Investigating the Accuracy and Completeness of an Artificial Intelligence Large Language Model About Uveitis: An Evaluation of ChatGPT.

PURPOSE: To assess the accuracy and completeness of ChatGPT-generated answers regarding uveitis description, prevention, treatment, and prognosis. METHODS: Thirty-two uveitis-related questions were generated by a uveitis specialist and inputted into ChatGPT 3.5. Answers were compiled into a survey and were reviewed by five uveitis specialists using standardized Likert scales of accuracy and completeness. RESULTS: In total, the median accuracy score for all the uveitis questions (n = 32) was 4.00 (between "more correct than incorrect" and "nearly all correct"), and the median completeness score was 2.00 ("adequate, addresses all aspects of the question and provides the minimum amount of information required to be considered complete"). The interrater variability assessment had a total kappa value of 0.0278 for accuracy and 0.0847 for completeness. CONCLUSION: ChatGPT can provide relatively high accuracy responses for various questions related to uveitis; however, the answers it provides are incomplete, with some inaccuracies. Its utility in providing medical information requires further validation and development prior to serving as a source of uveitis information for patients.

A latent profile analysis of tear cytokines and their association with severity of dry eye disease in the Dry Eye Assessment and Management (DREAM) study.

This study is to identify subgroups of DED patients with different tear cytokine profiles and compare their DED symptoms and signs among subgroups. Baseline tear cytokines (IL-1ß, IL-6, IL-8, IL-10, IL-17A, IFN-γ and TNF-α) were measured using a magnetic bead assay. DED symptoms were assessed by Ocular Surface Disease Index (OSDI) and signs were assessed by corneal and conjunctival staining, tear break-up time (TBUT), Schirmer's test, tear osmolarity and meibomian gland dysfunction (MGD). Latent profile analysis was performed to identify subgroups, and their scores of DED symptoms and signs were compared using generalized linear regression. Among 131 patients with total tear volume > 4 µl from both eyes, subgroup 1 (n = 23) significantly higher in IL-6 and IL-8 (all p < 0.001) and subgroup 2 (n = 108) significantly higher in IL-10 (p = 0.03), IL-17A (p < 0.001), and IFN-γ (p < 0.001). Both subgroups were similar in demographics and DED symptoms, but subgroup 1 had significantly more severe DED signs: higher conjunctival staining (3.38 vs. 2.69, p = 0.04), corneal staining (4.26 vs. 3.03, p = 0.03), lower Schirmer's test score (8.20 vs. 13.72, p < 0.001), and higher composite severity score of DED sign (0.62 vs. 0.45, p = 0.002). We identified two DED subgroups with different profiles of tear cytokines. Patients in these subgroups differed significantly in DED signs, supporting the inflammation's role in DED development and progression.

A Geodemographic Analysis of Travel Time to Uveitis Specialists in the United States.

PURPOSE: Travel time to a patient's medical provider represents a significant component of access to care. We examined travel time to the nearest uveitis specialist for the American population and characterize its impact on access to uveitis care. DESIGN: Observational studies using the American Community Survey and American Census Bureau population estimates. METHODS: Addresses of fellowship-trained uveitis specialists were collected from the American Uveitis Society (AUS) and the Ocular Immunology and Uveitis Foundation (OIUF) websites and geocoded using ArcGIS Pro 2.9. Service areas were defined as all locations within a 60-min drive time from each specialist's location. Demographic and population data for total population, racial groups, household poverty levels, population in dependent groups (younger than 18, older than 65), and health insurance status were overlaid. Data were aggregated for census tracts within and outside service areas and compared using chi-square analysis. Main Outcome Measures: Differences in population demographics for people within and outside service area coverage. RESULTS: 223 uveitis specialist addresses were geocoded into ArcGIS. Of specialist locations, 94% were found in urban areas. Of the total United States population, 63.3% were found to be within service areas. Of these, 55.9% were White and 14.7% were African American, whereas 70% of the people outside service areas were White and 10.1% were African American (p < 0.0001) (Table 1). Only 8.3% of the people within service area coverage had no health insurance compared with 9.5% outside service areas (p < 0.0001). Within coverage areas, 12.7% of the households had a total income below the federal poverty level versus 15.1% of the households outside service area coverage (p < 0.0001). Within service area coverage, 37.4% of the people were in a dependent age group compared with 40.4.% outside service area coverage (p < 0.0001). CONCLUSIONS: Our analysis shows a significant travel burden to the nearest uveitis specialist for a large proportion of Americans. More providers are needed in rural areas, as patients there are more likely to live under the poverty line, be uninsured, or belong to a dependent age group.

Comparing weight loss outcomes after conversion to Roux-en-Y gastric bypass versus duodenal switch from sleeve gastrectomy in a community hospital.

BACKGROUND: Recidivism after initial sleeve gastrectomy (SG) remains common. Revisional surgery to convert SG to Roux-en-Y gastric bypass (RYGB) or duodenal switch (DS) for additional weight loss is increasing. This study aims to compare the outcomes after conversion of SG to RYGB or DS. METHODS: A retrospective single-institution review was conducted from 2015 to 2021, identifying 75 patients who underwent conversion from prior SG to either RYGB (40) or DS (35). Mean excess body weight loss (EBWL) at 3, 6, 12, and 24 months was assessed and compared. Secondary measures of length of stay (LOS), procedure length, and 30-day readmission rate were also reviewed. RESULTS: Percentage EBWL for RYGB vs DS was 24.0% vs 18.8% at 3 months (N = 36 vs 26; P < 0.0491), 34.8% vs 29.0% at 6 months (N = 29 vs 17; P < 0.2192), 43.0% vs 40.1% at 12 months (N = 28 vs 12; P < 0.6828), and 36.2% vs 41.7% at 24 months (N = 27 vs 7; P < 0.5553). Average LOS was 2.6 days ± 1.4 for RYGB and 2.8 days ± 1.3 for DS (P < 0.6032). Average procedure length was 134.4 min for RYGB and 189.8 min for DS (P < 0.0001). 30-day readmission rate was 27.5% (N = 11) for RYGB and 14.3% (N = 5) for DS (P < 0.1645). Significant weight loss was observed in both subgroups up to 12 months, with no significant weight loss between 12 and 24 months (RYGB N = 21, P < 0.2961; DS N = 5, P < 0.7233). CONCLUSION: Both revisional RYGB and revisional DS procedures had significant and sustained weight loss in the first 12 months. There was no significant excess body weight loss difference between revisional RYGB and revisional DS patients at 6, 12, and 24 months, with only significant greater weight loss for RYGB patients at 3 months. Additionally, procedure length was significantly longer for DS compared to RYGB, with no significant differences in LOS and 30-day readmission rates.

Treatment of Pulmonary Hypertension With Angiotensin II Receptor Blocker and Neprilysin Inhibitor Sacubitril/Valsartan.

BACKGROUND: Angiotensin II has been implicated in maladaptive right ventricular (RV) hypertrophy and fibrosis associated with pulmonary hypertension (PH). Natriuretic peptides decrease RV afterload by promoting pulmonary vasodilation and inhibiting vascular remodeling but are degraded by neprilysin. We hypothesized that angiotensin receptor blocker and neprilysin inhibitor, sacubitril/valsartan (Sac/Val, LCZ696), will attenuate PH and improve RV function by targeting both pulmonary vascular and RV remodeling. METHODS: PH was induced in rats using the SU5416/hypoxia model (Su/Hx), followed by 6-week treatment with placebo, Sac/Val, or Val alone. There were 4 groups: CON-normoxic animals with placebo (n=18); PH-Su/Hx rats+placebo (n=34); PH+Sac/Val (N=24); and PH+Val (n=16). RESULTS: In animals with PH, treatment with Sac/Val but not Val resulted in significant reduction in RV pressure (mm Hg: PH: 62±4, PH+Sac/Val: 46±5), hypertrophy (RV/LV+S: PH: 0.74±0.06, PH+Sac/Val: 0.46±0.06), collagen content (µg/50 µg protein: PH: 8.2±0.3, PH+Sac/Val: 6.4±0.4), pressures and improvement in RVs (mm/s: PH: 31.2±1.8, PH+Sac/Val: 43.1±3.6) compared with placebo. This was associated with reduced pulmonary vascular wall thickness, increased lung levels of ANP (atrial natriuretic peptide), BNP (brain-type natriuretic peptide), and cGMP, and decreased plasma endothelin-1 compared with PH alone. Also, PH+Sac/Val animals had altered expression of PKC isozymes in RV tissue compared with PH alone. CONCLUSIONS: Sac/Val reduces pulmonary pressures, vascular remodeling, as well as RV hypertrophy in a rat model of PH and may be appropriate for treatment of pulmonary hypertension and RV dysfunction.