Comparison of the Usability and Reliability of Answers to Clinical Questions: AI-Generated ChatGPT versus a Human-Authored Resource.

OBJECTIVES: Our aim was to compare the usability and reliability of answers to clinical questions posed of Chat-Generative Pre-Trained Transformer (ChatGPT) compared to those of a human-authored Web source (www.Pearls4Peers.com) in response to "real-world" clinical questions raised during the care of patients. METHODS: Two domains of clinical information quality were studied: usability, based on organization/readability, relevance, and usefulness, and reliability, based on clarity, accuracy, and thoroughness. The top 36 most viewed real-world questions from a human-authored Web site (www.Pearls4Peers.com [P4P]) were posed to ChatGPT 3.5. Anonymized answers by ChatGPT and P4P (without literature citations) were separately assessed for usability by 18 practicing physicians ("clinician users") in triplicate and for reliability by 21 expert providers ("content experts") on a Likert scale ("definitely yes," "generally yes," or "no") in duplicate or triplicate. Participants also directly compared the usability and reliability of paired answers. RESULTS: The usability and reliability of ChatGPT answers varied widely depending on the question posed. ChatGPT answers were not considered useful or accurate in 13.9% and 13.1% of cases, respectively. In within-individual rankings for usability, ChatGPT was inferior to P4P in organization/readability, relevance, and usefulness in 29.6%, 28.3%, and 29.6% of cases, respectively, and for reliability, inferior to P4P in clarity, accuracy, and thoroughness in 38.1%, 34.5%, and 31% of cases, respectively. CONCLUSIONS: The quality of ChatGPT responses to real-world clinical questions varied widely, with nearly one-third or more answers considered inferior to a human-authored source in several aspects of usability and reliability. Caution is advised when using ChatGPT in clinical decision making.

Management of primary skin cancer during a pandemic: Multidisciplinary recommendations.

During the coronavirus disease 2019 (COVID-19) pandemic, providers and patients must engage in shared decision making regarding the pros and cons of early versus delayed interventions for localized skin cancer. Patients at highest risk of COVID-19 complications are older; are immunosuppressed; and have diabetes, cancer, or cardiopulmonary disease, with multiple comorbidities associated with worse outcomes. Physicians must weigh the patient's risk of COVID-19 complications in the event of exposure against the risk of worse oncologic outcomes from delaying cancer therapy. Herein, the authors have summarized current data regarding the risk of COVID-19 complications and mortality based on age and comorbidities and have reviewed the literature assessing how treatment delays affect oncologic outcomes. They also have provided multidisciplinary recommendations regarding the timing of local therapy for early-stage skin cancers during this pandemic with input from experts at 11 different institutions. For patients with Merkel cell carcinoma, the authors recommend prioritizing treatment, but a short delay can be considered for patients with favorable T1 disease who are at higher risk of COVID-19 complications. For patients with melanoma, the authors recommend delaying the treatment of patients with T0 to T1 disease for 3 months if there is no macroscopic residual disease at the time of biopsy. Treatment of tumors ≥T2 can be delayed for 3 months if the biopsy margins are negative. For patients with cutaneous squamous cell carcinoma, those with Brigham and Women's Hospital T1 to T2a disease can have their treatment delayed for 2 to 3 months unless there is rapid growth, symptomatic lesions, or the patient is immunocompromised. The treatment of tumors ≥T2b should be prioritized, but a 1-month to 2-month delay is unlikely to worsen disease-specific mortality. For patients with squamous cell carcinoma in situ and basal cell carcinoma, treatment can be deferred for 3 months unless the individual is highly symptomatic.

Coexisting Systemic Infections in Patients Hospitalized Because of a Fall: Prevalence and Risk Factors.

BACKGROUND: Infectious disease-related factors that may contribute to or complicate falls have received relatively little attention in the literature. OBJECTIVE: Our aim was to determine the prevalence of, and risk factors for, coexisting systemic infections (CSIs) in patients admitted to the hospital because of a fall or its complications. METHODS: We conducted a retrospective cohort study of adult patients seen at a tertiary care hospital emergency department and subsequently hospitalized because of a fall or its complications. RESULTS: Of 1,456 evaluable cases, 775 patients (53.2%) were female. Mean age was 71.6 years (range 18-104 years). CSI was diagnosed in 303 patients (20.8%), of which 166 (54.8%) were urinary tract infections and 108 (35.6%) were pneumonia cases; 14 patients (4.6%) were bacteremic. CSI was not initially suspected by providers in 98 (32.5%) subsequently diagnosed cases. Age ≥50 years (odds ratio [OR] 5.6; 95% confidence interval [CI] 1.2-24.9), inability to get up on own after the index fall (OR 2.1; 95% CI 1.2-3.6), preexisting symptom(s) (OR 3.0; 95% CI 1.8-5.2), and systemic inflammatory response syndrome (SIRS) (OR 2.9; 95% CI 1.5-5.4), or confusion at presentation (OR 3.0; 95% CI 1.5-6.0) were independently associated with CSI. In-hospital mortality rate was significantly higher among patients with CSI (6.9% vs. 3.8 %, OR 1.9; 95% CI 1.1-3.3). CONCLUSIONS: CSIs are common among patients admitted to the hospital after a fall or its complications. Age ≥ 50 years, inability to get up on own, preexisting symptom(s), and the presence of SIRS or confusion at presentation are potential predictors of CSI in this patient population.

Randomized controlled feasibility trial of supported self-management in adults with Type 2 diabetes mellitus and an intellectual disability: OK Diabetes.

AIMS: To undertake a feasibility randomized controlled trial of supported self-management vs treatment as usual in a population of adults with obesity, Type 2 diabetes and an intellectual disability. METHODS: We conducted an individually randomized feasibility trial. Participants were adults aged >18 years with a mild or moderate intellectual disability, living in the community with Type 2 diabetes, on any therapy other than insulin. Participants had mental capacity to consent to research and the intervention. Inclusion criteria included HbA1c > 48 mmol/mol (6.5%), BMI >25 kg/m2 , or self-reported physical activity below national guideline levels. The experimental intervention was standardized supported self-management delivered by diabetes specialist nurses plus treatment as usual, compared with treatment as usual alone. Feasibility outcomes included: recruitment and retention; intervention acceptability and feasibility; data collection and completeness for physiological state and values for candidate primary outcomes (HbA1c and BMI). RESULTS: A total of 82 participants (89% of those contacted and eligible) were randomized. All supported self-management sessions were completed by 35/41 participants (85%); only four completed no sessions. Data on the follow-up candidate primary outcomes HbA1c and BMI were obtained for 75/82 (91%) and 77/82 participants (94%), respectively. The mean baseline HbA1c was 56±16.5 mmol/mol (7.3±1.5%) and the mean BMI was 34±7.6 kg/m2 . CONCLUSIONS: Adherence to supported self-management and willingness to have blood taken for outcome measurement was good. A definitive randomized controlled trial is feasible in this population. (Trial registration: Current Controlled Trials ISRCTN41897033).

Characterizing adults with Type 2 diabetes mellitus and intellectual disability: outcomes of a case-finding study.

AIMS: To report the results of a case-finding study conducted during a feasibility trial of a supported self-management intervention for adults with mild to moderate intellectual disability and Type 2 diabetes mellitus, and to characterize the study sample in terms of diabetes control, health, and access to diabetes management services and support. METHODS: We conducted a cross-sectional case-finding study in the UK (March 2013 to June 2015), which recruited participants mainly through primary care settings. Data were obtained from medical records and during home visits. RESULTS: Of the 325 referrals, 147 eligible individuals participated. The participants' mean (sd) HbA1c concentration was 55 (15) mmol/mol [7.1 (1.4)%] and the mean (sd) BMI was 32.9 (7.9) kg/m2 , with 20% of participants having a BMI >40 kg/m2 . Self-reported frequency of physical activity was low and 79% of participants reported comorbidity, for example, cardiovascular disease, in addition to Type 2 diabetes. The majority of participants (88%) had a formal or informal supporter involved in their diabetes care, but level and consistency of support varied greatly. Post hoc exploratory analyses showed a significant association between BMI and self-reported mood, satisfaction with diet and weight. CONCLUSIONS: We found high obesity and low physical activity levels in people with intellectual disability and Type 2 diabetes. Glycaemic control was no worse than in the general Type 2 diabetes population. Increased risk of morbidity in this population is less likely to be attributable to poor glycaemic control and is probably related, at least in part, to greater prevalence of obesity and inactivity. More research, focused on weight management and increasing activity in this population, is warranted.

Efficacy of Oral Vancomycin in Preventing Recurrent Clostridium difficile Infection in Patients Treated With Systemic Antimicrobial Agents.

We compared rates of recurrent Clostridium difficile infection in patients receiving or not receiving oral vancomycin prophylaxis with systemic antimicrobial therapy. The incidence of C. difficile infection was significantly lower in patients receiving prophylaxis (4.2% vs 26.6% in those without prophylaxis; odds ratio, 0.12; 95% confidence interval, .04-.4; P < .001).

The role of postoperative factors in surgical site infections: time to take notice.

Surgical site infections (SSIs) continue to occur, in many instances despite high compliance with best practice measures primarily revolving around pre- and intraoperative periods. Postoperative factors have traditionally been considered to play a relatively minor role in the causation of SSIs. An increasing body of evidence, however, suggests that many SSIs occur as a result of pathogens gaining access to surgical wounds either hematogenously, through drains, or through slowly healing wounds due to systemic anticoagulation or other factors, particularly in the setting of high compliance with standard perioperative antibiotic prophylaxis. Evidence also supports frequent acquisition of methicillin-resistant Staphylococcus aureus (MRSA) during the postoperative period. These findings, coupled with lack of clear efficacy of various pre- and intraoperative interventions such as MRSA decolonization and use of vancomycin for prophylaxis against this organism, should force us to consider the important role that postoperative factors may play in the causation of SSIs in the current era.

Randomised controlled trial of a psychotherapeutic intervention to improve quality of life and other outcomes in people who repeatedly self-harm: FReSH START study protocol.

BACKGROUND: Self-harm is a major public health challenge, and repeated self-harm is common in those attending hospital following an episode. Evidence suggests psychological interventions could help people who self-harm, but few definitive studies have assessed their clinical and cost-effectiveness. Repeated self-harm is associated with poor quality of life, depression, suicide and increased health service costs which justify the development of psychotherapeutic interventions tailored for people with repeated self-harm. METHODS: FReSH START is a multicentre individually 1:1 randomised controlled trial evaluating the clinical and cost-effectiveness of standard care plus psychological therapy or standard care alone for adults (≥ 18 years) presenting at an emergency department (ED) with repeated self-harm. Recruiting 630 participants, it includes an internal pilot, economic evaluation and process evaluation. The intervention will be delivered by mental health staff working in acute settings, with experience of assessing and managing risk in people presenting to emergency services with self-harm. Staff will be trained and supervised to deliver one of three specially adapted therapies: psychodynamic interpersonal therapy, cognitive behavioural therapy or acceptance and commitment therapy. Participants allocated to the intervention will receive one of the adapted therapies according to therapist allocation for up to 6 months via 12 weekly, one to one, 45-50-min sessions. The primary outcome is quality of life measured by the Clinical Outcomes in Routine Evaluation Outcome Measure at 12 months post-randomisation. Secondary outcomes include suicidal intent, depression and cost-effectiveness. Data are collected using hospital attendance records and online/postal/telephone questionnaires at 6 and 12 months post-randomisation, with resource use additionally collected at 3 and 9 months. DISCUSSION: This protocol outlines a randomised controlled trial to investigate whether modified therapies are cost-effective and improve quality of life for people who repeatedly self-harm. Few interventions are proven to be deliverable in the NHS for this population. This study is strengthened by the involvement of qualified mental health workers experienced in managing risk as therapists. TRIAL REGISTRATION: Registered on August 03, 2021. IRAS number: 297939. ISRCTN: https://doi.org/10.1186/ISRCTN73357210 . REC reference: 21/EE/0145. SPONSOR: University of Leeds.

e-Consent in UK academic-led clinical trials: current practice, challenges and the need for more evidence.

BACKGROUND: During the COVID-19 pandemic, in-person healthcare visits were reduced. Consequently, trial teams needed to consider implementing remote methods for conducting clinical trials, including e-Consent. Although some clinical trials may have implemented e-Consent prior to the pandemic, anecdotes of uptake for this method increased within academic-led trials. When the increased use of this process emerged, representatives from several large academic clinical trial groups within the UK collaborated to discuss ways in which trialists can learn from one another when implementing e-Consent. METHODS: A survey of UKCRC-registered Clinical Trials Units (CTUs) was undertaken in April-June 2021 to understand the implementation of and their views on the use of e-Consent and experiences from the perspectives of systems programmers and quality assurance staff on the use of e-Consent. CTUs not using e-Consent were asked to provide any reasons/barriers (including no suitable trials) and any plans for implementing it in the future. Two events for trialists and patient and public involvement (PPI) representatives were then held to disseminate findings, foster discussion, share experiences and aid in the identification of areas that the academic CTU community felt required more research. RESULTS: Thirty-four (64%) of 53 CTUs responded to the survey, with good geographical representation across the UK. Twenty-one (62%) of the responding CTUs had implemented e-Consent in at least one of their trials, across different types of trials, including CTIMPs (Clinical Trial of Investigational Medicinal Product), ATIMPs (Advanced Therapy Medicinal Products) and non-CTIMPs. One hundred ninety-seven participants attended the two workshops for wide-ranging discussions. CONCLUSION: e-Consent is increasingly used in academic-led trials, yet uncertainties remain amongst trialists, patients and members of the public. Uncertainties include a lack of formal, practical guidance and a lack of evidence to demonstrate optimal or appropriate methods to use. We strongly encourage trialists to continue to share their own experiences of the implementation of e-Consent.

Fever, abnormal white blood cell count, neutrophilia, and elevated serum C-reactive protein in adult hospitalized patients with bacteremia.

OBJECTIVES: Relatively little is known about the prevalence of and factors associated with the absence of fever, normal white blood cell count (WBC), lack of neutrophilia, and the frequency of elevated serum C-reactive protein (CRP) in adult hospitalized patients with bacteremia. METHODS: Frequencies of fever (temperature ≥ 100°F), abnormal WBC count (≥ 10,000/µL or <4500/µL), neutrophilia (neutrophils ≥ 80%) and potential factors that may be associated with negative results, and frequency of elevated serum CRP were studied retrospectively in 622 bacteremic episodes in adult patients at a community teaching hospital. RESULTS: The frequencies of individual parameters were as follows: fever, 78.4% (95% confidence interval [CI] 74.9%-81.6%); abnormal WBC count, 65.2% (95% CI 61.3%-69.0%); neutrophilia, 67.5% (95% CI 63.7%-71.2%); and elevated CRP (>0.8 mg/dL), 98.4% (95% CI 97%-99.2%). Age 50 years and older, corticosteroid use, and isolation of either Staphylococcus sp not aureus or Enterococcus sp combined were independently associated with the absence of fever (P ≤ 0.01 for all). Isolation of Staphylococcus sp not aureus/Enterococcus and lack of corticosteroid use were independently associated with normal WBC (P < 0.05 for both). Nonwhite race and primary bacteremia were independently associated with lack of neutrophilia (P < 0.05 for both). CONCLUSIONS: Absence of fever, normal WBC, and lack of neutrophilia are common in patients with bacteremia and may be confounded by a variety of factors; however, serum CRP remains elevated in the majority (>98%) of bacteremic episodes.

Repurposing existing medications for coronavirus disease 2019: protocol for a rapid and living systematic review.

BACKGROUND: Coronavirus disease 2019 (COVID-19) has no confirmed specific treatments. However, there might be in vitro and early clinical data as well as evidence from severe acute respiratory syndrome and Middle Eastern respiratory syndrome that could inform clinicians and researchers. This systematic review aims to create priorities for future research of drugs repurposed for COVID-19. METHODS: This systematic review will include in vitro, animal, and clinical studies evaluating the efficacy of a list of 34 specific compounds and 4 groups of drugs identified in a previous scoping review. Studies will be identified both from traditional literature databases and pre-print servers. Outcomes assessed will include time to clinical improvement, time to viral clearance, mortality, length of hospital stay, and proportions transferred to the intensive care unit and intubated, respectively. We will use the GRADE methodology to assess the quality of the evidence. DISCUSSION: The challenge posed by COVID-19 requires not just a rapid review of drugs that can be repurposed but also a sustained effort to integrate new evidence into a living systematic review. TRIAL REGISTRATION: PROSPERO 2020 CRD42020175648.