ABSTRACT
Background: Many artificial intelligence (AI) studies have focused on development of AI models, novel techniques, and reporting guidelines. However, little is understood about clinicians' perspectives of AI applications in medical fields including ophthalmology, particularly in light of recent regulatory guidelines. The aim for this study was to evaluate the perspectives of ophthalmologists regarding AI in 4 major eye conditions: diabetic retinopathy (DR), glaucoma, age-related macular degeneration (AMD) and cataract. Methods: This was a multi-national survey of ophthalmologists between March 1st, 2020 to February 29th, 2021 disseminated via the major global ophthalmology societies. The survey was designed based on microsystem, mesosystem and macrosystem questions, and the software as a medical device (SaMD) regulatory framework chaired by the Food and Drug Administration (FDA). Factors associated with AI adoption for ophthalmology analyzed with multivariable logistic regression random forest machine learning. Results: One thousand one hundred seventy-six ophthalmologists from 70 countries participated with a response rate ranging from 78.8 to 85.8% per question. Ophthalmologists were more willing to use AI as clinical assistive tools (88.1%, n = 890/1,010) especially those with over 20 years' experience (OR 3.70, 95% CI: 1.10-12.5, p = 0.035), as compared to clinical decision support tools (78.8%, n = 796/1,010) or diagnostic tools (64.5%, n = 651). A majority of Ophthalmologists felt that AI is most relevant to DR (78.2%), followed by glaucoma (70.7%), AMD (66.8%), and cataract (51.4%) detection. Many participants were confident their roles will not be replaced (68.2%, n = 632/927), and felt COVID-19 catalyzed willingness to adopt AI (80.9%, n = 750/927). Common barriers to implementation include medical liability from errors (72.5%, n = 672/927) whereas enablers include improving access (94.5%, n = 876/927). Machine learning modeling predicted acceptance from participant demographics with moderate to high accuracy, and area under the receiver operating curves of 0.63-0.83. Conclusion: Ophthalmologists are receptive to adopting AI as assistive tools for DR, glaucoma, and AMD. Furthermore, ML is a useful method that can be applied to evaluate predictive factors on clinical qualitative questionnaires. This study outlines actionable insights for future research and facilitation interventions to drive adoption and operationalization of AI tools for Ophthalmology.
ABSTRACT
OBJECTIVES: The potential benefit of convalescent plasma (CP) therapy for coronavirus disease 2019 (COVID-19) is highest when administered early after symptom onset. Our objective was to determine the effectiveness of CP therapy in improving the disease course of COVID-19 among high-risk outpatients. METHODS: A multicentre, double-blind randomized trial was conducted comparing 300 mL of CP with non-CP. Patients were ≥50 years, were symptomatic for <8 days, had confirmed RT-PCR or antigen test result for COVID-19 and had at least one risk factor for severe COVID-19. The primary endpoint was the highest score on a 5-point ordinal scale ranging from fully recovered (score = 1) or not (score = 2) on day 7, over hospital admission (score = 3), intensive care unit admission (score = 4) and death (score = 5) in the 28 days following randomization. Secondary endpoints were hospital admission, symptom duration and viral RNA excretion. RESULTS: After the enrolment of 421 patients and the transfusion in 416 patients, recruitment was discontinued when the countrywide vaccination uptake in those aged >50 years was 80%. Patients had a median age of 60 years, symptoms for 5 days, and 207 of 416 patients received CP therapy. During the 28 day follow-up, 28 patients were hospitalized and two died. The OR for an improved disease severity score with CP was 0.86 (95% credible interval, 0.59-1.22). The OR was 0.58 (95% CI, 0.33-1.02) for patients with ≤5 days of symptoms. The hazard ratio for hospital admission was 0.61 (95% CI, 0.28-1.34). No difference was found in viral RNA excretion or in the duration of symptoms. CONCLUSIONS: In patients with early COVID-19, CP therapy did not improve the 5-point disease severity score.
ABSTRACT
Background Many artificial intelligence (AI) studies have focused on development of AI models, novel techniques, and reporting guidelines. However, little is understood about clinicians' perspectives of AI applications in medical fields including ophthalmology, particularly in light of recent regulatory guidelines. The aim for this study was to evaluate the perspectives of ophthalmologists regarding AI in 4 major eye conditions: diabetic retinopathy (DR), glaucoma, age-related macular degeneration (AMD) and cataract. Methods This was a multi-national survey of ophthalmologists between March 1st, 2020 to February 29th, 2021 disseminated via the major global ophthalmology societies. The survey was designed based on microsystem, mesosystem and macrosystem questions, and the software as a medical device (SaMD) regulatory framework chaired by the Food and Drug Administration (FDA). Factors associated with AI adoption for ophthalmology analyzed with multivariable logistic regression random forest machine learning. Results One thousand one hundred seventy-six ophthalmologists from 70 countries participated with a response rate ranging from 78.8 to 85.8% per question. Ophthalmologists were more willing to use AI as clinical assistive tools (88.1%, n = 890/1,010) especially those with over 20 years' experience (OR 3.70, 95% CI: 1.10–12.5, p = 0.035), as compared to clinical decision support tools (78.8%, n = 796/1,010) or diagnostic tools (64.5%, n = 651). A majority of Ophthalmologists felt that AI is most relevant to DR (78.2%), followed by glaucoma (70.7%), AMD (66.8%), and cataract (51.4%) detection. Many participants were confident their roles will not be replaced (68.2%, n = 632/927), and felt COVID-19 catalyzed willingness to adopt AI (80.9%, n = 750/927). Common barriers to implementation include medical liability from errors (72.5%, n = 672/927) whereas enablers include improving access (94.5%, n = 876/927). Machine learning modeling predicted acceptance from participant demographics with moderate to high accuracy, and area under the receiver operating curves of 0.63–0.83. Conclusion Ophthalmologists are receptive to adopting AI as assistive tools for DR, glaucoma, and AMD. Furthermore, ML is a useful method that can be applied to evaluate predictive factors on clinical qualitative questionnaires. This study outlines actionable insights for future research and facilitation interventions to drive adoption and operationalization of AI tools for Ophthalmology.
ABSTRACT
BACKGROUND: Critically ill COVID-19 patients have proven to be at risk for developing invasive fungal infections. However, the incidence and impact of possible/probable COVID-19-associated pulmonary aspergillosis (CAPA) in severe COVID-19 patients varies between cohorts. We aimed to assess the incidence, risk factors, and clinical outcome of invasive pulmonary aspergillosis in a regional cohort of COVID-19 intensive care patients. METHODS: We performed a regional, multicentre, retrospective cohort study in the intensive care units (ICUs) in North Brabant, The Netherlands. We included adult patients with rt-PCR-confirmed SARS-CoV-2 infection (COVID-19), requiring mechanical ventilation for acute respiratory distress syndrome. Demographics, clinical course, biomarker value, and treatment outcomes were compared between the groups with possible/probable CAPA from the main study centre and the regional centres, and without signs of CAPA from the main study centre as controls. The primary aim was to assess the regional impact of possible/probable CAPA in COVID-19 ICU patients, measured as all-cause mortality at 30 days after ICU admission. Secondary outcomes were risk factors for developing CAPA, based on underlying host factors and to identify the value of the mycological arguments for the diagnosing of CAPA. RESULTS: Between 1 March and 30 April 2020, we included 123 patients with severe COVID-19: 29 patients (30.9%) in the main ICU with possible/probable CAPA, and 65 (69.1%) with no signs of CAPA; 29 patients in the regional ICUs with signs of CAPA. Patients' characteristics and risk factors did not differ for CAPA and non-CAPA patients. Patients with COPD and/or chronic steroid medication developed CAPA more frequently, although this was not statistically significant. CAPA patients were admitted to the ICU earlier, had lower PF-ratios, and more often required renal replacement therapy. All-cause 30-day mortality was significantly higher in mechanically ventilated COVID-19 patients with possible/probable CAPA 39.7% (23/58) compared to patients without evidence for CAPA 16.9% (11/65) (OR 3.2 [95% CI 1.4-7.4] p = 0.005). CONCLUSION: The high incidence of possible and probable CAPA in critically ill COVID-19 patients is alarming. The increase in 30-day mortality in CAPA highlights the need for active surveillance and management strategies in critically ill COVID-19 patients.
ABSTRACT
AIM: To examine psychosocial and behavioural impacts of the novel coronavirus disease 2019 (COVID-19) pandemic and lockdown restrictions among adults with type 2 diabetes. METHODS: Participants enrolled in the PRogrEssion of DIabetic ComplicaTions (PREDICT) cohort study in Melbourne, Australia (n = 489 with a baseline assessment pre-2020) were invited to complete a phone/online follow-up assessment in mid-2020 (i.e., amidst COVID-19 lockdown restrictions). Repeated assessments that were compared with pre-COVID-19 baseline levels included anxiety symptoms (7-item Generalised Anxiety Disorder scale [GAD-7]), depressive symptoms (8-item Patient Health Questionnaire [PHQ-8]), diabetes distress (Problem Areas in Diabetes scale [PAID]), physical activity/sedentary behaviour, alcohol consumption and diabetes self-management behaviours. Additional once-off measures at follow-up included COVID-19-specific worry, quality of life (QoL), and healthcare appointment changes (telehealth engagement and appointment cancellations/avoidance). RESULTS: Among 470 respondents (96%; aged 66 ± 9 years, 69% men), at least 'moderate' worry about COVID-19 infection was reported by 31%, and 29%-73% reported negative impacts on QoL dimensions (greatest for: leisure activities, feelings about the future, emotional well-being). Younger participants reported more negative impacts (p < 0.05). Overall, anxiety/depressive symptoms were similar at follow-up compared with pre-COVID-19, but diabetes distress reduced (p < 0.001). Worse trajectories of anxiety/depressive symptoms were observed among those who reported COVID-19-specific worry or negative QoL impacts (p < 0.05). Physical activity trended lower (~10%), but sitting time, alcohol consumption and glucose-monitoring frequency remained unchanged. 73% of participants used telehealth, but 43% cancelled a healthcare appointment and 39% avoided new appointments despite perceived need. CONCLUSIONS: COVID-19 lockdown restrictions negatively impacted QoL, some behavioural risk factors and healthcare utilisation in adults with type 2 diabetes. However, generalised anxiety and depressive symptoms remained relatively stable.