Category: Statistics

JAMA Netw Open. 2024 Nov 4;7(11):e2448714. doi: 10.1001/jamanetworkopen.2024.48714.

ABSTRACT

IMPORTANCE: Understanding the association of artificial intelligence (AI) with physician burnout is crucial for fostering a collaborative interactive environment between physicians and AI.

OBJECTIVE: To estimate the association between AI use in radiology and radiologist burnout.

DESIGN, SETTING, AND PARTICIPANTS: This cross-sectional study conducted a questionnaire survey between May and October 2023, using the national quality control system of radiology in China. Participants included radiologists from 1143 hospitals. Radiologists reporting regular or consistent AI use were categorized as the AI group. Statistical analysis was performed from October 2023 to May 2024.

EXPOSURE: AI use in radiology practice.

MAIN OUTCOMES AND MEASURES: Burnout was defined by emotional exhaustion (EE) or depersonalization according to the Maslach Burnout Inventory. Workload was assessed based on working hours, number of image interpretations, hospital level, device type, and role in the workflow. AI acceptance was determined via latent class analysis considering AI-related knowledge, attitude, confidence, and intention. Propensity score-based mixed-effect generalized linear logistic regression was used to estimate the associations between AI use and burnout and its components. Interactions of AI use, workload, and AI acceptance were assessed on additive and multiplicative scales.

RESULTS: Among 6726 radiologists included in this study, 2376 (35.3%) were female and 4350 (64.7%) were male; the median (IQR) age was 41 (34-48) years; 3017 were in the AI group (1134 [37.6%] female; median [IQR] age, 40 [33-47] years) and 3709 in the non-AI group (1242 [33.5%] female; median [IQR] age, 42 [34-49] years). The weighted prevalence of burnout was significantly higher in the AI group compared with the non-AI group (40.9% vs 38.6%; P < .001). After adjusting for covariates, AI use was significantly associated with increased odds of burnout (odds ratio [OR], 1.20; 95% CI, 1.10-1.30), primarily driven by its association with EE (OR, 1.21; 95% CI, 1.10-1.34). A dose-response association was observed between the frequency of AI use and burnout (P for trend < .001). The associations were more pronounced among radiologists with high workload and lower AI acceptance. A significant negative interaction was noted between high AI acceptance and AI use.

CONCLUSIONS AND RELEVANCE: In this cross-sectional study of radiologist burnout, frequent AI use was associated with an increased risk of radiologist burnout, particularly among those with high workload or lower AI acceptance. Further longitudinal studies are needed to provide more evidence.

PMID:39576636 | DOI:10.1001/jamanetworkopen.2024.48714

Invest Ophthalmol Vis Sci. 2024 Nov 4;65(13):48. doi: 10.1167/iovs.65.13.48.

ABSTRACT

PURPOSE: The purpose of this study was to assess fine motor skills and reading proficiency in adults with amblyopia and/or strabismus, and to determine how these relate to clinical measures of vision and self-reported vision-related quality of life.

METHODS: Fine motor skills (Manual dexterity – Bruininks-Oseretsky Test of Motor Proficiency [BOT-2]) and reading performance (International Reading Speed Texts [IReST]) were assessed in 23 adults with non-strabismic amblyopia, 20 with non-amblyopic strabismus, 52 with both amblyopia and strabismus, and 19 with normal visual development. Visual acuity and binocular function score (BFS), obtained from stereoacuity and presence/absence of suppression, were also determined. Vision-related quality of life was assessed with the Amblyopia and Strabismus Questionnaire (A&SQ) in those with amblyopia and/or strabismus. Statistical analysis included ANOVA and multiple regression models.

RESULTS: Participants with amblyopia and/or strabismus exhibited poorer performance in all five manual dexterity sub-items and the overall standardized score (P < 0.05). The reading rate was significantly slower across all amblyopia/strabismus groups (P < 0.05). Poorer fine motor skills and slower reading performance were associated with each other (R = 0.29). Clinical visual characteristics (visual acuity [VA], BFS, and presence of strabismus) explained 39% of the variance in fine motor skills score (R2 = 0.39), however, these explained only 6% of the variance in reading speed (R2 = 0.06). Self-report of functional ability related most to BFS, whereas psychosocial impact related to the presence of strabismus. The clinical and functional characteristics predicted 4% of the variance in functional impact score (R2 = 0.038) and explained 16% of the variance in psychosocial impact score (R2 = 0.16).

CONCLUSIONS: The functional and psychosocial effects of amblyopia and strabismus are common and persist into adulthood, with outcomes inadequately accounted for by clinical measures of vision.

PMID:39576625 | DOI:10.1167/iovs.65.13.48

JAMA Health Forum. 2024 Nov 1;5(11):e244156. doi: 10.1001/jamahealthforum.2024.4156.

ABSTRACT

IMPORTANCE: Telemedicine is an increasingly used yet understudied vehicle to deliver pediatric primary care. Evidence detailing downstream health care utilization after telemedicine visits is needed.

OBJECTIVE: To compare pediatric primary care conducted via telemedicine (video or telephone) with in-person office visits with regard to physician medication prescribing and imaging and laboratory ordering and downstream follow-up office visits, emergency department (ED) visits, and hospitalizations.

DESIGN, SETTING, AND PARTICIPANTS: This cohort study included all patients younger than 18 years who had scheduled primary care appointments with a pediatrician from January 1 to December 31, 2022, in the Kaiser Permanente Northern California health system, a large integrated health care delivery system offering in-person office visits, video visits, or telephone visits for pediatric primary care.

EXPOSURE: Pediatric primary care in-person visit, telephone visit, or video visit.

MAIN OUTCOME AND MEASURES: Rates of physician medication prescribing and imaging and laboratory ordering during an index telemedicine or office visit and rates of in-person office visits, ED visits, and hospitalizations within 7 days after the visit, adjusted for patient and clinical characteristics.

RESULTS: Of 782 596 total appointments (51.1% male) among 438 638 patients, telemedicine was used for 332 153 visits (42.4%). After adjustment, there was more medication prescribing for in-person visits (39.8%) compared with video visits (29.5%; adjusted difference, -10.3%; 95% CI, -10.6% to -10.0%) or telephone visits (27.3%; adjusted difference, -12.5%; 95% CI, -12.5% to -12.7%). There was also more laboratory ordering for in-person visits (24.6%) compared with video visits (7.8%; adjusted difference, -16.8%; 95% CI, -17.0% to -16.6%) or telephone visits (8.5%; adjusted difference, -16.2%; 95% CI, -16.3% to -16.0%). There was more imaging ordering for in-person visits (8.5%) compared with video visits (4.0%; adjusted difference, -4.5%; 95% CI, -4.6% to -4.4%) and telephone visits (3.5%; adjusted difference, -5.0%; 95% CI, -5.1% to -4.9%). After adjustment, fewer in-person follow up visits occurred for index visits that were in-person (4.3%) compared with video (14.4%; adjusted difference, 10.1%; 95% CI, 9.9%-10.3%) or telephone (15.1%; adjusted difference, 10.8%; 95% CI, 10.7%-11.0%) visits. The rate of ED visits following an in-person visit was slightly lower (1.75%) compared with after video visits (2.04%; adjusted difference, 0.29%; 95% CI, 0.21%-0.38%) or telephone visits (2.00%; adjusted difference, 0.25%; 95% CI, 0.18%-0.33%). There was no statistically significant difference in the 7-day rate of hospitalizations.

CONCLUSIONS AND RELEVANCE: In this cohort study, telephone and video visits for pediatric primary care were associated with less prescribing and ordering than in-person visits. Telemedicine visits were associated with modestly higher rates of subsequent in-person visits and slightly higher rates of ED visits, and there was no difference in hospitalizations. Telemedicine appears to be a useful vehicle for health care delivery in the pediatric population, although it is not a universal substitute for in-person visits.

PMID:39576615 | DOI:10.1001/jamahealthforum.2024.4156

Methods Mol Biol. 2025;2873:145-165. doi: 10.1007/978-1-0716-4228-3_9.

ABSTRACT

The nuclear envelope (NE) defines the eukaryotic cell and functions in a myriad of fundamental cellular processes including but not limited to signal transduction, lipid metabolism, chromatin organization, and nucleocytoplasmic transportation. Although the general structure of the NE is well-conserved across eukaryotic kingdoms, its composition and functions vary substantially between species and remain largely unknown in plants. In this chapter, we describe a proximity-labeling-based proteomic approach to profile novel NE components in the model organism Arabidopsis. This method is generally suitable for the identification of protein components in subcellular compartments or protein complexes that are poorly accessible to traditional mass spectrometry approaches and can be easily applied to other plant species. In addition to giving a step-by-step detailed description of the proximity labeling proteomics procedure in plant samples, we also provide guidelines on the appropriate use of controls and statistical analysis to achieve a highly specific selection of probed candidates.

PMID:39576601 | DOI:10.1007/978-1-0716-4228-3_9

Methods Mol Biol. 2025;2867:201-218. doi: 10.1007/978-1-0716-4196-5_12.

ABSTRACT

Intrinsically disordered proteins (IDPs) that include one or more intrinsically disordered regions (IDRs) are abundant across all domains of life and viruses and play numerous functional roles in various cellular processes. Due to a relatively low throughput and high cost of experimental techniques for identifying IDRs, there is a growing need for fast and accurate computational algorithms that accurately predict IDRs/IDPs from protein sequences. We describe one of the leading disorder predictors, flDPnn. Results from a recent community-organized Critical Assessment of Intrinsic Disorder (CAID) experiment show that flDPnn provides fast and state-of-the-art predictions of disorder, which are supplemented with the predictions of several major disorder functions. This chapter provides a practical guide to flDPnn, which includes a brief explanation of its predictive model, descriptions of its web server and standalone versions, and a case study that showcases how to read and understand flDPnn’s predictions.

PMID:39576583 | DOI:10.1007/978-1-0716-4196-5_12

Methods Mol Biol. 2025;2867:43-53. doi: 10.1007/978-1-0716-4196-5_3.

ABSTRACT

Protein secondary structure prediction is useful for many applications. It can be considered a language translation problem, that is, translating a sequence of 20 different amino acids into a sequence of secondary structure symbols (e.g., alpha helix, beta strand, and coil). Here, we develop a novel protein secondary structure predictor called TransPross based on the transformer network and attention mechanism widely used in natural language processing to directly extract the evolutionary information from the protein language (i.e., raw multiple sequence alignment [MSA] of a protein) to predict the secondary structure. The method is different from traditional methods that first generate a MSA and then calculate expert-curated statistical profiles from the MSA as input. The attention mechanism used by TransPross can effectively capture long-range residue-residue interactions in protein sequences to predict secondary structures. Benchmarked on several datasets, TransPross outperforms the state-of-art methods. Moreover, our experiment shows that the prediction accuracy of TransPross positively correlates with the depth of MSAs, and it is able to achieve the average prediction accuracy (i.e., Q3 score) above 80% for hard targets with few homologous sequences in their MSAs. TransPross is freely available at https://github.com/BioinfoMachineLearning/TransPro .

PMID:39576574 | DOI:10.1007/978-1-0716-4196-5_3

Ann Biomed Eng. 2024 Nov 22. doi: 10.1007/s10439-024-03636-4. Online ahead of print.

ABSTRACT

Osteoporosis represents a major healthcare concern. The development of novel treatments presents challenges due to the limited cost-effectiveness of clinical trials and ethical concerns associated with placebo-controlled trials. Computational models for the design and assessment of biomedical products (In Silico Trials) are emerging as a promising alternative. In this study, a novel In Silico Trial technology (BoneStrength) was applied to replicate the placebo arms of two concluded clinical trials and its accuracy in predicting hip fracture incidence was evaluated. Two virtual cohorts (N = 1238 and 1226, respectively) were generated by sampling a statistical anatomy atlas based on CT scans of proximal femurs. Baseline characteristics were equivalent to those reported for the clinical cohorts. Fall events were sampled from a Poisson distribution. A multiscale stochastic model was implemented to estimate the impact force associated to each fall. Finite Element models were used to predict femur strength. Fracture incidence in 3 years follow-up was computed with a Markov chain approach; a patient was considered fractured if the impact force associated with a fall exceeded femur strength. Ten realizations of the stochastic process were run to reach convergence. Each realization required approximately 2500 FE simulations, solved using High-Performance Computing infrastructures. Predicted number of fractures was 12 ± 2 and 18 ± 4 for the two cohorts, respectively. The predicted incidence range consistently included the reported clinical data, although on average fracture incidence was overestimated. These findings highlight the potential of BoneStrength for future applications in drug development and assessment.

PMID:39576502 | DOI:10.1007/s10439-024-03636-4

Ir J Med Sci. 2024 Nov 22. doi: 10.1007/s11845-024-03842-y. Online ahead of print.

ABSTRACT

BACKGROUND: Dermatological conditions are the fourth most common of all diseases affecting humans, and affect almost one third of the world’s population, necessitating effective undergraduate dermatology education.

AIMS: 1. To enquire about the self-perceived confidence of newly qualified junior doctors in recognising and diagnosing dermatological conditions. 2. To ascertain information pertaining to newly qualified junior doctors’ undergraduate dermatology education.

METHODS: A mixed-methods survey on undergraduate dermatology education and self-perceived confidence levels of interns in relation to dermatological conditions was distributed to interns working within two Irish intern networks. Mann-Whitney U testing was used to test for differences between those who had a dermatology placement versus those who did not. Qualitative thematic analysis was used to analyse reflections and comments.

RESULTS: Fifty-seven interns completed our survey. A total of 60% of respondents were female (n = 34). The median age range of respondents was 25-29 years. Fifty-eight percent of interns had a clinical dermatology undergraduate placement (n = 33). Forty-nine percent of interns were dissatisfied with their undergraduate dermatology education (n = 28), while 26% (n = 15) were satisfied and 25% (n = 14) indicated neutral feelings. Confidence levels reported were reasonable where inflammatory and malignant dermatoses were concerned. Statistical significance was established across several areas in dermatology between those who had a dermatology placement versus those who did not. Thematic analysis revealed themes of ineffective dermatology education, of a basic knowledge of dermatology, and of dermatology knowledge via other specialties.

CONCLUSION: This study demonstrates that undergraduate dermatology education is not standardised across various universities in Ireland. It also revealed several areas within dermatology within which there was a statistically significant difference in confidence levels between those who had a clinical undergraduate dermatology placement versus those who did not. This raises the question: “Would mandating a dermatology clerkship may be beneficial to interns?”.

PMID:39576470 | DOI:10.1007/s11845-024-03842-y

Clin Pharmacokinet. 2024 Nov 22. doi: 10.1007/s40262-024-01443-7. Online ahead of print.

ABSTRACT

AIMS: This study aimed to quantify the effects of sodium-glucose co-transporter 2 (SGLT2) inhibitors on N-terminal pro-B-type natriuretic peptide (NT-proBNP) as a therapeutic approach for heart failure.

METHODS: A systematic literature review was conducted to collect pharmacokinetics (PK) and pharmacodynamics (PD) data on empagliflozin, dapagliflozin, and canagliflozin. Population pharmacokinetic models were developed separately for each drug, along with PK/PD turnover models for SGLT2 inhibitors, to describe the time course of NT-proBNP and simulate its changes over 52 weeks.

RESULTS: A total of 42 publications were included in this study. The results showed that baseline NT-proBNP levels, estimated glomerular filtration rate levels, and body weight significantly influenced the therapeutic effects of SGLT2 inhibitors. Among the studied drugs, canagliflozin demonstrated a greater reduction in NT-proBNP at comparable baseline levels.

CONCLUSIONS: Baseline NT-proBNP concentration, renal function, and body weight were covariates affecting the efficacy of SGLT2 inhibitors in reducing NT-proBNP. Canagliflozin showed the most favorable treatment outcomes at similar baseline levels. This model-based meta-analysis approach may support further drug development for SGLT2 inhibitors.

PMID:39576469 | DOI:10.1007/s40262-024-01443-7

Breast Cancer Res Treat. 2024 Nov 22. doi: 10.1007/s10549-024-07552-y. Online ahead of print.

ABSTRACT

PURPOSE: Chemotherapy in combination with trastuzumab is the standard neoadjuvant and adjuvant therapy for human epidermal growth factor receptor 2 (HER2)-positive breast cancer (BC). Assessing the regimens administered to patients with HER2-positive BC in the real world is lacking. We evaluated neoadjuvant and adjuvant regimen patterns among HER2-positive BC patients (2007 to 2021) identified in a health insurance claims database.

METHODS: Female BC patients ≥ 18 years who received chemotherapy, surgery, and trastuzumab were chosen from Optum’s de-identified Clinformatics® Data Mart database. Summary statistics, Joinpoint models, Kaplan-Meier survival curves, and Cox regression models were used to analyze the data.

RESULTS: We identified 6474 patients (median age 60 years), 71.7% were White, 10.9% were Black, 8.6% were Hispanic, 4.1% were Asian, and 4.7% had unknown race/ethnicity. About 33.8% received neoadjuvant therapy and neoadjuvant therapy use increased with an annual percent change of 10.24% (P < .001). The three most common regimens were adjuvant docetaxel, carboplatin, and trastuzumab (TCH; 29.0%); adjuvant paclitaxel and trastuzumab (17.7%); and neoadjuvant TCH with pertuzumab followed by adjuvant trastuzumab (17.7%). The 5-year overall survival (OS) was 96% (95% CI, 95-96%). Patients had an increased risk of death if they were ≥ 59 years at diagnosis, had a health maintenance organization or other insurance plan, had dual Medicare/Medicaid eligibility, had a mastectomy, did not receive 18 cycles of trastuzumab, or received regimens not recommended by the National Comprehensive Cancer Network.

CONCLUSION: Treatment regimen patterns for HER2-positive BC evolved in correspondence with the U.S. Food and Drug Administration’s approval of new drugs for this cancer and National Comprehensive Cancer Network treatment guidelines.

PMID:39576449 | DOI:10.1007/s10549-024-07552-y