Infectious Causes of 30-Day Unplanned Hospital Encounters and Readmissions After Hysterectomies: A Single Institutional Study

Introduction

Unplanned hospital readmissions represent a common and costly contributor to growing healthcare expenditures in the United States. ¹ Estimated costs associated with unplanned readmissions are > $17 billion annually, with ∼20% of Medicare beneficiaries requiring readmission within 30 days of hospital discharge. ¹ Increasingly, unplanned readmission rates are being used as a proxy for hospitals' quality of care, particularly in relation to surgical procedures.^1,2 In gynecologic surgery, previous studies have identified the unplanned readmission rate after hysterectomy to be between 2.8% and 3.8% for benign disease, and ∼6.1% following hysterectomy for uterine cancer.^3–5 Despite the growing interest in factors that affect readmission rates following certain medical conditions and surgical procedures, only a handful of studies have focused on the causes of readmissions in gynecologic surgery.

Across all surgical specialties, infectious etiologies have been identified as a significant a driver for increased complication and 30-day readmission rates. ⁴ In fact, infection rates following hysterectomy have been reported to be between 10.4% and 28.8%.^4,6 Furthermore, in a retrospective study by Catanzarite et al., the researchers showed that infectious complications were a leading cause 30-day readmissions following benign hysterectomy. ⁷ However, there remains a paucity of data identifying causes of postoperative infections that lead to unplanned 30-day readmission after elective hysterectomy.

The aim of this study was to investigate the causes of posthysterectomy infections 30 days after hospital discharge.

Materials and Methods

Queries were made to 3510 consecutive adult women (≥ 18 years old) undergoing hysterectomy at a major academic institution from 2010 to 2015. Of those, 453 (12.9%) patients were identified who had hospital encounters or readmissions 30 days after discharge, and the cases of 193 (42.6%) of these patients were reviewed retrospectively. Institutional review board approval was obtained prior to study initiation. The primary outcome investigated in this study was 30-day postdischarge infection rates and complications.

Baseline characteristics and demographic information collected for each patient included age, race, marital status, employment status, and body mass index (BMI). Comorbidities included depression, congestive heart failure, coronary artery disease, atrial fibrillation (A-fib), peripheral vascular disease, hypertension (HTN), diabetes, hyperlipidemia (HLD), anemia, history of myocardial infarction (MI), deep-vein thrombosis (DVT), and pulmonary embolism (PE). Other preoperative variables collected included smoking status and alcohol use.

Intraoperative variables included operating room (OR) time (including anesthesia time before hysterectomy, operative time of hysterectomy, and anesthesia time after hysterectomy), estimated blood loss (EBL), number of packed red blood cell (PRBC) transfusions, urine output (UOP), drain placement, and antibiotics administered. Type of procedure was collected, which included total abdominal hysterectomy (TAH), total vaginal hysterectomy (TVH), total laparoscopic hysterectomy (TLH), or robotically assisted total laparoscopic hysterectomy (RA-TLH). Uterine weight and pathology (benign versus malignant) were recorded for each patient. All patients with malignant pathology were elective procedures.

Postoperative complications recorded included unplanned intensive care unit (ICU) transfer, delirium, urinary-tract infection (UTI), fever, ileus, small-bowel obstruction (SBO), superficial surgical-site infection (SSI), deep SSI, HTN, hypotension, hematoma, MI, DVT, PE, anemia, weakness, sensory deficits, and urinary retention.

Thirty-day postdischarge infections evaluated included superficial SSI, intra-abdominal/pelvic abscess, UTI, wound dehiscence, pyelonephritis/acute kidney injury (AKI), cervicitis, and viral gastroenteritis.

Results

Discussion

In the current retrospective study of 193 adult women who underwent elective hysterectomy, it was found that the primary infectious causes of 30-day unplanned hospital encounters were superficial SSI, abscess formation, and UTI. Furthermore, approximately two-thirds of the patients who sought postdischarge care due to infections required hospital readmission.

Infection has previously been identified as a contributor to inferior outcomes following hysterectomy, with a few studies quantifying the prevalence. In a retrospective study of 64,918 abdominal hysterectomies performed at 270 California hospitals, Yokoe et al. showed that 1481 (2.3%) of patients received a diagnosis code for SSI within 30 days posthysterectomy. ⁸ Similarly, in another retrospective study 28,366 hysterectomy patients from the National Surgical Quality Improvement Program (NSQIP) database, Mahdi et al. found that 3% of patients developed an SSI. ⁹ Analogous to the aforementioned studies, the current study showed SSIs as a major infectious cause of 30-day unplanned hospital encounters.

Gynecologic surgery entails unique risk for infectious complications due to the fact that both vaginal and skin flora have the potential to cause infection. In a prospective study of 161 women undergoing abdominal hysterectomy, Soper et al. found that patients with bacterial vaginitis or trichomonas on preoperative evaluation were more likely than controls to experience cuff cellulitis, cuff abscess, or both. ¹⁰ Apart from this additional consideration, significant patient risk factors for SSI in hysterectomy appeared to resemble those of other surgical procedures, including obesity, lack of private insurance, chronic lung disease, and fluid/electrolyte imbalance.^11,12 In a study of 1369 women with endometrial carcinoma who underwent hysterectomy, Bakkum-Gamez et al. showed that independent predictors of superficial SSI were obesity, anemia (HCT <36%), American Society of Anesthesiology (ASA) score >2, smoking, laparotomy, and intraoperative transfusion. ¹³ Furthermore, these researchers found that the predictors of organ space SSI were older age, smoking, vascular disease, prior methicillin-resistant staphylococcal infection, greater EBL, and lymphadenectomy or bowel resection. ¹³ Further studies are necessary to risk-stratify patients with increased potential of experiencing unexpected SSI after elective hysterectomy.

There have only been a few studies that identified infectious etiology as a significant driver of unplanned hospital 30-day readmissions after hysterectomy. Roy et al. found that risk of 30-day readmission was three times greater among patients with SSI after hysterectomy, ranging from 16.0% to 23.5%, compared to 4.5% to 6.2% of patients without SSI. ¹⁴ Similarly, Catanzarite and colleagues reported that the most common complications leading to 30-day readmissions were infectious in nature, namely SSI (28.4%), sepsis/septic shock (12.8%), and UTI (9.7%). ⁷ Of note, this trend appeared to hold true regardless of surgical approach, with abdominal, laparoscopic, and vaginal hysterectomies all showing predominantly infectious complications among patients requiring 30-day readmission. ⁵ Analogously, the current study found that approximately two-thirds of the patients who sought postdischarge care due to infections required hospital readmission.

Several studies have investigated optimal antibiotic prophylaxis to reduce infection rates. In a retrospective review of 94 patients who underwent robotic radical hysterectomy, Avila et al. demonstrated that a solitary dose of vaginal metronidazole significantly reduced the risk of pelvic and genitourinary infection. ¹⁵ It is also necessary to weigh the potential benefits against the added expense and adverse effects inherent in providing additional prophylactic treatment. Among 310 patients who underwent laparoscopy-assisted vaginal hysterectomy, Chang et al. observed that a single dose of cefazolin, compared to multiple doses, administered during the perioperative period was similarly effective in preventing pelvic cellulitis or abscess, vaginal cuff abscess, wound infection and UTI both during the sentinel hospitalization and 30-days afterward. ¹⁶ Despite these encouraging findings, a 2017 Cochrane systematic review of antibiotic prophylaxis for elective hysterectomy found insufficient evidence to show which individual antibiotic (if any), dose regimen, or route of administration was most effective for preventing SSI. ¹⁷ Additional work is necessary to identify optimal prophylactic regimens for patients undergoing elective hysterectomies with the hope of reducing unplanned postoperative infections.

Elective hysterectomy procedures are a significant contributor to healthcare costs, mostly due to the high prevalence of operations performed annually. More than 600,000 of these operations are performed per year in the United States and approximately one-third of women have had the procedure by age 60. ¹⁵ Moreover, the related annual hospital expenditures are estimated to be >5 billion dollars, with infectious complications adding thousands of dollars to each encounter.^13,15 Therefore, there is a growing emphasis among the Centers for Medicare and Medicaid Services on providing potential modifiable interventions to reduce SSIs after elective hysterectomies.^8,15 The financial implications of postoperative infections are significant and warrant additional consideration as the field investigates ways to improve quality of care and reduce overall costs.

This study had several limitations that could potentially have implications regarding its interpretation. A relatively small patient sample size from only one academic center was used, making broad conclusions difficult and potentially biasing the results for particular patient population or treatment paradigms. Although all variables were recorded pre-, peri-, and postoperatively, they were reviewed retrospectively and, as such, were limited by the weaknesses inherent to retrospective analyses. Given that the aim of the study was to characterize patients who had subsequent infections, the differences with noninfected patients were not compared, which could have implications regarding the findings. Finally, this study did not control for benign versus malignant pathology, and this also could have implications regarding the results. Despite these limitations, the study suggests that the primary causes of 30-day hospital encounters and readmissions following hysterectomy are overwhelmingly infectious, underscoring the need for further studies to identify interventions and preventative measures for reducing posthysterectomy infection.

Conclusions

The current study suggests that the primary infectious causes of 30-day unplanned hospital encounters following hysterectomy are superficial SSI, abscess formation, and UTI. Furthermore, patients who seek postdischarge care due to infections are more likely to require readmission. Additional studies are necessary to corroborate these findings and to identify preventative measures for reducing posthysterectomy infections.