Validation of a simple screening tool for early diagnosis of advanced Parkinson's disease in daily practice: the CDEPA questionnaire

Early clinical diagnosis of advanced Parkinson's disease (APD) may be difficult. This study aimed to validate a simple screening tool, the CDEPA questionnaire ("Cuestionario De Enfermedad de Parkinson Avanzada" [Questionnaire for Advanced Parkinson's Disease]), for the identification of APD in daily practice. The study included 173 consecutively selected patients with PD (40% were women, mean age was 68.4 ± 10.5 years), stratified according to the Hoehn and Yahr (HY) scale. The CDEPA questionnaire defined APD as the presence of severe disability requiring help for activities of daily living (ADL), motor fluctuations with limitation or inability to perform ADL, severe dysphagia, recurrent falls, or dementia. The diagnostic performance of the questionnaire was assessed against the gold standard criterion based on clinical judgment. PD was categorized as advanced in 65 (38%) patients when using the gold standard and in 109 (63%) patients when the CDEPA questionnaire was used. The CDEPA questionnaire and the gold standard agreed moderately (kappa statistic of 0.48, P < 0.001). The CDEPA classified APD with a sensitivity of 97%; specificity of 57%; total accuracy of 72.3%; and area under the curve (for a binary classifier) of 77.2%. Significant differences were found between the groups created by the CDEPA in several usual PD evaluations (HY Scale, SCOPA Motor Scale, Non-motor Symptoms Scale for PD, Clinical Impression of Severity Index for PD, Clinical Global Impression-Severity Scale, and Patient Global Impression-Severity Scale). CDEPA showed satisfactory inter-rater agreement (kappa = 0.88) and test-retest concordance (kappa 0.83). In conclusion, the CDEPA questionnaire is a valid, reliable, and useful instrument for easily screening APD.The authors thank the Spanish Society of Neurology for promoting the study, Grupo SANED for logistic support, David Calbet, statistician, for analysis of data, and Marta Pulido, MD, for editing the manuscript and providing editorial assistance. John Wetmore, B.A., reviewed the English language of the manuscript.S