1. Australian and New Zealand Journal of Obstetrics and Gynaecology 2008; 48: 592–595 DOI: 10.1111/j.1479-828X.2008.00908.x
Original Article
Blackwell Publishing Ltd
Quality-initiated prophylactic antibiotic use in laparoscopic-assisted
vaginal hysterectomy
Wei-Chun CHANG,1,2 Meng-Chih LEE,2 Lian-Shung YEH,1 Yao-Ching HUNG,1
Cheng-Chieh LIN3,4 and Long-Yau LIN2
1
Department of Obstetrics and Gynecology, China Medical University and Hospital, 2Institute of Medicine, Chung-Shan Medical University,
3
Department of Family Medicine, China Medical University and Hospital, Taichung and 4Institute of Health Care Administration, College of
Health Science, Asia University, Taichung, Taiwan
Background: An evidence-based initiation of quality improvement activity for reducing the usage of prophylactic antibiotic in
laparoscopic-assisted vaginal hysterectomy (LAVH) in a tertiary hospital.
Aims: The authors investigated whether single or multiple doses of cefazoline were more cost-effective in preventing
postoperative infection associated with LAVH.
Methods: The study groups comprised of 310 patients who had undergone LAVH continuously in a medical centre. Patients
were divided into two groups on the basis of whether they received a single dose or multiple doses (range: two to four doses)
of cefazoline during the perioperative period. Postoperative infections such as pelvic cellulitis or abscess, vaginal cuff abscess,
wound infection and urinary tract infection that occurred either during hospitalisation or within one month after discharge
were observed and recorded. Incremental cost-effectiveness ratio (ICER) was calculated using the mean direct drug cost and
the prophylactic effect of infection in both groups.
Results: The prophylactic effect of infection was similar in the single-dose group and the multiple-dose group (94.6% vs
93.9%, P = 0.986). The ICER was significantly lower in the single-dose group (153.3 vs 460.4, P < 0.001).
Conclusions: The result revealed that a single dose of cefazoline is more cost-effective than multiple doses in the prevention
of infection associated with LAVH. It fulfils the goal of cost minimisation and quality of care in today's environment of medical
cost containment.
Key words: antibiotics, hysterectomy, laparoscopic surgical procedure, vaginal.
Current evidence-based clinical practice has put the major main interest is in identifying and choosing the least cost
emphasis on establishing the cost-effectiveness of interventions. option, and this is called a cost-minimisation study.
The shifting in the health-care system towards a more A clinical pathway for laparoscopic-assisted vaginal
managed environment has forced health-care providers in a hysterectomy (LAVH) started at our hospital in January
position to streamline resources and provide quality care in 1998, when medical expenditures were paid under the quota
the most cost-effective way. Cost-effectiveness analysis case–payment system. After a pilot study that proved a short
(CEA) is analytical techniques in health care that may assist course of combined prophylactic antibiotics (cephalothin +
with more rational, effective and economically sound medical gentamycin) was as efficacious as a longer course in preventing
decision-making.1–3 CEA, which assesses both the costs and postoperative infection, our department initiated a quality
the health outcomes of alternative health-care programs or improvement activity for encouragement of further reducing
strategies, can provide useful information about the relative prophylactic antibiotics usage in hospitalised LAVH patients
benefits and trade-offs of different health-care interventions. since May 2000.4 Previous studies have shown that the
The unit value of effectiveness in an incremental cost- postoperative infection rate following LAVH is 2–5.6%.5,6 In
effectiveness ratio (ICER) may be any unit such as quality- this study, a retrospective cohort study was conducted to
adjusted life years saved, a ratio of the difference in outcomes. prove the cost-minimisation result of a single dose versus
If there is information on the outcome or effectiveness of two multiple doses of single-agent prophylactic antibiotics to
alternatives, and they are known to be equivalent, then the prevent postoperative infection in LAVH.
Correspondence: Dr Long-Yau Lin, Institute of Medicine, Methods
Chung-Shan Medical University, No. 110 Sec. 1, Chien-Kuo A cohort of 310 eligible patients who had undergone LAVH
N. Road, Taichung 402, Taiwan. Email: wei66@iris.seed.net.tw
continuously in a medical centre in central Taiwan were
Received 21 January 2008; accepted 16 June 2008. studied. Patients were divided into two groups on the basis
592 © 2008 The Authors
Journal compilation © 2008 The Royal Australian and New Zealand College of Obstetricians and Gynaecologists

2. Quality-initiated prophylactic antibiotic use in LAVH
of whether they had received a single dose or multiple doses
(range: two to four doses) of prophylactic antibiotic during Results
the perioperative period. The prophylactic dosage was The single-dose group consisted of 147 patients (mean age,
ordered by attending physicians. The prophylactic antibiotic 42.4 ± 6.8 years). The multiple-dose group included 163
administered was cefazoline. A single dose was defined as a patients (mean age, of 41.2 ± 5.8 years). There were no
single 1-g intravenous push stat and a multiple doses differences in the American Society of Anaesthesiologists
consisted of 1-g injection of cefazoline stat and every six physical status scale, age, parity, duration of operation or
hours for one, two or three times. The first dose of antibiotic EBL during surgery between the two groups (Table 1).
had been injected within 30 min prior to the incision. The The single-dose group significantly influenced the number
single-dose group was composed of 147 patients and the of injected doses of antibiotics, the antibiotic fee and the total
multiple-dose group comprised 163 patients. The study admission fee. The average total fee decreased significantly,
period was from July 2002 to June 2005. Preoperative serum by 1%, from NT$59 325 to NT$58 726 (P = 0.05). The average
haemoglobin level > 9 g/dL was requisite for all patients in antibiotics fee dropped by NT$287, while the average vials
both groups. Patients who had developed complications and of cefazoline decreased by 1.9 (65.5%) (Table 2). Since
those who required therapeutic antibiotics or other associated LAVH cases are paid for under a case–payment system in
surgical procedures were excluded from the study. The top Taiwan, the total admission fee paid is fixed at NT$63 230.
four indications for LAVH in the single-dose group were Hospital managers may gain additional benefits because of
myoma uteri (70 patients), adenomyosis (49), cervical the related cost-saving in a case–payment system. There was
carcinoma in situ (12) and cervical carcinoma Ia1 (six). The no significant difference in the average hospital stay between
top four indications for LAVH in the multiple-dose group the two groups (Table 2).
were myoma uteri (76 patients), adenomyosis (53), cervical There were no significant difference in the rate of
carcinoma in situ (10) and benign adnexal cyst (eight). operative site infection and urinary tract infection during
The length of stay, antibiotic doses and fees, and total hospitalisation and within one month of discharge between
admission fee were collected from the hospital's electronic the two groups (Table 2). In the single-dose group, one
database. Patient characteristics (age, parity, diagnosis and trocar site wound infection occurred during hospitalisation,
the American Society of Anaesthesiologists physical status one infection occurred within one month of discharge, and
scale) and medical care process data (operation time, two cases of vaginal cuff abscess and four cases urinary tract
surgical estimated blood loss (EBL), operative site and infections occurred within one month of discharge. In the
urinary tract infection during hospitalisation and within one multiple-dose group, two trocar site wound infection
month of discharge) were collected from the patient charts. occurred during hospitalisation, two infections occurred
In order to evaluate the efficacy of the prophylactic within one month of discharge, and two cases of cuff abscess
antibiotic dose in the prevention of postoperative infection in and four cases urinary tract infections occurred within one
LAVH, we used the classification of operative site infection month of discharge.
by Shapiro et al. which includes pelvic cellulitis, vaginal cuff A single dose of prophylactic antibiotic was similarly
abscess, pelvic abscess and wound infection.7 Urinary tract effective at reducing infection than multiple doses (94.6% vs
infection was diagnosed based on the patients’ symptoms 93.9%, P = 0.986). Also, the ICER, which was calculated as
and signs, results of urinalysis and clinical improvement the mean direct antibiotic cost per patient divided by the
after treatment. prophylactic effect of infection was significantly lower in the
Data are presented as mean ± standard deviation. The single-dose group (153.3 vs 460.4, P < 0.001) (Table 3). It
statistical significance of the differences between continuous means that the cost is cheaper when per unit amount of
variables in the two groups of patients was determined by prophylactic effect of infection is achieved in the single-dose
Student’s t-test. The χ2 test was used to measure the statistical group. The power was 0.44 for finding an assumed 5%
significance of difference between nominal variables in the difference in the rate of successful prevention from operative
two groups of patients. site and urinary tract infections between the two groups to
Table 1 Patient characteristics and operative parameters
Characteristics Single-dose group (n = 147) Multiple-dose group (n = 163) P* value
Age (year, mean ± SD) 42.4 ± 6.8 41.2 ± 5.8 0.46
Parity (mean ± SD) 2.3 ± 1.1 2.5 ± 1.0 0.21
Operation time (min, mean ± SD) 142.4 ± 36 150.4 ± 44 0.57
EBL (mL, mean ± SD) 100.4 ± 56.6 108.6 ± 60.9 0.63
†American Society of Anaesthesiologists 1 (%) 103/212 (48.6%) 109/212 (51.4%)
American Society of Anaesthesiologists 2 (%) 44/98 (44.9%) 54/98 (55.1%)
*Student’s t-test for continuous variables, χ2 test for nominal variables.
†American Society of Anaesthesiologists physical status scale and case number (percentage) are shown.
EBL, estimated blood loss; SD, standard deviation.
© 2008 The Authors 593
Journal compilation © 2008 The Royal Australian and New Zealand College of Obstetricians and Gynaecologists; 48: 592–595

3. W.-C. Chang et al.
Table 2 Comparison of medical fees (NT$) and outcome of infection rate
Single-dose group Multiple-dose group
(n = 147) (n = 163) P* value
Antibiotic fee (mean ± SD) 145 ± 0 432 ± 83 < 0.001
Total admission fee (mean ± SD) 58 726 ± 3221 59 325 ± 3747 0.05
Cefazoline doses (vials, mean ± SD) 1±0 2.8 ± 0.5 < 0.001
Hospital stay (days, mean ± SD) 4.3 ± 1.2 4.4 ± 0.9 0.74
Rate of operative site infection 147 (2.7%) 6/163 (3.6%) 0.63
during hospitalisation and within
one month of discharge
Rate of urinary tract infection 4/147 (2.7%) 4/163 (2.4%) 0.88
during hospitalisation and within
one month of discharge
Exchange rate for 2007 NT$ 32 = US$1.
*Student’s t-test for continuous variables, χ2 test for nominal variables.
SD, standard deviation.
Table 3 Results of cost-effectiveness analysis
Single-dose group (n = 147) Multiple-dose group (n = 163) P* value
The prophylactic effect of infection (%) 139/147 (94.6%) 153/163 (93.9%) 0.986
Direct antibiotic fee per patient (mean ± SD) 145 ± 0 432.3 ± 82.8 < 0.001
Incremental cost-effectiveness ratio (mean ± SD) 153.3 ± 0 460.4 ± 88.3 < 0.001
*Student’s t-test for continuous variables, χ2 test for nominal variables.
SD, standard deviation.
be statistically significant at the 5% level of significance in contamination by pathogenic microorganisms indigenous to
this study because of sample size. To achieve a power level the genital tract during surgery, including gram-positive
of 0.8, about 1500 cases in each group would be necessary. and -negative aerobes and anaerobes, is the same as for
Also a prospective randomised controlled trial is the highest abdominal hysterectomy. A meta-analysis of 31 English-
priority of study design to have the least bias. language randomised-controlled trials published from 1972
to 1986 concluded that antibiotic prophylaxis reduced the
rate of serious infections after abdominal hysterectomy from
Discussion 21.1% to 9%.15 Another meta-analysis of 17 ‘controlled or
Prophylactic antibiotics refer to a brief course of an comparative’ trials was conducted between 1978 and 1990,
antimicrobial agent administered just prior to an operation in investigating single or one-day prophylactic regimens of
order to reduce intraoperative microbial contamination to a intravenous or intramuscular cephalosporins for abdominal
level that will not overwhelm host defences and result in hysterectomy.16 Again the results clearly favoured the use of
infection.8 There is substantial evidence in the literature that prophylaxis. Prophylactic antibiotic used routinely in total
prophylactic antibiotic use significantly decreases rates of abdominal hysterectomy is highly suggested.
postoperative febrile morbidity and infection.9–12 Such The most common prophylactic antibiotic is one that is
infections not only cause patient morbidity but also result in active against a wide range of bacteria (broad-spectrum),
additional costs, prolonged hospital stay and increased such as amoxicillin – clavulanic acid (Augmentin) or a
antibiotic use, which induces the emergence of antimicrobial cephalosporin. It is generally recommended that first- or
resistant organisms.13 second-generation cephalosporins be used for prophylaxis,
Even with the best surgical and perioperative care, as they appear to be equally effective for the purpose, less
hysterectomy is associated with a high risk of infection expensive and less likely to elucidate drug resistance.17,18 In
because the surgery breaches the genital tract, a region is this study, a type of antibiotic (cefazoline) was used to prevent
commonly colonised by a wide variety and large number of postoperative surgical site and urinary tract infection. The
microorganisms. Furthermore, most women who have timing of cefazoline administration was within 30 min prior
undergone hysterectomy require an indwelling urinary to surgical incision, when the antibiotic should be present in
catheter for the first 24 h, which increases the risk of urinary the tissue prior to opening the vaginal cuff, at which time
tract infection. Common sites of infection after hysterectomy vaginal organisms enter the pelvic cavity.
are the abdominal wound and the vaginal vault, the pelvic According to this study, there were no significant differences
floor, and the bladder.14 LAVH is performed through the in operative site infection or urinary tract infection during
abdomen and vagina simultaneously. The goal of preventing hospitalisation or within one month of discharge between the
594 © 2008 The Authors
Journal compilation © 2008 The Royal Australian and New Zealand College of Obstetricians and Gynaecologists; 48: 592–595

4. Quality-initiated prophylactic antibiotic use in LAVH
two groups (Table 2). The mean antibiotic cost in the multiple- 6 Harris WJ, Daniell JF. Early complications of laparoscopic
dose group was NT$432.3 for each patient. A single dose of hysterectomy. Obstet Gynecol Surv 1996; 51: 559–567.
cefazoline (1.0 g) cost only NT$145 per patient. The 7 Shapiro M, Munoz A, Tager TB, Schoenbaum SC, Polk BF.
incremental cost-effectiveness ratio was significantly lower in Risk factors for infection at the operative site after abdominal
the single-dose group (153.3 vs 460.4, P < 0.001). This result or vaginal hysterectomy. N Engl J Med 1982; 307: 1661–1666.
fulfils the goal of a cost-minimisation study that outcome or 8 Mangram AJ, Horan TC, Pearson ML, Silver LC, Jarvis WR.
effectiveness of two alternatives are found to be equivalent, Guideline for prevention of surgical site infection, 1999:
then the main interest is in identifying and choosing the least Centers for Disease Control and Prevention Hospital Infection
Control Practices Advisory Committee. Am J Infect Control
cost option.19 We suggest single-dose cefazoline as prophylaxis
1999; 27: 97–132.
of infection in LAVH.
9 Regidor PA, Bier UW, Preuss MJ et al. Efficacy and safety of
Interestingly, of the 18 infected cases in this study, only
two cephalosporins in the prerioperative prophylaxis in patients
three were found and diagnosed during hospitalisation. Most
undergoing abdominal or vaginal hysterectomies or gynaecological
infections (83%), especially urinary tract infections, were laparotomies: A prospective randomized study. Gynakologisch-
diagnosed within one month of discharge from the hospital. geburtshilfliche Rundschau 2000; 40: 153–158.
A study also pointed out that most patients with wound 10 Meltomaa SS, Makinen JI, Taalikka MO, Helenius HY. Incidence,
infection were diagnosed after discharge from the hospital.20 risk factors and outcome of infection in a 1-year hysterectomy
In their study population, transportation problems and remote cohort: A prospective follow-up study. J Hosp Infect 2000; 45:
residence were prevalent. There may be other risk factors 211–217.
such as inexperienced aseptic wound care, inadequate 11 Hayashi H, Yaginuma Y, Yamashita T et al. Prospective
nutrition and fluid supply to those whom being infected after randomized study of antibiotic prophylaxis for nonlaparotomy
discharge in our study. We suggest appropriate strategies for surgery in benign conditions. Chemotherapy 2000; 46: 213–
infection prevention and that surveillance be integral to 218.
discharge planning. 12 Varol N, Healey M, Tang P, Sheehan P, Maher P, Hill D. Ten-year
In conclusion, this cohort study revealed that a single dose review of hysterectomy morbidity and mortality: Can we change
of cefazoline is more cost-effective than multiple doses in the direction? Aust N Z J Obstet Gynaecol 2001; 41: 295–302.
prevention of operative site and urinary tract infection 13 Dellinger EP, Gross PA, Barrett TL, Krause PJ, Martone WJ,
associated with LAVH. This result fulfils the goal of cost McGowan JE. Quality standard for antimicrobial prophylaxis
minimisation and quality of care in today’s environment of in surgical procedures. Clin Infect Dis 1994; 18: 422–427.
medical cost containment. 14 Duff P, Park RC. Antibiotic prophylaxis in vaginal hysterectomy:
A review. Obstet Gynecol 1980; 55: 193–202.
15 Mittendorf R, Aronson MP, Berry RE et al. Avoiding serious
infections associated with abdominal hysterectomy: A meta-
References analysis of antibiotic prophylaxis. Am J Obstet Gynecol 1993;
1 Russell LB, Gold MR, Siegel JE, Daniels N, Weinstein MC. 69: 1119–1124.
The role of cost-effectiveness analysis in health and medicine. 16 Tanos V, Rojansky N. Prophylactic antibiotics in abdominal
JAMA 1996; 276: 1172–1177. hysterectomy. J Am Coll Surg 1994; 179: 593–600.
2 Weinstein MC, Siegel JE, Gold MR, Kamlet MS, Russell LB. 17 Fukatsu K, Saito H, Matsuda T et al. Influences of type and
Recommendations of the panel on cost-effectiveness in health duration of antimicrobial prophylaxis on an outbreak of
and medicine. JAMA 1996; 276: 1253–1258. methicillin-resistant Staphylococcus aureus and on the incidence
3 Siegel JE, Weinstein MC, Russell LB, Gold MR. Recommendations of wound infection. Arch Surg 1997; 132: 1320–1325.
for reporting cost-effectiveness analysis. JAMA 1996; 276: 18 Weed HG. Antimicrobial prophylaxis in the surgical patient.
1339–1341. Med Clin North Am 2003; 87: 59–75.
4 Chang WC, Hung YC, Li TC, Yang TC, Chen HY, Lin CC. 19 Gold MR, Siegel JE, Russell LB et al. Cost-Effectiveness in
Short course of prophylactic antibiotics in laparoscopically assisted Health and Medicine. New York: Oxford University Press, 1996.
vaginal hysterectomy. J Repord Med 2005; 50: 524–528. 20 Kamat AA, Brancazio L, Gibson M. Wound infection in
5 Liu CY, Reich H. Complications of total laparoscopic gynecologic surgery. Infect Dis Obstet Gynecol 2000; 8: 230–
hysterectomy in 518 cases. Gynecol Endosc 1994; 3: 203–208. 234.
© 2008 The Authors 595
Journal compilation © 2008 The Royal Australian and New Zealand College of Obstetricians and Gynaecologists; 48: 592–595