Diagnostic Tool to Detect Delirium Proves Superior to Commonly-Used Test

IN THIS ISSUE: ACLP President | Emergency Medicine | Delirium Webinar | Dlin/Fischer | What’s New on the Web | CLP2024 | A&E Abstracts

Diagnostic Tool to Detect Delirium Proves Superior to Commonly-Used Test

Research awarded the Academy’s Dlin/Fischer prize

The Dlin/Fischer Clinical Research Award—awarded for significant achievement in clinical research and the highest-scoring oral paper submitted for presentation at the Annual Meeting—has been won by José Maldonado, MD, FACLP, FACFE.

Jose Maldonado, MD, FACLP, FACFE, Professor & Chief, Division of Medical Psychiatry, Medical Director of Critical Care Psychiatry, Stanford University, Menlo Park, California
Jose Maldonado, MD, FACLP, FACFE

Dr. Maldonado, et al., Critical Care Psychiatry Service, Division of Medical Psychiatry, Stanford University School of Medicine, authored Assessing the Diagnostic Accuracy of the Stanford Proxy Test for Delirium (S-PTD) in a Critically Ill Population.

Their goal was to assess the accuracy of their proxy test against the most widely-used delirium detection tool in intensive care units, the Confusion Assessment Method for the ICU, then to compare the performance of both tools against a DSM-5 based neuropsychiatric assessment.

The Stanford tool demonstrated superior sensitivity regardless of phenotype—crucial in detecting delirium.

Delirium is the most common psychiatric disorder in critically ill patients with a reported incidence of 21.8% to 64% in intensive care units. In these settings, delirium is associated with an increased risk of prolonged mechanical ventilation, unplanned catheter removal, self-extubation, and the need for physical and/or pharmacological restraints.

“Despite its elevated prevalence and the availability of screening tools, delirium remains severely underdiagnosed across hospital environments, going unrecognized in 55% to 84% of patients,” say the authors. “This is particularly the case among those experiencing the most common type of delirium, the hypoactive phenotype.

“While many delirium screening tools report high sensitivities in controlled research environments, these seem to decrease under real-life conditions. Furthermore, the tools most commonly used currently are based on older versions of the Diagnostic and Statistical Manual of Mental Disorders (DSM).”

In their study of 200 patients aged 24-92, over a one-year period (July 2023-July 2024), just over 25% yielded a positive diagnosis of delirium as per the DSM-5-based assessment. The breakdown of identified delirium phenotypes was: hypoactive 60.12%; mixed 30.37%; and hyperactive 9.49%.

The CAM-ICU displayed a sensitivity of 35.29%, specificity of 99.55%, PPV of 94.74%, NPV of 87.13%, failing to detect 73/94 (77.7%) of hypoactive, 36/48 (75%) of mixed, and 12/15 (80%) of hyperactive delirium cases, respectively.

The S-PTD demonstrated a sensitivity of 89.73%, specificity of 82.05%, PPV of 62.38%, NPV of 96.01%, failing to detect 12/88 (13.6%) of hypoactive, 2/44 (4.5%) of mixed, and 1/14 (7.1%) of hyperactive delirium cases, respectively.

In addition to classic delirium phenotypes, the DSM-5 based assessment identified 32 cases of subsyndromal delirium. Against this reality, the S-PTD exams appropriately detected 73.33%; while the CAM-ICU failed to detect 100% of them.

At the authors’ institution, the S-PTD demonstrated diagnostic superiority over the CAM-ICU, including the ability to recognize cases of subsyndromal delirium.

“Its validation against DSM-5 criteria and ability to accommodate patients unable to participate in conventional assessments make it a preferable alternative for delirium screening,” say the authors. “The S-PTD’s ability to identify a broader spectrum of delirium subtypes suggests its potential in improving patient outcomes through timely and accurate diagnosis.”

 

 

 

« « All 2024 ACLP News« December 2024 ACLP News