Julio A. Esquivel-Tamayo, Consultorio Médico 16, Policlínico Manuel Fajardo Rivero, Las Tunas; Facultad de Ciencias Médicas Dr. Zoilo Enrique Marinello Vidaurreta, Universidad de Ciencias Médicas de Las Tunas, Las Tunas; Cuba
Arquímedes Montoya-Pedrón, Servicio de Neurofisiología Clínica, Hospital General Dr. Juan Bruno Zayas Alfonso, Santiago de Cuba; Facultad de Ciencias Médicas No. 1, Universidad de Ciencias Médicas de Santiago de Cuba, Santiago de Cuba. Cuba
Background: Predicting mild neurocognitive disorder (NCD) due to Alzheimer’s disease (AD) based on risk factors is the main option for introducing effective therapies. A predictive scale was developed to this end, but it requires validation to assess its effectiveness. Objective: To internally validate a predictive scale for mild NCD due to AD. Methods: A prospective analytical case-control study was conducted from January 1 to May 31, 2025. The sample included 49 subjects classified as having mild NCD due to AD according to the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) criteria and 98 controls with normal cognitive functions, using simple random sampling. The scale’s performance and construct validity were assessed using sensitivity and specificity, positive predictive value (PPV), negative predictive value (NPV), receiver operating curve, and area under the curve (AUC), and reliability was evaluated using the Kuder-Richardson test. Results: The scale had a PPV of 84.61%, a NPV of 94.37%, a sensitivity of 89.79%, a specificity of 91.83%, and an AUC of 0.908 (95% confidence dence interval: 0.850-0.966), which were not significantly different than those obtained in the design study and confirmed an accurate prediction. The Kuder-Richardson test checked the scale’s reliability with a coefficient of 0.719. Conclusions: Internal validation of the predictive scale demonstrated adequate performance.
Keywords: Cognitive dysfunction. Alzheimer disease. Validation study. Risk factors.