OBJECTIVE: Reporting bias, prevalent in biomedical fields, can undermine evidence credibility. Our objective was to evaluate the proportion of discrepancies between registered protocols and published manuscripts in randomized controlled trials (RCTs) on exercise interventions for patients with chronic low back pain (CLBP).
METHODS: We conducted a cross-sectional meta-research study, starting from the 2021 \"Exercise therapy for CLBP\" Cochrane Review. We selected all RCTs reporting a protocol registration on a primary register of the World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP) or in ClinicalTrials.gov. We extracted data from both registered protocol and published manuscript of RCTs, collecting recruitment and administrative information (eg, record dates) and details of trial characteristics (eg, outcomes, arms, statistical analysis plan details [SAPs]). Independent pairs of reviewers assessed discrepancies between registered protocol and published manuscript for the reporting of primary and secondary outcomes domains, measurement instruments, time-points, number of arms and SAPs(if attached). Outcome discrepancies were characterized as addition, omission, upgrade or downgrade.
RESULTS: We included 116 RCTs reporting an available protocol registration. Overall, 100 RCTs (86.2%) distinguished between primary and secondary outcomes. Of these, 39 RCTs (39.0%) reported one or more discrepancies in primary outcomes, and 78 RCTs (78.0%) reported one or more discrepancies in secondary outcomes. Focusing on discrepancies for the primary outcome, 64.5% of added, upgraded or downgraded outcomes favored statistically significant effects. Few RCTs (n = 6) reported discrepancies in the number of arms. SAPs were poorly reported in the registered protocols (n = 3) for being compared to the publications.
CONCLUSIONS: We found substantial outcome discrepancies comparing registered protocols and published manuscripts in RCTs assessing exercise interventions for patients with CLBP, with some impacting the statistical significance of the effects. Readers are encouraged to approach RCTs results in this field with caution.

OBJECTIVE: Heart failure is characterized as cardiac dysfunction resulting in elevated cardiac filling pressures with symptoms and signs of congestion. Distinguishing heart failure from other causes of similar presentations in patients with kidney failure is challenging but necessary, and is needed in randomized controlled trials (RCTs) to accurately estimate treatment effects. The objective of this study was to review heart failure events, their diagnostic criteria and adjudication in RCTs of patients with kidney failure treated with dialysis. We hypothesized that heart failure events, diagnostic criteria and adjudication were infrequently reported in RCTs in dialysis.
METHODS: We conducted a meta-epidemiologic systematic review of RCTs from high impact medical, nephrology and cardiology journals from 2000 to 2020. RCTs were eligible if they enrolled adults receiving maintenance dialysis for kidney failure and evaluated any intervention. Results. Of 561 RCTs in patients receiving dialysis, 36 (6.4%) reported heart failure events as primary (10, 27.8%) or secondary (31, 86.1%) outcomes. 10 of the 36 (27.8%) RCTs provided heart failure event diagnostic criteria and 5 of these 10 (50%) adjudicated heart failure events. These 10 RCTs included event diagnostic criteria for heart failure or heart failure hospitalizations, and their criteria included dyspnea (5/10), edema (2/10), rales/crackles (4/10), chest x-ray pulmonary edema or vascular redistribution (4/10), treatment in an acute setting (6/10) and ultrafiltration or dialysis (4/10). No study explicitly distinguished heart failure from volume overload secondary to non-adherence or underdialysis.
CONCLUSIONS: Overall, we found that heart failure events are infrequently reported in RCTs in dialysis and are heterogeneously defined. Further research is required to develop standardized diagnostic criteria that are practical and meaningful to patients and clinicians.

OBJECTIVE: In pursuit of health equity, the World Health Organization has recently called for more extensive monitoring of inequalities in eye health. Population-based eye health surveys can provide this information, but whether underserved groups are considered in the design, implementation, and reporting of surveys is unknown. We conducted a systematic methodological review of surveys published since 2000 to examine how many population-based eye health surveys have considered underserved groups in their design, implementation, or reporting.
METHODS: We identified all population-based cross-sectional surveys reporting the prevalence of objectively measured vision impairment or blindness. Using the PROGRESS + framework to identify underserved groups, we assessed whether each study considered underserved groups within 15 items across the rationale, sampling or recruitment methods, or the reporting of participation and prevalence rates.
RESULTS: 388 eye health surveys were included in this review. Few studies prospectively considered underserved groups during study planning or implementation, for example within their sample size calculations (n = 5, ∼1%) or recruitment strategies (n = 70, 18%). The most common way that studies considered underserved groups was in the reporting of prevalence estimates (n = 374, 96%). We observed a modest increase in the number of distinct PROGRESS + factors considered by a publication over the study period. Gender/sex was considered within at least one item by 95% (n = 367) of studies. Forty-three percent (n = 166) of included studies were conducted primarily on underserved population groups, particularly for subnational studies of people living in rural areas, and we identified examples of robust population-based studies in socially excluded groups.
CONCLUSIONS: More effort is needed to improve the design, implementation, and reporting of surveys to monitor inequality and promote equity in eye health. Ideally, national-level monitoring of vision impairment and service coverage would be supplemented with smaller-scale studies to understand the disparities experienced by the most underserved groups.

OBJECTIVE: To systematically investigate clinical applicability of the current prognostic prediction models for severe postpartum hemorrhage (SPPH).
METHODS: A meta-epidemiological study of prognostic prediction models was conducted for SPPH. A pre-designed structured questionnaire was adopted to extract the study characteristics, predictors and the outcome, modeling methods, predictive performance, the classification ability for high-risk individuals, and clinical use scenarios. The risk of bias among studies was assessed by the Prediction model Risk Of Bias ASsessment Tool (PROBAST).
RESULTS: Twenty-two studies containing 27 prediction models were included. The number of predictors in the final models varied from 3 to 53. However, one-third of the models (11) did not clearly specify the timing of predictor measurement. Calibration was found to be lacking in 10 (37.0%) models. Among the 20 models with an incidence rate of predicted outcomes below 15.0%, none of the models estimated the area under the precision-recall curve, and all reported positive predictive values were below 40.0%. Only two (7.4%) models specified the target clinical setting, while seven (25.9%) models clarified the intended timing of model use. Lastly, all 22 studies were deemed to be at high risk of bias.
CONCLUSIONS: Current SPPH prediction models have limited clinical applicability due to methodological flaws, including unclear predictor measurement, inadequate calibration assessment, and insufficient evaluation of classification ability. Additionally, there is a lack of clarity regarding the timing for model use, target users, and clinical settings. These limitations raise concerns about the reliability and usefulness of these models in real-world clinical practice.

BACKGROUND: Randomization and blinding are generally important in randomized trials. In neonatology, blinding of ventilation strategies is unfeasible if not impossible and we hypothesized that its importance has been overestimated, while the peculiarities of the neonatal patient and the specific outcomes have not been considered.
METHODS: For this meta-epidemiological review, we searched PubMed and Scopus databases in November 2023. We included all meta-analyses focusing on ventilation, published in past 5 years, and reporting either mortality or bronchopulmonary dysplasia (BPD) as an outcome. We extracted the information about how the authors had analyzed risk of bias and evidence certainty.
RESULTS: We screened 494 abstracts and included 40 meta-analyses. Overall, 13 of the 40 reviews assessed blinding properly. Australian and European authors were most likely to perform correct assessment of the blinding (p = 0.03) and the use of RoB 2.0 tool was also associated with proper assessment (p < 0.001). In multivariate regression, the use of RoB 2.0 was the only factor associated with a proper assessment (Beta 0.57 [95% confidence interval: 0.29-0.99]). GRADE ratings were performed in 25 reviews, and the authors downgraded the evidence certainty due to risk of bias in 19 of these and none of these reviews performed the blinding assessment correctly.
CONCLUSIONS: In past neonatal evidence syntheses, the role of blinding has been mostly overestimated, which has led to downgrading of evidence certainty. Objective outcomes (such as mortality and BPD) do not need to be downgraded due to lack of blinding, as the knowledge of the received intervention does not influence the outcome assessment.

OBJECTIVE: To assess to what extent the overall quality of evidence indicates changes to observe intervention effect estimates when new data become available.
METHODS: We conducted a meta-epidemiological study. We obtained evidence from meta-analyses of randomized trials of Cochrane reviews addressing the same health-care question that was updated with inclusion of additional data between January 2016 and May 2021. We extracted the reported effect estimates with 95% confidence intervals (CIs) from meta-analyses and corresponding GRADE (Grading of Recommendations Assessment, Development, and Evaluation) assessments of any intervention comparison for the primary outcome in the first and the last updated review version. We considered the reported overall quality (certainty) of evidence (CoE) and specific evidence limitations (no, serious or very serious for risk of bias, imprecision, inconsistency, and/or indirectness). We assessed the change in pooled effect estimates between the original and updated evidence using the ratio of odds ratio (ROR), absolute ratio of odds ratio (aROR), ratio of standard errors (RoSE), direction of effects, and level of statistical significance.
RESULTS: High CoE without limitations characterized 19.3% (n = 29) out of 150 included original Cochrane reviews. The update with additional data did not systematically change the effect estimates (mean ROR 1.00; 95% CI 0.99-1.02), which deviated 1.06-fold from the older estimates (median aROR; interquartile range [IQR]: 1.01-1.15), gained precision (median RoSE 0.87; IQR 0.76-1.00), and maintained the same direction with the same level of statistical significance in 93% (27 of 29) of cases. Lower CoE with limitations characterized 121 original reviews and graded as moderate CoE in 30.0% (45 of 150), low CoE in 32.0% (48 of 150), and very low CoE in 18.7% (28 of 150) reviews. Their update had larger absolute deviations (median aROR 1.12 to 1.33) and larger gains in precision (median RoSE 0.78-0.86) without clear and consistent differences between these categories of CoE. Changes in effect direction or statistical significance were also more common in the lower quality evidence, again with a similar extent across categories (without change in 75.6%, 64.6%, and 75.0% for moderate, low, very low CoE). As limitations increased, effect estimates deviated more (aROR 1.05 with zero, 1.11 with one, 1.25 with two, 1.24 with three limitations) and changes in direction or significance became more frequent (93.2% stable with no limitations, 74.5% with one, 68.2% with two, and 61.5% with three limitations).
CONCLUSIONS: High-quality evidence without methodological deficiencies is trustworthy and stable, providing reliable intervention effect estimates when updated with new data. Evidence of moderate and lower quality may be equally prone to being unstable and cannot indicate if available effect estimates are true, exaggerated, or underestimated.

OBJECTIVE: To investigate the prevalence of meta-analyses containing potentially redundant randomized controlled trials (RCTs) and the factors associated with the presence of redundancy.
METHODS: This is a cross-sectional study, based on a random sample of references (n = 4500) that were published during 2020 and 2021, indexed in PubMed, Embase, or the Cochrane Database of Systematic Reviews, and retrieved through comprehensive searches using terms about systematic reviews and meta-analysis. From each systematic review, one meta-analysis fulfilling all the following criteria, if available, was included in this study: (1) assessing the effect of the intervention on a primary outcome of the systematic review; (2) combining RCTs only. The primary outcome was prevalence of meta-analyses containing potentially redundant RCTs. Potentially redundant RCTs referred to the trials that started 1 year after the overall effect estimate from cumulative meta-analysis had been statistically robust, as determined by trial sequential analysis when appropriate. The number of potentially redundant trials (if any) in each eligible meta-analysis and the number of participants involved in those trials were documented and contrasted across groups. Logistic regression analysis was conducted to explore the factors associated with presence of potential redundancy.
RESULTS: Of the 448 eligible meta-analyses, 57 (12.7%, 95% confidence interval (CI) 9.8-16.2%) contained potentially redundant RCTs. When limited to the 333 low-heterogeneity meta-analyses, the prevalence was 17.1% (95% CI 13.5-21.5%). The total number of potentially redundant RCTs was 295 (involving 85,385 participants), accounting for 38.5% of the RCTs (and 30.3% of the participants) included in the 57 meta-analyses. In these meta-analyses, the median number of potentially redundant RCTs and the participants involved were 2 (range: 1-50) and 352 (range: 17-26997), respectively. Potentially redundant RCTs were more likely to be present in the meta-analyses evaluating pharmaceutical intervention (odds ratio [OR] 2.31, 95% CI 1.16-4.49), assessing efficacy outcomes (OR 7.25, 95% CI 0.85-61.87), containing more than 5 RCTs (OR 6.47, 95% CI 3.22-12.99), or with the earliest RCT reporting statistically significant effect estimate (OR 5.30, 95% CI 2.64-10.64).
CONCLUSIONS: This study found that 12.7% to 17.1% of recently published meta-analyses contained potentially redundant RCTs, highlighting the importance of conducting or examining systematic reviews of existing evidence to justify new RCTs. More importantly, the study identified some scenarios in which redundancy was more likely to occur and thus has implications for trialists, funding agencies, ethics committees, and journal editors.

To explore the relationships between the risk of bias and treatment effect estimates for exercise therapy interventions on pain intensity and physical functioning outcomes in randomized controlled trials (RCTs) involving patients with chronic low back pain.
A cross-sectional meta-epidemiological study of the 230 RCTs (31,674 participants) in the 2021 \'Exercise therapy for chronic low back pain\' Cochrane Review were included. Study design characteristics, sample size, prospective trial registration, flowchart information, interventions, and comparisons were extracted. Independent pairs of reviewers assessed the risk of bias using the Cochrane Risk of Bias 2 tool.
The metaregression included 220 (pain intensity) and 203 (physical functioning) effect sizes. Unadjusted and adjusted metaregression models showed no significant associations between the bias domains and pain intensity effect sizes. Only domain \'bias in the measurement of the outcome\' was significantly associated with physical functioning (standardized mean difference: -0.40, 95% confidence interval: -0.77 to -0.02) when adjusted for flowchart reported (yes/no), prospective trial registration, sample size, and comparator type.
The risk of bias in the measurement of the outcome could lead to slight overestimates of the effect size for physical functioning. Clinicians should consider this when they read and assess RCT results in this field. We encourage metaresearchers to replicate our findings using a consistent approach for evaluating the risk of bias (i.e., the RoB 2 tool) in other musculoskeletal conditions and interventions to investigate their generalizability.

Preprints became a major source of research communication during the COVID-19 pandemic. We aimed to evaluate whether summary treatment effect estimates differ between preprint and peer-reviewed journal trials.
A meta-epidemiological study. Data were derived from the COVID-NMA living systematic review (covid-nma.com) up to July 20, 2022. We identified all meta-analyses evaluating pharmacological treatments vs. standard of care or placebo for patients with COVID-19 that included at least one preprint and one peer-reviewed journal article. Difference in effect estimates between preprint and peer-reviewed journal trials were estimated by the ratio of odds ratio (ROR); ROR <1 indicated larger effects in preprint trials.
Thirty-seven meta-analyses including 114 trials (44 preprints and 70 peer-reviewed publications) were selected. The median number of randomized controlled trials (RCTs) per meta-analysis was 2 (interquartile range [IQR], 2-4; maximum, 11), median sample size of RCTs was 199 (IQR, 99-478). Overall, there was no statistically significant difference in summary effect estimates between preprint and peer-reviewed journal trials (ROR, 0.88; 95% CI, 0.71-1.09; I2 = 17.8%; τ2 = 0.06).
We did not find an important difference between summary treatment effects of preprints and summary treatment effects of peer-reviewed publications. Systematic reviewers and guideline developers should assess preprint inclusion individually, accounting for risk of bias and completeness of reporting.

We conducted a meta-epidemiological study on all non-specific low back pain (NSLBP) trial registrations on the World Health Organization International Clinical Trials Registry Platform and ClinicalTrials.gov. We aimed to 1) assess the uptake of the core outcome set (COS) for NSLBP in clinical trials; 2) assess the uptake of the core outcome measurement set for NSLBP in clinical trials; and 3) determine whether specific study characteristics are associated with the COS uptake. After applying the relevant filters for the condition, study type, and phase of the trial, 240 registry entries were included in this study. Only 50 (20.8%) entries showed a full COS uptake, and this rate did not increase over time. Most registry entries that planned to measure physical functioning (n = 152) used the Roland-Morris Disability Questionnaire (n = 74; 48.7%); a small percentage used the numeric rating scale (n = 60; 27.3%) or Short Form-12 (n = 5; 8.3%) if they planned to measure pain intensity (n = 220) or health-related quality of life (n = 60), respectively. Only the planned sample size (OR = 1.02; 95% CI = 1.01, 1.03) showed a significant but small association with COS uptake. The uptake of the COS for NSLBP is poor. Only 21% of the randomized controlled trials aimed to measure all COS domains in their study registration and COS uptake is not increased over time. Great heterogeneity in measurement instruments was also observed, revealing poor core outcome measurement set uptake. PERSPECTIVE: The Core Outcome Set (COS) for non-specific low back pain was published more than 20 years ago. We evaluated whether trial registrations are using this set of outcomes when testing interventions for low back pain. Full uptake was found only in 21% of the sample, and this is not increasing over time. Researchers should use the COS to ensure that trials measure relevant outcomes consistently.