Predictors of repeat emergency department visits by elders

OBJECTIVE

To determine which characteristics of older patients who use a hospital ED are associated with repeat visits during the 90 days following the index visit.

METHODS

The study was conducted in the ED of a 400-bed university-affiliated acute care community hospital in Montreal. Patients aged > or = 75 years who visited the ED between 08:00 and and 16:00 on a convenience sample of days over an 8-week period (July and August 1994) were assessed using a questionnaire, physical and cognitive status instruments, and a functional problem checklist. The hospital's administrative database was used to identify repeat visits during the 90 days following the ED visit. The representativeness of the sample was assessed by analyses of ED visits made by 4,466 persons aged > or = 65 years during a 12-month period (September 1993 to August 1994) using the hospital's administrative database.

RESULTS

256 patients aged > or = 75 years visited the ED during the study period and 167 were assessed. Of these, 54 (32%) were admitted to the hospital. Among the 113 patients released from the ED, 27 (24%) made repeat visits during the next 90 days. In univariate analyses, repeat visits were significantly associated with the number of functional problems, cognitive impairment, and previous ED visits. In multiple logistic regression, male gender, living alone, and number of functional problems were independent predictors of repeat visits. In the administrative data analyses, nighttime arrival to the ED for the index visit was significantly associated with repeat visits.

CONCLUSIONS

Self-reported risk factors can help to identify a group of elders likely to make repeated ED visits; the development of a screening instrument incorporating questions on these problems and implementation of appropriate interventions might improve these patients' quality of life and reduce the demand for further ED care in this age group.

The catalytic domain of lambda site-specific recombinase

The Escherichia coli phage lambda integrase protein (Int) belongs to the large Int family of site-specific recombinases. It is a heterobivalent DNA binding protein that makes use of a high energy covalent phosphotyrosine intermediate to catalyze integrative and excisive recombination at specific chromosomal sites (att sites). A 293-amino acid carboxy-terminal fragment of Int (C65) has been cloned, characterized, and used to further dissect the protein. From this we have cloned and characterized a 188-amino acid, protease-resistant, carboxy-terminal fragment (C170) that we believe is the minimal catalytically competent domain of Int. C170 has topoisomerase activity and converts att suicide substrates to the covalent phosphotyrosine complexes characteristic of recombination intermediates. However, it does not show efficient binding to att site DNA in a native gel shift assay. We propose that lambda Int consists of three functional and structural domains: residues 1-64 specify recognition of "arm-type" DNA sequences distant from the region of strand exchange; residues 65-169 contribute to specific recognition of "core-type" sequences at the sites of strand exchange and possibly to protein-protein interactions; and residues 170-356 carry out the chemistry of DNA cleavage and ligation. The finding that the active site nucleophile Tyr-342 is in a uniquely protease-sensitive region complements and reinforces the recently solved C170 crystal structure, which places Tyr-342 at the center of a 17-amino acid flexible loop. It is proposed that C170 is likely to represent a generic Int family domain that thus affords a specific route to studying the chemistry of DNA cleavage and ligation in these recombinases.

Effect of posture on vital capacity

The influence of some extreme body postures on vital capacity (VC) was examined in young adult humans. Two postures required full support of body weight by the arms: arms up, hanging from a bar, and arms down with hands gripping parallel bars. Three involved muscles that flex and extend the trunk: a partial sit-up position while supine and nearly maximal spinal extension and flexion while standing. Changes at the inspiratory and expiratory volume extremes were recognized by having the subjects do two VC efforts: the first standing and the second in the posture in question while continuing to breathe on the spirometer. Control observations in which the second of a VC pair was performed in an unstressed posture allowed correction for the influence of rebreathing. The changes in corrected VC were small, the greatest being an average reduction of approximately 8% in the partial sit-up position. During full support of body weight by the arms, the VC was slightly increased due to a significant increase in the inspiratory extreme and no change in the expiratory extreme. Spinal extension produced small increases in lung volume at both extremes with no significant change in VC, whereas spinal flexion did not influence the upper extreme but did increase lung volume at the lower extreme. The changes are discussed in terms of trunk muscle action.

Embedded Model Predictive Control for a Wearable Artificial Pancreas

While artificial pancreas (AP) systems are expected to improve the quality of life among people with type 1 diabetes mellitus (T1DM), the design of convenient systems that optimize the user experience, especially for those with active lifestyles, such as children and adolescents, still remains an open research question. In this work, we introduce an embeddable design and implementation of model predictive control (MPC) of AP systems for people with T1DM that significantly reduces the weight and on-body footprint of the AP system. The embeddable controller is based on a zone MPC that has been evaluated in multiple clinical studies. The proposed embedded zone MPC features a simpler design of the periodic safe zone in the cost function and the utilization of state-of-the-art alternating minimization algorithms for solving the convex programming problems inherent to MPC with linear models subject to convex constraints. Off-line closed-loop data generated by the FDA-accepted UVA/Padova simulator is used to select an optimization algorithm and corresponding tuning parameters. Through hardware-in-the-loop in silico results on a limited-resource Arduino Zero (Feather M0) platform, we demonstrate the potential of the proposed embedded MPC. In spite of resource limitations, our embedded zone MPC manages to achieve comparable performance of that of the full-version zone MPC implemented in a 64-bit desktop for scenarios with/without meal-disturbance compensations. Metrics for performance comparison included median percent time in the euglycemic ([70, 180] mg/dL range) of 84.3% vs. 83.1% for announced meals, with an equivalence test yielding p = 0.0013 and 66.2% vs. 66.0% for unannounced meals with p = 0.0028.

POS1106 MEDIATORS OF THE EFFECT OF THERAPEUTIC EXERCISE ON PAIN AND FUNCTION IN PATIENTS WITH KNEE AND HIP OSTEOARTHRITIS; AN INDIVIDUAL PATIENT DATA MEDIATION STUDY FROM THE OA TRIAL BANK

Background

Currently, there is no cure for osteoarthritis (OA). Therapeutic exercise is recommended in all international guidelines to improve pain and function (1). However, randomised controlled trials (RCTs) investigating therapeutic exercise showed that, on average, effect sizes are small to moderate compared to non-exercise controls (2-3). Better knowledge about how therapeutic exercise works to improve pain and function could help improve the design and delivery of future therapeutic exercise interventions and potentially improve patient outcomes.

Objectives

To evaluate the mediating effects of 1) muscle strength, proprioception, and range of motion (ROM) among patients with knee OA, and 2) muscle strength among patients with hip OA, using the procedures established for Individual Participant Data (IPD) analyses by the OA Trial Bank.

Methods

Relevant IPD were sought from all published RCTs comparing therapeutic exercise to a non-exercise control group among people with knee and/or hip OA, using previously published methods (4). For each potential mediator separately, the effect of exercise on the absolute change in pain/function directly after the intervention was determined, controlling for the relative change in mediator under investigation and potential mediator-outcome confounders, using the ‘counterfactual framework’ (5). Next, the effect of exercise on the relative change in the mediator was determined. The percentage mediated was calculated by dividing the natural indirect effect by the total effect.

Results

We obtained IPD from 12 RCTs and 1407 participants (knee OA = 1113; hip OA n=294) that were eligible for inclusion in one or more mediation analyses. Therapeutic exercise showed a significant effect on the relative change in muscle strength in knee OA, but not for proprioception or ROM in knee OA, nor for muscle strength in hip OA. Of all mediators, only the change in muscle strength had a significant effect on the change in pain in knee OA, and on function in knee OA and in hip OA. The percentage mediated for each mediator and each outcome was small (5.3% or less, see Table 1).

Table 1.

Percentage mediated for all potential mediators, by outcome and joint.

Knee OAPainKnee OAFunctionHip OAPainHip OAFunctionMuscle strength2.4%2.3%absentabsentProprioceptionabsent0.2%--Range of motion1.8%5.3%--

‘absent’: no percentage mediated calculated, since direct and indirect effect had opposite directions.

‘-‘: factors not considered a potential mediator for therapeutic exercise in hip OA.

Conclusion

In the first IPD mediation analysis to our knowledge of this scale, muscle strength, proprioception and ROM did mediate the effect of therapeutic exercise for pain and function in knee OA, but only to a very small degree. We observed no such mediating effect of muscle strength for therapeutic exercise in patients with hip OA. The challenge remains to better understand the key mechanisms of action of therapeutic exercise for knee and hip OA, so that exercise interventions can be designed and delivered to target these and therewith optimise the effects for patients.

References

[1]Bierma-Zeinstra S, et al. Nonpharmacological and nonsurgical approaches in OA. Best Pract Res Clin Rheumatol. 2020 Apr;34(2):101564.

[2]Fransen M, et al. Exercise for osteoarthritis of the knee. Cochrane Database Syst Rev. 2015 Jan 9;1:CD004376.

[3]Fransen M, et al. Exercise for osteoarthritis of the hip. Cochrane Database Syst Rev. 2014 Apr 22(4):CD007912.

[4]Holden MA, et al. Subgrouping and TargetEd Exercise pRogrammes for knee and hip OsteoArthritis (STEER OA): a systematic review update and individual participant data meta-analysis protocol. BMJ Open. 2017 Dec 22;7(12):e018971.

[5]VanderWeele TJ, Vansteelandt S. Mediation Analysis with Multiple Mediators. Epidemiol Methods. 2014 Jan;2(1):95-115.

Acknowledgements

We thank all members of the OA Trial Bank Exercise Collaborative, who shared IPD with the OA Trial Bank/STEER OA for making this study possible.

Disclosure of Interests

Jos Runhaar Grant/research support from: Fellowship from the Foundation for Research in Rheumatology (FOREUM), Melanie Holden Grant/research support from: Grant from the Chartered Society of Physiotherapy Charitable Trust (grant no PRF/16/A07), and the National Institute for Health Research (NIHR) School of Primary Care Research (grant no 531), Miriam Hattle: None declared, Johny Quicke Grant/research support from: JQ is part-funded by an NIHR CRN West Midlands Research Scholarship and part funded by the Haywood Foundation, Richard Riley: None declared, Emma Healey Grant/research support from: part funded by the National Institute for Health Research (NIHR) Applied Research Collaboration (ARC) West Midlands., Danielle van der Windt: None declared, Krysia Dziedzic: None declared, Marienke Van Middelkoop: None declared, S.M.A. Bierma-Zeinstra Grant/research support from: OA Trial Bank received long-term funding by the Dutch Arthritis Association, Nadine Foster: None declared

LLM-CGM: A Benchmark for Large Language Model-Enabled Querying of Continuous Glucose Monitoring Data for Conversational Diabetes Management

Over the past decade, wearable technology has dramatically changed how patients manage chronic diseases. The widespread availability of on-body sensors, such as heart rate monitors and continuous glucose monitoring (CGM) sensors, has allowed patients to have real-time data about their health. Most of these data are readily available on patients' smartphone applications, where patients can view their current and retrospective data. For patients with diabetes, CGM has transformed how their disease is managed. Many sensor devices interface with smartphones to display charts, metrics, and alerts. However, these metrics and plots may be challenging for some patients to interpret. In this work, we explore how large language models (LLMs) can be used to answer questions about CGM data. We produce an open-source benchmark of time-series question-answering tasks for CGM data in diabetes management. We evaluate different LLM frameworks to provide a performance benchmark. Lastly, we highlight the need for more research on how to optimize LLM frameworks to best handle questions about wearable data. Our benchmark is publicly available for future use and development. While this benchmark is specifically designed for diabetes care, our model implementation and several of the statistical tasks can be extended to other wearable device domains.

Model-Based Insulin Sensitivity and Beta-Cell Function Estimation from Daily Continuous Glucose Monitoring

Type 2 diabetes (T2D) is a prevalent chronic illness with many different options for treatment management. Continuous glucose monitors (CGM) offer physiological data that clinicians can access when making treatment decisions. However, the utility of CGM in management of T2D remains an active area of research. In our work, we demonstrate the feasibility of exploiting raw daily CGM data to estimate the physiological parameters of insulin sensitivity and beta-cell function that correlate with estimates derived from laboratory findings. We use a peak extraction algorithm to extract peaks from daily CGM data and implement a model-based approach to infer physiological parameters. We demonstrate that the inferred parameter estimates of insulin sensitivity and beta-cell function correlate to the ground truth measurements as determined by an oral glucose tolerance test (OGTT).

Reassessing the management of uncomplicated urinary tract infection: A retrospective analysis using machine learning causal inference

Background

Uncomplicated urinary tract infection (UTI) is a common indication for outpatient antimicrobial therapy. National guidelines for the management of uncomplicated UTI were published by the Infectious Diseases Society of America in 2011, however it is not fully known the extent to which they align with current practices, patient diversity, and pathogen biology, all of which have evolved significantly in the time since their publication.

Objective

We aimed to re-evaluate efficacy and adverse events for first-line antibiotics (nitrofurantoin, and trimethoprim-sulfamethoxazole), versus second-line antibiotics (fluoroquinolones) and versus alternative agents (oral β-lactams) for uncomplicated UTI in contemporary clinical practice by applying machine learning algorithms to a large claims database formatted into the Observational Medical Outcomes Partnership (OMOP) common data model.

Outcomes

Our primary outcome was a composite endpoint for treatment failure, defined as outpatient or inpatient re-visit within 30 days for UTI, pyelonephritis or sepsis. Secondary outcomes were the risk of 4 common antibiotic-associated adverse events: gastrointestinal symptoms, rash, kidney injury and C. difficile infection.

Statistical methods

We adjusted for covariate-dependent censoring and treatment indication using a broad set of domain-expert derived features. Sensitivity analyses were conducted using OMOP-learn, an automated feature engineering package for OMOP datasets.

Results

Our study included 57,585 episodes of UTI from 49,037 patients. First-line antibiotics were prescribed in 35,018 (61%) episodes, second-line antibiotics were prescribed in 21,140 (37%) episodes and alternative antibiotics were prescribed in 1,427 (2%) episodes. After adjustment, patients receiving first-line therapies had an absolute risk difference of -2.1% [95% CI -2.9% to -1.6%] for having a revisit for UTI within 30 days of diagnosis relative to second-line antibiotics. First-line therapies had an absolute risk difference of -6.6% [95% CI -9.4% to -3.8%] for 30-day revisit compared to alternative β-lactam antibiotics. Differences in adverse events were clinically similar between first and second line agents, but lower for first-line agents relative to alternative antibiotics (-3.5% [95% CI -5.9% to -1.2%]). Results were similar for models built with OMOPlearn.

Conclusion

Our study provides support for the continued use of first-line antibiotics for the management of uncomplicated UTI. Our results also provide proof-of-principle that automated feature extraction methods for OMOP formatted data can emulate manually curated models, thereby promoting reproducibility and generalizability.

Leveraging Large Language Models to Analyze Continuous Glucose Monitoring Data: A Case Study

Continuous glucose monitors (CGM) provide patients and clinicians with valuable insights about glycemic control that aid in diabetes management. The advent of large language models (LLMs), such as GPT-4, has enabled real-time text generation and summarization of medical data. Further, recent advancements have enabled the integration of data analysis features in chatbots, such that raw data can be uploaded and analyzed when prompted. Studying both the accuracy and suitability of LLM-derived data analysis performed on medical time series data, such as CGM data, is an important area of research. The objective of this study was to assess the strengths and limitations of using an LLM to analyze raw CGM data and produce summaries of 14 days of data for patients with type 1 diabetes. This study used simulated CGM data from 10 different cases. We first evaluated the ability of GPT-4 to compute quantitative metrics specific to diabetes found in an Ambulatory Glucose Profile (AGP). Then, using two independent clinician graders, we evaluated the accuracy, completeness, safety, and suitability of qualitative descriptions produced by GPT-4 across five different CGM analysis tasks. We demonstrated that GPT-4 performs well across measures of accuracy, completeness, and safety when producing summaries of CGM data across all tasks. These results highlight the capabilities of using an LLM to produce accurate and safe narrative summaries of medical time series data. We highlight several limitations of the work, including concerns related to how GPT-4 may misprioritize highlighting instances of hypoglycemia and hyperglycemia. Our work serves as a preliminary study on how generative language models can be integrated into diabetes care through CGM analysis, and more broadly, the potential to leverage LLMs for streamlined medical time series analysis.

A case study on using a large language model to analyze continuous glucose monitoring data

Continuous glucose monitors (CGM) provide valuable insights about glycemic control that aid in diabetes management. However, interpreting metrics and charts and synthesizing them into linguistic summaries is often non-trivial for patients and providers. The advent of large language models (LLMs) has enabled real-time text generation and summarization of medical data. The objective of this study was to assess the strengths and limitations of using an LLM to analyze raw CGM data and produce summaries of 14 days of data for patients with type 1 diabetes. We first evaluated the ability of GPT-4 to compute quantitative metrics specific to diabetes found in an Ambulatory Glucose Profile (AGP). Then, using two independent clinician graders, we evaluated the accuracy, completeness, safety, and suitability of qualitative descriptions produced by GPT-4 across five different CGM analysis tasks. GPT-4 performed 9 out of the 10 quantitative metrics tasks with perfect accuracy across all 10 cases. The clinician-evaluated CGM analysis tasks had good performance across measures of accuracy [lowest task mean score 8/10, highest task mean score 10/10], completeness [lowest task mean score 7.5/10, highest task mean score 10/10], and safety [lowest task mean score 9.5/10, highest task mean score 10/10]. Our work serves as a preliminary study on how generative language models can be integrated into diabetes care through data summarization and, more broadly, the potential to leverage LLMs for streamlined medical time series analysis.

Use of Machine Learning to Assess the Management of Uncomplicated Urinary Tract Infection

Importance

Uncomplicated urinary tract infection (UTI) is a common indication for outpatient antimicrobial therapy. National guidelines for the management of uncomplicated UTI were published in 2011, but the extent to which they align with current practices, patient diversity, and pathogen biology, all of which have evolved greatly in the time since their publication, is not fully known.

Objective

To reevaluate the effectiveness and adverse event profile for first-line antibiotics, fluoroquinolones, and oral β-lactams for treating uncomplicated UTI in contemporary clinical practice.

Design, Setting, and Participants

This retrospective, population-based cohort study used a claims dataset from Independence Blue Cross, which contains inpatient, outpatient, laboratory, and pharmacy claims that occurred between 2012 and 2021, formatted into the Observational Medical Outcomes Partnership (OMOP) common data model. Participants were nonpregnant female individuals aged 18 years or older with a diagnosis of uncomplicated, nonrecurrent UTI at an outpatient setting. Patients must also have been treated with first-line (nitrofurantoin or trimethoprim-sulfamethoxazole), fluoroquinolone (ciprofloxacin, levofloxacin, or ofloxacin), or oral β-lactam (amoxicillin-clavulanate, cefadroxil, or cefpodoxime) antibiotics. Data analysis was performed from November 2021 to August 2024.

Exposures

Patients exposed to first-line antibiotics were assigned to the treatment group, and those exposed to fluoroquinolone or β-lactam treatments were assigned to control groups.

Main Outcomes and Measures

The primary outcome was a composite end point for treatment failure, defined as outpatient or inpatient revisit within 30 days for UTI, pyelonephritis, or sepsis. Secondary outcomes were the risk of 4 common antibiotic-associated adverse events: gastrointestinal symptoms, rash, kidney injury, and Clostridium difficile infection.

Results

There were 57 585 episodes of UTI among 49 037 female patients (mean [SD] age, 51.7 [20.1]) years), with prescriptions for first-line antibiotics in 35 018 episodes (61%), fluoroquinolones in 21 140 episodes (37%), and β-lactams in 1427 episodes (2%). After adjustment, receipt of first-line therapies was associated with an absolute risk difference of -1.78% (95% CI, -2.37% to -1.06%) for having a revisit for UTI within 30 days of diagnosis vs fluoroquinolones. First-line therapies were associated with an absolute risk difference of -6.40% (95% CI, -10.14% to -3.24%) for 30-day revisit compared with β-lactam antibiotics. Differences in adverse events were similar between all comparators. Results were identical for models built with an automated OMOP feature extraction package.

Conclusions and Relevance

In this cohort study of patients with uncomplicated UTI derived from a large regional claims dataset, national treatment guidelines published almost 14 years ago continue to recommend optimal treatments. These results also provide proof-of-principle that automated feature extraction methods for OMOP formatted data can emulate manually curated models, thereby promoting reproducibility and generalizability.