- Email: [email protected]
Cardiopulmonary Resuscitation in Critically Ill Neurologic-Neurosurgical Patients
ORIGINAL ARTICLE
CPR IN NEUROLOGIC-NEUROSURGICAL PATIENTS
Cardiopulmonary Resuscitation in Critically Ill Neurologic-Neurosurgical Patients ALEJANDRO A. RABINSTEIN, MD; ROBYN L. MCCLELLAND, PHD; EELCO F. M. WIJDICKS, MD; EDWARD M. MANNO, MD; AND JOHN L. D. ATKINSON, MD
OBJECTIVES: To establish the rate of successful cardiopulmonary resuscitation (CPR) and to study outcome predictors in patients who experienced in-hospital cardiac arrest after being admitted to the neurologic-neurosurgical intensive care unit (ICU) with a primary neurologic diagnosis. PATIENTS AND METHODS: We identified patients admitted to the neurologic-neurosurgical ICU between 1994 and 2001 who experienced in-hospital cardiac arrest and received CPR. Functional outcome was assessed using the modified Rankin scale. RESULTS: During the study period, 38 consecutive patients experienced in-hospital cardiac arrest and received CPR. The median age of the patients was 65 years (range, 16-81 years), and the mean interval from admission to CPR was 12 days (range, 3 hours to 47 days). Acute intracranial disease was present in 32 patients (84%). Twenty-one patients (55%) were in the ICU at the time of the cardiac arrest; cardiac arrests in the wards occurred at a mean interval of 9 days (range, 1-45 days) after ICU discharge. Cardiopulmonary resuscitation achieved return of spontaneous circulation in 23 patients (61%). Seven patients (18%) were discharged from the hospital, 5 of whom later achieved a modified Rankin scale score of 2 or lower. Cardiac arrest after a deteriorating clinical course resulted in uniformly fatal outcomes. Duration of CPR shorter than 5 minutes and CPR in the ICU were associated with survival and good functional recovery. CONCLUSIONS: Cardiopulmonary resuscitation is a worthwhile procedure in severely ill neurologic-neurosurgical patients, regardless of the patient’s age. However, the outcome after CPR appears much worse in patients with a prior deteriorating clinical course.
Mayo Clin Proc. 2004;79(11):1391-1395 CPR = cardiopulmonary resuscitation; ICU = intensive care unit
C
ardiopulmonary resuscitation (CPR) is practiced throughout the entire hospital. However, studies on the validity of CPR for in-hospital cardiac arrest have included mixed populations, making it difficult to ferret out the appropriateness of CPR in the neurologicneurosurgical intensive care unit (ICU).1 Some studies have shown that CPR is ineffective in certain groups of
From the Department of Neurology (A.A.R., E.F.M.W., E.M.M.), Division of Biostatistics (R.L.M), and Department of Neurologic Surgery (J.L.D.A.), Mayo Clinic College of Medicine, Rochester, Minn. Dr Rabinstein is now with the University of Miami School of Medicine, Miami, Fla. Address reprint requests and correspondence to Eelco F. M. Wijdicks, MD, Department of Neurology, Mayo Clinic College of Medicine, 200 First St SW, Rochester, MN 55905 (e-mail: [email protected]). © 2004 Mayo Foundation for Medical Education and Research
Mayo Clin Proc.
•
patients, such as those with advanced cancer,2 refractory sepsis or severe pneumonia,3,4 renal failure,5 greater severity of acute illness,6,7 and poor preresuscitation functional status.8 Surprisingly, we found no such information on prognosis after CPR in critically ill neurologic and neurosurgical patients. Commonly, CPR is discussed in the neurologic-neurosurgical ICU, where, perhaps more than in any other ICU setting, families are concerned about not only the chances of survival but also functional outcome. Reasoned judgment is particularly necessary for deteriorating patients in whom meaningful neurologic recovery after CPR is considered unlikely. Therefore, information is needed on what factors influence survival and functional recovery after CPR in patients with critical neurologic and neurosurgical illness. The goals of this study were to define the rate of success of CPR in patients admitted to the neurologicneurosurgical ICU with a primary neurologic diagnosis who experienced in-hospital cardiac arrest and to study predictors of outcome in this patient population. PATIENTS AND METHODS We identified all patients admitted to the neurologic-neurosurgical ICU at the Mayo Clinic in Rochester, Minn, between January 1994 and December 2001 who experienced cardiac arrest during their hospital stay. All episodes of CPR in the hospital are thoroughly recorded and can be readily retrieved from our linkage system. In our institution, all episodes of CPR are reviewed in detail by a dedicated committee, and documentation of the interventions is standardized across the hospital. Cardiac arrest was defined by the presence of a hemodynamically unstable heart rhythm, including ventricular tachycardia or profound bradycardia with evidence of cerebral hypoperfusion, ventricular fibrillation, pulseless electrical activity, or cardiac asystole. We excluded patients with respiratory arrest not associated with cardiac arrest given their much more benign prognosis,9 patients who had seizures without unstable cardiac arrhythmia, patients who had hypotensive episodes without cardiac arrhythmia, and those who had episodes of cardiac arrhythmia without hemodynamic compromise.
November 2004;79(11):1391-1395
•
www.mayo.edu/proceedings
For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.
1391
CPR IN NEUROLOGIC-NEUROSURGICAL PATIENTS
TABLE 1. Primary Admission Diagnoses of 38 Consecutive Critically Ill Neurologic-Neurosurgical Patients Who Experienced In-hospital Cardiac Arrest Diagnosis
No. at risk
No. (%) who survived
Head trauma* Subarachnoid hemorrhage† Ischemic stroke Spinal disease Intracranial tumor Intracerebral hemorrhage Meningoencephalitis Status epilepticus Chronic subdural hematoma
8 6 5 5 4 3 3 2 2
2 (25) 2 (33) 1 (20) 1 (20) 1 (25) 0 (0) 0 (0) 0 (0) 0 (0)
*All patients with head trauma were admitted with a Glasgow Coma Scale sum score lower than 8, and there were no significant differences on Glasgow Coma Scale sum score at the time of cardiac arrest between survivors and nonsurvivors. †World Federation of Neurological Surgeons grades on admission in patients who did not survive were I in 2 patients, II in 1 patient, and III in 1 patient; admission grade was II in both survivors.
We recorded patients’ demographic information, medical and surgical history, reason for hospital and ICU admission, complications during the hospital course, and time of the cardiac arrest. Cause of cardiac arrest, initial heart rhythm at the time of onset of the resuscitation efforts, and duration of such efforts were noted. We categorized the arrests into unexpected or anticipated. We defined a cardiac arrest as anticipated when it was preceded by a gradual deterioration of the patient’s clinical status that was unresponsive to aggressive medical or surgical interventions. This deterioration could be due to worsening of the primary neurologic disease or development of neurologic or systemic complications. Outcome of survivors was assessed using the modified Rankin scale at the time of last followup. A modified Rankin scale score of 2 or lower (slight or no disability and preserved ability to look after own affairs TABLE 2. Causes of Cardiac Arrest in 38 Consecutive Critically Ill Neurologic-Neurosurgical Patients According to Outcome Cause
No. at risk
No. (%) who survived
Aspiration Myocardial ischemia Cardiac arrhythmia End-stage herniation Pulmonary embolism Sepsis Seizures Airway obstruction* Gastrointestinal bleeding Unclear
6 5 5 5 4 3 3 2 1 4
0 (0) 1 (20) 2 (40) 0 (0) 0 (0) 1 (33) 1 (33) 2 (100) 0 (0) 0 (0)
*We included patients with acute airway obstruction (obstruction of tracheostomy in 1 patient and upper airway obstruction after extubation in 1 patient) that led to cardiac arrest but excluded those with pure respiratory arrest.
1392
Mayo Clin Proc.
•
without assistance) was considered indicative of good functional outcome. Outcome predictors were compared using the Fisher exact test for nominal variables and the independent 2tailed t test for continuous variables, with modification for unequal variances. Age-adjusted evaluation of associations was performed using exact logistic regression models.10,11 Level of statistical significance was established at P<.05. RESULTS We identified 38 consecutive patients admitted to our neurologic-neurosurgical ICU during the study period who had experienced in-hospital cardiac arrest and received CPR. The median age of the patients was 65 years (range, 16-81 years), and 66% were men. Comorbidities were frequent, including smoking in 21 patients (55%); hypertension in 17 (45%); coronary artery disease in 14 (37%); hyperlipidemia in 9 (24%); previous stroke, diabetes mellitus, chronic obstructive pulmonary disease, and peptic ulcer disease in 7 each (18%); and history of cardiac arrhythmia in 6 (16%). Acute intracranial disease was present in 32 patients (84%). Primary admission diagnoses are listed in Table 1. Cardiac arrest occurred at a mean interval of 12 days after hospital admission (range, 3 hours to 47 days). Twenty-one patients (55%) were in the ICU at the time of the cardiac arrest; among those who experienced cardiac arrest in the wards, the mean interval between ICU discharge and cardiac arrest was 9 days (range, 1-45 days). The causes of cardiac arrest are presented in Table 2. The initial heart rhythms were asystole in 9 patients, bradycardia in 9 patients, and pulseless electrical activity, ventricular fibrillation, and unstable ventricular tachycardia in 7 patients each. Cardiac arrests were unexpected in 26 patients (68%). The median duration of the resuscitation efforts was 17 minutes (range, 2-60 minutes). Cardiopulmonary resuscitation achieved return of spontaneous circulation in 23 patients (61%). Of these 23 patients, 16 subsequently died of a recurrent cardiac arrest (5 patients) or had progression to brain death (11 patients). Seven patients (18% of our total study population) survived to be discharged from the hospital: 2 were discharged home, 2 to a rehabilitation facility, and 3 to nursing homes. At the time of last follow-up (mean follow-up period, 9 months; median, 7.5 months; range, 1-24 months), 5 (71%) of the 7 survivors had achieved a good functional recovery (modified Rankin scale score, ≤2). We analyzed predictive variables for survival in our study population (Tables 1, 2, and 3 and Figure 1). Surviving patients were older than those who died (mean ± SD age, 70±9 years vs 58±19 years; P=.03). There were no
November 2004;79(11):1391-1395
•
www.mayo.edu/proceedings
For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.
CPR IN NEUROLOGIC-NEUROSURGICAL PATIENTS
TABLE 3. Characteristics of In-hospital Cardiac Arrest and Resuscitation Efforts in 38 Consecutive Critically Ill Neurologic-Neurosurgical Patients* Variable Cardiac arrest type Unanticipated Anticipated Location ICU Ward Initial rhythm Asystole Bradycardia PEA VF Unstable tachycardia CPR duration (min) ≤5 >5
No. at risk
No. (%) who survived†
26 12
7 (27) 0 (0)
21 17
7 (33) 0 (0)
9 9 7 7 7
0 (0) 2 (22) 2 (29) 1 (14) 2 (29)
11 27
6 (55) 1 (4)
P value .07 .01‡ .54
.001‡
*CPR = cardiopulmonary resuscitation; ICU = intensive care unit; PEA = pulseless electrical activity; VF = ventricular fibrillation. †At the time of last follow-up. ‡Fisher exact test.
was precipitated by aspiration, pulmonary embolism, or brain herniation died in the hospital. Patients with cardiac arrest anticipated by a deteriorating clinical course (n=12) had uniformly fatal outcomes. After controlling for age, this association was no longer significant (P=.14). All surviving patients were in the ICU at the time of their cardiac arrest, and CPR in the ICU was significantly associated with survival (P=.01) and good functional recovery (P=.05). This association remained statistically significant even after adjustment for age (P=.003 for survival and P=.047 for good functional recovery). Although distributions of admission diagnoses and comorbidities were comparable between patients who experienced cardiac arrest in the ward and those who experienced cardiac arrest in the ICU, initial asystole was found more often in patients in the ward (40% vs 14%). Duration of CPR shorter than 5 minutes also predicted survival (P<.01) and good functional recovery (P=.03). DISCUSSION
major differences among outcome groups regarding comorbid conditions, primary admission diagnosis, cause of cardiac arrest, or initial heart rhythm (statistical analysis of these variables was impossible because of the small number of patients in each of the many subgroups for each variable). Nevertheless, all patients whose cardiac arrest
To our knowledge, this is the first systematic study of CPR in critically ill patients with neurologic or neurosurgical disease. Resuscitation was successful (ie, recovery of spontaneous circulation) in almost two thirds of patients, and 7 patients (18%) survived to be discharged from the hospital. Five of these 7 survivors achieved good functional recov-
60
80
50 60
Duration (min)
Age (y)
40
40
30
20 20 10
0
0
No
Yes
No
Survival
Yes Survival
FIGURE 1. Box plots of age and duration of cardiopulmonary resuscitation by survival. The center line within each box indicates median age or duration, with the box extending to cover the central 50% of the data. Whiskers extend to cover 95% of the data, with outliers shown as individual points.
Mayo Clin Proc.
•
November 2004;79(11):1391-1395
•
www.mayo.edu/proceedings
For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.
1393
CPR IN NEUROLOGIC-NEUROSURGICAL PATIENTS
ery. Being in the ICU at the time of the cardiac arrest and short duration of the CPR efforts, but not patient age, predicted functional outcome. Cardiopulmonary resuscitation was performed in the convalescence period even several days after patient transfer to the ward. Notably, no patient survived when the cardiac arrest was anticipated by a deteriorating clinical course. Successful resuscitation in this series is higher than rates reported in some previous cohorts,3,7,12-15 but similar results have been found by others.16-18 Moreover, the number of patients discharged from the hospital is comparable to most other studies in unselected cohorts.3,5,7-9,18 Favorable functional outcomes in survivors after CPR have been reported previously.3,5,19,20 We herein add that similar favorable functional recovery can be expected in patients with prior critical neurologic and neurosurgical illness. In our study, older age was not associated with lower chances of survival or functional recovery. In fact, survivors tended to be significantly older than nonsurvivors. On this issue, previously published literature has shown discrepancies. Although some reports suggest that increasing age predicts mortality in patients undergoing CPR,9,15,17,21-23 several other studies fail to confirm this finding.3,4,7,12,14,18,20,24-26 These studies have been difficult to perform because elderly patients are less likely to receive CPR,26 primarily because of a preference to favor do-notresuscitate directives.27,28 The relatively favorable outcome of elderly patients in our study could be at least partially explained by the exclusion of the sickest elderly patients who indicated a do-not-resuscitate wish in their advance directives. However, the prevalence of comorbid conditions in our elderly survivors did not differ from that of younger patients, who had slightly worse outcomes. In our study population, we believe that age alone should not be considered a motive to forgo CPR when neurologic condition is stable or improving. All survivors in our cohort experienced cardiac arrest while in the ICU. In other studies, location in the ICU has been found to be predictive of survival,9,15,29,30 but this correlation has not been found uniformly in all studies.4,7,16,31 A better outcome after CPR when performed in the neurologic-neurosurgical ICU is important because it has been argued that CPR in coronary care units but not in medical ICUs offers better chances of survival.9 The dismal outcomes observed after cardiac arrest in the ward have been attributed previously to a high proportion of poor CPR candidates and delayed recognition of the cardiac arrest.29 In our series, the first explanation does not appear applicable because patients who experienced cardiac arrest in the ward and in the ICU had similar admission diagnoses and comorbidities. However, the finding that asystole at onset was more often encountered in patients in the ward 1394
Mayo Clin Proc.
•
may suggest a greater delay in the recognition of the cardiac arrest. Response times may also be longer in wards, but we could not analyze this possibility with our data. Our finding that shorter CPR time is associated with improved outcomes is in clear agreement with the experience in other patient cohorts.3-5,7,9,18,21-23,32 Patients leaving the neurologicneurosurgical ICU are typically transferred to neurosurgical wards. Our findings may argue in favor of a transfer policy to step-down units with cardiac telemetry capabilities to ensure rapid detection of cardiac disturbances. None of our patients whose cardiac arrest was preceded by a declining clinical course survived to be discharged from the hospital. Similar findings have been reported in general medical cohorts32 and cancer patients.2 In our population, this finding seemed to be partially attributable to the fact that the patients with anticipated events tended to be 5 years younger on average, and younger patients tended to have lower cardiac arrest survival rates. This association should be explored further in a larger study. Critically ill neurologic patients may differ from other groups of patients with an acute illness because of a higher prevalence of do-not-resuscitate directives (patients with chronic neurologic disorders or severe acute brain damage may be more likely to decline CPR), lower incidence of multiorgan dysfunction, and more frequent progression to brain death. Therefore, extrapolating data from prior studies on nonneurologic medical or surgical patients may be inappropriate. The small size of our study population precludes a more thorough analysis of predictors of outcome. Variables that were not significantly different among the groups may have had an important effect that we were unable to detect. In other words, lack of significance in this study should not be interpreted to mean no effect. Our cohort included all consecutive patients admitted to the neurologic-neurosurgical ICU who experienced in-hospital cardiac arrest during a period of 8 years. Our tracking methods are reliable, and the small number of patients entering the study is likely representative of the actual frequency of CPR in this subset of patients. One should be cautious when extrapolating these data to other medical centers because remarkably disparate CPR outcomes have been reported across different centers.27 Cardiopulmonary resuscitation is a physically and emotionally traumatic intervention that should be reserved for patients with reasonable chances of a favorable outcome. In patients for whom CPR offers no potential benefit, withholding resuscitation measures after a thorough and compassionate discussion with the patient or family members may prevent unnecessary, additional suffering. Patients and families may be prepared to make a decision on resuscitation directives only after being adequately informed
November 2004;79(11):1391-1395
•
www.mayo.edu/proceedings
For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.
CPR IN NEUROLOGIC-NEUROSURGICAL PATIENTS
about the real chances of recovery in the event of CPR. Thus, it is imperative to define accurate prognostic indicators after CPR applicable to specific groups of neurologic and neurosurgical patients. CONCLUSIONS We believe this study provides novel and useful information that may help guide the counseling of families of critically ill neurologic-neurosurgical patients, regardless of patient age. The ominous prognosis of CPR in patients with a worsening clinical status despite maximal therapy should be communicated. Conversely, stable patients should not be denied resuscitation efforts solely on the basis of their advanced age or the severity of their underlying neurologic disease. REFERENCES 1. Blackhall LJ. Must we always use CPR? N Engl J Med. 1987;317:12811285. 2. Ewer MS, Kish SK, Martin CG, Price KJ, Feeley TW. Characteristics of cardiac arrest in cancer patients as a predictor of survival after cardiopulmonary resuscitation. Cancer. 2001;92:1905-1912. 3. Bedell SE, Delbanco TL, Cook EF, Epstein FH. Survival after cardiopulmonary resuscitation in the hospital. N Engl J Med. 1983;309:569-576. 4. Rozenbaum EA, Shenkman L. Predicting outcome of inhospital cardiopulmonary resuscitation. Crit Care Med. 1988;16:583-586. 5. Beuret P, Feihl F, Vogt P, Perret A, Romand JA, Perret C. Cardiac arrest: prognostic factors and outcome at one year. Resuscitation. 1993;25:171-179. 6. Bialecki L, Woodward RS. Predicting death after CPR: experience at a nonteaching community hospital with a full-time critical care staff. Chest. 1995;108:1009-1017. 7. Peterson MW, Geist LJ, Schwartz DA, Konicek S, Moseley PL. Outcome after cardiopulmonary resuscitation in a medical intensive care unit. Chest. 1991;100:168-174. 8. Ballew KA, Philbrick JT, Caven DE, Schorling JB. Predictors of survival following in-hospital cardiopulmonary resuscitation: a moving target. Arch Intern Med. 1994;154:2426-2432. 9. Saklayen M, Liss H, Markert R. In-hospital cardiopulmonary resuscitation: survival in 1 hospital and literature review. Medicine (Baltimore). 1995; 74:163-175. 10. Hirji KF, Mehta CR, Patel NR. Computing distributions for exact logistic regression. J Am Stat Assoc. 1987;82:1110-1117. 11. Stokes ME, Davis CS, Koch GG. Categorical Data Analysis Using the SAS System. Cary, NC: SAS Institute Inc; 1995.
Mayo Clin Proc.
•
12. Landry FJ, Parker JM, Phillips YY. Outcome of cardiopulmonary resuscitation in the intensive care setting. Arch Intern Med. 1992;152:2305-2308. 13. Messert B, Quaglieri CE. Cardiopulmonary resuscitation: perspectives and problems. Lancet. 1976;2:410-412. 14. Peatfield RC, Sillett RW, Taylor D, McNicol MW. Survival after cardiac arrest in hospital. Lancet. 1977;1:1223-1225. 15. Taffet GE, Teasdale TA, Luchi RJ. In-hospital cardiopulmonary resuscitation. JAMA. 1988;260:2069-2072. 16. DeBard ML. Cardiopulmonary resuscitation: analysis of six years’ experience and review of the literature. Ann Emerg Med. 1981;10:408-416. 17. Murphy DJ, Murray AM, Robinson BE, Campion EW. Outcomes of cardiopulmonary resuscitation in the elderly. Ann Intern Med. 1989;111:199205. 18. Zoch TW, Desbiens NA, DeStefano F, Stueland DT, Layde PM. Shortand long-term survival after cardiopulmonary resuscitation. Arch Intern Med. 2000;160:1969-1973. 19. Minuck M, Perkins R. Long-term study of patients successfully resuscitated following cardiac arrest. Can Med Assoc J. 1969;100:1126-1128. 20. Urberg M, Ways C. Survival after cardiopulmonary resuscitation for an in-hospital cardiac arrest. J Fam Pract. 1987;25:41-44. 21. O’Keeffe S, Redahan C, Keane P, Daly K. Age and other determinants of survival after in-hospital cardiopulmonary resuscitation. Q J Med. 1991;81: 1005-1010. 22. Schultz SC, Cullinane DC, Pasquale MD, Magnant C, Evans SR. Predicting in-hospital mortality during cardiopulmonary resuscitation. Resuscitation. 1996;33:13-17. 23. van Walraven C, Forster AJ, Stiell IG. Derivation of a clinical decision rule for the discontinuation of in-hospital cardiac arrest resuscitations. Arch Intern Med. 1999;159:129-134. 24. Di Bari M, Chiarlone M, Fumagalli S, et al. Cardiopulmonary resuscitation of older, inhospital patients: immediate efficacy and long-term outcome. Crit Care Med. 2000;28:2320-2325. 25. Gulati RS, Bhan GL, Horan MA. Cardiopulmonary resuscitation of old people. Lancet. 1983;2:267-269. 26. Schwenzer KJ, Smith WT, Durbin CG Jr. Selective application of cardiopulmonary resuscitation improves survival rates. Anesth Analg. 1993; 76:478-484. 27. Cook DJ, Guyatt G, Rocker G, et al. Cardiopulmonary resuscitation directives on admission to intensive-care unit: an international observational study. Lancet. 2001;358:1941-1945. 28. Murphy DJ, Burrows D, Santilli S, et al. The influence of the probability of survival on patients’ preferences regarding cardiopulmonary resuscitation. N Engl J Med. 1994;330:545-549. 29. Hershey CO, Fisher L. Why outcome of cardiopulmonary resuscitation in general wards is poor. Lancet. 1982;1:31-34. 30. Karetzky M, Zubair M, Parikh J. Cardiopulmonary resuscitation in intensive care unit and non-intensive care unit patients: immediate and long-term survival. Arch Intern Med. 1995;155:1277-1280. 31. Shultz JJ, Iskos D, Lurie KG. Alternative mechanical methods of cardiopulmonary resuscitation. Am J Ther. 1996;3:661-666. 32. Marik PE, Craft M. An outcomes analysis of in-hospital cardiopulmonary resuscitation: the futility rationale for do not resuscitate orders. J Crit Care. 1997;12:142-146.
November 2004;79(11):1391-1395
•
www.mayo.edu/proceedings
For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.
1395
CPR IN NEUROLOGIC-NEUROSURGICAL PATIENTS
Cardiopulmonary Resuscitation in Critically Ill Neurologic-Neurosurgical Patients ALEJANDRO A. RABINSTEIN, MD; ROBYN L. MCCLELLAND, PHD; EELCO F. M. WIJDICKS, MD; EDWARD M. MANNO, MD; AND JOHN L. D. ATKINSON, MD
OBJECTIVES: To establish the rate of successful cardiopulmonary resuscitation (CPR) and to study outcome predictors in patients who experienced in-hospital cardiac arrest after being admitted to the neurologic-neurosurgical intensive care unit (ICU) with a primary neurologic diagnosis. PATIENTS AND METHODS: We identified patients admitted to the neurologic-neurosurgical ICU between 1994 and 2001 who experienced in-hospital cardiac arrest and received CPR. Functional outcome was assessed using the modified Rankin scale. RESULTS: During the study period, 38 consecutive patients experienced in-hospital cardiac arrest and received CPR. The median age of the patients was 65 years (range, 16-81 years), and the mean interval from admission to CPR was 12 days (range, 3 hours to 47 days). Acute intracranial disease was present in 32 patients (84%). Twenty-one patients (55%) were in the ICU at the time of the cardiac arrest; cardiac arrests in the wards occurred at a mean interval of 9 days (range, 1-45 days) after ICU discharge. Cardiopulmonary resuscitation achieved return of spontaneous circulation in 23 patients (61%). Seven patients (18%) were discharged from the hospital, 5 of whom later achieved a modified Rankin scale score of 2 or lower. Cardiac arrest after a deteriorating clinical course resulted in uniformly fatal outcomes. Duration of CPR shorter than 5 minutes and CPR in the ICU were associated with survival and good functional recovery. CONCLUSIONS: Cardiopulmonary resuscitation is a worthwhile procedure in severely ill neurologic-neurosurgical patients, regardless of the patient’s age. However, the outcome after CPR appears much worse in patients with a prior deteriorating clinical course.
Mayo Clin Proc. 2004;79(11):1391-1395 CPR = cardiopulmonary resuscitation; ICU = intensive care unit
C
ardiopulmonary resuscitation (CPR) is practiced throughout the entire hospital. However, studies on the validity of CPR for in-hospital cardiac arrest have included mixed populations, making it difficult to ferret out the appropriateness of CPR in the neurologicneurosurgical intensive care unit (ICU).1 Some studies have shown that CPR is ineffective in certain groups of
From the Department of Neurology (A.A.R., E.F.M.W., E.M.M.), Division of Biostatistics (R.L.M), and Department of Neurologic Surgery (J.L.D.A.), Mayo Clinic College of Medicine, Rochester, Minn. Dr Rabinstein is now with the University of Miami School of Medicine, Miami, Fla. Address reprint requests and correspondence to Eelco F. M. Wijdicks, MD, Department of Neurology, Mayo Clinic College of Medicine, 200 First St SW, Rochester, MN 55905 (e-mail: [email protected]). © 2004 Mayo Foundation for Medical Education and Research
Mayo Clin Proc.
•
patients, such as those with advanced cancer,2 refractory sepsis or severe pneumonia,3,4 renal failure,5 greater severity of acute illness,6,7 and poor preresuscitation functional status.8 Surprisingly, we found no such information on prognosis after CPR in critically ill neurologic and neurosurgical patients. Commonly, CPR is discussed in the neurologic-neurosurgical ICU, where, perhaps more than in any other ICU setting, families are concerned about not only the chances of survival but also functional outcome. Reasoned judgment is particularly necessary for deteriorating patients in whom meaningful neurologic recovery after CPR is considered unlikely. Therefore, information is needed on what factors influence survival and functional recovery after CPR in patients with critical neurologic and neurosurgical illness. The goals of this study were to define the rate of success of CPR in patients admitted to the neurologicneurosurgical ICU with a primary neurologic diagnosis who experienced in-hospital cardiac arrest and to study predictors of outcome in this patient population. PATIENTS AND METHODS We identified all patients admitted to the neurologic-neurosurgical ICU at the Mayo Clinic in Rochester, Minn, between January 1994 and December 2001 who experienced cardiac arrest during their hospital stay. All episodes of CPR in the hospital are thoroughly recorded and can be readily retrieved from our linkage system. In our institution, all episodes of CPR are reviewed in detail by a dedicated committee, and documentation of the interventions is standardized across the hospital. Cardiac arrest was defined by the presence of a hemodynamically unstable heart rhythm, including ventricular tachycardia or profound bradycardia with evidence of cerebral hypoperfusion, ventricular fibrillation, pulseless electrical activity, or cardiac asystole. We excluded patients with respiratory arrest not associated with cardiac arrest given their much more benign prognosis,9 patients who had seizures without unstable cardiac arrhythmia, patients who had hypotensive episodes without cardiac arrhythmia, and those who had episodes of cardiac arrhythmia without hemodynamic compromise.
November 2004;79(11):1391-1395
•
www.mayo.edu/proceedings
For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.
1391
CPR IN NEUROLOGIC-NEUROSURGICAL PATIENTS
TABLE 1. Primary Admission Diagnoses of 38 Consecutive Critically Ill Neurologic-Neurosurgical Patients Who Experienced In-hospital Cardiac Arrest Diagnosis
No. at risk
No. (%) who survived
Head trauma* Subarachnoid hemorrhage† Ischemic stroke Spinal disease Intracranial tumor Intracerebral hemorrhage Meningoencephalitis Status epilepticus Chronic subdural hematoma
8 6 5 5 4 3 3 2 2
2 (25) 2 (33) 1 (20) 1 (20) 1 (25) 0 (0) 0 (0) 0 (0) 0 (0)
*All patients with head trauma were admitted with a Glasgow Coma Scale sum score lower than 8, and there were no significant differences on Glasgow Coma Scale sum score at the time of cardiac arrest between survivors and nonsurvivors. †World Federation of Neurological Surgeons grades on admission in patients who did not survive were I in 2 patients, II in 1 patient, and III in 1 patient; admission grade was II in both survivors.
We recorded patients’ demographic information, medical and surgical history, reason for hospital and ICU admission, complications during the hospital course, and time of the cardiac arrest. Cause of cardiac arrest, initial heart rhythm at the time of onset of the resuscitation efforts, and duration of such efforts were noted. We categorized the arrests into unexpected or anticipated. We defined a cardiac arrest as anticipated when it was preceded by a gradual deterioration of the patient’s clinical status that was unresponsive to aggressive medical or surgical interventions. This deterioration could be due to worsening of the primary neurologic disease or development of neurologic or systemic complications. Outcome of survivors was assessed using the modified Rankin scale at the time of last followup. A modified Rankin scale score of 2 or lower (slight or no disability and preserved ability to look after own affairs TABLE 2. Causes of Cardiac Arrest in 38 Consecutive Critically Ill Neurologic-Neurosurgical Patients According to Outcome Cause
No. at risk
No. (%) who survived
Aspiration Myocardial ischemia Cardiac arrhythmia End-stage herniation Pulmonary embolism Sepsis Seizures Airway obstruction* Gastrointestinal bleeding Unclear
6 5 5 5 4 3 3 2 1 4
0 (0) 1 (20) 2 (40) 0 (0) 0 (0) 1 (33) 1 (33) 2 (100) 0 (0) 0 (0)
*We included patients with acute airway obstruction (obstruction of tracheostomy in 1 patient and upper airway obstruction after extubation in 1 patient) that led to cardiac arrest but excluded those with pure respiratory arrest.
1392
Mayo Clin Proc.
•
without assistance) was considered indicative of good functional outcome. Outcome predictors were compared using the Fisher exact test for nominal variables and the independent 2tailed t test for continuous variables, with modification for unequal variances. Age-adjusted evaluation of associations was performed using exact logistic regression models.10,11 Level of statistical significance was established at P<.05. RESULTS We identified 38 consecutive patients admitted to our neurologic-neurosurgical ICU during the study period who had experienced in-hospital cardiac arrest and received CPR. The median age of the patients was 65 years (range, 16-81 years), and 66% were men. Comorbidities were frequent, including smoking in 21 patients (55%); hypertension in 17 (45%); coronary artery disease in 14 (37%); hyperlipidemia in 9 (24%); previous stroke, diabetes mellitus, chronic obstructive pulmonary disease, and peptic ulcer disease in 7 each (18%); and history of cardiac arrhythmia in 6 (16%). Acute intracranial disease was present in 32 patients (84%). Primary admission diagnoses are listed in Table 1. Cardiac arrest occurred at a mean interval of 12 days after hospital admission (range, 3 hours to 47 days). Twenty-one patients (55%) were in the ICU at the time of the cardiac arrest; among those who experienced cardiac arrest in the wards, the mean interval between ICU discharge and cardiac arrest was 9 days (range, 1-45 days). The causes of cardiac arrest are presented in Table 2. The initial heart rhythms were asystole in 9 patients, bradycardia in 9 patients, and pulseless electrical activity, ventricular fibrillation, and unstable ventricular tachycardia in 7 patients each. Cardiac arrests were unexpected in 26 patients (68%). The median duration of the resuscitation efforts was 17 minutes (range, 2-60 minutes). Cardiopulmonary resuscitation achieved return of spontaneous circulation in 23 patients (61%). Of these 23 patients, 16 subsequently died of a recurrent cardiac arrest (5 patients) or had progression to brain death (11 patients). Seven patients (18% of our total study population) survived to be discharged from the hospital: 2 were discharged home, 2 to a rehabilitation facility, and 3 to nursing homes. At the time of last follow-up (mean follow-up period, 9 months; median, 7.5 months; range, 1-24 months), 5 (71%) of the 7 survivors had achieved a good functional recovery (modified Rankin scale score, ≤2). We analyzed predictive variables for survival in our study population (Tables 1, 2, and 3 and Figure 1). Surviving patients were older than those who died (mean ± SD age, 70±9 years vs 58±19 years; P=.03). There were no
November 2004;79(11):1391-1395
•
www.mayo.edu/proceedings
For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.
CPR IN NEUROLOGIC-NEUROSURGICAL PATIENTS
TABLE 3. Characteristics of In-hospital Cardiac Arrest and Resuscitation Efforts in 38 Consecutive Critically Ill Neurologic-Neurosurgical Patients* Variable Cardiac arrest type Unanticipated Anticipated Location ICU Ward Initial rhythm Asystole Bradycardia PEA VF Unstable tachycardia CPR duration (min) ≤5 >5
No. at risk
No. (%) who survived†
26 12
7 (27) 0 (0)
21 17
7 (33) 0 (0)
9 9 7 7 7
0 (0) 2 (22) 2 (29) 1 (14) 2 (29)
11 27
6 (55) 1 (4)
P value .07 .01‡ .54
.001‡
*CPR = cardiopulmonary resuscitation; ICU = intensive care unit; PEA = pulseless electrical activity; VF = ventricular fibrillation. †At the time of last follow-up. ‡Fisher exact test.
was precipitated by aspiration, pulmonary embolism, or brain herniation died in the hospital. Patients with cardiac arrest anticipated by a deteriorating clinical course (n=12) had uniformly fatal outcomes. After controlling for age, this association was no longer significant (P=.14). All surviving patients were in the ICU at the time of their cardiac arrest, and CPR in the ICU was significantly associated with survival (P=.01) and good functional recovery (P=.05). This association remained statistically significant even after adjustment for age (P=.003 for survival and P=.047 for good functional recovery). Although distributions of admission diagnoses and comorbidities were comparable between patients who experienced cardiac arrest in the ward and those who experienced cardiac arrest in the ICU, initial asystole was found more often in patients in the ward (40% vs 14%). Duration of CPR shorter than 5 minutes also predicted survival (P<.01) and good functional recovery (P=.03). DISCUSSION
major differences among outcome groups regarding comorbid conditions, primary admission diagnosis, cause of cardiac arrest, or initial heart rhythm (statistical analysis of these variables was impossible because of the small number of patients in each of the many subgroups for each variable). Nevertheless, all patients whose cardiac arrest
To our knowledge, this is the first systematic study of CPR in critically ill patients with neurologic or neurosurgical disease. Resuscitation was successful (ie, recovery of spontaneous circulation) in almost two thirds of patients, and 7 patients (18%) survived to be discharged from the hospital. Five of these 7 survivors achieved good functional recov-
60
80
50 60
Duration (min)
Age (y)
40
40
30
20 20 10
0
0
No
Yes
No
Survival
Yes Survival
FIGURE 1. Box plots of age and duration of cardiopulmonary resuscitation by survival. The center line within each box indicates median age or duration, with the box extending to cover the central 50% of the data. Whiskers extend to cover 95% of the data, with outliers shown as individual points.
Mayo Clin Proc.
•
November 2004;79(11):1391-1395
•
www.mayo.edu/proceedings
For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.
1393
CPR IN NEUROLOGIC-NEUROSURGICAL PATIENTS
ery. Being in the ICU at the time of the cardiac arrest and short duration of the CPR efforts, but not patient age, predicted functional outcome. Cardiopulmonary resuscitation was performed in the convalescence period even several days after patient transfer to the ward. Notably, no patient survived when the cardiac arrest was anticipated by a deteriorating clinical course. Successful resuscitation in this series is higher than rates reported in some previous cohorts,3,7,12-15 but similar results have been found by others.16-18 Moreover, the number of patients discharged from the hospital is comparable to most other studies in unselected cohorts.3,5,7-9,18 Favorable functional outcomes in survivors after CPR have been reported previously.3,5,19,20 We herein add that similar favorable functional recovery can be expected in patients with prior critical neurologic and neurosurgical illness. In our study, older age was not associated with lower chances of survival or functional recovery. In fact, survivors tended to be significantly older than nonsurvivors. On this issue, previously published literature has shown discrepancies. Although some reports suggest that increasing age predicts mortality in patients undergoing CPR,9,15,17,21-23 several other studies fail to confirm this finding.3,4,7,12,14,18,20,24-26 These studies have been difficult to perform because elderly patients are less likely to receive CPR,26 primarily because of a preference to favor do-notresuscitate directives.27,28 The relatively favorable outcome of elderly patients in our study could be at least partially explained by the exclusion of the sickest elderly patients who indicated a do-not-resuscitate wish in their advance directives. However, the prevalence of comorbid conditions in our elderly survivors did not differ from that of younger patients, who had slightly worse outcomes. In our study population, we believe that age alone should not be considered a motive to forgo CPR when neurologic condition is stable or improving. All survivors in our cohort experienced cardiac arrest while in the ICU. In other studies, location in the ICU has been found to be predictive of survival,9,15,29,30 but this correlation has not been found uniformly in all studies.4,7,16,31 A better outcome after CPR when performed in the neurologic-neurosurgical ICU is important because it has been argued that CPR in coronary care units but not in medical ICUs offers better chances of survival.9 The dismal outcomes observed after cardiac arrest in the ward have been attributed previously to a high proportion of poor CPR candidates and delayed recognition of the cardiac arrest.29 In our series, the first explanation does not appear applicable because patients who experienced cardiac arrest in the ward and in the ICU had similar admission diagnoses and comorbidities. However, the finding that asystole at onset was more often encountered in patients in the ward 1394
Mayo Clin Proc.
•
may suggest a greater delay in the recognition of the cardiac arrest. Response times may also be longer in wards, but we could not analyze this possibility with our data. Our finding that shorter CPR time is associated with improved outcomes is in clear agreement with the experience in other patient cohorts.3-5,7,9,18,21-23,32 Patients leaving the neurologicneurosurgical ICU are typically transferred to neurosurgical wards. Our findings may argue in favor of a transfer policy to step-down units with cardiac telemetry capabilities to ensure rapid detection of cardiac disturbances. None of our patients whose cardiac arrest was preceded by a declining clinical course survived to be discharged from the hospital. Similar findings have been reported in general medical cohorts32 and cancer patients.2 In our population, this finding seemed to be partially attributable to the fact that the patients with anticipated events tended to be 5 years younger on average, and younger patients tended to have lower cardiac arrest survival rates. This association should be explored further in a larger study. Critically ill neurologic patients may differ from other groups of patients with an acute illness because of a higher prevalence of do-not-resuscitate directives (patients with chronic neurologic disorders or severe acute brain damage may be more likely to decline CPR), lower incidence of multiorgan dysfunction, and more frequent progression to brain death. Therefore, extrapolating data from prior studies on nonneurologic medical or surgical patients may be inappropriate. The small size of our study population precludes a more thorough analysis of predictors of outcome. Variables that were not significantly different among the groups may have had an important effect that we were unable to detect. In other words, lack of significance in this study should not be interpreted to mean no effect. Our cohort included all consecutive patients admitted to the neurologic-neurosurgical ICU who experienced in-hospital cardiac arrest during a period of 8 years. Our tracking methods are reliable, and the small number of patients entering the study is likely representative of the actual frequency of CPR in this subset of patients. One should be cautious when extrapolating these data to other medical centers because remarkably disparate CPR outcomes have been reported across different centers.27 Cardiopulmonary resuscitation is a physically and emotionally traumatic intervention that should be reserved for patients with reasonable chances of a favorable outcome. In patients for whom CPR offers no potential benefit, withholding resuscitation measures after a thorough and compassionate discussion with the patient or family members may prevent unnecessary, additional suffering. Patients and families may be prepared to make a decision on resuscitation directives only after being adequately informed
November 2004;79(11):1391-1395
•
www.mayo.edu/proceedings
For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.
CPR IN NEUROLOGIC-NEUROSURGICAL PATIENTS
about the real chances of recovery in the event of CPR. Thus, it is imperative to define accurate prognostic indicators after CPR applicable to specific groups of neurologic and neurosurgical patients. CONCLUSIONS We believe this study provides novel and useful information that may help guide the counseling of families of critically ill neurologic-neurosurgical patients, regardless of patient age. The ominous prognosis of CPR in patients with a worsening clinical status despite maximal therapy should be communicated. Conversely, stable patients should not be denied resuscitation efforts solely on the basis of their advanced age or the severity of their underlying neurologic disease. REFERENCES 1. Blackhall LJ. Must we always use CPR? N Engl J Med. 1987;317:12811285. 2. Ewer MS, Kish SK, Martin CG, Price KJ, Feeley TW. Characteristics of cardiac arrest in cancer patients as a predictor of survival after cardiopulmonary resuscitation. Cancer. 2001;92:1905-1912. 3. Bedell SE, Delbanco TL, Cook EF, Epstein FH. Survival after cardiopulmonary resuscitation in the hospital. N Engl J Med. 1983;309:569-576. 4. Rozenbaum EA, Shenkman L. Predicting outcome of inhospital cardiopulmonary resuscitation. Crit Care Med. 1988;16:583-586. 5. Beuret P, Feihl F, Vogt P, Perret A, Romand JA, Perret C. Cardiac arrest: prognostic factors and outcome at one year. Resuscitation. 1993;25:171-179. 6. Bialecki L, Woodward RS. Predicting death after CPR: experience at a nonteaching community hospital with a full-time critical care staff. Chest. 1995;108:1009-1017. 7. Peterson MW, Geist LJ, Schwartz DA, Konicek S, Moseley PL. Outcome after cardiopulmonary resuscitation in a medical intensive care unit. Chest. 1991;100:168-174. 8. Ballew KA, Philbrick JT, Caven DE, Schorling JB. Predictors of survival following in-hospital cardiopulmonary resuscitation: a moving target. Arch Intern Med. 1994;154:2426-2432. 9. Saklayen M, Liss H, Markert R. In-hospital cardiopulmonary resuscitation: survival in 1 hospital and literature review. Medicine (Baltimore). 1995; 74:163-175. 10. Hirji KF, Mehta CR, Patel NR. Computing distributions for exact logistic regression. J Am Stat Assoc. 1987;82:1110-1117. 11. Stokes ME, Davis CS, Koch GG. Categorical Data Analysis Using the SAS System. Cary, NC: SAS Institute Inc; 1995.
Mayo Clin Proc.
•
12. Landry FJ, Parker JM, Phillips YY. Outcome of cardiopulmonary resuscitation in the intensive care setting. Arch Intern Med. 1992;152:2305-2308. 13. Messert B, Quaglieri CE. Cardiopulmonary resuscitation: perspectives and problems. Lancet. 1976;2:410-412. 14. Peatfield RC, Sillett RW, Taylor D, McNicol MW. Survival after cardiac arrest in hospital. Lancet. 1977;1:1223-1225. 15. Taffet GE, Teasdale TA, Luchi RJ. In-hospital cardiopulmonary resuscitation. JAMA. 1988;260:2069-2072. 16. DeBard ML. Cardiopulmonary resuscitation: analysis of six years’ experience and review of the literature. Ann Emerg Med. 1981;10:408-416. 17. Murphy DJ, Murray AM, Robinson BE, Campion EW. Outcomes of cardiopulmonary resuscitation in the elderly. Ann Intern Med. 1989;111:199205. 18. Zoch TW, Desbiens NA, DeStefano F, Stueland DT, Layde PM. Shortand long-term survival after cardiopulmonary resuscitation. Arch Intern Med. 2000;160:1969-1973. 19. Minuck M, Perkins R. Long-term study of patients successfully resuscitated following cardiac arrest. Can Med Assoc J. 1969;100:1126-1128. 20. Urberg M, Ways C. Survival after cardiopulmonary resuscitation for an in-hospital cardiac arrest. J Fam Pract. 1987;25:41-44. 21. O’Keeffe S, Redahan C, Keane P, Daly K. Age and other determinants of survival after in-hospital cardiopulmonary resuscitation. Q J Med. 1991;81: 1005-1010. 22. Schultz SC, Cullinane DC, Pasquale MD, Magnant C, Evans SR. Predicting in-hospital mortality during cardiopulmonary resuscitation. Resuscitation. 1996;33:13-17. 23. van Walraven C, Forster AJ, Stiell IG. Derivation of a clinical decision rule for the discontinuation of in-hospital cardiac arrest resuscitations. Arch Intern Med. 1999;159:129-134. 24. Di Bari M, Chiarlone M, Fumagalli S, et al. Cardiopulmonary resuscitation of older, inhospital patients: immediate efficacy and long-term outcome. Crit Care Med. 2000;28:2320-2325. 25. Gulati RS, Bhan GL, Horan MA. Cardiopulmonary resuscitation of old people. Lancet. 1983;2:267-269. 26. Schwenzer KJ, Smith WT, Durbin CG Jr. Selective application of cardiopulmonary resuscitation improves survival rates. Anesth Analg. 1993; 76:478-484. 27. Cook DJ, Guyatt G, Rocker G, et al. Cardiopulmonary resuscitation directives on admission to intensive-care unit: an international observational study. Lancet. 2001;358:1941-1945. 28. Murphy DJ, Burrows D, Santilli S, et al. The influence of the probability of survival on patients’ preferences regarding cardiopulmonary resuscitation. N Engl J Med. 1994;330:545-549. 29. Hershey CO, Fisher L. Why outcome of cardiopulmonary resuscitation in general wards is poor. Lancet. 1982;1:31-34. 30. Karetzky M, Zubair M, Parikh J. Cardiopulmonary resuscitation in intensive care unit and non-intensive care unit patients: immediate and long-term survival. Arch Intern Med. 1995;155:1277-1280. 31. Shultz JJ, Iskos D, Lurie KG. Alternative mechanical methods of cardiopulmonary resuscitation. Am J Ther. 1996;3:661-666. 32. Marik PE, Craft M. An outcomes analysis of in-hospital cardiopulmonary resuscitation: the futility rationale for do not resuscitate orders. J Crit Care. 1997;12:142-146.
November 2004;79(11):1391-1395
•
www.mayo.edu/proceedings
For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.
1395