Serious reporting deficiencies exist in minimal important difference studies: current state and suggestions for improvement.

BACKGROUND AND OBJECTIVES: To evaluate reporting of minimal important difference (MID) estimates using anchor-based methods for patient-reported outcome measures (PROMs), and the association with reporting deficiencies on their credibility. METHODS: Systematic survey of primary studies empirically estimating MIDs. We searched Medline, EMBASE, PsycINFO, and the Patient-Reported Outcome and Quality of Life Instruments Database until October 2018. We evaluated study reporting, focusing on participants' demographics, intervention(s), characteristics of PROMs and anchors, and MID estimation method(s). We assessed the impact of reporting issues on credibility of MID estimates. RESULTS: In 585 studies reporting on 5,324 MID estimates for 526 distinct PROMs, authors frequently failed to adequately report key characteristics of PROMs and MIDs, including minimum and maximum values of PROM scale, measure of variability accompanying the MID estimate and number of participants included in the MID calculation. Across MID estimates (n = 5,324), the most serious reporting issues impacting credibility included infrequent reporting of the correlation between the anchor and PROM (66%), inadequate details to judge precision of MID point estimate (13%), and insufficient information about the threshold used to ascertain MIDs (16%). CONCLUSION: Serious issues of incomplete reporting in the MID literature threaten the optimal use of MID estimates to inform the magnitude of effects of interventions on PROMs.

Minimal important difference estimates for patient-reported outcomes: A systematic survey.

OBJECTIVES: The objective of the study was to develop an inventory summarizing all anchor-based minimal important difference (MID) estimates for patient-reported outcome measures (PROMs) available in the medical literature. STUDY DESIGN AND SETTING: We searched MEDLINE, EMBASE, CINAHL, PsycINFO, and the Patient-Reported Outcome and Quality of Life Instruments Database internal library (January 1989-October 2018). We included primary studies empirically calculating an anchor-based MID estimate for any PROM in adults and adolescents. Pairs of reviewers independently screened and selected studies, extracted data, and evaluated the credibility of the MIDs. RESULTS: We identified 585 eligible studies, the majority conducted in Europe (n = 211) and North America (n = 179), reporting 5,324 MID estimates for 526 distinct PROMs. Investigators conducted their studies in the context of patients receiving surgical (n = 105, 18%), pharmacological (n = 85, 15%), rehabilitation (n = 65, 11%), or a combination of interventions (n = 194, 33%). Of all MID estimates, 59% (n = 3,131) used a global rating of change anchor. Major credibility limitations included weak correlation (n = 1,246, 23%) or no information regarding the correlation (n = 3,498, 66%) between the PROM and anchor and imprecision in the MID estimate (n = 2,513, 47%). CONCLUSION: A large number of MIDs for assisting in the interpretation of PROMs exist. The MID inventory will facilitate the use of MID estimates to inform the interpretation of the magnitude of treatment effects in clinical research and guideline development.

Evaluating the credibility of anchor based estimates of minimal important differences for patient reported outcomes: instrument development and reliability study.

OBJECTIVE: To develop an instrument to evaluate the credibility of anchor based minimal important differences (MIDs) for outcome measures reported by patients, and to assess the reliability of the instrument. DESIGN: Instrument development and reliability study. DATA SOURCES: Initial criteria were developed for evaluating the credibility of anchor based MIDs based on a literature review (Medline, Embase, CINAHL, and PsycInfo databases) and the experience of the authors in the methodology for estimation of MIDs. Iterative discussions by the team and pilot testing with experts and potential users facilitated the development of the final instrument. PARTICIPANTS: With the newly developed instrument, pairs of masters, doctoral, or postdoctoral students with a background in health research methodology independently evaluated the credibility of a sample of MID estimates. MAIN OUTCOME MEASURES: Core credibility criteria applicable to all anchor types, additional criteria for transition rating anchors, and inter-rater reliability coefficients were determined. RESULTS: The credibility instrument has five core criteria: the anchor is rated by the patient; the anchor is interpretable and relevant to the patient; the MID estimate is precise; the correlation between the anchor and the outcome measure reported by the patient is satisfactory; and the authors select a threshold on the anchor that reflects a small but important difference. The additional criteria for transition rating anchors are: the time elapsed between baseline and follow-up measurement for estimation of the MID is optimal; and the correlations of the transition rating with the baseline, follow-up, and change score in the patient reported outcome measures are satisfactory. Inter-rater reliability coefficients (ĸ) for the core criteria and for one item from the additional criteria ranged from 0.70 to 0.94. Reporting issues prevented the evaluation of the reliability of the three other additional criteria for the transition rating anchors. CONCLUSIONS: Researchers, clinicians, and healthcare policy decision makers can consider using this instrument to evaluate the design, conduct, and analysis of studies estimating anchor based minimal important differences.

Public, health professional and legislator perspectives on the concept of psychiatric disease: a population-based survey.

OBJECTIVE: To assess which mental health-related states of being are perceived as diseases by psychiatrists, non-psychiatric physicians, nurses, parliament members and laypeople. DESIGN AND SETTING: A population-based, mailed survey in Finland. PARTICIPANTS: Respondents from a random sample of 3000 laypeople, 1500 physicians, 1500 nurses and all 200 members of the parliament (MPs) of Finland. PRIMARY OUTCOME MEASURES: Respondents' perspectives on 20 mental health-related states of being as diseases, measuring the extent of agreement with the claim: '[This state of being] is a disease'. RESULTS: Of the 6200 people approached, we received 3259 eligible responses (53%). Two conditions (schizophrenia and autism) were considered to be diseases by at least 75% and two states (grief and homosexuality) were considered not to be diseases by at least 75% in each group. A majority (at least 50% in each group) considered seven states as diseases (anorexia, attention deficit hyperactivity disorder, bulimia, depression, generalised anxiety disorder, panic disorder and personality disorder) and three not to be diseases (absence of sexual desire, premature ejaculation and transsexualism). In six states, there was a wide divergence of opinion (alcoholism, drug addiction, gambling addiction, insomnia, social anxiety disorder and work exhaustion). Psychiatrists were significantly more inclined to considering states of being as diseases relative to other groups, followed by non-psychiatric physicians, nurses, MPs and laypeople. CONCLUSIONS: Respondents agreed that some conditions, such as schizophrenia and autism, are diseases and other states, such as grief and homosexuality, are not; for others, there was considerable disagreement. Psychiatrists are more inclined to consider mental health-related states of being as diseases compared with other physicians, who, in turn, are more inclined than other constituencies. Understanding notions of disease may underlie important debates in public policy and practice in areas of mental health and behaviour, and have implications for resource allocation and stigma.

The Use of Artificially Intelligent Self-Diagnosing Digital Platforms by the General Public: Scoping Review.

BACKGROUND: Self-diagnosis is the process of diagnosing or identifying a medical condition in oneself. Artificially intelligent digital platforms for self-diagnosis are becoming widely available and are used by the general public; however, little is known about the body of knowledge surrounding this technology. OBJECTIVE: The objectives of this scoping review were to (1) systematically map the extent and nature of the literature and topic areas pertaining to digital platforms that use computerized algorithms to provide users with a list of potential diagnoses and (2) identify key knowledge gaps. METHODS: The following databases were searched: PubMed (Medline), Scopus, Association for Computing Machinery Digital Library, Institute of Electrical and Electronics Engineers, Google Scholar, Open Grey, and ProQuest Dissertations and Theses. The search strategy was developed and refined with the assistance of a librarian and consisted of 3 main concepts: (1) self-diagnosis; (2) digital platforms; and (3) public or patients. The search generated 2536 articles from which 217 were duplicates. Following the Tricco et al 2018 checklist, 2 researchers screened the titles and abstracts (n=2316) and full texts (n=104), independently. A total of 19 articles were included for review, and data were retrieved following a data-charting form that was pretested by the research team. RESULTS: The included articles were mainly conducted in the United States (n=10) or the United Kingdom (n=4). Among the articles, topic areas included accuracy or correspondence with a doctor's diagnosis (n=6), commentaries (n=2), regulation (n=3), sociological (n=2), user experience (n=2), theoretical (n=1), privacy and security (n=1), ethical (n=1), and design (n=1). Individuals who do not have access to health care and perceive to have a stigmatizing condition are more likely to use this technology. The accuracy of this technology varied substantially based on the disease examined and platform used. Women and those with higher education were more likely to choose the right diagnosis out of the potential list of diagnoses. Regulation of this technology is lacking in most parts of the world; however, they are currently under development. CONCLUSIONS: There are prominent research gaps in the literature surrounding the use of artificially intelligent self-diagnosing digital platforms. Given the variety of digital platforms and the wide array of diseases they cover, measuring accuracy is cumbersome. More research is needed to understand the user experience and inform regulations.

Examination of psychological risk factors for chronic pain following cardiac surgery: protocol for a prospective observational study.

INTRODUCTION: Approximately 400 000 Americans and 36 000 Canadians undergo cardiac surgery annually, and up to 56% will develop chronic postsurgical pain (CPSP). The primary aim of this study is to explore the association of pain-related beliefs and gender-based pain expectations on the development of CPSP. Secondary goals are to: (A) explore risk factors for poor functional status and patient-level cost of illness from a societal perspective up to 12 months following cardiac surgery; and (B) determine the impact of CPSP on quality-adjusted life years (QALYs) borne by cardiac surgery, in addition to the incremental cost for one additional QALY gained, among those who develop CPSP compared with those who do not. METHODS AND ANALYSES: In this prospective cohort study, 1250 adults undergoing cardiac surgery, including coronary artery bypass grafting and open-heart procedures, will be recruited over a 3-year period. Putative risk factors for CPSP will be captured prior to surgery, at postoperative day 3 (in hospital) and day 30 (at home). Outcome data will be collected via telephone interview at 6-month and 12-month follow-up. We will employ generalised estimating equations to model the primary (CPSP) and secondary outcomes (function and cost) while adjusting for prespecified model covariates. QALYs will be estimated by converting data from the Short Form-12 (version 2) to a utility score. ETHICS AND DISSEMINATION: This protocol has been approved by the responsible bodies at each of the hospital sites, and study enrolment began May 2015. We will disseminate our results through CardiacPain.Net, a web-based knowledge dissemination platform, presentation at international conferences and publications in scientific journals. TRIAL REGISTRATION NUMBER: NCT01842568.

Opioids for Chronic Noncancer Pain: A Systematic Review and Meta-analysis.

Importance: Harms and benefits of opioids for chronic noncancer pain remain unclear. Objective: To systematically review randomized clinical trials (RCTs) of opioids for chronic noncancer pain. Data Sources and Study Selection: The databases of CENTRAL, CINAHL, EMBASE, MEDLINE, AMED, and PsycINFO were searched from inception to April 2018 for RCTs of opioids for chronic noncancer pain vs any nonopioid control. Data Extraction and Synthesis: Paired reviewers independently extracted data. The analyses used random-effects models and the Grading of Recommendations Assessment, Development and Evaluation to rate the quality of the evidence. Main Outcomes and Measures: The primary outcomes were pain intensity (score range, 0-10 cm on a visual analog scale for pain; lower is better and the minimally important difference [MID] is 1 cm), physical functioning (score range, 0-100 points on the 36-item Short Form physical component score [SF-36 PCS]; higher is better and the MID is 5 points), and incidence of vomiting. Results: Ninety-six RCTs including 26 169 participants (61% female; median age, 58 years [interquartile range, 51-61 years]) were included. Of the included studies, there were 25 trials of neuropathic pain, 32 trials of nociceptive pain, 33 trials of central sensitization (pain present in the absence of tissue damage), and 6 trials of mixed types of pain. Compared with placebo, opioid use was associated with reduced pain (weighted mean difference [WMD], -0.69 cm [95% CI, -0.82 to -0.56 cm] on a 10-cm visual analog scale for pain; modeled risk difference for achieving the MID, 11.9% [95% CI, 9.7% to 14.1%]), improved physical functioning (WMD, 2.04 points [95% CI, 1.41 to 2.68 points] on the 100-point SF-36 PCS; modeled risk difference for achieving the MID, 8.5% [95% CI, 5.9% to 11.2%]), and increased vomiting (5.9% with opioids vs 2.3% with placebo for trials that excluded patients with adverse events during a run-in period). Low- to moderate-quality evidence suggested similar associations of opioids with improvements in pain and physical functioning compared with nonsteroidal anti-inflammatory drugs (pain: WMD, -0.60 cm [95% CI, -1.54 to 0.34 cm]; physical functioning: WMD, -0.90 points [95% CI, -2.69 to 0.89 points]), tricyclic antidepressants (pain: WMD, -0.13 cm [95% CI, -0.99 to 0.74 cm]; physical functioning: WMD, -5.31 points [95% CI, -13.77 to 3.14 points]), and anticonvulsants (pain: WMD, -0.90 cm [95% CI, -1.65 to -0.14 cm]; physical functioning: WMD, 0.45 points [95% CI, -5.77 to 6.66 points]). Conclusions and Relevance: In this meta-analysis of RCTs of patients with chronic noncancer pain, evidence from high-quality studies showed that opioid use was associated with statistically significant but small improvements in pain and physical functioning, and increased risk of vomiting compared with placebo. Comparisons of opioids with nonopioid alternatives suggested that the benefit for pain and functioning may be similar, although the evidence was from studies of only low to moderate quality.

GRADE guidelines 17: assessing the risk of bias associated with missing participant outcome data in a body of evidence.

OBJECTIVE: To provide GRADE guidance for assessing risk of bias across an entire body of evidence consequent on missing data for systematic reviews of both binary and continuous outcomes. STUDY DESIGN AND SETTING: Systematic survey of published methodological research, iterative discussions, testing in systematic reviews, and feedback from the GRADE Working Group. RESULTS: Approaches begin with a primary meta-analysis using a complete case analysis followed by sensitivity meta-analyses imputing, in each study, data for those with missing data, and then pooling across studies. For binary outcomes, we suggest use of "plausible worst case" in which review authors assume that those with missing data in treatment arms have proportionally higher event rates than those followed successfully. For continuous outcomes, imputed mean values come from other studies within the systematic review and the standard deviation (SD) from the median SDs of the control arms of all studies. CONCLUSIONS: If the results of the primary meta-analysis are robust to the most extreme assumptions viewed as plausible, one does not rate down certainty in the evidence for risk of bias due to missing participant outcome data. If the results prove not robust to plausible assumptions, one would rate down certainty in the evidence for risk of bias.

Premature Discontinuation of Pediatric Randomized Controlled Trials: A Retrospective Cohort Study.

OBJECTIVES: To determine the proportion of pediatric randomized controlled trials (RCTs) that are prematurely discontinued, examine the reasons for discontinuation, and compare the risk for recruitment failure in pediatric and adult RCTs. STUDY DESIGN: A retrospective cohort study of RCTs approved by 1 of 6 Research Ethics Committees (RECs) in Switzerland, Germany, and Canada between 2000 and 2003. We recorded trial characteristics, trial discontinuation, and reasons for discontinuation from protocols, corresponding publications, REC files, and a survey of trialists. RESULTS: We included 894 RCTs, of which 86 enrolled children and 808 enrolled adults. Forty percent of the pediatric RCTs and 29% of the adult RCTs were discontinued. Slow recruitment accounted for 56% of pediatric RCT discontinuations and 43% of adult RCT discontinuations. Multivariable logistic regression analyses suggested that pediatric RCT was not an independent risk factor for recruitment failure after adjustment for other potential risk factors (aOR, 1.22; 95% CI, 0.57-2.63). Independent risk factors were acute care setting (aOR, 4.00; 95% CI, 1.72-9.31), nonindustry sponsorship (aOR, 4.45; 95% CI, 2.59-7.65), and smaller planned sample size (aOR, 1.05; 95% CI 1.01-1.09, in decrements of 100 participants). CONCLUSION: Forty percent of pediatric RCTs were discontinued prematurely, owing predominately to slow recruitment. Enrollment of children was not an independent risk factor for recruitment failure.

Comparison of the Adverse Event Profile of TheraSphere® with SIR-Spheres® for the Treatment of Unresectable Hepatocellular Carcinoma: A Systematic Review.

To compare the safety profiles of TheraSphere® (glass) and SIR-Spheres® (resin) Y90 microspheres for the treatment of hepatocellular carcinoma. A systematic review was conducted using the databases MEDLINE, Embase, and Cochrane Trials Register to identify all relevant studies. Baseline characteristics and adverse events of all grades related to gastrointestinal, hepatobiliary, and respiratory systems were collected along with commonly reported outcomes related to post-embolization syndrome. For all outcomes, data from each study were tabulated for each intervention. Adverse events and patients were summed across studies on TheraSphere® and SIR-Spheres®, respectively, and the resulting proportion of patients experiencing an outcome for both interventions was calculated. Thirty-one observational studies were included in the review. In the adverse events of all grades, more patients treated with resin microspheres reported gastric ulcers, hepatic encephalopathy, cholecystitis, hepatic failure, and pleural effusion. Patients treated with resin microspheres also had more hepatobiliary adverse events of grade 3 or higher. In the events related to post-embolization syndrome, glass microspheres exhibited a similar safety profile compared to resin microspheres. Ascites and nausea grade 3 or higher were recorded more frequently with glass microsphere treatment. Based on this review of the published literature, glass microspheres exhibit a safety profile with fewer gastrointestinal and pulmonary adverse events compared to resin microspheres in the treatment of hepatocellular carcinoma.

Return-to-work coordination programmes for improving return to work in workers on sick leave.

BACKGROUND: To limit long-term sick leave and associated consequences, insurers, healthcare providers and employers provide programmes to facilitate disabled people's return to work. These programmes include a variety of coordinated and individualised interventions. Despite the increasing popularity of such programmes, their benefits remain uncertain. We conducted a systematic review to determine the long-term effectiveness of return-to-work coordination programmes compared to usual practice in workers at risk for long-term disability. OBJECTIVES: To assess the effects of return-to-work coordination programmes versus usual practice for workers on sick leave or disability. SEARCH METHODS: We searched the Cochrane Central Register of Controlled Trials (CENTRAL; 2016, Issue 11), MEDLINE, Embase, CINAHL and PsycINFO up to 1 November 2016. SELECTION CRITERIA: We included randomised controlled trials (RCTs) that enrolled workers absent from work for at least four weeks and randomly assigned them to return-to-work coordination programmes or usual practice. DATA COLLECTION AND ANALYSIS: Two review authors independently screened titles, abstracts and full-text articles for study eligibility; extracted data; and assessed risk of bias from eligible trials. We contacted authors for additional data where required. We conducted random-effects meta-analyses and used the GRADE approach to rate the quality of the evidence. MAIN RESULTS: We identified 14 studies from nine countries that enrolled 12,568 workers. Eleven studies focused on musculoskeletal problems, two on mental health and one on both. Most studies (11 of 14) followed workers 12 months or longer. Risk of bias was low in 10 and high in 4 studies, but findings were not sensitive to their exclusion.We found no benefits for return-to-work coordination programmes on return-to-work outcomes.For short-term follow-up of six months, we found no effect on time to return to work (hazard ratio (HR) 1.32, 95% confidence interval (CI) 0.93 to 1.88, low-quality evidence), cumulative sickness absence (mean difference (MD) -16.18 work days per year, 95% CI -32.42 to 0.06, moderate-quality evidence), the proportion of participants at work at end of the follow-up (risk ratio (RR) 1.06, 95% CI 0.86 to 1.30, low-quality evidence) or on the proportion of participants who had ever returned to work, that is, regardless of whether they had remained at work until last follow-up (RR 0.87, 95% CI 0.63 to 1.19, very low-quality evidence).For long-term follow-up of 12 months, we found no effect on time to return to work (HR 1.25, 95% CI 0.95 to 1.66, low-quality evidence), cumulative sickness absence (MD -14.84 work days per year, 95% CI -38.56 to 8.88, low-quality evidence), the proportion of participants at work at end of the follow-up (RR 1.06, 95% CI 0.99 to 1.15, low-quality evidence) or on the proportion of participants who had ever returned to work (RR 1.03, 95% CI 0.97 to 1.09, moderate-quality evidence).For very long-term follow-up of longer than 12 months, we found no effect on time to return to work (HR 0.93, 95% CI 0.74 to 1.17, low-quality evidence), cumulative sickness absence (MD 7.00 work days per year, 95% CI -15.17 to 29.17, moderate-quality evidence), the proportion of participants at work at end of the follow-up (RR 0.94, 95% CI 0.82 to 1.07, low-quality evidence) or on the proportion of participants who had ever returned to work (RR 0.95, 95% CI 0.88 to 1.02, low-quality evidence).We found only small benefits for return-to-work coordination programmes on patient-reported outcomes. All differences were below the minimal clinically important difference (MID). AUTHORS' CONCLUSIONS: Offering return-to-work coordination programmes for workers on sick leave for at least four weeks results in no benefits when compared to usual practice. We found no significant differences for the outcomes time to return to work, cumulative sickness absence, the proportion of participants at work at end of the follow-up or the proportion of participants who had ever returned to work at short-term, long-term or very long-term follow-up. For patient-reported outcomes, we found only marginal effects below the MID. The quality of the evidence ranged from very low to moderate across all outcomes.

Minimally Important Differences in Patient or Proxy-Reported Outcome Studies Relevant to Children: A Systematic Review.

CONTEXT: No study has characterized and appraised all anchor-based minimally important differences (MIDs) associated with patient-reported outcome (PRO) instruments in pediatric studies. OBJECTIVE: To complete a comprehensive systematic survey and appraisal of published anchor-based MIDs associated with PRO instruments used in children. DATA SOURCES: Medline, Embase, and PsycINFO (1989 to February 11, 2015). STUDY SELECTION: Studies reporting empirical ascertainment of anchor-based MIDs among PROs used in pediatric care. DATA EXTRACTION: All pertinent data items related to the characteristics of PRO instruments, anchors, and MIDs. RESULTS: Of 4179 unique citations, 30 studies (including 32 cohorts) proved eligible and reported on 28 unique PROs (8 generic, 13 disease-specific, 5 symptoms-specific, 2 function-specific), with 9 (32%) classified as patient-reported, 11 (39%) proxy-reported, and 8 (29%) both patient- and proxy-reported. Of the 30 studies, we rated 14 (44%) as providing highly credible estimates of the MID. Most cohorts (n = 20, 62%) recorded patients' direct response to the target PRO and the use of an independent standard of comparison (n = 25, 78%). Most, however, failed to effectively report measurement properties of the anchor (n = 24, 75%). LIMITATIONS: We have not yet addressed the measurement properties of instrument to measure credibility; our search was restricted to 3 electronic sources, and we used a single data abstractor. CONCLUSIONS: Our study found 28 PROs that have been developed for children, with fewer than half providing credible estimates. Clinicians, clinical trialists, systematic reviewers, and guideline developers seeking to effectively summarize and interpret results of studies addressing PROs in child health are likely to find our comprehensive compendium of MIDs of use, both in providing best estimates of MIDs and identifying credible estimates.

Authors seldom report the most patient-important outcomes and absolute effect measures in systematic review abstracts.

OBJECTIVES: Explicit reporting of absolute measures is important to ensure treatment effects are correctly interpreted. We examined the extent to which authors report absolute effects for patient-important outcomes in abstracts of systematic review (SR). STUDY DESIGN AND SETTING: We searched OVID MEDLINE and Cochrane Database of Systematic Reviews to identify eligible SRs published in the year 2010. Citations were stratified into Cochrane and non-Cochrane reviews, with repeated random sampling in a 1:1 ratio. Paired reviewers screened articles and recorded abstract characteristics, including reporting of effect measures for the most patient-important outcomes of benefit and harm. RESULTS: We included 96 Cochrane and 94 non-Cochrane reviews. About 117 (77.5%) relative measures were reported in abstracts for outcomes of benefit, whereas only 34 (22.5%) absolute measures were reported. Similarly, for outcomes of harm, 41 (87.2%) relative measures were provided in abstracts, compared with only 6 (12.8%) absolute measures. Eighteen (9.5%) abstracts reported both absolute and relative measures for outcomes of benefit, whereas only two (1.1%) abstracts reported both measures for outcomes of harm. Results were similar between Cochrane and non-Cochrane reviews. CONCLUSION: SR abstracts seldom report measures of absolute effect. Journal editors should insist that authors report both relative and absolute effects for patient-important outcomes.

Risk of medication overuse headache across classes of treatments for acute migraine.

BACKGROUND: The most commonly prescribed medications used to treat migraine acutely are single analgesics, ergots, opioids, and triptans. Due to varying mechanisms of action across drug classes, there is reason to believe that some classes may be less likely than others to elicit Medication Overuse Headache (MOH) than others. We therefore aimed to determine whether certain classes of acute migraine drugs are more likely to elicit MOH than others. METHODS: A comprehensive systematic literature was conducted to identify studies of varying designs that reported on MOH within the considered treatment classes. Only studies that reported MOH according to the International Classification of Headache Disorders (ICHD) were considered. Since no causal comparative design studies were identified; data from prevalence studies and surveys were retrieved. Prevalence-based relative risks between treatment classes were calculated by integrating both medication overuse and medication use from published studies. For each pair wise comparison, pooled relative risks were calculated as the inverse variance weighted average. RESULTS: A total of 29 studies informed the relative risk between treatment classes, all of which reported country-specific data. Five studies reported country-specific medication use data. For triptans versus analgesics the study relative risks generally favored triptans. The pooled relative risk was 0.65 (i.e., relative risk reduction of 35 %). For ergots versus analgesics, a similar trend was observed in favor of ergots with a relative risk of 0.41. For triptans versus ergots, the direction of effect was mixed, and the pooled relative risk was 1.07. Both triptans and ergots appeared favorable when compared to opioids, with pooled relative risks of 0.35 and 0.76, respectively. However, the evidence was limited for these comparisons. Analgesics and opioids also appeared to yield similar risk of MOH (pooled relative risk 1.09). CONCLUSION: Our study suggests that in patients receiving acute migraine treatment, analgesics and opioids are associated with a higher risk of developing MOH compared with other treatments. These findings provide incentive for better monitoring of use of analgesics and opioids for treating acute migraine, and suggest possible clinical preference for use of so-called "migraine-specific" treatments, that is, triptans and ergots.

Women's values and preferences and health state valuations for thromboprophylaxis during pregnancy: A cross-sectional interview.

BACKGROUND: Pregnant women with prior venous thromboembolism (VTE) are at risk of recurrence. Prophylaxis with low molecular weight heparin (LWMH) reduces that risk but is inconvenient, costly, and may be associated with increased risks of obstetrical bleeding. The views of pregnant women, crucial when making prophylaxis recommendations, are currently unknown. METHODS: Cross-sectional international multicenter study. We included women with a history of VTE who were either pregnant or planning pregnancy. We provided information regarding risk of VTE recurrence with and without LMWH and determined participant's willingness to receive LMWH prophylaxis through direct choice exercises, preference-elicitation (utilities) for health states (e.g. burden of LMWH prophylaxis), and a probability trade-off exercise. RESULTS: Of 123 women, more women at high risk than those at low risk of recurrence (86.4% vs. 60.0%; p = 0.003) chose to use LMWH. The median threshold reduction in VTE at which women were willing to accept use of LMWH, given a 16% risk of VTE without prophylaxis, was 3% (interquartile range: 1 to 6). Participants' evaluation of the relevant health states varied widely and was unrelated to their direct choices to use or not use LMWH. CONCLUSIONS: Although the majority of women with a previous VTE, pregnant or planning pregnancy choose to take LMWH during pregnancy, a minority ­and in low risk women, a large minority­ do not. Our results highlight the need for individualized shared decision-making (SDM) in the clinical encounter, and for guideline panels to make weak recommendations in favor of LMWH that make clear the need for SDM.

Randomized trials are frequently fragmented in multiple secondary publications.

OBJECTIVE: To assess the frequency and features of secondary publications of randomized controlled trials (RCTs). STUDY DESIGN AND SETTING: For 191 RCTs published in high-impact journals in 2009, we searched for secondary publications coauthored by at least one same author of the primary trial publication. We evaluated the probability of having secondary publications, characteristics of the primary trial publication that predict having secondary publications, types of secondary analyses conducted, and statistical significance of those analyses. RESULTS: Of 191 primary trials, 88 (46%) had a total of 475 secondary publications by 2/2014. Eight trials had >10 (up to 51) secondary publications each. In multivariable modeling, the risk of having subsequent secondary publications increased 1.32-fold (95% CI 1.05-1.68) per 10-fold increase in sample size, and 1.71-fold (95% CI 1.19-2.45) in the presence of a design article. In a sample of 197 secondary publications examined in depth, 193 tested different hypotheses than the primary publication. Of the 193, 43 tested differences between subgroups, 85 assessed predictive factors associated with an outcome of interest, 118 evaluated different outcomes than the original article, 71 had differences in eligibility criteria, and 21 assessed different durations of follow-up; 176 (91%) presented at least one analysis with statistically significant results. CONCLUSIONS: Approximately half of randomized trials in high-impact journals have secondary publications published with a few trials followed by numerous secondary publications. Almost all of these publications report some statistically significant results.

Agreements between Industry and Academia on Publication Rights: A Retrospective Study of Protocols and Publications of Randomized Clinical Trials.

BACKGROUND: Little is known about publication agreements between industry and academic investigators in trial protocols and the consistency of these agreements with corresponding statements in publications. We aimed to investigate (i) the existence and types of publication agreements in trial protocols, (ii) the completeness and consistency of the reporting of these agreements in subsequent publications, and (iii) the frequency of co-authorship by industry employees. METHODS AND FINDINGS: We used a retrospective cohort of randomized clinical trials (RCTs) based on archived protocols approved by six research ethics committees between 13 January 2000 and 25 November 2003. Only RCTs with industry involvement were eligible. We investigated the documentation of publication agreements in RCT protocols and statements in corresponding journal publications. Of 647 eligible RCT protocols, 456 (70.5%) mentioned an agreement regarding publication of results. Of these 456, 393 (86.2%) documented an industry partner's right to disapprove or at least review proposed manuscripts; 39 (8.6%) agreements were without constraints of publication. The remaining 24 (5.3%) protocols referred to separate agreement documents not accessible to us. Of those 432 protocols with an accessible publication agreement, 268 (62.0%) trials were published. Most agreements documented in the protocol were not reported in the subsequent publication (197/268 [73.5%]). Of 71 agreements reported in publications, 52 (73.2%) were concordant with those documented in the protocol. In 14 of 37 (37.8%) publications in which statements suggested unrestricted publication rights, at least one co-author was an industry employee. In 25 protocol-publication pairs, author statements in publications suggested no constraints, but 18 corresponding protocols documented restricting agreements. CONCLUSIONS: Publication agreements constraining academic authors' independence are common. Journal articles seldom report on publication agreements, and, if they do, statements can be discrepant with the trial protocol.

Glucagon-like peptide-1 receptor agonists and heart failure in type 2 diabetes: systematic review and meta-analysis of randomized and observational studies.

BACKGROUND: The effect of glucagon-like peptide-1(GLP-1) receptor agonists on heart failure remains uncertain. We therefore conducted a systematic review to assess the possible impact of GLP-1 agonists on heart failure or hospitalization for heart failure in patients with type 2 diabetes. METHODS: We searched MEDLINE, EMBASE, the Cochrane Central Register of Controlled Trials (CENTRAL) and ClinicalTrials.gov to identify randomized controlled trials (RCTs) and observational studies that addressed the effect of GLP-1 receptor agonists in adults with type 2 diabetes, and explicitly reported heart failure or hospitalization for heart failure. Two paired reviewers screened reports, collected data, and assessed the risk of bias. We pooled data from RCTs and observational studies separately, and used the GRADE approach to rate the quality of evidence. RESULTS: We identified 25 studies that were eligible for our review; 21 RCTs (n = 18,270) and 4 observational studies (n = 111,029). Low quality evidence from 20 RCTs suggested, if anything, a lower incidence of heart failure between GLP-1 agonists versus control (17/7,441 vs. 19/4,317; odds ratio (OR) 0.62, 95 % confidence interval (CI) 0.31 to 1.22; risk difference (RD) 19 fewer, 95 % CI 34 fewer to 11 more per 1000 over 5 years). Three cohort studies comparing GLP-1 agonists to alternative agents provided very low quality evidence that GLP-1 agonists do not increase the incidence of heart failure. One RCT provided moderate quality evidence that GLP-1 agonists were not associated with hospitalization for heart failure (lixisenatide vs placebo: 122/3,034 vs. 127/3,034; adjusted hazard ratio 0.96, 95 % CI 0.75 to 1.23; RD 4 fewer, 95 % CI 25 fewer to 23 more per 1000 over 5 years) and a case-control study provided very low quality evidence also suggesting no association (GLP-1 agonists vs. other anti-hyperglycemic drugs: 1118 cases and 17,626 controls, adjusted OR 0.67, 95 % CI 0.32 to 1.42). CONCLUSIONS: The current evidence suggests that GLP-1 agonists do not increase the risk of heart failure or hospitalization for heart failure among patients with type 2 diabetes.

Three challenges described for identifying participants with missing data in trials reports, and potential solutions suggested to systematic reviewers.

OBJECTIVE: To categorize the challenges in determining the extent of missing participant data in randomized trials and suggest potential solutions for systematic review authors. STUDY DESIGN AND SETTING: During the process of updating a series of Cochrane systematic reviews on the topic of anticoagulation in patients with cancer, we identified challenges and used an iterative approach to improve, and a consensus process to agree on the challenges identified, and to suggest potential ways of dealing with them. The five systematic reviews included 58 trials and 75 meta-analyses for patient-important dichotomous outcomes with 27,037 randomized participants. RESULTS: We identified three categories of challenges: (1) Although systematic reviewers require information about missing data to be reported by outcome, trialists typically report the information by participant; (2) It is not always clear whether the trialists followed up participants in certain categories (e.g., noncompliers), that is, whether some categories of participants did or did not have missing data; (3) It is not always clear how the trialists dealt with missing data in their analysis (e.g., exclusion from the denominator vs. assumptions made for the numerator). We discuss potential solutions for each one of these challenges and suggest further research work. CONCLUSION: Current reporting of missing data is often not explicit and transparent, and although our potential solutions to problems of suboptimal reporting may be helpful, reliable and valid characterization of the extent and nature of missing data remains elusive. Reporting of missing data in trials needs further improvement.

Dipeptidyl peptidase-4 inhibitors and risk of heart failure in type 2 diabetes: systematic review and meta-analysis of randomised and observational studies.

OBJECTIVES: To examine the association between dipeptidyl peptidase-4 (DPP-4) inhibitors and the risk of heart failure or hospital admission for heart failure in patients with type 2 diabetes. DESIGN: Systematic review and meta-analysis of randomised and observational studies. DATA SOURCES: Medline, Embase, Cochrane Central Register of Controlled Trials, and ClinicalTrials.gov searched up to 25 June 2015, and communication with experts. ELIGIBILITY CRITERIA: Randomised controlled trials, non-randomised controlled trials, cohort studies, and case-control studies that compared DPP-4 inhibitors against placebo, lifestyle modification, or active antidiabetic drugs in adults with type 2 diabetes, and explicitly reported the outcome of heart failure or hospital admission for heart failure. DATA COLLECTION AND ANALYSIS: Teams of paired reviewers independently screened for eligible studies, assessed risk of bias, and extracted data using standardised, pilot tested forms. Data from trials and observational studies were pooled separately; quality of evidence was assessed by the GRADE approach. RESULTS: Eligible studies included 43 trials (n=68,775) and 12 observational studies (nine cohort studies, three nested case-control studies; n=1,777,358). Pooling of 38 trials reporting heart failure provided low quality evidence for a possible similar risk of heart failure between DPP-4 inhibitor use versus control (42/15,701 v 33/12,591; odds ratio 0.97 (95% confidence interval 0.61 to 1.56); risk difference 2 fewer (19 fewer to 28 more) events per 1000 patients with type 2 diabetes over five years). The observational studies provided effect estimates generally consistent with trial findings, but with very low quality evidence. Pooling of the five trials reporting admission for heart failure provided moderate quality evidence for an increased risk in patients treated with DPP-4 inhibitors versus control (622/18,554 v 552/18,474; 1.13 (1.00 to 1.26); 8 more (0 more to 16 more)). The pooling of adjusted estimates from observational studies similarly suggested (with very low quality evidence) a possible increased risk of admission for heart failure (adjusted odds ratio 1.41, 95% confidence interval 0.95 to 2.09) in patients treated with DPP-4 inhibitors (exclusively sitagliptin) versus no use. CONCLUSIONS: The relative effect of DPP-4 inhibitors on the risk of heart failure in patients with type 2 diabetes is uncertain, given the relatively short follow-up and low quality of evidence. Both randomised controlled trials and observational studies, however, suggest that these drugs may increase the risk of hospital admission for heart failure in those patients with existing cardiovascular diseases or multiple risk factors for vascular diseases, compared with no use.