Surgery and Anesthesia in Sickle Cell Disease

By Mabel Koshy, Steven J . Weiner, Scott T. Miller, Lynn A. Sleeper, Elliott Vichinsky, Audrey K. Brown,
Yusuf Khakoo, Thomas R. Kinney, and The Cooperative Study of Sickle Cell Disease
From 1978 to 1988, The Cooperative Study of Sickle Cell
Disease observed 3,765 patients with a mean follow-up of
5.3 k 2.0 years. One thousand seventy-nine surgical proce-
dures were conducted on 717 patients (77% sickle cell ane-
mia [SS], 14% sickle hemoglobin C disease [SC], 5.7% Spo
thalassemia, 39'0 Sp+ thalassemia). Sixty-nine percent had a
single procedure, 219'0 had two procedures, and the re-
maining 11% had more than two procedures during the
study follow-up. The most frequent procedure was abdomi-
nal surgery for cholecystectomy or splenectomy (2496 of all
surgical procedures, N =258). Of these, 939'0 received blood
transfusion, and there was no association between preoper-
ative hemoglobin A level and complication rates (except re-
duction in pain crisis). Overall mortality within 30 days of a
surgical procedure was 1.1% (12 deaths after 1,079 surgical
procedures). Three deaths were considered to be related to
the surgical procedure andlor anesthesia (0.3%). No deaths
were reported in patients younger than 14 years of age.
ATIENTS WITH sickle cell disease (SCD; which in-
cludes sickle cell anemia [SS], sickle hemoglobin C
disease [SC], and the sickle 0 thalassemias) who undergo
surgery are generally considered to be at greater risk for
perioperative complications than otherwise healthy patients
without this hematologic disorder.'"' Favorable outcomes
have been reported without transfusion, but perioperative
transfusion is commonly used to prepare SCD patients for
surgery and to treat complications of sickle cell di sea~e.4,~.~~"~
Although the optimal level of sickle hemoglobin (Hb) to be
achieved is unknown, most sickle cell centers adhere to some
formof transfusion protocol for SCD patients undergoing
The Cooperative Study of Sickle Cell Disease (CSSCD)
was a natural history study that observed 3,765 patients from
1978 to 1988.26.27 This report analyzes the course and out-
come of the 1,079 surgical procedures performed on 717
patients during this time period.
P
sUrgery.l.18.20-25
From the University of Illinois, Chicago, IL; New England Re-
search Institutes, Watertown, MA; the Department of Pediatrics,
State University of New York Health Science Center at Brooklyn,
Brooklyn, NY; the Department of Pediatrics, Children's Hospital,
Oakland, CA; the Columbia University College of Physicians and
Surgeons and Harlem Hospital Center, New York, NY; the Depart-
ment of Pediatrics, Duke University Medical Center, Durham, NC.
Submitted February 21, 1995; accepted July 3, 1995.
Supported by the Division of Blood Diseases and Resources of
the National Heart, Lung, and Blood Institute of the National Insti-
tutes of Health.
Address reprint requests to Mabel Koshy, MD, University of Illi-
nois at Chicago, Division of Hematology (M/C 787), 840 S Wood
St, Room 314 CSB, Chicago, IL 60612.
The publication costs of this article were defrayed in part by page
charge payment. This article must therefore be hereby marked
"advertisement" in accordance with 18 U.S.C. section 1734 solely to
indicate this fact.
0 1995 by The American Society of Hematology.
0006-4971/95/8610-0012$3.00/0
3676
Sickle cell disease (SCD)-related complications after surgery
were more frequent in SS patients who received regional
compared with general anesthesia (adjusted for risk level of
the surgical procedure, patient age, and preoperative trans-
fusion status, P =.058). Non-SCD-related postoperative
complications were higher in both SS and SC patients who
received regional compared with those who received gen-
eral anesthesia (P =.095). Perioperative transfusion was as-
sociated with a lower rate of SCD-related postoperative
complications for SS patients undergoing low-risk proce-
dures (P =.006, adjusted for age and type of anesthesia),
with crude rates of 12.9% without transfusion compared
with 4.8% with transfusion. In SC patients, preoperative
transfusion was beneficial for all surgical risk levels ( P =
.0091. Thus, surgical procedures can be performed safely in
patients with SCD.
0 1995 by The American Society of Hematology.
MATERIALS AND METHODS
Patients
The goals, objectives, design, and enrollment procedures of the
Cooperative Study of Sickle Cell Disease have been described else-
where.26-28 FromOctober 1978 to October 1988, 3,765 patients from
23 clinical centers across the continental United States participated
in the CSSCD. The median length of patient follow-up was 6.0
years. Of the total study cohort, 67.5% of the patients had SS, 22.4%
had SC, 5.0% had sickle Po thalassemia (Soo thal), and 5.1% had
sickle p' thalassemia (So' thal). The Hb phenotype was established
by the Centers for Disease Control using standard laboratory
methods.
Data Collection
A standardized CSSCD data collection form was completed each
time a patient underwent a surgical procedure. Because of limitations
in data collection procedures, only one surgical procedure was re-
corded per operation. Medical history, laboratory data, and perioper-
ative course were recorded on this form. Separate forms were com-
pleted documenting blood transfusions and acute and chronic clinical
events. Transfusion data for the 30-day period before surgery and
acute event data for the 7-day period after surgery were used in this
report. All deaths within 30 days after a surgical procedure are
summarized in this report. Deaths occurring within 7 days of surgery
were defined as postoperative complications.
Class$cation of Surgeries, Anesthesia, Transfusion, and
Complications
Surgeries were defined using the International Classification of
Diseases (9th revision) diagnosis codes for procedures. For the pur-
poses of analysis, surgical procedures were categorized into three
groups by level of risk low, moderate, and high.29 Low-risk proce-
dures are those of the eyes, skin, nose, ears, and distal extremities
as well as those pertaining to the dental, perineal, and inguinal
areas (eg, inguinal hernia repair, myringotomy, and dilatation and
curettage). Moderate-risk procedures are those of the throat, neck,
spine, proximal extremities, genitourinary system, and intra-abdomi-
nal areas, such as tonsillectomy, Cesarean section, splenectomy, cho-
lecystectomy, and hip replacement. High-risk procedures are those
pertaining to the intracranial, cardiovascular, and intrathoracic sys-
tems (eg, craniotomy and heart valve replacement).
Blood, Vol 86, No 10 (November 15). 1995: pp 3676-3684
RISKS OF SURGERY AND ANESTHESIA I N SICKLE CELL DISEASE
3677
In addition to overall analyses stratified by risk level, six specific
classes of the most common surgical procedures were analyzed (1)
cholecystectomy or splenectomy (N =222 and 36, respectively;
23.9% of all surgical procedures); (2) dilation and curettage (N =
97; 9.0%); (3) Cesarean section or hysterectomy (N =87; 8.1%);
(4) tonsillectomy and/or adenoidectomy (N =46; 4.3%); ( 5 ) hip
replacement, removal, or revision (N =4 4 ; 4.1%); and (6) myrin-
gotomy (N =30; 2.8%). The frequencies of the remaining surgical
procedures (47.8%) are listed in the Appendix.
Anesthesia was classified as general, regional, and local. General
anesthesia refers to that induced by the inhalation of gas and balanced
intravenous methods. Regional anesthesia refers to spinal, epidural,
and nerve block anesthesia. The type of anesthesia and its method
of administration were not prescribed by the CSSCD protocol.
Patients were defined as preoperatively transfused if at least one
transfusion was administered within 30 days before surgery. Postop-
erative complication rates for patients who were perioperatively (ei-
ther preoperatively or intraoperatively) transfused were compared
with those who were not transfused. Total Hb concentrations and
Hb A percentages presented in this report were obtained after trans-
fusion and before surgery. The CSSCD was a natural history study,
thus no protocol was specified for perioperative management; all
patients were treated according to institutional practices.
Postoperative complications were defined as complications that
occurred within 7 days after surgery. These complications were cate-
gorized into three groups: (1) SCD-related, (2) non-SCD-related,
and (3) other. SCD-related complications were defined as painful
crisis, acute chest syndrome (ACS), and cerebrovascular accident
(CVA). Non-SCD-related complications were defined as fever, in-
fection (excluding ACS), bleeding, thrombosis, embolism, and death.
Other postoperative complications included transfusion reactions and
unspecified complications. Painful crisis was defined as pain in the
extremities, back, abdomen, chest, or head for which no other expla-
nation (eg, osteomyelitis or appendicitis) could be found. ACS was
defined as the new appearance of an infiltrate on chest radiograph
or abnormalities on a radioisotope lung scan in the presence of
symptoms.
Statistical Methods
Because many patients underwent more than one surgical proce-
dure, the surgery served as theunit of analysis; eg, percentages
reported refer to the percentage of surgeries with a particular charac-
teristic, rather than the percentage of patients. Complication rates are
computed as the number of surgeries with a particular complication
divided by the toal number of surgeries. Descriptive statistics are
presented as percentages and means -C 1 standard deviation. All
hypothesis tests and confidence intervals are two-sided. A two-sided
P value of .05 or less was considered to be a statistically significant
result. Postoperative complication rates with and without periopera-
tive transfusion were compared using logistic regression, with adjust-
ments for phenotype, type of anesthesia, surgical risk level, and
age. The logistic regression model provided robust standard error
estimates for the model parameters that accounted for the correlation
between different surgical procedures on the same patient.30 The
association between postoperative complication rates and anesthesia
was also examined using logistic regression. Where there were suf-
ficient data, the association between postoperative complication rates
and (1) total Hb concentration and (2) HbA percentage was exam-
ined using logistic regression. Mean total Hb concentrations of pa-
tients with and without postoperative complications were compared
using the Students t-test.
RESULTS
General Characteristics
There were 717 patients who had one or more surgical
procedures. Of these patients, 77.1% (N =553) were SS,
14.2% (N =102) were SC, 5.7% (N =41) were Spa thal, and
2.9% (N =21) were Sp+ thal. This group has proportionately
more SS and fewer SC and SO+ thal patients than the total
CSSCD cohort (see Materials and Methods). Sixty-nine per-
cent (N =495) of the patients underwent one, 20.5% (N =
147) had two, 6.0% (N =43) had three, and 4.5% (N =32)
had four or more surgical procedures during the follow-up
period. Forty-eight percent (N =520) of the 1,079 reported
surgical procedures were classified as low-risk, 50% (N =
543) as moderate-risk, and 2% (N =16) as high-risk proce-
dures. Seventy-five percent (N =806) were elective and
25% (N =271) were emergent surgical procedures. The
mean age at the time of surgery was 22.0 2 11.6 years, with
17% being less than 10 years of age, 25% being 10 to 19
years of age, 52% being 20 to 39 years of age, and 6% being
40 years of age and older. Female patients underwent 6 1 %
(N =660) of the surgical procedures and male patients
underwent the remaining 39% (N =419).
The reported sample includes surgical procedures of dif-
fering risk levels performed on patients of varying ages. The
risk of postoperative complications significantly increased
with age (estimated odds ratio, 1.3 times increased risk of
postoperative complications per 10 years of age, P <.mol ).
Comparisons of postoperative complication rates were there-
fore adjusted for patient age as well as surgical risk level to
correct for the potential confounding effects of these two
factors.
Postoperative Deaths
There were 12 postoperative deaths (10 SS, 1 SC, and 1
Spa thal) within 30 days of a surgical procedure (Table 1).
Notably, there were no deaths among patients under 14 years
of age and only 2 in patients between 14 and 20 years of
age. The mean age at death was 27.4 +- 10.4 years (range,
14 to 54 years). Eleven of these patients were transfused.
There were 8 minor and 4 major intra-abdominal procedures.
The deaths in the first 9 patients appear to be related to
comorbid medical complications and SCD-related multior-
gan failure. The deaths in the remaining 3 patients appear
to be related to the surgical procedure: patient no. 10 from
profound anemia secondary to delayed transfusion reaction;
patient no. 11 from an intra-abdominal hemorrhage requiring
18 U of packed red blood cells; and patient no. 12 from
rupture of the prosthetic mitral valve replaced 38 days earlier
(the surgical procedure was a diagnostic right heart catheter-
ization 1 day before death).
The overall 30-day postoperative mortality rate was 1 .l%
(12 deaths of 1,079 surgical procedures). The actual mortal-
ity rate of the 3 deaths related to the surgical procedure was
only 0.3%.
A Projile of Most Frequently Performed Procedures
A description of six classes of surgical procedures most
frequently performed during the course of the CSSCD is
displayed in Table 2. There were few procedures performed
on SC patients; thus, no formal statistical comparisons of
the outcome of SS and SC patients were made. Analyses
of the relationship between preoperative Hb A level and
3678
KOSHY ET AL
Table 1. Postoperative Deaths Occurring Within 30 Days of Surgery
No. of Units of Preop
Patient Age/ Hb Clinical Status at
Surgery
PRBC Transfused HbA to Death
No. Sex Diag. Surgery Surgical Procedure Within 30 d of Death % Anesthesia (d) Cause of Death
1
2
3
4
5
6
7
8
9
10
11
12
21lF
36lF
28lF
27lF
23lM
2OlF
54/F
17lF
26/F
14lF
32lM
2 4lF
ss
ss
ss
ss
ss
ss
ss
ss
sc
ss
SO0
ss
Cirrhosis DIC,
MOF, CRF
Fever
MOFlsepsis, CRFI
DIC peritonitis1
pneumonia
CRF/SP renal
transplant
CRF infected graft,
CVA, comatose
Acute abdomen,
Bili 26, CHF
Sepsislacidosis,
SepsisIMOF
Renal abscess
Liver disease, Bili
dialysis
MOF, DIC, ARF
30, sepsis
MV replacement
Bronchoscopy,
Tendon repair
Tenkhoff catheter
mediastinoscopy
Clotted venous graft
removal
Graft removal
Dental extraction
Exploratory cholecyst.
Exploratory laparotomy
A-V fistula
Drainage
Cholecystectomy
RH catheter
2
2
30
4
Dialysis
0
5
10
0
6
18
0
-
-
-
95
-
70
-
71
53
90
-
17
lnhal
lnhal
lnhal
Local
Local
Local
lnhal
lnhal
Local
lnhal
lnhal
Local
18 ARDS
26 ARF
9 Sepsis, DIC
15 DOA
29 Sepsis, M. TB
12 DOA
4 ARF
2 DIC, Sepsis
5 Sepsis
8 DTR, severe anemia
2 Intra-abdominal
hemorrhage
1 (38) MV rupture
Abbreviations: DIC, disseminated IN coagulation; ARF, acute renal failure; DOA, dead on arrival; MOF, multiorgan failure; TB, tuberculosis;
ARDS, adult respiratory distress syndrome; CRF, chronic renal failure; DTR, delayed transfusion reaction; PRBC, packed RBCs; MV, mitral valve;
CHF, congestive heart failure.
postoperative complications were conducted only for two
of the groups, ie, abdominal surgery (cholecystectomy and
splenectomy) and orthopedic procedures of the hip. Only for
these two groups were posttransfusion Hb A data available
for at least 80% of the surgical procedures on SS patients.
Cholecystectomy and splenectomy. Patients undergoing
open cholecystectomy had a mean age of 23.3 2 11. l years
(range, 5 to 64 years). Patients undergoing splenectomy had
a mean age of 7.8 8.5 years (range, 9 months to 30 years).
All procedures were performed under general anesthesia and
the majority of patients were preoperatively transfused. The
rate of SCD-related postoperative complications was similar
for S S and SC patients (8% and 9%, respectively). Rates of
non-SCD-related postoperative complications were 11% for
S S and 23% for SC patients. The most frequent non-SCD-
related complications were fever and infection (other
than ACS).
There was no difference in the overall rates of postoperative
complications in 203 transfused versus 13 untransfused SS
patients (21.6% v 33.3%, P =.229). When examined sepa-
rately, SCD-related and non-SCD-related complication rates
were again similar for transfused and untransfused SS patients.
Table 2. Profile of Six Surgical Procedures
Cholecystectomy
Hip Replacement,
Tonsillectomy Revision, and
and Dilation and Cesarean Section and Prosthesis
Splenectomy Curettage
and Hysterectomy
Adenoidectomy
Removal
Myringotomy
ss sc ss sc ss sc ss sc ss sc ss sc
N
Risk level
Mean age (yr)
Age range (yr)
Emergent (%)
Preoperative transfusion (%)
Perioperative transfusion (%)
General anesthetic (%)
Postoperative complications
SCD-related (%)
Non-SCD-related (%)
Other (%)
Any complications (%)
218 22
Moderate
20.7 28.3
0, 59 7, 64
18.4 13.6
94.0 81.8
94.5 81.8
100.0 100.0
7.8 9.1
11.0 22.7
7.9 4.6
22.2 36.4
70 14
Low
23.5 26.2
16,34 18,35
21.4 35.7
42.9 7.1
44.3 7.1
60.9 64.3
18.6 14.3
15.7 0.0
2.9 0.0
27.5 14.3
65 18
Moderate
25.9 27.7
17, 46 16,39
64.6 72.2
81.5 61.1
90.8 72.2
77.8 72.2
16.9 11.1
26.2 33.3
13.9 11.1
41.5 50.0
35 7
Moderate
11.4 15.5
2, 22 1, 31
5.7 0.0
82.9 100.0
82.9 100.0
100.0 100.0
0.0 14.3
5.7 14.3
2.9 0.0
5.7 14.3
34 6
Moderate
28.1 41.4
9, 46 25, 62
5.9 0.0
97.1 100.0
100.0 100.0
97.0 100.0
2.9 0.0
14.7 33.3
0.0 16.7
17.7 33.3
26 3
Low
9.1 5.3
0, 16 2, 7
3.9 0.0
53.8 33.3
53.8 33.3
100.0 100.0
3.9 0.0
7.7 0.0
0.0 0.0
11.5 0.0
RISKS OF SURGERY AND ANESTHESIA IN SICKLE CELL DISEASE
3679
Among preoperatively transfused SS patients, the risk of
postoperative painful crisis decreased with increasing levels
of Hb A (P =.054). The mean Hb A percentage in those
without pain (N =161) was 54.8% 2 23.3%, compared with
34.7 2 29.5% in those with pain (N =6). However, among
all S S patients (transfused and untransfused), total Hb con-
centration did not differ for those with and without postoper-
ative painful crisis.
There was no association between the development of
postoperative ACS in preoperatively transfused SS patients
and the mean Hb A percentage (P =.854). However, the
total Hb concentration of 7 Ss patients with postoperative
ACS was significantly lower than that of 205 S S patients
without ACS (9.3 2.3 v 11. 1 % 2.1 g/dL, P =.024),
even after adjustment for age. The risk of postoperative ACS
increased as total Hb concentration decreased (estimated
odds ratio, 1.7 times increased risk of ACS with each 1 g/
dL, P =.016).
Dilation and curettage. Only 44% of 70 S S patients and
7% of 14 SC patients received blood transfusion. There was
no difference in the overall rate of postoperative complica-
tions in SS patients by transfusion status (23.7% of N =38
untransfused v 32.3% of N =31 transfused, P =.662).
Except for 2 patients who developed ACS (1 patient was
transfused and the other not), all SCD-related complications
in S S and SC patients were painful crisis. The most common
non-SCD-related complications in S S patients were fever
and infection; none was reported in SC patients.
Cesarean section and hysterectomy. These procedures
were analyzed together. Cesarean sections comprised 55 of
the 65 S S procedures (85%) and 16 of the 18 SC procedures
(89%). The mean patient age was 25.3 2 5.3 years for Cesar-
ean section and 32.7 % 7.9 years for hysterectomy. Seventy-
three percent of the Cesarean sections were emergent.
Ninety-one percent of the SS and 72% of the SC patients
were transfused. The postoperative complication rates were
high (42% for SS patients and 50% for SC patients). Al-
though only 6 SS patients were untransfused, the overall
complication rate in SS patients did not differ by transfusion
status (44.1% of transfused v 16.7% of untransfused, P =
.793). The mean total Hb concentration of all SS patients was
not associated with the presence or absence of postoperative
complications (P =.836).
Tonsillectomy and adenoidectomy. All 7 SC patients and
83% of the 35 S S patients were transfused. Postoperative
complication rates were fairly low (6% for SS and 14%
for SC patients), with no SCD-related complications in SS
patients. The postoperative complication rates were similar
in transfused (N =29) and untransfused (N =6) SS patients
(3.5% of transfused v 16.7% of untransfused, P =.318).
Hip replacement, revision, and prosthesis removal.
Thirty-four patients were preoperatively transfused, and one
was transfused during the intraoperative period. The rate of
non-SCD-related complications was 15% for S S and 29%
for SC patients. The only SCD-related complication was
painful crisis in 1 SS patient. The mean Hb A percentage
was 55.3% 2 20.6% with no postoperative complications
(N =26) and 69.6% -C 9.8% with postoperative complica-
tions (N =5; P =.142) among the preoperatively transfused
S S patients.
Myringotomy. Three SC patients and 26 SS patients un-
derwent myringotomy. One SC patient and 54% of the SS
patients were transfused. There was one SCD-related com-
plication, ie, a cerebrovascular accident in a chronically
transfused SS patient with previous CVA. Non-SCD-related
complications (fever and bleeding) occurred in 2 other SS
patients (8%). There was no association between the overall
postoperative complication rate and transfusion status.
Anesthesia and Postoperative Complications
For the low-risk surgical procedures, 73.8%, 10.6%, and
15.6% were performed under general, regional, and local
anesthesia, respectively. For the moderate-risk procedures,
93.0%, 6.1%, and 1.0%, were performed under general, re-
gional, and local anesthesia, respectively. Thirteen of the 15
high-risk surgical procedures were performed under general
(86.7%) and 2 were performed under local anesthesia (drain-
age of a brain abscess and cardiac catheterization followed
by balloon valvuloplasty). The anesthesia data on 45 surgical
procedures (4.2%) were incomplete and excluded from anal-
ysis. The effect of type of anesthesia on postoperative com-
plication rates was examined with adjustment for Hb pheno-
type, age, surgery risk level (low v moderate), and
transfusion status. Crude postoperative complication rates
for S S patients are displayed in Fig 1.
Non-SCD-related complications. The most common
non-SCD-related postoperative complication was fever.
There was a marginally significant effect of anesthesia on
non-SCD-related postoperative complication rates (P =
.095). This effect was similar for SS and SC patients (P
=.199). These complication rates were lower for surgical
procedures with general anesthesia compared with those with
regional anesthesia (estimated odds ratio, 0.58; P =.095)
and compared with those with local anesthesia (estimated
odds ratio, 0.51; P =.100).
SCD-related complications. Painful crisis was the most
common SCD-related postoperative complication. Among
SS patients, the complication rate was associated with type
of anesthesia (P =.030), and rates were higher for surgical
procedures with regional anesthesia compared with those
with general anesthesia (estimated odds ratio, 2.32; P =
.058) and with those with local anesthesia (estimated odds
ratio, 4.65; P =.014). Among SC patients, complication
rates after general versus regional anesthesia did not differ
(P =%l), but this may be due to the small number of SC
patients who received regional anesthesia (N =18).
Perioperative Transhsion and Postoperative
Complications
The effect of blood transfusion on postoperative complica-
tion rates was examined with adjustment for Hb phenotype,
age, surgery risk level (low v moderate), and type of anesthe-
sia. Tables 3 and 4 present postoperative complication rates
by surgical risk level and perioperative transfusion status.
There were no complications in the 3 SC patients (2 trans-
fused and 1 untransfused) undergoing high-risk surgery (not
shown in Table 4).
Non-SCD-related complications. For non-SCD-re-
3680
25
U
C
0
23.8
f 20
U
.-
-
n
V
5 15 14.3 14 5
135
B
: 10
2
z 5
L
8.3
V
0
z
0
0.0
LOW Moderate
Risk Level of Surgical Procedure
, 6.9 6.5
LOW
Risk Level of Surgical Procedure
Moderate
KOSHY ET AL
Fig 1. Postoperative complication rate in SS pa-
tients by type of anesthesia. IO1General anesthesia;
1. 1 regional anesthesia; IO) local anesthesia.
lated postoperative complications, the effect of transfusion erative complications than did untransfused patients ( P =
depended on both surgery risk level and phenotype ( P = .004). The crude rates of non-SCD-related postoperative
.015). There was a significant effect of transfusion only in complications were 28.1 % for surgical procedures with peri-
SC patients undergoing low-risk procedures. When perioper-
operative transfusion and 2.1% for those without.
atively transfused, these patients had higher rates of postop- SCD-related complications. For S S patients undergoing
Table 3. PostoDerative ComDlication Rates bv PerioDerative Transfusion Status SS Surgeries
Low-Risk Moderate-Risk High-Risk
No TX TX No TX TX No TX TX
Postoperative Complications (%) (N =145) (N =248) (N =43) (N =390) (N =0) (N =12)
Pain
ACS
CVA
Any SCD complications
Non-ACS infection
Fever
Bleeding
Thrombosis
Embolism
Death
Any non-SCD complications
Other
Any postoperative complications
12.4
1.4
0.0
12.9
2.0
6.2
0.7
0.0
0.0
1.4
8.8
1.4
18.6
4.4
0.8
0.4
4.8
3.6
8.9
2.0
0.0
0.4
0.4
11.6
5.2
17.3
2.3
2.3
0.0
4.7
2.3
11.6
0.0
0.0
0.0
0.0
14.0
2.3
18.6
5.4
2.8
0.0
7.9
4.9
9.7
3.1
0.3
0.3
0.5
13.8
7.7
23.9
8.3
8.3
0.0
16.7
0.0
25.0
8.3
0.0
0.0
0.0
33.3
8.3
41.7
Of the SS surgeries, 4 low-risk and 2 moderate-risk surgeries are missing a response to the postoperative complication question.
~~
RISKS OF SURGERY AND ANESTHESIA IN SICKLE CELL DISEASE
3681
Table 4. Postoperative Cor npl i i i on Rates by Parioporative
Transfusion Status SC Surgeries
Low-Risk Moderate-Risk
NoTX TX NoTX TX
Postoperative Complications (N =48) (N =32) (N =21) (N =49)
Pain 8.3 0.0 9.5 4.1
ACS 0.0 0.0 14.3 0.0
CVA 0.0 0.0 0.0 0.0
Any SCD complications 8.3 0.0 19.1 4.1
Non-ACS infection 0.0 0.0 4.8 10.2
Fever 0.0 21.9 28.6 10.2
Bleeding
0.0 6.3 0.0 0.0
Thrombosis 0.0 0.0 0.0 0.0
Embolism 0.0 0.0 4.8 2.0
Death 2.1 0.0 0.0 0.0
Any non-SCD complications 2.1 28.1 28.6 16.3
Other 2.1 3.1 0.0 8.2
Any postoperative
complications 12.5 31.3 42.9 24.5
Of the SC surgeries, 1 low-risk surgery is missing a response to the
postoperative complication question.
low-risk surgical procedures, there was a beneficial effect of
transfusion (estimated odds ratio for untransfused v trans-
fused, 3.28; P =.006) on SCD-related postoperative compli-
cations, particularly painful crisis. The crude complication
rates for S S patients undergoing low-risk surgical procedures
were 12.9% for surgical procedures without transfusion and
4.8% for those with transfusion. Accordingly, the mean total
Hb concentration of transfused patients was significantly
higher (10.3 5 2.3 g/&) than that of untransfused patients
(8.4 -t 1.3 g/&; P =.OOOl). However, among SS patients
undergoing moderate-risk surgical procedures, no associa-
tion was found between transfusion and SCD-related postop-
erative complications ( P =3 1 ) . Mean total Hb concentra-
tions of S S patients undergoing moderate-risk surgical
procedures were similar for those with and without postoper-
ative painful crisis (10.6 2 1.9 v 10.5 ? 2.3 g/dL, P =
383). In contrast, the mean total Hb of 12 S S patients who
developed ACS was 9.3 ? 2.0 g/dL, compared with 10.5 rt
2.3 g/& for 413 SS patients who did not develop postopera-
tive ACS (P =.055). However, this difference did not remain
after adjusting for age.
Among SC patients there was a beneficial effect of periop-
erative transfusion regardless of surgery risk level (estimated
odds ratio for untransfused v transfused, 8.33; P =.009).
The crude rate of complications after low- and moderate-
risk surgeries was 2.5% with perioperative transfusion and
11.6% without. The mean total Hb concentration was 11.3
5 1.6 g/dL for untransfused SC patients and 12.0 ? 2.3 g/
dL for perioperatively transfused SC patients. Notably, 3
of 21 untransfused SC patients undergoing moderate-risk
surgical procedures had ACS, but none of the 49 transfused
patients in this group did.
Sf l o Thal and Sf l + Thal Patients
Fifty-two surgical procedures were performed on 41 S$
thal patients. There were 11 cholecystectomies, 5 splenecto-
mies, 2 hip replacements/revisions, 8 dilation and curettage
procedures, 4 circumcisions, 2 laparotomies, 2 laparoscopic
evaluations, 2 tubal ligations, 1 Cesarean section, 1 tonsillec-
tomy, 1 adenoidectomy, and 13 other surgical procedures.
The SCD-related postoperative complication rate was 9.6%
(3 painful crises, 1 ACS, and 1 with both). The non-SCD-
related complication rate was 11 S%.
Thirteen of 16 Soo thal patients were perioperatively trans-
fused for cholecystectomy or splenectomy (81.3%). Of these
transfused patients, 1 patient had fatal postoperative bleeding
(see Postoperative Deaths), and another had a painful crisis,
ACS, fever, and thrombosis after the surgical procedure. No
complications were reported for the 3 untransfused patients.
Twenty-nine surgical procedures were performed on 21
S o + thal patients. There were 2 splenectomies, 3 Cesarean
sections, 2 tonsillectomies, 5 dilation and curettage proce-
dures, 2 inguinal hernia repairs, 1 myringotomy, 1 hip revi-
sion, and 13 other surgical procedures. The perioperative
transfusion rate was 41.4%. The only SCD-related postoper-
ative complication was 1 ACS event after splenectomy with
intraoperative transfusion (preoperative total Hb, 13.2 g/dL).
No non-SCD-related complications were reported.
DISCUSSION
Because previous studies have shown significant compli-
cations for SCD patients undergoing surgical procedures,
most centers follow protocols for use of preoperative prepar-
reported a paucity of morbidity data in untransfused patients
undergoing minor These increased risks are
believed to be secondary both to acute tissue injury and
chronic organ damage produced by vaso-occlusion from
sickled red blood cells (RBCs). In addition, surgical proce-
dures may be complicated by hypoxia, acidosis, or hypother-
mia, adding to a greater likelihood of SCD patients experi-
encing an adverse event, because each of these factors
promotes erythrocyte sickling. Although RBC transfusions
have been advocated by many investigators to reduce periop-
erative complications, there are no controlled trials docu-
menting their benefit. There are also significant risks associ-
ated with RBC transfusions, including alloimmunization and
exposure to infectious diseases.
The CSSCD was designed to define the natural history
of SCD and its effects on health events. The recruitment
procedures insured that participants were representative of
the wide spectrum of clinical severity that is a hallmark of
this disease. Because this unique population was observed
prospectively for a median of 6 years, it is ideally suited
to define the types of surgical procedures and associated
complications that can be observed in patients with SCD.
There were several interesting observations noted in this
study. The overall mortality rate was very low (0.3%). There
were only 3 deaths attributed to the surgery or anesthesia.
No deaths were observed in patients under the age of 14
years, although many children had procedures associated
with moderate surgical risk.
The SCD-related complication rates were similar for SC
and SS patients undergoing abdominal surgeries and ortho-
pedic procedures. The level of Hb A in the transfused pa-
ative transfusion,'~4.'2~'3*'s-19 and only a few investigators have
3682
KOSHY ET AL
tients did not decrease postoperative complication rates, ex-
cept for the decrease in the rate of postoperative pain crisis
for abdominal surgeries. There were only 21 procedures
among S@+thal patients enrolled in the study, and there was
no report of cholecystectomy. Although the total number of
patients was small, this low number of surgeries may be
reflective of the more benign phenotype.
There was wide variation in the SCD-related complica-
tions associated with the more common surgeries. The rate
for SS patients was 0% for tonsillectomy and adenoidec-
tomy, 2.9% for hip surgery, 3.9% for myringotomy, 7.8%
for intra-abdominal surgery, 16.9% for cesarean section and
hysterectomy, and 18.6% for dilation and curettage. Reasons
for this wide variation were not identified, although the high
rate of sickle cell complications after Cesarean section prob-
ably is related to the increased morbidity associated with
pregnancy in women with sickle cell anemia irrespective of
transfusion practice^.^"'^At present, because laparoscopic
cholecystectomy has replaced the open procedure, the dura-
tion of hospitalization, transfusion requirements, and postop-
erative complications will most likely be lower than that
reported here.
We also observed that in the patients transfused before
intra-abdominal surgery, the mean level of Hb A was higher
for those patients with no painful crisis compared with those
who experienced painful crisis after surgery (58% v 35%,
respectively). However, prevention of painful crisis in the
postoperative period does not alone justify the use of preop-
erative transfusions in light of the recognized complications
of blood transfusion. The recently concluded randomized
controlled trial evaluating perioperative blood transfusion
has defined the role of preoperative blood transfusion for
sickle cell patient^.^'
This study also showed that non-SCD complication rates
for fever and infection were higher in patients receiving
regional anesthesia compared with those who had received
general anesthesia but unrelated to the preoperative transfu-
sion rates. Because regional anesthesia (specifically epidural)
is commonly used in Cesarean sections and other assisted
deliveries, the higher complication rates observed may be a
reflection specifically of the higher complication rates ob-
served in those obstetrical procedures.
In non-SCD patients undergoing surgery, perioperative
complications vary from general pulmonary complications
of 3% to 70% to serious ACS-like events of less than
40 Similarly, in-hospital complications for open cholecystec-
tomy occur in 22.4% (unadjusted rate) of patients4' Our
data show that SCD patients are at no greater risk for these
complications than the non-SCD patients.
Although not derived from a rigorously controlled trial,
the data do define the types of commonly performed surger-
ies on patients with SCD and provide new insights into
postoperative complications. Despite the variety of tech-
niques used to manage patients and the variations in methods
of inducing anesthesia, mortality was low and there were
relatively few serious perioperative complications related to
SCD. In part, this outcome can be attributed to careful atten-
tion to the details of patient management by the collaborative
efforts of the hematologist, surgeon, and anesthesiologist.
However, the role of transfusion in the perioperative period
remains to be defined, but the data do suggest that not all
patients undergoing surgery should routinely receive blood
transfusions.
During the study period, no protocol was specified for
preoperative transfusion practice. Each center continued to
follow its own preoperative preparative regimen. The pa-
tient's age, disease state, multiorgan disease status, and
American Surgical Association (ASA) risk level were not
predetermined.
For the minor procedures, many of the patients did not
receive blood transfusions and complication rates were low.
The postoperative complication rates were also low for the
42 patients who underwent tonsillectomy. Six SS patients
who were untransfused had no complications after tonsillec-
tomy. The paucity of complications after this procedure may
be related to the younger age of patients at the time of
surgery.
Recent advances in intraoperative techniques and new an-
esthetic agents have been touted for successful outcomes for
SCD patients in postoperative periods, rather than the use
of blood transfusion regimen alone.3~4~'2,20~22.25.29.35,42 Others
credit aggressive preoperative transfusion preparation as re-
sponsible for the successful outcome of SCD patients under-
going surgical procedures.'5342 The percent reduction of S
Hb required for transfusing patients is not well defined in
those receiving mandatory preoperative transfusion.
Surgical procedures can be performed successfully on
SCD patients. Careful assessment of Hb phenotype, past
medical history, and risk status should be documented.
Blood transfusion therapy will continue to be part of the
preoperative evaluation and preparation of patients. Simple
transfusion to increase the Hb level to 10 g/&, blood re-
placement for profound anemia of Hb less than 5 g/dL, and
intraoperative hemorrhage appears appropriate. Transfusion
still carries with it the complications of alloimmunization,
delayed transfusion reaction, transmission of viral infection,
hepatitis, and iron o~erload?~," These complications should
be considered when counseling patients and family before
the surgical procedure.
ACKNOWLEDGMENT
The authors are indebted to Pamela Moore, Shaheen Islam, and
Susan Weaver for their assistance in preparing this manuscript. The
following were senior investigators in the CSSCD: Clinical centers:
R. Johnson, Aka Bates Hospital (Berkeley, CA); L. McMahon, Bos-
ton City Hospital (Boston, MA); 0. Platt, Children's Hospital (Bos-
ton, MA); F. Gill and K. Ohene-Frempong, Children's Hospital
(Philadelphia, PA); G. Bray, J. Kelleher, and S. Leikin, Children's
National Medical Center (Washington, DC); E. Vichinsky and B.
Lubin, Children's Hospital (Oakland, CA); A. Bank and S. Piomelli,
Columbia Presbyterian Hospital (New York, NY); W. Rosse, J. Fd-
letta, and T. Kinney, Duke University (Durham, NC); L. Lessin,
George Washington University (Washington, DC); J. Smith and Y.
Khakoo, Harlem Hospital (New York, NY); R. Scott, 0. Castro, and
C. Reindorf, Howard University (Washington, DC); H. Dosik, S.
Diamond, and R. Bellevue, Interfaith Medical Center (Brooklyn,
NY); W. Wang and J. Wilimas, LeBonheur Children's Hospital
(Memphis, TN); P. Milner, Medical College of Georgia (Augusta,
GA); A. Brown, S. Miller, R. Rieder, and P. Gillette, State University
of New York Health Science Center at Brooklyn (Brooklyn, NY);
RISKS OF SURGERY AND ANESTHESIA IN SICKLE CELL DISEASE
3683
W. Lande, S . Embury, and W. Mentzer, San Francisco General
Hospital (San Francisco, CA); D. Wethers and R. Grover, St Lukes-
Roosevelt Medical Center (New York, NY); M. Koshy and N. Tali-
shy, University of Illinois (Chicago, L); C. Pegelow, P. Klug, and
J . Temple, University of Miami (Miami, FL); M. Steinberg, Univer-
sity of Mississippi (J ackson, MS); A. Kraus, University of Tennessee
(Memphis, TN); H. Zarkowsky, Washington University (St Louis,
MO); C. Dampier, Wyler Childrens Hospital (Chicago); H. Pearson
and A.K. Ritchey, Yale University (New Haven, CT); Statistical
Coordinating Centers: P. Levy, D. Gallagher, A. Koranda, Z. Flour-
noy-Gill, and E. J ones, University of Illinois School of Public Health
(Chicago, IL; 1979-89); S . McKinlay, O.Platt, D. Gallagher, D.
Brambilla, and L. Sleeper, New England Research Institutes (Water-
town, MA; 1989-1995); M. Espeland, Bowman-Gray School of
Medicine (Winston-Salem, NC); Program Administration: M. Gas-
ton, C. Reid, and J . Verter, National Heart, Lung, and Blood Institute
(Bethesda, MD).
APPENDIX
Descriptiodlocation of 1,079 surgical procedures. Two
hundred twenty-two cholecystectomy, 2 cholecystotomy, 3
obstruction of gall bladder, 36 splenectomy, 107 dilation
and curettage for pregnancy termination and miscellaneous
complications, 7 dilation and curettage postpartum and for
ectopic pregnancy, 75 Cesarean section, 12 hysterectomy,
46 tonsils and adenoids, 21 hip replacement, 19 hip revision,
3 hip prosthesis removal, 30 myringotomy, 5 craniotomy, 4
spinal canal, 4 miscellaneous nervous system, 6 retina, 4
scleral buckling, 12 miscellaneous eye, 1 myringoplasty, 1
tympanoplasty, 2 radical mastoidectomy, 3 sinus, 11 dental
exiraction and restoration, 2 cleft palate correction, 5 bron-
chial and pulmonary biopsy, 6 miscellaneous respiratory, 20
vascular access, 7 miscellaneous cardiovascular, 5 lymphatic
node biopsy, 10 appendectomy, 3 hemorrhoidectomy, 16
inguinal hernia repair, 8 umbilical hernia repair, 19 laparot-
omy and laparoscopy, 5 miscellaneous digestive system, 3
kidney transplant, 10 miscellaneous urinary system, 4 unde-
scended testes, 6 penile prosthesis, 17 priapism surgery, 22
circumcision, 4 miscellaneous male genital organ, 4 ovaries
and Fallopian tubes, 20 tubal ligation, 3 breast reduction/
enhancement, 8 miscellaneous breast surgery, 6 miscellane-
ous gynecological, 8 incision and drainage of long bones, 5
osteotomy, 18 bone excision and biopsy, 5 bone graft, 8
open reduction and treatment of fractures, 4 joint fusion, 2
shoulder and wrist replacement, 4 clubfoot release, 6 tendon
and sheath repair, 12 miscellaneous musculoskeletal, 31 skin
debridement, 21 skin and subcutaneous, 38 skin graft, 2
parathyroidectomy, 49 diagnostic procedures.
REFERENCES
1. J anik J , Seeler RA: Perioperative management of children with
sickle hemoglobinopathy. J Pediatr Surg 15:117, 1980
2. Spigelman A, Warden MJ: Surgery in patients with sickle cell
disease. Arch Surg 104:761, 1972
3. Holzmann L, Finn H, Lichtman HC, Harmel MH: Anesthesia
in patients with sickle cell disease: A review of 112 cases. Anesth
Analg 48:566, 1969
4. Charache S: The treatment of sickle cell anemia, in Creger
WP, Coggins CH, Hancock EW (eds): Annual Review of Medicine.
Palo Alto, CA, Annual Reviews, 1981, p 195
5. Seeler R: Intensive transfusion therapy for priapism in boys
with sickle cell anemia. J Urol 110360, 1973
6. Rifkind S , Waisman J , Thompson R, Goldfinger D: RBC ex-
change pheresis for priapism in sickle cell disease. J AMA 2422317,
1979
7. Schmalzer E, Chien S , Brown A: Transfusion therapy in sickle
cell disease. AmJ Pediatr Hematol Oncol 4:395, 1982
8. Russell MO, Goldberg HI, Hodson A, KimHC, Halus J , Rei-
vich M, Schwartz E: Effect of transfusion therapy on arteriographic
abnormalities and on recurrence of stroke in sickle cell disease.
Blood 63:162, 1984
9. Lanzkowsky P, Shende A, Karayalcin G, KimY, Aballi A:
Partial exchange transfusion in sickle cell anemia. AmJ Dis Child
132:1206, 1978
10. Brody JI, Goldsmith MH, Park SK, Soltys HD Symptomatic
crises of sickle cell anemia treated by limited exchange transfusion.
Ann Intern Med 72:327, 1970
1 I . Green M, Hall RJ C, Huntsman RG, Lawson A, Pearson TCF,
Wheeler PCG: Sickle cell crisis treated by exchange transfusion.
J AMA 231:948, 1975
12. US Department of Health and Human Services: Management
and Therapy of Sickle Cell Disease. NIH Publication No. 84-2117.
Washington, DC, National Institutes of Health, September 1984
13. Greenwalt TJ , Zelenski KR: Transfusion support for hemo-
globinopathies: Blood transfusion and blood banking, Clin Hematol
13:151, 1984
14. Morrison J C, Blake PG, McCoy C, Martin JN J r, Wiser WL:
Fetal health assessment in pregnancies complicated by sickle hemo-
globinopathies. Obstet Gynecol 61:22, 1983
15. Morrison J C, Schneider J M, Whybrew WD, Bucovaz ET,
Menzel DM: Prophylactic transfusions in pregnant patients with
sickle hemoglobinopathies: Benefit versus risk. Obstet Gynecol
56:274, 1980
16. Nagey DA, Garcia J , Welt S: Isovolumetric partial exchange
transfusion in the management of sickle cell disease in pregnancy.
AmJ Obstet Gynecol 141:403, 1981
17. Coker NJ , Milner PF: Elective surgery in patients with sickle
cell anemia. Arch Otolaryngol 108:574, 1982
18. Fullerton MW, Philippart AI, Samaik S , Lusher J M: Preoper-
ative exchange transfusion in sickle cell anemia. J Pediatr Surg
16:197, 1981
19. Burrington J D, Smith MD: Elective and emergency surgery
in children with sickle cell disease. Surg Clin North Am5655, 1976
20. Lagarde MC, Tunell WP: Surgery in patients with hemoglo-
bin-S disease. J Pediatr Surg 13:605, 1978
21. Homi J , Reynolds J , Skinner A, Hanna W, Serjeant G: General
anesthesia in sickle-cell disease. Br Med J 16:1599, 1979
22. Serjeant G: Sickle Cell Disease. London, UK, Oxford, 1985,
p 371
23. Davis J R, Vichinsky EP, Lubin BH: Current treatment of
sickle cell disease. Curr Probl Pediatr lO:l, 1980
24. Lessin LS, Kurantsin-Mills J, Klug PP, Weems HB: Determi-
nation of continuous flow blood cell separator. J Clin Apheresis
1:64, 1978
25. Griffin TC, Buchanan G: Elective surgery in children with
sickle cell disease without pre-operative blood transfusion. J Ped
Surg 28:681, 1993
26. Gaston M, Smith J , Gallagher D, Flournoy-Gill Z, West S ,
Bellevue R, Farber M, Grover R, Koshy M, Ritchey AK, Wilimas
J , Verter J , and the CSSCD Study Group: Recruitment in theCooper-
ative Study of Sickle Cell Disease (CSSCD). Controlled Clin Trials
8:131S, 1987
27. Farber MD, Koshy M, Kinney TR, and the Cooperative Study
of Sickle Cell Disease: Cooperative Study of Sickle Cell Disease:
Demographic and socioeconomic characteristics of patients and fam-
ilies with sickle cell disease. J Chron Dis 38:495, 1985
28. Gaston M, Rosse W: The Cooperative Study of Sickle Cell
3684 KOSHY ET AL
Disease: A review of study design and objectives. AmJ Pediatr
Hematol Oncol 4:196, 1982
29. Cohen MM, Duncan PG, Tate RB: Does anesthesia contribute
to operative mortality? J AMA 260:2859, 1988
30. Zeger SL, Liang KY: Longitudinal data analysis for discrete
and continuous outcomes. Biometrics 43:121, 1986
3 1. Koshy M, Ashenhurst J : Management of pregnancy in sickle
cell anemia. Tex Rep Biol Med 40:273, 1981
32. Koshy M, Burd L, Wallace D, Moawad A, Baron J: Prophy-
lactic red cell transfusion in pregnant patients with sickle cell disease:
A randomized cooperative study. N Engl J Med 319:1447, 1988
33. Powars DR, Sandhu M, Niland-Weiss J , J ohnson C, Bruce
S, Manning PR: Pregnancy in sickle cell disease. Obstet Gynecol
67:217, 1986
34. Charache S, Scott J , Niebyl J , Bonds J: Management of sickle
cell disease in pregnant patients. Obstet Gynecol 55:407, 1980
35. Vichinsky EP, Haberkem CM, Neumayr L, Earles A, Black
D, Koshy M, Pegelow C, Abboud M, Ohene-Frempong K, Iyer RV,
and the Preoperative Transfusion Study Group: A comparison of
conservative and aggressive transfusion regimens in the periopera-
tive management of sickle cell disease. N Engl J Med 333:206, 1995
36. Strandberg A, Tokics L, Brismar B, Lundquist H, Heden-
stiema G: Atelectasis during anaesthesia and in the postoperative
period. Acta Anaesthesiol Scand 30:154, 1986
37. J ayr C, Mollie A, Bourgain J L, Alarcon J , Masselot J , Lasser
P, Denjean A, Truffa-Bachi J , Henry-Amar M: Postoperative pulmo-
nary complications: General anesthesia with postoperative parenteral
morphine compared with epidural analgesia. Surgery 104:57, 1988
38. Engberg G, Wiklund L: Pulmonary complications after upper
abdominal surgery: Their prevention with intercostal blocks. Acta
Anaesthesiol Scand 32:1, 1988
39. Roukema J A, Carol EJ , Prins J G: The prevention of pulmo-
nary complications after upper abdominal surgery in patients with
noncompromised pulmonary status. Arch Surg 123:30, 1988
40. Vodinh J, Bonnet F, Touboul C, Lefloch J P, Becquemin J P,
Had A: Risk factors of postoperative pulmonary complications after
vascular surgery. Surgery 105:360, 1989
41. Kane RL, Lune N, Borbas C, Moms N, Flood S, McLaughlin
B, Nemanich G, Schultz A: The outcomes of elective laparoscopic
and open cholecystectomies. J AmCol1 Surg 180:136, 1995
42. Maduska AL, Guinee WS, Heaton J A, North WC, Barreras
LM: Sickling dynamics of red blood cells and other physiologic
studies during anesthesia. Anesth Analg 54:361, 1975
43. Orlina A, Unger P, Koshy M: Post-transfusion alloimmuniza-
tion in patients with sickle cell anemia. AmJ Hematol 5:101, 1978
44. Castro 0: Autotransfusion: A management option for alloim-
munized sickle cell patients?, in Scott R (ed): Advances in the Patho-
physiology, Diagnosis, and Treatment of Sickle Cell Disease. New
York, NY, Liss, 1982, p 117