Implementation of Artificial Intelligence-Based Diabetic Retinopathy Screening in a Tertiary Care Hospital in Quebec: Prospective Validation Study.

BACKGROUND: Diabetic retinopathy (DR) affects about 25% of people with diabetes in Canada. Early detection of DR is essential for preventing vision loss. OBJECTIVE: We evaluated the real-world performance of an artificial intelligence (AI) system that analyzes fundus images for DR screening in a Quebec tertiary care center. METHODS: We prospectively recruited adult patients with diabetes at the Centre hospitalier de l'Université de Montréal (CHUM) in Montreal, Quebec, Canada. Patients underwent dual-pathway screening: first by the Computer Assisted Retinal Analysis (CARA) AI system (index test), then by standard ophthalmological examination (reference standard). We measured the AI system's sensitivity and specificity for detecting referable disease at the patient level, along with its performance for detecting any retinopathy and diabetic macular edema (DME) at the eye level, and potential cost savings. RESULTS: This study included 115 patients. CARA demonstrated a sensitivity of 87.5% (95% CI 71.9-95.0) and specificity of 66.2% (95% CI 54.3-76.3) for detecting referable disease at the patient level. For any retinopathy detection at the eye level, CARA showed 88.2% sensitivity (95% CI 76.6-94.5) and 71.4% specificity (95% CI 63.7-78.1). For DME detection, CARA had 100% sensitivity (95% CI 64.6-100) and 81.9% specificity (95% CI 75.6-86.8). Potential yearly savings from implementing CARA at the CHUM were estimated at CAD $245,635 (US $177,643.23, as of July 26, 2024) considering 5000 patients with diabetes. CONCLUSIONS: Our study indicates that integrating a semiautomated AI system for DR screening demonstrates high sensitivity for detecting referable disease in a real-world setting. This system has the potential to improve screening efficiency and reduce costs at the CHUM, but more work is needed to validate it.

Experiences of Using Cochrane Systematic Reviews by Local HTA Units.

This study evaluated the use of Cochrane systematic reviews (CSRs) by Quebec's local health technology assessment (HTA) units to promote efficiency in hospital decision-making. An online survey was conducted to examine: Characteristics of the HTA units; Knowledge about works and services from the Cochrane Collaboration; Level of satisfaction about the use of CSRs; Facilitating factors and barriers to the implementation of CSRs evidence in a local context; Suggestions to improve the use of CSRs. Data accuracy was checked by 2 independent evaluators. Ten HTA units participated. From their implementation a total of 321 HTA reports were published (49.8% included a SR). Works and services provided by the Cochrane collaboration were very well-known and HTA units were highly satisfied with CSRs (80%-100%). As regards to applicability in HTA and use of CSRs, major strengths were as follow: Useful as resource for search terms and background material; May reduce the workload (eg, brief review instead of full SR); Use to update a current review. Major weaknesses were: Limited use since no CSRs were available for many HTA projects; Difficulty to apply findings to local context; Focused only on efficacy and innocuity; Cannot be used as a substitute to a full HTA report. This study provided a unique context of assessment with a familiar group of producers, users and disseminators of CSRs in hospital setting. Since they generally used other articles from the literature or produce an original SR in complement with CSRs, this led to suggestions to improve their use of CSRs. However, the main limit for the use of CRS in local HTA will remain its lack of contextualisation. As such, this study reinforces the need to consider the notion of complementarity of experimental data informing us about causality and contextual data, allowing decision-making adapted to local issues.

Increased glucose excursion in cystic fibrosis and its association with a worse clinical status.

BACKGROUND: Abnormal glucose tolerance is a frequent co-morbidity in cystic fibrosis patients (CF), and is associated with a worse prognosis. The objectives are to investigate (a) the relative contribution of insulinopenia and insulin resistance (IR) for glucose tolerance and (b) the association between various glucose parameters and CF clinical status. METHODS: Oral glucose tolerance tests were performed in 114 consecutive CF patients not known to be diabetic as well as 14 controls similar for age and BMI. RESULTS: Abnormal glucose tolerance was found in 40% of patients with CF: 28% had impaired glucose tolerance (IGT) and 12% had new cystic fibrosis related diabetes (CFRD). Compared to control subjects, all CF patients were characterized by an increased glucose excursion (AUC). While reduced early insulin release characterised CF, IGT and CFRD patients also present IR thus both mechanisms significantly contribute to glucose tolerance abnormalities. Increased glucose AUC and reduced early insulin release but not glucose tolerance categories were associated with a reduced pulmonary function (FEV(1)). CONCLUSION: In CF, early insulin secretion defect but also IR contribute to glucose intolerance. Early in the course of the disease, increased glucose AUC and reduced early insulin secretion are more closely associated with a worse clinical status than conventional glucose tolerance categories.

Variation of glucose tolerance in adult patients with cystic fibrosis: What is the potential contribution of insulin sensitivity?

BACKGROUND: Reduced insulin secretion is a key factor to explain high prevalence of glucose intolerance in patients with cystic fibrosis (CF). However, the role of insulin sensitivity remains unclear. The aim of this study is to investigate the association of insulin secretion and sensitivity with the evolution of glucose tolerance. METHODS: A total of 152 patients without known diabetes from the Montreal CF cohort underwent two 2-h oral glucose tolerance tests (OGTT) at baseline and again after 21.2±5.5months. Pulmonary function and anthropometric measurements were also collected at each visit. At both visits, based on their OGTT results, patients were categorized in glucose tolerance groups (normal glucose tolerance, impaired glucose tolerance or CF-related diabetes) and stratified in 3 groups according to the variation of their glucose tolerance: stable, improved or deteriorated. RESULTS: At baseline, patients in the deteriorated group had a better sensitivity to insulin than those in the improved group (P=0.029). At follow-up glucose tolerance remained stable in 55.3%, improved in 14.5% and deteriorated in 30.3% of patients. During follow-up, insulin secretion remained stable in all 3 groups. While insulin sensitivity remained stable in patients without changes in glucose tolerance it worsened in patients who deteriorated glucose tolerance (P<0.001) and improved in patients who improved their glucose tolerance (P=0.003). CONCLUSION: In a context of significantly reduced insulin secretion, variations of insulin sensitivity are associated with variations of glucose tolerance in adult patients with CF.

Speech recognition in the radiology department: a systematic review.

OBJECTIVE: To conduct a systematic review of the literature describing the impact of speech recognition systems on report error rates and productivity in radiology departments. METHODS: The search was conducted for relevant papers published from January 1992 to October 2013. Comparative studies reporting any of the following outcomes were selected: error rates, departmental productivity, and radiologist productivity. The retrieved studies were assessed for quality and risk of bias. RESULTS: The literature search identified 85 potentially relevant publications, but, based on the inclusion and exclusion criteria, only 20 were included. Most studies were before and after assessments with no control group. There was a large amount of heterogeneity due to differences in the imaging modalities assessed and the outcomes measured. The percentage of reports containing at least one error varied from 4.8% to 89% for speech recognition, and from 2.1% to 22% for transcription. Departmental productivity was improved with decreases in report turnaround times varying from 35% to 99%. Most studies found a lengthening of radiologist dictation time. CONCLUSION: Overall gains in departmental productivity were high, but radiologist productivity, as measured by the time to produce a report, was diminished.