Patient Safety Tip of the Week


March 31, 2009     Screening Patients for Risk of Delirium



In October 2008 we did back-to-back columns “Managing Delirium” and “Preventing Delirium”. We talked about high cost of delirium in both human terms and financial terms. The first column dealt with ways to identify delirium and manage it once it has occurred. The second dealt with what is probably the more important issue: identifying patients who are at risk for delirium in hopes of intervening to minimize the likelihood of actually developing delirium. We strongly advocated that any patient over the age of 65 have a delirium risk assessment prior to admission by a primary care physician or geriatrician who routinely does such screening or as part of a preoperative program staffed by a nurse or physician extender. We recommended doing the MMSE (or shorter versions) as the minimum assessment until future studies validate more complex instruments as well as looking for the other delirium risk factors noted in our October 21 column.


In the second of those two columns we mentioned some new work coming out of Duke on screening patients to predict the risk of postoperative delirium. That work has now been published (Greene et al 2009) along with a companion article (Smith et al 2009) and it provides some very encouraging insight into screening for delirium risk.


Greene et al screened 100 elderly patients prior to major noncardiac surgery, using a battery of tests of cognition and depression. A total of 16% of those patients subsequently developed postoperative delirium. They found 2 key predictors of postoperative delirium: impaired executive function (as measured by the Trail Making B Test) and depression (as measured by the Geriatric Depression Scale – Short Form or GDS-SF). Several other measures that had been noted to be predictive of delirium in other studies did not independently predict delirium in the Duke studies. Those included age, Charlson comorbidity index, history of depression, education level, ASA score, level of preoperative pain, alcohol use, BMI, or a measure of global cognitive function). However, the relatively small sample size may have been of insufficient power to demonstrate predictive power for several of these factors.


Specifically, of patients who scored above the cutoff on the Trails B test 44% developed postoperative delirium and of those who scored above the cutoff on the GDS-SF 35% developed postoperative delirium. Of the patients who scored above the cutoff on both tests, 83% developed postoperative delirium.


The two tests are easy to administer and are not very time-consuming. The GDS-SF is a 15-item yes/no test that has been validated in surgical populations. It takes about 5 minutes to administer. The Trail Making B test consists of 25 circles distributed over a sheet of paper, with numbers (1 – 13) and letters (A – L). The patient draws lines to connect the circles in an ascending pattern, alternating between the numbers and letters (i.e., 1-A-2-B-3-C, etc.). The patient is timed as he/she connects the circles as quickly as possible, without lifting the pen or pencil from the paper. It also takes less than 5 minutes to administer.


Interestingly, unlike several prior studies, the MMSE (mini-mental status exam) in their studies did not independently predict the occurrence of postoperative delirium. That may reflect that their population overall was functioning at a relatively higher cognitive level than the populations in the prior studies.


However, they also raise the interesting concept of “cognitive reserve”. By that they mean that many of their patients were functioning without obvious gross impairment, but their cognitive capacity was not sufficient in the presence of a new CNS insult. They speculate that the tests of executive function reflect higher order complex cognition and thus may be better markers than tests of more global cognitive function.


The companion article (Smith et al 2009) also performed a battery of tests prior to surgery on almost 1000 patients, though the population was quite different and the incidence of postoperative delirium was only 3.5%. They also found that impairment of executive function and presence of depressive symptoms were predictors of postoperative delirium. The greatest value of this paper is that it showed that tests of more complexity of executive function were more predictive of delirium. Specifically, the modified Stroop Color Word Interference Test was most predictive.


The science and our knowledge of the underlying pathophysiology of delirium are obviously still evolving. Though much more research is needed to improve our ability to prevent delirium, it would currently make sense for organizations to incorporate delirium screening into your preoperative protocols at least for the elderly. Based on the Greene study, use of the GDS-SF and the Trails B test rather than the MMSE or tests of more global cognitive function may make practical sense. These tests are short, easy to administer, and have good predictive value.





Greene, Nathaniel H. B.S. *; Attix, Deborah K. Ph.D. +; Weldon, B Craig M.D. ++; Smith, Patrick J. M.A. [S]; McDonagh, David L. M.D. [//]; Monk, Terri G. M.D., M.S. # Measures of Executive Function and Depression Identify Patients at Risk for Postoperative Delirium. Anesthesiology. 110(4):788-795, April 2009



Smith, Patrick J. M.A. *; Attix, Deborah K. Ph.D. +; Weldon, B Craig M.D. ++; Greene, Nathaniel H. B.S. [S]; Monk, Terri G. M.D., M.S. [//] Executive Function and Depression as Independent Risk Factors for Postoperative Delirium. Anesthesiology. 110(4):781-787, April 2009



Positive Aging Resource Center. Geriatric Depression Scale – Short Form.



University of Iowa. Trail Making Test Parts A and B.








Patient Safety Tip of the Week Archive


What’s New in the Patient Safety World Archive