- Email: [email protected]
Predictors of Electroconvulsive Therapy Postictal Delirium
Psychosomatics 2014:55:272–279
& 2014 The Academy of Psychosomatic Medicine. Published by Elsevier Inc. All rights reserved.
Original Research Reports Predictors of Electroconvulsive Therapy Postictal Delirium Irving M. Reti, M.B.B.S., Aparna Krishnan, B.S., Adam Podlisky, B.A., April Sharp, B.A., Melinda Walker R.N., Karin J. Neufeld, M.D., Matthew J. Hayat, Ph.D.
Background: Postictal delirium is a common adverse effect of electroconvulsive therapy (ECT) and can be dangerous to both patient and staff caring for them in the postanesthesia care unit. However, little is known about predictors of postictal delirium. Objectives: The aim of this study was to identify predictors of postictal delirium. We hypothesized that both patient and ECT treatment variables might influence the likelihood of postictal delirium. Methods: We prospectively monitored postictal delirium in the postanesthesia care unit using the Confusion Assessment Method for the Intensive Care Unit after the first ECT treatment of 96 consecutive patients. Patient and treatment variables were extracted
retrospectively by chart review. A multiple logistic regression model was developed to assess the effect of these variables on the likelihood of developing delirium. Results: Seizure length was found to be a statistically significant predictor of postictal delirium after adjusting for other covariates (p ¼ 0.003). No other variables were predictive. Conclusion: A long ECT seizure increases the likelihood of delirium in the postanesthesia care unit at the first treatment. This finding suggests that postanesthesia care unit staff may benefit from knowledge about seizure length for predicting postictal delirium and anticipating the best management of post-ECT patients. (Psychosomatics 2014; 55:272–279)
E
especially the elderly, may also be at increased risk of falls post-ECT.6 Additionally, postictal delirium is also significant as a predictor of later ECT-related cognitive side effects, including memory loss. Sobin et al. found that time to orientation post-ECT predicted the magnitude of retrograde amnesia in the week after the course of ECT and at 2-month follow-up.7 As postictal delirium is potentially dangerous to both patient and staff and is associated with later cognitive side
lectroconvulsive therapy (ECT) is a highly effective treatment for major depressive disorder and other psychiatric conditions. A common adverse effect of ECT immediately following the procedure is the emergence of postictal delirium. This state is characterized by a lack of awareness, disorientation, agitation, and sometimes erratic and even violent behavior lasting between 5 and 45 minutes,1 occurring in as many as 52% of patients.2 Although postictal delirium generally lasts less than 1 hour, overall consciousness remains blunted for several hours.3 In contrast, Katznelson et al., using the confusion assessment method for the intensive care unit (CAMICU), found only 11.9% were delirious after cardiac surgery, and Radtke et al. reported a postoperative delirium rate of 11% in the postanesthesia care unit (PACU) using the NuDeSC, another screening tool for delirium.4,5 The increased frequency and severity of postictal delirium is thought to be due to the seizure itself. During postictal delirium following ECT administration, patients become a hazard to themselves as well as others, such as nurses in the PACU.1 Delirious patients, 272
www.psychosomaticsjournal.org
Received January 21, 2013; revised March 14, 2013; accepted March 18, 2013. From Department of Psychiatry and Behavioral Sciences, Johns Hopkins University School of Medicine, Baltimore, MD, USA; Department of Neuroscience, Johns Hopkins University School of Medicine, Baltimore, MD, USA; Zanvyl Krieger School of Arts and Sciences, Johns Hopkins University, Baltimore, MD, USA; Johns Hopkins University School of Medicine, Baltimore, MD, USA; College of Nursing, Rutgers University, Newark, NJ, USA. Send correspondence and reprint requests to Irving Reti, M.B.B.S., Department of Psychiatry and Behavioral Sciences, Johns Hopkins University School of Medicine, 600N. Wolfe St., Meyer 3-181, Baltimore, MD 21205. e-mail: [email protected] & 2014 The Academy of Psychosomatic Medicine. Published by Elsevier Inc. All rights reserved.
Psychosomatics 55:3, May/June 2014
Reti et al. effects, it would be helpful to identify predictors of this state. Previous studies have yielded mixed results. Sackeim et al. examined electrode placement in 2 patients, 1 left-handed and the other right-handed, receiving ECT with right unilateral (RUL), left unilateral, and bilateral (BL) placement at different times. They found left unilateral placement was not associated with postictal delirium, unlike RUL and BL placements.1 Conversely, Leechuy and Abrams reported postictal delirium in a right-handed man receiving ECT with left unilateral placement.8 Devanand et al. analyzed potential predictors in a retrospective case control study consisting of 24 patients who experienced postictal delirium and 24 controls who did not.9 The 2 groups did not differ in age, gender, diagnosis, anesthesia and succinylcholine dosages, electrode placement, mean seizure duration, or clinical outcome.9 More recently, Sackeim et al. showed that BL placement was associated with both an increased rate of prolonged disorientation and a longer time to recover orientation compared with RUL placement.10 In contrast, Kikuchi et al. found that pretreatment catatonic features were the only significant predictor of postictal delirium severity. Kellner et al. also examined reorientation scores at 20 minutes post-ECT and did not find a difference due to electrode placement when averaging across the full treatment course, although there was a trend towards RUL and bitemporal placement being associated with higher scores than bifrontal placement.11 To help monitor postictal delirium in the PACU, we recently began assessing ECT patients with the CAM-ICU. The CAM-ICU is a brief delirium assessment tool for nurses, which has been demonstrated to have high interrater reliability, sensitivity, and specificity.12 Our nurses also monitor the RASS (Richmond Agitation-Sedation Scale), a sedation scale used to characterize the patient’s alertness and responsiveness.13 We assessed whether a variety of patient and ECT treatment variables might influence the likelihood of postictal delirium. METHODS Setting The setting for this study was The Johns Hopkins Hospital in Baltimore, Maryland, which has a large inpatient psychiatric unit attached to the general hospital. The ECT suite is open for treatment 3 days per week. Approximately 125 inpatients per year are treated with ECT at Johns Hopkins. The ECT PACU is Psychosomatics 55:3, May/June 2014
directly adjacent to the treatment room and is staffed by the ECT nurse coordinator, plus one other registered nurse of a small pool of 5 nurses who are PACUcredentialed. Design and Sample The study population was all consecutive inpatients beginning an acute series of ECT within the time frame October 2009–September 2010. Outpatients were not included as the CAM-ICU only began being administered to them at the end of 2010. Only data from the first ECT treatment of a series were considered in this report, as the presence of delirium at subsequent treatments might be affected by changes designed to minimize delirium at subsequent treatments. For example, if patients are agitated in the PACU after their first treatment, we may premedicate them with a low-dose neuroleptic for subsequent treatments or accelerate their taper from concurrent psychotropic medications. Ninety-six patients were included in the study. Of these, 95 had CAM-ICU assessments and 94 had RASS ratings. Data on both dependent and independent variables was gathered from a retrospective review of the electronic record of all included patients. This study was reviewed and approved by The Johns Hopkins Institutional Review Board. To protect patient confidentiality, medical record data were recorded using nonidentifiable codes and kept separate from patients’ names or medical history numbers. ECT Administration ECT was administered as previously described.14,15 Before each treatment, patients were prehydrated overnight with lactated Ringer solution. General anesthesia was induced by 1–1.5 mg/kg of methohexital, and 1–1.5 mg/kg of succinylcholine was used as the muscle relaxant. Charge was based on age, being 5 age in millicoulombs (mc) for RUL placement with brief pulses, 2.5 age for RUL with ultrabrief (UB) pulses, and 2.5 age for BL with bitemporal or bifrontal placement. Typical electrical parameters at the first treatment were as follows: pulse width—0.3 ms (UB) or 1 ms (brief, regardless of electrode placement); stimulus duration— 6 s (UB) or 4 s (brief, although usually less for BL placement); frequency varies with age so that total charge was a product of patient age as described above; and current was always 800 mA. Higher charge was used in www.psychosomaticsjournal.org
273
Predictors of Electroconvulsive Therapy Postictal delirium patients taking anticonvulsants, such as benzodiazepines. At the time of this study, only patients who had experienced significant cognitive side effects during a prior ECT course were started on UB pulses. Accordingly, only 7.3% (N ¼ 7) of patients were treated with UB pulses for their first treatment. Also, only 12.5% (N ¼ 12) of patients commenced treatment with BL placement, being those who had previously required such after failing ECT with RUL placement. Patients were hyperventilated to achieve a minute ventilation of 20 L/min, which was maintained for 60 seconds. ECT was administered with a Mecta Spectrum 5000Q unit (Mecta Corp, Tualatin, Ore). Seizure length and postictal suppression were monitored by electroencephalography. Postictal suppression is rated manually on a 4-point Likert scale from 0–3.16 A score of 3 signifies excellent postictal suppression and a score of 0 indicates no postictal suppression. Outcome Measures The 2 dependent variables assessed were the following1: presence or absence of delirium as assessed by the CAMICU and sedation level as assessed by the RASS2. Because a key feature of delirium is the presence or absence of inattention, the RASS is often assessed in conjunction with the CAM-ICU, as at The Johns Hopkins ECT PACU. However, delirium is considered to be present based solely on whether or not the CAM-ICU is positive. CAM-ICU The Confusion-Assessment Method tool was developed by Inouye et al. for monitoring delirium and was modified by Ely et al. to assess delirium specifically in ICU patients.12,17 The CAM-ICU uses similar criteria as Diagnostic and Statistical Manual of Mental Disorders-IV (DSM-IV) for diagnosing delirium. It is a binary variable indicating the presence or absence of delirium. The CAM-ICU is considered positive for delirium if there is an acute change in mental status (noting there is no change in mental status in our patients until the ECT treatment), plus more than 2 errors in response to an auditory stimulus, plus either the RASS score being anything other than 0 or more than 1 error in response to questions assessing for disorganized thinking.12 The CAMICU can be conducted within 2–3 minutes and is an easy assessment tool for nurses in the PACU. It has been validated in studies from ICUs around the world.18–23 For example, Guenther et al. found high sensitivity (88–92%), specificity (100%), and interrater reliability (k 0.96) compared with the reference standard DSM-IV definition of delirium.21 274
www.psychosomaticsjournal.org
RASS The RASS is a 10-point sedation scale ranging from5 to þ4. Unlike other sedation tests such as the Ramsey Scale and the Sedation-Agitation Scale, the RASS includes responses to both verbal and physical stimuli, and the patient’s level of arousal is graded in part according to the stimulus strength.13 Negative scores indicate increasing levels of sedation, with 5 corresponding to an unarousable patient, whereas positive scores indicate increasing levels of agitation with þ4 corresponding to a combative patient. A score of þ2 was considered to be threshold for dangerous levels of agitation (defined as ‘‘a patient with frequent nonpurposeful movements, who fights his or her ventilator’’), so scores of þ2 and higher were grouped together, whereas scores below þ2 were grouped as nonhazardous. The RASS takes less than 20 seconds to perform and has also been shown to have high interrater reliability and criterion, construct, and face validity.13 ECT PACU nurses at Hopkins rate both measures. The CAM-ICU is administered during the patient’s third set of vital sign assessments, approximately 12–15 minutes after the patient’s arrival to the PACU. Two RASS scores are recorded during the patient’s stay in the PACU: the first reflects the patient’s worst agitation state in the PACU, whereas the second RASS score reflects the level of agitation or sedation on leaving the PACU. The RASS scores included in this study are the former of these 2 scores. Independent Variables Independent variables assessed were as follows:
Patient variables: age, weight, gender, diagnosis, and
number of concomitant psychotropic medications taken within 48 hours before the first treatment. We included consumption of the following psychotropic medications: neuroleptics, antidepressants, benzodiazepines, and other anticonvulsants plus lithium. Note that if patients were taking 2 antidepressant medications, they were considered to be taking 2 concomitant medications. ECT treatment variables: dosages of methohexital and succinylcholine, electrode placement, total charge (mC), seizure length, and postictal suppression. Statistical Analysis
The data analyses for this paper were generated using SAS software, Version 9.1.3 of the SAS System for Psychosomatics 55:3, May/June 2014
Reti et al. Windows. Bivariate relationships between independent and dependent variables were examined using the Chisquare test, Fisher exact test, or independent samples t-test. Results of bivariate analyses were used to build a multiple logistic regression model for each dependent variable. In addition, a graphical depiction was made to display the functional relationship between seizure length and probability of delirium. The level of significance for this study was set at a ¼ 0.05. A Bonferroni correction was used and calculated as a divided by the number of variables used in multiple comparisons to derive a cutoff value (0.05/11 ¼ 0.0045). P values resulting from statistical tests need to be below this value for a comparison to be considered more than a chance result. RESULTS In Table 1, we present baseline demographic details for patients included in the study. Summary statistics for both delirium assessment and dichotomized RASS scores are provided. Sixty-two patients (65.3%) were positive for delirium, whereas 33 patients (34.7%) did not have delirium. The average seizure length for patients without delirium was 59.2 seconds, whereas the average seizure length for patients with delirium was 85.7 seconds (p o 0.0001, by independent samples t-test). In Table 2, we show results for multiple logistic regression models. Seizure length was found to still be a statistically significant predictor of postictal delirium after adjusting for other covariates (p ¼ 0.003). No other variables were predictive. The relationship between seizure length and delirium status is displayed in Fig. 1. This figure is based on the unadjusted model and shows an increased likelihood of delirium with an increasing seizure length. For example, a 30 second seizure is associated with an approximately 30% chance of delirium; an 80 second seizure is associated with a greater than 70% chance of delirium. None of the covariates studied were predictive of dichotomized RASS score. In addition, a linear regression analysis using the RASS score as a continuous dependent variable also failed to identify significant predictors amongst the independent variables (data not shown). DISCUSSION We report a significant positive association between seizure length and probability of postictal delirium at Psychosomatics 55:3, May/June 2014
the first treatment after adjusting for other covariates. Patients who had a seizure of 80 seconds or more were very likely to be delirious in the PACU. These findings are consistent with observations that seizure length also influences the severity of postictal delirium amongst epileptics.24 It is not clear why longer electroconvulsive seizures might lead to an increased chance of postictal delirium. Longer seizures may reflect recruitment of a greater number of neurons into the seizure. In addition, or alternatively, the same number of neurons may fire for a longer time. Under either scenario, a longer seizure might be associated with higher neurotransmitter release. For example, it is known that inhibitory neurotransmitters such as gamma-aminobutyric acid and adenosine are released by seizure activity, which could act to transiently dampen normal neuronal activity post treatment, and this might be enhanced by a longer seizure.25,26 We did not find an effect of electrode placement on risk of postictal delirium. In fact, in our sample, a Fisher exact test shows a higher rate of postictal delirium in patients receiving RUL compared with patients receiving BL treatment (Table 1). However the effect of placement was not significant in the logistic regression model (Table 2). Although age is a risk factor for delirium in many medical settings,27,28 we did not find it to be a significant factor for development of postictal delirium, consistent with previous studies.2,9 Regarding gender, seizure length is longer for female patients at the first treatment compared with subsequent treatments.29 Nonetheless, we did not find that gender predicted postictal delirium, consistent with Kikuchi et al.2 We were surprised that delirium was not associated with the number of medications taken by the patient in the 48 hours leading up to the treatment. Additionally, exploratory analyses by medication class (data not shown) failed to show an association between consumption of neuroleptics, antidepressants, benzodiazepines, other anticonvulsants plus lithium and either delirium or RASS score. We also did not find an association between seizure length and consumption of these medication classes, including benzodiazepines which might have been due to tapering of benzodiazepines in at least two-thirds of patients consuming them in the 48 hours before ECT treatment. We also did not find an association between the dose of methohexital and postictal delirium. All patients are adequately anesthetized before charge administration and accordingly, it is not surprising that methohexital dose is not associated with postictal delirium. However, in www.psychosomaticsjournal.org
275
Predictors of Electroconvulsive Therapy Postictal delirium
Baseline Demographics. A Bonferroni Correction was Applied to Adjust for Multiple Statistical Tests; P values o 0.0045 were Interpreted as Statistically Significant
Table 1.
Age (y) Weight (Kg) Methohexital (mg/kg) Succinylcholine dosage (mg/kg) Charge (mC) Seizure length (s) Postictal suppression score
Total N ¼ 96
No delirium N ¼ 33
Delirium N ¼ 62
Mean (SD)
Mean (SD)
Mean (SD)
52.8 73.6 1.2 1.2 269.0 76.5 1.9
52.3 77.0 1.2 1.1 262.6 59.6 1.8
53.6 71.9 1.2 1.2 274.7 85.7 2.0
(17.9) (18.7) (0.3) (0.2) (124.8) (36.8) (0.7)
N (%)
(15.3) (18.3) (0.2) (0.2) (140.4) (20.1) (0.8)
N (%)
(19.1) (18.9) (0.3) (0.3) (116.3) (40.8) (0.6)
p-Value
No agitation (RASS o 2) N ¼ 77 Mean (SD)
0.74* 0.20* 0.38* 0.37* 0.66* o0.0001* 0.47*
53.8 73.8 1.2 1.1 275.0 77.9 2.0
N (%)
Agitation (RASS Z 2) N ¼ 17 Mean (SD)
(18.9) (17.6) (0.2) (0.2) (130.8) (38.4) (0.7)
N (%)
50.2 75.5 1.2 1.2 250.7 73.2 1.7
(12.8) (22.5) (0.2) (0.2) (96.6) (30.4) (0.7)
41 (43) 55 (57)
15 (45) 18 (55)
25 (40) 37 (60)
Diagnosis Major depressive disorder Bipolar disorder Schizoaffective disorder
70 (73) 21 (22) 4 (5)
23 (70) 10 (30) 0 (0)
46 (75) 11 (18) 4 (7)
Placement RUL BL with bitemporal placement
84 (88) 12 (12)
25 (76) 8 (24)
58 (94) 4 (6)
Number of medications 0 1 2 3 4
11 22 37 22 4
3 7 14 7 2
8 15 22 15 2
0.90y 33 (43) 44 (57)
7 (41) 10 (59)
57 (75) 17 (22) 2 (3)
11 (65) 4 (23) 2 (12)
68 (88) 9 (12)
14 (82) 3 (18)
0.19z
0.21z
0.02z
0.45z
0.89z (11) (23) (38) (23) (5)
(9) (21) (43) (21) (6)
(13) (24) (35) (24) (4)
0.45* 0.73* 0.89* 0.63* 0.47* 0.64* 0.19*
N (%)
0.63y
Sex Male Female
p-Value
0.26z 10 14 30 20 3
(13) (18) (39) (26) (4)
1 7 6 2 1
(6) (41) (35) (12) (6)
BL ¼ bilateral; RUL ¼ right unilateral; SD ¼ standard deviation. n
Independent samples t-test. y Chi-square test. z Fisher exact test.
settings where ECT is administered unmodified with no anesthetic agent, postictal delirium occurs more frequently.30,31 Likewise, succinylcholine dose was not associated with postictal delirium, consistent with Devanand et al.9 It is important to note that a positive CAM-ICU does not necessarily denote delirium of sufficient severity that it places the staff or patient in danger. Two-thirds of our patients were positive on the CAM-ICU, however only 18% (17 out of 94) of patients scored Z2 on the RASS indicating a dangerous level of agitation. These results suggest that not all delirious patients were hyperactive and dangerous to nurses; however, even if the delirium is hypoactive, it is nonetheless potentially dangerous to the patient possibly placing them at increased risk of falls and 276
www.psychosomaticsjournal.org
aspiration. Our data are comparable to the findings of the Kikuchi sample2 that reported 52% of patients experienced at least mild delirium post-ECT. Limitations Although we found an association between seizure length and probability of postictal delirium, we have not demonstrated causality. Another possible explanation for our finding is that seizure length is a marker for an unidentified factor that could cause postictal delirium. Despite our analysis, the role of concomitant medications is difficult to evaluate. For example, benzodiazepine withdrawal could predispose to both longer seizures and to postictal delirium. We also did not evaluate Psychosomatics 55:3, May/June 2014
Reti et al. Table 2.
Logistic Regression Models for Delirium and RASS (Z2). A Bonferroni Correction was Applied to Adjust for Multiple Statistical Tests; p Values o 0.0045 were Interpreted as Statistically Significant Delirium—adjusted model
RASS—adjusted model
OR
OR
95% Confidence interval
Estimate Lower
Upper
95% Confidence interval
p-Value Estimate Lower
Upper
p-Value
Number of medications Age Weight (kg) Charge Seizure length Postictal suppression Factor construct (brevital and succinylcholine)
1.19 1.05 1.00 1.00 1.05 0.87 1.78
0.68 1.00 0.96 0.99 1.02 0.40 0.87
2.10 1.11 1.03 1.00 1.08 1.86 3.65
0.54 0.052 0.89 0.22 0.003 0.71 0.12
0.77 0.98 1.01 1.00 0.99 0.60 1.09
0.41 0.93 0.97 0.99 0.97 0.25 0.47
1.45 1.03 1.05 1.01 1.01 1.43 2.51
0.42 0.41 0.69 0.80 0.38 0.24 0.84
Sex Male Female
0.84 0.26 Reference
3.65
0.76
0.92 0.25 Reference
3.44
0.90
Diagnosis Major depressive disorder Bipolar or shizoaffective disorder
2.21 0.70 Reference
7.00
0.18
0.50 0.15 Reference
1.70
0.26
Placement RUL BL
5.70 0.90 Reference
36.26
0.07
0.73 0.09 Reference
6.34
0.78
BL ¼ bilateral; OR ¼ odds ratio; RASS ¼ Richmond Agitation-Sedation Scale; RUL ¼ right unilateral.
nonpsychotropic drugs or medical, especially neurologic, comorbidities. Finally, we did not evaluate the adequacy of oxygenation during the treatment or in recovery, or the extent of hyperventilation which could affect seizure duration. To decrease seizure threshold, patients are hyperventilated before electrical stimulation which
Figure 1.
causes a transient respiratory alkalosis. The blood alkalization in turn causes vasoconstriction which affects cerebral perfusion and potentially the chances of delirium following the seizure. Our ability to determine the effect of charge dose on the association between seizure length and postictal
Probability of Delirium as a Function of Seizure Length. 1 0.9 0.8 0.7 0.6
Probability 0.5 of Delirium 0.4 0.3 0.2 0.1 0
0
50
100
150
200
250
300
Seizure Length
Psychosomatics 55:3, May/June 2014
www.psychosomaticsjournal.org
277
Predictors of Electroconvulsive Therapy Postictal delirium delirium was limited by use of age to determine charge dose rather than seizure threshold. Adverse cognitive effects of ECT are attributable to the magnitude of the stimulus dosage relative to seizure threshold rather than to the absolute electrical charge.32,33 Likewise, the effect of charge on seizure length is mediated by the ratio of charge to seizure threshold, with brief seizures occurring at seizure threshold and long seizures occurring just above seizure threshold. Once significantly above seizure threshold, seizure duration seems to vary inversely with charge.34,35 We have only presented outcome data pertaining to the first of a course of ECT treatments and have not analyzed subsequent treatments, which could be altered by changes made after the first treatment if the patient were delirious. Therefore, it is not known whether the relationship we found between seizure length and delirium would hold for subsequent treatments. Both seizure length and risk of delirium change as the treatment series progresses. As mentioned before, seizure length typically decreases after the first treatment,29 and risk of cognitive side effects increases as the treatment series progresses.36,37 These changes may alter the seizure length-delirium association. Accordingly, the implications that can be drawn from our data for postindex treatments are limited.
Another limitation of the study pertains to the outcome measures themselves. As far as we are aware, neither the CAM-ICU nor RASS have ever been evaluated in an ECT PACU, so we have no data about their validity, reliability, sensitivity, and specificity for monitoring postictal delirium. However, 2 recent meta-analyses of the CAM-ICU, from studies in ICUs, found specificity of 95.8%38 and 95.9%.39 Furthermore, even in settings outside the ICU, the CAM-ICU does have excellent specificity.40 In conclusion, we have shown that a long ECT seizure increases the likelihood of delirium in the PACU at the first treatment. This finding suggests PACU staff may benefit from knowledge about seizure length for predicting postictal delirium and anticipating the best management of post-ECT patients. Disclosure: No financial support was provided for this study. Dr Reti has received grant support from NIH and HDRF and study supplies at no cost from Neuronetics Inc. Dr Reti is also site PI on a multisite rTMS trial sponsored by Brainsway Inc. The other authors report no proprietary or commercial interest in any product mentioned or concept discussed in this article.
References 1. Sackeim HA, Decina P, Malitz S, Hopkins N, Yudofsky SC, Prohovnik I: Postictal excitement following bilateral and rightunilateral ECT. Am J Psychiatry 1983; 140:1367–1368 2. Kikuchi A, Yasui-Furukori N, Fujii A, Katagai H, Kaneko S: Identification of predictors of post-ictal delirium after electroconvulsive therapy. Psychiatry Clin Neurosci 2009; 63:180–185 3. Fink M: Post-ECT delirium. Cogn Ther 1993; 9:326–330 4. Katznelson R, Djaiani G, Tait G, et al: Hospital administrative database underestimates delirium rate after cardiac surgery. Can J Anaesth 2010; 57:898–902 5. Radtke FM, Franck M, MacGuill M, et al: Duration of fluid fasting and choice of analgesic are modifiable factors for early postoperative delirium. Eur J Anaesthesiol 2010; 27:411–416 6. Rao SS, Daly JW, Sewell DD: Falls associated with electroconvulsive therapy among the geriatric population: a case report. J ECT 2008; 24:173–175 7. Sobin C, Sackeim HA, Prudic J, et al: Predictors of retrograde amnesia following ECT. Am J Psychiatry 1995; 152:995–1001 8. Leechuy I, Abrams R: Postictal delirium (and recovery from melancholia) after left-unilateral ECT. Convuls Ther 1987; 3:65–68 9. Devanand DP, Bricose KM, Sackeim HM: Clinical features and predictors of postictal excitement. Convuls Ther 1989; 5:140–146 10. Sackeim HA, Prudic J, Devanand DP, et al: Prospective, randomized, double-blind comparison of bilateral and right unilateral electroconvulsive therapy at different stimulus intensities. Arch Gen Psychiatry 2000; 57:425–434
278
www.psychosomaticsjournal.org
11. Kellner CH, Tobias KG, Wiegand J: Electrode placement in electroconvulsive therapy (ECT): a review of the literature. J ECT 2010; 26:175–180 12. Ely EW, Margolin R, Francis J, et al: Evaluation of delirium in critically ill patients: validation of the confusion assessment method for the intensive care unit (CAM-ICU). Crit Care Med 2001; 29:1370–1379 13. Ely EW, Truman B, Shintani A, et al: Monitoring sedation status over time in ICU patients: reliability of the richmond agitationsedation scale(RASS). JAMA 2003; 289:2983–2991 14. Wachtel LE, Contrucci-Kuhn SA, Griffin M, et al: ECT for selfinjury in an autistic boy. Eur Child Adolesc Psychiatry 2009; 18:458–463 15. Wachtel LE, Griffin M, Reti IM: Electroconvulsive therapy in a man with autism experiencing severe depression, catatonia, and self-injury. J ECT 2010; 26:70–73 16. Nobler MS, Sackeim HA, Solomou M, Lube B, Devanand DP, Prudic J: EEG manifestations during ECT: effects of electrode placement and stimulus intensity. Biol Psychiatry 1993; 34:321–330 17. Inouye SK, van Dyck CH, Alessi CA, Balkin S, Siegal AP, Horwitz RI: Clarifying confusion: the confusion assessment method. a new method for detection of delirium. Ann Intern Med 1990; 113:941–948 18. Pun BT, Gordon SM, Peterson JF, et al: Large-scale implementation of sedation and delirium monitoring in the intensive
Psychosomatics 55:3, May/June 2014
Reti et al.
19.
20.
21.
22.
23.
24. 25.
26.
27. 28.
29.
30.
care unit: a report from two medical centers. Crit Care Med 2005; 33:1199–1205 Chuang WL, Lin CH, Hsu WC, Ting YJ, Lin KC, Ma SC: Evaluation of the reliability and validity of the Chinese version of the confusion assessment method for the intensive care unit. Hu Li Za Zhi 2007; 54:45–52 Van Eijk MM, van Marum RJ, Klign IA, de Wit N, Kesecioglu J, Slooter AJ: Comparison of delirium assessment tools in a mixed intensive care unit. Crit Care Med 2009; 37:1881–1885 Guenther U, Popp J, Koecher L, et al: Validity and reliability of the CAM-ICU flowsheet to diagnose delirium in surgical ICU patients. J Crit Care 2010; 25:144–151 L¨utz A, Heymann A, Radtke FM, Spies CD: Different assessment tools for intensive care unit delirium: which score to use? Crit Care Med 2010; 38:409–418 Tobar E, Romero C, Galleguillos T, et al: Confusion assessment method for diagnosing delirium in ICU patients (CAM-ICU): cultural adapation and validation of the Spanish version. Med Intensiva 2010; 34:4–13 Krauss G, Theodore WH: Treatment strategies in the postictal state. Epilepsy Behav 2010; 19:188–190 During MJ, Spencer DD: Adenosine: a potential mediator of seizure arrest and postictal refractoriness. Ann Neurol 1992; 32:618–624 Dunwiddie TV, Masino SA: The role and regulation of adenosine in the central nervous system. Annu Rev Neurosci 2001; 24:31–55 Inouye SK: Delirum in older persons. N Engl J Med 2006; 354:1157–1165 Peterson JF, Pun BT, Dittus RS, et al: Delirium and its motoric subtypes: a study of 614 critically ill patients. J Am Geriatr Soc 2006; 54:479–484 Rasimas JJ, Stevens SR, Rasmussen KG: Seizure length in electroconvulsive therapy as a function of age, sex, and treatment number. J ECT 2007; 23:14–16 Gallinek A: Fear and anxiety in the course of electroshock therapy. Am J Psychiatry 1956; 113:428–434
Psychosomatics 55:3, May/June 2014
31. Gallegos J, Vaidya P, D’Agati D, et al: Decreasing adverse outcomes of unmodified electroconvulsive therapy: suggestions and possibilities. J ECT 2012; 28:77–81 32. Sackeim HA, Prudic J, Devanand DP, et al: Effects of stimulus intensity and electrode placement on the efficacy and cognitive effects of electroconvulsive therapy. N Engl J Med 1993; 328:839–846 33. McCall WV, Reboussin DM, Weiner RD, Sackeim HA: Titrated moderately suprathreshold vs fixed high-dose right unilateral electroconvulsive therapy: acute antidepressant and cognitive effects. Arch Gen Psychiatry 2000; 57:438–444 34. Sackeim HA, Devanand DP, Prudic J: Stimulus intensity, seizure threshold, and seizure duration: impact on the efficacy and safety of electroconvulsive therapy. Psychiatr Clin North Am 1991; 14:803–843 35. Frey R, Heiden A, Scharfetter J, et al: Inverse relation between stimulus intensity and seizure duration: implications for ECT procedure. J ECT 2001; 17:102–108 36. Ingram A, Saling MM, Schweitzer I: Cognitive side effects of brief pulse electroconvulsive therapy: a review. J ECT 2008; 24:3–9 37. Sackeim HA, Prudic J, Fuller R: The cognitive effects of electroconvulsive therapy in community settings. Neuropsychopharmacology 2007; 32:244–254 38. Neto AS, Nassar Jr AP, Cardoso SO, et al: Delirium screening in critically ill patients: a systematic review and meta-analysis. Crit Care Med 2012; 40:1946–1951 39. Gusmao-Flores D, Figueira Salluh JI, Chalhub RA, Quarantini LC: The confusion assessment method for the intensive care unit (CAM-ICU) and intensive care delirium screening checklist (ICDSC) for the diagnosis of delirium: a systematic review and meta-analysis of clinical studies. Crit Care Med 2012; 16:R115 40. Neufeld KJ, Hayat MJ, Coughlin JM, et al: Evaluation of two intensive care delirium screening tools for non-critically ill hospitalized patients. Psychosomatics 2011; 52:133–140
www.psychosomaticsjournal.org
279
& 2014 The Academy of Psychosomatic Medicine. Published by Elsevier Inc. All rights reserved.
Original Research Reports Predictors of Electroconvulsive Therapy Postictal Delirium Irving M. Reti, M.B.B.S., Aparna Krishnan, B.S., Adam Podlisky, B.A., April Sharp, B.A., Melinda Walker R.N., Karin J. Neufeld, M.D., Matthew J. Hayat, Ph.D.
Background: Postictal delirium is a common adverse effect of electroconvulsive therapy (ECT) and can be dangerous to both patient and staff caring for them in the postanesthesia care unit. However, little is known about predictors of postictal delirium. Objectives: The aim of this study was to identify predictors of postictal delirium. We hypothesized that both patient and ECT treatment variables might influence the likelihood of postictal delirium. Methods: We prospectively monitored postictal delirium in the postanesthesia care unit using the Confusion Assessment Method for the Intensive Care Unit after the first ECT treatment of 96 consecutive patients. Patient and treatment variables were extracted
retrospectively by chart review. A multiple logistic regression model was developed to assess the effect of these variables on the likelihood of developing delirium. Results: Seizure length was found to be a statistically significant predictor of postictal delirium after adjusting for other covariates (p ¼ 0.003). No other variables were predictive. Conclusion: A long ECT seizure increases the likelihood of delirium in the postanesthesia care unit at the first treatment. This finding suggests that postanesthesia care unit staff may benefit from knowledge about seizure length for predicting postictal delirium and anticipating the best management of post-ECT patients. (Psychosomatics 2014; 55:272–279)
E
especially the elderly, may also be at increased risk of falls post-ECT.6 Additionally, postictal delirium is also significant as a predictor of later ECT-related cognitive side effects, including memory loss. Sobin et al. found that time to orientation post-ECT predicted the magnitude of retrograde amnesia in the week after the course of ECT and at 2-month follow-up.7 As postictal delirium is potentially dangerous to both patient and staff and is associated with later cognitive side
lectroconvulsive therapy (ECT) is a highly effective treatment for major depressive disorder and other psychiatric conditions. A common adverse effect of ECT immediately following the procedure is the emergence of postictal delirium. This state is characterized by a lack of awareness, disorientation, agitation, and sometimes erratic and even violent behavior lasting between 5 and 45 minutes,1 occurring in as many as 52% of patients.2 Although postictal delirium generally lasts less than 1 hour, overall consciousness remains blunted for several hours.3 In contrast, Katznelson et al., using the confusion assessment method for the intensive care unit (CAMICU), found only 11.9% were delirious after cardiac surgery, and Radtke et al. reported a postoperative delirium rate of 11% in the postanesthesia care unit (PACU) using the NuDeSC, another screening tool for delirium.4,5 The increased frequency and severity of postictal delirium is thought to be due to the seizure itself. During postictal delirium following ECT administration, patients become a hazard to themselves as well as others, such as nurses in the PACU.1 Delirious patients, 272
www.psychosomaticsjournal.org
Received January 21, 2013; revised March 14, 2013; accepted March 18, 2013. From Department of Psychiatry and Behavioral Sciences, Johns Hopkins University School of Medicine, Baltimore, MD, USA; Department of Neuroscience, Johns Hopkins University School of Medicine, Baltimore, MD, USA; Zanvyl Krieger School of Arts and Sciences, Johns Hopkins University, Baltimore, MD, USA; Johns Hopkins University School of Medicine, Baltimore, MD, USA; College of Nursing, Rutgers University, Newark, NJ, USA. Send correspondence and reprint requests to Irving Reti, M.B.B.S., Department of Psychiatry and Behavioral Sciences, Johns Hopkins University School of Medicine, 600N. Wolfe St., Meyer 3-181, Baltimore, MD 21205. e-mail: [email protected] & 2014 The Academy of Psychosomatic Medicine. Published by Elsevier Inc. All rights reserved.
Psychosomatics 55:3, May/June 2014
Reti et al. effects, it would be helpful to identify predictors of this state. Previous studies have yielded mixed results. Sackeim et al. examined electrode placement in 2 patients, 1 left-handed and the other right-handed, receiving ECT with right unilateral (RUL), left unilateral, and bilateral (BL) placement at different times. They found left unilateral placement was not associated with postictal delirium, unlike RUL and BL placements.1 Conversely, Leechuy and Abrams reported postictal delirium in a right-handed man receiving ECT with left unilateral placement.8 Devanand et al. analyzed potential predictors in a retrospective case control study consisting of 24 patients who experienced postictal delirium and 24 controls who did not.9 The 2 groups did not differ in age, gender, diagnosis, anesthesia and succinylcholine dosages, electrode placement, mean seizure duration, or clinical outcome.9 More recently, Sackeim et al. showed that BL placement was associated with both an increased rate of prolonged disorientation and a longer time to recover orientation compared with RUL placement.10 In contrast, Kikuchi et al. found that pretreatment catatonic features were the only significant predictor of postictal delirium severity. Kellner et al. also examined reorientation scores at 20 minutes post-ECT and did not find a difference due to electrode placement when averaging across the full treatment course, although there was a trend towards RUL and bitemporal placement being associated with higher scores than bifrontal placement.11 To help monitor postictal delirium in the PACU, we recently began assessing ECT patients with the CAM-ICU. The CAM-ICU is a brief delirium assessment tool for nurses, which has been demonstrated to have high interrater reliability, sensitivity, and specificity.12 Our nurses also monitor the RASS (Richmond Agitation-Sedation Scale), a sedation scale used to characterize the patient’s alertness and responsiveness.13 We assessed whether a variety of patient and ECT treatment variables might influence the likelihood of postictal delirium. METHODS Setting The setting for this study was The Johns Hopkins Hospital in Baltimore, Maryland, which has a large inpatient psychiatric unit attached to the general hospital. The ECT suite is open for treatment 3 days per week. Approximately 125 inpatients per year are treated with ECT at Johns Hopkins. The ECT PACU is Psychosomatics 55:3, May/June 2014
directly adjacent to the treatment room and is staffed by the ECT nurse coordinator, plus one other registered nurse of a small pool of 5 nurses who are PACUcredentialed. Design and Sample The study population was all consecutive inpatients beginning an acute series of ECT within the time frame October 2009–September 2010. Outpatients were not included as the CAM-ICU only began being administered to them at the end of 2010. Only data from the first ECT treatment of a series were considered in this report, as the presence of delirium at subsequent treatments might be affected by changes designed to minimize delirium at subsequent treatments. For example, if patients are agitated in the PACU after their first treatment, we may premedicate them with a low-dose neuroleptic for subsequent treatments or accelerate their taper from concurrent psychotropic medications. Ninety-six patients were included in the study. Of these, 95 had CAM-ICU assessments and 94 had RASS ratings. Data on both dependent and independent variables was gathered from a retrospective review of the electronic record of all included patients. This study was reviewed and approved by The Johns Hopkins Institutional Review Board. To protect patient confidentiality, medical record data were recorded using nonidentifiable codes and kept separate from patients’ names or medical history numbers. ECT Administration ECT was administered as previously described.14,15 Before each treatment, patients were prehydrated overnight with lactated Ringer solution. General anesthesia was induced by 1–1.5 mg/kg of methohexital, and 1–1.5 mg/kg of succinylcholine was used as the muscle relaxant. Charge was based on age, being 5 age in millicoulombs (mc) for RUL placement with brief pulses, 2.5 age for RUL with ultrabrief (UB) pulses, and 2.5 age for BL with bitemporal or bifrontal placement. Typical electrical parameters at the first treatment were as follows: pulse width—0.3 ms (UB) or 1 ms (brief, regardless of electrode placement); stimulus duration— 6 s (UB) or 4 s (brief, although usually less for BL placement); frequency varies with age so that total charge was a product of patient age as described above; and current was always 800 mA. Higher charge was used in www.psychosomaticsjournal.org
273
Predictors of Electroconvulsive Therapy Postictal delirium patients taking anticonvulsants, such as benzodiazepines. At the time of this study, only patients who had experienced significant cognitive side effects during a prior ECT course were started on UB pulses. Accordingly, only 7.3% (N ¼ 7) of patients were treated with UB pulses for their first treatment. Also, only 12.5% (N ¼ 12) of patients commenced treatment with BL placement, being those who had previously required such after failing ECT with RUL placement. Patients were hyperventilated to achieve a minute ventilation of 20 L/min, which was maintained for 60 seconds. ECT was administered with a Mecta Spectrum 5000Q unit (Mecta Corp, Tualatin, Ore). Seizure length and postictal suppression were monitored by electroencephalography. Postictal suppression is rated manually on a 4-point Likert scale from 0–3.16 A score of 3 signifies excellent postictal suppression and a score of 0 indicates no postictal suppression. Outcome Measures The 2 dependent variables assessed were the following1: presence or absence of delirium as assessed by the CAMICU and sedation level as assessed by the RASS2. Because a key feature of delirium is the presence or absence of inattention, the RASS is often assessed in conjunction with the CAM-ICU, as at The Johns Hopkins ECT PACU. However, delirium is considered to be present based solely on whether or not the CAM-ICU is positive. CAM-ICU The Confusion-Assessment Method tool was developed by Inouye et al. for monitoring delirium and was modified by Ely et al. to assess delirium specifically in ICU patients.12,17 The CAM-ICU uses similar criteria as Diagnostic and Statistical Manual of Mental Disorders-IV (DSM-IV) for diagnosing delirium. It is a binary variable indicating the presence or absence of delirium. The CAM-ICU is considered positive for delirium if there is an acute change in mental status (noting there is no change in mental status in our patients until the ECT treatment), plus more than 2 errors in response to an auditory stimulus, plus either the RASS score being anything other than 0 or more than 1 error in response to questions assessing for disorganized thinking.12 The CAMICU can be conducted within 2–3 minutes and is an easy assessment tool for nurses in the PACU. It has been validated in studies from ICUs around the world.18–23 For example, Guenther et al. found high sensitivity (88–92%), specificity (100%), and interrater reliability (k 0.96) compared with the reference standard DSM-IV definition of delirium.21 274
www.psychosomaticsjournal.org
RASS The RASS is a 10-point sedation scale ranging from5 to þ4. Unlike other sedation tests such as the Ramsey Scale and the Sedation-Agitation Scale, the RASS includes responses to both verbal and physical stimuli, and the patient’s level of arousal is graded in part according to the stimulus strength.13 Negative scores indicate increasing levels of sedation, with 5 corresponding to an unarousable patient, whereas positive scores indicate increasing levels of agitation with þ4 corresponding to a combative patient. A score of þ2 was considered to be threshold for dangerous levels of agitation (defined as ‘‘a patient with frequent nonpurposeful movements, who fights his or her ventilator’’), so scores of þ2 and higher were grouped together, whereas scores below þ2 were grouped as nonhazardous. The RASS takes less than 20 seconds to perform and has also been shown to have high interrater reliability and criterion, construct, and face validity.13 ECT PACU nurses at Hopkins rate both measures. The CAM-ICU is administered during the patient’s third set of vital sign assessments, approximately 12–15 minutes after the patient’s arrival to the PACU. Two RASS scores are recorded during the patient’s stay in the PACU: the first reflects the patient’s worst agitation state in the PACU, whereas the second RASS score reflects the level of agitation or sedation on leaving the PACU. The RASS scores included in this study are the former of these 2 scores. Independent Variables Independent variables assessed were as follows:
Patient variables: age, weight, gender, diagnosis, and
number of concomitant psychotropic medications taken within 48 hours before the first treatment. We included consumption of the following psychotropic medications: neuroleptics, antidepressants, benzodiazepines, and other anticonvulsants plus lithium. Note that if patients were taking 2 antidepressant medications, they were considered to be taking 2 concomitant medications. ECT treatment variables: dosages of methohexital and succinylcholine, electrode placement, total charge (mC), seizure length, and postictal suppression. Statistical Analysis
The data analyses for this paper were generated using SAS software, Version 9.1.3 of the SAS System for Psychosomatics 55:3, May/June 2014
Reti et al. Windows. Bivariate relationships between independent and dependent variables were examined using the Chisquare test, Fisher exact test, or independent samples t-test. Results of bivariate analyses were used to build a multiple logistic regression model for each dependent variable. In addition, a graphical depiction was made to display the functional relationship between seizure length and probability of delirium. The level of significance for this study was set at a ¼ 0.05. A Bonferroni correction was used and calculated as a divided by the number of variables used in multiple comparisons to derive a cutoff value (0.05/11 ¼ 0.0045). P values resulting from statistical tests need to be below this value for a comparison to be considered more than a chance result. RESULTS In Table 1, we present baseline demographic details for patients included in the study. Summary statistics for both delirium assessment and dichotomized RASS scores are provided. Sixty-two patients (65.3%) were positive for delirium, whereas 33 patients (34.7%) did not have delirium. The average seizure length for patients without delirium was 59.2 seconds, whereas the average seizure length for patients with delirium was 85.7 seconds (p o 0.0001, by independent samples t-test). In Table 2, we show results for multiple logistic regression models. Seizure length was found to still be a statistically significant predictor of postictal delirium after adjusting for other covariates (p ¼ 0.003). No other variables were predictive. The relationship between seizure length and delirium status is displayed in Fig. 1. This figure is based on the unadjusted model and shows an increased likelihood of delirium with an increasing seizure length. For example, a 30 second seizure is associated with an approximately 30% chance of delirium; an 80 second seizure is associated with a greater than 70% chance of delirium. None of the covariates studied were predictive of dichotomized RASS score. In addition, a linear regression analysis using the RASS score as a continuous dependent variable also failed to identify significant predictors amongst the independent variables (data not shown). DISCUSSION We report a significant positive association between seizure length and probability of postictal delirium at Psychosomatics 55:3, May/June 2014
the first treatment after adjusting for other covariates. Patients who had a seizure of 80 seconds or more were very likely to be delirious in the PACU. These findings are consistent with observations that seizure length also influences the severity of postictal delirium amongst epileptics.24 It is not clear why longer electroconvulsive seizures might lead to an increased chance of postictal delirium. Longer seizures may reflect recruitment of a greater number of neurons into the seizure. In addition, or alternatively, the same number of neurons may fire for a longer time. Under either scenario, a longer seizure might be associated with higher neurotransmitter release. For example, it is known that inhibitory neurotransmitters such as gamma-aminobutyric acid and adenosine are released by seizure activity, which could act to transiently dampen normal neuronal activity post treatment, and this might be enhanced by a longer seizure.25,26 We did not find an effect of electrode placement on risk of postictal delirium. In fact, in our sample, a Fisher exact test shows a higher rate of postictal delirium in patients receiving RUL compared with patients receiving BL treatment (Table 1). However the effect of placement was not significant in the logistic regression model (Table 2). Although age is a risk factor for delirium in many medical settings,27,28 we did not find it to be a significant factor for development of postictal delirium, consistent with previous studies.2,9 Regarding gender, seizure length is longer for female patients at the first treatment compared with subsequent treatments.29 Nonetheless, we did not find that gender predicted postictal delirium, consistent with Kikuchi et al.2 We were surprised that delirium was not associated with the number of medications taken by the patient in the 48 hours leading up to the treatment. Additionally, exploratory analyses by medication class (data not shown) failed to show an association between consumption of neuroleptics, antidepressants, benzodiazepines, other anticonvulsants plus lithium and either delirium or RASS score. We also did not find an association between seizure length and consumption of these medication classes, including benzodiazepines which might have been due to tapering of benzodiazepines in at least two-thirds of patients consuming them in the 48 hours before ECT treatment. We also did not find an association between the dose of methohexital and postictal delirium. All patients are adequately anesthetized before charge administration and accordingly, it is not surprising that methohexital dose is not associated with postictal delirium. However, in www.psychosomaticsjournal.org
275
Predictors of Electroconvulsive Therapy Postictal delirium
Baseline Demographics. A Bonferroni Correction was Applied to Adjust for Multiple Statistical Tests; P values o 0.0045 were Interpreted as Statistically Significant
Table 1.
Age (y) Weight (Kg) Methohexital (mg/kg) Succinylcholine dosage (mg/kg) Charge (mC) Seizure length (s) Postictal suppression score
Total N ¼ 96
No delirium N ¼ 33
Delirium N ¼ 62
Mean (SD)
Mean (SD)
Mean (SD)
52.8 73.6 1.2 1.2 269.0 76.5 1.9
52.3 77.0 1.2 1.1 262.6 59.6 1.8
53.6 71.9 1.2 1.2 274.7 85.7 2.0
(17.9) (18.7) (0.3) (0.2) (124.8) (36.8) (0.7)
N (%)
(15.3) (18.3) (0.2) (0.2) (140.4) (20.1) (0.8)
N (%)
(19.1) (18.9) (0.3) (0.3) (116.3) (40.8) (0.6)
p-Value
No agitation (RASS o 2) N ¼ 77 Mean (SD)
0.74* 0.20* 0.38* 0.37* 0.66* o0.0001* 0.47*
53.8 73.8 1.2 1.1 275.0 77.9 2.0
N (%)
Agitation (RASS Z 2) N ¼ 17 Mean (SD)
(18.9) (17.6) (0.2) (0.2) (130.8) (38.4) (0.7)
N (%)
50.2 75.5 1.2 1.2 250.7 73.2 1.7
(12.8) (22.5) (0.2) (0.2) (96.6) (30.4) (0.7)
41 (43) 55 (57)
15 (45) 18 (55)
25 (40) 37 (60)
Diagnosis Major depressive disorder Bipolar disorder Schizoaffective disorder
70 (73) 21 (22) 4 (5)
23 (70) 10 (30) 0 (0)
46 (75) 11 (18) 4 (7)
Placement RUL BL with bitemporal placement
84 (88) 12 (12)
25 (76) 8 (24)
58 (94) 4 (6)
Number of medications 0 1 2 3 4
11 22 37 22 4
3 7 14 7 2
8 15 22 15 2
0.90y 33 (43) 44 (57)
7 (41) 10 (59)
57 (75) 17 (22) 2 (3)
11 (65) 4 (23) 2 (12)
68 (88) 9 (12)
14 (82) 3 (18)
0.19z
0.21z
0.02z
0.45z
0.89z (11) (23) (38) (23) (5)
(9) (21) (43) (21) (6)
(13) (24) (35) (24) (4)
0.45* 0.73* 0.89* 0.63* 0.47* 0.64* 0.19*
N (%)
0.63y
Sex Male Female
p-Value
0.26z 10 14 30 20 3
(13) (18) (39) (26) (4)
1 7 6 2 1
(6) (41) (35) (12) (6)
BL ¼ bilateral; RUL ¼ right unilateral; SD ¼ standard deviation. n
Independent samples t-test. y Chi-square test. z Fisher exact test.
settings where ECT is administered unmodified with no anesthetic agent, postictal delirium occurs more frequently.30,31 Likewise, succinylcholine dose was not associated with postictal delirium, consistent with Devanand et al.9 It is important to note that a positive CAM-ICU does not necessarily denote delirium of sufficient severity that it places the staff or patient in danger. Two-thirds of our patients were positive on the CAM-ICU, however only 18% (17 out of 94) of patients scored Z2 on the RASS indicating a dangerous level of agitation. These results suggest that not all delirious patients were hyperactive and dangerous to nurses; however, even if the delirium is hypoactive, it is nonetheless potentially dangerous to the patient possibly placing them at increased risk of falls and 276
www.psychosomaticsjournal.org
aspiration. Our data are comparable to the findings of the Kikuchi sample2 that reported 52% of patients experienced at least mild delirium post-ECT. Limitations Although we found an association between seizure length and probability of postictal delirium, we have not demonstrated causality. Another possible explanation for our finding is that seizure length is a marker for an unidentified factor that could cause postictal delirium. Despite our analysis, the role of concomitant medications is difficult to evaluate. For example, benzodiazepine withdrawal could predispose to both longer seizures and to postictal delirium. We also did not evaluate Psychosomatics 55:3, May/June 2014
Reti et al. Table 2.
Logistic Regression Models for Delirium and RASS (Z2). A Bonferroni Correction was Applied to Adjust for Multiple Statistical Tests; p Values o 0.0045 were Interpreted as Statistically Significant Delirium—adjusted model
RASS—adjusted model
OR
OR
95% Confidence interval
Estimate Lower
Upper
95% Confidence interval
p-Value Estimate Lower
Upper
p-Value
Number of medications Age Weight (kg) Charge Seizure length Postictal suppression Factor construct (brevital and succinylcholine)
1.19 1.05 1.00 1.00 1.05 0.87 1.78
0.68 1.00 0.96 0.99 1.02 0.40 0.87
2.10 1.11 1.03 1.00 1.08 1.86 3.65
0.54 0.052 0.89 0.22 0.003 0.71 0.12
0.77 0.98 1.01 1.00 0.99 0.60 1.09
0.41 0.93 0.97 0.99 0.97 0.25 0.47
1.45 1.03 1.05 1.01 1.01 1.43 2.51
0.42 0.41 0.69 0.80 0.38 0.24 0.84
Sex Male Female
0.84 0.26 Reference
3.65
0.76
0.92 0.25 Reference
3.44
0.90
Diagnosis Major depressive disorder Bipolar or shizoaffective disorder
2.21 0.70 Reference
7.00
0.18
0.50 0.15 Reference
1.70
0.26
Placement RUL BL
5.70 0.90 Reference
36.26
0.07
0.73 0.09 Reference
6.34
0.78
BL ¼ bilateral; OR ¼ odds ratio; RASS ¼ Richmond Agitation-Sedation Scale; RUL ¼ right unilateral.
nonpsychotropic drugs or medical, especially neurologic, comorbidities. Finally, we did not evaluate the adequacy of oxygenation during the treatment or in recovery, or the extent of hyperventilation which could affect seizure duration. To decrease seizure threshold, patients are hyperventilated before electrical stimulation which
Figure 1.
causes a transient respiratory alkalosis. The blood alkalization in turn causes vasoconstriction which affects cerebral perfusion and potentially the chances of delirium following the seizure. Our ability to determine the effect of charge dose on the association between seizure length and postictal
Probability of Delirium as a Function of Seizure Length. 1 0.9 0.8 0.7 0.6
Probability 0.5 of Delirium 0.4 0.3 0.2 0.1 0
0
50
100
150
200
250
300
Seizure Length
Psychosomatics 55:3, May/June 2014
www.psychosomaticsjournal.org
277
Predictors of Electroconvulsive Therapy Postictal delirium delirium was limited by use of age to determine charge dose rather than seizure threshold. Adverse cognitive effects of ECT are attributable to the magnitude of the stimulus dosage relative to seizure threshold rather than to the absolute electrical charge.32,33 Likewise, the effect of charge on seizure length is mediated by the ratio of charge to seizure threshold, with brief seizures occurring at seizure threshold and long seizures occurring just above seizure threshold. Once significantly above seizure threshold, seizure duration seems to vary inversely with charge.34,35 We have only presented outcome data pertaining to the first of a course of ECT treatments and have not analyzed subsequent treatments, which could be altered by changes made after the first treatment if the patient were delirious. Therefore, it is not known whether the relationship we found between seizure length and delirium would hold for subsequent treatments. Both seizure length and risk of delirium change as the treatment series progresses. As mentioned before, seizure length typically decreases after the first treatment,29 and risk of cognitive side effects increases as the treatment series progresses.36,37 These changes may alter the seizure length-delirium association. Accordingly, the implications that can be drawn from our data for postindex treatments are limited.
Another limitation of the study pertains to the outcome measures themselves. As far as we are aware, neither the CAM-ICU nor RASS have ever been evaluated in an ECT PACU, so we have no data about their validity, reliability, sensitivity, and specificity for monitoring postictal delirium. However, 2 recent meta-analyses of the CAM-ICU, from studies in ICUs, found specificity of 95.8%38 and 95.9%.39 Furthermore, even in settings outside the ICU, the CAM-ICU does have excellent specificity.40 In conclusion, we have shown that a long ECT seizure increases the likelihood of delirium in the PACU at the first treatment. This finding suggests PACU staff may benefit from knowledge about seizure length for predicting postictal delirium and anticipating the best management of post-ECT patients. Disclosure: No financial support was provided for this study. Dr Reti has received grant support from NIH and HDRF and study supplies at no cost from Neuronetics Inc. Dr Reti is also site PI on a multisite rTMS trial sponsored by Brainsway Inc. The other authors report no proprietary or commercial interest in any product mentioned or concept discussed in this article.
References 1. Sackeim HA, Decina P, Malitz S, Hopkins N, Yudofsky SC, Prohovnik I: Postictal excitement following bilateral and rightunilateral ECT. Am J Psychiatry 1983; 140:1367–1368 2. Kikuchi A, Yasui-Furukori N, Fujii A, Katagai H, Kaneko S: Identification of predictors of post-ictal delirium after electroconvulsive therapy. Psychiatry Clin Neurosci 2009; 63:180–185 3. Fink M: Post-ECT delirium. Cogn Ther 1993; 9:326–330 4. Katznelson R, Djaiani G, Tait G, et al: Hospital administrative database underestimates delirium rate after cardiac surgery. Can J Anaesth 2010; 57:898–902 5. Radtke FM, Franck M, MacGuill M, et al: Duration of fluid fasting and choice of analgesic are modifiable factors for early postoperative delirium. Eur J Anaesthesiol 2010; 27:411–416 6. Rao SS, Daly JW, Sewell DD: Falls associated with electroconvulsive therapy among the geriatric population: a case report. J ECT 2008; 24:173–175 7. Sobin C, Sackeim HA, Prudic J, et al: Predictors of retrograde amnesia following ECT. Am J Psychiatry 1995; 152:995–1001 8. Leechuy I, Abrams R: Postictal delirium (and recovery from melancholia) after left-unilateral ECT. Convuls Ther 1987; 3:65–68 9. Devanand DP, Bricose KM, Sackeim HM: Clinical features and predictors of postictal excitement. Convuls Ther 1989; 5:140–146 10. Sackeim HA, Prudic J, Devanand DP, et al: Prospective, randomized, double-blind comparison of bilateral and right unilateral electroconvulsive therapy at different stimulus intensities. Arch Gen Psychiatry 2000; 57:425–434
278
www.psychosomaticsjournal.org
11. Kellner CH, Tobias KG, Wiegand J: Electrode placement in electroconvulsive therapy (ECT): a review of the literature. J ECT 2010; 26:175–180 12. Ely EW, Margolin R, Francis J, et al: Evaluation of delirium in critically ill patients: validation of the confusion assessment method for the intensive care unit (CAM-ICU). Crit Care Med 2001; 29:1370–1379 13. Ely EW, Truman B, Shintani A, et al: Monitoring sedation status over time in ICU patients: reliability of the richmond agitationsedation scale(RASS). JAMA 2003; 289:2983–2991 14. Wachtel LE, Contrucci-Kuhn SA, Griffin M, et al: ECT for selfinjury in an autistic boy. Eur Child Adolesc Psychiatry 2009; 18:458–463 15. Wachtel LE, Griffin M, Reti IM: Electroconvulsive therapy in a man with autism experiencing severe depression, catatonia, and self-injury. J ECT 2010; 26:70–73 16. Nobler MS, Sackeim HA, Solomou M, Lube B, Devanand DP, Prudic J: EEG manifestations during ECT: effects of electrode placement and stimulus intensity. Biol Psychiatry 1993; 34:321–330 17. Inouye SK, van Dyck CH, Alessi CA, Balkin S, Siegal AP, Horwitz RI: Clarifying confusion: the confusion assessment method. a new method for detection of delirium. Ann Intern Med 1990; 113:941–948 18. Pun BT, Gordon SM, Peterson JF, et al: Large-scale implementation of sedation and delirium monitoring in the intensive
Psychosomatics 55:3, May/June 2014
Reti et al.
19.
20.
21.
22.
23.
24. 25.
26.
27. 28.
29.
30.
care unit: a report from two medical centers. Crit Care Med 2005; 33:1199–1205 Chuang WL, Lin CH, Hsu WC, Ting YJ, Lin KC, Ma SC: Evaluation of the reliability and validity of the Chinese version of the confusion assessment method for the intensive care unit. Hu Li Za Zhi 2007; 54:45–52 Van Eijk MM, van Marum RJ, Klign IA, de Wit N, Kesecioglu J, Slooter AJ: Comparison of delirium assessment tools in a mixed intensive care unit. Crit Care Med 2009; 37:1881–1885 Guenther U, Popp J, Koecher L, et al: Validity and reliability of the CAM-ICU flowsheet to diagnose delirium in surgical ICU patients. J Crit Care 2010; 25:144–151 L¨utz A, Heymann A, Radtke FM, Spies CD: Different assessment tools for intensive care unit delirium: which score to use? Crit Care Med 2010; 38:409–418 Tobar E, Romero C, Galleguillos T, et al: Confusion assessment method for diagnosing delirium in ICU patients (CAM-ICU): cultural adapation and validation of the Spanish version. Med Intensiva 2010; 34:4–13 Krauss G, Theodore WH: Treatment strategies in the postictal state. Epilepsy Behav 2010; 19:188–190 During MJ, Spencer DD: Adenosine: a potential mediator of seizure arrest and postictal refractoriness. Ann Neurol 1992; 32:618–624 Dunwiddie TV, Masino SA: The role and regulation of adenosine in the central nervous system. Annu Rev Neurosci 2001; 24:31–55 Inouye SK: Delirum in older persons. N Engl J Med 2006; 354:1157–1165 Peterson JF, Pun BT, Dittus RS, et al: Delirium and its motoric subtypes: a study of 614 critically ill patients. J Am Geriatr Soc 2006; 54:479–484 Rasimas JJ, Stevens SR, Rasmussen KG: Seizure length in electroconvulsive therapy as a function of age, sex, and treatment number. J ECT 2007; 23:14–16 Gallinek A: Fear and anxiety in the course of electroshock therapy. Am J Psychiatry 1956; 113:428–434
Psychosomatics 55:3, May/June 2014
31. Gallegos J, Vaidya P, D’Agati D, et al: Decreasing adverse outcomes of unmodified electroconvulsive therapy: suggestions and possibilities. J ECT 2012; 28:77–81 32. Sackeim HA, Prudic J, Devanand DP, et al: Effects of stimulus intensity and electrode placement on the efficacy and cognitive effects of electroconvulsive therapy. N Engl J Med 1993; 328:839–846 33. McCall WV, Reboussin DM, Weiner RD, Sackeim HA: Titrated moderately suprathreshold vs fixed high-dose right unilateral electroconvulsive therapy: acute antidepressant and cognitive effects. Arch Gen Psychiatry 2000; 57:438–444 34. Sackeim HA, Devanand DP, Prudic J: Stimulus intensity, seizure threshold, and seizure duration: impact on the efficacy and safety of electroconvulsive therapy. Psychiatr Clin North Am 1991; 14:803–843 35. Frey R, Heiden A, Scharfetter J, et al: Inverse relation between stimulus intensity and seizure duration: implications for ECT procedure. J ECT 2001; 17:102–108 36. Ingram A, Saling MM, Schweitzer I: Cognitive side effects of brief pulse electroconvulsive therapy: a review. J ECT 2008; 24:3–9 37. Sackeim HA, Prudic J, Fuller R: The cognitive effects of electroconvulsive therapy in community settings. Neuropsychopharmacology 2007; 32:244–254 38. Neto AS, Nassar Jr AP, Cardoso SO, et al: Delirium screening in critically ill patients: a systematic review and meta-analysis. Crit Care Med 2012; 40:1946–1951 39. Gusmao-Flores D, Figueira Salluh JI, Chalhub RA, Quarantini LC: The confusion assessment method for the intensive care unit (CAM-ICU) and intensive care delirium screening checklist (ICDSC) for the diagnosis of delirium: a systematic review and meta-analysis of clinical studies. Crit Care Med 2012; 16:R115 40. Neufeld KJ, Hayat MJ, Coughlin JM, et al: Evaluation of two intensive care delirium screening tools for non-critically ill hospitalized patients. Psychosomatics 2011; 52:133–140
www.psychosomaticsjournal.org
279