The effect of acute postoperative alcohol withdrawal (delirium tremens)
on fibular free flap survival.
Gallivan KH, Reiter D. Acute Alcohol Withdrawal and Free Flap Mandibular Reconstruction Outcomes. Arch Facial Plast Surg. 2001;3(4):264–266. doi:
From the Department of Otolaryngology–Head and Neck Surgery,
Jefferson Medical College and Thomas Jefferson University Hospital/Jefferson
Health System, Philadelphia, Pa.
Objective To evaluate the effect of acute postoperative alcohol withdrawal on
survival of vascularized fibular grafts for mandibular reconstruction.
Design Retrospective case series of 17 consecutive patients.
Main Outcomes Measure Relation between flap survival and alcohol withdrawal.
Results Flap survival rate was 25% for patients who experienced delirium tremens
and 85% in the other patients. Had all flaps in patients with postoperative
alcohol withdrawal survived, the success rate would have been 89%. Flap loss
was related to acute alcohol withdrawal (P = .02, χ2 analysis). The relationship between complication rate and alcohol
withdrawal was also significant, using the Fisher exact test.
Conclusions Fibular free flap reconstruction of the mandible is clearly cost-effective
when it facilitates return to social function and productivity. In our experience,
acute alcohol withdrawal in the first 72 hours after surgery is associated
with a high incidence of flap loss. Therefore, we believe that patients at
significant risk for alcohol withdrawal should undergo detoxification preoperatively.
Society's economic return for investing in free flap reconstruction comes
from minimizing convalescence and maximizing postoperative patient productivity.
This return will not be realized for poorly selected patients. We are looking
further into the effects of alcoholism on flap survival rates.
LITERATURE abounds on the definition of outcome.
General acceptance of the "4 D's" of medical management
(death, disease, disability, discomfort) has weakened in the face of ardent
consumerism. Lohr1 cites function, satisfaction,
and survival as more positive outcomes measures, whereas Lang and Marek2 describe 9 types of outcomes. These outcomes include
psychosocial, physiological, behavioral, functional, quality-of-life, knowledge,
resource utilization, financial, and satisfaction considerations. Tarlov et
al3 offered 4 categories, including clinical
end points (eg, laboratory values), functional status, general well-being,
and satisfaction (in which category they place access, convenience, and physician
coverage). As society's focus on the interface between socioeconomic factors
and the cost of health care sharpens, we must address issues that relate patient
behavior to clinical outcomes.
To date, no study has attempted to determine the effects of alcohol
withdrawal syndrome on the success of mandibular reconstruction with vascularized
We evaluated the medical records of 17 consecutive patients undergoing
mandibular resection with fibular free flap reconstruction by one of us (D.R.).
Defects ranged from 8 to 14 cm (mean, 11.3 cm). Nine defects were lateral
and 8 included the anterior arch. Five patients had received prior radiation
therapy. Nine patients had a history of regular employment, and 4 were nonsmokers
and nondrinkers. Age ranged from 35 to 72 years. Twelve patients were male
and 5 were female. Thirteen had squamous cell carcinoma (all stage IV), 2
had ameloblastoma, 1 had mucoepidermoid carcinoma, and 1 had osteoradionecrosis.
Medical records were reviewed for documentation of length of postoperative
hospital stay, complications, route of nutrition at discharge, integrity of
speech, need for nursing care on discharge, and return to work (if applicable).
Four of our patients developed acute alcohol withdrawal syndrome in
the initial postoperative period. These patients required 12, 19, 21, and
22 days of postoperative hospitalization. Three of the 4 in this group had
been unemployed before surgery. The fourth had been regularly and gainfully
employed, but was found retrospectively to have been consuming several alcoholic
beverages daily for years without attracting the attention of family or colleagues.
Only one flap survived acute delirium tremens.
Eight of 9 patients who were gainfully employed returned to work, including
the 1 employed patient who developed delirium tremens. These patients required
an average of 14.4 postoperative weeks to return to work. The indication for
surgery was squamous cell carcinoma in 5, ameloblastoma in 2, and mucoepidermoid
carcinoma in 1.
All 6 patients who had been unemployed for quite some time had postoperative
complications. Three of the 4 patients with alcohol withdrawal were in this
group, and the flap survival rate was 33% (2/6). Two unemployed patients had
retired after lives of active employment. Both were discharged from the hospital
with the ability to eat using their mouth, and neither had postdischarge care
Flap survival in patients experiencing delirium tremens was 25%, whereas
flap survival in the remaining group was 85% (Figure 1). Had all flaps in patients with alcohol withdrawal survived,
the success rate would have been 89%. A χ2 analysis of data
suggests a significant relationship between acute postoperative alcohol withdrawal
and flap survival (P = .02). Our patients who experienced
alcohol withdrawal syndrome had an average postoperative stay of 18.5 days
compared with an average of 15.5 days for those not experiencing alcohol withdrawal,
a relationship without statistical significance per the t test (P = .05).
Our patient group was otherwise similar to those in most prior published
series on vascularized fibular reconstruction of the mandible. Average length
of stay for patients with lateral defects was 12.3 days, whereas patients
with anterolateral defects averaged 9.7 postoperative days to discharge. Patients
with isolated anterior defects averaged 27 days to discharge. Cigarette smoking
correlated moderately well with postoperative complications (correlation coefficient,
0.662) and length of stay (correlation coefficient, 0.588). Twelve of the
13 patients with squamous cell carcinoma as the indication for surgery smoked
and drank. Delirium tremens developed in 4 of the 13 patients in the group
who admitted to drinking alcoholic beverages. (The numerical discrepancy between
total drinkers and drinkers with squamous cell carcinoma occurs because one
patient required surgery for osteoradionecrosis but was tumor free following
radiation therapy for prior squamous cell carcinoma.) One patient with squamous
cell carcinoma was a nonsmoker and nondrinker by his history. Rates of smoking
and drinking were not significantly different between the group that experienced
alcohol withdrawal and the remaining patients.
The postoperative period to resumption of work averaged 14.4 weeks.
However, the median was 8 weeks. The outliers were 2 and 3.2 SDs from the
mean. One patient had undergone preoperative chemotherapy and radiation, whereas
the other was one of the patients who experienced alcohol withdrawal and flap
Nine employed patients had an average hospital stay of 12.4 days, whereas
8 unemployed patients had an average postoperative hospital stay of 20.3 days,
a significant difference (P = .04) using the t test. Furthermore, 3 of the 4 patients who experienced
alcohol withdrawal had been unemployed for a long duration preoperatively.
Urken et al4 reported that delirium tremens
developed in 4 of 210 patients in their 1998 mandibular reconstruction series,
with "prolonged hospital stays" required for these patients (although the
reported flap survival rate is not stratified for delirium tremens). Alcohol
withdrawal is anecdotally reported to complicate care in patients undergoing
head and neck surgery, but no prior study attempts to relate reconstructive
failure to alcohol withdrawal syndrome or other social phenomena, such as
lifestyle. We observed that alcohol withdrawal syndrome seemed related causally
to reconstructive failure in 75% of our patients so afflicted, although it
did not correlate with prolonged length of postoperative hospital stay.
Our lack of strong correlation between cigarette smoking and postoperative
complications is consistent with the findings of Heinz et al5
in 18 patients with head and neck cancer who were undergoing free flap reconstruction.
Unfortunately, those of our patients who smoked also drank alcoholic beverages.
Therefore, the same statistic applies to that association as well, making
it impossible to stratify for either factor alone in our group. Neither smoking
nor drinking rates differed between the alcohol withdrawal group and the other
patients. A correlation between long-term unemployment and delirium tremens
was just shy of significance (P = .053, Fisher exact
Reconstruction of the mandible with vascularized fibular flap is a reliable
and cost-effective management method for patients not at risk for alcohol
withdrawal syndrome. Acute postoperative alcohol withdrawal was associated
with a high incidence of failure of vascularized fibular free flaps in our
series. Patients should undergo active detoxification from alcohol before
surgery, even if this requires residential observation in a treatment facility.
The cost of complications related to delirium tremens is high for patients
with free flaps, and detoxification is probably less costly than the management
of patients in whom reconstruction fails. We are undertaking a large-scale
study of this problem.
Accepted for publication August 21, 2000.
Presented at the 1998 Annual Meeting of the American Academy of Facial
Plastic and Reconstructive Surgery, San Antonio, Tex, September 10, 1998.
Corresponding author: David Reiter, MD, DMD, PO Box 770, Narberth,
PA 19072 (e-mail: firstname.lastname@example.org).