Tag: pubmed

Am J Epidemiol. 2023 Dec 5:kwad237. doi: 10.1093/aje/kwad237. Online ahead of print.

ABSTRACT

Studies have highlighted the potential importance of modeling interactions for suicide attempt prediction. This case-cohort study identified risk factors for suicide attempts among persons with depression in Denmark using statistical approaches that do (random forests) or do not model interactions (least absolute shrinkage and selection operator regression [LASSO]). Cases made a non-fatal suicide attempt (n = 6,032) between 1995 and 2015. The comparison subcohort was a 5% random sample of all persons in Denmark on January 1, 1995 (n = 11,963). We used random forests and LASSO for sex-stratified prediction of suicide attempts from demographic variables, psychiatric and somatic diagnoses, and treatments. Poisonings, psychiatric disorders, and medications were important predictors for both sexes. Area under the receiver operating characteristic curve (AUC) values were higher in LASSO models (0.85 [95% CI = 0.84, 0.86] in men; 0.89 [95% CI = 0.88, 0.90] in women) than random forests (0.76 [95% CI = 0.74, 0.78] in men; 0.79 [95% CI = 0.78, 0.81] in women). Automatic detection of interactions via random forests did not result in better model performance than LASSO models that did not model interactions. Due to the complex nature of psychiatric comorbidity and suicide, modeling interactions may not always be the optimal statistical approach to enhancing suicide attempt prediction in high-risk samples.

PMID:38055633 | DOI:10.1093/aje/kwad237

JMIR Med Educ. 2023 Dec 6;9:e52202. doi: 10.2196/52202.

ABSTRACT

BACKGROUND: The reliability of GPT-4, a state-of-the-art expansive language model specializing in clinical reasoning and medical knowledge, remains largely unverified across non-English languages.

OBJECTIVE: This study aims to compare fundamental clinical competencies between Japanese residents and GPT-4 by using the General Medicine In-Training Examination (GM-ITE).

METHODS: We used the GPT-4 model provided by OpenAI and the GM-ITE examination questions for the years 2020, 2021, and 2022 to conduct a comparative analysis. This analysis focused on evaluating the performance of individuals who were concluding their second year of residency in comparison to that of GPT-4. Given the current abilities of GPT-4, our study included only single-choice exam questions, excluding those involving audio, video, or image data. The assessment included 4 categories: general theory (professionalism and medical interviewing), symptomatology and clinical reasoning, physical examinations and clinical procedures, and specific diseases. Additionally, we categorized the questions into 7 specialty fields and 3 levels of difficulty, which were determined based on residents’ correct response rates.

RESULTS: Upon examination of 137 GM-ITE questions in Japanese, GPT-4 scores were significantly higher than the mean scores of residents (residents: 55.8%, GPT-4: 70.1%; P<.001). In terms of specific disciplines, GPT-4 scored 23.5 points higher in the “specific diseases,” 30.9 points higher in “obstetrics and gynecology,” and 26.1 points higher in “internal medicine.” In contrast, GPT-4 scores in “medical interviewing and professionalism,” “general practice,” and “psychiatry” were lower than those of the residents, although this discrepancy was not statistically significant. Upon analyzing scores based on question difficulty, GPT-4 scores were 17.2 points lower for easy problems (P=.007) but were 25.4 and 24.4 points higher for normal and difficult problems, respectively (P<.001). In year-on-year comparisons, GPT-4 scores were 21.7 and 21.5 points higher in the 2020 (P=.01) and 2022 (P=.003) examinations, respectively, but only 3.5 points higher in the 2021 examinations (no significant difference).

CONCLUSIONS: In the Japanese language, GPT-4 also outperformed the average medical residents in the GM-ITE test, originally designed for them. Specifically, GPT-4 demonstrated a tendency to score higher on difficult questions with low resident correct response rates and those demanding a more comprehensive understanding of diseases. However, GPT-4 scored comparatively lower on questions that residents could readily answer, such as those testing attitudes toward patients and professionalism, as well as those necessitating an understanding of context and communication. These findings highlight the strengths and limitations of artificial intelligence applications in medical education and practice.

PMID:38055323 | DOI:10.2196/52202

Sci Prog. 2023 Oct-Dec;106(4):368504231215591. doi: 10.1177/00368504231215591.

ABSTRACT

Introduction: Post-adenotonsillectomy (PAT) bleeding, a life-threatening surgical complication, remains unpredictable despite preoperative blood tests. Every surgeon would like predictive markers for this complication of one of the most common procedures performed in pediatric ear, nose, and throat (ENT). Objective: The purpose of the study is to see whether the results of the blood tests we perform routinely preoperatively in children undergoing adenotonsillectomy (AT) (lymphocyte count and percentage, C reactive protein, fibrinogen, or coagulation variables International Normalized Ratio and activated partial thromboplastin time) can potentially predict early post-AT bleeding. Focus has been placed on the presence of relative lymphocytosis (a value of lymphocyte percentage above 55%) in the blood cell count of the patients and its possible connection to postoperative hemorrhage. Method: We conducted an observational retrospective study on 801 children undergoing adenoidectomy, tonsillectomy, or AT over a period of 6 months in our ENT department. Statistical analysis was performed to compare the data. Results: we did not find a statistically significant correlation between preoperative blood markers (coagulation or inflammatory) and early post-AT bleeding. An important blood marker in relation to PAT bleeding appears to be relative lymphocytosis. Relative lymphocytosis has a weak predictive value of early postoperative bleeding in children with AT (sensitivity of only 31.58%, but acceptable specificity of above 80%). In other words, 80% of patients without relative lymphocytosis will not bleed in the first 24 h postoperatively. Children with relative lymphocytosis may need tighter surveillance in the first 24 h after AT. Conclusions: Relative lymphocytosis has a weak predictive value of early postoperative bleeding in children with AT children.

PMID:38055322 | DOI:10.1177/00368504231215591

JMIR Mhealth Uhealth. 2023 Dec 6;11:e46558. doi: 10.2196/46558.

ABSTRACT

BACKGROUND: There is growing interest in enhancing stroke self-management support using mobile health (mHealth) technology (eg, smartphones and apps). Despite this growing interest, “self-management support” is inconsistently defined and applied in the poststroke mHealth intervention literature, which limits efforts to synthesize and compare evidence. To address this gap in conceptual clarity, a scoping review was conducted.

OBJECTIVE: The objectives were to (1) identify and describe the types of poststroke mHealth interventions evaluated using a randomized controlled trial design, (2) determine whether (and how) such interventions align with well-accepted conceptualizations of self-management support (the theory by Lorig and Holman and the Practical Reviews in Self-Management Support [PRISMS] taxonomy by Pearce and colleagues), and (3) identify the mHealth functions that facilitate self-management.

METHODS: A scoping review was conducted according to the methodology by Arksey and O’Malley and Levac et al. In total, 7 databases were searched. Article screening and data extraction were performed by 2 reviewers. The data were analyzed using descriptive statistics and content analysis.

RESULTS: A total of 29 studies (26 interventions) were included. The interventions addressed 7 focal areas (physical exercise, risk factor management, linguistic exercise, activities of daily living training, medication adherence, stroke education, and weight management), 5 types of mobile devices (mobile phones or smartphones, tablets, wearable sensors, wireless monitoring devices, and laptops), and 7 mHealth functions (educating, communicating, goal setting, monitoring, providing feedback, reminding, and motivating). Collectively, the interventions aligned well with the concept of self-management support. However, on an individual basis (per intervention), the alignment was less strong.

CONCLUSIONS: On the basis of the results, it is recommended that future research on poststroke mHealth interventions be more theoretically driven, more multidisciplinary, and larger in scale.

PMID:38055318 | DOI:10.2196/46558

J Med Internet Res. 2023 Dec 6;25:e48145. doi: 10.2196/48145.

ABSTRACT

BACKGROUND: Electronic health records (EHRs) in unstructured formats are valuable sources of information for research in both the clinical and biomedical domains. However, before such records can be used for research purposes, sensitive health information (SHI) must be removed in several cases to protect patient privacy. Rule-based and machine learning-based methods have been shown to be effective in deidentification. However, very few studies investigated the combination of transformer-based language models and rules.

OBJECTIVE: The objective of this study is to develop a hybrid deidentification pipeline for Australian EHR text notes using rules and transformers. The study also aims to investigate the impact of pretrained word embedding and transformer-based language models.

METHODS: In this study, we present a hybrid deidentification pipeline called OpenDeID, which is developed using an Australian multicenter EHR-based corpus called OpenDeID Corpus. The OpenDeID corpus consists of 2100 pathology reports with 38,414 SHI entities from 1833 patients. The OpenDeID pipeline incorporates a hybrid approach of associative rules, supervised deep learning, and pretrained language models.

RESULTS: The OpenDeID achieved a best F₁-score of 0.9659 by fine-tuning the Discharge Summary BioBERT model and incorporating various preprocessing and postprocessing rules. The OpenDeID pipeline has been deployed at a large tertiary teaching hospital and has processed over 8000 unstructured EHR text notes in real time.

CONCLUSIONS: The OpenDeID pipeline is a hybrid deidentification pipeline to deidentify SHI entities in unstructured EHR text notes. The pipeline has been evaluated on a large multicenter corpus. External validation will be undertaken as part of our future work to evaluate the effectiveness of the OpenDeID pipeline.

PMID:38055317 | DOI:10.2196/48145

JMIR Form Res. 2023 Dec 6;7:e47630. doi: 10.2196/47630.

ABSTRACT

BACKGROUND: Adolescents with disabilities experience alarmingly higher rates of depression and isolation than peers without disabilities. There is a need to identify interventions that can improve mental health and isolation among this underserved population. Innovations in virtual reality (VR) gaming “standalone” headsets allow greater access to immersive high-quality digital experiences, due to their relatively low cost.

OBJECTIVE: This study had three purposes, which were to (1) examine the preliminary effects of a low-cost, home-based VR multiplayer recreation and socialization on depression, socialization, and loneliness; (2) quantify the acceptability of the program as measured by participant adherence, total play time, and exercise time; and (3) identify and describe behavioral mechanisms that affected participant engagement.

METHODS: This was a single-group, pre- to postdesign trial. The intervention was conducted at home. Participants were recruited from a children’s hospital. The intervention lasted 4 weeks and included 2×1-hour sessions per week of supervised peer-to-peer gaming. Participants used the Meta Quest 2 headset to meet peers and 2 coaches in a private party held digitally. Aim 1 was evaluated with the Children’s Depression Inventory 2 Short Form and the University of California, Los Angeles Loneliness Scale 20 items, which are measures of social isolation and loneliness, respectively. Aim 2 was evaluated through the following metrics: participant adherence, the types of games played, friendship building and playtime, and program satisfaction and enjoyment.

RESULTS: In total, 12 people enrolled (mean age 16.6, SD 1.8 years; male: n=9 and female: n=3), and 8 people completed the program. Mean attendance for the 8 participants was 77% (49 sessions of 64 total possible sessions; mean 6, SD 2 sessions). A trend was observed for improved Children’s Depression Inventory 2 Short Form scores (mean preintervention score 7.25, SD 4.2; mean postintervention score 5.38, SD 4.1; P=.06; effect size=0.45, 95% CI -0.15 to 3.9), but this was not statistically significant; no difference was observed for University of California, Los Angeles Loneliness Scale 20 items scores. Most participants (7/8, 88%) stated that they became friends with a peer in class; 50% (4/8) reported that they played with other people. Participants reported high levels of enjoyment and satisfaction with how the program was implemented. Qualitative analysis resulted in 4 qualitative themes that explained behavioral mechanisms that determined engagement in the program.

CONCLUSIONS: The study findings demonstrated that a brief VR group program could be valuable for potentially improving mental health among adolescents with physical disabilities. Participants built friendships with peers and other players on the web, using low-cost consumer equipment that provided easy access and strong scale-up potential. Study findings identified factors that can be addressed to enhance the program within a larger clinical trial.

TRIAL REGISTRATION: ClinicalTrials.gov NCT05259462; https://clinicaltrials.gov/study/NCT05259462.

INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): RR2-10.2196/42651.

PMID:38055309 | DOI:10.2196/47630

J Acquir Immune Defic Syndr. 2024 Jan 1;95(1):65-73. doi: 10.1097/QAI.0000000000003312.

ABSTRACT

BACKGROUND: Pregnancy represents a period of high HIV acquisition risk. Safety data for the monthly dapivirine vaginal ring (DVR) during pregnancy are limited. Here, we report data from the first 2 cohorts of pregnant participants in MTN-042/DELIVER, a phase 3b, randomized, open-label safety trial of DVR and oral tenofovir disoproxil fumarate/emtricitabine (TDF/FTC). MTN-042 is being conducted in 3 cohorts beginning with later gestational ages when risks of drug exposure are less.

METHODS: Eligible pregnant individuals aged 18-40 years in Malawi, South Africa, Uganda, and Zimbabwe were randomized 2:1 to monthly DVR or daily TDF/FTC. Participants in cohort 1 initiated product use between 36 weeks 0 days (36 0/7 weeks) and 37 6/7 weeks gestation; participants in cohort 2 initiated product use between 30 0/7 and 35 6/7 weeks gestation. All participants continued product use until delivery or 41 6/7 weeks gestation. Pregnancy outcomes and complications were assessed and summarized using descriptive statistics and compared with local background rates obtained through a separate chart review.

RESULTS: One-hundred and fifty participants were enrolled into cohort 1 with 101 randomized to DVR and 49 to TDF/FTC. One-hundred and fifty-seven participants were enrolled into cohort 2 with 106 randomized to DVR and 51 to TDF/FTC. In both cohorts, pregnancy complications were rare and similar to local background rates.

CONCLUSION: In this first study of a long-acting HIV prevention agent in pregnancy, adverse pregnancy outcomes and complications were uncommon when DVR and TDF/FTC were used in the third trimester of pregnancy, suggesting a favorable safety profile for both prevention products.

PMID:38055292 | DOI:10.1097/QAI.0000000000003312

Cancer. 2023 Dec 6. doi: 10.1002/cncr.35130. Online ahead of print.

ABSTRACT

BACKGROUND: Although the associations between genetic variations and lung cancer risk have been explored, the epigenetic consequences of DNA methylation in lung cancer development are largely unknown. Here, the genetically predicted DNA methylation markers associated with non-small cell lung cancer (NSCLC) risk by a two-stage case-control design were investigated.

METHODS: The genetic prediction models for methylation levels based on genetic and methylation data of 1595 subjects from the Framingham Heart Study were established. The prediction models were applied to a fixed-effect meta-analysis of screening data sets with 27,120 NSCLC cases and 27,355 controls to identify the methylation markers, which were then replicated in independent data sets with 7844 lung cancer cases and 421,224 controls. Also performed was a multi-omics functional annotation for the identified CpGs by integrating genomics, epigenomics, and transcriptomics and investigation of the potential regulation pathways.

RESULTS: Of the 29,894 CpG sites passing the quality control, 39 CpGs associated with NSCLC risk (Bonferroni-corrected p ≤ 1.67 × 10^-6 ) were originally identified. Of these, 16 CpGs remained significant in the validation stage (Bonferroni-corrected p ≤ 1.28 × 10^-3 ), including four novel CpGs. Multi-omics functional annotation showed nine of 16 CpGs were potentially functional biomarkers for NSCLC risk. Thirty-five genes within a 1-Mb window of 12 CpGs that might be involved in regulatory pathways of NSCLC risk were identified.

CONCLUSIONS: Sixteen promising DNA methylation markers associated with NSCLC were identified. Changes of the methylation level at these CpGs might influence the development of NSCLC by regulating the expression of genes nearby.

PLAIN LANGUAGE SUMMARY: The epigenetic consequences of DNA methylation in lung cancer development are still largely unknown. This study used summary data of large-scale genome-wide association studies to investigate the associations between genetically predicted levels of methylation biomarkers and non-small cell lung cancer risk at the first time. This study looked at how well larotrectinib worked in adult patients with sarcomas caused by TRK fusion proteins. These findings will provide a unique insight into the epigenetic susceptibility mechanisms of lung cancer.

PMID:38055287 | DOI:10.1002/cncr.35130

JAMA Surg. 2023 Dec 6. doi: 10.1001/jamasurg.2023.6254. Online ahead of print.

ABSTRACT

IMPORTANCE: A standardized definition and reporting of metabolic bariatric surgery (MBS) outcomes is not available for actual clinical practice and science.

OBJECTIVE: To assess the feasibility of using a clinically relevant and feasible MBS outcome score (Swiss-Finnish Bariatric Metabolic Outcome Score [SF-BARI Score]).

DESIGN, SETTING, AND PARTICIPANTS: This assessment of a bariatric surgery outcome score is based on the secondary analysis of merged 5-year individual patient data (N = 457) of 2 large randomized clinical trials (Swiss SM-BOSS [Swiss Multicenter Bypass or Sleeve Study], conducted from January 2007 to November 2011, and Finnish SLEEVEPASS [Laparoscopic Gastric Bypass Vs Sleeve Gastrectomy to Treat Morbid Obesity], conducted from March 2008 until June 2010) that compared laparoscopic sleeve gastrectomy with laparoscopic Roux-en-Y gastric bypass in the treatment of severe obesity. This secondary analysis was performed from January 2022 to January 2023.

MAIN OUTCOMES AND MEASURES: The main outcome was the feasibility of the SF-BARI Score and the SF-BARI Score QOL (quality of life) as tools to assess MBS outcomes. The score includes percentage of total weight loss (%TWL), 4 obesity-related comorbidities (type 2 diabetes, hypertension, dyslipidemia, and obstructive sleep apnea), complications, and QOL, if available.

RESULTS: Among the 457 included patients, 323 (70.7%) were female and 134 (29.3%) were male, and the mean (SD) age was 45.6 (10.7) years. Outcomes for the SF-BARI Score were available for 435 patients (95.2%) at 1 year and 398 patients (87.1%) at 5 years and for SF-BARI Score QOL in 289 (63.2%) patients at 1 year and 318 patients (69.6%) at 5 years. The SF-BARI Score was correlated with both the SF-BARI Score QOL (r = 0.96; 95% CI, 0.95-0.96; P < .001) and %TWL (r = 0.86; 95% CI, 0.84-0.89; P < .001) and with the Bariatric Analysis and Reporting Outcome System (r = 0.59; 95% CI, 0.51-0.65; P < .001). Score outcomes were categorized as excellent, very good, good, fair, and suboptimal response. There was a statistically significant difference in scores at 1 vs 5 years (4.0; 95% CI, 1.4-6.6; P = .003), and the gastric bypass group had a higher score compared with the sleeve gastrectomy group (7.4; 95% CI, 3.4-11.5; P < .001).

CONCLUSIONS AND RELEVANCE: These findings indicate that this metabolic bariatric surgery outcome score is a simple, relevant, and feasible composite tool to define and measure MBS outcomes, enabling standardized reporting.

TRIAL REGISTRATION: ClinicalTrials.gov Identifiers: NCT00356213 (SM-BOSS) and NCT00793143 (SLEEVEPASS).

PMID:38055284 | DOI:10.1001/jamasurg.2023.6254

JAMA Netw Open. 2023 Dec 1;6(12):e2334757. doi: 10.1001/jamanetworkopen.2023.34757.

ABSTRACT

IMPORTANCE: Incidence of colorectal cancer (CRC) is increasing among younger adults. However, data on precursor lesions in patients who are asymptomatic, especially those aged younger than 50 years, are lacking.

OBJECTIVE: To analyze the prevalence and number needed to screen (NNS) for adenomas, advanced adenomas, and serrated lesions, as well as the incidence of CRC in patients older than age 20 years.

DESIGN, SETTING, AND PARTICIPANTS: This cohort study was conducted among 296 170 patients who received a screening colonoscopy within a national screening colonoscopy registry from 2012 to 2018 in Austria, including 11 103 patients aged younger than 50 years. CRC incidence was analyzed using data from Statistic Austria from 1988 to 2018. Data were analyzed in September 2021.

MAIN OUTCOME AND MEASURES: The prevalence of adenomas and other lesions and the incidence of CRC in individuals aged 20 years or older were assessed.

RESULTS: Among 296 170 patients included in the study (median [IQR] age, 60 [54-68] years; 150 813 females [50.9%]), 11 103 patients (3.7%) were aged younger than 50 years and 285 067 patients (96.3%) were aged 50 years or older. Among patients younger than age 50 years, 1166 individuals (10.5%; NNS = 9) had adenomas and 389 individuals (3.9%; NNS = 26) had at least 1 advanced adenoma, while among those aged 50 years or older, 62 384 individuals (21.9%; NNS = 5) had adenomas and 19 680 individuals (6.9%; NNS = 15) had at least 1 advanced adenoma. Among 1128 males aged 40 to 44 years, 160 individuals (14.2%; NNS = 7) had at least 1 adenoma, and among 1398 females aged 40 to 44 years, 114 individuals (8.1%; NNS = 12) had at least 1 adenoma. The prevalence of adenomas for individuals aged 45 to 49 years vs 50 to 54 years was 490 of 2879 males (17.1%; NNS = 6) vs 8269 of 40 935 males (20.2%; NNS = 5) and 284 of 2792 females (10.2%; NNS = 10) vs 4997 of 40 303 females (12.4%; NNS = 8), respectively. Prevalence of adenomas changed from 61 of 498 individuals (12.4%) in 2008 to 150 of 1064 individuals (14.1%) in 2018 among those younger than 50 years and from 2646 of 12 166 individuals (21.8%) to 10 673 of 37 922 individuals (28.2%) among those aged 50 years and older. The prevalence of advanced adenomas changed from 20 individuals (4.0%) in 2008 to 55 individuals (5.2%) in 2018 in individuals younger than 50 years and from 888 individuals (7.3%) in 2008 to 2578 individuals (6.8%) in 2018 among those aged 50 years and older. Among individuals younger than age 50 years, CRC incidence per 100 000 individuals changed from 9.1 incidents in 1988 to 10.2 incidents in 2018 among males (average annual percentage change [AAPC], 0.5%; 95% CI, 0.1% to 1.0%) and from 9.7 incidents in 1988 to 7.7 incidents in 2018 among females, with a nonsignificant AAPC (-0.2%; 95% CI, -0.7% to 0.3%). Among individuals aged 50 years or older, CRC incidence per 100 000 individuals changed from 168 incidents in 1988 to 97 incidents in 2018 among females (AAPC, -1.8%; 95% CI, -1.9% to -1.6%), and 217 incidents in 1988 to 143 incidents in 2018 among males (AAPC, -1.2%; 95% CI, -1.3% to -1.1%).

CONCLUSION: In this study, CRC incidence decreased after 1988 in Austria among individuals older than 50 years, while among patients younger than 50 years, incidence increased among males but decreased among females. Prevalence of adenomas increased in all age groups, while advanced adenoma prevalence increased among patients younger than 50 years but decreased in patients aged 50 years and older.

PMID:38055281 | DOI:10.1001/jamanetworkopen.2023.34757