Reliability and Validity of the Spanish Adaptation of the Stanford Proxy Test for Delirium in Two Clinical Spanish-Speaking Communities

Abstract

Background: Delirium is the most prevalent neuropsychiatric syndrome experienced by patients admitted to inpatient clinical units, occurring in at least 20% of medically hospitalized patients and up to 85% of those admitted to critical care units. Although current guidelines recommend the implementation of universal prevention strategies, the use of management strategies largely depends on constant surveillance and screening. This allows for the timely diagnosis and correction of its underlying causes and implementation of management strategies.

Objective: It was to adapt and analyze the Spanish adaptation of the Stanford Proxy Test for Delirium (S-PTDsv) instrument for its use among Spanish-speaking populations. The S-PTD is an instrument consisting of 13 observational items to be completed by a clinician observer, usually the patient's nurse. The completion of the questionnaire takes about 1 minute and does not require the active participation of the person evaluated, which has important clinical advantages compared to other available instruments (e.g., the Confusion Assessment Method).

Methods: The psychometric properties of the S-PTDsv were evaluated in a population of 123 patients using a quantitative, cross-sectional design. All subjects were over 18 years of age and hospitalized in various inpatient medico-surgical and intensive care unit services, either at the Barcelona Clinical Hospital (Barcelona, Spain) or the UC-Christus Health Network Clinical Hospital (Santiago, Chile, S.A.). The ultimate diagnosis of delirium was made by a member of the Psychiatry Consult Service by means of an independent neuropsychiatric evaluation based on the Fifth Edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) criteria, published in 2013, which is the latest version of the diagnostic manual. All study tests were performed by study personnel who were blinded to each other's test results within an hour of each other.

Results: In the receiver operator characteristic (ROC) curve analysis, the S-PTDsv demonstrated excellent classification qualities when compared with the DSM-5 as the classification reference standard. Using a cutoff point of ≥3, the S-PTDsv had a sensitivity of 94% and a specificity of 97%. The area under the curve indicator was equal to 0.95, suggesting the S-PTDsv has an excellent overall performance in accurately identifying cases of delirium. Accordingly, the S-PTDsv's positive predictive value = 0.93, and the negative predictive value = 0.97. The internal reliability measured with Cronbach's alpha was 0.96. Confirmatory factor analysis revealed a 1-dimensional structure with high loadings (>0.72), demonstrating that all items similarly contribute to the total diagnostic dimension, suggesting adequate construct validity. This provided evidence of convergent validity.

Conclusions: The performance of the S-PTDsv, as compared to a blinded neuropsychiatric assessment based on DSM-5, indicates that it is an effective instrument for the detection of delirium, in the Spanish-speaking populations. These results are comparable and consistent with previously published studies in the English language version.

Keywords: Stanford proxy test for delirium; confusion assessment method; delirium; delirium and COVID; factor analysis.