SEARCH · Search PubMed

Results for “reliability”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

112 records · Page 7Linked to original sources

PubMed2026

Efficacy of a high-frequency repetitive transcranial magnetic stimulation for craving reduction in adolescents with gaming disorder: a 4-week randomized control trial with 24-week follow-up.

BACKGROUND: With the widespread popularity of online gaming, gaming addiction has come under scrutiny. While there is ongoing research on the diagnosis and treatment of gaming disorder among adolescents, the clinical robustness and reliability of intervention strategies remain uncertain. METHODS: A 4-week, double-blind, randomized, sham-controlled clinical trial was conducted to evaluate the efficacy of noninvasive, high-frequency repetitive transcranial magnetic stimulation (rTMS) in alleviating psychological craving in adolescents diagnosed with gaming disorder. Sham rTMS was administered to the control group using the tilted-coil method. Both groups of participants were treated with SSRI medications. A total of 80 adolescents with gaming disorder participated in this study, and 73 ultimately completed the 24-week follow-up. The primary outcome was the change in craving levels before and after the rTMS intervention, as assessed by the Visual Analogue Scale (VAS). Secondary outcomes included changes in anxiety and depression levels before and after the intervention, as assessed by the HAMA and HAMD. RESULTS: We assessed levels of psychological craving, anxiety, and depression among adolescents with gaming disorder at baseline, after completing a 4-week intervention, and at a 24-week follow-up post-intervention. Our repeated-measures MANOVA results, adjusted for course variables, revealed a significant main effect of rTMS intervention on psychological craving levels in adolescents addicted to online games (F(11, 781) = 11.238, P < 0.001, partial η² = 0.142), as well as significant main effects on time (F(11, 781) = 6.809; P < 0.001, partial η² = 0.091) and group effects (F(1, 71) = 26.707, P < 0.001, partial η² = 0.282). In addition, repeated-measures ANOVA results showed significant time effects for anxiety (F(2, 142) = 20.747, P < 0.001, partial η² = 0.234) and depression levels (F(2, 142) = 22.277, P < 0.001, partial η² = 0.285) among adolescents with gaming disorder, with nonsignificant between-group effects and no intergroup interaction. In the active stimulation group, changes in psychological craving levels after 4 weeks of treatment were significantly and positively correlated with changes in anxiety levels after 4 weeks of treatment in adolescents addicted to online games (r = 0.335, P < 0.05). CONCLUSION: Our findings indicate that high-frequency rTMS targeting the left dorsolateral prefrontal cortex may be a promising approach for reducing psychological craving in adolescents with gaming disorder. TRIAL REGISTRATION: ChiCTR2500102979 in chictr.org.cn, registered on May 22, 2025.

Humans↗

PubMed2026

Artificial intelligence (AI) uses in stereotactic radiosurgery (SRS): diagnosis with brain metastasis (BM) - A systematic review.

BACKGROUND: Brain metastases (BM) are the most common intracranial tumors in adults, and stereotactic radiosurgery (SRS) has become a mainstay of management. However, several diagnostic challenges persist in the SRS pathway, particularly the differentiation of radiation necrosis (RN) from true tumor progression, which conventional MRI and even advanced imaging techniques often cannot reliably resolve. Recent advances in artificial intelligence (AI) offer the potential to address these diagnostic limitations. This systematic review synthesizes current literature on AI applications for MRI-based diagnostic decision support in BM patients undergoing SRS, with a focus on radiomics and deep learning tools for distinguishing RN from progression, classifying molecular and histologic subtypes, and predicting treatment response. METHODS: A systematic review was performed in accordance with PRISMA guidelines. PubMed, Web of Science, and Scopus were searched using a targeted query combining terms related to AI, brain metastasis, diagnosis or imaging, and SRS. After screening 483 records and applying strict inclusion and exclusion criteria, 18 studies published between 2015 and 2025 were included. Data were extracted on study design, cohort characteristics, imaging modality, AI methodology, validation strategy, and reported diagnostic performance. RESULTS: Among the 18 included studies, AI models demonstrated strong performance across diagnostic tasks in the BM-SRS pathway. The differentiation of RN from true tumor progression was the most extensively studied application, addressed by 14 of 18 studies, with reported AUCs ranging from 0.71 to 0.94. Support vector machines, random-forest ensembles, convolutional neural networks, and transformer-based multimodal architectures were widely used. The literature evolved from single-sequence radiomic classifiers in 2018 to multimodal deep learning frameworks fusing imaging with clinical and genomic data in 2025. Contrast-enhanced T1-weighted MRI was the dominant imaging input, and texture-based radiomic features (GLCM, GLSZM, GLDM, and wavelet-derived features) were the most consistently predictive. The highest-performing models reached AUCs of 0.85-0.91 through multimodal integration of imaging with clinical and genomic features, and consistently outperformed expert neuroradiologist read on matched cases. Remaining studies addressed longitudinal segmentation-based detection of local failure and adverse radiation effects, BRAF mutation status in melanoma BM, early Gamma Knife treatment response, and primary tumor histology classification, with more variable performance. CONCLUSION: AI models, particularly those integrating MRI-derived radiomic features with clinical and genomic data, show high accuracy in supporting diagnostic decisions for BM patients treated with SRS. The post-SRS differentiation of radiation necrosis from true tumor progression has reached the greatest level of maturity and is closest to clinical translation, with potential to reduce unnecessary biopsies, personalize surveillance intervals, and rationalize treatment-pathway decisions. Other diagnostic applications, including molecular subtyping and primary tumor histology classification, remain exploratory and require further multicenter validation. Integration of AI tools into multidisciplinary tumor-board workflows, combined with prospective validation and standardized reporting, will be essential to realize the full clinical benefits of AI in SRS for brain metastases.

Humans↗

PubMed2026

Methods for defining equity-stratifying variables: a systematic review of validation studies.

BACKGROUND AND OBJECTIVE: Disease burden is often disproportionally higher among those who are socially disadvantaged by factors defined in the PROGRESS-Plus framework (ie, Place of residence, Race/ethnicity/culture/language, Occupation, Gender/sex, Religion, Education, Socioeconomic status, and Social capital, with "Plus" covering features like age and disability). The accuracy and applicability of case definitions to identify these variables from administrative and clinical health data are unknown. We conducted a systematic review to explore how equity-stratifying variables, as categorized by the PROGRESS-Plus framework, have been defined and validated in epidemiologic studies using administrative health, population-level, or electronic health record (EHR) data. METHODS: Medline, EMBASE, CINAHL, Web of Science, and Google Scholar were searched from the inception of the databases to 2024 for validation studies of equity-stratifying variables in adults using administrative health datasets, health registries, or EHR data. Titles and abstracts, followed by relevant full-text articles, were screened in duplicate by two reviewers for eligibility. The data sources utilized, algorithms employed, and their associated performance measures were extracted and synthesized from included studies. Given substantial heterogeneity in study design, equity-stratifying variable definition, and performance metrics, meta-analysis was not possible. RESULTS: Of the 9099 unique citations screened, 188 full texts were reviewed and 116 were included in this review. Most studies were published between 2019 and 2024 (n = 64, 55%) and were validation studies of race/ethnicity definitions that used race/ethnicity codes or surname list algorithms (n = 66, 57%). No studies examined religion. Regarding the reported performance measure estimates, the race/ethnicity/culture/language equity-stratifying variables category had the largest variability across sensitivity, positive predictive value (PPV), and Cohen's Kappa. Occupation validation studies had the lowest variation in sensitivity and PPV. CONCLUSION: Despite an increasing number of publications reporting on the validation of equity-stratifying variables relevant to the PROGRESS-Plus framework, performance measures varied widely across studies. The significant heterogeneity in equity-stratifying variable definitions and methods used to validate them support the need for further rigorous validation of equity-stratifying variables in administrative and clinical health data. PLAIN LANGUAGE SUMMARY: Disease burden is often higher in people who experience financial hardships, lower level of education, discrimination due to race/ethnicity, and unstable housing. These social factors can be considered health equity factors and are important for understanding health inequalities. Health researchers often use large datasets, such as hospital or electronic health records (EHRs), to study these health equity factors. However, it is not clear how accurately these data sources capture information about people's social circumstances and how these factors are defined. In this study, we reviewed existing research to understand how health equity factors have been defined across health data sources and how accurate they are at measuring aspects of health equity and social disadvantage. Of the more than 9000 studies we identified, we included 116 that met our criteria for this systematic review. Most included studies focused on identifying race and ethnicity, often using codes or surname-based methods. We found that the accuracy of these methods varied widely across studies, meaning results may not always be reliable or comparable. Overall, our findings show that there are inconsistencies in how social factors are defined and measured in health data. This makes it difficult to fully understand and address health inequalities using routinely collected health data. More work is needed to develop and validate better quality and more consistent methods for capturing these important social factors.

Humans↗

PubMed2026

Strategies to improve recruitment to randomised trials.

BACKGROUND: Recruiting participants to randomised controlled trials (RCTs) is challenging. Identifying effective recruitment strategies would benefit health research: poor recruitment leads to underpowered trials, reducing the reliability of findings and increasing the risk of wasted resources, ethical concerns, and trial failure. Evidence to inform recruitment strategies is increasingly generated through Studies Within A Trial (SWATs), which are methodological studies embedded within host RCTs. This is an update of a review last published in 2018. OBJECTIVES: Primary: to quantify the effects of strategies to improve recruitment of participants to RCTs. Secondary: to evaluate recruitment strategies' cost-effectiveness and impact on retention, and the equity, diversity, and inclusion (EDI) characteristics of recruited participants. SEARCH METHODS: We used MEDLINE, Embase, and six other databases to identify the studies included in the review. We also sought unpublished recruitment SWATs through social media and targeted email dissemination to trial methodology networks. The latest search date was 16 February 2023. SELECTION CRITERIA: We included randomised SWATs evaluating trial recruitment strategies embedded in healthcare and non-healthcare trials. We excluded quasi-randomised, hypothetical, questionnaire-only, retention-only, or clinician incentive studies. DATA COLLECTION AND ANALYSIS: Primary outcome: proportion of eligible participants or centres recruited. SECONDARY OUTCOMES: cost-effectiveness, retention rates, and EDI characteristics of included participants. We conducted random-effects meta-analysis for strategies evaluated in at least two studies; otherwise, we synthesised results narratively. We reported effects as risk differences (RDs) with 95% confidence intervals (CIs), and assessed between-trial heterogeneity. We used GRADE to assess the certainty of evidence for the primary outcome. We expressed cost-effectiveness as the incremental cost per additional participant recruited in pounds sterling (GBP). MAIN RESULTS: We identified 91 eligible studies (53 new to this update), providing 94 comparisons and involving at least 176,747 participants. Eighty-one studies involved strategies aimed at trial participants, while 10 evaluated strategies aimed at recruiters. All were healthcare studies. We found 65 recruitment strategies; 49 were evaluated in a single study. Only five strategies were supported by high-certainty evidence according to GRADE criteria, and we focus on these strategies in the summary below. Open-label trials versus blinded, placebo trials. Open-label trials recruited more participants than blinded trials (RD 10%, 95% CI 8% to 12%; 3 studies, 9004 participants), corresponding to approximately 10 additional participants per 100 approached. The studies involved mostly women in the UK and Estonia. No cost or retention data were reported. Telephone reminder versus no telephone reminder. Telephone reminders to people who did not respond to an initial postal invitation boosted recruitment by 6% (95% CI 3% to 9%; 2 studies, 1450 participants), in trials with low underlying recruitment (we are less certain for trials with over 10% recruitment). The studies involved people with a mean age of 58 years in Canada and Norway. No cost or retention data were reported. Recruitment primer letter versus no letter. Pre-recruitment letters and leaflets designed to encourage participation made little or no difference to recruitment (absolute improvement 1%, 95% CI -1% to 2%; 2 studies, 5376 participants), and were associated with increased costs compared to not sending a primer (incremental cost: GBP 2.08). The studies involved mostly older white people in the UK and Ireland. Multimedia information via a digital link/QR code plus paper participant information leaflet (PIL) versus paper PIL alone. This made little or no difference to recruitment (absolute improvement 0%, 95% CI -1% to 1%; 7 studies, 11,612 participants) and retention (absolute improvement 0%, 95% CI -2% to 3%; 5 studies, 7403 participants), and increased costs compared to not including multimedia information (incremental cost: GBP 0.78). The studies involved people in the UK. Optimised, user-tested PIL versus standard PIL. Optimising participant information leaflets (e.g. through user-testing the leaflet with the target population to shape its content, format, and appearance) made little or no difference to recruitment: absolute improvement was 0% (95% CI 0% to 1%; 6 studies, 27,805 participants). The studies involved people in the UK. Only one study reported EDI data; participants were mostly older women. No cost or retention data were reported. We had moderate-certainty evidence for 13 other strategies; confidence was often reduced because the results came from single studies. Seven strategies involved changes to how potential participants received information; four involved changes to trial conduct; one targeted the recruiter or recruitment site; and one tested non-monetary incentives. We had much less confidence in the other 47 comparisons because the studies had design flaws, were single studies, or had very uncertain results. Costs were reported in only 17 of 91 studies. Strategy impact on retention was reported in 15 studies. All but one study (99%) were from high-income countries. The most reported demographics were age (49 studies), sex (32 studies), gender (27 studies), and education level (16 studies). AUTHORS' CONCLUSIONS: The evidence on strategies to improve trial recruitment remains broad but lacks depth. Of 65 strategies evaluated, only five were supported by high-certainty evidence. Open-label trial designs and telephone reminders to non-responders increased recruitment, while optimised participant information leaflets, recruitment primer letters, and multimedia information provided alongside a paper participant information leaflet had little or no effect. Reporting of participant characteristics was poor, limiting assessment of equity, diversity, and inclusion across most studies. Evidence is heavily skewed toward high-income countries. Future research must prioritise evaluations in low-to-middle-income settings and consistently report cost, retention, and EDI outcomes. We strongly urge the methodology research community to strengthen the evidence base by prioritising replications of existing strategies over the development and testing of new ones. FUNDING: National Institute for Health and Care Research (Advanced Fellowship, Adwoa Parker, reference:NIHR302256). Health Research Board, Republic of Ireland, Evidence Synthesis Ireland (grant ESI-2021-001) REGISTRATION: This review updates an earlier Cochrane review, which was first published in 2002 and subsequently updated in 2007, 2010, and 2018. Previous versions of the review and their protocols are available at: https://doi.org/10.1002/14651858.MR000013.pub2 https://doi.org/10.1002/14651858.MR000013.pub3 https://doi.org/10.1002/14651858.MR000013.pub4 https://doi.org/10.1002/14651858.MR000013.pub5 https://doi.org/10.1002/14651858.MR000013.pub6.

Randomized Controlled Trials as Topic↗