Beyond Core Measures: Identifying Modifiable Risk Factors for Prevention of Surgical Site Infection after Elective Total Abdominal Hysterectomy

Abstract

Background:

Despite adherence to the Centers for Medicare and Medicaid Services (CMS) core measures for preventing surgical site infections (SSI), our institution has a >10% rate of SSI after total abdominal hysterectomy (TAH), higher than the 90^th percentile for SSI rates published in the 2009 National Healthcare Safety Network report.

Methods:

A retrospective chart review was performed for patients who underwent elective TAH at a public safety net hospital in Denver from December 30, 2005, to March 9, 2010. The primary outcome was development of SSI within 30 days. A secondary outcome was adherence to CMS core measures.

Results:

A total of 192 patients were included in the analysis, of whom 21 (10.9%) developed SSI. More than 95% had received antibiotics in the 60 min before surgical incision, and >90% received an appropriate antibiotic. Compliance with post-anesthesia care unit normothermia was equivalent in the SSI and non-SSI groups (81.0% vs. 75.2%; p=0.5588). Surgical site infection was associated with obesity (body mass index [BMI]≥30) (15.4% vs. 6.9%; p=0.0609), estimated blood loss≥500 mL (18.5% vs. 8.0%; p=0.0353), and receipt of a blood transfusion (28.6% vs. 10.5%; p=0.0183). In a multiple logistic regression model, obesity marginally increased the risk of SSI (odds ratio [OR] 2.55; 95% confidence interval [CI] 0.94–6.74), whereas blood transfusion was significantly associated with a higher risk of SSI (OR 3.58; 95% CI 1.21–10.62).

Conclusions:

Blood transfusion was associated with SSI after TAH in our population. As it is a modifiable risk factor, larger multi-center studies are needed to confirm this result and determine appropriate transfusion thresholds.

In August 2009, the State of Colorado began requiring reporting of the rates of surgical site infections (SSIs) after total abdominal hysterectomy (TAH) via the National Healthcare Safety Network (NHSN). In preparation for this requirement, we reviewed the rate of SSI after TAH at our institution and found it to be >10%, higher than the national average of <2% [1]. This high rate was despite attempts to ensure close adherence to the Centers for Medicare and Medicaid Services (CMS) core measures designed to prevent SSI. These measures include administration of peri-operative antibiotics within 60 min of surgical incision, appropriate choice of antibiotic (cefazolin, cefotetan, cefoxitin, cefuroxime, or ampicillin-sulbactam; or clindamycin plus an aminoglycoside, quinolone, or aztreonam; or metronidazole plus an aminoglycoside or quinolone), an appropriate method for removal of hair, discontinuation of antibiotics within 24 h of the surgical procedure, removal of any urinary catheter on post-operative day two, and prevention of peri-operative hypothermia [2].

We hypothesized that there are additional modifiable risk factors that contribute to the development of SSI after TAH that should be explored when core measures are pursued aggressively yet the rate of SSI remains unacceptably high.

Patients and Methods

Study design

This was a retrospective cohort study of patients who underwent TAH at a public safety net hospital in Denver between December 30, 2005, and March 9, 2010. Patients who had an emergency TAH, operations for tubo-ovarian abscess, or no followup visits in the 30 days after TAH were excluded. The primary outcome was SSI in the 30 days after TAH. A secondary outcome was adherence to the CMS core measures designed to prevent SSI.

Definitions

Surgical site infections are defined by the U.S. Centers for Disease Control and Prevention (CDC) as infections at the site of surgery within 30 days after an operation or within one year of an operation if a foreign body is implanted as part of the surgery. They are further divided into three classes: superficial incisional, deep incisional, and organ/space [3]. Superficial incisional SSIs involve only the skin and subcutaneous tissue at the incision site. There must be either purulent drainage; organisms isolated from a culture of fluid or tissue; a diagnosis of superficial incisional SSI by the surgeon or attending physician; or pain, swelling, redness, or heat in a wound that is opened deliberately by a surgeon and is culture-positive or not cultured. Deep incisional SSIs involve the fascia or muscle layers of the incision site and must have purulent drainage; an abscess; a diagnosis of deep incisional SSI by the surgeon or attending physician; or pain, swelling, redness, or heat in a wound that either dehisces spontaneously or is opened deliberately by a surgeon and is culture-positive or not cultured. Organ/space SSIs are defined as infections that involve any part of the body that is manipulated during an operation excluding the skin incision, fascia, or muscle layers. In addition, they must have either purulent drainage from a drain in the organ or space, organisms isolated from a culture of fluid or tissue, an abscess, or a diagnosis of an organ/space SSI by the surgeon or attending physician [3].

Data collection

Charts were abstracted by manual review to obtain information on variables that may be associated with SSI in the 30 days after TAH. These variables were age, body mass index (BMI), medical and surgical histories, surgeon, length of time in the operating room, concomitant operative procedures, estimated blood loss (EBL), duration of hospitalization, performance of blood transfusion, pre-operative complete blood count and creatinine assay, and post-operative hemoglobin and hematocrit. Data regarding the timing, presence, classification, and microbiology of the SSI also were noted.

Data regarding CMS core measures including the peri-operative antibiotic choice and timing and post-anesthesia care unit (PACU) arrival temperature were recorded. This hospital does not use razors to remove hair on any patients, so data were not collected for this core measurement. Information regarding discontinuation of peri-operative antibiotics and removal of urinary catheters was not collected, as it was not considered to be pertinent to the outcome of interest.

Statistical analysis

Descriptive statistics were used to analyze patient demographics and surgical characteristics. Chi-square tests of independence were used to determine associations between categorical variables, whereas differences-in-means t-tests were used to test for associations between continuous variables and the outcomes of interest. In addition, we examined potential associations between blood transfusion and PACU hypothermia, EBL, length of time in the operating room, and pre- and post-operative hemoglobin and hematocrit values. For patients who received a blood transfusion, objective markers of tissue hypoperfusion, including heart rate, blood pressure, orthostatic changes in vital signs, dizziness, and shortness of breath were noted, as well as the documented reason for the transfusion. Multiple logistic regression models were used to determine whether risk factors with significant bivariable associations were associated independently with the development of SSI. Variables were included in the multivariable logistic regression models if the p value corresponding to the bivariable test of association with SSI was <0.10. Otherwise, hypothesis tests were performed at the 0.05 significance level. Data were analyzed using SAS Version 9.2 for Windows (SAS Institute, Cary, NC). The study was approved by the Colorado Multiple Institutional Review Board.

Results

The records of 224 patients who underwent TAH at our institution were reviewed. Emergency operations (n=26), operations for tubo-ovarian abscess (n=4), and patients with no documented followup in the first 30 days after surgery (n=2) were excluded, leaving 192 records for analysis. Surgical site infections developed in 21 patients (10.9%) at a mean of 13.8 days (standard deviation 6.77) after surgery. The National Nosocomial Infections Surveillance (NNIS) classification was determined for all patients, and the rate of SSI in patients stratified by NNIS class is listed in Table 1. Our rate of SSI was higher than nationally reported rates for all NNIS risk index categories.

Table 1.

Observed vs. National Rate of Surgical Site Infection (SSI) after Total Abdominal Hysterectomy, Stratified by National Nosocomial Infections Surveillance Risk Index Category

Risk index category^a (n)	SSI n (%)	National rate ^b (%)
0 (30)	1 ( 3.33)	1.10
1 (116)	11 ( 9.48)	2.20
2 or 3 (46)	9 (19.57)	4.05

Calculated by adding one point for an American Society of Anesthesiologists Score≥3 points, one point for wound classification ≥3 (i.e., contaminated or dirty), and one point for operative time >60 min.

Data from Edwards J, Peterson K, Mu Y, et al. National Healthcare Safety Network (NHSN) report: Data summary for 2006 through 2008, issued December 2009. Am J Infect Control 2009;37:783–805.

With the exception of BMI, the medical and surgical histories of patients who developed SSI and those who did not were not different (Table 2). Surgical site infection was more common in obese (BMI≥30) than in non-obese patients (15.4% vs. 6.9%; p=0.0609). The most common reasons for TAH were removal of symptomatic leiomyomas (n=98; 51.0%), treatment of abnormal vaginal bleeding (n=97; 50.5%), management of a gynecologic malignant or pre-malignant condition (n=41; 21.4%), and evaluation of a pelvic mass (n=36; 18.8%). The indication for elective TAH was not associated with the development of SSI.

Table 2.

Characteristics of Elective Total Abdominal Hysterectomy Population, by Development of Surgical Site Infection vs. No Surgical Site Infection

	N without SSI (%)	N with SSI (%)	p value
Age (years)±SD	46.6±9.66	47.8±8.49	NS
Body mass index ≥30	76 (45.1)	14 (66.7)	0.0609
Diabetes mellitus	18 (10.5)	4 (19.1)	NS
Smoking	58 (33.9)	4 (19.1)	NS
Cancer
Gynecologic	30 (17.5)	5 (23.8)	NS
Non-gynecologic	7 ( 3.5)	0 ( 0.0)	NS
Pre-operative hematocrit (%)±SD	39.4±4.86	37.9±4.10	NS
Post-operative hematocrit (%)±SD	31.8±5.69	31.8±4.13	NS
Delta hematocrit (%)±SD	7.6±4.94	6.1±4.06	NS
Length of hospitalization (days)±SD	3.4±3.06	4.4±4.01	NS
Length of operating time (hours)±SD	2.8±1.28	3.0±0.87	NS
Estimated blood loss ≥500 mL	47 (25.7)	10 (47.6)	0.0353
Blood transfusion	18 (10.5)	6 (28.6)	0.0183

NS=not significant; SSI=surgical site infection; SD=standard deviation.

Key medical and surgical variables that were found to impact the development of SSI are described in Table 2. Neither diabetes mellitus nor smoking was associated with the development of SSI in this patient population. Surgical site infection was more common in patients who had EBL≥500 mL than in those with EBL<500 mL (18.5% vs. 8.0%; p=0.0353). Infection also was associated with blood transfusion (25.0% vs. 8.9%; p=0.0183). Blood transfusion was not associated with the pre- to post-operative change in the hematocrit, but it was associated with longer operative times, higher EBL, and both the pre- and post-operative hematocrit value (Table 3). The most commonly documented reasons for blood transfusion were anemia (n=24; 72.7%), high-volume intra-operative EBL (n=9; 27.3%), dizziness (n=8; 24.2%), and co-morbid cardiac conditions (n=6; 18.2%).

Table 3.

Characteristics of Patients Who Did and Did Not Receive Blood Transfusion

	No transfusion	Transfusion of ≥1 unit packed RBCs	p value
Pre-operative hematocrit (%)±SD	39.8±4.48	34.7±4.58	<0.0001
Post-operative hematocrit (%)±SD	32.5±5.30	26.7±4.37	<0.0001
Delta hematocrit (%)±SD	7.36±4.77	8.03±5.54	NS
Estimated blood loss ≥500 mL	39 (23.2)	15 (62.5)	<0.0001
Length of operation (h)±SD	2.7±1.01	4.1±1.89	<0.0001

NS=not significant; RBC=red blood cell; SD=standard deviation.

Multiple logistic regression models revealed that obesity was marginally associated with an increased risk of SSI, whereas blood transfusion was strongly associated with a higher risk (Table 4). We stratified the analysis by EBL≥500 mL vs. <500 mL and found that blood transfusion was a stronger predictor of SSI in patients with lower EBL.

Table 4.

Multiple Logistic Regression Models Predicting Odds Ratio (95% Confidence Interval) of Surgical Site Infection

	Model I ^a	Model II ^b
	(n=192)	EBL<500 mL (n=138)	EBL≥500 mL (n=54)
Blood transfusion	3.58 (1.21–10.62)	7.56 (1.59–35.98)	1.14 (0.25–5.16)
Body mass index ≥30	2.55 (0.96–6.74)	–	–

Multivariable logistic regression controlling for blood transfusion and obesity.

Multivariable logistic regression controlling for blood transfusion, stratified by EBL<500 mL vs. EBL≥500 mL.

Compliance with CMS core measures for prevention of SSI is reported in Figure 1. More than 95% of patients received antibiotics in the 60 min preceding surgical incision, and >90% of patients received an appropriate antibiotic by Surgical Care Improvement Project criteria [2]. There was no difference in SSI rates between patients who received appropriate vs. inappropriate antibiotic prophyloxis (7.7% vs. 9.5%; p=0.7691). Sixty-three patients had an operation that lasted >3 h, and 10 (15.9%) received a second dose of antibiotics between two and four hours after the initial dose. The patients who received a second dose appropriately because the operative time exceeded 3 h did not have a lower rate of SSI than those who had a longer operation but did not receive a second antibiotic dose (10.0% vs. 17.0%; p=0.5795. Compliance with PACU normothermia was suboptimal, with only 75.8% of patients having the lowest PACU temperature of >36°C. Hypothermia was not associated with a higher rate of SSI in our population.

FIG. 1.

Compliance with Centers for Medicare and Medicaid Services core measures for prevention of surgical site infections.

Discussion

This study was prompted by our unacceptably high rate of SSI after TAH in comparison with the national average. We became aware of this high rate as a result of an impending mandate for public reporting of SSI rates. The problem was brought to the attention of the Department of Obstetrics and Gynecology at our facility, and the rate of SSI decreased to 2% (n=50) over the next 12 mos in the absence of formal intervention relative to the risk factors identified in this study. There is a paucity of published data regarding the effect of active SSI surveillance and surgeon-specific feedback on the rate of SSI. Although some investigators support our findings that SSI surveillance and feedback produce lower SSI rates, others have found that these measures are not associated with lower rates of SSI [4 –7]. To the best of our knowledge, there are no published data on whether public reporting of SSI rates leads to better outcomes. Public reporting of SSI rates is a relatively new requirement in many states, and its impact on SSI rates will be an interesting area of future research.

Irrespective of this decrease in SSI after the gynecology faculty was alerted to the problem, the findings of our statistical analyses merit further discussion. Three factors, one of which is modifiable, were found to be associated with a higher rate of SSI after elective TAH: Obesity, EBL≥500 mL during surgery, and the need for blood transfusion.

Elevated BMI had a moderate association with a higher rate of SSI after TAH in our study, in agreement with the published literature [8,9]. This is not a newly described risk factor for SSI in surgical patients; obesity has been associated with a higher risk of SSI in abdominal and gynecologic procedures [10], revisions of total hip arthroplasties [11], renal transplantation [12], lower-extremity vascular bypass grafting [13], and others. The reason for the higher rates of SSI in obese patients is unclear. It is possible that these patients have poorer tissue perfusion than non-obese patients and therefore have less tissue oxygen, which may predispose to superficial and deep incisional SSIs. Alternatively, obese patients may have a higher rate of SSI because peri-operative antibiotic tissue concentrations are inadequate. At our institution, we do not change the antibiotic dose according to body weight; rather, all adult patients are given the same dose. This topic certainly merits additional study.

In addition to obesity, blood transfusion and EBL≥500 mL were associated with higher rates of SSI in our study. Published reports that greater intra-operative EBL is a risk factor for many types of infection after TAH, including SSI [14,15]. Blood transfusion was implicated in the development of SSI after TAH in one study [9] and in many non-gynecologic procedures, including general [16], colorectal [17], cardiothoracic [18], orthopedic [19], and vascular operations [16,20]. The relation between blood transfusion and SSI is being described increasingly, yet the biologic mechanism remains unknown. Further work is needed to establish causality.

It is not surprising that EBL≥500 mL and blood transfusion were linked closely in our analysis. Among patients with EBL<500 mL, blood transfusion increased the risk of SSI, an association not observed among patients with EBL≥500 mL. This finding certainly needs validation with larger data sets, although it raises interesting possibilities for implementation of strict criteria for blood transfusion in elective TAH.

Our hospital's performance on core measures was mixed. Although >95% of patients received antibiotics within 60 min of surgical incision, almost 8% received an inappropriate antibiotic by Surgical Care Improvement Project (SCIP) criteria, although the relevance of SCIP criteria to clinical outcomes is increasingly controversial [21 –23]. Most of these patients received clindamycin alone as peri-operative prophylaxis. We were surprised to find compliance with PACU normothermia was suboptimal; however, hypothermia in the PACU was not associated with SSI in this analysis. Previous studies have linked mild peri-operative hypothermia to a higher risk of SSI after colorectal operations and cholecystectomies [24,25]. Interestingly, one of these papers also noted that mildly hypothermic post-operative patients receive more blood transfusions than normothermic patients [24], a relation that was not replicated in our analysis.

The strengths of this study include the use of standardized definitions for SSI and the small number of patients lost to follow-up (n=2). As the study location is a public safety net hospital, it is a closed system; and patients perhaps have less difficulty obtaining followup surgical and primary care. Previous studies in this field have been limited by their ability to review only hospital records, a strategy known to underestimate the true number of SSIs [26 –28]. The weaknesses of the study include its retrospective design, use of data from a single center, and the small number of SSIs (n=21) over a five-year period.

In conclusion, we found blood transfusion to be associated with SSI after TAH in our population. Blood transfusion is a potentially modifiable risk factor that was not associated with the pre- to post-operative change in hematocrit; future studies that involve multiple institutions should be undertaken to ensure that these results are reproducible. If blood transfusion is confirmed to be a risk factor for SSI after TAH, we will need to both characterize surgeons' transfusion practices and determine appropriate thresholds for transfusion in this population. Obesity also was associated with SSI after TAH in our population. Larger peri-operative antibiotic doses may be required to achieve adequate tissue concentrations in such patients; alternatively, encouragement of weight loss before elective surgery may be considered to decrease the rate of SSI. Finally, we found that the process of measuring SSI rates and providing feedback to the gynecology faculty was sufficient to decrease the rate of SSI after elective TAH.

Author Disclosure Statement

No conflicting financial interests exist.

References

Edwards

, Peterson

, Mu

et al. National Healthcare Safety Network (NHSN) report: Data summary for 2006 through 2008, issued December 2009. Am J Infect Control, 2009; 37:783–805.

Specifications Manual for National Hospital Inpatient Quality Measures. Oak Brook, IL: The Joint Commission, 2010.

U.S. Centers for Disease Control and Prevention. Healthcare-Associated Infections (HAIs) Atlanta The Centers, 2010. www.cdc.gov/ncidod/dhqp/healthDis.html April 17, 2010.

Barwolff

, Sohr

, Geffers

et al. Reduction of surgical site infections after caesarean delivery using surveillance. J Hosp Infect, 2006; 64:156–161.

Gastmeier

, Sohr

, Brandt

et al. Reduction of orthopaedic wound infections in 21 hospitals. Arch Orthop Trauma Surg, 2005; 125:526–530.

Creamer

, Cunney

, Humphreys

, Smyth

. Sixteen years' surveillance of surgical sites in an Irish acute-care hospital. Infect Control Hosp Epidemiol, 2002; 23:36–40.

Kasatpibal

, Jamulitrat

, Chongsuvivatwong

et al. Impact of surgeon-specific feedback on surgical site infection rates in Thailand. J Hosp Infect, 2006; 63:148–155.

Soper

, Bump

, Hurt

. Wound infection after abdominal hysterectomy: Effect of the depth of subcutaneous tissue. Am J Obstet Gynecol, 1995; 173:465–469.

Olsen

, Higham-Kessler

, Yokoe

et al. Developing a risk stratification model for surgical site infection after abdominal hysterectomy. Infect Control Hosp Epidemiol, 2009; 30:1077–1083.

10.

Thomas

, Goldman

, Mangione

et al. Body mass index as a correlate of postoperative complications and resource utilization. Am J Med, 1997; 102:277–383.

11.

Lubbeke

, Moons

, Garavaglia

, Hoffmeyer

. Outcomes of obese and nonobese patients undergoing revision total hip arthroplasty. Arthritis Rheum, 2008; 59:738–745.

12.

Lynch

, Ranney

, Shijie

et al. Obesity, surgical site infection, and outcome following renal transplantation. Ann Surg, 2009; 250:1014–1020.

13.

Giles

, Hamdan

, Pomposelli

et al. Body mass index: Surgical site infections and mortality after lower extremity bypass from the National Surgical Quality Improvement Program 2005–2007. Ann Vasc Surg, 2010; 24:48–56.

14.

Persson

, Bergstrom

, Larsson

et al. Infections after hysterectomy: A prospective nationwide Swedish study. The Study Group on Infectious Diseases in Obstetrics and Gynecology within the Swedish Society of Obstetrics and Gynecology. Acta Obstet Gynecol Scand, 1996; 75:757–761.

15.

Meltomaa

, Makinen

, Taalikka

, Helenius

. Incidence, risk factors and outcome of infection in a 1-year hysterectomy cohort: A prospective follow-up study. J Hosp Infect, 2000; 45:211–217.

16.

Campbell

Jr , Henderson

, Englesbe

et al. Surgical site infection prevention: The importance of operative duration and blood transfusion—results of the first American College of Surgeons–National Surgical Quality Improvement Program Best Practices Initiative. J Am Coll Surg, 2008; 207:810–820.

17.

Tang

, Chen

, Want

et al. Risk factors for surgical site infection after elective resection of the colon and rectum: A single-center prospective study of 2,809 consecutive patients. Ann Surg, 2001; 234:181–189.

18.

Murphy

, Reeves

, Rogers

et al. Increased mortality, postoperative morbidity, and cost after red blood cell transfusion in patients having cardiac surgery. Circulation, 2007; 116:2544–2552.

19.

Innerhofer

, Klingler

, Klimmer

et al. Risk for postoperative infection after transfusion of white blood cell-filtered allogeneic or autologous blood components in orthopedic patients undergoing primary arthroplasty. Transfusion, 2005; 45:103–110.

20.

O'Keeffe

, Davenport

, Minion

et al. Blood transfusion is associated with increased morbidity and mortality after lower extremity revascularization. J Vasc Surg, 2010; 51:616–621.

21.

Barie

. SCIP to the Loo? Surg Infect, 2011; 12:161–167.

22.

Rosenberger

, Politano

, Sawyer

. The surgical care improvement project and prevention of post-operative infection, including surgical site infection. Surg Infect, 2011; 12:163–168.

23.

Edmiston

, Spencer

, Lewis

et al.

Reducing the risk of surgical site infections: Did we really think SCIP was going to lead us to the promised land?

Surg Infect, 2011; 17:169–177.

24.

Kurz

, Sessler

, Lenhardt

. Perioperative normothermia to reduce the incidence of surgical-wound infection and shorten hospitalization. Study of Wound Infection and Temperature Group. N Engl J Med, 1996; 334:1209–1215.

25.

Flores-Maldonado

, Medina-Escobedo

, Rios-Rodriguez

, Fernandez-Dominguez

. Mild perioperative hypothermia and the risk of wound infection. Arch Med Res, 2001; 32:227–231.

26.

Reilly

, Allardice

, Bruce

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

27.

Daneman

, Lu

, Redelmeier

. Discharge after discharge: Predicting surgical site infections after patients leave hospital. J Hosp Infect, 2010; 75:188–194.

28.

Petrosillo

, Drapeau

, Nicastri

et al. Surgical site infections in Italian hospitals: A prospective multicenter study. BMC Infect Dis, 2008; 8:34.