This post was contributed by Vanja Douglas, MD, and Stephanie Rogers, MD, both of the University of California San Francisco School of Medicine.
Delirium is increasingly being recognized as a potential area of hospital quality measurement. In several ways, delirium represents an ideal quality metric because standardization of care in this area is likely to lead to improvement in patient outcomes: delirium is a common, costly and harmful hospital-acquired condition that is preventable in at least a third of cases, with a practice gap in the application of prevention measures. How to measure quality of care for delirium, however, is a matter of debate. Some quality indicators specify outcome rates, such as mortality in acute myocardial infarction (MI), while others measure compliance with processes of care, such as primary percutaneous intervention received within 90 minutes of hospital arrival in acute MI. Should delirium quality metrics focus on outcomes or processes?
Recently, the American Academy of Neurology (AAN), the Neurohospitalist Society (NHS), and the Neurocritical Care Society (NCCS) jointly issued a set of quality metrics for inpatient neurology, two of which focus on delirium.(1) One calls for implementation of a delirium risk factor screening and prevention protocol, with measurement of the percentage of patients at high risk of developing delirium who had a preventative protocol instituted. The other emphasizes non-pharmacologic treatment of delirium, with measurement of the percent of patients with delirium that was not present on admission who were treated initially nonpharmacologically. These quality metrics are based in part on the delirium prevention, diagnosis, and management guidelines issued by the National Institute for Health and Care Excellence in 2010, which call for delirium risk factor assessment upon hospital admission for all patients, daily screening for indicators of delirium, specialist assessment for diagnosis, and nonpharmacologic prevention and initial treatment.(2)
These new proposed quality metrics are a tremendous step forward for the care of delirious patients and those at risk, but they raise questions about how delirium risk is assessed and how delirium is screened for, diagnosed, and reported. What is the best method to assess delirium risk upon admission? How should delirium be diagnosed and can we be confident in comparing incidence rates between hospitals? Several recent publications provide insight into our ability – or lack thereof – to currently answer these questions.
A number of delirium risk prediction scores have been published; these almost invariably include some combination of risk factors including older age, cognitive impairment, severity of illness, functional impairments, and in the elective surgery population, surgery-specific risk. Despite this, risk prediction scores can be difficult to implement into routine clinical practice due to the practical challenges associated with performing the necessary assessments on all admitted patients, and some argue that delirium prevention measures could more simply be applied to all patients above a certain age, such as 65 years. However, the downside of using age as the single determinant of delirium risk is that younger patients at high risk would be overlooked (such as those with chronic brain dysfunction from conditions like traumatic brain injury); highly functioning older patients with mild illness may unnecessarily be subjected to costly and time-intensive preventative measures; and more patients are likely to be identified as high risk leading to overuse of delirium prevention resources.
One approach to simplify implementation of delirium risk prediction tools is to use one that requires very little time and training. We have had success implementing such a tool at our institution and recently validated its performance when applied and charted at the time of admission by bedside nurses.(3) Another approach would be to develop prediction tools that exclusively use data already available in the electronic medical record; this is an area ripe for study.
As with delirium risk assessment, there are a number of different tools available to screen for and diagnose delirium. Last year, Hendry and colleagues published a comparison of a number of these assessment tools. The test characteristics of these tools were quite variable. For example, the abbreviated mental test-4, abbreviated mental test-10, months of the year backward, 4-AT and single question in delirium all had sensitivities ranging from 87% to 93% and specificities from 50 to 70%.(4) The brief confusion assessment method was more specific at 91%, but had a sensitivity of 70%. We examined and recently published the performance of the nursing delirium screening scale, which was only 42% sensitive but 98% specific at the standard cut-point of 2 or greater, but 67% sensitive and 93% specific at a cut-point of 1 or greater.(5) This is similar to what Neufeld and colleagues found when applying the nursing delirium screening scale in the post-anesthesia care unit.(6)
What seems clear from this literature is that the incidence of delirium at a given hospital is likely to depend on the screening tool used. Similarly, the number of patients at risk for delirium will vary depending on the approach to risk assessment. Because of this variability, the quality metrics designed by the AAN, the NHS, and NCCS sensibly focus on the process of delirium care rather than the outcome of delirium itself, namely the percentage of high risk and delirious patients receiving prevention and non-pharmacologic treatment, respectively. Much more work is needed to harmonize our approach to delirium screening and diagnosis before we can accurately compare delirium rates among hospitals. Until then, quality measurement in delirium care must focus on processes of care rather than outcome.
- Josephson SA, Ferro J, Cohen A, Webb A, Lee E, Vespa PM. Quality improvement in neurology: Inpatient and emergency care quality measure set: Executive summary. Neurology. 2017.
- O'Mahony R, Murthy L, Akunne A, Young J. Synopsis of the National Institute for Health and Clinical Excellence Guideline for Prevention of Delirium. Ann Intern Med. 2011;154(11):746-51.
- Brown EG, Josephson SA, Anderson N, Reid M, Lee M, Douglas VC. Predicting inpatient delirium: The AWOL delirium risk-stratification score in clinical practice. Geriatr Nurs. 2017.
- Hendry K, Quinn TJ, Evans J, Scortichini V, Miller H, Burns J, et al. Evaluation of delirium screening tools in geriatric medical inpatients: a diagnostic test accuracy study. Age Ageing. 2016;45(6):832-7.
- Hargrave A, Bastiaens J, Bourgeois JA, Neuhaus J, Josephson SA, Chinn J, et al. Validation of a Nurse-Based Delirium-Screening Tool for Hospitalized Patients. Psychosomatics. 2017.
- Neufeld KJ, Leoutsakos JS, Sieber FE, Joshi D, Wanamaker BL, Rios-Robles J, et al. Evaluation of two delirium screening tools for detecting post-operative delirium in the elderly. Br J Anaesth. 2013;111(4):612-8.