Objective
To evaluate whether surgical volume has an impact on short-term outcomes of laparoscopic hysterectomy.
Study Design
This is a retrospective analysis of 1016 laparoscopic hysterectomies.
Results
The surgeons were divided into 2 groups based on a cutoff of 30 cases. Patient characteristics, the rates of laparotomy (4.5% vs 6.7%), and serious complications (3.6% vs 5.5%) were similar between 9 “high” and the remaining 39 “low volume” gynecologists, respectively ( P < .05). Mean operating time was longer in the “low volume” group. Compared with their first 29 hysterectomies, the “high volume” surgeons decreased their operating time significantly in their subsequent cases. The “high volume” surgeons improved their conversion rate (9.2% vs 2.4%; P < .0001) over time but not their serious complications.
Conclusion
In laparoscopic hysterectomy, increasing the surgical volume can reduce the operating time and the risk for conversion to laparotomy but not the rate of serious complications.
As a minimally invasive approach, laparoscopic hysterectomy is gaining popularity among gynecologic surgeons. It offers a safe alternative to abdominal hysterectomy with less pain, reduced length of hospital stay, and faster postoperative recovery. Despite these clear advantages to laparotomy, it may be associated with serious complications and requires advanced skills. There has been an increasing interest in evaluation of the effect of surgeon’s experience on perioperative outcomes. We recently completed a large study evaluating the short-term surgical indices and complications of laparoscopic hysterectomies performed at our institution. Our overall complication rate was low; however, 1.3% of the patients had urinary tract injury, and 0.7% had vaginal cuff dehiscence. Serious complications, defined as life threatening or those requiring reoperation, occurred in 3.9% of all cases. Of all the surgeries, 4.8% were converted to abdominal hysterectomy. In this secondary analysis of the data from this previous study, we aimed to assess the influence of surgical volume on short-term outcomes and complication rates of laparoscopic hysterectomy.
Materials and Methods
This is a post hoc analysis of the database that was generated to compare the short-term outcomes between total (TLH) and supracervical laparoscopic hysterectomy (LSH). After obtaining an approval from the Institutional Review Board for this retrospective cohort study, we reviewed the medical records of all women who underwent a laparoscopic hysterectomy, TLH or LSH, for benign gynecologic problems at Baystate Medical Center between November 1999 and August 2008. During this period, slightly more than 5000 women underwent hysterectomy at our center, averaging 650 cases a year. Of those surgeries, approximately 30% were performed laparoscopically. The details regarding data collection and the results of that comparison can be found in our previous study. Exclusion criteria consisted of laparoscopically assisted vaginal hysterectomies, malignancies, and any other concomitant planned procedure with exception of adnexal surgery, adhesiolysis, and intraoperative cystoscopy. The choice between TLH and LSH for each surgeon was so variable that we decided to combine the data for both to evaluate the effect of surgical volume on short-term outcomes for laparoscopic hysterectomy.
To determine an appropriate cutoff for surgical volume, we categorized concurrent surgical cases in increments of 10 up to 100 cases and evaluated our data using a receiver operating characteristics (ROC) analysis for 2 of our main outcomes, serious complications, and conversions. Based on these results, we found that sensitivity and specificity were maximized at a cutoff of 40 for serious complications and 30 for conversions ( Figure , A and B). When we looked at the number of surgeons that fell above or below these cutoffs, there was only the difference of 1 surgeon who had 36 cases. A cutoff of 30 cases was therefore chosen. Reviewing the literature, we noted that this number was consistent with a previously reported cutoff point. Surgeons who performed 30 or more surgeries were classified as “high volume,” whereas those who did less than 30 surgeries as “low volume.” We also compared the cases that were performed by the “high-volume” surgeons before they reached the volume cutoff with their subsequent cases.
We evaluated baseline characteristics and comorbid conditions between the groups. We then examined the difference in operating time, rates of serious complications, and conversion to laparotomy. The definitions that were used for each complication in this study can be found in our previous report. “Serious complications” were defined as complications that were either life threatening such as thromboembolic event and bleeding requiring transfusion, or those necessitating reoperation, which included visceral injury and vaginal cuff dehiscence.
Data were analyzed using Stata version 10 (StataCorp LP, College Station, TX). All continuous variables are reported using means and standard deviations. Distributions were evaluated and determined to not follow a normal distribution, therefore the Wilcoxon rank-sum test was used to evaluate significant differences. Categorical variables are reported using number and percent. Significant differences were evaluated using Fisher’s exact test. Statistic significance was set at a P value of < .05 for all tests.
Results
There were 1110 consecutive laparoscopic hysterectomies identified. Of those, 94 cases were excluded because of concomitant major pelvic or abdominal surgery, leaving a total of 1016 cases that were accomplished by a total of 48 surgeons. Based on the cutoff of 30, we identified 9 surgeons as “high volume,” who performed 840 (83.4%) of all hysterectomies in this study. The remaining 39 surgeons were grouped as “low volume.” Of 9 “high volume” surgeons, 1 was fellowship trained in reproductive endocrinology, and the other 8 were general gynecologists.
The majority of the patient characteristics including age, parity, body mass index, uterine weight, menopausal status, and adnexal removal rates were similar between the groups ( Table 1 ). Race distribution was significantly different. More than 75% of all comparison groups were made up of white women. Among the comorbid conditions, endometriosis was the only one that significantly differed between the groups ( Table 2 ). It was found more frequently in the first 29 cases of the “high-volume” surgeons as compared with their subsequent cases (17.6% vs 9.7%, P = .001). When we focused on the indications for surgery ( Table 3 ), pelvic organ prolapse was more commonly recorded as an indication by the “high-volume” surgeons as compared with the “low-volume” surgeons (4.5% vs 0%; P = .002).
| Characteristic | Low volume (n = 165) | High volume (n = 851) | P value a | High volume first 29 cases (n = 261) | High volume subsequent cases (n = 590) | P value a |
|---|---|---|---|---|---|---|
| Age, y | 43.4 ± 6.2 | 44.3 ± 7.0 | .1080 | 44.0 ± 6.6 | 44.5 ± 7.2 | .2804 |
| Gravidity b | 2.4 ± 1.6 | 2.4 ± 1.6 | .4980 | 2.4 ± 1.6 | 2.5 ± 1.6 | .5331 |
| Parity c | 1.9 ± 1.3 | 1.9 ± 1.2 | .7200 | 1.8 ± 1.2 | 1.9 ± 1.2 | .5092 |
| Body mass index, kg/m 2 d | 28.6 ± 7.5 | 28.1 ± 6.6 | .7898 | 27.4 ± 5.8 | 28.4 ± 6.9 | .1911 |
| Uterine weight, g e | 176.3 ± 132.7 | 212.3 ± 194.3 | .2213 | 176.0 ± 125.7 | 227.0 ± 214.4 | .1586 |
| Postmenopausal, n (%) | 11 (6.7) | 60 (7.1) | > .99 | 23 (8.9) | 37 (6.3) | .192 |
| Race, n (%) | .027 | .030 | ||||
| White | 137 (83.1) | 660 (77.6) | 206 (78.9) | 454 (77.0) | ||
| Hispanic | 4 (2.4) | 73 (8.6) | 28 (10.7) | 45 (7.6) | ||
| Black | 17 (10.3) | 75 (8.8) | 13 (5.0) | 62 (7.6) | ||
| Other/unknown | 7 (4.3) | 43 (5.1) | 14 (5.4) | 29 (4.9) | ||
| Concomitant uni/bilateral adnexal removal, n (%) | 81 (49.1) | 441 (51.8) | .552 | 141 (54.0) | 300 (50.9) | .414 |
a P values calculated using Wilcoxon rank-sum test for continuous variables and Fisher’s exact test for categoric variables;
b Low volume n = 159, high volume n = 822;
c Low volume n = 159, high volume n = 823;
d Low volume n = 163, high volume n = 808;
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
