Upload
others
View
6
Download
0
Embed Size (px)
Citation preview
1238 Volume 3 ‘ Number 6 ‘ 1992
Outcomes of Cardiopulmonary Resuscitation in Dialysis1,2
Alvin H. Moss,3 Jean L. Holley, and Matthew B. Upton
A.H. Moss. Center for Health Ethics and Law, West
Virginia University Health Sciences Center, Morgan-town, WV
A.H. Moss, MB. Upton, Department of Medicine, Sec-tion of Nephrology, West Virginia University School ofMedicine, Morgantown, WV
J.L. Holley, Department of Medicine, Penal-ElectrolyteDivision, University of Pittsburgh School of Medicine,
Pittsburgh, PA
(J. Am. Soc. Nephrol. 1992; 3:1238-1243)
ABSTRACTPatients with renal failure are believed to have apoor survival rate after cardiopulmonary resuscita-tion, but there is little specific information about theoutcomes of cardiopulmonary resuscitation in di-alysis patients. To be better able to inform dialysispatients and assist them in decision making about
cardiopulmonary resuscitation, the eight-year ex-
perience with cardiopulmonary resuscitation in di-alysis patients at a university dialysis program was
analyzed and outcomes were compared with thoseof a control group of nondialysis patients undergoingcardiopulmonary resuscitation during the same timeperiod in the same hospital. Of 221 dialysis patientsexperiencing cardiopulmonary arrest, 74 (34%) hadCPR compared with 247 (21%) of 1,201 control pa-tients (P = 0.0002). Six of 74 (8%; 95% confidence
interval, 2 to 14%) dialysis patients survived to hos-
pital discharge compared with 30 of 247 (12%; 95%
confidence interval, 8 to 16%) control patients (P =
not significant). At 6 months after CPR, 2 (3%) of 74dialysis patients were still alive compared with 23
(9%) of 247 controls (P = 0.044); this difference wasnot explained by age or comorbid conditions.Twenty-one (78%) of the 27 successfully resuscitateddialysis patients died a mean of 4.4 days later; 95%
I Received April 14, 1992. Accepted July 20, 1992.2 Presented In part at the annual scientific meeting of the AmerIcan 5oclety of
Nephrology. Baltimore. MD, November 18. 1991.
3 Correspondence to Dr. A. H. Moss, center for Health Ethics and Law, West
Virginia University Health scIences center. 1354 H5N, Morgantown, WV 26506.
1046-6673/0306-1238$03.00/0Journal of the American 5ociety of NephrologyCopyright C 1992 by the American Society of Nephrology
were on mechanical ventilation in an intensive careunit at the time of death. It was concluded that
cardiopulmonary resuscitation is a procedure thatrarely results in extended survival for dialysis patients.In discussions about cardiopulmonary resuscitation
with dialysis patients, nephrologists should providethis information.
Key Words: Survival, advance directives, decision making
O nly 1 5% of all patients who undergo cardlopul-monary resuscitation (CPR) survive to hospital
discharge ( 1 ). Survival after CPR for selected groupsof patients such as elderly nursing home residentsand those with metastatic cancer is much worse (2-
4). PhysicIans have been encouraged to Inform suchpatients of their likely poor outcomes after CPR andto refrain from offering It to them (2,4,5).
The Information In the medical literature aboutoutcomes of resuscitation in patients with renal fall-ure Is limited (6- 1 2). However, the available studiessuggest that 1 0% or fever of renal failure patientsundergoing CPR survive to hospital discharge. Data
on CPR outcomes in dialysis patients are even morelimited; in one small study, 2 of 1 0 peritoneal dialysis
patients left the hospital alive after CPR (1 3). NIne ofthese patients had major complications of the CPR;five had flail chests, and four had multiple rib frac-tures. We know that the attitudes of hemodlalysispatients toward CPR do not differ from those of otherpatients (1 4). Unfortunately, aside from the abovestudy, there is little specific Information about CPRoutcomes In dialysis patients. Thus, the risks andbenefits of CPR are difficult for nephrologlsts to enu-merate In discussions with their dialysis patients.
We conducted this study to answer the followingthree questions about CPR in dialysis patients. (1)
What percentage of dialysis patients undergo CPR,which factors distinguish those who do from those
who do not, and how does this percentage comparewith that of a control group of nondlalysis patients?
(2) What are the causes of cardiopulmonary arrestpreceding CPR in dialysis patients? (3) What are theoutcomes of CPR in dialysis patients and how do theycompare with those of nondialysis patients?
METHODS
Study PopulationDialysis Patients. All adult chronic dialysis pa-
tients of the West Virginia Health Care Cooperative
Moss et al
Journal of the American Society of Nephrology 1239
who experienced a cardiopulmonary arrest between
January 1983 and May 1991 were included. Patientswere identified from the dialysis unit and West Vir-
ginla University Hospitals records of CPR proceduresand patient deaths. All of these patients underwentCPR unless a “Do Not Resuscitate” order had been
written, or unless they were found dead with rigormortis. They were identified retrospectively far theyears 1 983 to 1 986 and prospectively far 1 987 to
1991.The dialysis program of the West Virginia Health
Care Cooperative provided dialysis of 1 00 patients onaverage per year during the study and was treating1 30 patIents at the end of the study. Patients were
dialyzed In one of two free-standing units. Dialysiscare at each unit was provided by registered nurseswho were certified in Advanced Cardiac Life Supportby the American Heart Association. All inpatient carewas provided at West Virginia University Hospitals.
Control Patients. Control group patients were in-eluded In this study If they were adults and met the
following criteria: ( 1 ) experienced a cardiopulmonaryarrest while an inpatient at West Virginia UniversityHospitals, (2) underwent CPR between March 1987
and May 1989, and (3) dId not have acute or chronicrenal failure requiring dialysis. Control group pa-
tlents were Identified retrospectively from hospitalrecords of consecutive CPR procedures betweenMarch 1987 and May 1989.
Each patient chart was reviewed for the followingpatient parameters: age: gender; dialysis treatment(peritoneal, hemodlalysis, or none); survival afterCPR; camorbid conditions including diabetes,chronic pulmonary disease, cardiac disease, malig-
nancy, cerebrovascular disease, and sepsis; andcause of cardiopulmonary arrest. Charts were alsoreviewed for CPR parameters: performance of CPR,duration of procedure, location, and complications.
The percentage of dialysis patients undergoing CPRwas calculated by dividing the number of patients onwham CPR was performed by the number of dialysispatients experiencing cardiopulmonary arrest during
the study period. The percentage of control grouppatients undergoing CPR was calculated by dividingthe number of patients on whom CPR was performedby the number of hospital patients experiencing car-dlopulmonary arrest during the study period. Pa-tients who were dead on arrival In the emergencydepartment, stillborn births, and patients dying inthe operating room were not Included because CPRwas not attempted In the first two groups and CPRprocedure records were not submitted far the lastgroup.
Follow-Up
Survival to 6 months after CPR in dialysis andcontrol group patients was determined by reviewing
dialysis and hospital records and, when these recordswere incomplete, by contacting the primary physi-clan.
For the purpose of this study, CPR was definedaccording to the Guidelines far Cardiopulmonary Re-suscltatian and Emergency Cardiac Care of theAmerican Heart Association (15).
Statistical Analysis
Comparisons of population proportions were per-formed by the x2 test with Yates’ correction and the
Fisher’s exact test where appropriate. Parametricdata were analyzed by the t test for independentsamples. P values of hess than 0.05 were consideredsignificant. Data are presented as the mean ± stand-ard deviation.
This study protocol was approved by the West Vir-gmnla University Institutional Review Board far theProtection of Human Subjects.
RESULTS
Patients and Rates of CPR
Between 1 983 and 1 99 1 , 22 1 dialysIs patients ex-perienced cardiopulmonary arrest, and 74 (34%)
underwent CPR. The patients who underwent CPRwere younger. and there was a trend for the patientsto more often be men. A smaller percentage also hadcancer (Table 1 ). The diabetics who underwent CPRwere significantly younger than those who did not(56 ± 1 3 versus 62 ± 1 2 yr: P 0.007). The meanage of the men and women undergoing CPR was the
same for bath diabetics and nondlabetics. There wasno difference in the percentage of patients undergo-ing CPR on the basis of their dialysis modality: 49(34%) of 143 hemodlalysls patients and 14 (31%) of
45 peritoneal dialysis patients. Eleven (31%) of 35patients who had been on both In the recent pastunderwent CPR. Sixty-three (85%) patIents experl-enced cardiopulmonary arrest In the hospital, six
experienced It in their homes, and five experiencedIt In the outpatient dialysis units.
From 1987 to 1989, 1,201 control group patientsexperienced cardiopulmonary arrest, and 247 (2 1 %)
TABLE I . Demographics of dialysis patientsexperiencing cardiopulmonary arrest
�PR(N=74)
NoCPR(N=147) ‘#{176}
Mean Age (yr) 58 ± 15 65 ± 12 0.001No. of Men (%) 41 (55) 63 (43) 0.08No. of Diabetics (%) 39 (53) 81 (55) NSNo. of Diagnosis of Cancer (%) 8 (1 1) 43 (29) 0.007Mean Time on Dialysis (months) 24 ± 26 20 ± 26 NS
2711 (1)#{176}
74
664(1)3(1)�3 (2)d22
2(1)
74 (6)
CPR in Dialysis Patients
1240 Volume 3 ‘ Number 6 ‘ 1992
TABLE 2. Comparison of dialysis and control patients undergoing CPR
Dialysis(N= 74)
Control Group(N= 247)
Mean Age (yr) 58 ± 15 64 ± 16 0.004No. of Men (%) 41 (55) 147 (60) NSNo. of Diabetics (%) 39 (53) 57 (23) 0.001Mean Serum Creatinine (,�mol/L) 740 ± 290 160 ± I 10 0.001
underwent CPR. The percentage of control grouppatients who received CPR was less than that ofdialysis patients (21 versus 33%; P 0.0002). Thesepatients were alder, and fewer were diabetics cam-
pared with the dialysis patients undergoing CPR(Table 2). The diabetics in the control group were
significantly older than the diabetic dialysis patients
who underwent CPR (68 ± 1 1 versus 56 ± 13 yr; P <
0.00 1 ). There was no difference in the mean age ofthe control group compared with that of the entiredialysis patient population (63 ± 1 6 versus 62 ± 14yr; P = not significant [NS]). The mean age of the menin the control group was significantly less than thatof the women (61 ± 17 versus 67 ± 15 yr: P = 0.004).The percentage of patients with comorbid condi-
tions-chronic pulmonary disease, cardiac disease,
malignancy. cerebrovascular disease, neurologic dis-ease, and sepsis-did not differ between dialysispatients undergoing CPR and control group patients.
Causes of Cardiopulmonary Arrest
The causes of cardiopulmonary arrest In the dl-alysls patients are listed In Table 3. Fewer than 10%of the arrests (7 of 74) occurred while patients werereceiving a hemodlalysis treatment, yielding a cardla-pulmonary arrest rate on dialysis of roughly 1 per 82patient yr. All of these arrests were likely precipitatedin some way by the dialysis process: In six cases, It
Is likely that the cardiovascular stress of hemodi-alysis contributed to the arrests. Five of these pa-
tients had known severe cardiac disease (four hadlschemlc disease, and one who had a cardiomyapathy
also had an allergic reaction to a dialyzer). and thesixth had amyloldosls and developed refractory hy-potenslon. In the seventh patient, a dialysis-relatedair embolus caused the arrest. Two (3%) of the arrestswere the result of hyperkalemia. One patient withatherosclerotic cardiovascular disease suffered a car-diac arrest from electromechanical dissociation afterreceiving phenytoin at a rate greater than 25 mg/mm.
Resuscitation Outcomes and Survival
The survival after CPR in the dialysis and controlgroup patients Is listed In Table 4. There was no
TABLE 3. Causes of cardiopulmonary arrest indialysis patients
Undetermined Cardiac ArrhythmiaVentricular FibrillationCardiac Arrest During Dialysis
Secondary to cardiac arrhythmiaSecondary to air emboliSecondary to hypotensionbSecondary to dialyzer reactionc
Myocardial InfarctionAsystoleElectromechanical DissociationSepsisRespiratory Arrest Secondary to AspirationCardiac Arrest Secondary to HemorrhageCardiac Arrest Secondary to Pulmonary
EdemaCardiac Arrest Secondary to HyperkalemiaCardiac Arrest Secondary to Fat EmboliTotal
No. ofPatients
0 Numbers in parentheses indicate number of patients surviving to
discharge.b Refractory hypotension in a patient with amyloidosis.
C Allergic reaction to a dialyzer in a patient with a severe cardiomy-
opathy.
d One patient in each group survived 6 months after CPR.
difference in initial successful resuscitation of di-alysis and control group patients (27 [37%] of 74
versus 98 [40%] of 247; P = NS). The mean durationof CPR was significantly less In dialysis and controlgroup patients who were successfully resuscitatedthan in those who were not (dialysis patients, 22 ±1 7 versus 37 ± 1 8 mm; P 0.008; control group, 18± 19 versus 36 ± 21 mm; P < 0.001). There was nosignificant difference In the mean duration of resus-citation for successfully resuscitated dialysis andcontrol group patients.
After CPR, 6 (8%) of 74 (95% confIdence interval,
2 to 14%) dialysis patients, three men and threewomen, survived to hospital discharge comparedwith 30 (12%) of 247 (95% confIdence interval, 8 to
16%) control group patIents (P = NS). None of the
Moss et al
Journal of the American Society of Nephrology I241
TABLE 4. Survival After CPR in Dialysis and ControlPatients
Dialysis(N= 74)
ControlGroup
(N= 247)P
No. of Resuscitations (%) 27 (37) 98 (40) NSNo. of Discharges (%) 6 (8) 30 (12) NSNo. SurvivIng at Six Months (%) 2 (3) 23 (9) 0.044
dialysis patients who were resuscitated at home or inthe dialysis unit survived to discharge. At 6 monthsafter CPR, 2 (3%) of 74 dIalysis patients were stillalive compared with 23 (9%) of 247 controls (P =
0.044).The presence of diabetes did not affect survival to
hospItal discharge after CPR In dialysis or controlgroup patients. Three of 39 diabetic dialysis patientslived to dIscharge compared with 3 of 35 nondlabeticdialysis patients (8 versus 9%: P NS). Six of 51
diabetic control group patients lived to dischargecompared with 24 of 1 76 nondiabetic control grouppatients (12 versus 14%: P = NS).
Four of the six dialysis patients who survived tohospital discharge after CPR died within 4 months ofthe resuscitation. One patIent with end-stage heart
disease who also had diabetes and a prior strokediscontinued dialysis and died within 3 months ofher CPR. The second patient was a 30-yr-old type Idiabetic who had had a previous myocardial infarc-tion and who was found dead In bed 3 months and 1day after resuscitation. The third patient, who alsohad end-stage heart disease, experienced a secondcardIac arrest 3 months and 8 days after his firstcardiac arrest and died 3 days later. The fourth pa-tient, who had arrested after a phenytoin Infusionand who was hospitalized several months later forweight loss, was found dead In bed 4 months and 1
day after resuscitation.Two dialysis patients lived for more than 6 months
after CPR. The first, who had experienced a respira-tory arrest after aspiration, was debIlitated and re-malned in the hospital for 2/2 months after CPR. Hedied 1 4 months after CPR from complications of astroke. His baseline weight was 55 kg, and heweighed 36 kg shortly before his death. The second,who had a respiratory arrest as a complication ofsepsis, remained on mechanical ventilation for 23days and was discharged to a nursing home 3#{189}months later. She has survived 1 9 months, has re-quired oxygen by nasal cannula 24 h/day for hersevere chronic obstructive pulmonary disease, andhas been confined to a wheelchair. She continues tolive in a nursing home, and although she Is happy tobe alive, she has requested not to be resuscitated
again.
Twenty-one (78%) of the 27 dialysIs patients who
were successfully resuscitated did not live to dis-charge: 20 (95%) of 21 requIred Intubatlon and me-chanlcal ventilation for a mean of 4.4 days beforedeath. Only three of the successfully resuscitated
patients who did not hive to discharge had autopsies.so data on other complications of CPR in these pa-
tients are unavailable.
DISCUSSION
ThIs study had several noteworthy findings. First,the percentage of dialysis patients undergoing CPRwas more than that of the control group. It Is possiblethat the previously described Inclination of nephroh-ogists to use CPR more than other Internists, both indialysis patients and In other patients with Impaired
functional status, may account for this fInding (16).Second, the dialysis patients who underwent CPR
were significantly younger than those who did not.Because, in our study, the patients who did not haveCPR had requested not to be resuscitated, thIs fIndingis consistent with reported preferences of older dl-
ahysls patients to decline CPR more often thanyounger ones (13,17).
Third, despite the fact that the dialysis patientswere signIficantly younger than the control group
patients, there was a lower long-term survIval In thedialysis patients. Overall, fewer than 10% of the
dialysis patients survived to hospital discharge andonly 3% were stIll alive 6 months after CPR. Although
many more of the dIalysis patients were diabetics,diabetes did not account for the worse survival ratein dialysis patients. Other comorbid conditions alsodid not account for this difference. It would appearthat the underlying ESRD Is responsible for the poorsurvival after CPR in dialysIs patients.
Even those six dialysIs patients who did survive to
hospItal dIscharge had limited hong-term benefIt from
the CPR. Two thirds were dead within 4 months, andtwo of these had a progressively downhIll course, one
dying after the discontinuation of dialysIs. Both pa-tients who survived longer than 6 months sustaineda marked decrease In their functional status com-
pared wIth prearrest status. These findIngs are sim-ilar to those for nondlalysis patIents who were man-itored for 6 months after resuscItatIon (18).
Fourth, the dialysis patients who were successfullyresuscitated experienced sIgnifIcant short-term mar-
bldlty and mortality. Over three quarters of them diedon average within a week. Of these. 95% were Intu-bated and on mechanical ventilatIon in an Intensivecare unit at the time of death.
SIx lImitations of this study need to be acknowl-edged. The outcomes In fewer than 100 dialysis pa-tients undergoing CPR are reported. It is possIble that
with large numbers a signifIcant decrease in survival
CPI? in Dialysis Patients
I242 Volume 3 . Number 6 ‘ 1992
to hospital discharge might be seen in dialysis pa-tients compared with others. However, this studyreports 8 yr of observation of CPR outcomes and, inthe last 4 yr. there has been no trend toward astatistically significant difference. The hack of differ-ence in either duration of resuscitation or percentageof patients successfully resuscitated supports theobservation that the short-term results of CPR in
dIalysis and control patients do not vary. Yet, despitethe small numbers, this study shows clearly that
dialysis patients undergoing CPR have a significantlyworse 6-month survival compared with that of non-
dialysis patIents.A second limitation is that the outcomes of 247
patients In the control group who all had CPR in thehospital are compared with those of 74 dialysis pa-tlents, 1 1 of wham had CPR in their homes or thedialysIs unit. Despite this dissimilarity in location,the percentage of patients successfully resuscitatedand surviving to discharge did not differ between the
two groups.Third, It could be argued that the how survival rates
for dialysis and control group patients are the resultof poor CPR technique at West Virginia University
Hospitals. However, the survival to discharge and 6-month survival of our control group patients are
comparable to those previously reported (18).
Fourth, in this study, the outcomes of CPR in di-alysls patients aver an 8-yr period, from 1 983 to1 99 1 , are compared with those far control patientsover a more recent part of this period. from 1987 to1 989. It would be desirable if the dialysis patientsand control patients undergoing CPR were comparedover the same period of time, but West Virginia Uni-versity Hospitals’ system for tracking all resuscita-
tions was Implemented in 1 987 and a retrospectiveanalysis of CPR far the years before 1 987 would notbe accurate. Ta examine this limitation, the age,gender. and percentage of dialysis patients undergo-Ing CPR after 1 986 were compared with those ofdIalysis patients undergoing CPR in the earlier years
of the study. There were no differences in age, gen-der, or rate of CPR, and the percentage of patients
surviving to discharge were similar: 9% for the years1983 to 1986 and 7% for the years 1987 to 1991.The 6-month survival far bath groups was very law:none of the 33 patients In the earlier years survived,and 2 (5%) of 41 patIents in the later years survived.
Fifth, although it was possible to verify the dialysis
unit records of deaths and CPR with hospital records,It Is likely that same control patients undergoing CPRduring the study period were missed because hospital
CPR documentation forms were not submitted.Sixth, multiple statistical comparisons were made
in thIs study and same of the results may have arisenby chance alone.
Despite the paucity of information about CPR out-
comes in dialysis patients and the inference from theexperience of CPR in renal failure patients that theexpected outcomes would be poor, mast dialysis pa-tients indicate that they want CPR in the event ofcardiopulmonary arrest (13,14,17,19).
Our data show that CPR is a procedure that rarelyresults in extended survival far dialysis patients.Kjehlstrand has written that nephrolagists need to be
more open with their patients and more realisticabout what they can and cannot do (20). Such dis-cussians need to contain useful information far pa-tients that allows them to make informed decisions,and nephralogists have been encouraged to obtainadvance directives from patients as part of thesediscussions (2 1 ,22). The poor outcomes of CPR indialysis patients and the automatic presumption that
CPR will be performed unless a “Do Nat Resuscitate”order has been written should be included in theinformation that nephrolagists provide during can-versatians about CPR with their dialysis patients. Inthis way, nephrahagists can assist patients in making
informed decisions about their future treatment.
ACKNOWLEDGMENTS
The authors thank Mary Hoizer, RN.. Jennifer Domico, RN.. John
L. Burkhard, Sue Ann Crowder Upton, M.D. , the nurses of the West
Virginia Health Care Cooperative, and Conrad Pesyna of the Medical
Records Department of West Virginia University Hospitals for as-
sistance in data collection; Paul B. Hshieh, Ph.D. , for assistance in
statistical analysis; Cynthia F. Jamison for assistance in manuscript
preparation; and two anonymous reviewers for helpful comments.
A. H. Moss is supported in part by a grant from the Henry J. Kaiser
Family Foundation and the Pew Charitable Trusts. The opinions
expressed are those of the authors and should not be taken to
represent necessarily those of the supporting groups.
REFERENCES
1 . McGrath RB: In-house cardiopulmonary resus-citation-after a quarter of a century. AnnEmerg Med 1987:16:1365-1368.
2. Applebaum GE, King JE, Finucane TE: Theoutcome of CPR initiated in nursing homes. JAm GeriatrSac 1990:38:197-200.
3. Murphy DJ, Murray AM, Robinson BE, Cam-pion EW: Outcomes of cardiopulmonary resus-citation in the elderly. Ann Intern Med 1989;111:199-205.
4. Faber-Langendoen K: Resuscitation of patientswith metastatic cancer: Is transient benefit stillfutile? Arch Intern Med 1 99 1 ; 151:235-239.
5. Blackhall U: Must we always use CPR? N EnglJ Med 1987:317:1281-1285.
6. Moss AH: Informing the patient about cardio-pulmonary resuscitation: When the risks out-weigh the benefits. J Gen Intern Med 1989:4:346-355.
7. Saphir R: External cardiac massage: Prospectiveanalysis of 123 cases and review of the hitera-ture. Medicine (Baltimore) 1968:47:73-87.
8. Hollingsworth JH: The results of cardiapulma-nary resuscitation: A 3-year university hospitalexperience. Ann Intern Med 1969:71:459-466.
Moss et al
Journal of the American Society of Nephrology I243
9. Kyff J, Pun VK, Raheja R, Ireland T: Cardlo-pulmonary resuscitatIon In hospitalized pa-tients: Continuing problems of decision-making.Crit Care Med 1987;15:41-43.
1 0. Scott RPF: CardIopulmonary resuscitatIon in ateaching hospital. Anaesthesia 1981:36:526-530.
1 1 . Klassen GA, Broadhurst C, Peretz DI, JohnsonAL: Cardiac resuscitation In 1 26 medIcal pa-tlents using external cardiac massage. Lancetl963;l:1290-1292.
1 2. Johnson AL, Tanser PH, Ulan RA, Wood TE:Results of cardiac resuscitation in 522 patients.Am J Cardlol 1967:20:832-835.
1 3. Tzamaloukas AH, Zager PG, Quintana BJ, Nev-arez M, Rogers K, Murata GH: Mechanical car-diopulmonary resuscitation choice of patients onchronic peritoneal dialysis. Peritoneal Dial Int1990:10:299-302.
14. Holley JL, Finucane TE, Moss AH: Dialysis pa-tients’ attitudes about cardiopulmonary resus-citatIon and stopping dialysis. Am J Nephrol1 989;9:245-25 1.
1 5. Standards and guidelines for cardiopulmonaryresuscitation (CPR) and emergency cardiac
care (ECC). JAMA 1986:255:2905-2989.16. Foulks CJ, Holley JL, Moss AH: The use of
cardiopulmonary resuscitation: how nephralo-gists and Internists dIffer. Am J Kidney Dis1991:18:379-383.
1 7. Quintana BJ, Nevarez M, Rogers K, Murata GH,Tzamaloukas AH: Reaction of patients onchronic dialysis to dIscussions about cardlopul-monary resuscitation. ANNA J 1991:18:29-32.
1 8. Bedell SE, Delbanco TL, Cook EF, Epstein FH:Survival after cardiopulmonary resuscitation Inthe hospital. N Engl J Med 1983;309:569-576.
1 9. Rogers K, Crawford DR. Tzamaloukas AH: Car-diopulmonary resuscitation preferences of pa-tients on chronic hemodlalysis. Dial Transplant1990: 19: 182-184.
20. Kjellstrand CM: Who should decide about yourdeath? JAMA 1992:267:103-104.
2 1 . Singer PA: Nephrohogists’ experience with andattitudes towards decisions to forego dialysis. JAm Soc Nephrol 1992:2:1235-1240.
22. Sehgal AS, Galbraith A, Chesney M, Schoen-feld P, Charles G, Lo B: How strictly do dialysispatients want their advance directives followed?JAMA 1992:267:59-63.