SciELO - Scientific Electronic Library Online

vol.21 número4Influence of affective personality on cognition-mediated emotional processing: Need for empowerment índice de autoresíndice de materiabúsqueda de artículos
Home Pagelista alfabética de revistas  

Servicios Personalizados



Links relacionados

  • En proceso de indezaciónCitado por Google
  • No hay articulos similaresSimilares en SciELO
  • En proceso de indezaciónSimilares en Google


The European Journal of Psychiatry

versión impresa ISSN 0213-6163

Eur. J. Psychiat. vol.21 no.4  oct./dic. 2007




Safe performance of ECT in severely ill patients: A retrospective study



A. Sartorius MD*, W. Hewer MD**,***

* Central Institute of Mental Health, Mannheim
** Department of Geriatric Psychiatry and Psychotherapy, Vinzenz von Paul Hospital, Rottweil
*** Central Institute of Mental Health, Mannheim. GERMANY





Background and Objectives: Electroconvulsive therapy (ECT) is nowadays known as a first line therapy in many certain illness conditions. Despite the fact that psychotic depression and treatment resistant depression are more common in geriatric psychiatry, the use of ECT is not. The aim of our study was to show that ECT can be safely performed even if patients show high comorbidity and are therefore per se at a higher risk for experiencing severe side effects.
Methods: We examined 25 ECT treated and severely ill patients of advanced age (mean 66 years) by chart review.
Results: Mean age corrected Charlson Comorbidity Index (CCI) was 4.1, mean Cumulative Illness Rating Scale for Geriatrics (CIRS-G) 10.5. Generally, ECT- related complications were rated as mild and short termed, 14 patients showed no complications at all. Complications did not correlate with age or comorbidity. Post hoc, we noted a significant advantage for the use of propofol or etomidate compared with thiopental as narcotic agents.
Conclusions: Under optimized somatic treatment conditions ECT can be performed safely in comorbid patients of advanced age. However, a risk/benefit analysis should always be performed individually.

Key words: Electroconvulsive therapy, Geriatric psychiatry, Medical comorbidity, Adverse effects.



Electroconvulsive therapy (ECT) is nowadays known as a first line therapy in many certain illness conditions. Among these, and occurring with high incidence in the elderly, are psychotic depression and treatment resistant depression. Despite the fact that psychotic depression and treatment resistant depression are more common in geriatric psychiatry, the use of ECT is not. This is somewhat surprising since medication side effects frequently limit therapeutic approaches in geriatrics and ECT efficacy increases with age. The widespread belief that ECT cannot be performed safely or at least with relative safety in this patient group might partially explain this phenomenon. Higher comorbidity or even multimorbidity underlines the problem since in such cases not only psychopharmacology but also ECT has to be done with great care.

There is a body of evidence that ECT can be performed safely in highly comorbid or old patients, even in the old-old1-12. Nevertheless, to the best of our knowledge, there are only a few retrospective studies dealing with efficacy and safety of highly comorbid patients or geriatric patients with at least some relevant comorbidity.

Tew et al. investigated efficacy and clinical characteristics of ECT in 268 patients11. Adult patients (59 years and younger) showed a lower ECT response rate compared with young-olds and old-olds (older than 75 years). Despite a higher level of comorbidity and cognitive impairment in older patients no differences in safety of ECT was observed.

Philibert et al. investigated 192 geriatric patients within the framework of a retrospective study13. 108 of them had undergone ECT, and showed greater clinical improvement and were more likely to be alive at follow-up.

Gormley et al. examined 93 ECT-courses in patients older than 75 years14. Ten patients suffered from complications (most common were hypomania and confusion) which all ceased within 2 weeks. The response rate in this cohort was 85 %.

Zisselman et al. reported 13 geriatric patients with high comorbidity15, which was quantified by the Cumulative Illness Rating Scale for Geriatrics (CIRS-G)16. Of these patients nine were rated as improved and complications were limited to atrial fibrillation in one patient and delirium in another. Despite an average age of 81 years and a mean CIRS-G of almost 20, cognitive impairment was stated in only five patients and was transient in all of them.

The study by Zisselman is limited by the small number of participants. Its strength lies in a well described cohort of psychogeriatric patients with high comorbidity all of which were known personally to the first author.

To assess the problem of quantifying somatic comorbidity we decided to rate the patients' comorbidity using two scales: The Charlson Comorbidity Index (CCI) including age related risk points17, and the Cumulative Illness Rating Scale for Geriatrics (CIRS-G)16. Comorbidity rating scales and indices can be used to quantify the burden of current and chronic illnesses in the older adult. CCI and CIRS have shown to be a valid predictor of risk of mortality in longitudinal studies.

With our naturalistic retrospective study we would like to add a cohort of 25 severely ill patients with either old age and/or high comorbidity undergoing ECT. Our retrospective study was conducted to prove the concept of safe ECT performance even under conditions of high comorbidity. To achieve this goal we tried to optimize pre and concurrent ECT treatment at our site taking into account anesthesiological and internal medicine problems.


Subjects and methods

We examined the records of 25 patients with high age and/or high comorbidity who were treated with ECT in the Department of Psychiatry, Central Institute of Mental Health, Mannheim between 1988 and 1999. Within this period a total of 162 patients completed an ECT course. Patients in a severe or life threatening condition (for psychiatric or somatic reasons) before onset of ECT were included in our study. The majority of our cohort (88%) was treated on the psychiatric intensive care unit of the Central Institute of Mental Health, a unit which is specialized in the treatment of psychiatric patients with severe medical comorbidity. The ward is headed by a physician qualified in psychiatry and internal medicine (W.H.). Likewise, the nursing staff has special qualifications and is upsized. The psychiatric intensive care unit includes a specialized emergency room where all ECT sessions are performed. Equipment for diagnostic workup and treatment monitoring with regard to medical comorbidity is available (ECG, Holter monitoring, long term blood pressure recording, ultrasound, including echocardiography, EEG, X-ray equipment, including CT-scanning, clinical chemistry and hematology, on a 24-h basis, if necessary, and monitoring devices for cardiovascular and respiratory function). The head of the unit was responsible for conducting all ECT and internal medicine treatments of all patients. Prior to ECT patients´ diagnostic workup and treatment with regard to comorbid general medical disorders were optimized with greatest care. Particular attention was paid to concurrent medication which was reduced to an essential minimum to avoid interactions and additional side effects. Until 1989 ECT was conducted with a Siemens Convulsator (only the first 2 patients), afterwards with a Thymatron stimulator with short pulse technique (1989-1999 Thymatron DG, Somatics, Inc.). In most of the cases the amount of the released charge of the ECT device was not documented and therefore not implemented into the study.

Comorbidity was rated for all patients with the CCI including age related risk points17 and additionally with the CIRS-G16. Relevant comorbidity is visualized in table I. Basically, the CIRS-G assesses a total of 13 items (each coded with '0' = "none", '1' = "mild", '2' = "moderate", '3' = "severe", '4' = "life threatening"). These include items from the cardiovascular-respiratory system (4 items), gastrointestinal system (3 items), genitourinary system (2 items), musculo-skeletal-integumentary system (1 item), neuropsychiatric system (2 items), and general system (endocrine/metabolic) (1 item). The CCI accumulates the following items (1 point each): myocardial infarct, congestive heart failure, peripheral vascular disease, cerebrovascular disease (except hemiplegia), dementia, chronic pulmonary disease, connective tissue disease, ulcer disease, mild liver disease, and diabetes (without complications). 2 points are added for diabetes with end organ damage, hemiplegia, moderate or severe renal disease, second solid tumor (non metastatic), leukemia, lymphoma, metastases, 3 points for moderate or severe liver disease and 6 points for a 2nd metastatic solid tumor or AIDS.

Onset of complications during ECT course was assessed as follows (analogue to CIRS-G):

'0' = "none", '1' = "mild", '2' = "moderate", '3' = "severe", '4' = "life threatening". Very short termed cardiac arrhythmias such as postictal bigeminy or compensatory sinus bradycardias were not stated as complications (and had occasionally been probably ignored or not documented). Urgency of ECT indication was stated as "acute" = '1' and "subacute" = '0'. ECT treatment itself was classified into right unilateral (RUL) and bilateral (BIL) treatment. For ECT narcosis we used thiopental, etomidate or propofol. Number of individual ECT sessions and duration of the complete ECT course were noted. ECT treatment success was clinically defined as "remission" = '2', "response" = '1', or "non-response" = '0'. Patients´ primary psychiatric illness was additionally described (as far as possible from chart review) with years since first diagnosis, duration of the current episode, and number of former episodes. Number of concurrent psychiatric and non-psychiatric drugs were noted as well as number of psychiatric and non-psychiatric drugs taken 7 days before the initiation of the ECT course. All statistical analyses were calculated with STATA (version 9.0).



Mean age of our cohort was 66.4 (+/-12) years, the youngest patient was 38 years, the oldest 88 years of age. 6 patients were male. Our sample of 25 patients showed a mean age corrected Charlson Comorbidity Index of 4.1 and a Cumulative Illness Rating Scale for Geriatrics of 10.5. Both indices correlated well F (1,23) = 28.9, r2 = 0.56, corr. coeff. = 0.65, p < 0.001). 14 patients experienced no complications at all, and mean complications were rated as mild (= 0.8), with no significant difference for older subgroups (e.g. age >70 years, n = 10, mean = 1.0 ± 0.9, range 0-2) or for the old-old patients (age >75 years, n = 6, mean = 1.1 ± 1.0, range 0-2). (Table 2)

There was no correlation between complications and age corrected Charlson Comorbidity Index or the Cumulative Illness Rating Scale for Geriatrics (F (1,23) = 0.14, r2 = 0.006, p = 0.71, and F (1,23) = 0.001, r2 = 0.00, p = 0.99, respectively).

In 14 cases indication for ECT was acute. In all, 14 patients remitted, with eleven showing a response to ECT treatment. Five patients were initially treated bilaterally, another five were switched during the course to bilateral treatment. Six patients received etomidate as narcotic agent, 3 propofol and 10 thiopental, another 4 switched during the course (for two cases the narcotic agent was not documented).

There was also no correlation between age and remission rates, as well as no correlation between age and complications (F (1,23) = 0.52, r2 = 0.02, p = 0.48, and F(1,23) = 2.25, r2 = 0.09, p = 0.15, respectively).

A multiple regression analysis including degree of complications and remission as dependent variables revealed only one significant result: Remission rates were positively correlated with an increasing number of concurrent psychiatric medications. A post hoc regression for remission and the number of concurrent psychiatric drugs confirmed the significant correlation (F (1,20) = 6.23, r2 = 0.24, correlation coefficient = 0.28, p = 0.02.). Other independent variables (duration of acute episode, number of concurrent non-psychiatric drugs, number of concurrent psychiatric and non-psychiatric drugs taken 7 days before) showed no significances. Including diagnosis or subsyndromes (e.g. psychotic symptoms) in the regression analysis did not influence any of the above results.

Post hoc, a regression analysis for narcotic agents showed a significant impact on complications: All 9 patients receiving propofol or etomidate showed no complications at all, whereas 10 patients treated with thiopental had an average degree of complications of 1.4 ± 1.1 (range 0-3). Comparing complications of thiopental treated patients with patients treated with etomidate or propofol reveals p = 0.002 (two-sided unpaired t-test). All 4 patients switching during the course were changed from thiopental to propofol or etomidate.

Remission rates were slightly higher in patients treated with bilateral ECT (p = 0.1, one-sided t-test). Complication rates did not differ between RUL and BIL treatment.


Illustrative example

Patient #19 (see Table I) may serve as an illustrative example: This 70 year old female patient had been suffering from a bipolar affective disorder for the last 51 years. She also had been diagnosed with plasmacytoma (or multiple myeloma) two years earlier. Three weeks before admission she became abasic. At admission she showed symptoms of a severe depressive episode with psychotic features, but without suicidality. She presented with a transverse spinal cord syndrome (T10) with severe paraparesis (initially atonic and later spastic), as well as hypaesthesia and hypalgesia of both legs. Total spine NMR revealed a spinal cord compression (spinal canal diameter of 7 mm). By biopsy the tumor was diagnosed as a plasmacytomic tumor. Hemoglobin concentration was below 10g/dl. Electrophysiology revealed a severe axonal motor polyneuropathy, probably caused by a critical illness syndrome. Because our patient refused food and medication we decided to start an ECT course taking the vertebral fracture risk into account. In parallel radiation of the vertebral tumor was initiated. All of the 12 ECT sessions were well tolerated, no side effects occurred with carefully adapted muscle relaxation. Psychopathology initially improved, but a psychotic decompensation occurred while dexamethasone and morphine were introduced. Reducing the doses of dexamethasone and morphine finally resulted in a remission of the depressive episode and our patient was discharged with a medication of valproate, haloperidol and nortriptyline.



Both, age corrected CCI and CIRS-G are well validated scales for comorbidity18. In our cohort a highly significant correlation between both measures was shown.

ECT was performed safely independent of age, age corrected CCI and CIRS-G. In a patient group with high comorbidity and/or advanced age only short termed side effects occurred.

Compared with a study by von Ammon Cavanaugh19 the mean CCI in our cohort was high. In her study of 241 patients hospitalized with MDE 20 died while in hospital. This subgroup had a mean CCI of 5.4 ± 2.6, while discharged patients had a CCI of 3.1 ± 2.0. Our patients (60% diagnosed with MDE) suffered from illnesses amounting to a mean CCI of 4.1, indicating the severity of comorbidity.

Surprisingly, the difference in complication rates of thiopental and etomidate/propofol anesthesia was striking. It is known that sinus bradycardia and premature ventricular contractions appear more often with thiopental even when compared with methohexital. Additionally, middle cerebral artery flow velocities and arterial pressure increase with thiopental compared with propofol20,21. However, it is difficult to explain the generally higher complication rate by these properties. Nevertheless, any clinical significant cognitive deterioration and all cardiac arrhythmias were observed in thiopental treated patients. There is a great lack of clinical data comparing side effects or complications with thiopental /methohexital versus propofol/etomidate in older patients22. From our patient group it can be suggested that if induction of seizure is not problematic propofol should be preferred. If induction of seizures is a problem, however, etomidate may be the anaesthetic of first choice.

Data on concurrent use of psychotropic medication are heterogeneous and at times contradictory (e.g. for lithium23). In general, additional medication leads to more interaction and side effects. It is widely accepted that patients suffering from a schizophrenic disorder can benefit from concurrent psychotropic drugs compared with ECT alone24. Evidence showing an advantage for concurrent medication in patients with affective disorders is less extensive25,26. In our study concurrent psychiatric medication (in most cases 1 benzodiazepine and 1 antidepressant or 1 antipsychotic) was associated with significantly higher remission rates. Additionally, we would like to emphasize that our patients did not suffer from more complications when taking a greater number of psychotropic or non-psychotropic drugs (with a mean of 1.7 and 4.0 drugs per patient, respectively). This may be interpreted as a probably u-shaped complication rate over number of medication curve. A low but well chosen number of drugs is necessary to lower risks in multimorbid old patients while a further increase of prescribed medication again rises the probability of interactions and side effects. This probably holds true for both, psychiatric medication as well as non-psychiatric medication. In our opinion great care must be taken to optimize individual medication. At least in our patients this proved very benefical, since our data suggest that they were neither over- nor undermedicated.

Cardiac arrhythmias were rare and characterized as short termed and not serious. In three cases arrhythmia was stated as mild, moderate and severe (ventricular tachycardia in one case), respectively. This is in good agreement with recent ECT investigations that included patients with cardiac risk factors4,8,27. Mild cardiac arrhythmias were seen in 10-50% of all patients, but severe side effects in just 1.5-20%. It is noteworthy that in the study of Zielinski et al. severe complications were found in 20% of all patients, but ECT was still much better tolerated than drugs. In his study tricyclic antidepressants had been discontinued (due to intolerable side effects) in patients who then completed an ECT course. In the naturalistic study of Agelink et al. only 3 patients showed minor complications, which was attributed to an optimized somatic pretreatment of patients at cardiac risk.

Distinguishing between RUL and BIL treatment did not reveal significant findings in our cohort. On the one hand BIL was more efficient (not reaching the significance level, however), which is not surprising28. On the other hand our sample size is too small for comparison owing to the fact that stimulation energy was not standardized (e.g. by seizure threshold titration). RUL and BIL treatment did not differ regarding clinically significant cognitive side effects. Most studies dealing with MDE prefer high dose RUL to BIL because of comparable efficacy and better cognitive outcome29, 31. Thus our results do not contradict findings of other studies dealing with RUL/BIL efficacy and side effects, but are most likely due to non standardization of stimulation energy and other confounds in our relatively small sample size.



Optimized somatic diagnostics and pre-treatment made it possible to safely perform ECT in 25 severely ill patients with high age and/or high comorbidity. Only short termed complications occurred. Neither age nor comorbidity correlated with number and severity of complications. Compared to thiopental propofol and etomidate may be preferable as narcotic agents.

Interpretation is also limited by the naturalistic design of our study and by the small sample size. Within this context registration of very short termed and minor events was probably not always sufficient. Furthermore specific interpretations are limited by the heterogeneity of psychiatric and differentiated comorbid diagnoses in our cohort.



1. Christopher EJ. Electroconvulsive therapy in the medically ill. Curr Psychiatry Rep 2003; 5(3): 225-230.        [ Links ]

2. Greenberg RM. ECT in the elderly. New Dir Ment Health Serv 1997; (76): 85-96.        [ Links ]

3. Casey DA, Davis MH. Electroconvulsive therapy in the very old. Gen Hosp Psychiatry 1996; 18(6): 436-439.        [ Links ]

4. Rice EH, Sombrotto LB, Markowitz JC, Leon AC. Cardiovascular morbidity in high-risk patients during ECT. Am J Psychiatry 1994; 151(11): 1637-1641.        [ Links ]

5. Mulsant BH, Rosen J, Thornton JE, Zubenko GS. A prospective naturalistic study of electroconvulsive therapy in late-life depression. J Geriatr Psychiatry Neurol 1991; 4(1): 3-13.        [ Links ]

6. Rasmussen KG, Rummans TA, Richardson JW. Electroconvulsive therapy in the medically ill. Psychiatr Clin North Am 2002; 25(1): 177-193.        [ Links ]

7. Kelly KG, Zisselman M. Update on electroconvulsive therapy (ECT) in older adults. J Am Geriatr Soc 2000; 48(5): 560-566.        [ Links ]

8. Agelink MW, Dammers S, Malessa R, Leonhardt M, Zitzelsberger A, Ullrich H et al. [Benefits and risks of electroconvulsive therapy (ECG) in elderly patients with cardiovascular risk factors]. Nervenarzt 1998; 69(1): 70-75.        [ Links ]

9. Hay DP. Electroconvulsive Therapy in the Medically Ill Elderly. Convuls Ther 1989; 5(1): 8-16.        [ Links ]

10. Manly DT, Oakley SP, Jr., Bloch RM. Electroconvulsive therapy in old-old patients. Am J Geriatr Psychiatry 2000; 8(3): 232-236.        [ Links ]

11. Tew JD, Jr., Mulsant BH, Haskett RF, Prudic J, Thase ME, Crowe RR et al. Acute efficacy of ECT in the treatment of major depression in the old-old. Am J Psychiatry 1999; 156(12): 1865-1870.        [ Links ]

12. Flint AJ, Gagnon N. Effective use of electroconvulsive therapy in late-life depression. Can J Psychiatry 2002; 47(8): 734-741.        [ Links ]

13. Philibert RA, Richards L, Lynch CF, Winokur G. Effect of ECT on mortality and clinical outcome in geriatric unipolar depression. J Clin Psychiatry 1995; 56(9): 390-394.        [ Links ]

14. Gormley N, Cullen C, Walters L, Philpot M, Lawlor B. The safety and efficacy of electroconvulsive therapy in patients over age 75. Int J Geriatr Psychiatry 1998; 13(12): 871-874.        [ Links ]

15. Zisselman MH, Kelly KG, Cutillo-Schmitter T, Payne D, Denman SJ. Successful ECT in long-term care residents. J Am Med Dir Assoc 2001; 2(1): 22-25.        [ Links ]

16. Miller MD, Paradis CF, Houck PR, Mazumdar S, Stack JA, Rifai AH et al. Rating chronic medical illness burden in geropsychiatric practice and research: application of the Cumulative Illness Rating Scale. Psychiatry Res 1992; 41(3): 237-248.        [ Links ]

17. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987; 40(5): 373-383.        [ Links ]

18. de Groot V, Beckerman H, Lankhorst GJ, Bouter LM. How to measure comorbidity. a critical review of available methods. J Clin Epidemiol 2003; 56(3): 221-229.        [ Links ]

19. von Ammon CS, Furlanetto LM, Creech SD, Powell LH. Medical illness, past depression, and present depression: a predictive triad for in-hospital mortality. Am J Psychiatry 2001; 158(1): 43-48.        [ Links ]

20. Ding Z, White PF. Anesthesia for electroconvulsive therapy. Anesth Analg 2002; 94(5): 1351-1364.        [ Links ]

21. Scheepstra GL, Booij LH, de Lange JJ. Influence of equipotent doses propofol and thiopentone on arterial pressure and heart rate at induction of anesthesia. A comparison between young and old patients. Acta Anaesthesiol Belg 1989; 40(3): 167-174.        [ Links ]

22. Eilers H, Niemann C. Clinically important drug interactions with intravenous anaesthetics in older patients. Drugs Aging 2003; 20(13): 969-980.        [ Links ]

23. Sartorius A, Wolf J, Henn FA. Lithium and ECT--concurrent use still demands attention: three case reports. World J Biol Psychiatry 2005; 6(2): 121-124.        [ Links ]

24. Braga RJ, Petrides G. The combined use of electroconvulsive therapy and antipsychotics in patients with schizophrenia J ECT 2005; 21(2): 75-83.        [ Links ]

25. Weiner RD. The Practice of Electroconvulsive Therapy: Recommendations for Treatment, Training, and Privileging: A Task Force Report of the American Psychiatric Association. 2nd ed. Washington, DC: American Psychiatric Association, 2001.        [ Links ]

26. Baghai TC, Marcuse A, Brosch M, Schule C, Eser D, Nothdurfter C et al. The influence of concomitant antidepressant medication on safety, tolerability and clinical effectiveness of electroconvulsive therapy. World J Biol Psychiatry 2006; 7(2): 82-90.        [ Links ]

27. Zielinski RJ, Roose SP, Devanand DP, Woodring S, Sackeim HA. Cardiovascular complications of ECT in depressed patients with cardiac disease. Am J Psychiatry 1993; 150(6): 904-909.        [ Links ]

28. Petrides G, Fink M, Husain MM, Knapp RG, Rush AJ, Mueller M et al. ECT remission rates in psychotic versus nonpsychotic depressed patients: a report from CORE. J ECT 2001; 17(4): 244-253.        [ Links ]

29. Tew JD, Jr., Mulsant BH, Haskett RF, Dolata D, Hixson L, Mann JJ. A randomized comparison of high-charge right unilateral electroconvulsive therapy and bilateral electroconvulsive therapy in older depressed patients who failed to respond to 5 to 8 moderate-charge right unilateral treatments. J Clin Psychiatry 2002; 63(12): 1102-1105.        [ Links ]

30. Sackeim HA, Prudic J, Devanand DP, Nobler MS, Lisanby SH, Peyser S et al. A prospective, randomized, double-blind comparison of bilateral and right unilateral electroconvulsive therapy at different stimulus intensities. Arch Gen Psychiatry 2000; 57(5): 425-434.        [ Links ]

31. Lisanby SH, Maddox JH, Prudic J, Devanand DP, Sackeim HA. The effects of electroconvulsive therapy on memory of autobiographical and public events. Arch Gen Psychiatry 2000; 57(6): 581-590.        [ Links ]



M.D. A. Sartorius
Central Institute of Mental Health, J5, 68159 Mannheim, Germany
Phone: +49-621-1703-2722, fax: +49-621-1703-3165

M.D W. Hewer
Department of Geriatric Psychiatry and Psychotherapy,
Vinzenz von Paul Hospital, Rottweil, Germany.

Received 19 July 2006
Revised 24 January 2007
Accepted 8 March 2007

Creative Commons License Todo el contenido de esta revista, excepto dónde está identificado, está bajo una Licencia Creative Commons