The Association of Preoperative Length of Stay on Surgical Site Infection After Lower Extremity Bypass
Luke M Stewart, Benjamin J Pearce, Danielle C Sutzko, Graeme E McFarland, Emily L Spangler, Zdenek Novak, Adam W Beck
University of Alabama at Birmingham, Birmingham, AL
BACKGROUND: Lower extremity bypass (LEB) for treatment of arterial occlusive disease remains a commonly performed procedure with 126 performed per 100,000 U.S. citizens annually. Surgical site infection (SSI) remains an important source of patient morbidity and increased cost to the healthcare system following LEB. Despite the rate of postoperative SSI varying widely from 4-31% in previous studies, there is consensus that it remains a significant complication following LEB and a potential source of future quality improvement. Postoperative SSI has been associated with an increase in healthcare costs of over $10,000 as well as extended hospital length of stay. Additionally, SSI has been associated with increased rates of reoperation, loss of patency, limb loss, and decreased patient quality of life. Prior studies have identified a number of factors associated with increased SSI including increased procedure length, blood transfusion, iodine surgical prep in comparison to chlorhexidine, increased age, female gender, obesity, diabetes, dialysis-dependence, preoperative anticoagulant use, and adjunctive femoral endarterectomy. As issues of hospital-acquired infections and resistant organisms have become targets of quality initiatives, we postulated duration in hospital prior to surgery may place patients at additional infectious risk perioperatively. Increased rates of resistant infections such as methicillin-resistant Staphylococcus aureus (MRSA) have been seen with increasing hospital length of stay in non-surgical patients. The effect of preoperative length of stay (LOS) on SSI rate has previously been understudied in the field of Vascular Surgery. In other domains of surgery, increased preoperative LOS has been associated with increased rate of postoperative complications. In Cardiac surgery, it has been shown that prolonged preoperative LOS is associated with increased development of postoperative mediastinitis while in Neurosurgery it has been shown that inpatient status at the time of spine surgery is associated with increased rates of postoperative SSI. We therefore sought to examine the effect of preoperative LOS on postoperative SSI rate following LEB and studied factors which may lead to increased preoperative LOS.
METHODS: A retrospective analysis of the Society for Vascular Surgery Vascular Quality Initiative (VQI) infrainguinal bypass registry was used to identify patients undergoing elective LEB for critical limb ischemia (rest pain or tissue loss) from 2003-2020. Patients undergoing LEB for acute limb ischemia, urgent or emergent procedures, infrainguinal aneurysm, had a bypass origin distal to the above-knee popliteal artery, distal bypass target at the ankle or foot level, or who had concomitant suprainguinal bypass were excluded. Additionally, patients who were transferred from another institution were excluded as the total preoperative length of stay could not be determined in this group. 17,822 LEB were selected for inclusion and classified into preoperative LOS groups of 0 days (12,428 LEB), 1-2 days (1,701 LEB), and 3-14 days (3,693 LEB). Postoperative SSI was defined as the occurrence of documented wound infection or graft infection during index hospital stay for LEB. Demographic and perioperative variables were analyzed using X2 and ANOVA, where appropriate. Variables with p<.10 were included in multivariate logistic regression analysis.
RESULTS: Surgical site infection rate increased with increasing preoperative LOS (0 days: 2.7%, 1-2 days: 3.5%, 3-14 days: 4.1%; p<.01) (Figure 1). 
Rate of return to operating room for infection during index admission also increased with increasing preoperative LOS (0 days: 0.7%, 1-2 days: 0.9%, 3-14 days: 2.3%; p<.01).
Among demographic and comorbid factors, dialysis, diabetes requiring insulin, CHF, ASA class 4 (compared to 3 or less), a bedridden functional status, or residing in a nursing home preoperatively were present with increased frequency in 1-2 and 3-14 preop LOS groups (Table 1).
| Factor | 0 Days Preop LOS | 1-2 Days Preop LOS | 3-14 Days Preop LOS | p |
| Age | 67.6±10.8 | 68.2±11.2 | 67.4±11.3 | .04 |
| Diabetes on Insulin | 27.6% | 34.1% | 38.7% | <.01 |
| CHF | 17.5% | 19.3% | 25.1% | <.01 |
| Dialysis | 4.7% | 8.6% | 11.2% | <.01 |
| ASA Class 4 | 21.0% | 26.3% | 31.3% | <.01 |
| Nursing Home Preop | 2.6% | 4.6% | 4.6% | <.01 |
| Ambulatory without assistance Preop | 70.3% | 61.3% | 60.2% | <.01 |
| Procedure length >290 min | 26.7% | 27.3% | 32.1% | <.01 |
| Transfused Postop | 24.5% | 35.0% | 44.2% | <.01 |
The rates of preoperative angiography increased with increasing preoperative LOS, (0 days: 64.3%, 1-2 days: 67.4%, 3-14 days: 70.5%; p<.01), as did having vein mapping (0 days: 63.6%, 1-2 days: 63.4%, 3-14 days: 72.2%; p<.01), and preoperative CT-A or MR-A (0 days: 34.4%, 1-2 days: 34.0%, 3-14 days: 37.8%; p<.01).
In regards to cardiac testing, greater rates of preoperative cardiac stress testing was seen with increasing preoperative LOS (0 days: 35.3%, 1-2 days: 31.6%, 3-14 days: 39.9%; p<.01) while a trend towards patients with 0 days preop LOS having a normal cardiac stress test without ischemia or myocardial infarction existed (0 days: 71.1% normal, 1-2 days: 69.0%, 3-14 days: 68.1%; p=.08).
A greater proportion of bypasses were performed for indications of tissue loss with increasing preop LOS (0 days Preop LOS: 52.4%, 1-2 days: 62.8%, 3-14 days: 76.0%; p<.01).
Intraoperatively there was no significant difference across preop LOS groups with regard to vein conduit use (0 days Preop LOS: 58.9%, 1-2 days: 55.9%, 3-14 days: 58.8%; p=.09), though patients with increased preoperative LOS were more likely to have increased total procedure length (0 days Preop LOS: 26.7% > 290 min, 1-2 days: 27.3%, 3-14 days: 32.1%, p<.01).
Additionally, patients with increased preoperative LOS were more likely to require postoperative blood transfusion (0 days Preop LOS: 24.5%, 1-2 days: 35.0%, 3-14 days: 44.2%; p<.01)(Table 1).
Multivariable logistic regression considering the above identified factors associated with increased preoperative LOS demonstrated that preoperative length of stay 3-14 days was associated with an increased rate of SSI (OR=1.36, 95% CI: 1.09-1.74; p=.01), however a preoperative length of stay of 1-2 days was not significant for greater postoperative SSI (adjusted odds ratio [OR]=1.26, 95% Confidence Interval: 0.93-1.71; p=.14).
Other significant operative and perioperative factors remaining significantly associated with SSI included prolonged procedure length >290 min (OR=1.67, 95% CI:1.37-2.03; p<.01), and need for postoperative transfusion of 1-2 units pRBC (OR=1.45, 95% CI: 1.14-1.85; p<.01), or 3+ units pRBC (OR=2.80, 95%CI: 2.21-3.56; p<.01).
CONCLUSIONS: Although preoperative comorbid status and operative complexity undoubtedly play a role, increasing preoperative length of stay is associated with an increase in the rate of surgical site infection following lower extremity bypass for critical limb ischemia. Preoperative length of stay represents a potentially modifiable variable for quality improvement in the future.
Back to 2020 Posters