The longitudinal relationship between fear of movement and physical activity after cardiac hospitalization: A cross lagged panel model.

BACKGROUND: Little is known about the association between fear of movement (kinesiophobia) and objectively measured physical activity (PA), the first 12 weeks after cardiac hospitalization. PURPOSE: To assess the longitudinal association between kinesiophobia and objectively measured PA and to assess the factor structure of kinesiophobia. METHODS: We performed a longitudinal observational study. PA was continuously measured from hospital discharge to 12 weeks using the Personal Activity Monitor. The PAM measures time spent per day in PA-intensity categories: light, moderate and heavy. Kinesiophobia was assessed with the Tampa Scale for Kinesiophobia (TSK) at four time points (hospital discharge, 3, 6 and 12 weeks). The longitudinal association between PA-intensity and kinesiophobia was studied with a random intercept cross lagged panel model (RI-CLPM). A RI-CLPM estimates effects from kinesiophobia on objectively measured PA and vice versa (cross-over effects), and autoregressive effects (e.g. kinesiophobia from one occasion to the next). RESULTS: In total, 116 patients (83.6% male) with a median age of 65.5 were included in this study. On no occasion did we find an effect of kinesiophobia on PA and vice versa. Model fit for the original model was poor (X2: = 44.646 P<0.001). Best model fit was found for a model were kinesiophobia was modelled as a stable between factor (latent variable) and PA as autoregressive component (dynamic process) (X2 = 27.541 P<0.12). CONCLUSION: Kinesiophobia and objectively measured PA are not associated in the first 12 weeks after hospital discharge. This study shows that kinesiophobia remained relatively stable, 12 weeks after hospital discharge, despite fluctuations in light to moderate PA-intensity.

Automatic categorization of self-acknowledged limitations in randomized controlled trial publications.

OBJECTIVE: Acknowledging study limitations in a scientific publication is a crucial element in scientific transparency and progress. However, limitation reporting is often inadequate. Natural language processing (NLP) methods could support automated reporting checks, improving research transparency. In this study, our objective was to develop a dataset and NLP methods to detect and categorize self-acknowledged limitations (e.g., sample size, blinding) reported in randomized controlled trial (RCT) publications. METHODS: We created a data model of limitation types in RCT studies and annotated a corpus of 200 full-text RCT publications using this data model. We fine-tuned BERT-based sentence classification models to recognize the limitation sentences and their types. To address the small size of the annotated corpus, we experimented with data augmentation approaches, including Easy Data Augmentation (EDA) and Prompt-Based Data Augmentation (PromDA). We applied the best-performing model to a set of about 12K RCT publications to characterize self-acknowledged limitations at larger scale. RESULTS: Our data model consists of 15 categories and 24 sub-categories (e.g., Population and its sub-category DiagnosticCriteria). We annotated 1090 instances of limitation types in 952 sentences (4.8 limitation sentences and 5.5 limitation types per article). A fine-tuned PubMedBERT model for limitation sentence classification improved upon our earlier model by about 1.5 absolute percentage points in F1 score (0.821 vs. 0.8) with statistical significance (p<.001). Our best-performing limitation type classification model, PubMedBERT fine-tuning with PromDA (Output View), achieved an F1 score of 0.7, improving upon the vanilla PubMedBERT model by 2.7 percentage points, with statistical significance (p<.001). CONCLUSION: The model could support automated screening tools which can be used by journals to draw the authors' attention to reporting issues. Automatic extraction of limitations from RCT publications could benefit peer review and evidence synthesis, and support advanced methods to search and aggregate the evidence from the clinical trial literature.

Automatically listing senior members of departments as co-authors is highly prevalent in health sciences: meta-analysis of survey research.

A systematic review with meta-analysis was conducted to assess the prevalence of automatically listing (a) senior member(s) of a department as co-author(s) on all submitted articles in health sciences and the prevalence of degrees of support on a 5-point justification scale. Survey research was searched in PubMed, Lens.org, and Dimensions.ai. until January 5 2023. We assessed the methodological quality of studies and conducted quantitative syntheses. We identified 15 eligible surveys, that provided 67 results, all of which were rated as having low quality. A pooled estimate of 20% [95% CI 16-25] (10 surveys, 3619 respondents) of researchers in various health sciences reported that a senior member of their department was automatically listed as an author on all submitted articles. Furthermore, 28% [95% CI 22-34] of researchers (10 surveys, 2180 respondents) felt that this practice was 'never', 24% [95% CI 22-27] 'rarely', 25% [95% CI 23-28] 'sometimes', 13% [95% CI 9-17] 'most of the time', and 8% [95% CI 6-9] 'always justified'. The practice of automatically assigning senior members of departments as co-authors on all submitted manuscripts may be common in the health sciences; with those admitting to this practice finding it unjustified in most cases.Registration of the protocol The protocol was registered in Open Science Framework. Link: https://osf.io/4eywp/ .

Honorary authorship is highly prevalent in health sciences: systematic review and meta-analysis of surveys.

A systematic review and meta-analysis of survey research was conducted to estimate honorary authorship prevalence in health sciences. We searched PubMed, Lens.org, and Dimensions.ai. until January 5 2023. Methodological quality was assessed and quantitative syntheses were conducted. Nineteen surveys were included and rated as having low methodological quality. We found a pooled prevalence of 26% [95% CI 21-31] (6 surveys, 2758 respondents) of researchers that perceived co-author(s) as honorary on the publication at issue (when they were not referred to any authorship criteria). That prevalence was 18% [95% CI 15-21] (11 surveys, 4272 respondents) when researchers were referred to Committee of Medical Journal Editors (ICMJE) authorship criteria, and 51% [95% CI 47-56] (15 surveys, 5111 respondents) when researchers were asked to declare their co-author(s) contributions on the publication at issue (and these were then compared to ICMJE criteria). 10% of researchers [95% CI 9-12] (11 surveys, 3,663 respondents) reported being approached by others to include honorary author(s) on the publication at issue and 16% [95% CI 13-18] (2 surveys, 823 respondents) admitted adding (an) honorary author(s). Survey research consistently indicates that honorary authorship in the health sciences is highly prevalent, however the quality of the surveys' methods and reporting needs improvement.

Transparency in conducting and reporting research: A survey of authors, reviewers, and editors across scholarly disciplines.

Calls have been made for improving transparency in conducting and reporting research, improving work climates, and preventing detrimental research practices. To assess attitudes and practices regarding these topics, we sent a survey to authors, reviewers, and editors. We received 3,659 (4.9%) responses out of 74,749 delivered emails. We found no significant differences between authors', reviewers', and editors' attitudes towards transparency in conducting and reporting research, or towards their perceptions of work climates. Undeserved authorship was perceived by all groups as the most prevalent detrimental research practice, while fabrication, falsification, plagiarism, and not citing prior relevant research, were seen as more prevalent by editors than authors or reviewers. Overall, 20% of respondents admitted sacrificing the quality of their publications for quantity, and 14% reported that funders interfered in their study design or reporting. While survey respondents came from 126 different countries, due to the survey's overall low response rate our results might not necessarily be generalizable. Nevertheless, results indicate that greater involvement of all stakeholders is needed to align actual practices with current recommendations.

Impact of kinesiophobia on initiation of cardiac rehabilitation: a prospective cohort path analysis.

OBJECTIVES: To identify factors associated with kinesiophobia (fear of movement) after cardiac hospitalisation and to assess the impact of kinesiophobia on cardiac rehabilitation (CR) initiation. DESIGN: Prospective cohort study. SETTING: Academic Medical Centre, Department of Cardiology. PARTICIPANTS: We performed a prospective cohort study in cardiac patients recruited at hospital discharge. In total, 149 patients (78.5% male) with a median age of 65 years were included, of which 82 (59%) were referred for CR. PRIMARY AND SECONDARY OUTCOME MEASURES: We assessed kinesiophobia with the Tampa Scale for Kinesiophobia (TSK). For this study, the total score was used (range 13-52). We assessed baseline factors (demographics, cardiac disease history, questionnaire data on anxiety, biopsychosocial complexity and self-efficacy) associated with kinesiophobia using linear regression with backward elimination. For linear regression, the standardised beta (ß) was reported. Prospectively, the impact of kinesiophobia on probability of CR initiation, in the first 3 months after hospital discharge (subsample referred for CR), was assessed with logistic regression. For logistic regression, the OR was reported. RESULTS: Moderate and severe levels of kinesiophobia were found in 22.8%. In the total sample, kinesiophobia was associated with cardiac anxiety (ß=0.33, 95% CI: 0.19 to 0.48), social complexity (ß=0.23, 95% CI: 0.06 to 0.39) and higher education (ß=-0.18, 95% CI: -0.34 to -0.02). In those referred for CR, kinesiophobia was negatively associated with self-efficacy (ß=-0.29, 95% CI: -0.47 to -0.12) and positively with cardiac anxiety (ß=0.43, 95% CI: 0.24 to 0.62). Kinesiophobia decreased the probability of CR initiation (OR Range13-52 points =0.92, 95% CI: 0.85 to 0.99). CONCLUSION: In patients hospitalised for cardiovascular disease, kinesiophobia is associated with cardiac anxiety, social complexity, educational level and self-efficacy. Kinesiophobia decreased the likelihood of CR initiation with 8% per point on the TSK.

Prevalence of responsible research practices among academics in The Netherlands.

Background: Traditionally, research integrity studies have focused on research misbehaviors and their explanations. Over time, attention has shifted towards preventing questionable research practices and promoting responsible ones. However, data on the prevalence of responsible research practices, especially open methods, open codes and open data and their underlying associative factors, remains scarce. Methods: We conducted a web-based anonymized questionnaire, targeting all academic researchers working at or affiliated to a university or university medical center in The Netherlands, to investigate the prevalence and potential explanatory factors of 11 responsible research practices. Results: A total of 6,813 academics completed the survey, the results of which show that prevalence of responsible practices differs substantially across disciplines and ranks, with 99 percent avoiding plagiarism in their work but less than 50 percent pre-registering a research protocol. Arts and humanities scholars as well as PhD candidates and junior researchers engaged less often in responsible research practices. Publication pressure negatively affected responsible practices, while mentoring, scientific norms subscription and funding pressure stimulated them. Conclusions: Understanding the prevalence of responsible research practices across disciplines and ranks, as well as their associated explanatory factors, can help to systematically address disciplinary- and academic rank-specific obstacles, and thereby facilitate responsible conduct of research.

Is the future of peer review automated?

The rising rate of preprints and publications, combined with persistent inadequate reporting practices and problems with study design and execution, have strained the traditional peer review system. Automated screening tools could potentially enhance peer review by helping authors, journal editors, and reviewers to identify beneficial practices and common problems in preprints or submitted manuscripts. Tools can screen many papers quickly, and may be particularly helpful in assessing compliance with journal policies and with straightforward items in reporting guidelines. However, existing tools cannot understand or interpret the paper in the context of the scientific literature. Tools cannot yet determine whether the methods used are suitable to answer the research question, or whether the data support the authors' conclusions. Editors and peer reviewers are essential for assessing journal fit and the overall quality of a paper, including the experimental design, the soundness of the study's conclusions, potential impact and innovation. Automated screening tools cannot replace peer review, but may aid authors, reviewers, and editors in improving scientific papers. Strategies for responsible use of automated tools in peer review may include setting performance criteria for tools, transparently reporting tool performance and use, and training users to interpret reports.

Association between characteristics of nursing teams and patients' aggressive behavior in closed psychiatric wards.

PURPOSE: Estimate the effect of nursing, shift, and patient characteristics on patients' aggression. DESIGN AND METHODS: Follow-up study on a closed psychiatric ward was performed to estimate the effect of nursing team characteristics and patient characteristics on the incidence of aggression. FINDINGS: The incidence of aggression (n = 802 in sample) was lower in teams with >75% male nurses. Teams scoring high on extraversion experienced more verbal aggression and teams scoring high on neuroticism experienced more physical aggression. Younger patients and/or involuntarily admitted patients were more frequently aggressive. PRACTICE IMPLICATIONS: These findings could stimulate support for nurses to prevent aggression.

Honorary authorship in health sciences: a protocol for a systematic review of survey research.

BACKGROUND: Honorary authorship refers to the practice of naming an individual who has made little or no contribution to a publication as an author. Honorary authorship inflates the output estimates of honorary authors and deflates the value of the work by authors who truly merit authorship. This manuscript presents the protocol for a systematic review that will assess the prevalence of five honorary authorship issues in health sciences. METHODS: Surveys of authors of scientific publications in health sciences that assess prevalence estimates will be eligible. No selection criteria will be set for the time point for measuring outcomes, the setting, the language of the publication, and the publication status. Eligible manuscripts are searched from inception onwards in PubMed, Lens.org , and Dimensions.ai. Two calibrated authors will independently search, determine eligibility of manuscripts, and conduct data extraction. The quality of each review outcome for each eligible manuscript will be assessed with a 14-item checklist developed and piloted for this review. Data will be qualitatively synthesized and quantitative syntheses will be performed where feasible. Criteria for precluding quantitative syntheses were defined a priori. The pooled random effects double arcsine transformed summary event rates of five outcomes on honorary authorship issues with the pertinent 95% confidence intervals will be calculated if these criteria are met. Summary estimates will be displayed after back-transformation. Stata software (Stata Corporation, College Station, TX, USA) version 16 will be used for all statistical analyses. Statistical heterogeneity will be assessed using Tau2 and Chi2 tests and I2 to quantify inconsistency. DISCUSSION: The outcomes of the planned systematic review will give insights in the magnitude of honorary authorship in health sciences and could direct new research studies to develop and implement strategies to address this problem. However, the validity of the outcomes could be influenced by low response rates, inadequate research design, weighting issues, and recall bias in the eligible surveys. SYSTEMATIC REVIEW REGISTRATION: This protocol was registered a priori in the Open Science Framework (OSF) link: https://osf.io/5nvar/ .

Prevalence of questionable research practices, research misconduct and their potential explanatory factors: A survey among academic researchers in The Netherlands.

Prevalence of research misconduct, questionable research practices (QRPs) and their associations with a range of explanatory factors has not been studied sufficiently among academic researchers. The National Survey on Research Integrity targeted all disciplinary fields and academic ranks in the Netherlands. It included questions about engagement in fabrication, falsification and 11 QRPs over the previous three years, and 12 explanatory factor scales. We ensured strict identity protection and used the randomized response method for questions on research misconduct. 6,813 respondents completed the survey. Prevalence of fabrication was 4.3% (95% CI: 2.9, 5.7) and of falsification 4.2% (95% CI: 2.8, 5.6). Prevalence of QRPs ranged from 0.6% (95% CI: 0.5, 0.9) to 17.5% (95% CI: 16.4, 18.7) with 51.3% (95% CI: 50.1, 52.5) of respondents engaging frequently in at least one QRP. Being a PhD candidate or junior researcher increased the odds of frequently engaging in at least one QRP, as did being male. Scientific norm subscription (odds ratio (OR) 0.79; 95% CI: 0.63, 1.00) and perceived likelihood of detection by reviewers (OR 0.62, 95% CI: 0.44, 0.88) were associated with engaging in less research misconduct. Publication pressure was associated with more often engaging in one or more QRPs frequently (OR 1.22, 95% CI: 1.14, 1.30). We found higher prevalence of misconduct than earlier surveys. Our results suggest that greater emphasis on scientific norm subscription, strengthening reviewers in their role as gatekeepers of research quality and curbing the "publish or perish" incentive system promotes research integrity.

Cross-classified multilevel models improved standard error estimates of covariates in clinical outcomes - a simulation study.

OBJECTIVE: To compare estimates of effect and variability resulting from standard linear regression analysis and hierarchical multilevel analysis with cross-classified multilevel analysis under various scenarios. STUDY DESIGN AND SETTING: We performed a simulation study based on a data structure from an observational study in clinical mental health care. We used a Markov chain Monte Carlo approach to simulate 18 scenarios, varying sample sizes, cluster sizes, effect sizes and between group variances. For each scenario, we performed standard linear regression, multilevel regression with random intercept on patient level, multilevel regression with random intercept on nursing team level and cross-classified multilevel analysis. RESULTS: Applying cross-classified multilevel analyses had negligible influence on the effect estimates. However, ignoring cross-classification led to underestimation of the standard errors of the covariates at the two cross-classified levels and to invalidly narrow confidence intervals. This may lead to incorrect statistical inference. Varying sample size, cluster size, effect size and variance had no meaningful influence on these findings. CONCLUSION: In case of cross-classified data structures, the use of a cross-classified multilevel model helps estimating valid precision of effects, and thereby, support correct inferences.

Effects of a transitional care programme on medication adherence in an older cardiac population: A randomized clinical trial.

AIMS: Medication non-adherence post-discharge is common among patients, especially those suffering from chronic medical conditions, and contributes to hospital admissions and mortality. This study aimed to evaluate the effect of the Cardiac Care Bridge (CCB) intervention on medication adherence post-discharge. METHODS: We performed a secondary analysis of the CCB randomized single-blind trial, a study in patients ≥70 years, at high risk of functional loss and admitted to cardiology departments in six hospitals. In this multi-component intervention study, community nurses performed medication reconciliation and observed medication-related problems (MRPs) during post-discharge home visits, and pharmacists provided recommendations to resolve MRPs. Adherence to high-risk medications was measured using the proportion of days covered (PDC), using pharmacy refill data. Furthermore, MRPs were assessed in the intervention group. RESULTS: For 198 (64.7%) of 306 CCB patients, data were available on adherence (mean age: 82 years; 58.9% of patients used a multidose drug dispensing [MDD] system). The mean PDC before admission was 92.3% in the intervention group (n = 99) and 88.5% in the control group (n = 99), decreasing to 85.2% and 84.1% post-discharge, respectively (unadjusted difference: -2.6% (95% CI -9.8 to 4.6, P = .473); adjusted difference -3.3 (95% CI -10.3 to 3.7, P = .353)). Post-hoc analysis indicated that a modest beneficial intervention effect may be restricted to MDD non-users (Pinteraction = .085). In total, 77.0% of the patients had at least one MRP post-discharge. CONCLUSIONS: Our findings indicate that a multi-component intervention, including several components targeting medication adherence in older cardiac patients discharged from hospital back home, did not benefit their medication adherence levels. A modest positive effect on adherence may potentially exist in those patients not using an MDD system. This finding needs replication.

Systematic review and meta-analyses of studies analysing instructions to authors from 1987 to 2017.

To gain insight into changes of scholarly journals' recommendations, we conducted a systematic review of studies that analysed journals' Instructions to Authors (ItAs). We summarised results of 153 studies, and meta-analysed how often ItAs addressed: 1) authorship, 2) conflicts of interest, 3) data sharing, 4) ethics approval, 5) funding disclosure, and 6) International Committee of Medical Journal Editors' Uniform Requirements for Manuscripts. For each topic we found large between-study heterogeneity. Here, we show six factors that explained most of that heterogeneity: 1) time (addressing of topics generally increased over time), 2) country (large differences found between countries), 3) database indexation (large differences found between databases), 4) impact factor (topics were more often addressed in highest than in lowest impact factor journals), 5) discipline (topics were more often addressed in Health Sciences than in other disciplines), and 6) sub-discipline (topics were more often addressed in general than in sub-disciplinary journals).

Influence of nursing staff characteristics on seclusion in acute mental health care-A prospective two-year follow-up study.

INTRODUCTION: Seclusion still occurs on mental health wards, despite absence of therapeutic efficacy and high risks of adverse patient effects. Literature on the effect of nursing teams, and the role of psychological characteristics in particular, on frequency of seclusion is scarce. AIM: To explore the influence of demographic, professional or psychological, nursing team-level, and shift characteristics on the frequency of use of seclusion. METHODS: Prospective two-year follow-up study. RESULTS: We found that the probability of seclusion was lower when nursing teams with at least 75% males were on duty, compared to female only teams, odds ratio (OR = 0.283; 95% CrI 0.046-0.811). We observed a trend indicating that teams scoring higher on the openness personality dimension secluded less, (OR = 0.636; 95% CrI 0.292-1.156). DISCUSSION: Higher proportions of male nurses in teams on duty were associated with lower likelihood of seclusion. We found an indication that teams with a higher mean openness personality trait tended to seclude less. These findings, if causal, could serve as an incentive to reflect on staff mix if circumstances demand better prevention of seclusion.

Personalising add-on treatment with inhaled corticosteroids in patients with chronic obstructive pulmonary disease: a benefit-harm modelling study.

BACKGROUND: Since the benefit-harm balance of adding inhaled corticosteroids to long-acting ß2-agonists (LABA) and long-acting muscarinic antagonists (LAMA) for patients with chronic obstructive pulmonary disease is unclear, we evaluated this addition for a range of patient profiles. METHODS: Analyses considered the effects of low-to-moderate doses of inhaled corticosteroids, LABA, and LAMA compared with LABA and LAMA alone, outcome incidences, and preference weights assigned to averted moderate-to-severe exacerbations (benefit) and severe pneumonia, candidiasis, and dysphonia (harm). Using exponential models, we estimated the preference weight-adjusted 2-year net clinical benefit (ie, benefits outweighing harms) indices. Exacerbation risk thresholds for triggering inhaled corticosteroids, LABA, and LAMA were established when the probability of a 2-year net clinical benefit reached 60%. We estimated the proportion of patients benefiting from added inhaled corticosteroids using an externally validated prediction model for acute exacerbations in primary care. FINDINGS: Adding low-to-moderate dose inhaled corticosteroids to LABA and LAMA provided a net clinical benefit in patients with a 2-year baseline exacerbation risk of 54-83%. Low-dose inhaled corticosteroids showed a net clinical benefit if the baseline risk was 40-91%, but not at higher doses. The benefit was modified by blood eosinophil count (BEC) and age. Although no net benefit was associated with a BEC of less than 150 cells per µL, patients with a BEC of 150 cells per µL or more had a net benefit from low-dose inhaled corticosteroids with a 2-year exacerbation risk of 32-95% in those aged 40-79 years and 41-93% in those older than 80 years. A moderate dose of inhaled corticosteroids showed a net benefit in patients younger than 80 years with a BEC of 150 cells per µL or more at 52-86% 2-year exacerbation risk. Depending on the subgroups, the proportion of patients with a net benefit from added inhaled corticosteroids ranged from 0 to 68%. INTERPRETATION: The net clinical benefit of adding different inhaled corticosteroid doses to LABA and LAMA varies greatly with exacerbation risk, BEC, and age. Personalised treatment decisions based on these factors and predicted exacerbation risks might reduce overtreatment and undertreatment with inhaled corticosteroids. FUNDING: None.

Active Ingredients and Mechanisms of Change in Motivational Interviewing for Smoking Cessation in Patients With Coronary Artery Disease: A Mixed Methods Study.

Background: For patients with coronary artery disease (CAD), smoking is an important risk factor for the recurrence of a cardiovascular event. Motivational interviewing (MI) may increase the motivation of the smokers to stop smoking. Data on MI for smoking cessation in patients with CAD are limited, and the active ingredients and working mechanisms of MI in smoking cessation are largely unknown. Therefore, this study was designed to explore active ingredients and working mechanisms of MI for smoking cessation in smokers with CAD, shortly after a cardiovascular event. Methods: We conducted a qualitative multiple case study of 24 patients with CAD who participated in a randomized trial on lifestyle change. One hundred and nine audio-recorded MI sessions were coded with a combination of the sequential code for observing process exchanges (SCOPE) and the motivational interviewing skill code (MISC). The analysis of the cases consisted of three phases: single case analysis, cross-case analysis, and cross-case synthesis. In a quantitative sequential analysis, we calculated the transition probabilities between the use of MI techniques by the coaches and the subsequent patient statements concerning smoking cessation. Results: In 12 cases, we observed ingredients that appeared to activate the mechanisms of change. Active ingredients were compositions of behaviors of the coaches (e.g., supporting self-efficacy and supporting autonomy) and patient reactions (e.g., in-depth self-exploration and change talk), interacting over large parts of an MI session. The composition of active ingredients differed among cases, as the patient process and the MI-coaching strategy differed. Particularly, change talk and self-efficacy appeared to stimulate the mechanisms of change "arguing oneself into change" and "increasing self-efficacy/confidence." Conclusion: Harnessing active ingredients that target the mechanisms of change "increasing self-efficacy" and "arguing oneself into change" is a good MI strategy for smoking cessation, because it addresses the ambivalence of a patient toward his/her ability to quit, while, after the actual cessation, maintaining the feeling of urgency to persist in not smoking in the patient.

Noncollapsibility and its role in quantifying confounding bias in logistic regression.

BACKGROUND: Confounding bias is a common concern in epidemiological research. Its presence is often determined by comparing exposure effects between univariable- and multivariable regression models, using an arbitrary threshold of a 10% difference to indicate confounding bias. However, many clinical researchers are not aware that the use of this change-in-estimate criterion may lead to wrong conclusions when applied to logistic regression coefficients. This is due to a statistical phenomenon called noncollapsibility, which manifests itself in logistic regression models. This paper aims to clarify the role of noncollapsibility in logistic regression and to provide guidance in determining the presence of confounding bias. METHODS: A Monte Carlo simulation study was designed to uncover patterns of confounding bias and noncollapsibility effects in logistic regression. An empirical data example was used to illustrate the inability of the change-in-estimate criterion to distinguish confounding bias from noncollapsibility effects. RESULTS: The simulation study showed that, depending on the sign and magnitude of the confounding bias and the noncollapsibility effect, the difference between the effect estimates from univariable- and multivariable regression models may underestimate or overestimate the magnitude of the confounding bias. Because of the noncollapsibility effect, multivariable regression analysis and inverse probability weighting provided different but valid estimates of the confounder-adjusted exposure effect. In our data example, confounding bias was underestimated by the change in estimate due to the presence of a noncollapsibility effect. CONCLUSION: In logistic regression, the difference between the univariable- and multivariable effect estimate might not only reflect confounding bias but also a noncollapsibility effect. Ideally, the set of confounders is determined at the study design phase and based on subject matter knowledge. To quantify confounding bias, one could compare the unadjusted exposure effect estimate and the estimate from an inverse probability weighted model.