Quality indicators in type 2 diabetes patient care: analysis per care-complexity level

Advancing healthcare with artificial intelligence: diagnostic accuracy of machine learning algorithm in diagnosis of diabetic retinopathy in the Brazilian population

BACKGROUND

In healthcare systems in general, access to diabetic retinopathy (DR) screening is limited. Artificial intelligence has the potential to increase care delivery. Therefore, we trained and evaluated the diagnostic accuracy of a machine learning algorithm for automated detection of DR.

METHODS

We included color fundus photographs from individuals from 4 databases (primary and specialized care settings), excluding uninterpretable images. The datasets consist of images from Brazilian patients, which differs from previous work. This modification allows for a more tailored application of the model to Brazilian patients, ensuring that the nuances and characteristics of this specific population are adequately captured. The sample was fractionated in training (70%) and testing (30%) samples. A convolutional neural network was trained for image classification. The reference test was the combined decision from three ophthalmologists. The sensitivity, specificity, and area under the ROC curve of the algorithm for detecting referable DR (moderate non-proliferative DR; severe non-proliferative DR; proliferative DR and/or clinically significant macular edema) were estimated.

RESULTS

A total of 15,816 images (4590 patients) were included. The overall prevalence of any degree of DR was 26.5%. Compared with human evaluators (manual method of diagnosing DR performed by an ophthalmologist), the deep learning algorithm achieved an area under the ROC curve of 0.98 (95% CI 0.97-0.98), with a specificity of 94.6% (95% CI 93.8-95.3) and a sensitivity of 93.5% (95% CI 92.2-94.9) at the point of greatest efficiency to detect referable DR.

CONCLUSIONS

A large database showed that this deep learning algorithm was accurate in detecting referable DR. This finding aids to universal healthcare systems like Brazil, optimizing screening processes and can serve as a tool for improving DR screening, making it more agile and expanding care access.

Organizational intervention to improve access to retinopathy screening for patients with diabetes mellitus: health care service improvement project in a tertiary public hospital

INTRODUCTION

The aim of this study was to improve the access to diabetic retinopathy (DR) screening with retinographies in a tertiary care center through a process conducted by an interdisciplinary group driven by a nurse.

MATERIALS AND METHODS

This is a quality improvement study that evaluated the flow for DR screening conducted by an interdisciplinary group using the Plan-Do-Study-Act methodology. As a result measure, we assessed the quantity of retinographies performed after the project implementation, the proportion of patients who presented abnormal retinographies, and the proportion of patients referred to a specialist.

RESULTS

The new patient screening flow and the reinforcement of available human resources led to an increase in the number of retinographies performed and screened patients. A total of 1184 retinographies were conducted, and it was possible to observe DR alterations in 37.8 % patients, of which only 6 % needed to be referred to the DR reference center.

CONCLUSIONS

This study showed a significant increase in the number of retinographies performed. The Plan-Do-Study-Act methodology was an important tool to improve the flow of patient access to fundus images, helping to consistently and continuously improve these processes.

Predictors of traffic events due to hypoglycemia in adults with type 1 diabetes: A Brazilian prospective cohort study

BACKGROUND

Individuals with type 1 diabetes (T1D) are exposed to an elevated risk of automobile accidents especially because of hypoglycemia that impairs physiological and defense responses.

OBJECTIVES

To assess local risk factors for traffic events in T1D adult Brazilian patients.

METHODS

This is a prospective study and 12-month follow-up to assess predictors for traffic events on a cohort of drivers with T1D (n = 168) in Brazil. The inclusion criteria for participants were Brazilian nationality, age ≥ 18 years-old, diagnosis of T1D for more than one year, driving license B, C or D categories (four-wheel vehicles), driving three-times per week or more, and checking blood glucose twice-daily or more. The primary outcome was hypoglycemia driving mishaps assessed by a seven-query questionnaire about the past 30 days. Secondary outcomes included driving mishaps not related to hypoglycemia. Statistical analysis was performed through Poisson regression models with robust variance estimarion, in which the measure of association is the relative risk.

RESULTS

A total of 109 participants completed the 12-month follow-up. Most of them were men (66%) and 37 ± 11 years-old, and had a mean HbA1c of 8.2% (66 mmol/mol). In the follow up, the incidence of traffic events was high (70.6%); however, only a minority was attributed to hypoglycemia as the cause of the reported event (19.3%). The best predictors for new traffic events due to hypoglycemia were those related to driving characteristics. The best of them was a history of episodes of hypoglycemia while driving [RR 3.40 (1.22-9.43); p < 0.05].

CONCLUSIONS

We found that previous episodes of hypoglycemia while driving significantly increase the risk of new traffic events and are the best predictor for it. This highlights the need to assess the risks of traffic accidents especially in people who have had experienced episodes of hypoglycemia while driving.

Integrated quality assessment for diabetes care in Portuguese primary health care using prevention quality indicators

AIMS

This study evaluates the prevention quality indicators (PQI) for Diabetes Mellitus (DM) in Portugal using contemporary data and explores their variability according to Primary Health Care (PHC) quality indicators.

METHODS

We conducted a retrospective observational analysis of secondary data comprising Portuguese PHC indicators by health centres group (ACES) and the National Hospital Morbidity Database. We calculated and analysed age-sex-adjusted rates for each PQI. Worse-performing ACES were identified using the 2017 median PQI values as an assessment cut-off. A multivariate logistic analysis was carried to find variables associated with the likelihood of being a worse-performing ACES for the biennium.

RESULTS

The median values of the indicator PQI93 - Prevention Quality Diabetes Composite were 79 and 65.2 hospitalizations per 100 000 pop, in 2016 and 2017 respectively. Diabetes long term complications (PQI 03) accounted for most of the hospitalizations. The quality indicator in PHC with greater influence on PQI93 was the proportion of DM patients with <65 years with test results for HbA1c < = 6.5%.

CONCLUSIONS

This study shows that some PHC quality indicators are closely related to DM care, and so their monitoring is of high importance. Diabetes long term complications (PQI 03) demand greater attention from PHC professionals.