BMC Medical Research Methodology最新文献

Background: Method comparison studies are conducted to examine the level of agreement between two instruments measuring physiological continuous parameters. The inclusion of repeated measurements in such studies poses additional challenges. The Bland & Altman limits of agreement (LOA) approach has been adapted to account for the correlation between repeated measurements and is widely used in method comparison studies. Alternate statistical methods are not always appropriate for the analysis of such data, and there is a paucity of evidence and guidelines pertaining to statistical methods that inform the analysis of method comparison studies that include repeated measurements. This scoping review aimed to identify methodological publications that propose statistical methods to inform the analysis of method comparison studies that include repeated measurements of continuous clinical data and that may be compared with the LOA method.

Methods: Six online databases were searched from inception to November 2022 using a peer-reviewed search strategy. Searching of grey literature and books, as well as backward citation searching were performed to identify additional sources of evidence. Screening and data abstraction were done by two independent reviewers. Results were synthesized narratively.

Results: Twenty-nine publications were included in this review. Thirty-two independent statistical methods were identified from the included publications, including variants of the LOA method. Four included publications compared findings from different versions of the LOA method. Four different approaches to handling repeated measurements in the context of method comparison studies were identified and were used to group findings from the included publications. Reported strengths and limitations of the LOA method were summarized.

Conclusion: This scoping review provides a synthesis of existing statistical approaches to inform the analysis of method comparison studies with repeated measurements of clinical data, as well as how the various statistical methods perform when compared with various version of the LOA method. Based on the findings, it is generally advisable to consider using adjusted LOAs or modified mixed-effect LOAs in analyzing method comparison studies with repeated measurements.

Trial registration: The protocol was registered on Open Science Framework (https://osf.io/4p8ut).

Background: In public health, epidemiology and medical research, it is considered a relatively effective practice to report treatment effect estimates from stratified regressions alongside significance test results from interaction regressions. However, since the two approaches differ in their theoretical foundations, statistical properties, interpretability, and conditions for appropriate application, combining them in reporting may raise methodological concerns. This study aims to evaluate and compare the effectiveness of these two methods in estimating heterogeneous treatment effects using Monte Carlo simulations.

Methods: We conducted a series of Monte Carlo simulations to compare the performance of interaction regression and stratified regression across various scenarios. These scenarios varied in total sample size, the distribution of subgroup proportions, heterogeneity in the associations between covariates and the outcome variable across subgroups, and the degree of correlation among covariates. We simulated 10 covariates and used a generalised linear model to generate the outcome variable. The performance of both methods was evaluated using mean squared error (MSE), empirical coverage rate (ECR), and empirical statistical power (ESP).

Results: The results demonstrated that stratified regression based on bootstrap resampling effectively controlled Type I error under large sample sizes and balanced group proportions. Interaction regression struggled to control Type I error when the coefficients of baseline characteristics differed across groups, and covariates were only weakly correlated. In other scenarios, the interaction regression outperforms stratified regression, primarily due to its relatively higher statistical power while maintaining acceptable Type I error rates. As an illustrative application, the two methods were applied to real-world data from the International Social Survey Programme (ISSP) to demonstrate how interaction regression and stratified regression may yield different conclusions regarding the association between fruit or vegetable consumption and body mass index.

Conclusion: This study highlights the strengths and limitations of interaction regression and stratified regression in estimating heterogeneous treatment effects. Researchers should avoid adopting a hybrid strategy for reporting subgroup analysis results, such as presenting treatment effect estimates from stratified regressions and significance tests from interaction regressions, because the two methods may exhibit substantial differences in Type I error control under certain conditions.

Starting with the Nuremberg Code in 1947, several guidelines were developed to formulate rules to guide research on humans and safeguard the rights and well-being of subjects participating in clinical research. In recent years, retrospective observational studies based on disease and drug registries, surveillance systems, hospital-based data lakes and platforms, and unstructured data have gained progressively greater attention in the medical literature. Although several guidelines and checklists are currently available to develop and evaluate a protocol for observational studies, issues concerning ethical considerations, data protection and data access have been often ignored. We propose the Data Protection and Good Epidemiologic Standard (DP_GOES) checklist for the development and evaluation of the protocol of observational, retrospective studies based on secondary data. The checklist is divided into four parts, 9 sections and 68 items, and should help to verify whether the study protocol respects the constraints of the regulatory requirements and provisions of data protection authorities, while ensuring that the study may generate robust evidence potentially useful to promote health, supplying more effective healthcare, and guaranteeing system sustainability. The DP_GOES checklist represents a novel and integrative contribution, as it systematically combines epidemiological research standards with data protection principles. Its practical value lies in offering a structured and operational tool that supports both researchers and evaluators in conducting and assessing retrospective observational studies based on secondary data in a rigorous, transparent, and ethically accepted manner.

Background: Integrating artificial intelligence (AI) into literature searching has the potential to enhance research synthesis by improving the identification of conceptually rich or otherwise difficult-to-locate evidence. Theoretical or conceptual literature reviews, including realist reviews, often involve resource-intensive searches because they aim to trace nuanced ideas, mechanisms, or conceptual relationships across multiple sources. This case study illustrates the use of AI-powered tools to support and streamline such literature searching, using a realist review as an example.

Methods: We applied AI tools-Scite and Undermind-in the context of a realist review to facilitate the identification of relevant studies. Seed papers and key informant papers guided the search, and a novel classification system (grandparent, parent, and child papers) was used to systematically organise studies for developing and refining theoretical constructs. Transparent screening procedures and decision-making frameworks were employed to ensure methodological rigour and reproducibility.

Results: The integration of AI tools supported the retrieval of conceptually relevant literature and helped manage complex datasets. The classification system enabled structured organisation of studies, supporting iterative testing and refinement of theoretical constructs. The workflow demonstrated flexibility and adaptability, suggesting potential applicability beyond realist review.

Conclusions: Our findings suggest that AI-powered tools can support literature searching, particularly in identifying conceptually relevant studies. However, these tools do not replace the critical interpretive work required by researchers. Human judgement remains essential to assess relevance, evaluate nuanced concepts, and make informed decisions throughout the search process, with AI serving as a valuable adjunct rather than a substitute.