Delirium, an acute disturbance in mental ability, is a common, costly and often overlooked complication of hospitalization in older adults. The incidence of delirium in the acute care setting varies, ranging from 11 to 50% outside of the intensive care unit (ICU), and up to 83% in the ICU.1 When compared to patients who do not develop delirium, those with delirium experience longer lengths of hospitalization, more frequent readmissions, and incur greater costs of care.1 Despite its common, and costly occurrence, researchers report that delirium often goes unrecognized or misdiagnosed.2, 3, 4 Researchers have historically and consistently supported the importance of early identification of those at risk, combined with prompt initiation of prevention strategies, as an effective way to reduce incidence and ameliorate severity of delirium.5, 6, 7

There are a number of delirium prediction models reported in the literature. Many have been validated in a specific patient population (i.e. hip fracture, stroke patients, and patients in intensive care) and most include a comprehensive assessment for mental status as essential to identify baseline cognitive status. One of the challenges of initiating delirium prediction models in the acute care setting, is that they can be cumbersome to use, are time consuming to complete and require the clinician be educated on their use to ensure completion and accuracy. Lindroth et al.8 conducted a systematic review in which the authors suggested prediction models “should consider inclusion of variables and assessments that are readily available in clinical practice and are feasible to administer without extensive training”. Results of a recent study examining delirum prevalence rates across services that manage delirium found 84.2% of the cases of delirium were undiagnosed. The authors call for a novel approach to delirium screening.9

The Delirium Elderly at Risk (DEAR) screening tool was designed to predict the risk of delirium in older surgical patients.10, 11, 12 The tool consists of five key predisposing risk factors for delirium in adults age 65 and older: (1). Cognitive impairment as defined by a Mini-Mental Status Examination (MMSE) score of <23; (2). Vision or hearing impairment (regardless of correction); (3). Dependence in at least one activity of daily living; (4). Substance use, defined as chronic alcohol or benzodiazepine use and (5). Age 80 or older. Scores range from 0 to 5, with 1 point for the presence of each risk factor on admission. A cutoff score of 2 is predictive of delirium in pre-surgical and post-surgical patients.11,12 The tool has been validated in surgical patients aged 65 and older undergoing either elective hip arthroplasty or emergent hip fracture repair.10,11

The original studies of the DEAR used the Mini Mental State Examination (MMSE) to determine cognitive impairment. The MMSE is a 30 item assessment that is often used as a measure of cognitive impairment in clinical practice and research.13 However, the MMSE requires time and training to administer. Currently use of the MMSE, which had been free for decades, is now proprietary and requires a fee to use adding to its impracticality in the acute care setting).14 It is unknown how well the DEAR would perform when used in a combined medical and surgical population since prior studies were performed only with surgical patients. It is also unknown if the DEAR, when modified using nursing's admission assessment of baseline cognition rather than the MMSE, would be a valid predictor of delirium. The purpose of this study was to test the psychometric properties of a modified version of the DEAR (mDEAR), with retrospective data from the nursing admission documentation for hospitalized medical and surgical patients aged 65 and older.

The overall aim of this study was to examine the predictive validity of the mDEAR using routinely collected medical record data at admission, for a population of hospitalized medical and surgical patients aged 65 and older. The specific aims of the study were:

Determine if a delirium risk score generated by the mDEAR on admission is associated with the occurrence of delirium during hospitalization in an adult population of medical and surgical patients aged 65 and older, when all factors are weighted equally (as described in Freter12).

Determine an optimum risk prediction model for delirium using the mDEAR in a combined population of surgical and non-surgical patients.

Test the validity of the final model in the combined medical and surgical inpatient population