Patients’ details
Twenty-five pelvic fracture patients with vaginal injury met the inclusion criterion during the study period. The mean age was 32.8±12.4 years old (range, 17-55). The mean ISS was 31.4±14.9 (range, 10-57). Nine patients in this series had an ISS of greater than 25. The mean time from injury to emergency department was 2.02±0.8 hours (range, 1-3.5). Most injuries were caused by traffic accidents (22 patients, 88%) while 2 patients falling from height and one patient encountered with crush injury.
Pattern of pelvic fractures and vaginal injury
Six patients (24%) with mucosal injury were categorized as first degree. Twelve patients (48%) suffered second degree injury, presented with involvement of vaginal muscularis but not penetrating through rectum or urethra. Seven patients (28%) suffered third degree injury. Four of them suffered vaginal-rectal penetrating injury and vaginal-urethral penetrating wound was presented in 3 patients. All diagnoses were made by a gynecologist and vaginal speculum was used when necessary.
According to Tile classification, the most common injury pattern was type B fracture, occurring in 18 patients (72%). Another 7 patients (28%) sustained a type C fracture, and none had type A fracture. The severity of vaginal injury didn’t differ between type B and C fractures according to our classification (p=0.208, Table 1). In Young-Burgess system, twenty patients sustained anterior-posterior compression (APC), five patients sustained vertical shear (VS) injuries and no one showed lateral compression (LC) injuries. The VS fracture patients showed higher severity of vaginal injury compared to patients with APC fracture (p=0.034, Table 1).
All patients suffered disruption of anterior pelvic ring. Most patients suffered pubic ramus fracture (20 patients, 80%). Among them, 10 patients presented unilateral fracture and 10 presented bilateral fracture. Four patients were also combined with pubic symphysis separation. Thirteen patients had a compromised pubic symphysis including 9 patients with pubic symphysis separation and 4 patients presented a rather rare floating pubic symphysis which means fracture of the bilateral superior and inferior pubic rami and ischial rami. A compromised pubic symphysis was related to more severe vaginal injury. (p=0.024, Table 1)
Treatment of pelvic fracture
Twenty-four patients survived initial resuscitation phase. External fixation was performed for 11 patients and 5 patients didn’t receive further surgery. Fourteen patients had open reduction and internal fixation (ORIF), among them first-stage ORIF was only performed for 5 patients. The other 9 patients had second-stage or delayed surgery due to unstable hemodynamic condition and abscess formation (7-14 days). In these 9 patients, external fixation was applied in 6 patients at first stage and pelvic binder was used for 2 patients. Conservative treatment was done in the other 5 patients.
Treatment of vaginal injury
No special treatment but only gauze packing was done in the 6 patients with mucosal injury and one patient with second degree injury. All the other patients with second or third degree injury went through surgical repair. Fourteen patients had primary closure and 4 patients had secondary vaginal repair. The wound was sutured through interrupted suture by a gynecologist with 1-0 absorbable sutures
Treatment of associated injury
Anorectal injury occurred in 15 patients, all of them were performed colostomy except for one patient died in emergency room. Another patient with severe vaginal-perineal laceration but no injury of rectum also had protective colostomy. All these patients were performed primary debridement and colostomy with anal sphincter repaired and distal lumen rinsed to avoid secondary contamination. Fifteen patients were complicated with urinary tract injuries. Of these 3 were bladder injuries. One patient acquired laparotomy with surgical repair of the bladder due to serious damage. Two patients had slight damage and were treated with indwelling catheter. Twelve patients suffered urethral disruption. Six of them required primary suprapubic drainage and subsequent delayed repair. Two patients with vaginal-urethral penetration had urethral realignment. The others were treated conservatively.
Outcome evaluation
Totally 24 patients were followed up at mean 17.7 months (range, 10-36). The function evaluation was done at least 12 months after discharge. No patient showed signs of radiographic nonunion. The average time of fracture healing was 4.3±1.2 months (range 3-7 months). The radiologic outcome was assessed using Tornetta and Matta criteria. Only 2 patients showed unsatisfactory results. One was fair and one was poor. Factors associated with pelvic outcomes were shown in Table 2. Eighteen patients were assessed as excellent, four patients were good, two patients were fair and none was poor. Pelvic outcome was better significantly among younger patients (p=0.043) and patients without urethral injury (p=0.02). Infected patients showed worse results (p=0.005). Though management of pelvic fractures has insignificant influence on pelvic function (p=0.055), patients received external fixation in isolation seemed to present worse pelvic function.
Four patients complained of pain in sex intercourse at last follow-up but did not demand further treatment. Nine patients were still virgin at the last follow-up. Sexual function was assessed in the other 15 patients using Fisher exact test but no significant factor related to sexual pain was found (Table 3). All surviving patients had normal menstruation while one patient in menopause at last follow-up. Nine patients had at least one child delivered and one patient was pregnant at the time of the latest follow-up. Six patients were performed cesarean section and three patients had natural birth.
Complications
Infection occurred in 6 patients. Four patients developed pelvic abscess and were treated with abscess incision and drainage. Other two patients were treated with debridement and changing dressings. Sensitive antibiotic drugs were chosen for infected patients based on the drug sensitivity tests. Vaginal stenosis occurred in 2 patients. No patient showed associated vessel or nerve injuries. Mechanical complications like instrumentation failure did not occur.
Limitations
There were several limitations in this study. The sample size was comparatively small and there was probably high number of undetected vaginal injuries. It was a retrospective study with long time span and long-term follow-ups were not able to acquire from all patients. This study also lacked a more concrete evaluation of sexual function.