Original Article

Relevance of Uterine Weight for Predicting Surgical Complications in

Minimally Invasive Benign Hysterectomy
Steven Mouro, DO, Jenna L. Carter Hamed, BS, James L. Whiteside, MD, MA, MHA, and
Dmitry Tumin, PhD
From the Department of Obstetrics and Gynecology (Drs. Mouro and Whiteside), East Carolina University, Greenville, North Carolina, Department of
Pediatrics and Department of Academic Affairs (Dr. Tumin), East Carolina University, Greenville, North Carolina, and Brody School of Medicine (Ms.
Carter Hamed), East Carolina University, Greenville, North Carolina

ABSTRACT Study Objectives: To describe the uterine weight threshold for increasing risk of complications after a laparoscopic hyster-
ectomy using the American College of Surgeons National Surgical Quality Improvement Program (NSQIP) database.
Design: Cross-sectional analysis using the American College of Surgeons NSQIP database from 2016 to 2021.
Setting: American College of Surgeons NSQIP database.
Patients: Patients undergoing minimally invasive hysterectomy for benign indications (N = 64 289).
Interventions: None.
Measurements and Main Results: Uterine weight was entered in grams and 30-day complications were abstracted from
patient charts. In the analytic sample, median uterine weight was 135 grams (interquartile range, 90−215) and 6% of
patients (n = 4085) experienced complications. Uterine weight performed very poorly in predicting complications on bivari-
ate analysis (area under the receiver operating characteristics curve, 0.53; 95% confidence interval, 0.53−0.54). On multi-
variable analysis, a uterine weight cutoff of 163 grams was associated with higher odds of complications (odds ratio, 1.11;
95% confidence interval, 1.03−1.19; p = .003), but this threshold achieved only a 43% sensitivity and 62% specificity for
predicting complications.
Conclusions: Uterine weight alone possessed negligible utility for predicting the risk of perioperative complications in min-
imally invasive hysterectomy. Journal of Minimally Invasive Gynecology (2023) 00, 1−7. © 2023 AAGL. All rights
reserved.
Keywords: Vaginal hysterectomy; Robotic-assisted hysterectomy; Laparoscopic hysterectomy; Uterine weight; Laparoscopy; Postoper-
ative complication; Minimally invasive hysterectomy; NSQIP

Uterine size has been cited as a barrier to minimally inva- hysterectomies for uteri up to a 300-gram threshold [3]. Other
sive hysterectomy. However, recent case studies [1,2] have studies have reported improved outcomes with laparoscopic
reported successful minimally invasive hysterectomies with hysterectomy regardless of uterine weight, with proper
uteri as large as 11 kg. Notwithstanding these reports, some surgical experience and expertise [5−8].
studies document an association of greater peri- and postoper- In previous research on the relationship between uterine
ative complications with increasing uterine size [3,4], yet other weight and hysterectomy complications, uterine weights
studies find no difference between vaginal and laparoscopic from 100 grams to 500 grams have been identified as
thresholds associated with higher complications rates
[1,3,9]. In contrast, Louie et al [4] found no distinct thresh-
Dmitry Tumin discloses quality improvement and research funding project old weight but indicated that complication risk linearly
funding unrelated to this submission from Kate B. Reynolds Charitable
Trust and Lilly and Co., Inc. The other authors declare that they have no
increased with weights >150 grams for vaginal hysterec-
conflict of interest. tomy and >300 grams for laparoscopic hysterectomy.
Corresponding author: Steven Mouro, DO, Department of Obstetrics and Current Procedural Terminology (CPT)-based reim-
Gynecology, East Carolina University, 600 Moye Blvd, Greenville, North bursement recognizes a 250-gram threshold for higher pay-
Carolina 27834. ment, but the ambiguity of the evidence base for any
E-mail: [email protected]
specific threshold offers an opportunity for further study of
Submitted May 31, 2023, Revised August 9, 2023, Accepted for publication the significance of uterine weight in minimally invasive
August 10, 2023. hysterectomy. Although surgical expertise, patient safety,
Available at www.sciencedirect.com and www.jmig.org

1553-4650/$ — see front matter © 2023 AAGL. All rights reserved.

https://doi.org/10.1016/j.jmig.2023.08.005

Downloaded for Anonymous User (n/a) at Hualien Tzu Chi Medical Center from ClinicalKey.com by Elsevier on October 26,
2023. For personal use only. No other uses without permission. Copyright ©2023. Elsevier Inc. All rights reserved.
2 Journal of Minimally Invasive Gynecology. Vol 00, No 00, 00 2023

and cost considerations also contribute to preoperative procedure. This composite outcome included one or more
counseling, the need to assess the independent utility of of 17 postoperative complications captured in the NSQIP:
uterine weight presents an opportunity for further evalua- bleeding/transfusion, superficial surgical site infection
tion. Identification of an evidence-based uterine weight (SSI), deep incisional SSI, organ space SSI, wound disrup-
threshold could also inform risk stratification and thus pre- tion, sepsis, septic shock, pneumonia, unplanned intubation,
operative counseling. In this study, we used a large multi- ventilator dependence >48 hours after surgery, urinary tract
institutional registry to analyze the independent association infection, deep vein thrombosis/thrombophlebitis requiring
of uterine weight with the risk of perioperative complica- treatment, pulmonary embolism, cardiac arrest requiring
tions after minimally invasive hysterectomy for benign cardiopulmonary resuscitation, myocardial infarction, pro-
indications and to estimate the uterine weight cutoff most gressive renal insufficiency, and stroke [10]. Unplanned
predictive of increased complications. reoperation was queried separately as a secondary outcome.
The primary exposure was uterine weight, entered in
grams by the NSQIP data abstractor at each participating
Materials and Methods
center. Uterine weights recorded as zero were regarded as
This study was a secondary analysis of the American missing. To limit the influence of potentially implausible
College of Surgeons (ACS) National Surgical Quality outliers, nonzero weights <30 grams were rounded up to 30
Improvement Program (NSQIP) Participant Use Data Files grams (approximately the 1st percentile), and weights
(PUFs) from 2016 to 2021. Patient-level data contained in >1 kg were rounded down to 1 kg (approximately the 99th
the PUFs are completely deidentified, and therefore, this percentile) to account for potential data entry error. Patient
study did not constitute human subjects research requiring characteristics extracted from the NSQIP database included
local institutional review board approval. More than 600 the year of procedure, patient age in years, race and ethnic-
teaching and community hospitals across the United States ity (Hispanic or Latino, non-Hispanic Black, non-Hispanic
participate in data collection for the NSQIP. At each institu- White, or any other race and ethnicity not mentioned
tion, clinically trained reviewers abstract information on above), body mass index (classified as underweight, <18.5;
patient demographics, diagnoses, procedures, and out- normal weight, 18.5 to <25; overweight, 25 to <30; or
comes, including intraoperative and 30-day postoperative obese, ≥30), parity (0, 1, 2, 3+), American Society of Anes-
complications. Because NSQIP data are deidentified, this thesiologists classification (I, II, or III+), comorbidities
study does not involve human subjects and was deemed (diabetes, current smoking, hypertension treated with medi-
exempt from review by the local institutional review board. cation, bleeding disorders, pelvic inflammatory disease),
For this study, we used CPT codes to identify patients previous abdominal surgery, and previous pelvic surgery
who underwent any of the following elective procedures: [11] Surgical characteristics included approach (TVH,
Total vaginal hysterectomy (TVH) (CPT codes 58260, LAVH, TLH) and operative time (minutes).
58262, 58263, 58267, 58270, 58290, 58291, 58293, Categorical variables were summarized using counts and
58294), laparoscopically assisted vaginal hysterectomy percentages, whereas continuous variables were summa-
(LAVH) (CPT codes 58550, 58552, 58553, 58554), or total rized as medians with interquartile range. Receiver operat-
laparoscopic hysterectomy (TLH) with or without robotic ing characteristics (ROC) curve analysis [12] was used to
assistance (CPT codes 58570, 58571, 58572, 58573). Infor- characterize the utility of uterine weight for predicting the
mation on uterine weight was provided as part of the tar- composite primary study outcome, with area under the
geted hysterectomy PUF, so we limited the sample to ROC curve (AUROC) used as the measure of predictive
patients included in the targeted PUF. We then excluded utility (range, 0.5−1.0, with higher values indicating better
patients with a principal diagnosis of cancer based on Inter- predictive utility). The optimal uterine weight threshold for
national Classification of Diseases, Tenth Revision, codes predicting this outcome was defined based on Youden’s J
C51−C58, C796, C7982, D070, or D39, and any patients index, maximizing the sum of sensitivity and specificity.
whose hysterectomy was identified as being performed for The ROC curve analysis was repeated in subsamples
an indication of malignancy based on the targeted hysterec- defined by surgical approach, as well as tertiles of the over-
tomy PUF. Patients who underwent supracervical hysterec- all sample stratified by predicted risk of complications. Pre-
tomy (CPT codes 58541, 58542, 58543, 58544) were dicted risk was calculated using multivariable logistic
excluded to avoid unaccounted for variance in gross speci- regression, controlling for all covariates except for uterine
men weight given lack of cervix. We also excluded cases weight, and complication risk was classified as low (bottom
with concomitant surgery other than lysis of adhesions tertile), moderate (middle tertile), or high (top tertile) based
(CPT codes 44180, 58660, 44005)), ureterolysis (CPT on the linear prediction from this model. In the entire sam-
codes 50715, 50722, 50725), or cystoscopy (CPT code ple, we evaluated the odds ratio (OR) of any complications
52000). Finally, we excluded patients with missing data on and unplanned reoperation associated with the optimal uter-
uterine weight, study outcomes, or study covariates. ine weight cutoff using multivariable logistic regression,
The primary study outcome was experiencing any surgi- controlling for all covariates in the study. Data analysis was
cal complications within 30 days of the hysterectomy completed using Stata/SE 16.1 (College Station, TX:

Downloaded for Anonymous User (n/a) at Hualien Tzu Chi Medical Center from ClinicalKey.com by Elsevier on October 26,
2023. For personal use only. No other uses without permission. Copyright ©2023. Elsevier Inc. All rights reserved.
Mouro et al. Uterine Weight for Predicting Surgical Complications 3

StataCorp, LP), and p <.05 was considered statistically sig- Table 1

nificant.
Descriptive statistics of study variables (N = 64 289)

Variable n (%) Median (IQR)

Results
Uterine weight (g) 135 (90−215)
The 2016−2021 targeted PUF included 139 143 patients Any complication 4085 (6)
undergoing elective minimally invasive hysterectomy. We Unplanned reoperation 736 (1)
Procedure yr 2018 (2016−2021)
limited the sample to 109 916 benign cases and excluded
Age (yrs) 45 (39−51)
34 521 cases involving concurrent ineligible procedures. We Race and ethnicity
then excluded a further 2678 cases missing data on uterine Hispanic or Latino 6530 (10)
weight and 8428 cases missing data on study covariates. In Non-Hispanic Black 10 071 (15)
the final sample (Fig. 1) of 64 289 patients, the median uter- Non-Hispanic White 43 075 (67)
None of the above 4613 (7)
ine weight was 135 grams (interquartile range, 90−215), 6%
BMI
of patients (n = 4085) experienced complications, and 1% Underweight 445 (0.69)
(n = 736) required unplanned reoperation. Other study varia- Normal weight 12 042 (20)
bles are summarized for the overall sample in Table 1. Overweight 16 942 (26)
The ROC curve analysis for the entire sample is illus- Obese 33 860 (52)
Parity
trated in Fig. 2. Uterine weight performed very poorly in
0 12 581 (19)
predicting the risk of any complications (AUROC, 0.53, 1 10 079 (15)
95% confidence interval [CI], 0.53−0.54). The best cutoff 2 21 225 (33)
3+ 20 404 (31)
ASA status
I 5744 (8)
Fig. 1 II 42 484 (66)
Flowchart of study subject selection with exclusions. III+ 16 061 (24)
Diabetes 5208 (8)
Current smoker 9964 (15)
Hypertension 15 965 (24)
Bleeding disorder 482 (0.75)
Pelvic inflammatory disease 1458 (2.2)
Previous abdominal surgery 18 169 (28)
Previous pelvic surgery 38 613 (60)
Approach
TLH 47 296 (73)
LAVH 9309 (14)
TVH 7684 (11)
Operative time (min) 114 (85−153)

ASA = American Society of Anesthesiologists; BMI = body mass index;

IQR = interquartile range; LAVH = laparoscopically assisted vaginal hysterec-
tomy; TLH = total laparoscopic hysterectomy; TVH = total vaginal hysterec-
tomy.

identified was 163 grams, accounting for 38% of the sample

and achieving only 43% sensitivity and 62% specificity for
predicting the composite outcome of any complications.
When stratifying the sample by approach, predictive utility
of uterine weight was consistently poor across all 3
approaches (AUROC ranging from 0.53 to 0.56) and the
optimal cutoff for uterine weight ranged from 113 to 191 g
(Table 4). Similarly, after calculating complication risk
based on other study covariates (model shown in Table 5),
AUROC ranged from 0.50 to 0.52, and the optimal cutoff
for uterine weight ranged from 86 to 208 in each risk tercile
(Table 6).
In the entire sample, multivariable analyses demon-
strated that a uterine weight cutoff of 163 grams was inde-
pendently associated with higher odds of the composite

Downloaded for Anonymous User (n/a) at Hualien Tzu Chi Medical Center from ClinicalKey.com by Elsevier on October 26,
2023. For personal use only. No other uses without permission. Copyright ©2023. Elsevier Inc. All rights reserved.
4 Journal of Minimally Invasive Gynecology. Vol 00, No 00, 00 2023

Fig. 2 Table 2
Area under the receiver operating characteristics curve, predicting risk
Multivariable logistic regression of any perioperative complication
of any postoperative complications based on uterine weight.
(N = 64 289)

Variable OR 95% CI p value

Uterine weight >163 g 1.11 1.03−1.19 .003
Procedure yr 1.07 1.05−1.09 <.001
Age (yrs) 0.98 0.98−0.98 <.001
Race and ethnicity
Hispanic or Latino 1.20 1.09−1.33 <.001
Non-Hispanic Black 1.17 1.07−1.28 <.001
Non-Hispanic White Ref.
None of the above 1.03 0.91−1.18 .603
BMI
Underweight 1.18 0.81−1.73 .386
Normal weight Ref.
Overweight 0.95 0.86−1.04 .269
Obese 0.96 0.87−1.05 .379
Parity
0 Ref.
measure of any complications (OR, 1.11; 95% CI, 1.03
1 1.14 1.01−1.27 .029
−1.19; p = .003) (Table 2), but not with the secondary out- 2 1.13 1.02−1.25 .015
come of unplanned readmissions (OR, 1.04; 95% CI, 0.88 3+ 1.29 1.17−1.42 <.001
−1.22; p = .658) (Table 3). Perioperative complications ASA status
were also more common for more recent procedures, proce- I Ref.
II 1.25 1.09−1.42 .001
dures involving younger patients, and procedures involving
III+ 1.63 1.41−1.89 <.001
Black or Hispanic patients (compared with non-Hispanic Diabetes 1.12 1.00−1.26 .053
White). Higher parity, higher American Society of Anes- Current smoker 1.18 1.08−1.28 <.001
thesiologists status, smoking, comorbid bleeding disorder, Hypertension 0.99 0.92−1.08 .899
and history of abdominal or pelvic surgery were also associ- Bleeding disorder 1.64 1.23−2.21 .001
Pelvic inflammatory disease 0.96 0.77−1.19 .695
ated with increased complication risk. Longer surgery dura-
Previous abdominal surgery 1.18 1.10−1.26 <.001
tion and use of LAVH or TVH approaches (compared with Previous pelvic surgery 1.09 1.02−1.17 .014
TLH) were associated with increased odds of the composite Approach
outcome of perioperative complications. Considering the TLH Ref.
secondary outcome, younger age, comorbid hypertension or LAVH 1.16 1.06−1.27 .001
TVH 1.49 1.36−1.65 <.001
bleeding disorder, previous abdominal surgery, longer oper-
Operative time (£ 10 min) 1.04 1.04−1.05 <.001
ative time, and LAVH or TVH approaches (compared with
TLH) were also associated with higher odds of unplanned ASA = American Society of Anesthesiologists; BMI = body mass index;
CI = confidence interval; LAVH = laparoscopically assisted vaginal hysterec-
reoperation.
tomy; OR = odds ratio; TLH = total laparoscopic hysterectomy; TVH = total
vaginal hysterectomy.

Discussion
This study investigated the correlation between uterine
weight and rates of intra- and postoperative complications of identified in ours or other studies, longer operative times
benign hysterectomy. Using a large multicenter registry, we and greater blood loss do occur more frequently among
found that uterine weight possessed negligible utility for pre- patients undergoing a hysterectomy with an enlarged
dicting the risk of perioperative complications for minimally uterus [9,13,17,18]. Studies seeking to identify a predic-
invasive benign hysterectomy. Minimally invasive hysterec- tive uterine threshold weight note that greater uterine
tomy for enlarged uteri can be performed with a low risk of weight may be associated with statistically significant
complications and it does not seem to have any basis in sur- increases in the risk of complications, but these pre-
gical complexity as determined by rate of perioperative com- dicted differences in the risk of complications do not
plications [13]. Higher reimbursement in these settings could seem to be large enough to be clinically relevant. Uter-
be related to increased operative time given that many studies ine weight alone does not account for overall surgical
have demonstrated that uterine weight greater than 250 grams complexity, given that patient anatomy, uterine shape,
has been consistently shown to increase operative time for and surgeon expertise (a function of skill, volume, and
laparoscopic cases [7,14−16]. the surgical team) also influence surgical outcomes.
Although no uterine weight threshold that would reli- Overall, it is unlikely that uterine weight alone is an
ably predict greater risk of complications has been adequate proxy for surgical difficulty.

Downloaded for Anonymous User (n/a) at Hualien Tzu Chi Medical Center from ClinicalKey.com by Elsevier on October 26,
2023. For personal use only. No other uses without permission. Copyright ©2023. Elsevier Inc. All rights reserved.
Mouro et al. Uterine Weight for Predicting Surgical Complications 5

Table 3 Table 5
Multivariable logistic regression of unplanned reoperation Multivariable logistic regression predicting risk of any complica-
(N = 64 289) tions, independent of uterine weight (N = 64 289)

Variable OR 95% CI p value Variable OR 95% CI p value

Uterine weight >163 g 1.04 0.88−1.22 .658 Procedure yr 1.07 1.05−1.09 <.001
Procedure yr 1.02 0.97−1.06 .507 Age (yrs) 0.98 0.97−0.98 <.001
Age (yrs) 0.97 0.96−0.98 <.001 Race and ethnicity
Race and ethnicity Hispanic or Latino 1.22 1.10−1.35 <.001
Hispanic or Latino 0.89 0.69−1.15 .358 Non-Hispanic Black 1.20 1.10−1.31 <.001
Non-Hispanic Black 1.12 0.91−1.37 .283 Non-Hispanic White Ref.
Non-Hispanic White Ref. None of the above 1.05 0.92−1.19 .468
None of the above 0.98 0.73−1.31 .882 BMI
BMI Underweight 1.17 0.80−1.82 .410
Underweight 0.68 0.25−1.83 .442 Normal weight Ref.
Normal weight Ref. Overweight 0.95 0.86−1.05 .301
Overweight 0.85 0.69−1.05 .136 Obese 4 0.97 0.88−1.07 .476
Obese 0.69 0.56−0.84 <.001 Parity
Parity 0 Ref.
0 Ref. 1 1.14 1.01−1.28 .023
1 0.86 0.65−1.12 .264 2 1.13 1.03−1.25 .011
2 1.10 0.88−1.37 .407 3+ 1.30 1.17−1.43 <.001
3+ 1.08 0.87−1.36 .482 ASA status
ASA status I Ref.
I Ref. II 1.24 1.09−1.41 .001
II 1.09 0.83−1.44 .522 III+ 1.61 1.39−1.87 <.001
III+ 1.21 0.88−1.67 .232 Diabetes 1.11 0.99−1.25 .059
Diabetes 0.90 0.67−1.21 .485 Current smoker 1.16 1.07−1.27 <.001
Current smoker 1.01 0.83−1.23 .904 Hypertension 0.99 0.91−1.07 .875
Hypertension 1.43 1.19−1.71 <.001 Bleeding disorder 1.64 1.22−2.20 .001
Bleeding disorder 2.53 1.47−4.35 .001 Pelvic inflammatory disease 0.95 0.76−1.18 .681
Pelvic inflammatory disease 0.96 0.59−1.56 .861 Previous abdominal surgery 1.17 1.08−1.25 <.001
Previous abdominal surgery 1.32 1.13−1.55 .001 Previous pelvic surgery 1.08 1.01−1.16 .021
Previous pelvic surgery 1.13 0.96−1.33 .130 Approach
Approach TLH Ref.
TLH Ref. LAVH 1.15 1.05−1.26 .001
LAVH 1.32 1.08−1.61 .006 TVH 1.47 1.33−1.61 <.001
TVH 1.70 1.38−2.09 <.001 Operative time 1.004 1.004−1.005 <.001
Operative time (£ 10 min) 1.02 1.01−1.04 <.001
ASA = American Society of Anesthesiologists; BMI = body mass index;
ASA = American Society of Anesthesiologists; BMI = body mass index; CI = confidence interval; LAVH = laparoscopically assisted vaginal hysterec-
CI = confidence interval; LAVH = laparoscopically assisted vaginal hysterec- tomy; OR = odds ratio; TLH = total laparoscopic hysterectomy; TVH = total
tomy; OR = odds ratio; TLH = total laparoscopic hysterectomy; TVH = total vaginal hysterectomy.
vaginal hysterectomy.

Table 4
AUROC, optimal cutoff, sensitivity, and specificity for predicting complications based on uterine weight, stratified by approach

Approach N AUROC (95% CI) Optimal uterine Sensitivity Specificity

weight cutoff (g)
TLH 47 296 0.53 (0.52−0.54) 163 45% 60%
LAVH 9309 0.55 (0.52−0.57) 191 37% 72%
TVH 7684 0.56 (0.54−0.59) 113 59% 50%

AUROC = area under the receiver operating characteristics curve; CI = confidence interval; LAVH = laparoscopically assisted vaginal hysterectomy; TLH = total laparo-
scopic hysterectomy; TVH = total vaginal hysterectomy.

Downloaded for Anonymous User (n/a) at Hualien Tzu Chi Medical Center from ClinicalKey.com by Elsevier on October 26,
2023. For personal use only. No other uses without permission. Copyright ©2023. Elsevier Inc. All rights reserved.
6 Journal of Minimally Invasive Gynecology. Vol 00, No 00, 00 2023

Table 6
AUROC, optimal cutoff, sensitivity, and specificity for predicting complications based on uterine weight, stratified by predicted risk tertile

Predicted risk tertile N AUROC (95% CI) Optimal uterine Sensitivity Specificity
weight cutoff (g)
Low risk 21 430 0.50 (0.48−0.52) 86 72% 30%
Moderate risk 21 430 0.52 (0.50−0.53) 208 29% 74%
High risk 21 429 0.52 (0.51−0.54) 131 63% 41%

AUROC = area under the receiver operating characteristics curve; CI = confidence interval; LAVH = laparoscopically assisted vaginal hysterectomy; TLH = total laparo-
scopic hysterectomy; TVH = total vaginal hysterectomy.

Many studies have noted that surgical expertise and peri- correlation between uterine weight and hysterectomy com-
operative planning directly affect the outcome of hysterec- plications. Surgeon-level factors are known to primarily
tomies, particularly in the setting of large uteri [4,13,18]. drive the cost variance across hysterectomy approach [19],
Studies have examined both single surgeons at single hospi- and surgical approach is also most strongly determined by
tals and larger patient populations across different hospital the surgeon [20]. Although some data point to surgeon-level
systems and surgeons [13]. The lack of association between factors correlating with complications in the context of this
increased risk of complications and larger uterine weight procedure, additional research is needed to specifically
identified in some studies may reflect a higher degree of characterize how increased surgeon expertise with perform-
surgical expertise reflected in those study samples [4,18]. ing minimally invasive hysterectomy with large uteri can
This has been demonstrated in studies where documented translate to reducing the operative risk in these cases.
intraoperative complications were negatively correlated
with increasing expertise of the surgeon [4,13,18]. The Acknowledgments
NSQIP dataset does not provide a way to quantify surgeon
expertise, presenting a limitation to this study, nor does not The ACS NSQIP and the hospitals participating in the
include information on specific institutional practices that ACS NSQIP are the source of the data used herein; they
may influence preoperative decision making. have not verified and are not responsible for the statistical
The strengths of our study include the quality and detail validity of the data analysis or the conclusions derived by
of the NSQIP data, facilitating the generalizability of our the authors.
study findings. In addition, the NSQIP measurement of 30-
day complications was standardized across participating
institutions, providing a clinically relevant measure of oper- References
ative complications. Limitations of our study include the 1. Kehde BH, van Herendael BJ, Tas B, Jain D, Helsen K, Jochems L.
potential for uterine weight measurement error or other Large uterus: what is the limit for a laparoscopic approach? Autops
potential errors in the source data, including complications Case Rep. 2016;6:51–56.
2. Maccio A, Chiappe G, Kotsonis P, et al. Feasibility and safety of total
and reoperations not included in the dataset. In addition, laparoscopic hysterectomy for uteri weighing from 1.5 kg to 11.000
estimated uterine volume was not available within this data- kg. Arch Gynecol Obstet. 2021;303:169–179.
set, which would likely have been used as part of preopera- 3. Wu KY, Lertvikool S, Huang KG, Su H, Yen CF, Lee CL. Laparo-
tive counseling. The large sample size can identify scopic hysterectomies for large uteri. Taiwan J Obstet Gynecol.
statistically significant differences in risk that are not clini- 2011;50:411–414.
4. Louie M, Strassle PD, Moulder JK, Dizon AM, Schiff LD, Carey ET.
cally significant. We suspect this was the case for our logis- Uterine weight and complications after abdominal, laparoscopic, and
tic regression analysis, where a dichotomized measure of vaginal hysterectomy. Am J Obstet Gynecol. 2018;219:480.e1–480.e8.
uterine weight was associated with a slight increase in com- 5. Fiaccavento A, Landi S, Barbieri F, et al. Total laparoscopic hysterec-
plication risk, despite bivariate ROC analysis showing vir- tomy in cases of very large uteri: a retrospective comparative study. J
tually no predictive utility when regressing risk of Minim Invasive Gynecol. 2007;14:559–563.
6. Uccella S, Cromi A, Bogani G, Casarin J, Formenti G, Ghezzi F. Sys-
complications on uterine weight alone. Finally, the NSQIP tematic implementation of laparoscopic hysterectomy independent of
database does not include information on use of the surgical uterus size: clinical effect. J Minim Invasive Gynecol. 2013;20:505–
robot, so TLH cases in this study included a combination of 516.
robotic and nonrobotic procedures. 7. O’Hanlan KA, McCutcheon SP, McCutcheon JG. Laparoscopic hys-
In conclusion, our study demonstrates that uterine terectomy: impact of uterine size. J Minim Invasive Gynecol.
2011;18:85–91.
weight is generally irrelevant for predicting complications 8. Sinha R, Swarnasree G, Rupa B, Madhumathi S. Laparoscopic hyster-
after minimally invasive hysterectomy. Our conclusions ectomy for large uteri: outcomes and techniques. J Minim Access
bolster existing literature documenting the absence of a Surg. 2019;15:8–13.

Downloaded for Anonymous User (n/a) at Hualien Tzu Chi Medical Center from ClinicalKey.com by Elsevier on October 26,
2023. For personal use only. No other uses without permission. Copyright ©2023. Elsevier Inc. All rights reserved.
Mouro et al. Uterine Weight for Predicting Surgical Complications 7

9. Tanacan A, Gunes AC, Unal C, Usubutun A, Beksac MS. Impact of associated with increased 30-day perioperative complications. J Minim
uterine weight on the surgical outcomes of vaginal hysterectomy. J Invasive Gynecol. 2015;22:1049–1058.
Gynecol Surg. 2019;35:184–189. 15. Wattiez A, Soriano D, Fiaccavento A, et al. Total laparoscopic hyster-
10. Bronsert M, Singh AB, Henderson WG, Hammermeister K, Meguid ectomy for very enlarged uteri. J Am Assoc Gynecol Laparosc.
RA, Colborn KL. Identification of postoperative complications using 2002;9:125–130.
electronic health record data and machine learning. Am J Surg. 16. Uccella S, Morosi C, Marconi N, et al. Laparoscopic versus open hys-
2020;220:114–119. terectomy for benign disease in uteri weighing >1 kg: a retrospective
11. Stewart KA, Tessier KM, Lebovic DI. Comparing characteristics of analysis on 258 patients. J Minim Invasive Gynecol. 2018;25:62–69.
and postoperative morbidity after hysterectomy for endometriosis ver- 17. Akazawa M, Lee SL, Liu WM. Impact of uterine weight on robotic
sus other benign indications: a NSQIP study. J Minim Invasive Gyne- hysterectomy: analysis of 500 cases in a single institute. Int J Med
col. 2022;29. 884−890.e2. Robot. 2019;15:e2026.
12. Carter JV, Pan J, Rai SN, Galandiuk S. ROC-ing along: evaluation and 18. Ito TE, Vargas MV, Moawad GN, et al. Minimally invasive hysterec-
interpretation of receiver operating characteristic curves. Surgery. tomy for uteri greater than one kilogram. J Soc Laparoendosc Surg.
2016;159:1638–1645. 2017;21. e2016.00098.
13. Macci o A, Chiappe G, Kotsonis P, et al. Surgical outcome and compli- 19. Whiteside JL, Tumin D, Hohmann SF, Harris A. Determinants of cost
cations of total laparoscopic hysterectomy for very large myomatous for outpatient hysterectomy for benign indications in a nationwide
uteri in relation to uterine weight: a prospective study in a continuous sample. Obstet Gynecol. 2023;141:765–772.
series of 461 procedures. Arch Gynecol Obstet. 2016;294:525–531. 20. Albright BB, Heyward QD, Erkanli A, et al. Geographic variation in the rate
14. Catanzarite T, Saha S, Pilecki MA, Kim JYS, Milad MP. Longer oper- and route of hysterectomy for benign disease in the USA: a retrospective
ative time during benign laparoscopic and robotic hysterectomy is cross-sectional study. BJOG. 2023 May 2. [Epub ahead of print].

Downloaded for Anonymous User (n/a) at Hualien Tzu Chi Medical Center from ClinicalKey.com by Elsevier on October 26,
2023. For personal use only. No other uses without permission. Copyright ©2023. Elsevier Inc. All rights reserved.