Postoperative Outcomes Associated with Topical Skin Adhesives among Women Having Hysterectomies

Abstract

Background:

Multiple options are available for closure of hysterectomy incisions. This study compared postoperative clinical and economic outcomes using topical skin adhesive (2-octyl cyanoacrylate; OCA) vs. conventional skin closure in women undergoing total abdominal hysterectomy.

Methods:

A multi-hospital administrative database was used to identify women discharged in 2005 who had undergone total abdominal hysterectomy. Patients, classified by skin closure as suture (n = 21,201), staples (n = 23,441), OCA (n = 880), or staples + OCA (n = 489), were compared on length of inpatient stay (LOS), total inpatient cost, and non-prophylactic antibiotic treatment after day four.

Results:

The unadjusted mean LOS was 3.9, 4.5, 3.7, and 5.2 days for suture, staples, OCA, and staples + OCA, respectively; and the percentages of patients having antibiotic treatment were 12.93, 17.51, 11.14, and 23.72. There were overall differences in adjusted mean LOS, mean total cost, and antibiotic treatment (p < 0.0001). Pairwise comparisons indicated no difference between sutures and OCA, whereas the outcomes for each of the non-staple groups were more favorable than those for the staple group (p < 0.01). Results were similar in adjusted comparisons, with pairwise comparisons between OCA and staples at or near the threshold for significance.

Conclusions:

2-octyl cyanoacrylate appears to be a safe and cost-effective alternative to topical sutures for patients having total abdominal hysterectomy. There were less favorable outcomes in groups receiving staples.

With multiple options available, essential considerations in selecting a device for surgical skin closure are to maintain the integrity of the closure and to minimize the risk of infection; other aspects are patient risk factors, cost, physician time, patient comfort, ambulation, and surgical site appearance [1]. In many types of surgery, when evaluated in relation to sutures and staples, topical skin adhesive or cyanoacrylates have demonstrated comparable effectiveness and safety [1 –14], with potential advantages such as shorter operative time, formation of a protective antimicrobial barrier, simplified incision care, and avoidance of the need for suture removal [15] Although hysterectomy is the second most common surgical procedure performed on women in the U.S. [16], little is known about the extent of cyanoacrylate use as a topical skin closure method or the outcomes associated with its use.

Despite the lack of clinical trials and rigorous evaluation supporting topical skin adhesive use in obstetrical and gynecological procedures, a preliminary analysis indicated their value for skin closure in total abdominal hysterectomy. This study, conducted on a large multi-hospital administrative database, is the first to compare clinical and economic outcomes of total abdominal hysterectomy using topical skin adhesives vs. conventional skin closure techniques.

Patients and Methods

Data source

The source of the data was the Premier's Perspective™ Comparative Database, Premier, Inc., San Diego, CA, which incorporates clinical and financial information from all discharges from more than 400 hospitals in the U.S. The database is compliant with Health Insurance Portability and Accountability Act (HIPAA) guidelines. Patient information includes admission and discharge months, length of stay (LOS), and primary and secondary diagnoses related to the patient's admission using the ninth version of the International Classification of Diseases (ICD-9) coding system [17]. Patient charges, which include room and board, pharmacy, and procedures, are categorized by service day. In addition, the database describes hospital characteristics, such as region, academic status and urban versus rural location.

Sample eligibility

Among all patients discharged from participating hospitals in 2005, the sample comprised all women, aged 18 years and older, who had total abdominal hysterectomies (ICD-9: 68.4). Patients undergoing subtotal or radical hysterectomies (ICD-9: 68.3, 68.6) were excluded to increase the homogeneity of the sample. Laparoscopic hysterectomies were not included because of the minimal incision length. The most common indications for total hysterectomy in the Premier database for 2005, accounting for at least 5% of the distribution, were uterine leiomyoma (34%), disorders of menstruation (15%), other symptoms (10%), malignant neoplasm (10%), pain and other symptoms (9%), and endometriosis (8%). There was no exclusion of patients on the basis of the reason for the procedure.

Key grouping variable

The main independent variable was treatment group: (1) Sutures, (2) staples, (3) cyanoacrylate, or (4) staples plus cyanoacrylate. The skin closure procedure was identified according to the billing for surgical supplies (i.e., staples, suture material, or cyanoacrylate). At the time of the study, although two cyanoacrylates were available (2-octyl cyanoacrylate [Dermabond®, Ethicon Inc., Somerville, NJ] [OCA]) and butylcyanoacrylate [Indermil®, US Surgical, Norwalk, CT]), there were no patients in the database whose skin closure involved butylcyanoacrylate. Consequently, this analysis was limited to OCA.

Outcome variables

Groups were compared on the following continuous outcomes: Length of inpatient stay (LOS), total inpatient costs, and several indicators of antibiotic treatment. A prior chart review study, which evaluated antibiotic exposure after three types of surgery, concluded that use of postoperative antibiotic treatment was a reliable and efficient way to identify surgical site infections (SSIs) [18]. Both LOS and total cost were calculated by Premier and included as distinct variables in the database. The number of antibiotic treatment days was determined by the last service day after admission indicating antibiotic treatment. This last day was a proxy for the number of days of antibiotic treatment. It was defined as the last service day after admission indicating antibiotic treatment.

Treatment groups also were compared on three categorical outcomes: Rates of dehiscence, SSI, and non-prophylactic antibiotic treatment. Specific postoperative complications were assessed by the reported relevant ICD-9 codes—dehiscence (ICD-9: 998.3) and SSI (ICD-9: 998.5). Antibiotic treatment was defined as antibiotic administration on day four or later. This conservative measurement approach bypasses the recommended use of antibiotics as a prophylactic measure to reduce postoperative infection rates for abdominal hysterectomy [19,20]. Our exclusion of antibiotic use in the initial days of hospitalization parallels the method used by Yokoe et al. [18] in a retrospective database study.

Patient characteristics

Demographics

Patients were described in terms of age, marital status, and ethnicity/race. The seven age categories in years were: 18–24, 25–34, 35–44, 45–54, 55–64, 65–74, and 75 plus. Marital status was coded as married, other, separated/divorced, single, or widowed, whereas ethnicity/race was described as black, white, Hispanic, and other.

Clinical characteristics

Patients were assessed with respect to illness severity, risk of death, and specific co-morbid conditions. A severity of illness rating was available in the database, which was scored in accordance with the All Patient Refined Diagnostic Related Groups (APR DRGs), a widely established and externally validated standard within the industry developed by 3M Health Information Systems [21]. As applied to each inpatient stay, the severity of illness rating describes the extent of physiologic decompensation or organic system loss of function and thereby reflects both the underlying problem and the multiplicity of problems.

Each of three co-morbid conditions was coded dichotomously—obesity, diabetes mellitus, and gynecological cancer. The presence of obesity was limited to an ICD-9 diagnosis (ICD-9: 278.00), whereas diabetes was either ICD-9: 250.x, 790.2, or treatment with insulin or oral hypoglycemic agents. In addition, the presence of a cancer, either as a primary or a metastatic condition, was identified according to ICD-9 codes. Studies show that wound healing may be influenced by the presence of obesity or diabetes mellitus [22, 23].

Hospital characteristics

Hospital characteristics identified in the database and linked to each patient record were teaching hospital (yes/no), urban local (yes/no), actual number of hospital beds, and region (Midwest, Northeast, South, or West).

Statistical analysis

Treatment groups were compared on demographic, clinical, and hospital characteristics. Unadjusted comparisons also were conducted on all outcome variables. One-way analysis of variance (ANOVA) and chi-square tests were performed for continuous and categorical outcomes, respectively. For continuous outcomes (e.g., LOS), analysis of covariance (ANCOVA) was used to examine the effect of treatment group while controlling for all relevant covariates. For dichotomous outcomes (i.e., dehiscence, SSI, and prophylactic antibiotic use), logistic regression was conducted to examine the effect of treatment group while adjusting for other dimensions. Adjusted odds ratios (ORs) and 95 percent confidence intervals (CIs) were calculated as part of the logistic regression results. Final models derived from multivariable statistical analyses included grouping variable, as well as demographic, clinical, and hospital dimensions that demonstrated significant effects in a preliminary multivariable evaluation.

Because of the multiple statistical comparisons, an alpha of ≤0.01 (two-tailed) was used. For all continuous outcome dimensions, when the overall p value was significant, pairwise group comparisons were conducted using t-tests. For all categorical outcomes, pairwise comparisons were conducted using chi-square tests or logistic regression.

Results

In the total sample of eligible patients (n = 46,011), the distribution of skin closures was 21,201 sutures, 23,441 staples, 880 OCA, and 489 staples + OCA. As shown in Table 1, all assessed comparisons among the four treatment groups on demographic, clinical, and hospital characteristics were significantly different (p < 0.001) because of the large sample size. In this section, we highlight the differences that seem most relevant.

Table 1.

Characteristics of Patients Hospitalized for Abdominal Hysterectomy by Skin Closure Group

	Sutures(n = 21,201)	Staples(n = 23,441)	OCA(n = 880)	Staples + OCA(n = 489)	P value
Demographics
Age, mean (SD)(years)	46.2 (10.8)	48.0 (12.3)	45.4 (11.1)	48.5 (12.2)	<0.001
Age categories (years), % (n)
18–24	0.6 (118)	0.6 (134)	0.8 (7)	0 (0)
25–34	9.7 (2,059)	9.2 (2,160)	12.2 (107)	9.6 (47)
35–44	38.8 (8,216)	34.4 (8,070)	37.8 (333)	31.5 (154)
45–54	35.2 (7,468)	33.3 (7,810)	34.7 (333)	35.8 (175)
55–64	8.4 (1,778)	11.0 (2,578)	7.3 (64)	10.8 (53)
65–74	4.6 (969)	6.8 (1,584)	5.1 (45)	7.2 (35)
≥75	2.8 (593)	4.7 (1,105)	2.2 (19)	5.1 (25)
Marital status, % (n)					<0.001
Married	60.2 (11,848)	59.4 (13,089)	70.1 (587)	68.5 (324)
Other	5.1 (1,002)	5.5 (1,213)	1.7 (14)	1.9 (9)
Separated/divorced	11.3 (2,232)	11.3 (2,499)	7.9 (66)	9.7 (46)
Single	19.4 (3,824)	18.1 (3,982)	16.8 (141)	15.2 (72)
Widowed	4.0 (789)	5.7 (1,251)	3.6 (30)	4.7 (22)
Ethnicity/race, % (n)					<0.001
Black	20.0 (4,170)	18.9 (4,437)	19.7 (173)	14.1 (69)
Hispanic	5.5 (1,160)	4.9 (1,141)	3.3 (29)	3.7 (18)
Other	15.5 (3,070)	13.5 (3,153)	7.1 (62)	7.6 (37)
White	60.4 (12,801)	62.8 (14,710)	70.0 (616)	74.6 (365)
Clinical
Severity,^a mean (SD)	1.3 (0.5)	1.4 (0.6)	1.3 (0.5)	1.5 (0.7)	<0.001
Gynecologic cancer, % (n) yes	9.4 (1,989)	17.2 (4,035)	7.5 (66)	18.2 (89)	<0.001
Obesity, % (n) yes	7.3 (1,553)	9.1 (2,137)	5.5 (48)	9.0 (44)	<0.001
Diabetes, % (n) yes	7.9 (1,665)	10.7 (2,499)	7.3 (64)	9.4 (46)	<0.001
Hospital
Teaching hospital,% (n) yes	41.7 (8,833)	41.4 (9,702)	51.8 (456)	54.2 (265)	<0.001
Urban hospital, % (n) yes	89.1 (18,886)	87.6 (20,536)	95.9 (844)	81.4 (447)	<0.001
Region, % (n)					<0.001
Midwest	19.6 (4,144)	16.2 (3,795)	18.5 (163)	18.8 (92)
Northeast	13.5 (2,853)	9.3 (2,180)	10.0 (88)	2.7 (13)
South	55.3 (11,733)	65.0 (15,239)	67.1 (590)	67.9 (332)
West	11.7 (2,471)	9.5 (2,227)	4.4 (39)	10.6 (52)
Number of hospital beds, mean (SD)	420.8 (217)	444.5 (232)	392.6 (266)	427.8 (228)	<0.001

Severity of illness scored in accordance with the All Patient Refined Diagnostic Related Groups (APR DRGs). See reference 23.

OCA = 2-octyl cyanoacrylate. There was no other cyanoacylate usage in the database.

Demographic and clinical characteristics

Patients receiving sutures or OCA were two to three years younger (mean age 46.2 and 45.4 years, respectively) than patients receiving staples or staples + OCA (mean age 48.0 and 48.5 years, respectively). Compared with patients who had sutures, patients who had OCA had similar disease severity but with somewhat lower proportions of gynecological cancer, diabetes, and obesity. In contrast, patients in each of the two staples groups had higher severity and higher proportions with cancer, diabetes, and obesity than patients in the non-staples groups.

Length of stay and total costs

Table 2 summarizes the unadjusted and adjusted results for group comparisons on mean LOS and total costs. In unadjusted results, the mean LOS ranged from 3.7 days for OCA to 5.2 days for staples + OCA. A similar pattern was found for mean total costs, which ranged from $5,816 for OCA to $9,434 for staples + OCA. For the unadjusted results for mean LOS and mean total costs, pairwise comparisons indicated that the means for staples and staples + OCA were significantly higher than the means for sutures or OCA. However, there were no significant differences between sutures and OCA in these two outcomes. As described in the table (see footnote 4), results of pairwise comparisons of adjusted results were similar. Although the adjusted comparison between OCA and staples on total costs did not meet the criterion for significance, the pattern suggested lower costs for OCA (p = 0.039).

Table 2.

Unadjusted and Adjusted Outcomes for Abdominal Hysterectomy by Skin Closure Group

Outcomes	Sutures	Staples	OCA	Staples + OCA	Overall P value ¹	Significant pairwise comparisons ²	Overall P value ¹	Significant pairwise comparisons ²
Continuous	Unadjusted means				Unadjusted analysis ³		Adjusted analysis ^3,4
Length of stay (days)⁵	3.9	4.5	3.7	5.2	<0.001	1^a,3^a,4^a,5^a,6^a	<0.001	1^a,3^a,4^a,5^a,6^a
Total costs (dollars)	5,862	6,965	5,816	9,434	<0.001	1^a,3^a,4^a,5^a,6^a	<0.001	1^a,3^a,5^a,6^a
Total antibiotic days⁶	1.9	2.3	1.8	3.2	<0.001	1^a,3^a,4^a,5^a,6^a	<0.001	1^a,3^a,5^a,6^a
Total antibiotic cost	52.9	78.6	48.3	113.2	<0.001	1^a,3^a,4^a,5^a,6^a	<0.001	1^a,3^a,5^a,6^a
N of antibiotic claims	2.5	3.0	2.3	4.0	<0.001	1^a,3^a,4^a,5^a,6^a	<0.001	1^a,3^a,4^b,5^a,6^a

Categorical	Percent				Unadjusted analysis ³		Adjusted analysis ³
Dehiscence	0.17	0.33	0.34	0.61	0.003	1	0.294	–
Surgical site infection	0.61	0.90	0.33	1.02	0.002	1	0.382	–
Non-prophylactic antibiotic treatment	12.93	17.51	11.14	23.72	<0.001	1^a,3^a,4^a,5^a,6^a	<0.001	1^a,3^a,5^a,6^a

Pairwise comparisons conducted only if overall p value ≤0.01.

Adjusted analysis controlling for significant demographic, clinical, and hospital dimensions.

Significant pairwise comparisons: a = p ≤ 0.001; b = p ≤ 0.01.

1 sutures versus staples

2 sutures versus OCA

3 sutures versus staples + OCA

4 staples versus OCA

5 staples versus staples + OCA

6 OCA versus staples + OCA.

Least square (adjusted) means by group in following order: sutures, staples, OCA, staples + OCA.

Length of stay: 10.0, 10.3, 9.9, 10.8

Total costs in dollars: 19,346, 19,901, 19,518, 21,816

Total antibiotic days: 7.5, 7.7, 7.5, 8.3

Total antibiotic costs: 327.4, 343.7, 327.0, 369.4

Number of antibiotic claims: 9.7, 10.0, 9.6, 10.7.

In the four groups, the proportions hospitalized for four or more days (vs. three days or less) was sutures 49.3%, staples 59.7%, OCA 41.5%, staples + OCA 64.8%.

Total number of antibiotic days calculated as mean of last antibiotic service day.

OCA = 2-octyl cyanoacrylate. There was no other cyanoacrylate usage in the database.

Clinical outcomes

In the sample as a whole, there were 120 dehiscences and 348 SSIs. Table 2 examines the rates of these events by treatment group. In unadjusted analyses, rates of dehiscence ranged from 0.17% in patients with sutures to 0.61% in patients with staples + OCA (p = 0.003). For SSI, rates ranged from 0.34% in patients with OCA to 1.02% in patients with staples + OCA (p = 0.002). In unadjusted pairwise tests, the rates of dehiscence and SSI were significantly higher for staples than for sutures (p < 0.0001), whereas all other (unadjusted) comparisons failed to meet the significance criterion. In adjusted analyses, overall comparisons failed to meet the significance criterion, and therefore, no pairwise tests were conducted.

Table 2 also presents results related to non-prophylactic antibiotic use. In unadjusted results, the proportions of use were 11.14%, 12.93%, 17.51%, and 23.72% for patients receiving OCA, sutures, staples, and staples + OCA, respectively (p < 0.001). In unadjusted analyses, all pairwise comparisons indicated significantly less antibiotic use for patients receiving OCA at p ≤ 0.01 with the exception of OCA vs. sutures. In adjusted analyses, the pattern of less antibiotic use for patients receiving OCA remained significant at p ≤ 0.01 with two exceptions: OCA vs. sutures (p = 0.011) and OCA vs. staples (p = 0.030). Moreover, when the risk of non-prophylactic antibiotic treatment was calculated relative to the sutures group, the risk was the same in the group having OCA (OR_adj 1.0; 95% CI 0.8, 1.3), 30% higher in the group having staples (OR_adj 1.3; 95% CI 1.2, 1.4), and 100% higher in the group having staples + OCA (OR_adj 2.0; 95% CI 1.6, 2.5).

In the subsample of patients who had an LOS of four days or more, antibiotic usage was significantly different by treatment group. The proportions were 26.21%, 26.85%, 29.31%, and 36.59% for patients receiving sutures, OCA, staples, and staples + OCA, respectively (p < 0.001).

Other markers of antibiotic use showed a pattern of similar results. In unadjusted results, the mean number of antibiotic claims was 2.3, 2.5, 3.0 and 4.0, while the total mean cost of antibiotics was $48.30, $52.90, $78.60, and $113.20 for patients receiving OCA, sutures, staples and staples + OCA, respectively (p < 0.001). Moreover, in unadjusted analyses of number of antibiotic claims and costs, all pairwise comparisons met the significance criterion of p ≤ 0.01 with the exception of the difference between OCA and sutures. In adjusted analyses of these two variables, the pattern of the results was similar (see footnote 6 in Table 2). Although there was no difference between OCA and sutures (p = 0.570 and 0.876, respectively), other pairwise comparisons were at or near the significance criterion of 0.01.

Discussion

This study, conducted using a large multi-hospital administrative database, demonstrated similar infection-related outcomes for patients whose incisions were closed with OCA and patients whose skin was sutured. These comparisons were confirmed in adjusted multivariable analyses that controlled for hospital characteristics and patient attributes. Furthermore, these results are consistent with randomized trials of skin closure that found similar outcomes for adhesives and suture [1,4,9].

In contrast, patients treated with staples had poorer infection-related outcomes than patients treated with sutures or OCA. In adjusted analyses, the differences between staples and sutures were uniformly significant, whereas differences between staples and OCA showed a consistent pattern of less favorable infection-related outcomes for staples, although this did not always meet the designated significance threshold. When the sutured incision group was used as the reference group, adjusted analyses showed that the risk of non-prophylactic antibiotic treatment was the same in the group having OCA, 30% higher in the group having staples, and 100% higher in the group having staples + OCA. In a review based on five randomized, controlled trials, use of topical sutures was superior to staples with respect to the complication rate for chest and leg incisions in patients undergoing cardiovascular surgery [24]. This pattern of results has not emerged in prior trials that compared OCA with staples [3,10] and merits further study.

An advantage of the use of staples rather than sutures for surgical closure is the time saved by the faster application [3,25]. However, in total hysterectomy, this time saving appears to increase the risk of infection and may add to LOS and increase the hospital costs.

Patients who had staples + OCA had the poorest infection-related outcomes among the four groups. The proportion of patients treated in this way (∼1%) was modest. Although the adjusted analyses attempted to control for baseline differences between treatment groups, there may be unmeasured factors that differentiated the groups. The use of OCA in addition to staples in hysterectomy procedures may warrant further investigation in view of these initial findings.

Because SSIs lead to greater morbidity and more deaths among surgical patients and add to health care costs, the incidence is receiving more attention as an institutional marker of quality assurance [26]. Although the price of OCA is higher than that of sutures, in this analysis, total LOS and hospital costs for patients who received OCA were similar to those in patients receiving sutures and lower than (i.e., superior to) results among patients receiving staples. It is possible that decreased use of anti-infectives and related savings in these medication costs is contributing to the offset of OCA acquisition costs.

The overall rates of dehiscence and SSI were low, 0.3% and 0.8%, respectively. These results may underestimate the risk of SSI because follow-up usually was limited to the inpatient stay. Although the assessment of SSI was augmented by identification of in-hospital, non-prophylactic treatment with antibiotics, a recognized strategy for monitoring SSI [18], the method does not extend information-gathering beyond the inpatient phase. A recent study in Scotland reported SSI rates of 6.03% and 3.14% among patients with and without post-discharge surveillance [27], suggesting that the SSI rate almost doubles when follow-up continues after discharge. Although hospital stays were about one day longer in the Scottish study than in the current study (mean 5.4 days vs. 4.23 days), the infection rate during the hospital period in the former study was markedly higher [27].

In the United States, approximately 600,000 hysterectomies are performed annually. In the late 1990s, national data indicated that the proportion of total and subtotal abdominal hysterectomies was 63% [28]. Despite the increasing use of laparoscopy for hysterectomy [4], a recent study suggests that the proportion of abdominal hysterectomies remains above 60% [29]. For this reason, examination of factors that address safety and costs relevant to this high-volume procedure remains pertinent.

A strength of conducting a study in a database including large numbers of patients in community-based care is that it is possible to examine pre-specified relationships that cannot be addressed in smaller clinical samples. Although large samples are heterogeneous, the results can be adjusted for differences using multivariable statistical approaches, as in this study. Nevertheless, potentially useful information, such as the National Nosocomial Infection Surveillance Surgical-Site Infection (NNIS SSI) risk index, was not available [30]. There are other limitations that should be considered when interpreting the study results. With this dataset, it was not possible to ascertain whether topical skin adhesives were used for primary closure of the subcuticular layer or applied over topical suture for additional strength, as a dressing, or both. In addition, there was no way to determine the method used for closure of the layer of subcutaneous fat above the fascia. Variability in surgical techniques within each group would make it more difficult to identify differences across treatments.

With respect to postoperative antibiotic treatment, there is no certainty that all prescriptions were administered for SSI. However, other conditions for use should be distributed equally across treatment groups. In this study, the definition of non-prophylactic antibiotic treatment started at day four, a more conservative definition than the day two employed by Yokoe et al. and Stevenson et al. [18,31]. Although adjusted results for continuous variables confirmed patterns in unadjusted analyses, the actual numbers were not interpretable.

The physician's choice of closure methods depends on his/her assessment of the patient's condition and other risk factors. Vertical vs. Pfannenstiel incision, early ambulation, and other attributes also may impact the outcomes selected. Despite adjustments for multiple variables, it is possible that other patient or hospital characteristics not assessed in this database influenced the poorer outcomes for skin closure with staples.

Although comparative cohort studies conducted in claims data lack the rigor of randomized assignment to treatment and uniform criteria for treatment and follow-up, their main advantage is large numbers, inclusion of patients with co-morbid conditions who may be excluded from trials, and treatment conditions that reflect patterns of care in the community. A strength of the data is the large numbers of patients treated with sutures and staples, as well as the substantial number treated with OCA (n = 880). When samples are large, even small correlations can be significant. Nevertheless, the magnitude of the among-group differences in outcomes and the consistency of the findings appear to be clinically meaningful, and the differences in LOS have implications for patients and payers. Because the number of patients who received OCA was much smaller than the numbers who received sutures or staples, it was difficult to obtain significant results for comparisons between OCA and other skin closure procedures. Although OCA was superior to staples in bivariable analyses of most outcomes, at the multivariable level, a significant advantage for OCA was found only in relation to two outcomes (LOS and number of antibiotic claims).

In conclusion, in this study of more than 46,000 women having total abdominal hysterectomies, when compared with sutures, staples appeared to increase the incidence of infection, add to the LOS, and increase total costs as well as the costs of anti-infective treatment. In contrast, skin closure with OCA appeared to be a safe and cost-effective alternative to topical sutures.

Footnotes

Acknowledgments

The authors thank Sheri Dodd, MSc, Anuprita Patkar, PhD, and Brian B. Vaughn, MPA, for their editorial contributions.

Author Disclosure Statement

This study was funded by Ethicon Inc., Somerville, NJ. All authors were either partly funded by the sponsoring company as consultants (SGM, JSM, EMG) or were company employees (GM). One author (SGM) has received speaking honoraria from Ethicon.

An abstract based on a portion of the results described in this manuscript was presented as a poster at the 56^th Annual Clinical Meeting of the American College of Obstetricians and Gynecologists in New Orleans, Louisiana, May 3–7, 2008.

References

Coulthard

, Worthington

, Esposito

et al. Tissue adhesives for closure of surgical incisions (review). The Cochrane Collaboration. John Wiley & Sons, Ltd., 2006.

Blondeel

PNV

, Murphy

, Debrosse

et al. Closure of long surgical incisions with a new formulation of 2-octylcyanoacrylate tissue adhesive versus commercially available methods. Am J Surg, 2004; 188:307–313.

Khan

, Fick

, Yao

et al. A comparison of three methods of wound closure following arthroplasty: A prospective, randomized, controlled trial. J Bone Joint Surg Br, 2006; 88:238–242.

Gennari

, Rotmensz

, Ballardini

et al. A prospective, randomized, controlled clinical trial of tissue adhesive (2-octylcyanoacrylate) versus standard wound closure in breast surgery. Surgery, 2004; 136:593–599.

Greene

, Kock

, Goode

. Efficacy of octyl-2-cyanoacrylate tissue glue in blepharoplasty. Arch Facial Plast Surg, 1999; 1:292–296.

Nowobilski

, Dobosz

, Wojciechowicz

, Mionskowska

. Lichtenstein inguinal hernioplasty using butyl-2-cyanoacrylate versus sutures: Preliminary experience of a prospective randomized trial. Eur Surg Res, 2004; 36:367–370.

Jallali

, Haji

, Watson

. A prospective randomized trial comparing 2-octylcyanoacrylate to conventional suturing in closure of laparoscopic cholecystectomy incisions. J Laparoendosc Adv Surg Tech A, 2004; 14:209–211.

Maartense

, Bemelman

, Dunker

et al. Randomised study of the effectiveness of closing laparoscopic trocar wound with octylcyanoacrylate, adhesive papertape or poliglecaprone. Br J Surg, 2002; 89:1370–1375.

Matin

. Prospective randomized trial of skin adhesive versus sutures for closure of 217 laparoscopic port-site incisions. J Am Coll Surg, 2003; 196:845–853.

10.

Ridgeway

, Mahmood

, Moore

et al. A blinded, randomized, controlled trial of stapled versus tissue glue closure of neck surgery incisions. Ann R Coll Surg Engl, 2007; 89:242–246.

11.

Shamiyeh

, Schrenk

, Stelzer

, Wayand

. Prospective randomized blind controlled trial comparing sutures, tape, and octylcyanoacrylate tissue adhesive for skin closure after phlebectomy. Dermatol Surg, 2001; 27:877–880.

12.

Ozturan

, Miman

, Aktas

, Oncel

. Butylcyanoacrylate tissue adhesive for columellar incision closure. J Laryngol Otol, 2001; 115:535–540.

13.

Sinha

, Naik

, Wright

et al. A single blind, prospective, randomized trial comparing n-butyl 2-cyanoacrylate tissue adhesive (Indermil) and sutures for skin closure in hand surgery. J Hand Surg, 2001; 26B:264–265.

14.

Toriumi

, O'Grady

, Desai

, Bagal

. Use of octyl-2-cyanoacrylate for skin closure in facial plastic surgery. Plastic Reconstruct Surg, 102:2209–2219.

15.

Magee

, Ajkay

, Githae

, Rosenblum

. Use of octyl-2-cyanoacrylate in cleft lip repair. Ann Plast Surg, 2003; 50:1–5.

16.

Keshavarz

, Hillis

, Kiele

, Marchbanks

. Hysterectomy surveillance–United States, 1994–1999. MMWR CDC Surveillance Sum, 2002; 51,SS05:1–8.

17.

International Classification of Diseases. 9^th Revision, Clinical Modification, Sixth Edition (ICD-9-CM) Salt Lake City, UT: Ingenix/St. Anthony Publishing/Medicode, 2002.

18.

Yokoe

, Noskin

, Cunningham

et al. Enhanced identification of postoperative infections among inpatients. Emerg Infect Dis, 2004; 10:1924–1930. www.cdc.gov/ncidod/eid/vol10no11/04-0572.htm. 2007 March 15.

19.

Ledger

. Prophylactic antibiotics in obstetrics–gynecology: A current asset, a future liability? Expert Rev Anti-Infect Ther, 2006; 4:957–964.

20.

Lofgren

, Poromaa

, Stjerndahl

, Renstrom

. Postoperative infections and antibiotic prophylaxis for hysterectomy in Sweden: A study by the Swedish National Register for Gynecologic Surgery. Acta Obstet Gynecol Scand, 2004; 83:1202–1207.

21.

Averill

, Goldfield

, Hughes

et al. What are APR-DRGs? An introduction to severity of illness and risk of mortality adjustment methodology. 3M Health Information Systems. http://multimedia.mmm.com/mws/mediawebserver.dyn?6666660Zjcf6lVs6EVs666IupCOrrrrQ-. 2007 March 15.

22.

Hibbard

, Gilbert

, Landon

et al. Trial of labor or repeat cesarean delivery in women with morbid obesity and previous cesarean delivery. Obstet Gynecol, 2006; 108:125–133.

23.

Takoudes

, Weitzen

, Slocum

, Malee

. Risk of cesarean wound complications in diabetic gestations. Am J Obstet Gynecol, 2004; 191:958–963.

24.

Sanni

, Dumming

. Staples or sutures for chest and leg wounds following cardiovascular surgery. Interact Cardiovasc Thorac Surg, 2007; 6:243–246.

25.

Ranaboldo

, Rowe-Jones

. Closure of laparotomy wounds: Skin staples versus sutures. Br J Surg, 1992; 79:1172–1173.

26.

Hedrick

, Anastacio

, Sawyer

. Prevention of surgical site infections. Summ Exp Rev Anti-infect Ther, 2006; 4:223–233.

27.

Reilly

, Allardice

, Bruce

et al. Procedure-specific surgical site infection rates and postdischarge surveillance in Scotland. Infect Control Hosp Epidemiol, 2006; 27:1318–1323.

28.

Farquhar

, Steiner

. Hysterectomy rates in the United States 1990–1997. Obstet Gynecol, 2002; 99:229–234.

29.

Diouf

, Greenberg

, McElrath

. Do laparoscopic hysterectomies supplant abdominal or vaginal hysterectomies? Obstet Gynecol, 2007; 109,4 Suppl:19S–20S.

30.

Gaynes

, Culver

, Haran

et al. Surgical site infection (SSI) rates in the US, 1992–1998; the National Nosocomial Infections Surveillance System Basic SSI Risk Index. Clin Infect Dis, 2001; 33,2 Suppl:S69–S77.

31.

Stevenson

, Freshwater

, Santangelo

et al.

Using electronic health data as predictors of surgical site infections (SSI)

Poster presented at Society for Healthcare Epidemiology of America. Orlando, FL, April 5–8, 2008.