The Results of a Propensity Score Matching Analysis of the Efficacy of Abdominal Fascia and Skin Closure Using PDS ® Plus Antibacterial (Polydioxanone) Sutures on the Incidence of Superficial Incisional Surgical Site Infections after Gastroenterologic Surgery

Abstract

Background:

Surgical site infections (SSIs) lead to prolonged hospitalization and increased cost of hospital stay after surgery. Therefore, the prevention of SSIs is one of the most critical tasks facing surgeons and nursing staff. In the present study, the efficacy of using triclosan-coated polidioxanone sutures (PDS^® Plus Antibacterial Suture, (Ethicon Inc., Somerville, NJ) for abdominal closure was analyzed retrospectively using a propensity score matching analysis.

Patients and Methods:

Of 1,768 patients who underwent gastroenterologic surgery at Fukuoka University Hospital between January 2009 and September 2013, 812 underwent abdominal closure using PDS Plus. These patients were compared retrospectively regarding the incidence of SSIs using a propensity score matching method with 956 patients treated in the previous period without abdominal closure using PDS Plus sutures. The propensity score was calculated from the age, gender, body mass index, smoking status, diabetes mellitus, use of steroid medication, malignant or benign disease, organ location, emergency or planned surgery, wound classification, the American Society of Anesthesiologists score, open or laparoscopic surgery, the length of the operation, and blood loss.

Results:

Nine hundred sixty-six patients (483 matched sets) were enrolled by the propensity score matching method. No parameter used for the propensity score was different between the PDS Plus and control groups. Closure using PDS Plus could reduce the incidence of SSIs compared with that in the control group (p = 0.022). Of the parameters used for the propensity score, malignant disease (p = 0.0002), open surgery (p = 0.0020), a prolonged operation (p < 0.0001), high blood loss (p < 0.0001), the need for a transfusion (p = 0.019), and gastrointestinal tract surgery (p = 0.0059) were significant risk factors for the development of SSIs in the univariable analysis. In a multivariable regression model, open surgery (p < 0.0001), prolonged operation (p < 0.0001), gastrointestinal tract surgery (p = 0.001), and abdominal closure without PDS Plus (p < 0.0001) were the independent risk factors for the development of an SSI. The development of an SSI prolonged the hospital stay (p < 0.0001) and the use of antibiotic medication (p < 0.0001); abdominal closure using PDS Plus affected the antibiotic medication period (p = 0.013) but not the hospital stay (p = 0.40).

Conclusion:

Although abdominal fascia and skin closure using PDS Plus was compared with variable abdominal closure, the present findings suggest that abdominal fascia and skin closure using PDS Plus sutures could help prevent the development of SSIs after gastroenterologic surgery, as determined by a propensity score matching analysis.

Surgical site infections (SSIs) are one of the common complications after gastroenterologic surgery, and the incidence of SSIs has been reported to range from 10% to 30%. Surgeons have attempted to decrease the incidence of SSIs, because they are linked to prolonged hospital stay and increased costs, as well as being dangerous to patients [1 –4]. To date, less intra-operative blood loss, frequent glove changes, syringe pressure irrigation of subcutaneous tissue, the use of subcuticular sutures and placing drains have been adopted as a measure to reduce the risk of SSIs after abdominal closure [2,5,6].

Triclosan-coated polyglactin 910 sutures (Vicryl^® Plus, Ethicon Inc., Somerville, NJ) were developed recently as a new material to reduce the risk of SSIs [7,8]. Previous studies at our institute [9] and by Galal et al. [10] demonstrated the superiority of Vicryl Plus, with our institute demonstrating that the new sutures reduced the incidence of SSI from 12.2% to 6.6% after gastroenterologic surgery. However, incision closure using Vicryl Plus sutures did not reduce the incidence of SSIs after breast cancer surgery or coronary artery bypass grafting [11,12]. The efficacy of the antimicrobial-coated sutures therefore seems to depend on the type of surgery.

Polydioxanone sutures (PDS) were reported to be able to inhibit bacterial adhesion in vitro better than Vicryl, and triclosan-coated PDS (PDS® Plus, Ethicon Inc., Somerville, NJ) might be a promising suture material for reducing the incidence of SSIs further [13]. However, there is currently not enough evidence available concerning the efficacy of triclosan-coated PDS for preventing SSIs, because the few studies that have been reported have been controversial [14 –16]. Thus, we examined the efficacy of PDS Plus to reduce superficial incisional SSIs in abdominal closure after gastroenterologic surgery using our institutional data.

Patients and Methods

Patients

The end point of the present study was to clarify the efficacy of abdominal closure using PDS Plus sutures in reducing the incidence of superficial incisional SSI after gastroenterologic surgery. We examined the correlation between abdominal closure using PDS Plus and the incidence of superficial incisional SSI after gastroenterologic surgery by performing an historic control study using a propensity score matched analysis. The present study was approved by Fukuoka University Hospital Clinical Research Assist Center on August 22, 2012 (no. 12-7-06).

A total of 1,768 patients underwent gastroenterologic surgery at Fukuoka University Hospital between January 2009 and September 2013 and were enrolled in the present study. The patients with deeper organ/space SSIs were excluded from the present study in order to evaluate the local control efficacy of abdominal fascia and skin closure using PDS Plus. The characteristics of these patients are summarized in Table 1. We compared the incidence of superficial incisional SSI statistically using a propensity score matching method retrospectively between 812 patients treated in the latter period (September 2012 to September 2013) with abdominal skin and fascia closure using PDS Plus sutures and 956 patients treated during the former period (January 2009 to June 2011) without PDS Plus suturing.

Table 1.

The Correlation between the Clinical Parameters and the Incidence of Surgical Site Infections for the Total Patients Population

	SSI (−)	SSI (+)	p
Age (years)	64.6 ± 15.1	65.1 ± 13.4	0.65^a
Gender
Male	942	121	0.011^b
Female	650	54
BMI (kg/m²)	21.6 ± 3.6	21.5 ± 3.6	0.77^a
Diabetes mellitus
(+)	233	25	0.89^b
(−)	1355	150
Smoking
(+)	414	52	0.31^b
(−)	1143	123
Steroid use
(+)	23	8	0.0028^c
(−)	1566	167
Disease
Malignant	883	123	0.0002^b
Benign	708	52
Surgery
Planned	1351	130	0.0003^b
Emergency	239	45
Laparoscopic	449	18	0.0001^b
Open	960	153
Transfusion
(+)	146	27	0.0065^b
(−)	1432	144
Blood loss	262.8 ± 660.1	482.8 ± 791.1	<0.0001^a
Length of operation	216.5 ± 168.3	297.1 ± 203.6	<0.0001^a
Wound classification
1	404	11	<0.0001^c
2	1057	130
3	91	23
4	30	11
ASA score
1	384	20	<0.0001^c
2	985	98
3	168	43
4	48	11
5	4	3
Organ location
Upper GI	433	30	0.0007^b
Lower GI	590	83
HBP	261	18
Others	305	44
PDS -Plus
(+)	593	35	<0.0001^b
(−)	988	140

Student t-test.

Chi-square test.

Mann Whitney U-test.

SSI = surgical site infections; BMI = body mass index; ASA = American Society of Anesthesiologists; GI = gastrointestinal; HBP = hepato-biliary-pancreatic duct.

Surgical technique and evaluation of SSI

During the earlier period, the abdominal fascia was closed with the interrupting sutures using 1-Vicryl Plus in 467 patients and 1-Vicryl in 489 patients, and the type of skin closure was left to the discretion of the surgeon, however, triclosan-coated sutures were not used. PDS Plus sutures were introduced at Fukuoka University Hospital in September 2012, and not only the fascia but also the skin was closed with interrupting sutures using 1 and 4-0 PDS Plus sutures, respectively, during the later period. Interrupted subcutaneous sutures were used for skin closure in the later period. The abdominal closure in both groups usually was performed by residents who were well trained at our institute not to vary technically, and their technical variance was controlled minimum. The incidence of SSI was evaluated according to the U.S. Centers for Disease Control and Prevention criteria [17].

Propensity score matching

Propensity score matching was performed retrospectively by creating similar case (later period) and control (earlier period) sets using our institutional data. The propensity score was calculated from the age, gender, body mass index, smoking status, diabetes mellitus, steroid medication, malignant or benign disease, organ location, emergency or planned surgery, wound classification, American Society of Anesthesiologists (ASA) score, open or laparoscopic surgery, length of the operation, and blood loss.

Of 1,768 patients, 483 pairs were matched 1:1 and were included in the statistical analysis. Of matched patients in the earlier period, 1-Vicryl Plus was used for abdominal fascia closure in 273 and 1-Vicryl was used in 210 patients. The clinical-pathologic characteristics of the patients after matching are summarized in Table 2. There were no significant differences in any of these characteristics between the two groups in the earlier and later periods.

Table 2.

The Efficacy of Triclosan-Coated Absorbable Sutures as Determined by Propensity Score Matching

	PDS Plus	Control	p
Age (years)	64.8 ± 14.6	66.0 ± 13.9	0.20^a
Gender
Male	261	255	0.58^b
Female	222	228
BMI	21.6 ± 3.6	21.5 ± 3.5	0.79^a
Diabetes mellitus
(+)	72	69	0.70^b
(−)	411	414
Smoking
(+)	140	136	0.69^b
(−)	343	347
Steroid use
(+)	6	3	0.082^b
(−)	477	480
Disease
Malignant	305	285	0.064^b
Benign	178	198
Surgery
Planned	440	440	0.99^b
Emergency	43	43
Laparoscopic	145	134	0.26^b
Open	338	349
Transfusion
(+)	46	38	0.18^b
(−)	437	445
Blood loss	222.7 ± 483.7	284.6 ± 563.1	0.067^a
Length of operation	236.2 ± 177.6	225.7 ± 167.6	0.35^a
Wound classification
1, 2	466	462	0.37^b
3, 4	17	21
ASA score
1, 2	459	458	0.84^b
3, 4	24	25
Organ location
Upper GI	91	88	0.42^b
Lower GI	178	173
HBP	131	147
Others	83	75

Student t-test.

Chi-square test.

PDS = polidioxanone sutures; BMI = body mass index; ASA = American Society of Anesthesiologists; GI = gastrointestinal; HBP = hepato-biliary-pancreatic duct.

Statistical analysis

The efficacy of using the triclosan-coated polydioxanone sutures for abdominal closure was examined in univariable and multivariable analyses using Student t-test, χ², and a multivariable logistic regression model. Values of p < 0.05 were considered to be statistically significant. The statistical analyses were performed using the Statistical Package for the Social Sciences software program, version 22 (SPSS Inc., Chicago, IL).

Results

Retrospective analysis of correlation between SSI and clinical characteristics

The correlation between SSIs and clinical characteristics are summarized in Table 1. Abdominal closure without PDS Plus was one of risk factors in the univariable analysis as well as male gender, steroid medication, malignant disease, emergency operation, open abdominal surgery, transfusion, blood loss, prolonged operation time, wound classification, ASA score, and organ location (Table 1).

Propensity score matching

Of the 1,768 patients, 483 sets were matched and analyzed after propensity score matching. The characteristics used to calculate the propensity score are summarized in Table 2. There were no statistically significant differences between the PDS Plus and control groups.

Efficacy of PDS Plus against SSI

After the propensity score matching, the incidence of SSI after surgery was 5.8% in the PDS Plus group and 9.7% in the control group. PDS Plus reduced the incidence of SSIs compared with the control group (p = 0.022; Table 3). The development of an SSI prolonged the hospital stay (p < 0.0001) and the duration of treatment with antibiotics (p < 0.0001; Table 4), regardless of whether the incisions were closed with or without PDS Plus sutures (p < 0.0001, p < 0.0001, p < 0.0001, p = 0.0004, respectively). Although abdominal closure using PDS Plus sutures could reduce the length of the antibiotic treatment (p = 0.013), it did not reduce the hospital stay (p = 0.40; Table 5) and abdominal closure using PDS Plus sutures had no impact on the hospital stay or antibiotic treatment period in patients with or without SSIs.

Table 3.

The Efficacy of PDS Plus for Preventing Surgical Site Infections after Propensity Score Matching

	PDS Plus	Control	p
SSI
(+)	28	47	0.022^a
(−)	455	436

Chi-square test.

PDS = polidioxanone sutures; SSI = surgical site infection.

Table 4.

The Clinical Impact of Surgical Site Infections on the Hospital Stay or the Need for Antibiotic Medication after Propensity Score Matching

	SSI (−)	SSI (+)	p
Hospital stay (days)	17.3 ± 14.6	43.0 ± 50.3	p < 0.0001^a
Antibiotic treatment (days)	3.7 ± 2.0	7.3 ± 9.9	p < 0.0001^a

Chi-square test.

SSI = surgical site infection.

Table 5.

The Clinical Impact of PDS Plus on the Hospital Stay and Use of Antibiotics after Propensity Score Matching

	PDS Plus	Control	p
Hospital stay (days)	18.8 ± 16.4	19.9 ± 24.7	0.40^a
Antibiotic treatment (days)	3.7 ± 1.9	4.3 ± 4.6	0.013^a

Chi-square test.

PDS = polidioxanone sutures.

Factors influencing the risk of SSIs

After the propensity score matching, as risk factors that were associated with the development of an SSI, in addition to abdominal closure without PDS Plus, malignant disease (p = 0.0002), open surgery (p = 0.002), a prolonged operation (p < 0.0001), increased blood loss (p < 0.0001), transfusion (p = 0.019), and gastrointestinal tract surgery (p = 0.0059) were revealed to correlate with a high incidence of SSIs in a univariable analysis (Table 6). Of these risk factors for SSI, open surgery (p < 0.0001), prolonged operation (p < 0.0001), gastrointestinal tract surgery (p = 0.001), and abdominal closure without PDS Plus (p < 0.0001) were the independent risk factors for SSIs identified in the multivariable logistic regression model (Table 7).

Table 6.

The Correlation between the Clinical Parameters and Incidence of Surgical Site Infections after Propensity Score Matching

	SSI (−)	SSI (+)	p
Age (years)	65.3 ± 14.5	67.1 ± 11.2	0.29^a
Gender
Male	471	45	0.23^b
Female	420	30
BMI (kg/m²)	21.6 ± 3.6	21.6 ± 3.6	0.97^a
Diabetes
(+)	129	12	0.72^b
(−)	762	63
Smoking
(+)	248	28	0.080^a
(−)	643	47
Steroid use
(+)	7	2	0.15^c
(−)	884	73
Disease
Malignant	529	61	0.0002^b
Benign	362	14
Surgery
Planned	814	66	0.33^b
Emergency	77	9
Laparoscopic	269	10	0.0020^b
Open	622	65
Transfusion
(+)	72	12	0.019^b
(−)	819	63
Blood loss	233.4 ± 479.6	494 ± 878.0	<0.0001^a
Length of operation	222.8 ± 166.1	327.6 ± 215.6	<0.0001^a
Wound classification
1, 2	859	69	0.059^b
3. 4	32	6
ASA score
1, 2	849	68	0.080^b
3, 4	42	7
Organ location
Upper GI	165	14	0.0059^b
Lower GI	327	24
HBP	268	4
Others	151	7

Student t-test.

Chi-square test.

Upper gastrointestinal tract.

SSI = surgical site infection; BMI = body mass index; ASA = American Society of Anesthesiologists; GI = gastrointestinal; HBP = hepato-biliary-pancreatic duct.

Table 7.

The Results of the Univariable and Multivariable Analyses of Factors Correlated with the Development of Surgical Site Infections after Propensity Score Matching

	p (univariable)	Hazard ratio	Confidence interval	p (multivariable)
Disease (malignant/benign)	0.0002	0.887	0.585–1.346	0.57
Surgery (open/laparoscopic)	0.002	0.27	0.160–0.456	<0.0001
Blood loss	<0.0001	1	1.000–1.000	0.35
Transfusion	0.019	0.963	0.537–1.725	0.9
Length of operation	<0.0001	1.002	1.001–1.003	<0.0001
Organ	0.0059	0.711	0.585–0.864	0.001
Suture (PDS Plus/control)	0.022	0.404	0.269–0.607	p < 0.0001

PDS = polidioxanone sutures.

Discussion

The efficacy of PDS Plus sutures for the prevention of SSI was demonstrated previously in vivo and in vitro, and the colonization by Staphylococcus aureus and Escherichia coli was inhibited [18]. Our previous study using Vicryl Plus suggested that these sutures had the potential to reduce SSIs when used for abdominal fascia closure [9]. In addition, PDS Plus seemed to reduce SSIs better than Vicryl Plus, and the superiority of PDS compared with Vicryl in terms of the bacterial adherence was demonstrated previously by Chu and Williams [13]. However, the recent randomized controlled studies could not reach a consensus regarding whether abdominal closure using PDS Plus sutures reduced the risk of SSIs [14 –16]. Justinger et al. [15] reported that abdominal closure using PDS Plus reduced the incidence of SSI from 11.3% to 6.4%. In contrast, there was no significant difference between the PDS Plus (14.8%) and uncoated suture (PDS II) (16.1%) groups in the study by Diener et al. [16], and Bracs et al. [14] also suggested no superiority of PDS Plus compared with PDS II in colorectal surgery [14,16].

In the present study, significant efficacy of PDS Plus sutures for the prevention for SSIs was demonstrated. One reason for this finding might depend on the differences in study design, wherein not only the abdominal fascia but also the skin was closed using PDS Plus sutures in the present study. This is in contrast to the previous studies in which only abdominal fascia closure was performed using PDS Plus sutures and the skin was closed using staples. Interestingly, Tsujinaka et al. [19] reported no superiority of reducing SSIs for abdominal skin closure using PDS II compared with staples (8.4% versus 11.5%) using a randomized controlled study with a large series.

Therefore, taking the present results and the findings of previous reports into consideration, a single improvement such as closing only the fascia using PDS Plus or closing only the skin using PDS Plus might be of little use for reducing the incidence of SSIs. However, performing a combination of such treatments, such as closing both the skin and fascia using PDS Plus, might help to reduce the incidence of SSIs.

In addition, malignant disease, open laparotomy, prolonged operation, more extensive blood loss, blood transfusion, and gastrointestinal tract surgery, as well as abdominal closure without PDS Plus, were suggested to be risk factors for SSI in the univariable analysis. These factors were consistent with the previous literature regarding the risk factors for SSIs [2,20]. Of these risk factors, open laparotomy, a prolonged operation, gastrointestinal tract surgery, and abdominal closure without PDS Plus sutures were also independent risk factors in the multivariable analysis.

Abdominal closure using PDS Plus sutures might be a promising procedure to reduce SSIs in the abdominal wall, because PDS Plus sutures were one of the independent factors found to reduce the risk of SSI. It is noteworthy that gastrointestinal surgery was an independent risk factor for SSI in the present study, which is in agreement with a previous report [10]. Our results suggested that abdominal fascia and skin closure using PDS Plus sutures may be useful to reduce the risk of SSIs after gastrointestinal surgery.

The hospitalization period could not be shortened directly by abdominal wall closure using PDS Plus sutures in the present study, although the incidence of SSIs was reduced by PDS Plus. We believe that this was because other critical respiratory or circulatory complications after surgery, other than superficial incisional SSIs in the abdominal wall, might have prolonged the hospital stay, although the use of PDS Plus sutures might correlate indirectly with a shorter hospital because the development of a superficial incisional SSI was correlated with a prolonged hospital stay.

In the present study, propensity score matching was adopted in order to eliminate the statistical bias, although there might be a limitation in that it was an historic control study [21]. As the weaknesses of the present analysis, abdominal closure in the control group was manifold and might mean few comparisons such as standard technique versus random technique, monofilament suture versus braid suture, and triclosan-coated sutures used in abdominal fascia and skin closure versus fascia or skin. However, the results of the present study suggest that abdominal closure using PDS Plus sutures for the skin as well as the fascia might be a promising procedure to reduce the risk of SSIs that can be implemented immediately without the need for special techniques. A randomized control study would be helpful to confirm the present findings, but such a study might be difficult to plan and might be ethically unsuitable because the superiority of PDS Plus for the prevention for SSIs was already verified in vitro and in vivo during the process of development for the market [18]. Therefore, we believe that a propensity score matching study might be suitable to clarify the efficacy of PDS Plus. In addition, the present historic control study using propensity score matching might surpass a randomized control study in some senses. First, the existing data at the institute can be used without the need to collect any additional data or to recruit new patients; second, it can be performed inexpensively. It is indisputable that propensity score matching is a useful statistical method for use in a retrospective study.

In conclusion, the present findings suggest that abdominal fascia and skin closure using PDS Plus sutures might help to reduce the risk of SSIs after gastroenterologic surgery. A more detailed analysis of the efficacy of the present procedure to determine what kind(s) of surgery or what status of patients would be best suited for this treatment would lead to the optimal application of PDS Plus, because these sutures are more expensive than the sutures without an antimicrobial coating.

Footnotes

Author Disclosure Statement

The authors have not been supported by any organization in the present study.

References

Smith

, Bohl

, McElearney

, et al. Wound infection after elective colorectal resection. Ann Surg, 2004; 239:599–605.

Watanabe

, Kohnoe

, Shimabukuro

, et al. Risk factors associated with surgical site infection in upper and lower gastrointestinal surgery. Surg Today, 2008; 38:404–412.

Perencevich

, Sands

, Cosgrove

, et al. Health and economic impact of surgical site infections diagnosed after hospital discharge. Emerg Infect Dis, 2003; 9:196–203.

Fry

. A systems approach to the prevention of surgical infections. Surg Clin North Am, 2009; 89:521–537.

Cervantes-Sánchez

, Gutiérrez-Vega

, Vázquez-Carpizo

, et al. Syringe pressure irrigation of subdermic tissue after appendectomy to decrease the incidence of postoperative wound infection. World J Surg, 2000; 24:38–41.

Inotsume-Kojima

, Uchida

, Abe

, et al. A combination of subcuticular sutures and a drain for skin closure reduces wound complications in obese women undergoing surgery using vertical incisions. J Hosp Infect, 2011; 77:162–165.

Rothenburger

, Spangler

, Bhende

, Burkley

. In vitro antimicrobial evaluation of Coated VICRYL* Plus Antibacterial Suture (coated polyglactin 910 with triclosan) using zone of inhibition assays. Surg Infect, 2002; 3(Suppl 1):S79–87.

Storch

, Rothenburger

, Jacinto

. Experimental efficacy study of coated VICRYL plus antibacterial suture in guinea pigs challenged with Staphylococcus aureus. Surg Infect, 2004; 5:281–288.

Hoshino

, Yoshida

, Tanimura

, et al. A study of the efficacy of antibacterial sutures for surgical site infection: A retrospective controlled trial. Int Surg, 2013; 98:129–132.

10.

Galal

, El-Hindawy

. Impact of using triclosan-antibacterial sutures on incidence of surgical site infection. Am J Surg, 2011; 202:133–138.

11.

Williams

, Sweetland

, Goyal

, et al. Randomized trial of antimicrobial-coated sutures to prevent surgical site infection after breast cancer surgery. Surg Infect, 2011; 12:469–474.

12.

Seim

, Tønnessen

, Woldbaek

. Triclosan-coated sutures do not reduce leg wound infections after coronary artery bypass grafting. Interact Cardiovasc Thorac Surg, 2012; 15:411–415.

13.

Chu

, Williams

. Effects of physical configuration and chemical structure of suture materials on bacterial adhesion. A possible link to wound infection. Am J Surg, 1984; 147:197–204.

14.

Baracs

, Huszár

, Sajjadi

, Horváth

. Surgical site infections after abdominal closure in colorectal surgery using triclosan-coated absorbable suture (PDS Plus) vs. uncoated sutures (PDS II): A randomized multicenter study. Surg Infect, 2011; 12:483–489.

15.

Justinger

, Slotta

, Ningel

, et al. Surgical-site infection after abdominal wall closure with triclosan-impregnated polydioxanone sutures: Results of a randomized clinical pathway facilitated trial (NCT00998907). Surgery, 2013; 154:589–595.

16.

Diener

, Knebel

, Kieser

, et al. Effectiveness of triclosan-coated PDS Plus versus uncoated PDS II sutures for prevention of surgical site infection after abdominal wall closure: The randomised controlled PROUD trial. Lancet, 2014; 384:142–152.

17.

Horan

, Gaynes

, Martone

, et al. CDC definitions of nosocomial surgical site infections, 1992: A modification of CDC definitions of surgical wound infections. Infect Control Hosp Epidemiol, 1992; 13:606–608.

18.

Ming

, Rothenburger

, Nichols

. In vivo and in vitro antibacterial efficacy of PDS plus (polidioxanone with triclosan) suture. Surg Infect, 2008; 9:451–457.

19.

Tsujinaka

, Yamamoto

, Fujita

, et al. Subcuticular sutures versus staples for skin closure after open gastrointestinal surgery: A phase 3, multicentre, open-label, randomised controlled trial. Lancet, 2013; 382:1105–1112.

20.

Hawn

, Vick

, Richman

, et al. Surgical site infection prevention: Time to move beyond the surgical care improvement program. Ann Surg, 2011; 254:494–499.

21.

D'Agostino

Jr.

Propensity score methods for bias reduction in the comparison of a treatment to a non-randomized control group. Stat Med, 1998; 17:2265–2281.