- Email: [email protected]
Predictors of Overnight Admission After Laparoscopic Myomectomy in a High-Volume Minimally Invasive Gynecologic Surgery Setting
Original Article
Predictors of Overnight Admission After Laparoscopic Myomectomy in a High-Volume Minimally Invasive Gynecologic Surgery Setting Gaby Moawad, MD, Daniel Park, PhD, Marloes Maasen, MD, Anja Frost, MD, and Paul Tyan, MD From the Division of Minimally Invasive Gynecologic Surgery, Department of Obstetrics and Gynecology, George Washington University, Washington, DC (Drs. Moawad and Maasen), Department of Epidemiology and Biostatistics, Milken Institute School of Public Health, George Washington University, Washington, DC (Dr. Park), Department of Obstetrics and Gynecology, Johns Hopkins Hospital, Baltimore, Maryland (Dr. Frost), and Division of Minimally Invasive Gynecologic Surgery, Department of Obstetrics and Gynecology, University of North Carolina School of Medicine, Chapel Hill, North Carolina (Dr. Tyan).
ABSTRACT Study Objective: Our primary goal was to uncover preoperative and intraoperative risk factors that prevented same-day discharge (SDD) after myomectomy in a setting where SDD was the standard of care. Uncovered predictors would serve to enhance patient counseling and medical optimization before surgery. Design: Single-center retrospective cohort study. Setting: Urban university hospital center, by fellowship-trained minimally invasive gynecologic surgeons. Patients: A total of 315 consecutive patients undergoing minimally invasive myomectomy between March 2012 and May 2018. Intervention: Minimally invasive myomectomy. Measurements and Main Results: Preoperative and intraoperative characteristics were collected for analysis as predictors of SDD vs overnight admission. Comparisons of demographic, clinical, imaging, and operative characteristics were made using appropriate statistical methods for normally distributed, skewed, and categorical variables. Length of stay was analyzed as a categorical variable, SDD vs overnight admission. A logistic regression model was used to evaluate SDD vs any overnight stay. Out of the 315 patients undergoing laparoscopic myomectomy, 208 (66.03%) were discharged on the same day, and 107 (33.96%) were admitted for at least 1 day after surgery. The patients were more likely to be admitted overnight if they were of Asian ethnicity (p = .01), or if they had a lower preoperative Hct (36.15 vs 37.57; p < .003). An increase in any myoma characteristic metric was associated with overnight stay after surgery; these included mean myoma weight (512.0 g vs 310.1 g; p < .001), estimated size of the largest myoma on imaging (9.01 cm vs 7.77 cm; p < .001), and number of myomas removed (6.59 vs 5.57; p = .021). Other statistically significant differences between the overnight admission and SDD groups were mean estimated blood loss (599.4 mL vs 221.9 mL; p < .001), operative time (224.4 minutes vs 140.9 minutes; p < .001), and surgery end time (15:02 hours vs 12:43 hours; p < .001). Intraoperative complications associated with overnight admission were estimated blood loss >1 L (p < .001) and any intraoperative transfusion (p < .001). The adjusted logistic regression model identified an increase in operative time (by 60 minutes) and later surgery end time (by 60 minutes) as predictors of an overnight stay, whereas an 5% increase in preoperative hematocrit was associated with a 34% decrease in odds for an overnight stay. Conclusion: Perioperative factors, such as preoperative hematocrit, and myoma characteristics, as well as intraoperative factors, such as prolonged operative time and surgery end-time, are independent predictors of overnight hospital admission after minimally invasive myomectomy. Our present data can be used to provide better patient counseling before surgery. Journal of Minimally Invasive Gynecology (2019) 00, 1−5. © 2019 Published by Elsevier Inc. on behalf of AAGL. Keywords:
Same-day discharge; Minimally invasive myomectomy; Predictors; Safety
Dr. Moawad is a speaker for Intuitive Surgical. The other authors have nothing to disclose. Corresponding author: Paul Tyan, MD, 101 Manning Drive, Chapel Hill, NC 27415. E-mail: [email protected] Available at www.sciencedirect.com and www.jmig.org
Uterine myomas are the most common benign tumors of the female genital tract, with a lifetime risk of approximately 70% to 80% [1]. They can cause significant morbidity, including heavy menstrual bleeding, bulk symptoms, pelvic pain, subfertility, and complications related to pregnancy [2]. Surgery is the mainstay of treatment for women with large or
1553-4650/$ — see front matter © 2019 Published by Elsevier Inc. on behalf of AAGL. https://doi.org/10.1016/j.jmig.2019.03.022
2
symptomatic myomas refractory to medical or procedural interventions. Myomectomy, via a hysteroscopic or abdominal route, is the preferred uterine-sparing option among women desiring retention of fertility [3]. The benefits of a minimally invasive approach to myomectomy include less postoperative pain, reduced blood loss, fewer surgical complications, shorter hospitalization, decreased medical expenditure, and faster recovery time [3,4]. The safety and feasibility of a laparoscopic approach for large and complex myomectomies have been well established in the literature [5−7]. The field of minimally invasive myomectomy is undergoing radical changes, including advances in diagnostic and imaging studies, preoperative planning, energy sources, suturing technology, and surgical equipment specifically designed to improve the technical aspect of the procedure. Same-day discharge (SDD) after minimally invasive surgery is the focus of significant research effort in the field of gynecologic surgery; however, published data supporting the safety or optimization of that practice in the myomectomy setting are limited. Numerous articles published over the past decade investigating the safety of SDD after minimally invasive hysterectomy have identified various patient-specific, surgeon-specific, and surgery-specific variables affecting SDD. Those include patient demographics, preoperative hematocrit (Hct), surgeon caseload volume, hospital processes, surgery end time, and surgical findings. Specifically, for minimally invasive myomectomy, large myoma diameter, the presence of 3 or more myomas, nonuse of vasopressin, and preoperative anemia are known to be predictive of complications [7,8]. However, the current literature provides scant evidence regarding which factors are associated with an unplanned overnight admission. According to the most recent data, the mean length of hospital stay after minimally invasive myomectomy, varies profoundly from 7.5 hours to 2.9 days [4,9]. A recent study from a specialized center quoted SDD rates upward of 88% with a low readmission rate (0.6%) within 48 hours [6]. Experts in the field of minimally invasive gynecologic surgery advocate for the shift of laparoscopic myomectomy to the outpatient surgical setting for carefully selected patients with limited tumor burden [10]. Our primary goal in the present study was to combine preoperative and intraoperative variables from medical history and demographics, preoperative laboratory tests and imaging, and intraoperative findings in a single dataset to investigate predictive factors for an unplanned overnight admission. We designed this study as a single-site investigation in a highly specialized setting to minimize the considerable operator and staff-dependent confounders known to affect SDD rates. We also investigated the effect of SDD on postoperative complications and readmission rates. Materials and Methods Patients undergoing minimally invasive myomectomy by a minimally invasive gynecologic surgery fellowship-
Journal of Minimally Invasive Gynecology. Vol 00, No 00, 00 2019
trained surgeon between March 2012 and May 2018 were included for analysis. All consecutive cases within that time period were collected, and no omissions were allowed. Patients undergoing myectomy by fellowshiptrained staff at an ancillary location or patients undergoing myomectomy by non−fellowship-trained providers were excluded. In accordance with our division’s protocol, all patients scheduled for minimally invasive myomectomy, either via robot-assisted or a conventional straight-stick laparoscopic approach, were counseled on SDD at a dedicated preoperative visit. Fellows in minimally invasive gynecologic surgery and obstetrics and gynecology residents participated in all cases and performed their respective parts of the procedure under direct supervision of the attending surgeon. Starting in 2017, a standardized enhanced recovery protocol was instituted for all patients undergoing gynecologic surgery. Patients were thoroughly advised on the benefits and risks of SDD and were given clear expectations for the immediate recovery period and later at home. Institutional Review Board approval was obtained before retrospective chart analysis. Data extracted from the electronic medical records were categorized into 3 main groups. Preoperatively, we looked at age, race, body mass index, history of smoking, history of previous abdominal surgery, preoperative Hct level, and estimated size of the largest myoma on imaging. For intraoperative variables, we collected operative time, total number and weight of myomas removed, estimated blood loss (EBL), surgical modality (robot-assisted vs conventional laparoscopy), any organ injury, or blood loss >1 L. Postoperative variables collected were length of stay, reoperation, readmission within 1 month, intensive care unit admission, transfusion, surgical site infection or other incisional complications. Operative time was defined as the time from the first skin incision to the time of completed skin closure. For robotassisted cases, this included robot docking time but excluded time needed to assemble the robot which was done preoperatively. The analysis was designed to evaluate factors influencing overnight stay after surgery as predictors for same-day discharge. Those predictors focused on demographics, clinical history, imaging, procedural characteristics, and intraoperative complications. Comparisons of demographic and clinical characteristics were made using the Student or Satterthwaite t test for normally distributed continuous variables, the Wilcoxon test for skewed variables (EBL, myoma weight, number of myomas removed), and the Pearson x2 test or Fisher exact test for categorical variables, as appropriate. Length of stay was analyzed as a categorical variable, SDD vs overnight admission. Logistic regression was used to evaluate SDD vs any overnight stay. For all analyses, a p value < .05 was considered to indicate significance. Analyses were done using SAS version 9.4 (SAS Institute, Cary, NC).
Moawad et al.
Predictors of Overnight Admission After Laparoscopic
Results Out of the 315 patients who underwent laparoscopic myomectomy, 208 (66.03%) were discharged on the same day, and 107 (33.96%) were admitted for at least 1 day after surgery. Preoperative and demographic variables are presented in Table 1. The patients admitted overnight were more likely to have a lower preoperative Hct (36.1 vs 37.5; p < .003) and to be of Asian ethnicity. Smoking history and body mass index were not significantly correlated with overnight admission. Intraoperative variables of the 2 groups are compared in Table 2. An increase in any myoma
3
metric was associated with an overnight stay after surgery. Those included mean myoma weight (512.0 g vs 310.1 g; p < .001), estimated size of the largest myoma on imaging (9.0 cm vs 7.7 cm; p < .001), and number of myomas removed (6.5 vs. 5.5; p = .021). Other statistically significant differences between the overnight admission and SDD groups were EBL (599.4 mL vs 221.9 mL; p < .001), operative time (224.4 minutes vs 140.9 minutes; p < .001), and surgery end time (15:02 vs 12:43; p < .001). Intraoperative complications associated with an overnight admission were EBL >1 L (p < .001) and intraoperative transfusion
Table 1 Patient demographic and clinical characteristics Characteristic Age, yr, mean (SD) Body mass index, kg/m2, mean (SD) Race/ethnicity, n (%) Black or African American White or Caucasian Asian Other/unknown Preoperative hematocrit, %, mean (SD) History of smoking, n (%)
Same-day discharge (N = 208) 35.30 (5.21) 27.67 (6.06)
Overnight admission (N = 107) 36.28 (4.85) 29.03 (6.59)
136 (65.40) 49 (23.60) 6 (2.90) 17 (8.20) 37.57 (3.64) 25 (12.20)
84 (78.50) 9 (8.40) 5 (4.70) 9 (8.40) 36.15 (4.58) 15 (14.30)
p value* .106 .067 .011
.003 .603
Significant p values are in bold type. * p values for continuous variables are calculated using the t test, and those for categorical variables are compared using Pearson’s x2 test.
Table 2 Clinical, operative, and imaging predictors Predictor* Operative characteristics Modality: robot-assisted, n (%) Estimated blood loss, mL, mean (SD) Total surgery time, min, mean (SD) Surgery end time, mean (SD) Myoma characteristics Weight, g, mean (SD) Estimated size on imaging, cm, mean (SD) Number removed, mean (SD) Intraoperative complications, n (%) Organ injury Estimated blood loss >1000 mL Transfusion Conversion to laparotomy Surgical history, n (%) Caesarean section Previous myomectomy Adnexal surgery Appendectomy Nonspecified laparoscopy Nonspecified laparotomy
Same-Day Discharge (N = 208)
Overnight Admission (N = 107)
p value
160 (76.90) 221.9 (228.2) 140.90 (57.64) 12:43 (2:41)
89 (83.20) 599.4 (630.3) 224.45 (78.28) 15:02 (3:20)
.131 <.001 <.001 <.001
310.1 (374.9) 7.77 (3.88) 5.57 (5.66)
512.0 (405.3) 9.01 (3.76) 6.59 (5.60)
<.001 <.001 .021
0 (0) 3 (1.40) 2 (1.00) 0 (0)
1 (0.90) 17 (15.90) 7 (6.50) 1 (0.90)
.34 <.001 .008 .34
12 (5.80) 16 (7.70) 5 (2.40) 5 (2.40) 26 (12.50) 10 (4.80)
5 (4.70) 15 (14.00) 4 (3.70) 0 (0) 14 (13.10) 9 (9.40)
.683 .074 .495 .171 .883 .118
Significant p values are in bold type. * p values for continuous variables are calculated using the t test for normal variables and the Wilcoxon test for nonnormal variables (estimated blood loss, number of myomas, myoma weight, size on imaging). Categorical variables are compared using Pearson’s x2 test or Fisher’s exact test as appropriate (when n < 5).
4
Journal of Minimally Invasive Gynecology. Vol 00, No 00, 00 2019
Table 3 Postoperative Complications Postoperative complications Any postoperative complicationy Reoperation Intensive care unit admission Transfusion Readmission Incisional seroma/cellulitis/hematoma/separation/hernia
Same-Day Discharge (N = 208) 6 (2.90) 1 (0.50) 0 (0) 1 (0.50) 4 (1.90) 2 (1.00)
Overnight Admission (N = 107) 16 (17.12) 1 (0.90) 2 (1.90) 10 (9.40) 3 (2.80) 2 (1.90)
p value* <.001 1 .115 <.001 .694 .608
Significant p values are in bold type. * p values for categorical variables are compared using Pearson’s x2 test or Fisher’s exact test as appropriate (when n < 5). y Excluding minor incisional complications.
(p < .001). Surgical modality for the myomectomy (robotassisted vs conventional laparoscopy, p = .131) was not statistically significantly different between the 2 groups. When looking at postoperative complications, the only notable difference between the 2 groups was postoperative transfusion (Table 3). Patients who experienced an overnight admission were more likely to receive a blood transfusion after surgery (9.4% vs 0.5%; p < .001). There were no significant between-group differences in the rate of readmission, reoperation, intensive care unit admission, or incisional complications. Table 4 shows the multivariate logistic regression model for clinical and operative predictors of any overnight stay compared with SDD. Increased operative time and later surgery end time (by 60 minutes for each) were found to be independent predictors of overnight admission, whereas a 5% increase in preoperative Hct was associated with a 34% decrease in odds of an overnight stay. After adjustment, the robot-assisted surgical modality became significantly associated with an overnight admission. The model was adjusted
for all the other predictors included using a logistic regression model. We excluded the number of myomas from our regression model owing to the high correlation with operative time. Discussion Our findings suggest that several preoperative and intraoperative variables are associated with same-day discharge after minimally invasive myomectomy. We included most variables shown to influence SDD in a single regression model to provide a holistic, patient-centered approach for comparison. Lower preoperative Hct, longer operative time, and later surgery end time were all predictive of an overnight stay. Our findings are in line with several publications confirming that a later surgery end time, lower preoperative Hct, and longer operative times were predictive of inpatient admission for at least 1 day in the setting of minimally invasive hysterectomy [11]. Previous myomectomy studies suggest that myoma characteristics, such as number and diameter, are risk factors for complications after
Table 4 Logistic regression: Adjusted odds ratios for predictors of overnight stay following myomectomy Effect Age (5 yr) Race/ethnicity White or Caucasian Asian Black or African American Other/unknown Estimated blood loss (100 mL) Operative time (1 h) Surgery end time (1 h) Any intraoperative complications Preoperative hematocrit (5%) Previous myomectomy Nonspecified laparotomy Modality: robot-assisted Myoma weight (100 g) Size on imaging (5-cm increase) Significant p values are in bold type.
Estimate 0.907
95% confidence interval 0.647−1.272
p value .572
Reference 3.593 3.316 3.068 1.13 2.586 1.129 1.159 0.66 1.931 2.298 2.636 0.978 0.942
− 0.619−20.853 1.175−9.358 0.699−13.469 0.982−1.3 1.75−3.821 1.01−1.261 0.228−5.897 0.445−0.978 0.686−5.438 0.73−7.229 1.116−6.222 0.883−1.084 0.549−1.617
− .529 .325 .653 .088 <.0001 .033 .859 .039 .213 .155 .027 .675 .829
Moawad et al.
Predictors of Overnight Admission After Laparoscopic
surgery and thus might lead to an extended hospital stay [7,8,12,13]. When evaluating each myoma characteristic separately, we could not identify a statistically significant correlation between any imaging metric (size or number) and overnight admission; however, when stratifying by surgical modality, we found a significant independent correlation between robot-assisted surgery and overnight stay, which could possibly be explained by the greater complexity of surgeries and greater myoma tumor burden excision achieved via the robotic platform. Our results support the safety of same-day discharge for patients undergoing minimally invasive myomectomy. In fact, the patients admitted overnight had a higher rate of postoperative transfusion, with no significant differences in other postoperative complications. However, this postoperative finding is likely related to the greater intraoperative blood loss, which motivated the surgeon to admit those patients overnight. In this large retrospective cohort analysis of 315 cases over a 6-year span, 1 patient experienced an organ injury, 1 patient had conversion to laparotomy, and 2 patients underwent reoperation. Large population studies from national databases clearly document the cost savings associated with SDD after surgery in general and after minimally invasive gynecologic surgery in particular [14]. SDD also falls in line with the surgeon’s determination to provide the highest quality care and the patient’s desires to improve satisfaction. For all the aforementioned reasons, minimally invasive myomectomy has become standard practice in our division. Our study has several strengths, including the strict inclusion criteria that control for the significant confounders imposed by surgeon preference, surgeon case volume, support from the hospital and ancillary staff, and preoperative counseling. Another strength of our study is our comprehensive approach to collecting intraoperative and postoperative variables. This study also has some limitations. A retrospective study design that evaluates surgical outcomes has patient selection bias inherent in its design. In the study setting, all surgeons are fellowship-trained and have high surgical caseloads. Even though our results may generalize to a similar specialized environment, they are less likely to be generalizable to general gynecologic practice. Another weakness is the limited number of postoperative complications that compromise the study’s power. This finding stems from the low number of occurrences compared with the literature. Performing the same research using larger national databases will improve the power associated with postoperative findings but will jeopardize the setting in which this study was conducted and the number of variables used when building the regression model. Our study can aid the gynecologic surgeon in identifying variables at the preoperative visit that function as predictors
5
of overnight admission when SDD is planned. Patients with lower Hct, heavy bleeding, or multiple and large myomas can be optimized by preoperative iron intake and hormonal therapy in preparation for surgery. In addition, given the clear effect of operative time on SDD, these data should incentivize surgeons to operate more efficiently and minimize downtime during surgical cases. When planning the operative block, surgeons could benefit from operating on patients with multiple risk factors for failed SDD first, to mitigate the cumulative risk achieved through an earlier surgical end time. Our study also confirms that SDD after a minimally invasive myomectomy is feasible and does not jeopardize patient safety.
References 1. Baird DD, Dunson DB, Hill MC, Cousins D, Schectman JM. High cumulative incidence of uterine leiomyoma in black and white women: ultrasound evidence. Am J Obstet Gynecol. 2003;188:100–107. 2. Stewart EA. Uterine fibroids. Lancet. 2001;357:293–298. 3. Bhave Chittawar P, Franik S, Pouwer AW, Farquhar C. Minimally invasive surgical techniques versus open myomectomy for uterine fibroids. Cochrane Database Syst Rev. 2014;10:CD004638. 4. Hurst BS, Matthews ML, Marshburn PB. Laparoscopic myomectomy for symptomatic uterine myomas. Fertil Steril. 2005;83:1–23. 5. Vargas MV, Robinson JK, Schwab TD, Opoku-Anane J, Marfori CQ, Moawad GN. Feasibility and safety of a minimally invasive surgical (MIS) approach in complex myomectomies. J Minim Invasive Gynecol. 2015;22(6S):S38. 6. Alton K, Sullivan S, Udaltsova N, Yamamoto M, Zaritsky E. Sameday discharge after minimally invasive myomectomy. Obstet Gynecol. 2016;127:539–544. 7. Vargas MV, Moawad GN, Sievers C, et al. Feasibility, safety, and prediction of complications for minimally invasive myomectomy in women with large and numerous myomata. J Minim Invasive Gynecol. 2017;24:315–322. 8. Saridogan E. Surgical treatment of fibroids in heavy menstrual bleeding. Womens Health (Lond). 2016;12:53–62. 9. G€oc¸ men A, Sanl{kan F, U¸c ar MG. Comparison of robotic-assisted laparoscopic myomectomy outcomes with laparoscopic myomectomy. Arch Gynecol Obstet. 2013;287:91–96. 10. Gunnala V, Setton R, Pereira N, Huang JQ. Robot-assisted myomectomy for large uterine myomas: a single center experience. Minim Invasive Surg. 2016;2016:4905292. 11. Korsholm M, Mogensen O, Jeppesen MM, Lysdal VK, Traen K, Jensen PT. Systematic review of same-day discharge after minimally invasive hysterectomy. Int J Gynaecol Obstet. 2017;136:128–137. 12. Vargas MV, Schwab TD, Marfori JA, Moawad GN, Marfori CQ, Robinson JK. Predictors of hemorrhage and transfusion for women undergoing laparoscopic and robotic-assisted myomectomy. J Minim Invasive Gynecol. 2015;22(6S):S233. 13. Bean EM, Cutner A, Holland T, Vashisht A, Jurkovic D, Saridogan E. Laparoscopic myomectomy: a single-center retrospective review of 514 patients. J Minim Invasive Gynecol. 2017;24:485–493. 14. Schiavone MB, Herzog TJ, Ananth CV, et al. Feasibility and economic impact of same-day discharge for women who undergo laparoscopic hysterectomy. Am J Obstet Gynecol. 2012;207. 382.e1−e9.
Predictors of Overnight Admission After Laparoscopic Myomectomy in a High-Volume Minimally Invasive Gynecologic Surgery Setting Gaby Moawad, MD, Daniel Park, PhD, Marloes Maasen, MD, Anja Frost, MD, and Paul Tyan, MD From the Division of Minimally Invasive Gynecologic Surgery, Department of Obstetrics and Gynecology, George Washington University, Washington, DC (Drs. Moawad and Maasen), Department of Epidemiology and Biostatistics, Milken Institute School of Public Health, George Washington University, Washington, DC (Dr. Park), Department of Obstetrics and Gynecology, Johns Hopkins Hospital, Baltimore, Maryland (Dr. Frost), and Division of Minimally Invasive Gynecologic Surgery, Department of Obstetrics and Gynecology, University of North Carolina School of Medicine, Chapel Hill, North Carolina (Dr. Tyan).
ABSTRACT Study Objective: Our primary goal was to uncover preoperative and intraoperative risk factors that prevented same-day discharge (SDD) after myomectomy in a setting where SDD was the standard of care. Uncovered predictors would serve to enhance patient counseling and medical optimization before surgery. Design: Single-center retrospective cohort study. Setting: Urban university hospital center, by fellowship-trained minimally invasive gynecologic surgeons. Patients: A total of 315 consecutive patients undergoing minimally invasive myomectomy between March 2012 and May 2018. Intervention: Minimally invasive myomectomy. Measurements and Main Results: Preoperative and intraoperative characteristics were collected for analysis as predictors of SDD vs overnight admission. Comparisons of demographic, clinical, imaging, and operative characteristics were made using appropriate statistical methods for normally distributed, skewed, and categorical variables. Length of stay was analyzed as a categorical variable, SDD vs overnight admission. A logistic regression model was used to evaluate SDD vs any overnight stay. Out of the 315 patients undergoing laparoscopic myomectomy, 208 (66.03%) were discharged on the same day, and 107 (33.96%) were admitted for at least 1 day after surgery. The patients were more likely to be admitted overnight if they were of Asian ethnicity (p = .01), or if they had a lower preoperative Hct (36.15 vs 37.57; p < .003). An increase in any myoma characteristic metric was associated with overnight stay after surgery; these included mean myoma weight (512.0 g vs 310.1 g; p < .001), estimated size of the largest myoma on imaging (9.01 cm vs 7.77 cm; p < .001), and number of myomas removed (6.59 vs 5.57; p = .021). Other statistically significant differences between the overnight admission and SDD groups were mean estimated blood loss (599.4 mL vs 221.9 mL; p < .001), operative time (224.4 minutes vs 140.9 minutes; p < .001), and surgery end time (15:02 hours vs 12:43 hours; p < .001). Intraoperative complications associated with overnight admission were estimated blood loss >1 L (p < .001) and any intraoperative transfusion (p < .001). The adjusted logistic regression model identified an increase in operative time (by 60 minutes) and later surgery end time (by 60 minutes) as predictors of an overnight stay, whereas an 5% increase in preoperative hematocrit was associated with a 34% decrease in odds for an overnight stay. Conclusion: Perioperative factors, such as preoperative hematocrit, and myoma characteristics, as well as intraoperative factors, such as prolonged operative time and surgery end-time, are independent predictors of overnight hospital admission after minimally invasive myomectomy. Our present data can be used to provide better patient counseling before surgery. Journal of Minimally Invasive Gynecology (2019) 00, 1−5. © 2019 Published by Elsevier Inc. on behalf of AAGL. Keywords:
Same-day discharge; Minimally invasive myomectomy; Predictors; Safety
Dr. Moawad is a speaker for Intuitive Surgical. The other authors have nothing to disclose. Corresponding author: Paul Tyan, MD, 101 Manning Drive, Chapel Hill, NC 27415. E-mail: [email protected] Available at www.sciencedirect.com and www.jmig.org
Uterine myomas are the most common benign tumors of the female genital tract, with a lifetime risk of approximately 70% to 80% [1]. They can cause significant morbidity, including heavy menstrual bleeding, bulk symptoms, pelvic pain, subfertility, and complications related to pregnancy [2]. Surgery is the mainstay of treatment for women with large or
1553-4650/$ — see front matter © 2019 Published by Elsevier Inc. on behalf of AAGL. https://doi.org/10.1016/j.jmig.2019.03.022
2
symptomatic myomas refractory to medical or procedural interventions. Myomectomy, via a hysteroscopic or abdominal route, is the preferred uterine-sparing option among women desiring retention of fertility [3]. The benefits of a minimally invasive approach to myomectomy include less postoperative pain, reduced blood loss, fewer surgical complications, shorter hospitalization, decreased medical expenditure, and faster recovery time [3,4]. The safety and feasibility of a laparoscopic approach for large and complex myomectomies have been well established in the literature [5−7]. The field of minimally invasive myomectomy is undergoing radical changes, including advances in diagnostic and imaging studies, preoperative planning, energy sources, suturing technology, and surgical equipment specifically designed to improve the technical aspect of the procedure. Same-day discharge (SDD) after minimally invasive surgery is the focus of significant research effort in the field of gynecologic surgery; however, published data supporting the safety or optimization of that practice in the myomectomy setting are limited. Numerous articles published over the past decade investigating the safety of SDD after minimally invasive hysterectomy have identified various patient-specific, surgeon-specific, and surgery-specific variables affecting SDD. Those include patient demographics, preoperative hematocrit (Hct), surgeon caseload volume, hospital processes, surgery end time, and surgical findings. Specifically, for minimally invasive myomectomy, large myoma diameter, the presence of 3 or more myomas, nonuse of vasopressin, and preoperative anemia are known to be predictive of complications [7,8]. However, the current literature provides scant evidence regarding which factors are associated with an unplanned overnight admission. According to the most recent data, the mean length of hospital stay after minimally invasive myomectomy, varies profoundly from 7.5 hours to 2.9 days [4,9]. A recent study from a specialized center quoted SDD rates upward of 88% with a low readmission rate (0.6%) within 48 hours [6]. Experts in the field of minimally invasive gynecologic surgery advocate for the shift of laparoscopic myomectomy to the outpatient surgical setting for carefully selected patients with limited tumor burden [10]. Our primary goal in the present study was to combine preoperative and intraoperative variables from medical history and demographics, preoperative laboratory tests and imaging, and intraoperative findings in a single dataset to investigate predictive factors for an unplanned overnight admission. We designed this study as a single-site investigation in a highly specialized setting to minimize the considerable operator and staff-dependent confounders known to affect SDD rates. We also investigated the effect of SDD on postoperative complications and readmission rates. Materials and Methods Patients undergoing minimally invasive myomectomy by a minimally invasive gynecologic surgery fellowship-
Journal of Minimally Invasive Gynecology. Vol 00, No 00, 00 2019
trained surgeon between March 2012 and May 2018 were included for analysis. All consecutive cases within that time period were collected, and no omissions were allowed. Patients undergoing myectomy by fellowshiptrained staff at an ancillary location or patients undergoing myomectomy by non−fellowship-trained providers were excluded. In accordance with our division’s protocol, all patients scheduled for minimally invasive myomectomy, either via robot-assisted or a conventional straight-stick laparoscopic approach, were counseled on SDD at a dedicated preoperative visit. Fellows in minimally invasive gynecologic surgery and obstetrics and gynecology residents participated in all cases and performed their respective parts of the procedure under direct supervision of the attending surgeon. Starting in 2017, a standardized enhanced recovery protocol was instituted for all patients undergoing gynecologic surgery. Patients were thoroughly advised on the benefits and risks of SDD and were given clear expectations for the immediate recovery period and later at home. Institutional Review Board approval was obtained before retrospective chart analysis. Data extracted from the electronic medical records were categorized into 3 main groups. Preoperatively, we looked at age, race, body mass index, history of smoking, history of previous abdominal surgery, preoperative Hct level, and estimated size of the largest myoma on imaging. For intraoperative variables, we collected operative time, total number and weight of myomas removed, estimated blood loss (EBL), surgical modality (robot-assisted vs conventional laparoscopy), any organ injury, or blood loss >1 L. Postoperative variables collected were length of stay, reoperation, readmission within 1 month, intensive care unit admission, transfusion, surgical site infection or other incisional complications. Operative time was defined as the time from the first skin incision to the time of completed skin closure. For robotassisted cases, this included robot docking time but excluded time needed to assemble the robot which was done preoperatively. The analysis was designed to evaluate factors influencing overnight stay after surgery as predictors for same-day discharge. Those predictors focused on demographics, clinical history, imaging, procedural characteristics, and intraoperative complications. Comparisons of demographic and clinical characteristics were made using the Student or Satterthwaite t test for normally distributed continuous variables, the Wilcoxon test for skewed variables (EBL, myoma weight, number of myomas removed), and the Pearson x2 test or Fisher exact test for categorical variables, as appropriate. Length of stay was analyzed as a categorical variable, SDD vs overnight admission. Logistic regression was used to evaluate SDD vs any overnight stay. For all analyses, a p value < .05 was considered to indicate significance. Analyses were done using SAS version 9.4 (SAS Institute, Cary, NC).
Moawad et al.
Predictors of Overnight Admission After Laparoscopic
Results Out of the 315 patients who underwent laparoscopic myomectomy, 208 (66.03%) were discharged on the same day, and 107 (33.96%) were admitted for at least 1 day after surgery. Preoperative and demographic variables are presented in Table 1. The patients admitted overnight were more likely to have a lower preoperative Hct (36.1 vs 37.5; p < .003) and to be of Asian ethnicity. Smoking history and body mass index were not significantly correlated with overnight admission. Intraoperative variables of the 2 groups are compared in Table 2. An increase in any myoma
3
metric was associated with an overnight stay after surgery. Those included mean myoma weight (512.0 g vs 310.1 g; p < .001), estimated size of the largest myoma on imaging (9.0 cm vs 7.7 cm; p < .001), and number of myomas removed (6.5 vs. 5.5; p = .021). Other statistically significant differences between the overnight admission and SDD groups were EBL (599.4 mL vs 221.9 mL; p < .001), operative time (224.4 minutes vs 140.9 minutes; p < .001), and surgery end time (15:02 vs 12:43; p < .001). Intraoperative complications associated with an overnight admission were EBL >1 L (p < .001) and intraoperative transfusion
Table 1 Patient demographic and clinical characteristics Characteristic Age, yr, mean (SD) Body mass index, kg/m2, mean (SD) Race/ethnicity, n (%) Black or African American White or Caucasian Asian Other/unknown Preoperative hematocrit, %, mean (SD) History of smoking, n (%)
Same-day discharge (N = 208) 35.30 (5.21) 27.67 (6.06)
Overnight admission (N = 107) 36.28 (4.85) 29.03 (6.59)
136 (65.40) 49 (23.60) 6 (2.90) 17 (8.20) 37.57 (3.64) 25 (12.20)
84 (78.50) 9 (8.40) 5 (4.70) 9 (8.40) 36.15 (4.58) 15 (14.30)
p value* .106 .067 .011
.003 .603
Significant p values are in bold type. * p values for continuous variables are calculated using the t test, and those for categorical variables are compared using Pearson’s x2 test.
Table 2 Clinical, operative, and imaging predictors Predictor* Operative characteristics Modality: robot-assisted, n (%) Estimated blood loss, mL, mean (SD) Total surgery time, min, mean (SD) Surgery end time, mean (SD) Myoma characteristics Weight, g, mean (SD) Estimated size on imaging, cm, mean (SD) Number removed, mean (SD) Intraoperative complications, n (%) Organ injury Estimated blood loss >1000 mL Transfusion Conversion to laparotomy Surgical history, n (%) Caesarean section Previous myomectomy Adnexal surgery Appendectomy Nonspecified laparoscopy Nonspecified laparotomy
Same-Day Discharge (N = 208)
Overnight Admission (N = 107)
p value
160 (76.90) 221.9 (228.2) 140.90 (57.64) 12:43 (2:41)
89 (83.20) 599.4 (630.3) 224.45 (78.28) 15:02 (3:20)
.131 <.001 <.001 <.001
310.1 (374.9) 7.77 (3.88) 5.57 (5.66)
512.0 (405.3) 9.01 (3.76) 6.59 (5.60)
<.001 <.001 .021
0 (0) 3 (1.40) 2 (1.00) 0 (0)
1 (0.90) 17 (15.90) 7 (6.50) 1 (0.90)
.34 <.001 .008 .34
12 (5.80) 16 (7.70) 5 (2.40) 5 (2.40) 26 (12.50) 10 (4.80)
5 (4.70) 15 (14.00) 4 (3.70) 0 (0) 14 (13.10) 9 (9.40)
.683 .074 .495 .171 .883 .118
Significant p values are in bold type. * p values for continuous variables are calculated using the t test for normal variables and the Wilcoxon test for nonnormal variables (estimated blood loss, number of myomas, myoma weight, size on imaging). Categorical variables are compared using Pearson’s x2 test or Fisher’s exact test as appropriate (when n < 5).
4
Journal of Minimally Invasive Gynecology. Vol 00, No 00, 00 2019
Table 3 Postoperative Complications Postoperative complications Any postoperative complicationy Reoperation Intensive care unit admission Transfusion Readmission Incisional seroma/cellulitis/hematoma/separation/hernia
Same-Day Discharge (N = 208) 6 (2.90) 1 (0.50) 0 (0) 1 (0.50) 4 (1.90) 2 (1.00)
Overnight Admission (N = 107) 16 (17.12) 1 (0.90) 2 (1.90) 10 (9.40) 3 (2.80) 2 (1.90)
p value* <.001 1 .115 <.001 .694 .608
Significant p values are in bold type. * p values for categorical variables are compared using Pearson’s x2 test or Fisher’s exact test as appropriate (when n < 5). y Excluding minor incisional complications.
(p < .001). Surgical modality for the myomectomy (robotassisted vs conventional laparoscopy, p = .131) was not statistically significantly different between the 2 groups. When looking at postoperative complications, the only notable difference between the 2 groups was postoperative transfusion (Table 3). Patients who experienced an overnight admission were more likely to receive a blood transfusion after surgery (9.4% vs 0.5%; p < .001). There were no significant between-group differences in the rate of readmission, reoperation, intensive care unit admission, or incisional complications. Table 4 shows the multivariate logistic regression model for clinical and operative predictors of any overnight stay compared with SDD. Increased operative time and later surgery end time (by 60 minutes for each) were found to be independent predictors of overnight admission, whereas a 5% increase in preoperative Hct was associated with a 34% decrease in odds of an overnight stay. After adjustment, the robot-assisted surgical modality became significantly associated with an overnight admission. The model was adjusted
for all the other predictors included using a logistic regression model. We excluded the number of myomas from our regression model owing to the high correlation with operative time. Discussion Our findings suggest that several preoperative and intraoperative variables are associated with same-day discharge after minimally invasive myomectomy. We included most variables shown to influence SDD in a single regression model to provide a holistic, patient-centered approach for comparison. Lower preoperative Hct, longer operative time, and later surgery end time were all predictive of an overnight stay. Our findings are in line with several publications confirming that a later surgery end time, lower preoperative Hct, and longer operative times were predictive of inpatient admission for at least 1 day in the setting of minimally invasive hysterectomy [11]. Previous myomectomy studies suggest that myoma characteristics, such as number and diameter, are risk factors for complications after
Table 4 Logistic regression: Adjusted odds ratios for predictors of overnight stay following myomectomy Effect Age (5 yr) Race/ethnicity White or Caucasian Asian Black or African American Other/unknown Estimated blood loss (100 mL) Operative time (1 h) Surgery end time (1 h) Any intraoperative complications Preoperative hematocrit (5%) Previous myomectomy Nonspecified laparotomy Modality: robot-assisted Myoma weight (100 g) Size on imaging (5-cm increase) Significant p values are in bold type.
Estimate 0.907
95% confidence interval 0.647−1.272
p value .572
Reference 3.593 3.316 3.068 1.13 2.586 1.129 1.159 0.66 1.931 2.298 2.636 0.978 0.942
− 0.619−20.853 1.175−9.358 0.699−13.469 0.982−1.3 1.75−3.821 1.01−1.261 0.228−5.897 0.445−0.978 0.686−5.438 0.73−7.229 1.116−6.222 0.883−1.084 0.549−1.617
− .529 .325 .653 .088 <.0001 .033 .859 .039 .213 .155 .027 .675 .829
Moawad et al.
Predictors of Overnight Admission After Laparoscopic
surgery and thus might lead to an extended hospital stay [7,8,12,13]. When evaluating each myoma characteristic separately, we could not identify a statistically significant correlation between any imaging metric (size or number) and overnight admission; however, when stratifying by surgical modality, we found a significant independent correlation between robot-assisted surgery and overnight stay, which could possibly be explained by the greater complexity of surgeries and greater myoma tumor burden excision achieved via the robotic platform. Our results support the safety of same-day discharge for patients undergoing minimally invasive myomectomy. In fact, the patients admitted overnight had a higher rate of postoperative transfusion, with no significant differences in other postoperative complications. However, this postoperative finding is likely related to the greater intraoperative blood loss, which motivated the surgeon to admit those patients overnight. In this large retrospective cohort analysis of 315 cases over a 6-year span, 1 patient experienced an organ injury, 1 patient had conversion to laparotomy, and 2 patients underwent reoperation. Large population studies from national databases clearly document the cost savings associated with SDD after surgery in general and after minimally invasive gynecologic surgery in particular [14]. SDD also falls in line with the surgeon’s determination to provide the highest quality care and the patient’s desires to improve satisfaction. For all the aforementioned reasons, minimally invasive myomectomy has become standard practice in our division. Our study has several strengths, including the strict inclusion criteria that control for the significant confounders imposed by surgeon preference, surgeon case volume, support from the hospital and ancillary staff, and preoperative counseling. Another strength of our study is our comprehensive approach to collecting intraoperative and postoperative variables. This study also has some limitations. A retrospective study design that evaluates surgical outcomes has patient selection bias inherent in its design. In the study setting, all surgeons are fellowship-trained and have high surgical caseloads. Even though our results may generalize to a similar specialized environment, they are less likely to be generalizable to general gynecologic practice. Another weakness is the limited number of postoperative complications that compromise the study’s power. This finding stems from the low number of occurrences compared with the literature. Performing the same research using larger national databases will improve the power associated with postoperative findings but will jeopardize the setting in which this study was conducted and the number of variables used when building the regression model. Our study can aid the gynecologic surgeon in identifying variables at the preoperative visit that function as predictors
5
of overnight admission when SDD is planned. Patients with lower Hct, heavy bleeding, or multiple and large myomas can be optimized by preoperative iron intake and hormonal therapy in preparation for surgery. In addition, given the clear effect of operative time on SDD, these data should incentivize surgeons to operate more efficiently and minimize downtime during surgical cases. When planning the operative block, surgeons could benefit from operating on patients with multiple risk factors for failed SDD first, to mitigate the cumulative risk achieved through an earlier surgical end time. Our study also confirms that SDD after a minimally invasive myomectomy is feasible and does not jeopardize patient safety.
References 1. Baird DD, Dunson DB, Hill MC, Cousins D, Schectman JM. High cumulative incidence of uterine leiomyoma in black and white women: ultrasound evidence. Am J Obstet Gynecol. 2003;188:100–107. 2. Stewart EA. Uterine fibroids. Lancet. 2001;357:293–298. 3. Bhave Chittawar P, Franik S, Pouwer AW, Farquhar C. Minimally invasive surgical techniques versus open myomectomy for uterine fibroids. Cochrane Database Syst Rev. 2014;10:CD004638. 4. Hurst BS, Matthews ML, Marshburn PB. Laparoscopic myomectomy for symptomatic uterine myomas. Fertil Steril. 2005;83:1–23. 5. Vargas MV, Robinson JK, Schwab TD, Opoku-Anane J, Marfori CQ, Moawad GN. Feasibility and safety of a minimally invasive surgical (MIS) approach in complex myomectomies. J Minim Invasive Gynecol. 2015;22(6S):S38. 6. Alton K, Sullivan S, Udaltsova N, Yamamoto M, Zaritsky E. Sameday discharge after minimally invasive myomectomy. Obstet Gynecol. 2016;127:539–544. 7. Vargas MV, Moawad GN, Sievers C, et al. Feasibility, safety, and prediction of complications for minimally invasive myomectomy in women with large and numerous myomata. J Minim Invasive Gynecol. 2017;24:315–322. 8. Saridogan E. Surgical treatment of fibroids in heavy menstrual bleeding. Womens Health (Lond). 2016;12:53–62. 9. G€oc¸ men A, Sanl{kan F, U¸c ar MG. Comparison of robotic-assisted laparoscopic myomectomy outcomes with laparoscopic myomectomy. Arch Gynecol Obstet. 2013;287:91–96. 10. Gunnala V, Setton R, Pereira N, Huang JQ. Robot-assisted myomectomy for large uterine myomas: a single center experience. Minim Invasive Surg. 2016;2016:4905292. 11. Korsholm M, Mogensen O, Jeppesen MM, Lysdal VK, Traen K, Jensen PT. Systematic review of same-day discharge after minimally invasive hysterectomy. Int J Gynaecol Obstet. 2017;136:128–137. 12. Vargas MV, Schwab TD, Marfori JA, Moawad GN, Marfori CQ, Robinson JK. Predictors of hemorrhage and transfusion for women undergoing laparoscopic and robotic-assisted myomectomy. J Minim Invasive Gynecol. 2015;22(6S):S233. 13. Bean EM, Cutner A, Holland T, Vashisht A, Jurkovic D, Saridogan E. Laparoscopic myomectomy: a single-center retrospective review of 514 patients. J Minim Invasive Gynecol. 2017;24:485–493. 14. Schiavone MB, Herzog TJ, Ananth CV, et al. Feasibility and economic impact of same-day discharge for women who undergo laparoscopic hysterectomy. Am J Obstet Gynecol. 2012;207. 382.e1−e9.