Showing papers by "Gordon H. Guyatt published in 2009"

Grading Quality of Evidence and Strength of Recommendations for Diagnostic Tests and Strategies

Abstract: The GRADE system can be used to grade the quality of evidence and strength of recommendations for diagnostic tests or strategies. This article explains how patient-important outcomes are taken into account in this process

Can trial sequential monitoring boundaries reduce spurious inferences from meta-analyses?

Abstract: Background Results from apparently conclusive meta-analyses may be false. A limited number of events from a few small trials and the associated random error may be under-recognized sources of spurious findings. The information size (IS, i.e. number of participants) required for a reliable and conclusive meta-analysis should be no less rigorous than the sample size of a single, optimally powered randomized clinical trial. If a meta-analysis is conducted before a sufficient IS is reached, it should be evaluated in a manner that accounts for the increased risk that the result might represent a chance finding (i.e. applying trial sequential monitoring boundaries). Methods We analysed 33 meta-analyses with a sufficient IS to detect a treatment effect of 15% relative risk reduction (RRR). We successively monitored the results of the meta-analyses by generating interim cumulative meta-analyses after each included trial and evaluated their results using a conventional statistical criterion (a ¼ 0.05) and two-sided Lan-DeMets monitoring boundaries. We examined the proportion of false positive results and important inaccuracies in estimates of treatment effects that resulted from the two approaches.

Association between change in high density lipoprotein cholesterol and cardiovascular disease morbidity and mortality: systematic review and meta-regression analysis

Abstract: Objective To investigate the association between treatment induced change in high density lipoprotein cholesterol and total death, coronary heart disease death, and coronary heart disease events (coronary heart disease death and non-fatal myocardial infarction) adjusted for changes in low density lipoprotein cholesterol and drug class in randomised trials of lipid modifying interventions. Design Systematic review and meta-regression analysis of randomised controlled trials. Data sources Medline, Embase, Central, CINAHL, and AMED to October 2006 supplemented by contact with experts in the field. Study selection In teams of two, reviewers independently determined eligibility of randomised trials that tested lipid modifying interventions to reduce cardiovascular risk, reported high density lipoprotein cholesterol and mortality or myocardial infarctions separately for treatment groups, and treated and followed participants for at least six months. Data extraction and synthesis Using standardised, pre-piloted forms, reviewers independently extracted relevant information from each article. The change in lipid concentrations for each trial and the weighted risk ratios for clinical outcomes were calculated. Results The meta-regression analysis included 108 randomised trials involving 299 310 participants at risk of cardiovascular events. All analyses that adjusted for changes in low density lipoprotein cholesterol showed no association between treatment induced change in high density lipoprotein cholesterol and risk ratios for coronary heart disease deaths, coronary heart disease events, or total deaths. With all trials included, change in high density lipoprotein cholesterol explained almost no variability ( Conclusions Available data suggest that simply increasing the amount of circulating high density lipoprotein cholesterol does not reduce the risk of coronary heart disease events, coronary heart disease deaths, or total deaths. The results support reduction in low density lipoprotein cholesterol as the primary goal for lipid modifying interventions.

Quality of care in for-profit and not-for-profit nursing homes: systematic review and meta-analysis

Abstract: Objective To compare quality of care in for-profit and not-for-profit nursing homes. Design Systematic review and meta-analysis of observational studies and randomised controlled trials investigating quality of care in for-profit versus not-for-profit nursing homes. Results A comprehensive search yielded 8827 citations, of which 956 were judged appropriate for full text review. Study characteristics and results of 82 articles that met inclusion criteria were summarised, and results for the four most frequently reported quality measures were pooled. Included studies reported results dating from 1965 to 2003. In 40 studies, all statistically significant comparisons (P Conclusions This systematic review and meta-analysis of the evidence suggests that, on average, not-for-profit nursing homes deliver higher quality care than do for-profit nursing homes. Many factors may, however, influence this relation in the case of individual institutions.

The diabetes mellitus medication choice decision aid: a randomized trial.

Abstract: Background Patient involvement in the choice of antihyperglycemic agents could improve adherence and optimize glycemic control in patients with type 2 diabetes mellitus. Methods We conducted a pilot, cluster randomized trial of Diabetes Medication Choice, a decision aid that describes 5 antihyperglycemic drugs, their treatment burden (adverse effects, administration, and self-monitoring demands), and impact on hemoglobin A 1c (HbA 1c ) levels. Twenty-one clinicians were randomized to use the decision aid during the clinical encounter and 19 to dispense usual care and an educational pamphlet. We used surveys and video analysis to assess postvisit decisional outcomes, and medical and pharmacy records to assess 6-month medication adherence and HbA 1c levels. Results Compared with usual care patients (n = 37), patients receiving the decision aid (n = 48) found the tool more helpful (clustered-adjusted mean difference [AMD] in a 7-point scale, 0.38; 95% confidence interval [CI], 0.04-0.72); had improved knowledge (AMD, 1.10 of 10 questions; 95% CI, 0.11-2.09); and had more involvement in making decisions about diabetes medications (AMD, 21.8 of 100; 95% CI, 13.0-30.5). At 6-month follow-up, both groups had nearly perfect medication use (median, 100% of days covered), with better adherence (AMD, 9% more days covered; 95% CI, 4%-14%) and persistence (AMD, 12 more days covered; 95% CI, 3-21 days) in the usual care group, and no significant impact on HbA 1c levels (AMD, 0.01; 95% CI, −0.49 to 0.50). Conclusion An innovative decision aid effectively involved patients with type 2 diabetes mellitus in decisions about their medications but did not improve adherence or HbA 1c levels. Trial Registration clinicaltrials.gov Identifier:NCT00388050

Design, analysis, and presentation of crossover trials.

Abstract: Although crossover trials enjoy wide use, standards for analysis and reporting have not been established. We reviewed methodological aspects and quality of reporting in a representative sample of published crossover trials. We searched MEDLINE for December 2000 and identified all randomized crossover trials. We abstracted data independently, in duplicate, on 14 design criteria, 13 analysis criteria, and 14 criteria assessing the data presentation. We identified 526 randomized controlled trials, of which 116 were crossover trials. Trials were drug efficacy (48%), pharmacokinetic (28%), and nonpharmacologic (30%). The median sample size was 15 (interquartile range 8–38). Most (72%) trials used 2 treatments and had 2 periods (64%). Few trials reported allocation concealment (17%) or sequence generation (7%). Only 20% of trials reported a sample size calculation and only 31% of these considered pairing of data in the calculation. Carry-over issues were addressed in 29% of trial's methods. Most trials reported and defended a washout period (70%). Almost all trials (93%) tested for treatment effects using paired data and also presented details on by-group results (95%). Only 29% presented CIs or SE so that data could be entered into a meta-analysis. Reports of crossover trials frequently omit important methodological issues in design, analysis, and presentation. Guidelines for the conduct and reporting of crossover trials might improve the conduct and reporting of studies using this important trial design.

Epistemologic Inquiries in Evidence-Based Medicine

Abstract: Background: Since the term “evidence-based medicine” (EBM) first appeared in the scientific literature in 1991, the concept has had considerable influence in many parts of the world. Most professional societies, the public, and funding agencies have accepted EBM with remarkable enthusiasm. The concept of evidence-based practice is now applied in management, education, criminology, and social work. Yet, EBM has attracted controversy: its critics allege that EBM uses a narrow concept of evidence and a naive conception of the relationships between evidence, theory, and practice. They also contend that EBM presents itself as a radical restructuring of medical knowledge that discredits more traditional ways of knowing in medicine, largely in the interests of people with a particular investment in the enterprise of large-scale clinical trials. Because EBM proposes a specific relationship between theory, evidence, and knowledge, its theoretical basis can be understood as an epistemological system. Undertaking epistemological inquiry is important because the adoption of a particular epistemological view defines how science is conducted. Methods: In this paper, we challenge this critical view of EBM by examining how EBM fits into broad epistemological debates within the philosophy of science. We consider how EBM relates to some classical debates regarding the nature of science and knowledge. We investigate EBM from the perspective of major epistemological theories (logical-positivism/inductivism, deductivism/falsificationism/theory-ladeness of observations, explanationism/holism, instrumentalism, underdetermination theory by evidence). Results: We first explore the relationship between evidence and knowledge and discuss philosophical support for the main way that evidence is used in medicine: (1) in the philosophical tradition that “rational thinkers respect their evidence,” we show that EBM refers to making medical decisions that are consistent with evidence, (2) as a reliable sign, symptom, or mark to enhance reasonableness or truthfulness of some particular claim (“evidence as a guide to truth”), and (3) to serve as a neutral arbiter among competing views. Our analysis indicates that EBM does not have a rigorous epistemological stance. In fact, EBM enthusiastically draws on all major traditions of philosophical theories of scientific evidence. Conclusions: Our findings indicate that EBM should not be construed as a new scientific or philosophical theory that changes the nature of medicine or our understanding thereof. Rather, we should consider EBM as a continuously evolving heuristic structure for optimizing clinical practice.

How to use an article about genetic association: B: Are the results of the study valid?

Abstract: In the first article of this series, we reviewed the basic genetics concepts necessary to understand genetic association studies. In this second article, we enumerate the major issues in judging the validity of these studies, framed as critical appraisal questions. Was the disease phenotype properly defined and accurately recorded by someone blind to the genetic information? Have any potential differences between disease and nondisease groups, particularly ethnicity, been properly addressed? In genetic studies, one potential cause of spurious associations is differences between cases and controls in ethnicity, a situation termed population stratification. Was measurement of the genetic variants unbiased and accurate? Methods for determining DNA sequence variation are not perfect and may have some measurement error. Do the genotype proportions observe Hardy-Weinberg equilibrium? This simple mathematic rule about the distribution of genetic groups may be one way to check for errors in reading DNA information. Have the investigators adjusted their inferences for multiple comparisons? Given the thousands of genetic markers tested in genome-wide association studies, the potential for false-positive and false-negative results is much higher than in traditional medical studies, and it is particularly important to look for replication of results.

Evaluating agreement: conducting a reliability study.

Abstract: Instruments that are useful in clinical or research practice will, when the object of measurement is stable, yield similar results when applied at different times, in different situations, or by different users. Studies that measure the relation of differences between patients or subjects and measurement error (reliability studies) are becoming increasingly common in the orthopaedic literature. In this paper, we identify common aspects of reliability studies and suggest features that improve the reader's confidence in the results. One concept serves as the foundation for all further consideration: in order for a reliability study to be relevant, the patients, raters, and test administration in the study must be similar to the clinical or research context in which the instrument will be used. We introduce the statistical measures that readers will most commonly encounter in reliability studies, and we suggest an approach to sample-size estimation. Readers interested in critically appraising reliability studies or in developing their own reliability studies may find this review helpful.

Low intensity pulsed ultrasonography for fractures: systematic review of randomised controlled trials

Abstract: Objective To determine the efficacy of low intensity pulsed ultrasonography for healing of fractures. Design Systematic review of randomised controlled trials. Data sources Electronic literature search without language restrictions of CINAHL, Embase, Medline, HealthSTAR, and the Cochrane Central Registry of Controlled Trials, from inception of the database to 10 September 2008. Review methods Eligible studies were randomised controlled trials that enrolled patients with any kind of fracture and randomly assigned them to low intensity pulsed ultrasonography or to a control group. Two reviewers independently agreed on eligibility; three reviewers independently assessed methodological quality and extracted outcome data. All outcomes were included and meta-analyses done when possible. Results 13 randomised trials, of which five assessed outcomes of importance to patients, were included. Moderate quality evidence from one trial found no effect of low intensity pulsed ultrasonography on functional recovery from conservatively managed fresh clavicle fractures; whereas low quality evidence from three trials suggests benefit in non-operatively managed fresh fractures (faster radiographic healing time mean 36.9%, 95% confidence interval 25.6% to 46.0%). A single trial provided moderate quality evidence suggesting no effect of low intensity pulsed ultrasonography on return to function among non-operatively treated stress fractures. Three trials provided very low quality evidence for accelerated functional improvement after distraction osteogenesis. One trial provided low quality evidence for a benefit of low intensity pulsed ultrasonography in accelerating healing of established non-unions managed with bone graft. Four trials provided low quality evidence for acceleration of healing of operatively managed fresh fractures. Conclusion Evidence for the effect of low intensity pulsed ultrasonography on healing of fractures is moderate to very low in quality and provides conflicting results. Although overall results are promising, establishing the role of low intensity pulsed ultrasonography in the management of fractures requires large, blinded trials, directly addressing patient important outcomes such as return to function.

Combined corticosteroid and antiviral treatment for bell palsy: a systematic review and meta-analysis

Abstract: Context New evidence has emerged regarding the use of corticosteroids and antiviral agents in Bell palsy. Objective To estimate the association of corticosteroids and antiviral agents with the risk of unsatisfactory facial recovery in patients with Bell palsy. Data Sources The search included MEDLINE, EMBASE, CENTRAL, PsychInfo, CINAHL, Web of Science, PAPERSFIRST, PROCEEDINGSFIRST, and PROQUEST to identify studies up to March 1, 2009. Study Selection and Data Extraction Eligible studies were randomized controlled trials comparing treatment with either corticosteroids or antiviral agents with a control and measuring at least 1 of the following outcomes: unsatisfactory facial recovery (≥4 months), unsatisfactory short-term recovery (6 weeks to Results Eighteen trials involving 2786 patients were eligible. Regression analysis identified a synergistic effect when corticosteroids and antiviral agents were administered in combination compared with alone (odds ratio for interaction term, 0.54 [95% confidence interval {CI}, 0.35-0.83]; P = .004). Meta-analysis using a random-effects model showed corticosteroids alone were associated with a reduced risk of unsatisfactory recovery (relative risk [RR], 0.69 [95% CI, 0.55-0.87]; P = .001) (number needed to treat to benefit 1 person, 11 [95% CI, 8-25]), a reduced risk of synkinesis and autonomic dysfunction (RR, 0.48 [95% CI, 0.36-0.65]; P Conclusions In Bell palsy, corticosteroids are associated with a reduced risk of unsatisfactory recovery. Antiviral agents, when administered with corticosteroids, may be associated with additional benefit.

How to use an article about genetic association: A: Background concepts.

Abstract: This is the first in a series of 3 articles serving as an introduction to clinicians wishing to read and critically appraise genetic association studies. We summarize the key concepts in genetics that clinicians must understand to review these studies, including the structure of DNA, transcription and translation, patterns of inheritance, Hardy-Weinberg equilibrium, and linkage disequilibrium. We review the types of DNA variation, including single-nucleotide polymorphisms (SNPs), insertions, and deletions, and how these can affect protein function. We introduce the idea of genetic association for both single-candidate gene and genome-wide association studies, in which thousands of genetic variants are tested for association with disease. We use the APOE polymorphism and its association with dementia as a case study to demonstrate the concepts and introduce the terminology used in this field. The second and third articles will focus on issues of validity and applicability.

A treatment decision aid may increase patient trust in the diabetes specialist. The Statin Choice randomized trial

Abstract: Aims Decision aids in practice may affect patient trust in the clinician, a requirement for optimal diabetes care. We sought to determine the impact of a decision aid to help patients with diabetes decide about statins (Statin Choice) on patients’ trust in the clinician. Methods We randomized 16 diabetologists and 98 patients with type 2 diabetes referred to a subspecialty diabetes clinic to use the Statin Choice decision aid or a patient pamphlet about dyslipidaemia, and then to receive these materials from either the clinician during the visit or a researcher prior to the visit. Providers and patients were blinded to the study hypothesis. Immediately after the clinical encounter, patients completed a survey including questions on trust (range 0 to total trust = 100), knowledge, and decisional conflict. Researchers reviewed videotaped encounters and assessed patient participation (using the OPTION scale) and visit length. Results Overall mean trust score was 91 (median 97.2, IQR 86, 100). After adjustment for patient characteristics, results suggested greater total trust (trust = 100) with the decision aid [odds ratio (OR) 1.77, 95% CI 0.94, 3.35]. Total trust was associated with knowledge (for each additional knowledge point, OR 1.3, 95% CI 1.1, 1.6), patient participation (for each additional point in the OPTION scale, OR 1.1, 95% CI 1.1, 1.2), and decisional conflict (for every 5-point decrease in conflict, OR 1.5, 95% CI 1.2, 1.9). Total trust was not associated with visit length, which the decision aid did not significantly affect. There was no significant effect interaction across the trial factors. Conclusions Preliminary evidence suggests that decision aids do not have a large negative impact on trust in the physician and may increase trust through improvements in the decision-making process.

LOST to follow-up Information in Trials (LOST-IT): a protocol on the potential impact

Abstract: Background Incomplete ascertainment of outcomes in randomized controlled trials (RCTs) is likely to bias final study results if reasons for unavailability of patient data are associated with the outcome of interest. The primary objective of this study is to assess the potential impact of loss to follow-up on the estimates of treatment effect. The secondary objectives are to describe, for published RCTs, (1) the reporting of loss to follow-up information, (2) the analytic methods used for handling loss to follow-up information, and (3) the extent of reported loss to follow-up.

How to Use an Article About Genetic Association: C: What Are the Results and Will They Help Me in Caring for My Patients?

Abstract: In the first 2 articles of this series, we reviewed the basic genetics concepts necessary to understand genetic association studies, and we enumerated the major issues in judging the validity of these studies. In this third article, we review the issues relating to the applicability of the results in the clinical situation. How large and precise are the associations? Many genetic effects are expected to be smaller in magnitude than traditional risk factors. Does the genetic association improve predictive power beyond easily measured clinical variables? In some cases, the additional genetic information adds only a small increment in the predictive ability of a diagnostic or prognostic test. What are the absolute vs relative effects? Even if the genetic risk is high in relative terms, the baseline risk may be very low in absolute terms. Is the risk-associated allele likely to be present in my patient? A risk allele may have a strong effect but be rare in a particular ethnic group. Is the patient likely better off knowing the genetic information? Given that genes cannot be modified, one must weigh whether the genetic information is likely to be helpful in planning other health interventions or initiating behavior change.

Psychological intervention for premenstrual syndrome: a meta-analysis of randomized controlled trials.

Abstract: Background: We conducted a systematic review and meta-analysis to determine the efficacy of psychological interventions for premenstrual syndrome. Methods: We sys

The more the better?: the impact of surgeon and hospital volume on in-hospital mortality following colorectal resection.

Abstract: Objective: To determine the in-hospital mortality rates for patients undergoing colorectal resection for malignant or benign conditions, and to identify risk factors for in-hospital death, particularly the relationships with surgeon and hospital volume. Background: Although there is strong evidence that complex cancer operations are best performed at specialized high-volume centers and by high-volume surgeons, the relationship between surgeon and hospital volume and perioperative outcomes is less well defined for more common procedures such as colorectal resections, particularly for benign diseases. Methods: We obtained data from the Canadian Institute for Health Information Discharge Abstract Database on all adult patients who underwent colorectal resection between April 1, 2005 and March 31, 2006. We performed a logistic regression to identify variables associated with a higher likelihood of in-hospital death. Results: Twenty-one thousand seventy-four patients underwent colorectal resection, with the majority being elective (59.4%). Malignancy represented the most common indication for resection (56.8%), followed by diverticular disease (16.2%) and inflammatory bowel disease (7.1%). The overall in-hospital mortality rate among patients undergoing colorectal resection was 5.3%. Increased age (adjusted Odds Ratio [OR]: 1.97 per 10 years, P < 0.001), urgent operation (OR: 2.63, P < 0.001), indication for resection (P < 0.001), nature of the surgery (P < 0.001), and several comorbidities were all independently associated with an increased risk of death. Surgeons with higher volumes of colorectal resections achieved significantly lower mortality rates (OR: 0.92 per 20 cases/y, P = 0.003), corresponding to an adjusted mortality rate of 5.6% for surgeons in the bottom decile (1 case per year) compared with 4.5% for surgeons in the top decile (greater than 43 cases per year). Hospital volume was not associated with mortality (OR: 1.00 per 10 cases, P = 0.504). Conclusions: This large, population-based study suggests that surgeons who perform high volumes of colorectal resections achieve lower in-hospital mortality rates than surgeons with low volumes, whereas the hospital volume does not influence mortality.

Citrus fruit intake and pancreatic cancer risk: a quantitative systematic review.

Abstract: Objectives:The purpose of this systematic review was to investigate the association between dietary intake of citrus fruits and pancreatic cancer risk.Methods:Authors searched electronic databases and the reference lists of publications of studies addressing diet and pancreatic cancer up to December

Subgroup Analysis of Trials Is Rarely Easy (SATIRE): a study protocol for a systematic review to characterize the analysis, reporting, and claim of subgroup effects in randomized trials

Abstract: Subgroup analyses in randomized trials examine whether effects of interventions differ between subgroups of study populations according to characteristics of patients or interventions. However, findings from subgroup analyses may be misleading, potentially resulting in suboptimal clinical and health decision making. Few studies have investigated the reporting and conduct of subgroup analyses and a number of important questions remain unanswered. The objectives of this study are: 1) to describe the reporting of subgroup analyses and claims of subgroup effects in randomized controlled trials, 2) to assess study characteristics associated with reporting of subgroup analyses and with claims of subgroup effects, and 3) to examine the analysis, and interpretation of subgroup effects for each study's primary outcome. We will conduct a systematic review of 464 randomized controlled human trials published in 2007 in the 118 Core Clinical Journals defined by the National Library of Medicine. We will randomly select journal articles, stratified in a 1:1 ratio by higher impact versus lower impact journals. According to 2007 ISI total citations, we consider the New England Journal of Medicine, JAMA, Lancet, Annals of Internal Medicine, and BMJ as higher impact journals. Teams of two reviewers will independently screen full texts of reports for eligibility, and abstract data, using standardized, pilot-tested extraction forms. We will conduct univariable and multivariable logistic regression analyses to examine the association of pre-specified study characteristics with reporting of subgroup analyses and with claims of subgroup effects for the primary and any other outcomes. A clear understanding of subgroup analyses, as currently conducted and reported in published randomized controlled trials, will reveal both strengths and weaknesses of this practice. Our findings will contribute to a set of recommendations to optimize the conduct and reporting of subgroup analyses, and claim and interpretation of subgroup effects in randomized trials.

Immunotherapy improves health‐related quality of life of adult patients with dermal reactions following yellow jacket stings

Abstract: In many European centers insect venom allergic patients with a reaction confined to the skin are only offered an epinephrine auto-injector and not venom immunotherapy (VIT). Previously we showed that VIT improves health-related quality of life (HRQL) of yellow jacket allergic patients with more than dermal reactions. To examine whether HRQL of dermal reactors is impaired and to examine the influence of VIT on HRQL in comparison with the EpiPen((R)). Patients with solely dermal reactions were asked if they were willing to be randomized either to VIT or EpiPen((R)), after receiving patient information. Before and 1 year after enrollment, patients completed the Vespid allergy Quality of Life Questionnaire (VQLQ), Burden of Treatment and Expectation of Outcome. Of 55 patients eligible for the study, 29 consented to randomization: 15 to VIT, 14 to EpiPen((R)). The remaining 26 patients preferred to choose their treatment: 11 VIT and 15 EpiPen((R)). The VQLQ score of patients randomized to VIT improved (mean change 0.83 (SD 0.87), in contrast to patients randomized to the EpiPen((R)) whose scores deteriorated (mean change -0.42 (SD 0.64), P <0.0001), resulting in an overall difference of 1.25 [95% confidence interval (CI): 0.63-1.87]. With a minimal important difference of 0.5 indicating a clinically significant improvement, VIT generated an number needed to treat (NNT) of 1.7. Dermal reactors did not consider VIT burdensome and rated this treatment as being superior to the EpiPen((R)). VIT results in a clinically significant improvement of HRQL in most patients with reactions limited to the skin following yellow jacket stings. Prescription of an EpiPen((R)) in patients not choosing this treatment is associated with deterioration in HRQL and should therefore be avoided as definitive treatment in these patients.

Incorporating multiple interventions in meta-analysis: an evaluation of the mixed treatment comparison with the adjusted indirect comparison.

Abstract: Comparing the effectiveness of interventions is now a requirement for regulatory approval in several countries. It also aids in clinical and public health decision-making. However, in the absence of head-to-head randomized trials (RCTs), determining the relative effectiveness of interventions is challenging. Several methodological options are now available. We aimed to determine the comparative validity of the adjusted indirect comparisons of RCTs with the mixed treatment comparison approach. Using systematic searching, we identified all meta-analyses evaluating more than 3 interventions for a similar disease state with binary outcomes. We abstracted data on each clinical trial including population n and outcomes. We conducted fixed effects meta-analysis of each intervention versus mutual comparator and then applied the adjusted indirect comparison. We conducted a mixed treatment meta-analysis on all trials and compared the point estimates and 95% confidence/credible intervals (CIs/CrIs) to determine important differences. We included data from 7 reviews that met our inclusion criteria, allowing a total of 51 comparisons. According to the a priori consistency rule, we found 2 examples where the analytic comparisons were statistically significant using the mixed treatment comparison over the adjusted indirect comparisons and 1 example where this was vice versa. We found 6 examples where the direction of effect differed according to the indirect comparison method chosen and we found 9 examples where the confidence intervals were importantly different between approaches. In most analyses, the adjusted indirect comparison yields estimates of relative effectiveness equal to the mixed treatment comparison. In less complex indirect comparisons, where all studies share a mutual comparator, both approaches yield similar benefits. As comparisons become more complex, the mixed treatment comparison may be favoured.

Impact of quality of evidence on the strength of recommendations: an empirical study

Abstract: Background Evidence is necessary but not sufficient for decision-making, such as making recommendations by clinical practice guideline panels. However, the fundamental premise of evidence-based medicine (EBM) rests on the assumed link between the quality of evidence and "truth" and/or correctness in making guideline recommendations. If this assumption is accurate, then the quality of evidence ought to play a key role in making guideline recommendations. Surprisingly, and despite the widespread penetration of EBM in health care, there has been no empirical research to date investigating the impact of quality of evidence on the strength of recommendations made by guidelines panels.