Influence of Patient and Intraoperative Characteristics on Length of Hospital Stay After Robotic Hysterectomy

Abstract

Objective: The aim of this research was to determine patient and intraoperative characteristics that may predict feasibility of same-day discharge after robotic hysterectomy. Materials and Methods: Patients undergoing robotic hysterectomy from October 2007 to April 2011 were included in the study. Patient demographics were retrospectively collected and analyzed, include: age (years), height (inches), weight (kg), body mass index (BMI, measured in kg/m²), procedure indication, number of previous abdominal procedures, and American Society of Anesthesiologists (ASA) physical classification system score. Procedure data were collected as well as transfusion status, postoperative length of hospital stay (LOS), procedure type, presence of abdominal adhesive disease, conversion to laparotomy, estimated blood loss (EBL), procedure duration, intraoperative complications, and short-term complications. Results: Patient age (p<0.0001), malignant diagnosis (p<0.001), transfusion (p=0.001), EBL (p<0.001), intraoperative complications (p=0.038), ASA score (p<0.001), procedure performed (p<0.0001), and adhesive disease (p=0.024) were significantly associated with LOS. Case time, surgeon, BMI, and number of previous abdominal surgeries were not associated with LOS. Conclusions: Patient and intraoperative characteristics can be used to predict the likelihood of successful same-day discharge after robotic-assisted hysterectomy. Same-day discharge after robotic-assisted laparoscopic hysterectomy should be regarded as an essential component of cost containment. (J GYNECOL SURG 30:96)

Introduction

Minimally invasive approaches to hysterectomy have many advantages, including reduced blood loss, reduced postoperative hospital stay, and shorter recovery time. Because of the increasing trend toward earlier postoperative discharge, several studies have examined the feasibility of same-day discharge after laparoscopic hysterectomy.^1–7 These studies demonstrated an acceptable safety profile and readmission rates as well as high patient satisfaction in appropriately selected populations.^1–7 Selection criteria in these populations are variable, with one study requiring only “good general health” and a social network¹ and another study using specific criteria, including surgical indication, patient health status, and patient social network.³

Despite ample data demonstrating the safety and feasibility of early discharge after laparoscopic hysterectomy, no studies have examined the role of patient and surgical characteristics to predict which patients have a greater likelihood of successful same-day discharge. Within the last 2 years at the Crozer-Chester Medical Center, in Upland, PA, patients are frequently discharged on the same day after undergoing robotic-assisted hysterectomy. Therefore, this study was conducted to determine what patient and operative characteristics were associated with postoperative length of hospital stay (LOS) prior to this change in discharge timing. In undertaking this project, the aim was to develop a better understanding of patient characteristics that may predict ability for same-day discharge after robotic hysterectomy and to improve preoperative patient counseling regarding discharge expectations.

Materials and Methods

The Crozer Keystone Health System's institutional review board approved this study and it was in compliance with the rules set forth by the Declaration of Helsinki.⁸ A retrospective chart review of all patients undergoing robotic-assisted hysterectomy for benign and malignant indications was conducted. Hospital electronic health records (EHR), physician office EHR, and physician office paper charts were reviewed. EHR and paper charts were both reviewed in this study because, physicians at Crozer-Chester Medical Center were transitioning from a paper-based system to an electronic-based record system during the study period.

Patients undergoing robotic-assisted hysterectomy from October 2007 to April 2011 were included in the study. This review started at the establishment of a robotics program at the Crozer-Chester Medical Center. All surgeons who were participating in the program began their robotic experience at the program's initiation. Five surgeons performed all procedures; 2 gynecologic oncologists and 3 general gynecologists comprised this group. All procedures were performed on the da Vinci Standard Surgical System (Intuitive Surgical, Sunnyvale, CA).

Data collection

Patient demographics were collected including: age (years); height (inches); weight (kg); body mass index (BMI, measured in kg/m²), procedure indication, number of previous abdominal procedures, and American Society of Anesthesiologists (ASA) physical classification system score (ASA). Procedure and postoperative data were collected as well as transfusion status, LOS, procedure type, presence of abdominal adhesive disease, conversion to laparotomy, estimated blood loss (EBL, in mL), procedure duration (in minutes), intraoperative complications, and short-term complications. Conversion to laparotomy was defined as laparotomy performed prior to completion of hysterectomy. Laparotomy only performed to deliver a large specimen was not included. Intraoperative complications included bladder, uretral, intestinal, or vascular injury, or other intraoperative injury. Short-term complications were limited to the first 30 postoperative days and included wound complications, bacteremia or sepsis, abscess, hematoma, venous thromboembolism, myocardial infarction, cardiopulmonary arrest, acute renal failure, respiratory failure, cerebrovascular accident, shock, or need to return to the operating room.

Inclusion criteria were women undergoing robotic-assisted hysterectomy for any indication during the study period. Exclusion criteria were hysterectomy not initiated robotically or missing patient data.

Statistical analysis

Descriptive statistics, including medians and ranges, were computed for measured continuous variables, and marginal frequency distributions were summarized for categorical variables. Generalized linear models with Poisson regression were used to predict the logarithm of average length of postoperative hospital stay with one predictor at a time. All analyses were performed using SAS/STAT software, version 9.3 (Cary, NC). A level of p<0.05 was selected as the threshold for statistically significant differences.

Results

Study population

During the study period, 238 patients underwent robotic-assisted hysterectomy. Of these patients, 155 (65%) procedures were performed for benign indications. The median age of the study population was 52 years (range, 32–91). The median BMI of the population was 30 (range, 18–72). One (0.4%) patient was underweight (BMI<18.5), 55 (23.1%) patients were of normal weight (BMI 18.5–24.9), 56 (23.5%) patients were overweight (BMI 25–29.9), 84 (35.3%) patients were obese (BMI 30–39.9), and 42 (17.6%) patients were morbidly obese (BMI≥40). The median ASA score was 2 (range, 1–4). Nine patients had an ASA score of 1 (3%), 140 patients had an ASA score of 2 (59%), 80 patients had an ASA score of 3 (34%), and 9 patients had an ASA score of 4 (3%). One hundred and forty-six (61%) patients had undergone previous abdominal surgery; the median number of previous abdominal surgeries was 1 (maximum, 6). See Tables 1 and 2 for additional patient characteristics and procedures performed.

Table 1.

Characteristics of the Study Population

Characteristic	Median	Range
Age (years)	52	32–91
Weight (kg)	80	40–190
BMI (m/kg²)	30	18–72
LOS (days)	1	0–19
EBL (mL)	100	Minimal to 1000
Operative time (minutes)	226	104–540
ASA score^a	2	1–4

	Number	%
Adhesive disease	108	45
Previous abdominal surgery	146	61

ASA physical classification system score

BMI, body mass index; LOS, length of stay; EBL, estimated blood loss; ASA, American Society of Anesthesiologists.

Table 2.

Procedures Performed

Procedure	Number (%)
Total or subtotal robotic hysterectomy	26 (10.9%)
Total robotic hysterectomy with BSO and/or adnexal surgery	119 (50%)
Total robotic hysterectomy with BSO and cancer staging	61 (25.6%)
Total or subtotal robotic-assisted hysterectomy with or without adnexal surgery or cancer staging with pelvic-floor repair	23 (9.7%)
Robotic radical hysterectomy with BSO and staging	1 (0.4)
Total abdominal hysterectomy	8 (3.4%)

BSO, bilateral salpingo-oophorectomy.

Operative characteristics

Ten (4.2%) procedures were converted to laparotomy. Indications for laparotomy included: extensive adhesive disease; suspected bowel injury; gross metastatic cancer; incarcerated umbilical hernia not reducible laparoscopically; inflammatory changes in the pelvis; distorted pelvic anatomy with limited uterine mobility; a large pelvic mass; and limited visualization (in 2 patients). Of the cases in which procedures were converted to laparotomy, 7 patients had a malignant diagnosis. One patient underwent a Pfannenstiel incision to deliver a large uterus, but this was not counted as a laparotomy as indicated above.

Median case time was 226 minutes (range, 104–540 minutes). Six (2.5%) patients underwent intraoperative or postoperative transfusion of packed red blood cells (PRBCs), and 108 (45.3%) patients had abdominal adhesive disease. Twenty-six (10.9%) patients experienced a short-term complication. Thirteen (5.4%) patients experienced an intraoperative complication. Median EBL was 100 mL (range, minimal to 1000 mL).

Median LOS was 1 day (range, 0–19 days), 4 patients went home on the day of the procedure. The patient with a LOS of 19 days developed ventilator-dependent respiratory failure and was not extubated until postoperative day 14.

Postoperative LOS

Patient age was significantly associated with LOS (p<0.0001). On average, LOS increases 1.8% for each 1-year increase in patient age. On average, the expected LOS is 220% longer in patients ages 80 years and older; in this study, there were 13 patients in this category. See Table 3.

Table 3.

Patient and Intraoperative Characteristics Related to LOS

Predictor	β (95% CI)	p-Value
Age (years)	0.018 (0.010–0.026)	<0.0001
Case time (minutes)	0.057 (−0.013 to 0.127)	0.114
Malignant diagnosis	0.374 (0.159 to 0.588)	0.0006
Transfusion	0.827 (0.364 to 1.289)	0.0018
Intraoperative complications	0.479 (0.0981 to 0.8592)	0.021
EBL (mL)	0.0026 (0.0015 to 0.0038)	<0.0001
Procedure performed		<0.0001
Surgeon		0.329
Abdominal adhesive disease	0.090 (−0.1384 to 0.3186)	0.024
BMI (kg/m²)	0.0070 (−0.0034 to 0.0173)	0.272
Previous abdominal surgery	0.037 (−0.0410 to 0.1147)	0.358
ASA score^a	0.375 (0.212 to 0.537)	<0.0001

ASA physical classification system score.

LOS, length of stay; CI, confidence interval; EBL, estimated blood loss; BMI, body mass index; ASA, American Society of Anesthesiologists.

Malignant diagnosis was significantly associated with LOS (p<0.001). On average, the length of stay was 45% longer for patients having a malignant diagnosis, compared to patients with benign disease. In this study, there were 83 patients with a malignant diagnosis.

Transfusion was significantly associated with LOS (p=0.001). On average, the length of stay was 128% longer in patients who received transfusion of PRBCs intraoperatively or postoperatively. EBL was significantly associated with LOS (p<0.001). On average, the length of stay is 14% longer for an additional 50-mL blood loss.

Intraoperative complications were significantly associated with LOS (p=0.038). On average, the LOS is ∼134% longer in patients with an intraoperative complication.

ASA score was significantly associated with LOS (p<0.001). On average, LOS is 45% longer for each 1-unit increase in ASA score. There was no significant difference between patients with an ASA score 1 versus an ASA score of 2, but differences were observed for a score of 2 versus a score of 3, a score of 2 versus a score of 4, and a score of 3 versus a score of 4. Eighty-nine patients had an ASA score >2.

Procedure performed is also significantly associated with LOS (p<0.0001). As expected, the procedure of total abdominal hysterectomy is associated with a longer LOS than all the other procedures, with the exceptions of robotic-assisted radical hysterectomy with bilateral salpingo-oopherectomy (BSO), and systematic pelvic and paraaortic lymph node dissection (only 1 patient was in this category). There was no statistically significant difference in length of stay among the other procedures.

Adhesive disease was significantly associated with LOS (p=0.024). On average, the LOS was 27% longer for patients with adhesive disease.

Case time, surgeon, BMI, and number of previous abdominal surgeries were not associated with LOS.

Discussion

Use of robotic-assisted laparoscopic hysterectomy has grown significantly in recent years. In 2013 Wright et al. published the largest study to date comparing robotic-assisted hysterectomy with abdominal, vaginal, and conventional laparoscopic surgery in more than 200,000 patients.⁹ This analysis revealed an increase in robotic-assisted hysterectomy rates from 0.5% in 2007 to 9.5% in 2010. Hospitalization for <2 days was more common in the robotic-assisted hysterectomy group; otherwise, there were no differences in outcomes between robotic-assisted and traditional laparoscopic hysterectomies. This analysis also highlighted an increased cost of robotic hysterectomy. However, with increased surgeon experience, total cost per case showed an overall downward trend. Other studies have also documented the increased cost of robotic assistance.¹⁰

Data from both prospective and retrospective studies demonstrate that early discharge after laparoscopic hysterectomy is safe and feasible without compromising patient outcome or patient satisfaction.^1–7 In 2011 Perron-Burdick et al. demonstrated safety in a retrospective case series of 527 patients who underwent laparoscopic hysterectomy followed by same-day discharge. Readmission rates were 0.6% at 48 hours, 3.6% at 3 months, and 4% at 12 months.⁴ Another retrospective project of 359 laparoscopic hysterectomies, demonstrated no differences in complication or readmission rates for patients discharged on the day of surgery versus those who were discharged later.⁵ In a prospective study of 23 patients undergoing laparoscopic hysterectomy, 88% were discharged on the day of surgery, with a median rating of 9/10 regarding patient satisfaction.¹ In a retrospective study of 66 patients, 91% were discharged successfully on the same day as surgery, with 95% patient satisfaction.²

Because of the increased cost of robotic assistance and the exponential growth of the robotic-assisted approach, it is imperative to make robotic surgery more cost-effective. Marketing strategies and patient perception of superiority of robotic assistance ensures that this technology will continue to be utilized despite cost disadvantages, at least for the short-term. The current authors believe that several strategies must be used to produce cost containment. As a result of trends toward decreased cost with increasing surgeon experience, it may be most appropriate for robotic surgery to be concentrated in referral centers with robotic experts and high case volumes per surgeon. Pasic et al. showed increased cost when robotic hysterectomy was performed as an inpatient procedure.¹⁰ Therefore, early discharge is an important component of cost containment. Unfortunately, cost data related to LOS in the current study's population is not currently available. Patient safety and satisfaction are well-established for same-day discharge.

This current project defined patient characteristics that can be used to predict successful same-day discharge and to tailor preoperative counseling related to patients' discharge expectations. At the Crozer-Chester Medical Center, discharge counseling preoperatively appears to largely influence patients' expectations and acceptance of discharge timing. These results indicate that procedure planned, patient BMI, and patient history of previous abdominal procedures should not limit willingness to counsel patients for same-day discharge after robotic-assisted hysterectomy. In addition, the current authors note that caution should be used when considering counseling patients about early discharge in those with increasing age. The current authors also suggest that malignant procedure indication and an ASA score >2 should create some caution with respect to preoperative counseling about same-day discharge.

When these counseling guidelines regarding age, ASA score, and malignant procedure indication were applied to the study population, a total of 185 patients were included. When overlap among categories was considered, a total of 121 patients were represented. Based on overlap of these higher-risk categories, the current authors believe that these factors may work in concert to prolong LOS after robotic-assisted hysterectomy. While the current authors do not advocate for firm guidelines to counsel patients against same-day discharge, if any of these factors are present, it is recommended that these patient characteristics be taken into account during the preoperative discussion, especially when several factors are present.

The results also indicate intraoperative findings that may influence a change in preoperative discharge planning. Specifically, increased EBL, intraoperative complications, and presence of abdominal adhesive disease point toward a longer LOS. In contrast, case time should not be used to alter previous discharge instructions.

As this analysis showed no relationship between surgeon and LOS, these counseling guidelines can be applied universally. As each of the surgeons in this analysis began their robotic experience at the same time, one could expect a similar level of skill based on number of cases performed. While the current authors cannot predict similarities between the studied patient population and other population, these results were specific to the patients and not the surgeons in this study.

There are several strengths of this project. The study involved a diverse population, including patients with both benign and malignant surgical indications. To the current authors' knowledge, this study offers the only published data regarding predictors for LOS after robotic-assisted hysterectomy. The project included 5 surgeons and, therefore, was representative of the entire group of robotic surgeons; thus the results were generalizable to the entire department of the institution. In addition, these data were collected from the initiation of a robotics program at the institution and provided a unique opportunity for comparison to a subsequent patient group to determine the influence of physician learning curves on outcomes.

This project was limited by its retrospective nature. While patient data were collected from hospital and physician records, it cannot be guaranteed that all postoperative complications resulted in presentation to the Crozer-Chester Medical Center. While these results show that BMI was not related to LOS, the majority of the studied patient population have abnormal BMIs. Therefore, this study may not have had a diverse enough population, with respect to BMI to detect its influence on LOS.

Conclusions

Findings from this project can be used to determine the likelihood of successful same-day discharge and can facilitate patient counseling regarding discharge. Successful same-day discharge after robotic-assisted laparoscopic hysterectomy should be regarded as an essential component of cost containment.

Footnotes

Disclosure Statement

No competing financial conflicts exist.

References

Lassen

, Moeller-Sarsen

, DE Nully

. Same-day discharge after laparoscopic hysterectomy. Acta Obstet Gynecol Scand, 2012; 11:1339.

Thiel

, Gamelin

. Outpatient total laparoscopic hysterectomy. J Am Assoc Gynecol Laparosc, 2003; 4:481.

de Lapasse

, Rabischong

, Bolandard

, et al. Total laparoscopic hysterectomy and early discharge: Satisfaction and feasibility study. J Am Assoc Gynecol Laparosc, 2008; 15:20.

Perron-Burdick

, Yamamoto

, Zaritsky

. Same-day discharge after laparoscopic hysterectomy. Obstet Gynecol, 2008; 117:1136.

Hoffman

, Kennedy

, Borschel

, Burchette

, Kidd

. Laparoscopic hysterectomy: The Kaiser Permanente San Diego experience. J Minim Invasive Gynecol, 2005; 12:16.

Taylor

. Outpatient laparoscopic hysterectomy with discharge in 4 to 6 hours. J Am Assoc Gynecol Laparosc, 1994; 1(suppl):S35.

Chou

, Rosen

, Cario

, et al. Home within 24 hours of laparoscopc hysterectomy. Aust N Z J Obstet Gynecol, 1999; 39:234.

World Medical Association. World Medical Association Declaration of Helsinki: Ethical Principles for Medical Research Involving Human Subjects. JAMA, 2013; 310:2191.

Wright

, Ananth

, Lewin

, et al. Robotically assisted vs. laparoscopic hysterectomy among women with benign gynecologic disease. JAMA, 2013; 309:689.

10.

Pasic

, Rizzo

, Fang

, Ross

, Moore

, Gunnarsson

. Comparing robot-assisted with conventional laparoscopic hysterectomy: Impact on cost and clinical outcomes. J Minim Invasive Gynecol, 2010; 17:730.