Evaluation of the Effectiveness of Implementing a Screening Strategy for the Diagnosis of Prediabetes and Type 2 Diabetes Mellitus in Primary Health Care in a Brazilian Municipality (ARISE)

Background: Type 2 diabetes mellitus (T2DM) is a major cause of morbidity, disability, and premature mortality. A significant proportion of people with T2DM remain undiagnosed, and prediabetes is even less frequently identified. Although Primary Health Care has guidelines for screening, systematic identification of eligible individuals is hindered by data fragmentation, incomplete records, interoperability limitations, and the absence of mechanisms to operationalize the criteria of the Clinical Protocol and Therapeutic Guidelines for T2DM in teams' routine practice.

Primary objective: To evaluate the effectiveness of the ARISE strategy on the proportion of individuals with confirmed T2DM diagnosis, according to the criteria defined in the protocol and based on the PCDT-DM2 (Clinical Protocol and Therapeutic Guidelines for T2DM), among adults eligible for screening registered at the UBS (Primary Health Care Units) of Afogados da Ingazeira, during the 12 months following implementation, compared to a reference value estimated from a historical 12-month period in the same municipality.

Methodology: This is a pragmatic, quasi-experimental, single-arm effectiveness and implementation study using a historical comparator. The following periods will be defined: a historical reference period, lasting 12 months; a preparation and implementation period, lasting up to 30 days; and a post-intervention period, lasting 12 months, beginning on the activation date of the digital analytics layer. In each period, the analytic population will be defined based on eligibility existing on the respective index date. In the historical period, data will be used to estimate the reference proportion of the primary outcome and the historical values of applicable secondary outcomes. In the post-intervention period, individuals eligible on the index date will be followed for 12 months through care records. Individuals who become eligible after the post-intervention index date may receive the care actions provided for in the strategy and be included in operational, implementation, and exploratory analyses, but will not be included in the denominators of outcomes whose population was fixed at the index date.

The study will be conducted in the municipality of Afogados da Ingazeira, Pernambuco, including the 15 Primary Health Care Units (UBS) with an affiliated Family Health Team. The municipality has an estimated population of 40,241 inhabitants and has used epHealth digital solutions integrated into Primary Health Care routine since 2018. The municipality's participation will be voluntary and formalized through a Technical Cooperation Agreement, which will establish the parties' responsibilities, the conditions for data access and processing, and the integration of the strategy into the municipal care routine. Implementation of the research-specific technological components will not incur any additional cost to the municipality.

The ARISE strategy will consist of two components:

1. Digital analytics layer. This will be configured on top of the epHealth infrastructure already deployed in the municipality. The layer will use structured data extracted from the e-SUS PEC (Primary Care Information System) to apply predefined eligibility rules for T2DM screening. Identified individuals will be made available in operational lists and may generate alerts for UBS managers, community health workers, and community health worker technicians. The alerts will support the organization of visits, contacts, scheduling, test requests, follow-up of pending items, and patients' return to the unit. The layer may also flag: delays exceeding 60 days between test request and completion; availability of laboratory results; and the need to organize active case-finding. Alerts will be issued during business hours and will not display laboratory result values. No test, consultation, diagnosis, prescription, referral, or treatment will be carried out automatically. Clinical decision-making will remain the responsibility of primary care professionals and managers.
2. Professional training. UBS managers, community health workers, community health worker technicians, physicians, nurses, and nursing technicians will receive training on the Clinical Protocol and Therapeutic Guidelines for T2DM (PCDT-DM2) and on the operational workflows of the ARISE strategy. Training will be conducted prior to activation of the digital analytics layer. Professionals will be considered research participants exclusively for the components evaluating educational impact, acceptability, and usability of the strategy. Eligible professionals will be those aged 18 years or older, currently working in the municipality, directly involved in the implementation or operationalization of the strategy, who attend the training corresponding to their role and agree by signing the Informed Consent Form (ICF). Participation in institutional training will not be contingent on participation in the research. Refusal to authorize the use of questionnaire data will not result in any functional, hierarchical, or institutional penalty.

A census-based approach will be used for population size. The primary analysis will include all individuals meeting eligibility criteria on the post-intervention index date. The final number will depend on the population actually identified as eligible in the municipal databases. Power calculations considered scenarios ranging from 8,000 to 30,000 eligible individuals and historical proportions of confirmed T2DM diagnosis between 0.80% and 0.90%. A population of approximately 14,000 to 15,000 eligible individuals provides power close to 80% to detect a relative increase of approximately 25%, in scenarios with a historical proportion between 0.85% and 0.90%. For the components involving professionals, a census-based approach will be used across eligible categories and roles, according to the updated list of professionals in active service who are directly involved in the strategy.

For the statistical analysis, the proportion of confirmed T2DM diagnosis observed in the post-intervention period will be compared to the historical reference proportion using a two-tailed one-proportion test, considering a 5% significance level. Point estimates, 95% confidence intervals, absolute differences, and, where applicable, proportion ratios will be presented. Adjusted analyses and assessments of associated factors may use logistic regression models. Change in professionals' knowledge will be analyzed in a paired manner, using the intra-individual difference between post- and pre-training scores. RE-AIM indicators will be evaluated predominantly in a descriptive manner. The blocks assessing acceptability, reaction to training, and perception of workflow will be analyzed separately, using scores from 1 to 5. Usability will be assessed using the System Usability Scale, with scores from 0 to 100, among professionals who directly used the digital component.

The intervention will last 12 months, preceded by a preparation and training period of up to 30 days. The historical reference period will also comprise 12 months. After the end of the post-intervention period, database consolidation, statistical analyses, report preparation, and scientific dissemination of results will be carried out.

The study was approved by the Research Ethics Committee (REC), under Certificate of Presentation for Ethical Consideration (CAAE) No. 00260026.5.0000.5330 and approval opinion No. 8.705.723.

Primary outcome: The primary outcome will be confirmed diagnosis of T2DM during the observation period, treated as a binary individual indicator. The numerator will consist of eligible individuals classified as screening-positive for T2DM according to the criteria defined in the protocol. The denominator will consist of all individuals eligible for screening on the index date of the post-intervention period. The post-intervention proportion will be compared to the historical reference value, estimated by applying the same eligibility rules and outcome definition to the historical data.