Five Variables Separate Low-Risk from High-Risk Older ED Patients
Delirium moves quickly in the emergency department, and older adults often arrive before anyone recognizes the syndrome. In a prospective cohort from a tertiary hospital in Thailand, investigators built and internally validated a bedside clinical risk score designed to flag delirium within the first 12 hours after arrival in patients aged 65 years and older who stayed in the ED for more than four hours. The result was intentionally simple: five readily available variables, a total score ranging from 0 to 118, and a discrimination profile that gives clinicians a practical way to stratify risk while the patient is still under emergency care.
Among 380 enrolled patients, 102 had delirium, a prevalence of 26.8%. That figure alone is a reminder that delirium is not a rare complication at the margins of emergency medicine; it is a common bedside problem in older adults, especially when acute illness, physiologic stress, and baseline cognitive vulnerability collide. The study’s logic was straightforward and clinically grounded: if delirium can be identified early, preventive measures, closer observation, and medication review can begin before agitation, falls, restraint use, or prolonged hospitalization compound the harm.
Age, Organ Failure, and Baseline Cognition Dominated the Final Model
Five independent variables remained in the final multivariable model. Age 70–79 years carried 1 point. Age 80–89 years, stage III–V chronic kidney disease, or prior cerebrovascular accident each carried 12 points. Age 90 years or older carried 29 points, cirrhosis 24 points, and dementia 41 points. The weighting is clinically intuitive even before a score is calculated at the bedside. Advanced age increases brain vulnerability to acute metabolic and inflammatory insults, chronic kidney disease raises the risk of medication accumulation and metabolic disturbance, cirrhosis can alter neurotransmission through hepatic encephalopathy pathways, and dementia lowers the threshold for an acute confusional state.
The heavy contribution assigned to dementia is especially telling. Baseline cognitive impairment does not merely coexist with delirium; it often creates the substrate on which delirium emerges when infection, dehydration, pain, hypoxia, unfamiliar surroundings, or polypharmacy tip the balance. A patient with dementia may have less reserve to compensate for sleep disruption, sensory overload, or even relatively small physiologic shifts that would not provoke delirium in a cognitively intact adult.
Discrimination Was Strong Enough to Separate Bedside Risk Groups
The score performed well in internal validation, with an area under the receiver operating characteristic curve of 0.79, and the 95% confidence interval ranged from 0.74 to 0.84. In practical terms, that level of discrimination suggests the tool can distinguish between patients who are more likely and less likely to develop delirium, even if it is not a substitute for serial clinical assessment. The score’s strongest use is likely to be triage support rather than diagnosis: it identifies the older ED patient who needs more attentive monitoring, more deliberate medication review, and a lower threshold for reorientation strategies and delirium precautions.
The risk separation was clinically meaningful. Delirium occurred in 17.97% of patients in the low-risk group, defined as scores of 0 to 24, and in 57.65% of patients in the high-risk group, defined as scores of 25 or higher. Those proportions make the score feel less like an abstract academic exercise and more like a usable emergency tool. A patient crossing into the high-risk range is not guaranteed to become delirious, but the probability rises sharply enough to justify action.
A score that can move delirium risk from roughly one in five to well over one in two has real operational consequences in the ED.
Why the Score Fits Emergency Care
The appeal of the model lies in its restraint. It relies on variables already known or rapidly obtainable during routine emergency assessment: age, prior cerebrovascular disease, cirrhosis, dementia, and chronic kidney disease stage III to V. No specialized neurocognitive battery, biomarker panel, or advanced imaging was required. That matters in a setting where time is compressed, staffing is stretched, and patients may be boarding for hours before disposition is finalized.
From a pathophysiologic standpoint, the selected variables all point toward reduced cerebral resilience. Chronic vascular brain injury after stroke, uremic burden in chronic kidney disease, neurocognitive frailty in dementia, and metabolic instability in cirrhosis all lower the threshold for acute brain dysfunction. In older adults, delirium often emerges when that already-fragile equilibrium is disturbed by pain, dehydration, infection, sedatives, anticholinergic drugs, or sleep deprivation. A risk score built around these conditions therefore aligns with the biology of delirium rather than merely with administrative convenience.
What the Score Can and Cannot Do Yet
This was a prospective cohort study, not a randomized intervention trial, so the score predicts risk rather than proving that score-guided care improves outcomes. Internal validation is useful, but it does not settle how the model will behave in different EDs, among different patient populations, or in systems with different workflows, language patterns, and baseline rates of cognitive impairment. External validation remains necessary before broad adoption.
The study also used the Thai version of the Confusion Assessment Method intensive care unit within 12 hours of arrival, which gives the project a concrete diagnostic framework but also anchors it to the local clinical environment. Even so, the core signal is likely to travel well: older adults with dementia, very advanced age, cirrhosis, prior stroke, and significant chronic kidney disease are the patients most likely to need early delirium surveillance.
Retrospective trial registration, completed after enrollment, is another detail that should be read carefully. It does not erase the findings, but it does shape how confidently the score can be viewed as a preplanned development effort rather than a fully prospectively registered prediction project. The authors have provided a useful starting point. The next step is to test whether the same five-variable score performs as well outside a single tertiary hospital and whether early ED-driven prevention changes delirium-related outcomes that matter at the bedside.