Postoperative delirium is an acute change in cognitive status characterized by fluctuating consciousness and inattention within 30 days after an operation (1). The reported incidence of postoperative delirium in hip-fracture patients ranges from 28% to 41% (2,3). Delirium is associated with increased morbidity and mortality among hospitalized older patients (4,5). It is important, therefore, to identify predictors of postoperative delirium. If patients who are predisposed to postoperative delirium can be identified, then interventions to prevent this complication can be implemented.
Delirium can occur either on emergence from anesthesia or in the recovery room. Emergence delirium is an acute change in cognitive status characterized by fluctuating consciousness and inattention in the immediate recovery phase from general anesthesia. The incidence of emergence delirium in adults is unclear; reports range from 3% to 20% (6). The incidence of recovery room delirium, whether related to emergence delirium or as a separate entity, is unclear. Furthermore, the prognostic implications of both emergence delirium and recovery room delirium, especially with respect to continuing postoperative delirium, have not been studied. The aim of this study was to determine the incidence of delirium in the recovery room in elderly patients having hip-fracture repair under general anesthesia and to discover whether recovery room delirium is associated with continuing postoperative delirium.
The study design was a prospective blinded case series. The IRB of the Johns Hopkins Bayview Medical Center approved this study and written informed consent was obtained from all study participants. The study population consisted of patients undergoing hip-fracture repair at Johns Hopkins Bayview Medical Center. The intervention studied was a standardized protocol for general anesthesia. The outcome measurements were recovery room delirium and continuing postoperative delirium. Exclusion criteria were severe pulmonary disease, severe cardiac disease, severe dementia characterized by Mini-Mental State Examination (MMSE) of <10/30 (7), and patients requesting regional anesthesia.
Before their operation, all patients underwent standard preoperative evaluation. In addition, we performed preoperative interviews and used the MMSE (7) and Confusion Assessment Method (CAM) scores (8) to assess dementia and delirium, respectively. During hip-fracture repair, all patients were anesthetized using a standardized protocol. Standardized monitors included a five-lead electrocardiogram, oscillometric arterial blood pressure, pulse oximetry, inspired and expired gases, and temperature. Invasive monitors were used as required by the patient’s condition. We administered no premedication. Propofol (0.5–1 mg/kg) was the induction drug, with succinylcholine (1 mg/kg) for endotracheal intubation. Anesthesia was maintained with a mixture of isoflurane (1%–1.5% inspired) and oxygen until completion of the operation. Patients were allowed to breathe spontaneously throughout the procedure. Intraoperatively, we titrated morphine according to respiratory rate, with a target of 14–16 breaths/min. We recorded the total intraoperative doses of propofol (mg/kg), succinylcholine (mg/kg), isoflurane (MAC-h), and morphine (mg/kg), as well as the time from the discontinuation of isoflurane until endotracheal extubation.
The presence of preoperative dementia was defined by the MMSE score <24/30 (9). Moderate dementia was defined by an MMSE score of 10/30 to 21/30 (7). The presence of recovery room delirium and postoperative delirium was defined by CAM score. The CAM score uses a diagnostic algorithm consisting of 4 features: 1) acute and fluctuating change in mental status, 2) inattention, 3) disorganized or incoherent thinking, and 4) altered level of consciousness. A CAM score is considered positive if the patient displays the combination of features: (1, 2, and 3) or (1, 2, and 4) or (1, 2, 3, and 4). For determining the presence of recovery room delirium, feature 1 was determined to be present primarily by report from the primary postanesthesia care unit (PACU) nurse or family member. Features 2, 3, and 4 were determined to be present by the attending anesthesiologist using a standardized battery of questions (8). For recovery room delirium, an attending anesthesiologist examined the patient at 60 min after discontinuation of isoflurane. We defined recovery room delirium as a positive CAM score at this time. We chose a 60-min time interval to obtain a CAM score because previous studies of the elimination pharmacokinetics of isoflurane show that the ratio FA/FA0 (FA, end-tidal concentration; FA0, the last FA during anesthetic administration) reproducibly decreases to <0.1 by this time interval (10). Postoperative pain management in the PACU was managed via a standardized algorithm, with 1-mg IV morphine boluses (11). We recorded the total PACU dose of morphine (mg/kg).
During the postoperative period, we interviewed each patient daily until hospital discharge to evaluate for postoperative delirium by use of the CAM score (8) as described above. For determining the presence of postoperative delirium, feature 1 was determined primarily from report of the floor nurse taking care of the patient or a family member. Features 2, 3, and 4 were determined by an attending anesthesiologist after asking a standardized battery of questions (8). The attending anesthesiologist performing the postoperative CAM assessment was blinded to the presence or absence of recovery room delirium. Postoperative pain management on the hospital ward, including the intensive care unit (ICU), was standardized. Upon discharge from the PACU, all patients received morphine through patient-controlled analgesia (PCA). The IV morphine PCA dose was standardized to 0.5–1 mg, with a lockout interval of 6 min and no basal rate. All patients received acetaminophen and/or oxycodone for breakthrough pain when they were able to take oral medicines. We recorded the total dose administered on the hospital ward of narcotics (morphine equivalents in mg·kg−1·d−1) and acetaminophen (mg·kg−1·d−1). We also obtained pain scores at rest each postoperative day using an 11-point Likert scale, 0 (no pain) through 10 (worse pain) (12).
To examine the risk factor of recovery room delirium and its predictive value for postoperative delirium, we based our sample size calculations on an assumed relative risk. In our practice, when performing observational studies, we do not accept a relative risk of <3 as significant. We choose to take a rigorous view of relative risk values. Therefore, in this observational study, we assumed a strong relationship between recovery room delirium and postoperative delirium would exist and chose a relative risk of 4 on which to base our sample size calculations. We defined postoperative delirium in this study as a positive CAM score at any evaluation point (24, 48, 72 h, etc.). We further assumed that the expected frequency of postoperative delirium in patients without recovery room delirium would be ≤10%. Our initial sample size calculation for α = 0.05 and β = 0.10 was 76 patients, with equal groups (recovery room delirium and no recovery room delirium).
The IRB at our hospital requires that all human subject research have a data safety plan for intermittent analysis and early stopping rules. Because our initial sample size calculation was based on an assumption of relative risk, our prospective plan was to perform an interim data analysis at 40 patients at which time we would reassess our initial sample size calculation. Our early stopping rules included reaching statistical significance.
Sensitivity was calculated as “true positives” divided by (“true positives” + “false negatives”) multiplied by 100 to convert it into percentage. Specificity was calculated as “true negatives” divided by (“true negatives” + “false positives”) multiplied by 100 to convert it into percentage. The relative risk was calculated directly from the χ2 table and was defined as the probability of an event in the active group divided by the probability of the event in the control group.
The statistical software used was SPSS 12.0.1 for Windows. Difference between groups was tested by Fisher’s exact test and Student’s t-test. We considered a value of P < 0.05 to be significant.
In 1 yr, 150 patients were admitted to the Johns Hopkins Bayview Medical Center for primary surgical repair of hip fracture. Of these 150 patients, 50 patients consented to the study, and 47 were included in the analysis (1 operation was cancelled after anesthetic induction, and there were 2 cases of nonadherence to protocol). An interim data analysis performed at 40 patients demonstrated that postoperative delirium only occurred in patients who had recovery room delirium. If recovery room delirium was absent, postoperative delirium did not occur. In fact, recovery room delirium predicted postoperative delirium (P < 0.001, Fisher’s exact test) with an infinite relative risk. We revised our initial sample size calculation using these new data. In our new sample size calculation, the expected frequency of postoperative delirium in patients without recovery room delirium was 0% (0.01% was used in calculations). The ratio of patients with recovery room delirium to those without recovery room delirium was 16:24. The percentage of patients with recovery room delirium who had postoperative delirium was 75%. Our revised sample size calculation for α = 0.05 and β = 0.10 required 15 patients total (9 without recovery room delirium and 6 with recovery room delirium). Another 10 patients were subsequently recruited to further substantiate the study’s initial conclusion. Then the study was stopped, as per our early stopping rules.
Table 1 shows the preoperative patient demographics. The majority of patients were classified as ASA physical status III. The most common underlying comorbidity was hypertension. The average age of patients included in the study was 77 ± 1 (mean ± se) yr, with a range of 56–98 yr. Seventy-seven percent of the study population was ASA class III or more. Of the preoperative variables noted in Table 1, only dementia and ASA class III or more were associated with postoperative delirium. We found a 76% incidence of dementia in patients with postoperative delirium, versus a 37% incidence in those without (P = 0.015, Fisher’s exact test). We also found a 94% incidence of ASA class ≥III in patients with postoperative delirium, versus a 67% incidence in those without (P = 0.039, Fisher’s exact test).
Table 2 shows the number of patients with and without delirium in the recovery room and postoperatively. The prevalence of recovery room delirium was 45%. The prevalence of postoperative delirium was 36%. Recovery room delirium predicted postoperative delirium (P < 0.001, Fisher’s exact test) with a sensitivity of 100% and a specificity of 85%. Postoperative delirium lasted 3.7 ± 0.6 days (mean ± se), with a range of 1–10 days.
Doses of drugs administered intraoperatively were similar in patients with and without postoperative delirium (Table 3). Similarly, analgesic drug doses were comparable in the PACU and on the hospital ward. All patients received isoflurane, propofol, and morphine intraoperatively. Not all patients, however, received analgesics in the PACU and on the hospital ward (Table 3). Daily pain scales on the hospital ward were similar in patients with and without postoperative delirium (2 ± 1 versus 4 ± 1 on postoperative day 1, 2 ± 1 versus 3 ± 1 on postoperative day 2, and 1 ± 1 versus 2 ± 1 on postoperative day 3 in patients with and without postoperative delirium, respectively; t-test). The length of the surgical procedure was similar in patients with and without postoperative delirium (155 ± 14 and 135 ± 10 min, respectively). Postoperative delirium was associated with postoperative ICU admission (41% versus 7% admission rate to ICU in patients with and without postoperative delirium, respectively; P = 0.007, Fisher’s exact test). However, ICU admission was based on underlying medical condition and not recovery room delirium or postoperative delirium. Acute length of stay, however, did not differ between patients with or without postoperative delirium (6.6 ± 0.9 and 5.5 ± 0.5 days in patients with or without recovery room delirium, respectively).
Table 4 shows the number of patients that were CAM positive (delirious) and CAM negative (nondelirious) preoperatively, in the recovery room, and on each postoperative day until time of discharge. Table 4 suggests that there is a gradual decrease in the number of confused patients during their stay in the hospital. Most of the confused patients are clustered around 1–3 days postoperatively. Delirium resolved in most patients by the fourth postoperative day.
In our patient population, 23 patients had an MMSE score of ≥24/30 and showed no evidence of preoperative dementia. Five patients had mild dementia as detected by an MMSE score of 22/30 or 23/30. Nineteen patients had moderate dementia as evidenced by the MMSE scores of 12/30 to 21/30. Four patients in the moderately demented group had preoperative delirium. Two of these patients with preoperative delirium had MMSE scores of 19/30 and the other 2 patients had MMSE scores of 12/30.
This study shows that, in patients undergoing hip-fracture repair, recovery room delirium is a strong predictor of postoperative delirium when using a standardized protocol for general anesthesia and postoperative pain management. Our results suggest that short-term recovery from general anesthesia might provide an index of functional recovery and functional reserve of the brain during the stress of the perioperative period.
The definition of delirium during recovery from anesthesia has varied among studies, and this led to a wide range of reported incidence. In the pediatric population, delirium on recovery from anesthesia has been equated with agitation (13). Other studies in adults also used agitation as a marker for delirium (14). Studies of postoperative delirium in the geriatric population, however, report that hypoactive forms of delirium are more prevalent than hyperactive forms (15). This suggests that previous studies might have under-recognized recovery room delirium. The current study used the CAM score to diagnose recovery room delirium. This method is more likely than others to identify hypoactive forms of delirium (15). The 45% incidence of recovery room delirium in our study is high compared with previous reports of incidences ranging from 3% to 20% (6). This difference might be related both to methodology and to the elderly population studied.
Numerous predictors have been documented for the development of postoperative delirium (16–18). Many of these predictors are related to underlying comorbidity. In particular, preoperative cognitive dysfunction is a strong predictor of postoperative delirium (1,2,16,18). Likewise, in the current study, dementia was associated with postoperative delirium. It is unlikely that preoperative delirium had an important role in determining recovery room delirium or postoperative delirium. There were 4 patients with preoperative delirium, of which 3 remained delirious in the recovery room and postoperatively, whereas 1 patient showed no further evidence of delirium at any time postoperatively. However, drugs have been heavily implicated in the onset of delirium (19). Clinical studies have implicated sevoflurane (14), meperidine, benzodiazepines (20), and anticholinergics (21) as drugs frequently administered during general anesthesia that might be associated with postoperative delirium. However, we do not know what role the anesthetic protocol used in the current study may have had in postoperative delirium.
The strength of this study is that it is a prospective blinded case series that examined one type of surgical procedure, i.e., hip-fracture repair. A standardized protocol for general anesthesia was used, with a well-defined recovery profile. In addition, pain management was standardized in the PACU and on the hospital ward. Delirium, both recovery room and postoperative, was diagnosed using a standardized screening tool, the CAM score, which has an excellent sensitivity and specificity (9).
Our study has several shortcomings. For one, measuring anything subjective in the short interval immediately after discontinuing anesthesia in the elderly can be difficult. To decrease variability in our study, anesthesia was administered via a standardized protocol. In our study, when patients were given a standardized anesthesia and for the same type of surgery, the recovery profile was highly reproducible. In this study, we performed the CAM score at 30 minutes. We found 22 patients were CAM positive at 30 minutes, as compared with 21, which were CAM positive at 60 minutes. Thus, the measurement seems to be reproducible over time, although more studies would be needed to verify this. Second, although we used a standardized anesthetic protocol and the doses of drugs administered were comparable among patients, we did not directly measure the depth of anesthesia. Thus, we cannot be certain that differences in recovery room delirium were not secondary to differences in anesthetic depth. Third, although the postoperative analgesic administration was standardized and based on a pain scale, the study population was primarily elderly, with a 51% incidence of dementia. Thus, the pain scale might have been inaccurate in those patients with preexisting cognitive dysfunction, and this in turn might have led to inappropriate analgesic dosing. Postoperative pain at rest has been associated with delirium (22). In addition, we may have underestimated the central complications of PCA morphine.
We recommend caution in generalizing the current study results to other anesthetic techniques. A technique of propofol, succinylcholine, isoflurane in oxygen, and morphine with spontaneous respiration for >2 hours would be uncommon in most busy orthopedic centers dealing with elderly patients. This anesthetic combination administered by the authors provided cardiovascular stability. However, in 2 of 47 patients, phenylephrine was needed for arterial blood pressure support to maintain a mean arterial blood pressure within 20% of baseline. It is unclear whether our findings are specific to the drugs used or whether they apply to other types of anesthetics. Nonetheless, the importance of this study is that it defines a new, strong predictor of postoperative delirium. Our results suggest that the presence or absence of recovery room delirium might give us an index of functional recovery and functional reserve of the brain during the stress of the perioperative period.
1.Dyer CB, Ashton CM, Teasdale TA. Postoperative delirium: a review of 80 primary data-collection studies. Arch Intern Med 1995;155:461–5.
2.Edlund A, Lundstrom M, Lundstrom G, et al. Clinical profile of delirium in patients treated for femoral neck fractures. Dement Geriatr Cogn Disord 1999;10:325–9.
3.Marcantonio ER, Flacker JM, Michaels M, Resnick NM. Delirium is independently associated with poor functional recovery after hip fracture. J Am Geriatr Soc 2000;48:618–24.
4.Francis J, Martin D, Kapoor WN. A prospective study of delirium in hospitalized elderly. JAMA 1990;263:1097–101.
5.Magaziner J, Simonsick EM, Kashner TM, et al. Survival experience of aged hip fracture patients. Am J Public Health 1989;79:274–8.
6.Haynes C. Emergence delirium: a literature review. Br J Theatre Nurs 1999;9:502–3, 6–10.
7.Folstein MF, Folstein SE, McHugh PR. “Mini-mental state”: a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12:189–98.
8.Inouye SK, van Dyck CH, Alessi CA, et al. Clarifying confusion: the confusion assessment method—a new method for detection of delirium. Ann Intern Med 1990;113:941–8.
9.Cockrell JR, Folstein MF. Mini-Mental State Examination (MMSE). Psychopharmacol Bull 1988;24:689–692.
10.Yasuda N, Lockhart SH, Eger EI 2nd, et al. Kinetics of desflurane, isoflurane, and halothane in humans. Anesthesiology 1991;74:489–98.
11.Buss HE, Melderis K. PACU pain management algorithm. J Perianesth Nurs 2002;17:11–20.
12.Farley DR, Greenlee SM. et al. Double-blind, prospective, randomized study of warmed, humidified carbon dioxide insufflation vs standard carbon dioxide for patients undergoing laparoscopic cholecystectomy. Arch Surg 2004;139:739–43.
13.Aono J, Ueda W, Mamiya K, et al. Greater incidence of delirium during recovery from sevoflurane anesthesia in preschool boys. Anesthesiology 1997;87:1298–300.
14.Nunn JF. Isoflurane as a routine anaesthetic in general surgical practice. Br J Anaesth 1985;57:461–75.
15.Marcantonio E, Ta T, Duthie E, Resnick NM. Delirium severity and psychomotor types: their relationship with outcomes after hip fracture repair. J Am Geriatr Soc 2002;50:850–7.
16.Galanakis P, Bickel H, Gradinger R, et al. Acute confusional state in the elderly following hip surgery: incidence, risk factors and complications. Int J Geriatr Psychiatry 2001;16:349–55.
17.Marcantonio ER, Goldman L, Mangione CM, et al. A clinical prediction rule for delirium after elective noncardiac surgery. JAMA 1994;271:134–9.
18.Williams-Russo P, Urquhart BL, Sharrock NE, Charlson ME. Post-operative delirium: predictors and prognosis in elderly orthopedic patients. J Am Geriatr Soc 1992;40:759–67.
19.Brown TM. Drug-induced delirium. Semin Clin Neuropsychiatry 2000;5:113–24.
20.Marcantonio ER, Juarez G, Goldman L, et al. The relationship of postoperative delirium with psychoactive medications. JAMA 1994;272:1518–22.
21.Parikh SS, Chung F. Postoperative delirium in the elderly. Anesth Analg 1995;80:1223–32.
22.Lynch EP, Lazor MA, Gellis JE, et al. The impact of postoperative pain on the development of postoperative delirium. Anesth Analg 1998;86:781–5.