Health system-scale language models are all-purpose prediction engines.

Physicians make critical time-constrained decisions every day. Clinical predictive models can help physicians and administrators make decisions by forecasting clinical and operational events. Existing structured data-based clinical predictive models have limited use in everyday practice owing to complexity in data processing, as well as model development and deployment1-3. Here we show that unstructured clinical notes from the electronic health record can enable the training of clinical language models, which can be used as all-purpose clinical predictive engines with low-resistance development and deployment. Our approach leverages recent advances in natural language processing4,5 to train a large language model for medical language (NYUTron) and subsequently fine-tune it across a wide range of clinical and operational predictive tasks. We evaluated our approach within our health system for five such tasks: 30-day all-cause readmission prediction, in-hospital mortality prediction, comorbidity index prediction, length of stay prediction, and insurance denial prediction. We show that NYUTron has an area under the curve (AUC) of 78.7-94.9%, with an improvement of 5.36-14.7% in the AUC compared with traditional models. We additionally demonstrate the benefits of pretraining with clinical text, the potential for increasing generalizability to different sites through fine-tuning and the full deployment of our system in a prospective, single-arm trial. These results show the potential for using clinical language models in medicine to read alongside physicians and provide guidance at the point of care.

Effect of a high value care curriculum on standardized patient exam in the Core Clerkship in Internal Medicine.

BACKGROUND: With almost 20% unnecessary spending on healthcare, there has been increasing interest in high value care defined as the best care for the patient, with the optimal result for the circumstances, delivered at the right price. The American Association of Medical Colleges recommend that medical students are proficient in concepts of cost-effective clinical practice by graduation, thus leading to curricula on high value care. However little is published on the effectiveness of these curricula on medical students' ability to practice high value care. METHODS: In addition to the standard curriculum, the intervention group received two classroom sessions and three virtual patients focused on the concepts of high value care. The primary outcome was number of tests and charges for tests on standardized patients. RESULTS: 136 students enrolled in the Core Clerkship in Internal Medicine and 70 completed the high value care curriculum. There were no significant differences in ordering of appropriate tests (3.1 vs. 3.2 tests/students, p = 0.55) and inappropriate tests (1.8 vs. 2.2, p = 0.13) between the intervention and control. Students in the intervention group had significantly lower median Medicare charges ($287.59 vs. $500.86, p = 0.04) and felt their education in high value care was appropriate (81% vs. 56%, p = 0.02). CONCLUSIONS: This is the first study to describe the impact of a high value care curriculum on medical students' ordering practices. While number of inappropriate tests was not significantly different, students in the intervention group refrained from ordering expensive tests.

Human papillomavirus clearance among males is associated with HIV acquisition and increased dendritic cell density in the foreskin.

BACKGROUND: The association between human papillomavirus (HPV) infection and the risk of human immunodeficiency virus (HIV) seroconversion is unclear, and the genital cellular immunology has not been evaluated. METHODS: A case-control analysis nested within a male circumcision trial was conducted. Cases consisted of 44 male HIV seroconverters, and controls were 787 males who were persistently negative for HIV. The Roche HPV Linear Array Genotype Test detected high-risk HPV (HR-HPV) and low-risk HPV (LR-HPV) genotypes. Generalized estimating equations logistic regression was used to estimate adjusted odds ratios (aORs) of HIV seroconversion. In addition, densities of CD1a(+) dendritic cells, CD4(+) T cells, and CD8(+) T cells were measured using immunohistochemistry analysis in foreskins of 79 males randomly selected from participants in the circumcision trial. RESULTS: HR-HPV or LR-HPV acquisition was not significantly associated with HIV seroconversion, after adjustment for sexual behaviors. However, HR-HPV and LR-HPV clearance was significantly associated with HIV seroconversion (aOR, 3.25 [95% confidence interval {CI}, 1.11-9.55] and 3.18 [95% CI, 1.14-8.90], respectively). The odds of HIV seroconversion increased with increasing number of HPV genotypes cleared (P < .001, by the test for trend). The median CD1a(+) dendritic cell density in the foreskin epidermis was significantly higher among males who cleared HPV (72.0 cells/mm(2) [interquartile range {IQR}, 29.4-138.3 cells/mm(2)]), compared with males who were persistently negative for HPV (32.1 cells/mm(2) [IQR, 3.1-96.2 cells/mm(2)]; P = .047), and increased progressively with the number of HPV genotypes cleared (P = .05). CONCLUSIONS: HPV clearance was associated with subsequent HIV seroconversion and also with increased epidermal dendritic cell density, which potentially mediates HIV seroconversion.

Understanding provider use of a new clinical decision support tool aimed at reducing excess telemetry in an academic health system: A retrospective study.

RATIONALE: Cardiac monitoring has often been identified as an area of overutilization and remains a limited resource in many hospitals. With the aim of reducing telemetry overuse, we added clinical decision support to our health system's telemetry order with guidance on appropriate indications for monitoring. The new order requires selection of an appropriate clinical indication. AIMS AND OBJECTIVES: In this study, we aimed to understand provider engagement with this tool by assessing concordance between selected indications within the order and the clinical presence of those conditions as documented within the patient chart. METHODS: We randomly selected 100 telemetry orders from July to October 2022 across four different hospitals at NYU Langone Health. Two independent, blinded reviewers used a structured protocol to identify documentation of actual indications for telemetry in each selected chart. We calculated the rate of concordance between indications selected in the order and indications that were determined to be clinically present on chart review. RESULTS: There were 30,839 telemetry orders placed during the study timeframe. Overall concordance between the selection within the order and the actual indication was 48% (95% confidence interval [CI], 38.21%-57.79%). We observed especially low concordance rates for vague indications, such as 'Other', and for 'Confirmed Stroke', which was the only indication allowing for indefinite telemetry. CONCLUSION: The overall low concordance suggests a disconnect between the support tool and clinical practice. Providers are more likely to select an indication that reduces downstream work regardless of a patient's true clinical indication.

High-risk human papillomavirus prevalence is associated with HIV infection among heterosexual men in Rakai, Uganda.

OBJECTIVES: Human papillomavirus (HPV) infection causes genital warts, penile cancer and cervical cancer. Africa has one of the highest rates of penile and cervical cancers, but there are little data on high-risk human papillomavirus (HR-HPV) prevalence in heterosexual men. Knowledge of HR-HPV prevalence, risk factors and genotype distribution among heterosexual men is important to establish risk-reduction prevention strategies. METHODS: 1578 uncircumcised men aged 15-49 years who enrolled in male circumcision trials in Rakai, Uganda, were evaluated for HR-HPV from swabs of the coronal sulcus/glans using Roche HPV Linear Array. Adjusted prevalence risk ratios (adjPRRs) were estimated using modified Poisson multivariable regression. RESULTS: HPV prevalence (either high risk or low risk) was 90.7% (382/421) among HIV-positive men and 60.9% (596/978) among HIV-negative men (PRR 1.49, 95% CI 1.40 to 1.58). HIV-positive men had a significantly higher risk of infection with three or more HR-HPV genotypes (PRR=5.76, 95% CI 4.27 to 7.79). Among HIV-positive men, high-risk sexual behaviours were not associated with increased HR-HPV prevalence. Among HIV-negative men, HR-HPV prevalence was associated with self-reported genital warts (adjPRR 1.57, 95% CI 1.07 to 2.31). Among all men (both HIV negative and HIV positive), HR-HPV prevalence was associated with more than 10 lifetime sexual partners (adjPRR 1.30, 95% CI 1.01 to 1.66), consistent condom use (adjPRR 1.31, 95% CI 1.08 to 1.60) and HIV infection (adjPRR 1.80, 95% CI 1.60 to 2.02). HR-HPV prevalence was lower among men who reported no sexual partners during the past year (adjPRR 0.47, 95% CI 0.23 to 0.94). CONCLUSION: The burden of HR-HPV infection is high among heterosexual men in sub-Saharan Africa and most pronounced among the HIV-infected individuals.

Evaluating Whether an Inpatient Initiative to Time Lab Draws in the Evening Reduces Anemia.

BACKGROUND: Hospital acquired anemia is common during admission and can result in increased transfusion and length of stay. Recumbent posture is known to lead to lower hemoglobin measurements. We tested to see if an initiative promoting evening lab draws would lead to higher hemoglobin measurements due to more time in upright posture during the day and evening. METHODS: We included patients hospitalized on 2 medical units, beginning March 26, 2020 and discharged prior to January 25, 2021. On one of the units, we implemented an initiative to have routine laboratory draws in the evening rather than the morning starting on August 26, 2020. There were 1217 patients on the control unit and 1265 on the intervention unit during the entire study period. First we used a linear mixed-effects model to see if timing of blood draw was associated with hemoglobin level in the pre-intervention period. We then compared levels of hemoglobin before and after the intervention using a difference-in-difference analysis. RESULTS: In the pre-intervention period, evening blood draws were associated with higher hemoglobin compared to morning (0.28; 95% CI, 0.22-0.35). Evening blood draws increased with the intervention (10.3% vs 47.9%, P > 0.001). However, the intervention floor was not associated with hemoglobin levels in difference-in-difference analysis (coefficient of -0.15; 95% CI, -0.51-0.21). CONCLUSIONS: While evening blood draws were associated with higher hemoglobin levels, an intervention that successfully changed timing of routine labs to the evening did not lead to an increase in hemoglobin levels.

Effect of circumcision of HIV-negative men on transmission of human papillomavirus to HIV-negative women: a randomised trial in Rakai, Uganda.

BACKGROUND: Randomised trials show that male circumcision reduces the prevalence and incidence of high-risk human papillomavirus (HPV) infection in men. We assessed the efficacy of male circumcision to reduce prevalence and incidence of high-risk HPV in female partners of circumcised men. METHODS: In two parallel but independent randomised controlled trials of male circumcision, we enrolled HIV-negative men and their female partners between 2003 and 2006, in Rakai, Uganda. With a computer-generated random number sequence in blocks of 20, men were assigned to undergo circumcision immediately (intervention) or after 24 months (control). HIV-uninfected female partners (648 of men from the intervention group, and 597 of men in the control group) were simultaneously enrolled and provided interview information and self-collected vaginal swabs at baseline, 12 months, and 24 months. Vaginal swabs were tested for high-risk HPV by Roche HPV Linear Array. Female HPV infection was a secondary endpoint of the trials, assessed as the prevalence of high-risk HPV infection 24 months after intervention and the incidence of new infections during the trial. Analysis was by intention-to-treat. An as-treated analysis was also done to account for study-group crossovers. The trials were registered, numbers NCT00425984 and NCT00124878. FINDINGS: During the trial, 18 men in the control group underwent circumcision elsewhere, and 31 in the intervention group did not undergo circumcision. At 24-month follow-up, data were available for 544 women in the intervention group and 488 in the control group; 151 (27·8%) women in the intervention group and 189 (38·7%) in the control group had high-risk HPV infection (prevalence risk ratio=0·72, 95% CI 0·60-0·85, p=0·001). During the trial, incidence of high-risk HPV infection in women was lower in the intervention group than in the control group (20·7 infections vs 26·9 infections per 100 person-years; incidence rate ratio=0·77, 0·63-0·93, p=0·008). INTERPRETATION: Our findings indicate that male circumcision should now be accepted as an efficacious intervention for reducing the prevalence and incidence of HPV infections in female partners. However, protection is only partial; the promotion of safe sex practices is also important. FUNDING: The Bill & Melinda Gates Foundation, National Institutes of Health, and Fogarty International Center.

Male circumcision and anatomic sites of penile high-risk human papillomavirus in Rakai, Uganda.

Male circumcision (MC) reduces penile high-risk human papillomavirus (HR-HPV) on the coronal sulcus and urethra. HR-HPV varies by anatomic site, and it is unknown whether MC decreases HR-HPV on the penile shaft. We assessed the efficacy of MC to reduce HR-HPV on the penile shaft and compared it to known efficacy of MC to reduce HR-HPV on the coronal sulcus. HIV-negative men randomized to receive immediate circumcision (intervention) or circumcision delayed for 24 months (control) were evaluated for HR-HPV at 12 months postenrollment using the Roche HPV Linear Array assay. Among swabs with detectable ß-globin or HPV, year 1 HR-HPV prevalence on the coronal sulcus was 21.5% in the intervention arm and 36.3% in the control arm men [adjusted prevalence risk ratios (PRRs) = 0.57, 95% CI 0.39-0.84, p = 0.005]. On the shaft, year 1 HR-HPV prevalence was 15.5% in the intervention and 23.8% in the control arm (adjusted PRR = 0.66, 95% CI 0.39-1.12, p = 0.12). Efficacy of MC to reduce HR-HPV on the shaft was similar to efficacy on the coronal sulcus (p = 0.52). In a sensitivity analysis in which swabs without detectable ß-globin or HPV were included as HPV negative, prevalence of HR-HPV on the shaft was lower in the intervention arm (7.8%) than control arm (13.6%; PRR 0.57, 95% CI 0.33-0.99, p < 0.05). HR-HPV was more frequently detected on the coronal sulcus than penile shaft among uncircumcised men (36.3% vs. 23.8%, respectively, p = 0.02) and circumcised men (21.5% vs. 15.5%, respectively, p = 0.24). MC reduced HR-HPV prevalence on both the coronal sulcus and shaft.

Dapsone-Induced Hypoxia.

Dapsone is a common medication that is utilized in the treatment of dermatological conditions, pneumocystis pneumonia, and toxoplasmosis. Methemoglobinemia is a known but rare complication of dapsone therapy that can result in cyanosis. We present a case of a patient on dapsone therapy who developed hypoxia due to methemoglobinemia. This case emphasizes the importance of knowledge of drugs likely to cause methemoglobinemia which requires clinicians to have a high degree of suspicion especially when the patient's oxygen saturation does not improve with treatment.

Characterization of gamma-butyrolactone autoregulatory signaling gene homologs in the angucyclinone polyketide WS5995B producer Streptomyces acidiscabies.

Organisms belonging to the genus Streptomyces produce numerous important secondary metabolites and undergo a sophisticated morphological differentiation program. In many instances these processes are under the control of gamma-butyrolactone (GBL) autoregulatory systems. Streptomyces acidiscabies strain 84.104 produces the secondary metabolite aromatic angucyclinone polyketide WS5995B. In order to explore the role of GBL regulatory circuitry in WS5995B production and morphogenesis in S. acidiscabies, a gene cluster encoding GBL autoregulatory signaling homologs was identified and characterized. Two GBL receptor homologs, sabR and sabS, were found flanking a GBL synthase homolog sabA. Strains carrying mutations in sabS produced elevated levels of WS5995B and displayed conditional morphological defects reminiscent of defects seen in Streptomyces bldA mutants. Notably, sabS possesses a TTA codon predicted to be recognized by tRNA(leu). sabA mutants produced higher levels of WS5995B than the wild-type strain but to a lesser extent than the levels of WS5995B seen in sabS mutants. Purified recombinant SabR and SabS were tested for their abilities to bind predicted AT-rich autoregulatory element (ARE) boxes within the sabRAS region. SabS did not bind any DNA sequences in this region, while SabR bound an ARE box in the region upstream of sabS. Quantitative reverse transcription-PCR analysis revealed higher levels of sabS transcript in sabR mutants than in the wild-type strain, suggesting that sabS expression is repressed by SabR. Based on these data, we propose that the S. acidiscabies sabRAS genes encode components of a signaling pathway which participates in the regulation of WS5995B production and morphogenesis.

Impact of nonintrusive clinical decision support systems on laboratory test utilization in a large academic centre.

BACKGROUND: The near-universal prevalence of electronic health records (EHRs) has made the utilization of clinical decision support systems (CDSS) an integral strategy for improving the value of laboratory ordering. Few studies have examined the effectiveness of nonintrusive CDSS on inpatient laboratory utilization in large academic centres. METHODS: Red blood cell folate, hepatitis C virus viral loads and genotypes, and type and screens were selected for study. We incorporated the appropriate indications for these labs into text that accompanied the laboratory orders in our hospital's EHR. Providers could proceed with the order without additional clicks. An interrupted time-series analysis was performed, and the primary outcome was the rate of tests ordered on all inpatient medicine floors. RESULTS: The rate of folate tests ordered per monthly admissions showed no significant level change at the time of the intervention with only a slight decrease in rate of 0.0109 (P = .07). There was a 43% decrease in the rate of hepatitis C virus tests per monthly admissions immediately after the intervention with a decrease of 0.0135 tests per monthly admissions (P = .02). The rate of type and screens orders per patient days each month had a significant downward trend by 0.114 before the intervention (P = .04) but no significant level change at the time of the intervention or significant change in rate after the intervention. DISCUSSION: Our study suggests that nonintrusive CDSS should be evaluated for individual laboratory tests to ensure only effective alerts continue to be used so as to avoid increasing EHR fatigue.

Evidence-Based Guidelines to Eliminate Repetitive Laboratory Testing.

Routine daily laboratory testing of hospitalized patients reflects a wasteful clinical practice that threatens the value of health care. Choosing Wisely initiatives from numerous professional societies have identified repetitive laboratory testing in the face of clinical stability as low value care. Although laboratory expenditure often represents less than 5% of most hospital budgets, the impact is far-reaching given that laboratory tests influence nearly 60% to 70% of all medical decisions. Excessive phlebotomy can lead to hospital-acquired anemia, increased costs, and unnecessary downstream testing and procedures. Efforts to reduce the frequency of laboratory orders can improve patient satisfaction and reduce cost without negatively affecting patient outcomes. To date, numerous interventions have been deployed across multiple institutions without a standardized approach. Health care professionals and administrative leaders should carefully strategize and optimize efforts to reduce daily laboratory testing. This review presents an evidence-based implementation blueprint to guide teams aimed at improving appropriate routine laboratory testing among hospitalized patients.

Human papillomavirus incidence and clearance among HIV-positive and HIV-negative men in sub-Saharan Africa.

OBJECTIVES: High-risk human papillomavirus (HR-HPV) infection is the most common sexually transmitted infection. Penile and cervical cancer rates are highest in sub-Saharan Africa. However, little is known about the impact of HIV infection on HR-HPV acquisition and clearance among heterosexual men. DESIGN: HR-HPV incidence and clearance were evaluated in 999 men (776 HIV-negative and 223 HIV-positive) aged 15-49 years who participated in male circumcision trials in Rakai, Uganda. METHODS: Penile swabs were tested for HR-HPV by Roche HPV Linear Array. A Poisson multivariable model was used to estimate adjusted incidence rate ratios (adjIRRs) and clearance risk ratios (adjRRs). RESULTS: HR-HPV incidence was 66.5/100 person-years in HIV-positive men and 32.9/100 person-years among HIV-negative men [IRR=2.02, 95% confidence interval (CI) 1.67-2.44]. Incidence was higher in the unmarried men (adjIRR=1.73, 95% CI 1.19-2.52), and decreased with age (adjIRR for men >35 years=0.64, 95% CI 0.43-0.94) and male circumcision (adjIRR=0.70, 95% CI 0.55-0.89). HR-HPV clearance was 114.7/100 person-years for HIV-positive men and 170.2/100 person-years for HIV-negative men (risk ratio=0.67, 95% CI 0.59-0.77). HR-HPV clearance in HIV-negative men increased with circumcision (adjRR=1.48, 95% CI 1.26-1.74), HSV-2 infection (adjRR=1.20, 95% CI 1.01-1.44), and symptoms of urethral discharge (adjRR=1.35, 95% CI 1.06-1.73). Clearance of HR-HPV was significantly lower for unmarried men (adjRR 0.76, 95% CI 0.59-0.98). CONCLUSION: HR-HPV is common among heterosexual Ugandan men, particularly the HIV-infected. HIV infection increases HR-HPV acquisition and reduces HR-HPV clearance. Promotion of male circumcision and additional prevention measures, such as HPV vaccination, is critical in sub-Saharan Africa.

Factors associated with incorrect identification of recent HIV infection using the BED capture immunoassay.

The BED capture enzyme immunoassay (BED-CEIA) was developed for estimating HIV incidence from cross-sectional data. This assay misclassifies some individuals with nonrecent HIV infection as recently infected, leading to overestimation of HIV incidence. We analyzed factors associated with misclassification by the BED-CEIA. We analyzed samples from 383 men who were diagnosed with HIV infection less than 1 year after a negative HIV test (Multicenter AIDS Cohort Study). Samples were collected 2-8 years after HIV seroconversion, which was defined as the midpoint between the last negative and first positive HIV test. Samples were analyzed using the BED-CEIA with a cutoff of OD-n ≤ 0.8 for recent infection. Logistic regression was used to identify factors associated with misclassification. Ninety-one (15.1%) of 603 samples were misclassified. In multivariate models, misclassification was independently associated with highly active antiretroviral treatment (HAART) for >2 years, HIV RNA <400 copies/ml, and CD4 cell count <50 or <200 cells/mm(3); adjusted odds ratios (OR) and 95% confidence intervals (CI) were 4.72 (1.35-16.5), 3.96 (1.53-10.3), 6.85 (2.71-17.4), and 11.5 (3.64-36.0), respectively. Among 220 men with paired samples, misclassification 2-4 years after seroconversion was significantly associated with misclassification 6-8 years after seroconversion [adjusted OR: 25.8 (95% CI: 8.17-81.5), p<0.001] after adjusting for race, CD4 cell count, HIV viral load, and HAART use. Low HIV viral load, low CD4 cell count, and >2 years of HAART were significantly associated with misclassification using the BED-CEIA. Some men were persistently misclassified as recently infected up to 8 years after HIV seroconversion.

Circumcision of HIV-infected men and transmission of human papillomavirus to female partners: analyses of data from a randomised trial in Rakai, Uganda.

BACKGROUND: Male circumcision reduces the transmission of high-risk human papillomavirus (HPV) in HIV-uninfected men and their female partners. We assessed whether circumcision of HIV-infected men would reduce the transmission of high-risk HPV to their female partners. METHODS: Female partners of HIV-infected men (aged 15-49 years) in Rakai, Uganda, with CD4 counts of greater than 350 cells per mL who were randomly assigned to undergo circumcision immediately (intervention group) and after 24 months (control group) were assessed for infection with high-risk HPV. Randomisation was done in blocks of 20, stratified by community, with computer-generated random numbers. Laboratory technicians and female fieldworkers were masked to the circumcision status of male participants. The main outcome assessed in this study was the effects of circumcision of HIV-infected men on transmission of HPV to their female partners. Analysis was by intention to treat. The trial is registered with ClinicalTrials.gov, number NCT00124878. FINDINGS: 474 men were assigned to the intervention group, 448 to the control group. 211 women were in consensual relationships with 193 men in the intervention group, and 171 women were in consensual unions with 155 men in the control group. High-risk HPV at the 2-year follow-up was prevalent in 88 female partners (55%) of 159 men in the intervention group and 68 (52%) of 131 female partners of men in the control group (prevalence risk ratio 1·07, 95% CI 0·86-1·32, p=0·64). Incidence of high-risk HPV over 2 years was 32·0 per 100 person-years in the female partners of men in the intervention group and 30·6 per 100 person-years in the female partners of men in the control group (incidence rate ratio 1·05, 0·77-1·43, p=0·78). No difference was noted in the clearance of genotype-specific high-risk HPV between the intervention group (196 [46%] of 424) and control group (167 [48%] of 347; rate ratio 0·96, 0·83-1·12; p=0·61). INTERPRETATION: Because circumcision of HIV-infected men did not affect transmission of high-risk HPV to their female partners, promotion of consistent safe sexual practices for HIV-infected men remains important. FUNDING: Bill & Melinda Gates Foundation and National Institutes of Health.

C-reactive protein levels increase during HIV-1 disease progression in Rakai, Uganda, despite the absence of microbial translocation.

INTRODUCTION: Microbial translocation has been implicated as a contributing factor to the heightened immune activation observed during HIV-1 disease progression. When examined in a longitudinal study of HIV-1 seroconverters in Rakai, Uganda, microbial translocation was not associated with HIV-1 disease progression. However, the role of general immune activation in HIV disease progression in this population was not fully examined. METHODS: Longitudinal serum samples of HIV-1 seroconverters in three HIV-1 disease progression groups [long-term nonprogressors (LTNP), standard progressors (SP), and rapid progressors (RP)] from Rakai, Uganda, were tested for levels of C-reactive protein (CRP), a marker for immune activation. RESULTS: CRP levels significantly increased in the SP group (P < 0.0001) but not in the RP group or the LTNP group. CRP levels during the first year post-HIV seroconversion in the RP group were significantly higher than those observed in the LTNP group (P < 0.05). For the entire population, CRP levels negatively correlated with lipopolysaccharide levels (P < 0.05) and were not associated with endotoxin antibody levels. CONCLUSIONS: This study suggests that in this population, increased immune activation is significantly associated with HIV-1 disease progression but not microbial translocation.