgical community. Meta-analysis is a highly accurate method that permits results from multiple well-designed research studies to be quantitatively compared. Meta-analysis can be performed in many settings, such as the evaluation of treatment or of a diagnostic test or prognostic factor. Meta-analyses of randomized controlled treatment trials are well known, but there is a paucity of papers describing the ways to perform a meta-analysis of a diagnostic test. The aim of this paper is to improve neurosurgeons' familiarity with the meta-analysis of diagnostic test accuracy by describing and detailing each stage leading to publication.

The impact of threat appeals on fear arousal and driver behavior: a meta-analysis of experimental research 1990-2011

The existing empirical research exploring the impact of threat appeals on driver behavior has reported inconsistent findings. In an effort to provide an up-to-date synthesis of the experimental findings, meta-analytic techniques were employed to examine the impact of threat-based messages on fear arousal and on lab-based indices of driving behavior. Experimental studies (k = 13, N = 3044), conducted between 1990 and 2011, were included in the analyses. The aims of the current analysis were (a) to examine whether or not the experimental manipulations had a significant impact on evoked fear, (b) to examine the impact of threat appeals on three distinct indices of driving, and (c) to identify moderators and mediators of the relationship between fear and driving outcomes. Large effects emerged for the level of fear evoked, with experimental groups reporting increased fear arousal in comparison to control groups (r = .64, n = 619, p<.01). The effect of threat appeals on driving outcomes, however, was not significant (r = .03, p = .17). This analysis of the experimental literature indicates that threat appeals can lead to increased fear arousal, but do not appear to have the desired impact on driving behavior. We discuss these findings in the context of threat-based road safety campaigns and future directions for experimental research in this area.

Sex bias in infectious disease epidemiology: patterns and processes

BACKGROUND

Infectious disease incidence is often male-biased. Two main hypotheses have been proposed to explain this observation. The physiological hypothesis (PH) emphasizes differences in sex hormones and genetic architecture, while the behavioral hypothesis (BH) stresses gender-related differences in exposure. Surprisingly, the population-level predictions of these hypotheses are yet to be thoroughly tested in humans.

METHODS AND FINDINGS

For ten major pathogens, we tested PH and BH predictions about incidence and exposure-prevalence patterns. Compulsory-notification records (Brazil, 2006-2009) were used to estimate age-stratified ♂:♀ incidence rate ratios for the general population and across selected sociological contrasts. Exposure-prevalence odds ratios were derived from 82 published surveys. We estimated summary effect-size measures using random-effects models; our analyses encompass ∼0.5 million cases of disease or exposure. We found that, after puberty, disease incidence is male-biased in cutaneous and visceral leishmaniasis, schistosomiasis, pulmonary tuberculosis, leptospirosis, meningococcal meningitis, and hepatitis A. Severe dengue is female-biased, and no clear pattern is evident for typhoid fever. In leprosy, milder tuberculoid forms are female-biased, whereas more severe lepromatous forms are male-biased. For most diseases, male bias emerges also during infancy, when behavior is unbiased but sex steroid levels transiently rise. Behavioral factors likely modulate male-female differences in some diseases (the leishmaniases, tuberculosis, leptospirosis, or schistosomiasis) and age classes; however, average exposure-prevalence is significantly sex-biased only for Schistosoma and Leptospira.

CONCLUSIONS

Our results closely match some key PH predictions and contradict some crucial BH predictions, suggesting that gender-specific behavior plays an overall secondary role in generating sex bias. Physiological differences, including the crosstalk between sex hormones and immune effectors, thus emerge as the main candidate drivers of gender differences in infectious disease susceptibility.

Use of a commercial enzyme-linked immunosorbent assay for rapid detection of Giardia duodenalis in dog stools in the environment

Giardia duodenalis is considered a potentially zoonotic protozoan. Some animal species, including infected dogs, may play an important role in the spread of Giardia cysts through environmental contamination with their feces. In the present study, a commercial enzyme-linked immunosorbent assay (ELISA) was used to examine 143 samples of dog feces collected in urban areas as an indicator of the risk of field contamination. Using a Bayesian statistical approach, the ELISA showed a sensitivity of 88.9% and a specificity of 95.8% with positive and negative predictive values of 89.6% and 95.4%, respectively. The test affords the advantage of rapid processing of fecal samples without a complex technical structure and extensive costly labor. Moreover, the present results show that the assay provides public health veterinarians with a practical tool that can be used in screening programs, as a valid alternative or as an adjunct to other tests, in order to assess the biological risk of exposure to G. duodenalis cysts from dogs in human settlements. However, the test may not be completely accurate for human health risk evaluation, as it does not discriminate between zoonotic and non-zoonotic isolates.

Individual participant data meta-analysis for a binary outcome: one-stage or two-stage?

BACKGROUND

A fundamental aspect of epidemiological studies concerns the estimation of factor-outcome associations to identify risk factors, prognostic factors and potential causal factors. Because reliable estimates for these associations are important, there is a growing interest in methods for combining the results from multiple studies in individual participant data meta-analyses (IPD-MA). When there is substantial heterogeneity across studies, various random-effects meta-analysis models are possible that employ a one-stage or two-stage method. These are generally thought to produce similar results, but empirical comparisons are few.

OBJECTIVE

We describe and compare several one- and two-stage random-effects IPD-MA methods for estimating factor-outcome associations from multiple risk-factor or predictor finding studies with a binary outcome. One-stage methods use the IPD of each study and meta-analyse using the exact binomial distribution, whereas two-stage methods reduce evidence to the aggregated level (e.g. odds ratios) and then meta-analyse assuming approximate normality. We compare the methods in an empirical dataset for unadjusted and adjusted risk-factor estimates.

RESULTS

Though often similar, on occasion the one-stage and two-stage methods provide different parameter estimates and different conclusions. For example, the effect of erythema and its statistical significance was different for a one-stage (OR = 1.35, [Formula: see text]) and univariate two-stage (OR = 1.55, [Formula: see text]). Estimation issues can also arise: two-stage models suffer unstable estimates when zero cell counts occur and one-stage models do not always converge.

CONCLUSION

When planning an IPD-MA, the choice and implementation (e.g. univariate or multivariate) of a one-stage or two-stage method should be prespecified in the protocol as occasionally they lead to different conclusions about which factors are associated with outcome. Though both approaches can suffer from estimation challenges, we recommend employing the one-stage method, as it uses a more exact statistical approach and accounts for parameter correlation.

Meta-analysis and subgroups

Subgroup analysis is the process of comparing a treatment effect for two or more variants of an intervention-to ask, for example, if an intervention's impact is affected by the setting (school versus community), by the delivery agent (outside facilitator versus regular classroom teacher), by the quality of delivery, or if the long-term effect differs from the short-term effect. While large-scale studies often employ subgroup analyses, these analyses cannot generally be performed for small-scale studies, since these typically include a homogeneous population and only one variant of the intervention. This limitation can be bypassed by using meta-analysis. Meta-analysis allows the researcher to compare the treatment effect in different subgroups, even if these subgroups appear in separate studies. We discuss several statistical issues related to this procedure, including the selection of a statistical model and statistical power for the comparison. To illustrate these points, we use the example of a meta-analysis of obesity prevention.

Comparative benefits of statins in the primary and secondary prevention of major coronary events and all-cause mortality: a network meta-analysis of placebo-controlled and active-comparator trials

BACKGROUND

The extent to which individual statins vary in terms of clinical outcomes across all populations, in addition to secondary and primary prevention has not been studied extensively in meta-analyses.

METHODS

We systematically studied 199,721 participants in 92 placebo-controlled and active-comparator trials comparing atorvastatin, fluvastatin, lovastatin, pravastatin, rosuvastatin, and simvastatin in participants with, or at risk of developing, cardiovascular disease. We performed pairwise and network meta-analyses for major coronary events and all-cause mortality outcomes, taking into account the dose differences across trials. Systematic review registration: PROSPERO 2011:CRD42011001470.

RESULTS

There were only a few trials that evaluated fluvastatin. Most frequent comparisons occurred between pravastatin and placebo, atorvastatin and placebo, and rosuvastatin and atorvastatin. No trial directly compared all six statins to each other. Across all populations, statins were significantly more effective than control in reducing all-cause mortality (OR 0.87, 95% credible interval 0.82-0.92) and major coronary events (OR 0.69, 95% CI 0.64-0.75). In terms of reducing major coronary events, atorvastatin (OR 0.66, 95% CI 0.48-0.94) and fluvastatin (OR 0.59, 95% CI 0.36-0.95) were significantly more effective than rosuvastatin at comparable doses. In participants with cardiovascular disease, statins significantly reduced deaths (OR 0.82, 95% CI 0.75-0.90) and major coronary events (OR 0.69, 95% CI 0.62-0.77). Atorvastatin was significantly more effective than pravastatin (OR 0.65, 95% CI 0.43-0.99) and simvastatin (OR 0.68, 95% CI 0.38-0.98) for secondary prevention of major coronary events. In primary prevention, statins significantly reduced deaths (OR 0.91, 95% CI 0.83-0.99) and major coronary events (OR 0.69, 95% CI 0.61-0.79) with no differences among individual statins. Across all populations, atorvastatin (80%), fluvastatin (79%), and simvastatin (62%) had the highest overall probability of being the best treatment in terms of both outcomes. Higher doses of atorvastatin and fluvastatin had the highest number of significant differences in preventing major coronary events compared with other statins. No significant heterogeneity or inconsistency was detected.

CONCLUSIONS

Statins significantly reduce the incidence of all-cause mortality and major coronary events as compared to control in both secondary and primary prevention. This analysis provides evidence for potential differences between individual statins, which are not fully explained by their low-density lipoprotein cholesterol-reducing effects. The observed differences between statins should be investigated in future prospective studies.

Do changes in coping style explain the effectiveness of interventions for psychological morbidity in family carers of people with dementia? A systematic review and meta-analysis

BACKGROUND

Observational studies find that family carers of people with dementia who use more emotional support and acceptance-based coping, and less dysfunctional coping, are less depressed and anxious. We hypothesized that interventions effective in reducing psychological symptoms would increase emotional support and acceptance-based coping, or decrease dysfunctional coping.

METHODS

We systematically reviewed randomized controlled trials published up to July 2011, of interventions for carers of people with dementia measuring coping and psychological morbidity. We rated study validity and reported findings. We conducted fixed-effect meta-analyses for interventions where possible.

RESULTS

Eight of 433 papers identified by the search met inclusion criteria. All measured coping immediately after intervention. Two interventions significantly decreased depressive or anxiety symptoms: the smaller study found no change in dysfunctional coping. Neither measured emotional support and acceptance-based coping. Meta-analysis found that both group coping skills interventions alone (SMD = -0.39, 95% CI = -0.75 to -0.03, p = 0.04) and with behavioral activation (SMD = -0.26, 95% CI = -0.48 to -0.04, p = 0.02) significantly increased dysfunctional coping, while significantly reducing depressive symptoms. Positive coping (a mix of emotional and solution-focused strategies) increased (SMD = 0.28, 95% CI = 0.05-0.51, p = 0.02) with group coping skills interventions and behavioral activation.

CONCLUSIONS

Contrary to our hypothesis, dysfunctional coping increased when carer depressive symptoms improved. There was preliminary evidence that emotional support and acceptance-based coping increased, as positive coping increased although solution-focused coping alone did not. More research is needed to elucidate whether successful interventions work through changing coping strategies immediately and in the longer term.

Prevalence of low back pain in children and adolescents: a meta-analysis

BACKGROUND

Low back pain (LBP) is common in children and adolescents, and it is becoming a public health concern. In recent years there has been a considerable increase in research studies that examine the prevalence of LBP in this population, but studies exhibit great variability in the prevalence rates reported. The purpose of this research was to examine, by means of a meta-analytic investigation, the prevalence rates of LBP in children and adolescents.

METHODS

Studies were located from computerized databases (ISI Web of Knowledge, MedLine, PEDro, IME, LILACS, and CINAHL) and other sources. The search period extended to April 2011. To be included in the meta-analysis, studies had to report a prevalence rate (whether point, period or lifetime prevalence) of LBP in children and/or adolescents (≤ 18 years old). Two independent researchers coded the moderator variables of the studies, and extracted the prevalence rates. Separate meta-analyses were carried out for the different types of prevalence in order to avoid dependence problems. In each meta-analysis, a random-effects model was assumed to carry out the statistical analyses.

RESULTS

A total of 59 articles fulfilled the selection criteria. The mean point prevalence obtained from 10 studies was 0.120 (95% CI: 0.09 and 0.159). The mean period prevalence at 12 months obtained from 13 studies was 0.336 (95% CI: 0.269 and 0.410), whereas the mean period prevalence at one week obtained from six studies was 0.177 (95% CI: 0.124 and 0.247). The mean lifetime prevalence obtained from 30 studies was 0.399 (95% CI: 0.342 and 0.459). Lifetime prevalence exhibited a positive, statistically significant relationship with the mean age of the participants in the samples and with the publication year of the studies.

CONCLUSIONS

The most recent studies showed higher prevalence rates than the oldest ones, and studies with a better methodology exhibited higher lifetime prevalence rates than studies that were methodologically poor. Future studies should report more information regarding the definition of LBP and there is a need to improve the methodological quality of studies.

Systematic review and meta-analysis of the efficacy and safety of perampanel in the treatment of partial-onset epilepsy

INTRODUCTION

Perampanel is a first-in-class antiepileptic drug approved for adjunctive treatment of partial-onset seizure in patients aged 12 years or older. Published randomised controlled trials (RCTs) had small sample sizes, and meta-analyses have included too few studies to draw conclusive results for the assessment of tolerability, efficacy and safety of perampanel. There is a need to conduct a meta-analysis with a larger dataset and an appropriate study design.

OBJECTIVE

The aim of this study was to systematically review the efficacy and safety of perampanel in the treatment of partial-onset epilepsy.

METHODS

Electronic and clinical trials databases were searched for RCTs of perampanel published up to March 2013. Outcomes of interest were 50 % responder rates, seizure freedom, treatment-emergent adverse events (TEAEs) and incidence of withdrawal. Meta-analysis was performed to investigate the outcomes of interest.

RESULTS

Five RCTs with a total of 1,678 subjects were included. The 50 % responder rates were significantly greater in patients receiving 4, 8 and 12 mg perampanel versus placebo, with risk ratios of 1.54 (95 % CI 1.11-2.13), 1.80 (95 % CI 1.38-2.35) and 1.72 (95 % CI 1.17-2.52), respectively. There was no statistical evidence of a difference in seizure freedom between 8 or 12 mg perampanel and placebo. Of the five commonly reported TEAEs included, both dizziness and somnolence were statistically associated with 8 mg perampanel, whilst dizziness was statistically associated with 12 mg perampanel. Incidences of withdrawal due to adverse events were significantly higher in the 8 mg and 12 mg perampanel groups versus placebo.

CONCLUSION

The use of perampanel resulted in a statistically significant reduction of seizure frequency with respect to the 50 % responder rate in patients with partial-onset epilepsy. Perampanel is well tolerated at 4 mg and reasonably tolerated at 8 and 12 mg. Further clinical and pharmacovigilance studies are required to investigate the long-term efficacy and safety of perampanel in the management of other types of epilepsy.

Meta-analysis of clinical trial safety data in a drug development program: answers to frequently asked questions

BACKGROUND

Meta-analyses of clinical trial safety data have risen in importance beyond regulatory submissions. During drug development, sponsors need to recognize safety signals early and adjust the development program accordingly, so as to facilitate the assessment of causality. Once a product is marketed, sponsors add postapproval clinical trial data to the body of information to help understand existing safety concerns or those that arise from other postapproval data sources, such as spontaneous reports.

PURPOSE

This article focuses on common questions encountered when designing and performing a meta-analysis of clinical trial safety data. Although far from an exhaustive set of questions, they touch on some basic and often misunderstood features of conducting such meta-analyses.

METHODS

The authors reviewed the current literature and used their combined experience with regulatory and other uses of meta-analysis to answer common questions that arise when performing meta-analyses of safety data.

RESULTS

We addressed the following topics: choice of studies to pool, effects of the method of ascertainment, use of patient-level data compared to trial-level data, the need (or not) for multiplicity adjustments, heterogeneity of effects and sources of it, and choice of fixed effects versus random effects.

LIMITATIONS

The list of topics is not exhaustive and the opinions offered represent only our perspective; we recognize that there may be other valid perspectives.

CONCLUSIONS

Meta-analysis can be a valuable tool for evaluating safety questions, but a number of methodological choices need to be made in designing and conducting any meta-analysis. This article provides advice on some of the more commonly encountered choices.

Evidence-based sample size estimation based upon an updated meta-regression analysis

A traditional meta-analysis examines the overall effectiveness of an intervention by producing a pooled estimate of treatment efficacy. In contrast to this, a meta-regression model seeks to determine whether a study-level covariate (X) is a plausible source of heterogeneity in a set of treatment effects. Upon performing such an analysis, the results may suggest the presence of a meaningful amount of variation in the treatment effects because of the covariate; however, the current set of trials may not provide sufficient statistical power for such a conclusion. The proposed approach provides quantitative insight into the amount of support that a new trial may provide to the hypothesis that X is a meaningful source of variation in an updated meta-regression model, which includes both the previously completed and the proposed trial. This empirical algorithm allows examination of the potential feasibility of a planned study of various sizes to further support or refute the hypothesis that X is a statistically significant source of variation. A detailed example illustrates the sample size estimation algorithm for both a planned individually or cluster randomized trial to investigate the now commonly accepted impact of geographical latitude on the observed effectiveness of the Bacillus Calmette-Guérin vaccine in the prevention of tuberculosis. Copyright © 2012 John Wiley & Sons, Ltd.

The relationship between sexual abuse and risky sexual behavior among adolescent boys: a meta-analysis

PURPOSE

Childhood and adolescent sexual abuse has been shown to lead to increased odds of sexual behaviors that lead to sexually transmitted infections and early pregnancy involvement. Research, meta-analyses, and interventions, however, have focused primarily on girls and young women who have experienced abuse, yet some adolescent boys are also sexually abused. We performed a meta-analysis of the existing studies to assess the magnitudes of the link between a history of sexual abuse and each of the three risky sexual behaviors among adolescent boys in North America.

METHODS

The three outcomes were (a) unprotected sexual intercourse, (b) multiple sexual partners, and (c) pregnancy involvement. Weighted mean effect sizes were computed from ten independent samples, from nine studies published between 1990 and 2011.

RESULTS

Sexually abused boys were significantly more likely than nonabused boys to report all three risky sexual behaviors. Weighted mean odds ratios were 1.91 for unprotected intercourse, 2.91 for multiple sexual partners, and 4.81 for pregnancy involvement.

CONCLUSIONS

Our results indicate that childhood and adolescent sexual abuse can substantially influence sexual behavior in adolescence among male survivors. To improve sexual health for all adolescents, even young men, we should strengthen sexual abuse prevention initiatives, raise awareness about male sexual abuse survivors' existence and sexual health issues, improve sexual health promotion for abused young men, and screen all people, regardless of gender, for a history of sexual abuse.

Soil-transmitted helminth reinfection after drug treatment: a systematic review and meta-analysis

Background

Soil-transmitted helminth (STH) infections (i.e., Ascaris lumbricoides, hookworm, and Trichuris trichiura) affect more than a billion people. Preventive chemotherapy (i.e., repeated administration of anthelmintic drugs to at-risk populations), is the mainstay of control. This strategy, however, does not prevent reinfection. We performed a systematic review and meta-analysis to assess patterns and dynamics of STH reinfection after drug treatment.

Methodology

We systematically searched PubMed, ISI Web of Science, EMBASE, Cochrane Database of Systematic Reviews, China National Knowledge Infrastructure, WanFang Database, Chinese Scientific Journal Database, and Google Scholar. Information on study year, country, sample size, age of participants, diagnostic method, drug administration strategy, prevalence and intensity of infection pre- and posttreatment, cure and egg reduction rate, evaluation period posttreatment, and adherence was extracted. Pooled risk ratios from random-effects models were used to assess the risk of STH reinfection after treatment. Our protocol is available on PROSPERO, registration number: CRD42011001678.

Principal Findings

From 154 studies identified, 51 were included and 24 provided STH infection rates pre- and posttreatment, whereas 42 reported determinants of predisposition to reinfection. At 3, 6, and 12 months posttreatment, A. lumbricoides prevalence reached 26% (95% confidence interval (CI): 16–43%), 68% (95% CI: 60–76%) and 94% (95% CI: 88–100%) of pretreatment levels, respectively. For T. trichiura, respective reinfection prevalence were 36% (95% CI: 28–47%), 67% (95% CI: 42–100%), and 82% (95% CI: 62–100%), and for hookworm, 30% (95% CI: 26–34%), 55% (95% CI: 34–87%), and 57% (95% CI: 49–67%). Prevalence and intensity of reinfection were positively correlated with pretreatment infection status.

Conclusion

STH reinfections occur rapidly after treatment, particularly for A. lumbricoides and T. trichiura. Hence, there is a need for frequent anthelmintic drug administrations to maximize the benefit of preventive chemotherapy. Integrated control approaches emphasizing health education and environmental sanitation are needed to interrupt transmission of STH.

Infections with soil-transmitted helminths (the roundworm Ascaris lumbricoides, the whipworm Trichuris trichiura, and hookworm) affect over 1 billion people, particularly rural communities in the developing world. The global strategy to control soil-transmitted helminth infections is ‘preventive chemotherapy’, which means large-scale administration of anthelmintic drugs to at-risk populations. However, because reinfection occurs after treatment, ‘preventive chemotherapy’ must be repeated regularly. Our systematic review and meta-analysis found that at 3, 6, and 12 months after treatment, A. lumbricoides prevalence reached 26% (95% confidence interval (CI): 16–43%), 68% (95% CI: 60–76%) and 94% (95% CI: 88–100%) of pretreatment levels, respectively. For T. trichiura, respective reinfection prevalence at these time points were 36% (95% CI: 28–47%), 67% (95% CI: 42–100%), and 82% (95% CI: 62–100%); and for hookworm, 30% (95% CI: 26–34%), 55% (95% CI: 34–87%), and 57% (95% CI: 49–67%). Prevalence and intensity of reinfection were positively correlated with pretreatment infection status. Our results suggest a frequent anthelmintic drug administration to maximize the benefit of preventive chemotherapy. Moreover, an integrated control strategy, consisting of preventive chemotherapy combined with health education and environmental sanitation is needed to interrupt transmission of soil-transmitted helminths.

Central venous access sites for the prevention of venous thrombosis, stenosis and infection

BACKGROUND

Central venous access (CVA) is widely used. However, its thrombotic, stenotic and infectious complications can be life-threatening and involve high-cost therapy. Research revealed that the risk of catheter-related complications varied according to the site of CVA. It would be helpful to find the preferred site of insertion to minimize the risk of catheter-related complications. This review was originally published in 2007 and was updated in 2011.

OBJECTIVES

1. Our primary objective was to establish whether the jugular, subclavian or femoral CVA routes resulted in a lower incidence of venous thrombosis, venous stenosis or infections related to CVA devices in adult patients.2. Our secondary objective was to assess whether the jugular, subclavian or femoral CVA routes influenced the incidence of catheter-related mechanical complications in adult patients; and the reasons why patients left the studies early.

SEARCH METHODS

We searched CENTRAL (The Cochrane Library 2011, Issue 9), MEDLINE, CINAHL, EMBASE (from inception to September 2011), four Chinese databases (CBM, WANFANG DATA, CAJD, VIP Database) (from inception to November 2011), Google Scholar and bibliographies of published reviews. The original search was performed in December 2006. We also contacted researchers in the field. There were no language restrictions.

SELECTION CRITERIA

We included randomized controlled trials comparing central venous catheter insertion routes.

DATA COLLECTION AND ANALYSIS

Three authors assessed potentially relevant studies independently. We resolved disagreements by discussion. Dichotomous data on catheter-related complications were analysed. We calculated relative risks (RR) and their 95% confidence intervals (CI) based on a random-effects model.

MAIN RESULTS

We identified 5854 citations from the initial search strategy; 28 references were then identified as potentially relevant. Of these, we Included four studies with data from 1513 participants. We undertook a priori subgroup analysis according to the duration of catheterization, short-term (< one month) and long-term (> one month) defined according to the Food and Drug Administration (FDA).No randomized controlled trial (RCT) was found comparing all three CVA routes and reporting the complications of venous stenosis.Regarding internal jugular versus subclavian CVA routes, the evidence was moderate and applicable for long-term catheterization in cancer patients. Subclavian and internal jugular CVA routes had similar risks for catheter-related complications. Regarding femoral versus subclavian CVA routes, the evidence was high and applicable for short-term catheterization in critically ill patients. Subclavian CVA routes were preferable to femoral CVA routes in short-term catheterization because femoral CVA routes were associated with higher risks of catheter colonization (14.18% or 19/134 versus 2.21% or 3/136) (n = 270, one RCT, RR 6.43, 95% CI 1.95 to 21.21) and thrombotic complications (21.55% or 25/116 versus 1.87% or 2/107) (n = 223, one RCT, RR 11.53, 95% CI 2.80 to 47.52) than with subclavian CVA routes. Regarding femoral versus internal jugular routes, the evidence was moderate and applicable for short-term haemodialysis catheterization in critically ill patients. No significant differences were found between femoral and internal jugular CVA routes in catheter colonization, catheter-related bloodstream infection (CRBSI) and thrombotic complications, but fewer mechanical complications occurred in femoral CVA routes (4.86% or 18/370 versus 9.56% or 35/366) (n = 736, one RCT, RR 0.51, 95% CI 0.29 to 0.88).

AUTHORS' CONCLUSIONS

Subclavian and internal jugular CVA routes have similar risks for catheter-related complications in long-term catheterization in cancer patients. Subclavian CVA is preferable to femoral CVA in short-term catheterization because of lower risks of catheter colonization and thrombotic complications. In short-term haemodialysis catheterization, femoral and internal jugular CVA routes have similar risks for catheter-related complications except internal jugular CVA routes are associated with higher risks of mechanical complications.

Meta-analysis of safety for low event-rate binomial trials

This article focuses on meta-analysis of low event-rate binomial trials. We introduce two forms of random effects: (1) 'studies at random' (SR), where we assume no more than independence between studies; and (2) 'effects at random' (ER), which forces the effect size distribution to be independent of the study design. On the basis of the summary estimates of proportions, we present both unweighted and study-size weighted methods, which, under SR, target different population parameters. We demonstrate mechanistically that the popular DerSimonian-Laird (DL) method, as DL actually warned in their paper, should never be used in this setting. We conducted a survey of the major cardiovascular literature on low event-rate studies and found that DL using odds ratios or relative risks to be the clear method of choice. We looked at two high profile examples from diabetes and cancer, respectively, where the choice of weighted versus unweighted methods makes a large difference. A large simulation study supports the accuracy of the coverage of our approximate confidence intervals. We recommend that before looking at their data, users should prespecify which target parameter they intend to estimate (weighted vs. unweighted) but estimate the other as a secondary analysis. Copyright © 2012 John Wiley & Sons, Ltd.

Meta-analysis: an introduction using regression models

Research synthesis of evaluation findings is a multistep process. An investigator identifies a research question, acquires the relevant literature, codes findings from that literature, and analyzes the coded data to estimate the average treatment effect and its distribution in a population of interest. The process of estimating the average treatment effect is meta-analysis. Meta-analysis is gaining popularity across the social sciences, but introductions and even advanced treatments of the subject are often formula-driven. Researchers, who apply formulas without understanding, risk both misapplication and misinterpretation. This article derives formulas used in meta-analysis using standard regression models so this article seeks to introduce readers with an understanding of regression analysis to meta-analysis, intending thereby to motivate and facilitate a deeper reading of the meta-analysis literature.

A note on the use of the generalized odds ratio in meta-analysis of association studies involving bi- and tri-allelic polymorphisms

Background

The generalized odds ratio (GOR) was recently suggested as a genetic model-free measure for association studies. However, its properties were not extensively investigated. We used Monte Carlo simulations to investigate type-I error rates, power and bias in both effect size and between-study variance estimates of meta-analyses using the GOR as a summary effect, and compared these results to those obtained by usual approaches of model specification. We further applied the GOR in a real meta-analysis of three genome-wide association studies in Alzheimer's disease.

Findings

For bi-allelic polymorphisms, the GOR performs virtually identical to a standard multiplicative model of analysis (e.g. per-allele odds ratio) for variants acting multiplicatively, but augments slightly the power to detect variants with a dominant mode of action, while reducing the probability to detect recessive variants. Although there were differences among the GOR and usual approaches in terms of bias and type-I error rates, both simulation- and real data-based results provided little indication that these differences will be substantial in practice for meta-analyses involving bi-allelic polymorphisms. However, the use of the GOR may be slightly more powerful for the synthesis of data from tri-allelic variants, particularly when susceptibility alleles are less common in the populations (≤10%). This gain in power may depend on knowledge of the direction of the effects.

Conclusions

For the synthesis of data from bi-allelic variants, the GOR may be regarded as a multiplicative-like model of analysis. The use of the GOR may be slightly more powerful in the tri-allelic case, particularly when susceptibility alleles are less common in the populations.

Conducting indirect-treatment-comparison and network-meta-analysis studies: report of the ISPOR Task Force on Indirect Treatment Comparisons Good Research Practices: part 2

Evidence-based health care decision making requires comparison of all relevant competing interventions. In the absence of randomized controlled trials involving a direct comparison of all treatments of interest, indirect treatment comparisons and network meta-analysis provide useful evidence for judiciously selecting the best treatment(s). Mixed treatment comparisons, a special case of network meta-analysis, combine direct evidence and indirect evidence for particular pairwise comparisons, thereby synthesizing a greater share of the available evidence than traditional meta-analysis. This report from the International Society for Pharmacoeconomics and Outcomes Research Indirect Treatment Comparisons Good Research Practices Task Force provides guidance on technical aspects of conducting network meta-analyses (our use of this term includes most methods that involve meta-analysis in the context of a network of evidence). We start with a discussion of strategies for developing networks of evidence. Next we briefly review assumptions of network meta-analysis. Then we focus on the statistical analysis of the data: objectives, models (fixed-effects and random-effects), frequentist versus Bayesian approaches, and model validation. A checklist highlights key components of network meta-analysis, and substantial examples illustrate indirect treatment comparisons (both frequentist and Bayesian approaches) and network meta-analysis. A further section discusses eight key areas for future research.

The effectiveness of group medical visits on diabetes mellitus type 2 (dm2) specific outcomes in adults: a systematic review

BACKGROUND

The prevalence of diabetes mellitus is ubiquitous. Complications and costs of diabetes are rising and are depleting limited resources. It is imperative for healthcare professionals and patients alike to confront this chronic condition by exploring new interventions. An innovative health care delivery model has emerged in the last 25 years in the form of the group medical visit. Group medical visits can range from of six to twenty patients scheduled together with time allotted for individual care as well as in the group setting.

OBJECTIVES

The review objective was to conduct a systematic review and meta-analysis to synthesize the best available evidence related to effectiveness of group medical visits on HbA1c, blood pressure and cholesterol measurements/levels for adult patients with type II diabetes in outpatient settings.

SEARCH STRATEGY

A three-step literature search for studies in English language from 1990 to 2010 was conducted utilising (a) a primary search of Medline, CINAHL, PsycINFO and Cochrane Central Register of Controlled Trials, (b) a secondary search of non-indexed databases, and (c) a search of the grey literature. In addition, a manual review of the reference lists of all identified reports and articles was performed to identify additional studies.

INCLUSION CRITERIA

All randomised and quasi-experimental studies of adult patients (>18) with type II diabetes mellitus seen in outpatient health settings were considered if they met one or more of the following diagnostic outcome measurements: haemoglobin A1c, systolic and diastolic blood pressure, and low density lipoprotein cholesterol.

CRITICAL APPRAISAL, DATA COLLECTION AND ANALYSIS

Each of the eligible articles was reviewed by two independent reviewers. Disagreements between the reviewers were resolved through discussion, or with a third reviewer. Studies that met the inclusion criteria were assessed for methodological quality using the JBI standardized critical appraisal tools. Data extraction was undertaken using the standardised data extraction tool from JBI-MAStARI.

MAIN RESULTS

The search strategy identified 2,040 articles in the published and unpublished literature. Of these, 11 randomised controlled trials and 4 quasi-experimental trials met the inclusion criteria and represented 2240 patients included in the final review.There are clear benefits of group medical visits for patients' HbA1c levels which are consistent in the post-intervention and change from baseline effect sizes. The most significant effect observed is with the change from baseline results. Some evidence suggests post-intervention and change from baseline systolic blood pressure improvement at the nine to twelve month interval and change from baseline improvement at the 4 year timeframe. There is no evidence that group visits improve LDL cholesterol values of the group visit participants.

CONCLUSIONS

Group medical visits should be considered by clinicians as an effective non-pharmacologic intervention that can have a positive impact on biologic markers such as haemoglobin A1c and systolic blood pressure.

IMPLICATIONS FOR PRACTICE

The evidence suggests that the most powerful model of Group medical visits (GMV) is when a clinician prescriber is present during or immediately after GMV sessions for medication reconciliation or individual patient needs.

IMPLICATIONS FOR RESEARCH

Future research should include this GMV intervention in randomised controlled trials across different health systems and socio-economic and ethnic groups.

A systematic review of the effectiveness of nurse coordinated transitioning of care on readmission rates for patients with heart failure

BACKGROUND

Readmission rates for patients with heart failure are a major concern for hospitals worldwide. The importance of patient education and a structured care plan to ease the transition from hospital to home has been the focus of many intervention strategies to reduce readmission rates. The use of transitioning of care plans is believed to improve medication reconciliation, communication, patient education, and follow-up. To date, the evidence has not been systematically evaluated to support the effectiveness of a nurse coordinated transitioning of care for patients with heart failure in reducing readmission rates.

OBJECTIVE

The objective of the systematic review was to identify the best available evidence on the effectiveness of nurse coordinated transitioning of care between hospital and home on hospital readmission rates for all causes in adult patients hospitalised with heart failure.

SEARCH STRATEGY

The search strategy aimed to find both published and unpublished studies in the English language from January 1975 through July 2010. A search of MEDLINE, CINAHL, PsycINFO, Healthsource Nursing/academic edition, EMBASE, the Cochrane Library, and the Joanna Briggs Institute Library of Systematic Reviews was conducted followed by a reference search of relevant studies. The initial key words searched were: heart failure, readmission, and transitional care.

INCLUSION CRITERIA

Randomised controlled trials that evaluated the effect of nurse coordinated transitioning of care from hospital to home in adult patients with heart failure on readmission rates were selected. The outcome was defined as hospital readmissions for all causes following an initial admission for heart failure.

DATA COLLECTION AND ANALYSIS

Studies selected for retrieval were critically evaluated by two independent reviewers for methodological validity using standardised critical appraisal instruments from the Joanna Briggs Institute Meta Analysis of Statistics Assessment and Review Instrument (JBI-MAStARI). Data were extracted and analysed using the JBI-MAStARI program.

RESULTS

A total of 16 randomised controlled studies were included. Ten of the 16 studies included in the review show that a nurse led transitioning of care intervention can reduce the rate of readmission for patients with heart failure. Interventions utilising home visits, or home visits coupled with telephone follow-up, show a more favourable reduction in readmission rates.

CONCLUSIONS

Reduced readmissions occur when transitioning of care interventions are carried out by a heart failure trained nurse who conducts at least one home visit and follows the patient at least weekly for a minimum of 30 days post discharge with either additional home visits or telephone contact.

IMPLICATIONS FOR PRACTICE

This review supports the development of a nurse coordinated transitioning of care plan which will require improvements in communication, in addition to changes in health policy and payment systems that align incentives and performance measures in caring for patients with heart failure.

IMPLICATIONS FOR RESEARCH

Future research should evaluate the effect of the intensity and duration of the transitioning of care intervention on readmission rates in a large randomised control trial on an adult population with heart failure to determine the ideal frequency and duration of the post discharge interventions.

Identifying Systematic Disobedience in Milgram's Obedience Experiments: A Meta-Analytic Review

A meta-analysis of data from eight of Mil-gram's obedience experiments reveals previously undocumented systematicity in the behavior of disobedient participants. In all studies, disobedience was most likely at 150 v, the point at which the shocked "learner" first requested to be released. Further illustrating the importance of the 150-v point, obedience rates across studies covaried with rates of disobedience at 150 v, but not at any other point; as obedience decreased, disobedience at 150 v increased. In contrast, disobedience was not associated with the learner's escalating expressions of pain. This analysis identifies a critical decision point in the obedience paradigm and suggests that disobedient participants perceived the learner's right to terminate the experiment as overriding the experimenter's orders, a finding with potential implications for the treatment of prisoners.