Diane Heels-Ansdell

Problems with use of composite end points in cardiovascular trials: systematic review of randomised controlled trials

Objective To explore the extent to which components of composite end points in randomised controlled trials vary in importance to patients, the frequency of events in the more and less important components, and the extent of variability in the relative risk reductions across components. Design Systematic review of randomised controlled trials. Data sources Cardiovascular randomised controlled trials published in the Lancet, Annals of Internal Medicine, Circulation, European Heart Journal, JAMA, and New England Journal of Medicine, from 1 January 2002 to 30 June 2003. Component end points of composite end points were categorised according to importance to patients as fatal, critical, major, moderate, or minor. Results Of 114 identified randomised controlled trials that included a composite end point of importance to patients, 68% (n=77) reported complete component data for the primary composite end point; almost all (98%; n=112) primary composite end points included a fatal end point. Of 84 composite end points for which component data were available, 54% (n=45) showed large or moderate gradients in both importance to patients and magnitude of effect across components. When analysed by categories of importance to patients, the most important components were associated with lower event rates in the control group (medians of 3.3-3.7% for fatal, critical, and major outcomes; 12.3% for moderate outcomes; and 8.0% for minor outcomes). Components of greater importance to patients were associated with smaller treatment effects than less important ones (relative risk reduction of 8% for death and 33% for components of minor importance to patients). Conclusion The use of composite end points in cardiovascular trials is frequently complicated by large gradients in importance to patients and in magnitude of the effect of treatment across component end points. Higher event rates and larger treatment effects associated with less important components may result in misleading impressions of the impact of treatment.

Dalteparin versus unfractionated heparin in critically ill patients.

BACKGROUND The effects of thromboprophylaxis with low-molecular-weight heparin, as compared with unfractionated heparin, on venous thromboembolism, bleeding, and other outcomes are uncertain in critically ill patients. METHODS In this multicenter trial, we tested the superiority of dalteparin over unfractionated heparin by randomly assigning 3764 patients to receive either subcutaneous dalteparin (at a dose of 5000 IU once daily) plus placebo once daily (for parallel-group twice-daily injections) or unfractionated heparin (at a dose of 5000 IU twice daily) while they were in the intensive care unit. The primary outcome, proximal leg deep-vein thrombosis, was diagnosed on compression ultrasonography performed within 2 days after admission, twice weekly, and as clinically indicated. Additional testing for venous thromboembolism was performed as clinically indicated. Data were analyzed according to the intention-to-treat principle. RESULTS There was no significant between-group difference in the rate of proximal leg deep-vein thrombosis, which occurred in 96 of 1873 patients (5.1%) receiving dalteparin versus 109 of 1873 patients (5.8%) receiving unfractionated heparin (hazard ratio in the dalteparin group, 0.92; 95% confidence interval [CI], 0.68 to 1.23; P=0.57). The proportion of patients with pulmonary emboli was significantly lower with dalteparin (24 patients, 1.3%) than with unfractionated heparin (43 patients, 2.3%) (hazard ratio, 0.51; 95% CI, 0.30 to 0.88; P=0.01). There was no significant between-group difference in the rates of major bleeding (hazard ratio, 1.00; 95% CI, 0.75 to 1.34; P=0.98) or death in the hospital (hazard ratio, 0.92; 95% CI, 0.80 to 1.05; P=0.21). In prespecified per-protocol analyses, the results were similar to those of the main analyses, but fewer patients receiving dalteparin had heparin-induced thrombocytopenia (hazard ratio, 0.27; 95% CI, 0.08 to 0.98; P=0.046). CONCLUSIONS Among critically ill patients, dalteparin was not superior to unfractionated heparin in decreasing the incidence of proximal deep-vein thrombosis. (Funded by the Canadian Institutes of Health Research and others; PROTECT ClinicalTrials.gov number, NCT00182143.).

Credibility of claims of subgroup effects in randomised controlled trials: systematic review

Objective To investigate the credibility of authors’ claims of subgroup effects using a representative sample of recently published randomised controlled trials. Design Systematic review. Data source Core clinical journals, as defined by the National Library of Medicine, in Medline. Study selection Randomised controlled trials published in 2007. Using prespecified criteria, teams of trained reviewers independently judged whether authors claimed subgroup effects and the strength of their claims. Reviewers assessed each of these claims against 10 predefined criteria, developed through a search of existing criteria and a consensus process. Results Of 207 randomised controlled trials reporting subgroup analyses, 64 (31%) made claims for the primary outcome. Of those, 20 were strong claims and 28 claims of a likely effect. Authors included subgroup variables measured at baseline in 60 (94%) trials, used subgroup variable as a stratification factor at randomisation in 13 (20%), clearly prespecified their hypotheses in 26 (41%), correctly prespecified direction in 4 (6%), tested a small number of hypotheses in 28 (44%), carried out a test of interaction that proved statistically significant in 6 (9%), documented replication of a subgroup effect with previous related studies in 21 (33%), identified consistency of a subgroup effect across related outcomes in 19 (30%), and provided a compelling indirect evidence for the effect in 14 (22%). In the 19 trials making more than one claim, only one (5%) checked the independence of the interaction. Of the 64 claims, 54 (84%) met four or fewer of the 10 criteria. For strong claims, more than 50% failed each of the individual criteria, and only three (15%) met more than five criteria. Conclusion Authors often claim subgroup effects in their trial report. However, the credibility of subgroup effects, even when claims are strong, is usually low. Users of the information should treat claims that fail to meet most criteria with scepticism. Trial researchers should report the conduct of subgroup analyses and provide sufficient evidence when claiming a subgroup effect or suggesting a possible effect.

Is a pre-operative brain natriuretic peptide or N-terminal pro-B-type natriuretic peptide measurement an independent predictor of adverse cardiovascular outcomes within 30 days of noncardiac surgery? A systematic review and meta-analysis of observational studies.

OBJECTIVES We conducted a systematic review and meta-analysis to determine if pre-operative brain natriuretic peptide (BNP) (i.e., BNP or N-terminal pro-B-type natriuretic peptide [NT-proBNP]) is an independent predictor of 30-day adverse cardiovascular outcomes after noncardiac surgery. BACKGROUND Pre-operative clinical cardiac risk indices have only modest predictive power. BNP predicts adverse cardiovascular outcomes in a variety of nonsurgical settings and may similarly predict these outcomes in the perioperative setting. METHODS We employed 5 search strategies (e.g., searching bibliographic databases), and we included all studies that assessed the independent prognostic value of pre-operative BNP measurement as a predictor of cardiovascular complications after noncardiac surgery. We determined study eligibility and conducted data abstraction independently and in duplicate. We calculated a pooled odds ratio using a random effects model. RESULTS Nine studies met eligibility criteria, and included a total of 3,281 patients, among whom 314 experienced 1 or more perioperative cardiovascular complications. The average proportion of patients with elevated BNP was 24.8% (95% confidence interval [CI]: 20.1 to 30.4%; I(2) = 89%). All studies showed a statistically significant association between an elevated pre-operative BNP level and various cardiovascular outcomes (e.g., a composite of cardiac death and nonfatal myocardial infarction; atrial fibrillation). Data pooled from 7 studies demonstrated an odds ratio (OR) of 19.3 (95% CI: 8.5 to 43.7; I(2) = 58%). The pre-operative BNP measurement was an independent predictor of perioperative cardiovascular events among studies that only considered the outcomes of death, cardiovascular death, or myocardial infarction (OR: 44.2, 95% CI: 7.6 to 257.0, I(2) = 51.6%), and those that included other outcomes (OR: 14.7, 95% CI: 5.7 to 38.2, I(2) = 62.2%); the p value for interaction was 0.28. CONCLUSIONS These results suggest that an elevated pre-operative BNP or NT-proBNP measurement is a powerful, independent predictor of cardiovascular events in the first 30 days after noncardiac surgery.

Prevalence and underdiagnosis of chronic obstructive pulmonary disease among patients at risk in primary care

Background: People with known risk factors for chronic obstructive pulmonary disease (COPD) are important targets for screening and early intervention. We sought to measure the prevalence of COPD among such individuals visiting a primary care practitioner for any reason. We also evaluated the accuracy of prior diagnosis or nondiagnosis of COPD and identified associated clinical characteristics. Methods: We recruited patients from three primary care sites who were 40 years or older and had a smoking history of at least 20 pack-years. Participants were asked about respiratory symptoms and underwent postbronchodilator spirometry. COPD was defined as a ratio of forced expiratory volume in the first second of expiration to forced vital capacity (FEV1/FVC) of less than 0.7 and an FEV1 of less than 80% predicted. Results: Of the 1459 patients who met the study criteria, 1003 (68.7%) completed spirometry testing. Of these, 208 were found to have COPD, for a prevalence of 20.7% (95% confidence interval 18.3%–23.4%). Of the 205 participants with COPD who completed the interview about respiratory symptoms before spirometry, only 67 (32.7%) were aware of their diagnosis before the study. Compared with patients in whom COPD had been correctly diagnosed before the study, those in whom COPD had been over-diagnosed or undiagnosed were similar in terms of age, sex, current smoking status and number of visits to a primary care practitioner because of a respiratory problem. Interpretation: Among adult patients visiting a primary care practitioner, as many as one in five with known risk factors met spirometric criteria for COPD. Underdiagnosis of COPD was frequent, which suggests a need for greater screening of at-risk individuals. Knowledge of the prevalence of COPD will help plan strategies for disease management.

Prognostic value of troponin and creatine kinase muscle and brain isoenzyme measurement after noncardiac surgery: a systematic review and meta-analysis.

Background:There is uncertainty regarding the prognostic value of troponin and creatine kinase muscle and brain isoenzyme measurements after noncardiac surgery. Methods:The current study undertook a systematic review and meta-analysis. The study used six search strategies and included noncardiac surgery studies that provided data from a multivariable analysis assessing whether a postoperative troponin or creatine kinase muscle and brain isoenzyme measurement was an independent predictor of mortality or a major cardiovascular event. Independent investigators determined study eligibility and abstracted data in duplicate. Results:Fourteen studies, enrolling 3,318 patients and 459 deaths, demonstrated that an increased troponin measurement after surgery was an independent predictor of mortality (odds ratio [OR] 3.4, 95% confidence interval [CI] 2.2–5.2), but there was substantial heterogeneity (I2 = 56%). The independent prognostic capabilities of an increased troponin value after surgery in the 10 studies that assessed intermediate-term (≤ 12 months) mortality was an OR = 6.7 (95% CI 4.1–10.9, I2 = 0%) and in the 4 studies that assessed long-term (more than 12 months) mortality was an OR = 1.8 (95% CI 1.4–2.3, I2 = 0%; P < 0.001 for test of interaction). Four studies, including 1,165 patients and 202 deaths, demonstrated an independent association between an increased creatine kinase muscle and brain isoenzyme measurement after surgery and mortality (OR 2.5, 95% CI 1.5–4.0, I2 = 4%). Conclusions:An increased troponin measurement after surgery is an independent predictor of mortality, particularly within the first year; limited data suggest an increased creatine kinase muscle and brain isoenzyme measurement also predicts subsequent mortality. Monitoring troponin measurements after noncardiac surgery may allow physicians to better risk stratify and manage their patients.

Prophylaxis Against Deep Vein Thrombosis in Critically Ill Patients With Severe Renal Insufficiency With the Low-Molecular-Weight Heparin Dalteparin An Assessment of Safety and Pharmacodynamics: The DIRECT Study

BACKGROUND Use of low-molecular-weight heparins is avoided in patients with renal insufficiency because of concerns about an excessive anticoagulant effect and increased bleeding risk. To challenge this premise, we evaluated if deep vein thrombosis (DVT) prophylaxis with dalteparin sodium confers an excessive anticoagulant effect in critically ill patients with severe renal insufficiency. METHODS We conducted a multicenter, single-arm clinical trial of DVT prophylaxis with dalteparin sodium, 5000 IU once daily in critically ill patients with a creatinine clearance lower than 30 mL/min (to convert to milliliters per second, multiply by 0.0167). Bioaccumulation was defined by a trough anti-Xa level higher than 0.40 IU/mL, measured twice weekly. The pharmacodynamic properties of dalteparin were assessed by serial anti-Xa levels measured on days 3, 10, and 17. RESULTS We enrolled 156 patients with a mean (SD) creatinine clearance of 18.9 (6.5) mL/min; 18 were excluded because they died or were discharged before testing (n = 3) or had prevalent DVT (n = 15). Of 138 patients included, the median (interquartile range [IQR]) duration of dalteparin exposure was 7 (4-12) days. In 120 patients who had at least 1 trough anti-Xa level (427 total measurements), no patient had bioaccumulation (0%; 95% confidence interval [CI]: 0%-3.0%); the median (IQR) trough anti-Xa level was undetectable (<0.10 IU/mL [<0.10 to <0.10 IU/mL]). Based on serial measurements, peak anti-Xa levels were 0.29 to 0.34 IU/mL and trough levels were lower than 0.06 IU/mL. Deep vein thrombosis occurred in 7 of 138 patients (5.1%; 95% CI, 2.5%-10.1%); major bleeding occurred in 10 patients (7.2%; 95% CI, 4.0%-12.8%), all with trough anti-Xa levels of 0.18 IU/mL or lower. CONCLUSION In critically ill patients with severe renal insufficiency, DVT prophylaxis with dalteparin sodium, 5000 IU once daily, is not associated with an excessive anticoagulant effect due to drug bioaccumulation and is unlikely to contribute to bleeding. TRIAL REGISTRATION clinicaltrials.gov Identifier: NCT00138099.

Fluid Resuscitation in Sepsis: A Systematic Review and Network Meta-analysis

Resuscitation with crystalloids compared with colloids for critically ill patients has been evaluated in large randomized, controlled trials (16) and meta-analyses (713). One meta-analysis(10) including 74 trials reported no difference in mortality between critically ill patients resuscitated with crystalloids and albumin (relative risk [RR], 1.01 [95% CI, 0.93 to 1.10]), hydroxyethyl starch (HES) (RR, 1.10 [CI, 0.91 to 1.32]), gelatin (RR, 0.91 [CI, 0.49 to 1.72]), or dextran (RR, 1.24 [CI, 0.94 to 1.65]). Another meta-analysis (8) reported that resuscitation with an albumin-containing solution in patients with sepsis may decrease mortality compared with solutions containing no albumin (RR, 0.82 [CI, 0.67 to 1.00]). Recent evidence suggests that starches, compared with other fluids and regardless of molecular weight, may be associated with acute kidney injury in the general population of critically ill patients and in those with sepsis (5, 11, 1315). A recent large pragmatic trial comparing colloids (mostly starches) with crystalloids (mostly 0.9% sodium chloride) suggested a 90-day mortality benefit with colloids (RR, 0.92 [CI, 0.86 to 0.99]) (16). Crystalloids can be characterized on the basis of tonicity and electrolyte content. The presence of an organic anion (for example, lactate, acetate, or gluconate) and correspondingly lower chloride content that more closely resembles the composition of plasma suggest that a crystalloid is balanced (for example, Ringer lactate and acetate solutions) (17). The most commonly used crystalloid, normal saline (0.9% sodium chloride), is far from normal, with a pH much less than 7.0 and a supraphysiologic chloride content of 154 mmol/L (18, 19). Compared with a balanced crystalloid solution, normal saline predisposes patients to hyperchloremic metabolic acidosis, decreased renal blood flow to the glomerulus, and impaired smooth-muscle contractility (20). Investigators have not done randomized, controlled trials (RCTs) comparing balanced and unbalanced crystalloids. However, 1 large beforeafter study of critically ill patients showed that balanced versus unbalanced fluid solution was associated with a lower incidence of acute kidney injury (8.4% vs. 14%; P< 0.01) and renal replacement therapy (6.3% vs. 10%; P= 0.05) but no differences in hospital mortality (18). Colloids include natural compounds, such as albumin, and synthetic compounds of HES, gelatin, or dextran. Expansion of plasma volume increases in proportion to the osmotic or oncotic potential, and colloids theoretically require less volume than crystalloids to achieve equivalent hemodynamic effect (19). Limitations of colloids include development of acute kidney injury and coagulation disorders with starches (14) and albumin creates risk for exposure to blood products (19). Another important consideration is the biochemical properties of the crystalloid solution in which the colloid is dissolved. For example, the chloride concentrations in HES may vary between 154 mmol/L (Voluven, Fresenius Kabi) and 118 mmol/L (Tetraspan, B. Braun Medical) (21). Whether any of these fluid properties translate into a survival advantage remains unclear, particularly regarding the optimal fluid for resuscitation in patients with sepsis. Fluid resuscitation, in addition to antibiotics and source control, is a cornerstone of initial management of sepsis (22). However, fluid management in patients with sepsis varies widely in practice (16, 23, 24). Meta-analyses of fluid resuscitation have been limited by not focusing on patients with sepsis (7, 9, 10), not considering electrolyte composition (5, 8, 10, 11), considering only 2 or 3 categories of fluid (25), not including direct and indirect comparisons in the same model, and omission of recent large RCTs(35, 16). Therefore, we did a network meta-analysis (NMA) considering direct and indirect comparisons of all types of fluid resuscitation tested in RCTs in patients with severe sepsis and septic shock, focusing on the effect of these interventions on mortality. Methods Data Sources and Searches This review was done using a predefined protocol. Initially, we searched MEDLINE (1948 to December 2012), EMBASE (1980 to December 2012), ACP Journal Club (1991 to December 2012), the Cochrane Central Register of Controlled Trials, HealthSTAR, the Allied and Complementary Medicine Database, and CINAHL. We updated the MEDLINE and EMBASE searches in August 2013 and March 2014. We screened previously published meta-analyses for relevant citations. Supplement 1 presents the search terms used. Supplement 1. WinBUGS Code for NMA Six reviewers working in 3 pairs screened the titles and abstracts to determine potential eligibility, and entries identified by any reviewer proceeded to the full-text eligibility review. Pretested eligibility forms were used for full-text review, which was also done in duplicate. A third adjudicator helped to resolve disagreements through consensus. Study Selection We selected parallel-group RCTs, including factorial designs, but excluded quasi-randomized and crossover trials. We excluded all studies published by Dr. Joachim Boldt because of suspected lack of integrity (26, 27). We did not apply restrictions on language or publication date. We included studies that involved adult (aged 16 years) critically ill patients with severe sepsis or septic shock as defined by the investigators and who required fluid resuscitation (defined as the administration of a bolus of intravenous fluid exceeding the amount required for maintenance or replacement fluids). We included studies with mixed critically ill populations whenever separate data for patients with sepsis were available. We excluded studies in which most patients had the systemic inflammatory response syndrome secondary to other causes (such as burn, pancreatitis, and trauma) without a clear sepsis subgroup and those focusing on patients after elective surgery. Interventions studied included any fluid or fluid strategy used for resuscitation compared with another fluid or fluid strategy. We excluded studies in which the primary goal was to assess short-term hemodynamic response. Our outcome was 90-day mortality or, if not available, 30-day, intensive care unit, or hospital mortality, whichever was longest. Data Extraction Pairs from the same 6 reviewers abstracted data in duplicate. Another clinician reviewed disagreements, and consensus was reached by discussion. We contacted authors of primary publications for missing or unclear information. Risk of Bias Independently and in duplicate, reviewers assessed risk of bias using a modified version of the Cochrane Collaboration assessment tool (28, 29). We judged each included study as having low, probably low, probably high, or high risk of bias for randomization-sequence generation, randomization concealment, blinding, incomplete data, selective reporting, and free of other bias (including intention-to-treat analysis). The overall rating of risk of bias for each study was the lowest rating for any of the criteria (Appendix Table). Appendix Table. Risk of Bias, by Study Data Synthesis and Analysis Our analysis classified fluids as crystalloids (divided into balanced and unbalanced solutions) and colloids (divided into albumin, gelatin, and low- and high-molecular-weight HES [threshold molecular weight, 150000 kDa]). We considered fluid balanced if it contained an anion of a weak acid (buffer) and its chloride content was correspondingly less than in 0.9% sodium chloride (21). The relevant analyses were a 4-node NMA (crystalloids vs. albumin vs. HES vs. gelatin), a 6-node NMA (crystalloids vs. albumin vs. HES vs. gelatin, with crystalloids divided into balanced or unbalanced and HES divided into low or high molecular weight), and a conventional direct frequentist fixed-effects meta-analytic comparison of crystalloids versus colloids. To calculate direct estimates of treatment effect for each pair of treatments in the 4- and 6-node networks, we did a frequentist fixed-effects meta-analysis. We reported the results as odds ratios (ORs) and corresponding 95% CIs. We evaluated heterogeneity by estimating the variance between studies (chi-square test and I 2 statistic) (30, 31). Using a Bayesian framework, we did 4- and 6-node fixed-effects NMAs for each treatment. We reported the results as ORs and corresponding 95% credibility intervals (CrIs), which are the Bayesian analogue of 95% CIs(32). The ORs reported are relative effects of compared fluids. The models are based on 80000 iterations with a burn-in of 40000 and a thin of 10. We used a random seed and vague priors. We assessed nonconvergence on the basis of BrooksGelmanRubin plots (33). We used the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) approach to assess confidence in estimates of effect (quality of evidence) associated with specific comparisons, including estimates from direct, indirect, and final NMAs (Supplement 2) (34). Our confidence assessment addressed risk of bias, incoherence, imprecision, inconsistency, indirectness, and publication bias. Supplement 2. GRADE Confidence Explanations for All Point Estimates The starting point for confidence in direct and indirect estimates was high. However, indirect estimates were potentially rated down for intransitivity (that is, differences in patients, co-interventions, or settings that could lead to effect modification and thus a misleading comparison of fluid management strategies). We inferred confidence in indirect estimates by examining the connecting loops associated with the particular comparison. The confidence rating chosen was the lowest of the direct estimates contributing to the indirect comparison. For example, consider a comparison of A versus B that is informed by comparisons of A versus C and B versus C. If A versus C was rated as high confidence and B versus C as moderate confidence, the overall indirect confidence rating was ini

Effects of higher- versus lower-protein diets on health outcomes: a systematic review and meta-analysis

Background/Objectives:Numerous randomised controlled trials (RCTs) published in first tier medical journals have evaluated the health effects of diets high in protein. We conducted a rigorous systematic review of RCTs comparing higher- and lower-protein diets.Methods:We searched several electronic databases up to July 2011 for studies focusing on patient-important outcomes (for example, cardiovascular disease) and secondary outcomes such as risk factors for chronic disease (for example, adiposity).Results:We identified 111 articles reporting on 74 trials. Pooled effect sizes using standardised mean differences (SMDs) were small to moderate and favoured higher-protein diets for weight loss (SMD −0.36, 95% confidence interval (CI) −0.56 to −0.17), body mass index (−0.37, CI −0.56 to 0.19), waist circumference (−0.43, CI −0.69 to −0.16), blood pressure (systolic: −0.21, CI −0.32 to −0.09 and diastolic: −0.18, CI −0.29 to −0.06), high-density lipoproteins (HDL 0.25, CI 0.07 to 0.44), fasting insulin (−0.20, CI −0.39 to −0.01) and triglycerides (−0.51, CI −0.78 to −0.24). Sensitivity analysis of studies with lower risk of bias abolished the effect on HDL and fasting insulin, and reduced the effect on triglycerides. We observed nonsignificant effects on total cholesterol, low-density lipoproteins, C-reactive protein, HbA1c, fasting blood glucose, and surrogates for bone and kidney health. Adverse gastrointestinal events were more common with high-protein diets. Multivariable meta-regression analysis showed no significant dose response with higher protein intake.Conclusions:Higher-protein diets probably improve adiposity, blood pressure and triglyceride levels, but these effects are small and need to be weighed against the potential for harms.

Meta-analysis of flavonoids for the treatment of haemorrhoids†

The aim of the study was to evaluate the impact of flavonoids on those symptoms important to patients with symptomatic haemorrhoids.