Vaginal cuff dehiscence can present as a serious complication after hysterectomy. Although the overall incidence of dehiscence is low, with reported rates ranging from 0.03% to 0.28%, recent data show that postoperative dehiscence complications vary by mode of hysterectomy.1,2 Specifically, total laparoscopic hysterectomies have been associated with higher rates of dehiscence compared with abdominal or vaginal routes.2–5
In 2007, we reported a large case series of 7,039 total hysterectomies performed over a 6-year period (January 2000 to March 2006) at Magee-Womens Hospital of the University of Pittsburgh Medical Center health system.5 Although our 0.14% incidence of dehiscence after all modes of hysterectomy was similar to prior publications, we noted a heightened annual incidence of dehiscence (0.70%) in 2005. After analyzing dehiscences by mode of hysterectomy, we reported a 4.93% incidence of dehiscence among total laparoscopic hysterectomies from January 2005 to March 2006.5
Although we agree total laparoscopic hysterectomies have a greater risk of vaginal cuff dehiscence, we believe our prior report was disproportionately elevated from the unexplained concentration of dehiscences in 2005. The objective of this follow-up study was to update the incidence of vaginal cuff dehiscence after total hysterectomy and to compare the incidence by different modes of surgery (abdominal, vaginal, and laparoscopic). We also aimed to report patient and surgical characteristics of those presenting with a dehiscence complication.
MATERIALS AND METHODS
This was an observational cohort study of women who underwent laparoscopic hysterectomies over a 10-year period. Institutional review board approval was obtained from the University of Pittsburgh Medical Center before initiation of the review. Pertinent Physicians' Current Procedural Terminology Coding System, 4th edition (CPT-4) procedure codes and International Classification of Diseases, 9th Revision diagnostic codes were used to identify all patients who underwent a hysterectomy or repair of a vaginal cuff dehiscence at Magee-Womens Hospital between January 2006 and December 2009. The data from the follow-up study period (January 2006 to December 2009) were analyzed separately and in combination with the data from the initial study period (January 2000 to March 2006) for a 10-year cumulative analysis (January 2000 to December 2009). Of note, for the purposes of the overall analysis, the initial study period was redefined from January 2000 to December 2005 to avoid the 3-month overlap between the two study periods.
Vaginal cuff dehiscence was defined as partial or complete separation of the vaginal cuff with or without evisceration that required surgical intervention. Supracervical hysterectomy (SCH) was defined as surgical removal of the uterine corpus alone. Total hysterectomy was defined as surgical removal of the uterine corpus and cervix. Total hysterectomies were categorized by mode of incision as total abdominal hysterectomies (TAHs), total vaginal hysterectomies, laparoscopically assisted vaginal hysterectomies (LAVHs), total laparoscopic hysterectomies, or total robotic hysterectomies. All total laparoscopic hysterectomies in our study were type IVE laparoscopic hysterectomies as outlined in the American Association of Gynecologic Laparoscopists classification system.6 Type IVE laparoscopic hysterectomy is defined as a hysterectomy performed entirely through a laparoscopic approach, including a laparoscopic colpotomy and cuff closure. Hysterectomies performed laparoscopically but with a vaginal colpotomy or vaginal cuff closure were categorized as LAVH in our study. Because separate CPT-4 codes distinguishing total laparoscopic hysterectomy and TRH from LAVH did not exist for the majority of the study period, a chart review of the LAVH-coded hysterectomies was conducted for the full 10-year period.
Only patients with vaginal cuff dehiscence clinically significant enough to require surgical repair were included in the study. Specific cases of cuff dehiscence were excluded if the hysterectomy was not performed at Magee-Womens Hospital. In addition, only total hysterectomies were included in the dehiscence analysis because SCHs are not at risk for this complication. TRHs were excluded from the dehiscence analysis because they were performed only during the last 2 months of the study period.
The overall and annual incidence of vaginal cuff dehiscence was calculated by mode of hysterectomy. The risk of dehiscence after total laparoscopic hysterectomy was then compared with other modes of hysterectomy using a risk ratio calculation. All analyses were conducted using SAS 9.2. Risk ratios and 95% confidence intervals (CIs) were calculated using a log-binomial regression model.7 All P values were calculated from a Fisher's exact test with P<.05 considered statistically significant. This was a sample of convenience to describe the incidence of vaginal dehiscence at one institution; therefor, a power calculation was not done before data collection. The power of the sample was calculated during data analysis.
As a secondary assessment, the medical records of all patients with a vaginal dehiscence were reviewed to identify characteristics associated with this complication. Patient information collected included age, race, medical history, parity, menopausal status, smoking history, body mass index, and immune status (immunocompromised compared with healthy patient). Surgical characteristics included mode of hysterectomy, indication for surgery, any additional procedures performed at the time of hysterectomy, technique for colpotomy and vaginal cuff closure (including suture material), use of prophylactic antibiotics at the time of hysterectomy, intraoperative and postoperative complications (ie, vaginal cuff cellulitis), and pathology results. The clinical presentation and management of these dehiscences, including the interval time between hysterectomy and dehiscence, trigger events to onset of dehiscence, presenting symptoms at the time of vaginal dehiscence, presenting or prolapsing organ if evisceration present, and type of dehiscence repair (ie, mode of surgery for repair such as vaginal compared with abdominal compared with laparoscopic compared with a combined approach) were analyzed.
Between January 2000 and December 2009, there were 12,472 hysterectomies. Seven robotic cases and 859 SCHs were excluded, resulting in 11,606 total hysterectomies included in this analysis. Among the total hysterectomies, there were 7,392 (63.7%) TAHs, 2,534 (21.8%) total vaginal hysterectomies, and 1,687 (14.5%) laparoscopic hysterectomies. A chart review of the laparoscopic hysterectomies confirmed 958 (8.2%) total laparoscopic hysterectomies and 722 (6.2%) LAVHs.
During the 10-year analysis, 28 vaginal cuff dehiscences were repaired at Magee-Womens Hospital with the following breakdown by mode of hysterectomy: 13 total laparoscopic hysterectomies (46.3%), 11 TAHs (39.3%), two LAVHs (7.1%), and two total vaginal hysterectomies (7.1%). The 10-year cumulative incidence of dehiscence after all modes of hysterectomy was 0.24% (95% CI 0.15–0.33) with an annual incidence ranging from 0.00% to 0.70% (Tables 1 and 2). When stratifying total hysterectomies by mode of surgery, the 10-year cumulative incidence of dehiscence is as follows: total laparoscopic hysterectomy 1.35% (95% CI 0.72–2.3), LAVH 0.28% (95% CI 0.03–1.0), TAH 0.15% (95% CI 0.07–0.27), and total vaginal hysterectomy 0.08% (95% CI 0.01–0.28) (Table 2). The difference in incidence by mode of hysterectomy was statistically significant (P<.001). The risk ratio for dehiscence after a total laparoscopic hysterectomy compared with other modes of hysterectomy was 9.1 (95% CI 4.1–20.3) compared with TAH, 17.2 (95% CI 3.5–75.9) compared with total vaginal hysterectomy, and 4.9 (95% CI 1.1–21.5) compared with LAVH (Table 3).
A comparison between the two study periods is presented in Tables 2 and 3. The results of our prior analysis from January 2000 to March 2006 are reported in detail in our previous article.5 The readjusted data analysis of the initial study period is described in this article. Between January 2000 and December 2005, a total of 11 vaginal cuff dehiscences were repaired at Magee-Womens Hospital. Two of these cases were complications after hysterectomies performed at another institution. Therefore, only nine dehiscences were included in the analysis. The cumulative incidence of vaginal dehiscence after all modes of hysterectomy during this readjusted 6-year period was 0.13%. The incidence of dehiscence varied considerably by mode of hysterectomy: total laparoscopic hysterectomy 2.76% (95% CI 0.87–4.64), total vaginal hysterectomy 0.06% (95% CI 0.0–0.18), TAH 0.00%, and LAVH 0.00% (Table 2). Eight of the nine dehiscences occurred in 2005, establishing the highest annual incidence of dehiscence at 0.70% that year (Table 1). Among the eight dehiscences in 2005, seven were associated with total laparoscopic hysterectomy, resulting in a 6.73% incidence of dehiscence after total laparoscopic hysterectomy in 2005 (Table 1). In the redefined initial study period, the risk ratio of dehiscence after total laparoscopic hysterectomy compared with total vaginal hysterectomy was 44.6 (95% CI 5.6–355.3). There were no dehiscences observed among TAHs or LAVHs during this time period, so we were unable to calculate a risk ratio comparing total laparoscopic hysterectomy with these subsets with zero events (Table 3).
During the 4-year follow-up study period (January 2006 to December 2009), 28 vaginal cuff dehiscences were repaired. Only 19 of the 28 dehiscences had their initial hysterectomy at Magee-Womens Hospital. Of the 19 vaginal cuff dehiscences between January 2006 to December 2009, 11 (57.9%) were complications of TAH, 5 (26.3%) were complications of total laparoscopic hysterectomy, two (10.5%) were complications of LAVH, and one (5.3%) was a complication of total vaginal hysterectomy (Table 2). The overall cumulative incidence of vaginal cuff dehiscence after all routes of total hysterectomy during this 4-year period was 0.39% (95% CI 0.21–0.56). The breakdown by route of surgery is as follows: total laparoscopic hysterectomy 0.75% (95% CI 0.09–1.40), LAVH 0.46% (95% CI 0.0–1.10), TAH 0.38% (95% CI 0.16–0.61), and total vaginal hysterectomy 0.11% (95% CI 0.0–0.32) (Table 2). No significant difference was appreciated by mode of hysterectomy (P=.24). However, the sample size of the follow-up study period was not powered. When comparing the risk of dehiscence after total laparoscopic hysterectomy to other modes of hysterectomy, the risk ratio was 6.9 (95% CI 0.8–58.6) compared with total vaginal hysterectomy, 2.0 (95% CI 0.7–5.6) compared with TAH, and 1.6 (95% CI 0.3–8.3) compared with LAVH (Table 3).
Patient characteristics, details of primary hysterectomy, clinical presentation at the time of dehiscence, and type of repair were reviewed for all 28 dehiscence patients and evaluated by mode of hysterectomy. The data are presented in Tables 4 and 5.
Historically, studies examining vaginal cuff dehiscence have been limited given the challenge of calculating a true incidence of dehiscence after hysterectomy. This is because total hysterectomy numbers are often not reported with the dehiscence events. Furthermore, mode of hysterectomy was difficult to analyze because a CPT-4 code distinguishing total laparoscopic hysterectomy from LAVH did not exist until 2008. Overcoming such difficulties, we present a large study population to examine the incidence of dehiscence after varying modes of hysterectomy.
Prior studies have reported a low incidence of dehiscence ranging from 0.03% to 0.28%1,2 with conflicting conclusions regarding factors contributing to this complication. Although some studies are unable to comment on the risk of mode of surgery, others have concluded that route of hysterectomy has no effect on evisceration, but rather patient characteristics such as postoperative coitus, menopausal status, and pelvic floor prolapse affect dehiscence rates.1,2,8–10 These studies contained few if any laparoscopic hysterectomies, because the procedure was still in its early stages of practice.
In 2006, Iaco et al published a 0.28% incidence of evisceration after 3,593 hysterectomies. This was the first study to compare different modes of hysterectomy that included the laparoscopic route. Although a greater rate of evisceration after laparoscopic hysterectomies was reported (0.26% abdominal, 0.25% vaginal, 0.79% laparoscopic), the authors concluded that route of surgery does not influence risk of dehiscence because the difference was not statistically significant.2 Although Iaco et al's study included 127 laparoscopic hysterectomies, it was most likely not powered to detect a difference. Therefore, the lack of significance may reflect the lack of power to detect a difference rather than a true lack of difference in risk.
In 2007, we initially reported a 4.93% cumulative incidence of dehiscence among total laparoscopic hysterectomies between January 2005 and March 2006.5 In our readjusted analysis of the redefined initial study period (January 2000 to December 2005), we found a 2.76% (95% CI 0.87–4.64) incidence of dehiscence among total laparoscopic hysterectomies, which was higher than all other modes of hysterectomy (total vaginal hysterectomy 0.06%, 95% CI 0.0–0.18; TAH 0%; LAVH 0%) (Table 2). Although total laparoscopic hysterectomies continued to have the highest incidence of dehiscence in the follow-up period (0.75%; 95% CI 0.09–1.4) and the combined 10-year study period (1.35%; 95% CI 0.72–2.3), the incidence was appreciably lower than previously reported (4.93% in 2007 publication, 2.76% readjusted calculation).
In contrast to Iaco et al, we found the increase in dehiscences among total laparoscopic hysterectomies to be significant. In our 10-year analysis, the risk of dehiscence was 4.9 times more likely among total laparoscopic hysterectomies compared with LAVHs (95% CI 1.1–21.5), 9.1 times more likely compared with TAH (95% CI 4.1–20.3), and 17.2 times more likely compared with total vaginal hysterectomy (95% CI 3.9–75.9) (Table 3). The conflicting results between the two studies may be explained by the sample size. Because the incidence of dehiscence complications is so rare, a large sample size is required to detect any significant differences. A power calculation for the overall time period in our study revealed 98.7% power to detect differences between total laparoscopic hysterectomy and TAH, 98.1% power between total laparoscopic hysterectomy and total vaginal hysterectomy and 65.0% power between total laparoscopic hysterectomy and LAVH. When comparing our two study periods, we found the initial study period was powered to detect significant differences (99.1% power for total laparoscopic hysterectomy compared with TAH, 97.2% power for total laparoscopic hysterectomy compared with total vaginal hysterectomy, 81.5% power for total laparoscopic hysterectomy compared with LAVH), whereas the follow-up study period was not (14.5% power for total laparoscopic hysterectomy compared with TAH, 37.4% power for total laparoscopic hysterectomy compared with total vaginal hysterectomy, 5.5% power for total laparoscopic hysterectomy compared with total vaginal hysterectomy).
Another explanation for the conflicting results may be differences in methodologies. The Iaco et al study did not specify the type(s) of laparoscopic hysterectomies performed (LAVHs compared with total laparoscopic hysterectomy). We believe this distinction is critical because LAVHs use different colpotomy and cuff closure techniques compared with total laparoscopic hysterectomy.
The cause for our increased dehiscences among total laparoscopic hysterectomies in 2005 remains unknown. Advanced fellowship-trained minimally invasive gynecologic surgeons performed the vast majority of total laparoscopic hysterectomies in the initial study period, making the learning curve an unlikely explanation. Interestingly, total laparoscopic hysterectomy dehiscence rates decreased during the follow-up study period when total laparoscopic hysterectomies became more widely adopted. With more gynecologic surgeons learning to perform total laparoscopic hysterectomies, one may have anticipated an increased incidence of total laparoscopic hysterectomy-related dehiscences associated with the learning curve of new procedures; however, we saw the opposite effect.
The only change identified during the initial study period was a change in the Polysorb suture composition by the manufacturer in 2005. Polysorb suture was used for cuff closure for the majority of hysterectomies at Magee-Womens Hospital. Since 2006, after the findings of our initial study, the minimally invasive gynecologic surgeons now use a delayed absorbable monofilament suture for cuff closure and advise delayed resumption of coitus postoperatively.
Since 2006, it appears the incidence of dehiscence decreased among total laparoscopic hysterectomies, whereas it increased among all other modes of hysterectomy (TAH, LAVH, and total vaginal hysterectomy). In the follow-up study period, TAHs had the greatest absolute increase in dehiscences (0 in 2000–2005 compared with 11 in 2006–2009); however, LAVHs had the greatest increase in the incidence of dehiscence (0% in 2000–2005 compared with 0.46% in 2006–2009) (Table 2). The change in dehiscence patterns may reflect the increasing use of electrosurgery (eg, Bovie, Bovie Medical Corporation) across all modes of surgery or a change in surgical technique by those performing total laparoscopic hysterectomy in response to our previous report. Unfortunately, our study is not equipped to address this clinically interesting observation. Although patient and surgical characteristics are described for the dehiscence patients, the value of this information is indeterminate because the same information is unknown for the remaining hysterectomy patients. The exact cause for cuff dehiscence remains unknown and is beyond the scope of our study.
There are several limitations to our study. First, the analysis is limited to one major academic institution, which may not reflect hysterectomy practices in other regions. Second, patients with cuff dehiscence complications after hysterectomy at Magee-Womens Hospital may present to another institution for care, resulting in an underreported incidence. Third, this was a sample of convenience with varying patient follow-up. Fourth, the retrospective nature of this study lacks randomization of mode of hysterectomy and relies on accurate CPT-4 coding and chart review. Ideally, all hysterectomy patients, and not just the 1,687 LAVH-coded cases and 28 dehiscences, would have been reviewed. Finally, the 10-year study period allows opportunities for historical contamination-like changes in surgical practice that can influence the incidence of complications.
Although there are limitations to this study, we feel the updated report is of clinical importance because the practice trend at Magee-Womens Hospital supports that total laparoscopic hysterectomy has replaced abdominal and not vaginal mode of hysterectomy. Although the relative risk of dehiscence is higher among total laparoscopic hysterectomies compared with other modes of hysterectomy, the well-known benefits of total laparoscopic hysterectomy11 weighed against the small absolute risk of total laparoscopic hysterectomy-related dehiscence may argue for a laparoscopic mode of hysterectomy from a patient's overall risk and benefit perspective.
In summary, although the incidence of dehiscence is highest among total laparoscopic hysterectomies (1.35%), it is not as high as previously reported. Specific patient characteristics, methods for colpotomy, and vaginal cuff closure techniques may be important influences for dehiscence outcomes. However, given the overall rarity of a dehiscence event, we are unable to make any firm conclusions about patient and surgical characteristics at this time. Total laparoscopic hysterectomy remains an acceptable mode of hysterectomy to offer our patients.
1. Croak AJ, Gebhart JB, Klingele CJ, Schroeder G, Lee RA, Podratz KC. Characteristics of patients with vaginal rupture and evisceration. Obstet Gynecol 2004;103:572–6.
2. Iaco PD, Ceccaroni M, Alboni C, Roset B, Sansovini M, D'Alessandro L, et al.. Transvaginal evisceration after hysterectomy: is vaginal cuff closure associated with a reduced risk? Eur J Obstet Gynecol Reprod Biol 2006;125:134–8.
3. Agdi M, Al-Ghafri W, Antolin R, Arrington J, O'Kelley K, Thomson AJ, et al.. Vaginal vault dehiscence after hysterectomy. J Minim Invasive Gynecol 2009;16:313–7.
4. Doganay M, Yildiz Y, Tonguc E, Var T, Karayalcin R, Eryiotalmaz OG, et al.. Abdominal, vaginal and total laparoscopic hysterectomy: perioperative morbidity. Arch Gynecol Obstet 2011;284:385–9.
5. Hur HC, Guido RS, Mansuria SM, Hacker MR, Sanfilippo JS, Lee TT. Incidence and patient characteristics of vaginal cuff dehiscence after different modes of hysterectomies. J Minim Invasive Gynecol 2007;14:311–7.
6. Olive DL, Parker WH, Cooper JM, Levine RL. The AAGL classification system for laparoscopic hysterectomy. Classification committee of the American Association of Gynecologic Laparoscopists. J Am Assoc Gynecol Laparosc 2000;7:9–15.
7. Spiegelman D, Hertzmark E. Easy SAS calculations for risk or prevalence ratios and differences. Am J Epidemiol 2005;162:199–200.
8. Kowalski LD, Seski JC, Timmins PF, Kanbour AI, Kunschner AJ, Kanbour-Shakir A. Vaginal evisceration: presentation and management in postmenopausal women. J Am Coll Surg 1996;183:225–9.
9. Somkuti SG, Vieta PA, Daugherty JF, Hartley LW, Blackmon EB Jr. Transvaginal evisceration after hysterectomy in premenopausal women: a presentation of three cases. Am J Obstet Gynecol 1994;171:567–8.
10. Cardosi RJ, Hoffman MS, Roberts WS, Spellacy WN. Vaginal evisceration after hysterectomy in premenopausal women. Obstet Gynecol 1999;94:859.
11. Nezhat C, Nezhat F, Admon D, Nezhat AA. Proposed classification of hysterectomies involving laparoscopy. J Am Assoc Gynecol Laparosc 1995;2:427–9.
12. Johnson N, Barlow D, Lethaby A, Tavender E, Curr E, Garry R. Surgical approach to hysterectomy for benign gynecologic disease. The Cochrane Database of Systematic Reviews 2006, Issue 2. Art. No.: CD003677. DOI: 10.1002/14651858.CD003677.pub3.