RESUMO
PURPOSE: To evaluate serum beta-D-glucan (BDG) as a biomarker for endogenous fungal eye infection. METHODS: Retrospective case-control study of 88 patients with a BDG test and eye examination at UPenn (2013-2018). Cases had endogenous fungal chorioretinitis or endophthalmitis diagnosed by eye examination and confirmed with positive culture; controls were without these fungal eye findings. Charts were reviewed for BDG values, blood/vitreous cultures, and eye examinations. Outcomes were BDG sensitivity, specificity, positive predictive value, and negative predictive value for fungal chorioretinitis or endophthalmitis, using prespecified BDG cut-off points of ≥80, ≥250, and ≥500 pg/mL as test positive. RESULTS: Cases included six chorioretinitis and four endophthalmitis patients. Controls included 78 patients without chorioretinitis or endophthalmitis. Defining BDG ≥80 pg/mL as test positive, the BDG sensitivity (95% confidence interval) was 66.7% (22.3%-95.7%) for chorioretinitis and 100% (39.8%-100%) for endophthalmitis. The specificity was 74.4% (63.2%-83.6%) when BDG values ≥80 pg/mL were test positive, and 85.9% (76.2%-92.7%) when values ≥250 pg/mL were test positive. For a 1% endophthalmitis prevalence and BDG cut-off value of ≥80 pg/mL, the positive predictive value was 3.8% (2.4%-5.2%) and negative predictive value was 100% (99.1%-100%). CONCLUSION: For endogenous fungal endophthalmitis, BDG's sensitivity and specificity seem good and the negative predictive value is high; a larger ophthalmic study is indicated.
Assuntos
Endoftalmite/diagnóstico , Infecções Oculares Fúngicas/diagnóstico , beta-Glucanas/sangue , Adulto , Biomarcadores/sangue , Estudos de Casos e Controles , Endoftalmite/sangue , Infecções Oculares Fúngicas/sangue , Feminino , Humanos , Masculino , Pessoa de Meia-Idade , Valor Preditivo dos Testes , Curva ROC , Estudos RetrospectivosRESUMO
OBJECTIVES: This work presents the development and evaluation of coordn8, a web-based application that streamlines fax processing in outpatient clinics using a "human-in-the-loop" machine learning framework. We demonstrate the effectiveness of the platform at reducing fax processing time and producing accurate machine learning inferences across the tasks of patient identification, document classification, spam classification, and duplicate document detection. METHODS: We deployed coordn8 in 11 outpatient clinics and conducted a time savings analysis by observing users and measuring fax processing event logs. We used statistical methods to evaluate the machine learning components across different datasets to show generalizability. We conducted a time series analysis to show variations in model performance as new clinics were onboarded and to demonstrate our approach to mitigating model drift. RESULTS: Our observation analysis showed a mean reduction in individual fax processing time by 147.5 s, while our event log analysis of over 7000 faxes reinforced this finding. Document classification produced an accuracy of 81.6%, patient identification produced an accuracy of 83.7%, spam classification produced an accuracy of 98.4%, and duplicate document detection produced a precision of 81.0%. Retraining document classification increased accuracy by 10.2%. DISCUSSION: coordn8 significantly decreased fax-processing time and produced accurate machine learning inferences. Our human-in-the-loop framework facilitated the collection of high-quality data necessary for model training. Expanding to new clinics correlated with performance decline, which was mitigated through model retraining. CONCLUSION: Our framework for automating clinical tasks with machine learning offers a template for health systems looking to implement similar technologies.
RESUMO
OBJECTIVES: Acute care ophthalmic clinics often suffer from inefficient triage, leading to suboptimal patient access and resource utilization. This study reports the preliminary results of a novel, symptom-based, patient-directed, online triage tool developed to address the most common acute ophthalmic diagnoses and associated presenting symptoms. METHODS: A retrospective chart review of patients who presented to a tertiary academic medical center's urgent eye clinic after being referred for an urgent, semi-urgent, or nonurgent visit by the ophthalmic triage tool between January 1, 2021 and January 1, 2022 was performed. Concordance between triage category and severity of diagnosis on the subsequent clinic visit was assessed. RESULTS: The online triage tool was utilized 1,370 and 95 times, by the call center administrators (phone triage group) and patients directly (web triage group), respectively. Of all patients triaged with the tool, 8.50% were deemed urgent, 59.2% semi-urgent, and 32.3% nonurgent. At the subsequent clinic visit, the history of present illness had significant agreement with symptoms reported to the triage tool (99.3% agreement, weighted kappa = 0.980, p < 0.001). The triage algorithm also had significant agreement with the severity of the physician diagnosis (97.0% agreement, weighted kappa = 0.912, p < 0.001). Zero patients were found to have a diagnosis on exam that should have corresponded to a higher urgency level on the triage tool. CONCLUSION: The automated ophthalmic triage algorithm was able to safely and effectively triage patients based on symptoms. Future work should focus on the utility of this tool to reduce nonurgent patient load in urgent clinical settings and to improve access for patients who require urgent medical care.
Assuntos
Oftalmologia , Triagem , Humanos , Triagem/métodos , Estudos Retrospectivos , Serviço Hospitalar de Emergência , Cuidados CríticosRESUMO
Introduction: A methodology for safe recovery of an ophthalmology department during a pandemic does not currently exist. This study describes successful recovery strategies for an urban, multi-specialty ophthalmology department serving a high-risk patient population. Methods: The study took place at a large multi-specialty tertiary care academic ophthalmology department in a metropolitan city during a seven-month period (March-October 2020). Five recovery ad hoc committees were charged with formulating metrics and initiatives to manage clinical volumes while maintaining safe practices, providing patient access, and minimizing financial damage. A six-tier system was created to resume non-urgent appointments in May 2020. Educational and research activities were maintained through the development of virtual curricula and research platforms. Results: The number of clinical and surgical visits per month in 2020 compared to 2019 and the time to reach ≥95% of pre-COVID patient volumes were monitored. In October 2020, ≥95% of pre-COVID volumes were attained (11,975 vs 12,337 patient visits in October 2019; 266 vs 272 surgical cases in October 2019). Despite significant financial losses, the department surpassed December 2019 collections in December 2020. No faculty, staff, or trainees received furloughs or pay cuts. There was no COVID-19 transmission between faculty, staff, and patients. Discussion: With strategic implementation of recovery strategies following CDC safety measures, it was possible to safely deliver care to patients with urgent and non-urgent eye conditions. Patient volumes were fully recovered in an ambulatory urban healthcare setting within a high-risk COVID-19 population within seven months while educational and research missions were successfully sustained.
RESUMO
PRCIS: In primary angle closure suspects (PACS), self-identified Black race was a risk factor for intraocular pressure (IOP) elevation and iritis following laser peripheral iridotomy (LPI). Laser type was not associated with either immediate post-LPI IOP elevation or iritis in multivariate analysis. PURPOSE: The aim was to determine the impact of laser type and patient characteristics on the incidence of IOP elevation and iritis after LPI in PACS. MATERIALS AND METHODS: The electronic medical records of 1485 PACS (2407 eyes) who underwent either neodymium-doped yttrium-aluminum-garnet or sequential argon and neodymium-doped yttrium-aluminum-garnet LPI at the University of Pennsylvania between 2010 and 2018 were retrospectively reviewed. Average IOP within 30 days before LPI (baseline IOP), post-LPI IOP within 1 hour, laser type, laser energy, and the incidence of new iritis within 30 days following the procedure were collected. Multivariate logistic regression accounting for intereye correlation was used to assess factors associated with incidence of post-LPI IOP elevation and iritis, adjusted by age, sex, surgeon, and histories of autoimmune disease, diabetes, and hypertension. RESULTS: The incidence of post-LPI IOP elevation and iritis were 9.3% (95% confidence interval: 8.1%-10.5%) and 2.6% (95% CI: 1.9%-3.2%), respectively. In multivariate analysis, self-identified Black race was a risk factor for both IOP elevation [odds ratio (OR): 2.08 compared with White; P=0.002] and iritis (OR: 5.07; P<0.001). Higher baseline IOP was associated with increased risk for post-LPI IOP elevation (OR: 1.19; P<0.001). Laser type and energy were not associated with either post-LPI IOP elevation or iritis (P>0.11 for all). CONCLUSIONS: The incidence of immediate IOP elevation and iritis following prophylactic LPI was higher in Black patients independent of laser type and energy. Heightened vigilance and increased medication management before and after the procedure are suggested to help mitigate these risks.
Assuntos
Glaucoma de Ângulo Fechado , Irite , Terapia a Laser , Glaucoma de Ângulo Fechado/diagnóstico , Glaucoma de Ângulo Fechado/cirurgia , Humanos , Pressão Intraocular , Iridectomia/métodos , Iris/cirurgia , Irite/cirurgia , Terapia a Laser/métodos , Estudos Retrospectivos , Fatores de RiscoRESUMO
OBJECTIVE: No method of grading visual field (VF) defects has been widely accepted throughout the glaucoma community. The SCHEIE (Systematic Classification of Humphrey visual fields-Easy Interpretation and Evaluation) grading system for glaucomatous visual fields was created to convey qualitative and quantitative information regarding visual field defects in an objective, reproducible, and easily applicable manner for research purposes. METHODS: The SCHEIE grading system is composed of a qualitative and quantitative score. The qualitative score consists of designation in one or more of the following categories: normal, central scotoma, paracentral scotoma, paracentral crescent, temporal quadrant, nasal quadrant, peripheral arcuate defect, expansive arcuate, or altitudinal defect. The quantitative component incorporates the Humphrey visual field index (VFI), location of visual defects for superior and inferior hemifields, and blind spot involvement. Accuracy and speed at grading using the qualitative and quantitative components was calculated for non-physician graders. RESULTS: Graders had a median accuracy of 96.67% for their qualitative scores and a median accuracy of 98.75% for their quantitative scores. Graders took a mean of 56 seconds per visual field to assign a qualitative score and 20 seconds per visual field to assign a quantitative score. CONCLUSION: The SCHEIE grading system is a reproducible tool that combines qualitative and quantitative measurements to grade glaucomatous visual field defects. The system aims to standardize clinical staging and to make specific visual field defects more easily identifiable. Specific patterns of visual field loss may also be associated with genetic variants in future genetic analysis.