Forty-eight patients met the inclusion criteria. No patients were excluded (i.e., none had a clinical hernia on abdominal examination). Characteristics of the 48 patients are summarized in . Of note, 38 patients (79%) had had a previous laparoscopy, and of these, almost all (n = 35) had had a previous laparoscopy for the same pain. The procedures performed at the previous laparoscopy are also summarized in .
At the time of empiric laparoscopic inguinal exploration and mesh placement, there was a laparoscopic diagnosis of an ipsilateral patent processus vaginalis in 7 patients (15%) (; ); in an additional 5 patients (10%), there were laparoscopic findings after exploration of the ipsilateral inguinal internal ring (). Empiric ipsilateral inguinal exploration and mesh placement was done for all patients, regardless of whether these findings were present or not. In addition, 4 patients (8%) had a laparoscopic abnormality of the contralateral inguinal region (the side with no pain or tenderness; ), which was not explored or repaired. Four patients (8%) had a concurrent excision of endometriosis classified as symptomatic, while 13 patients (27%) had concurrent excision of endometriosis classified as incidental ().
Five patients did not return for a postoperative visit, and therefore 43 patients were informative for the primary outcome (pain level at the last postoperative visit). The average time to the last postoperative visit was 12.6 ± 14.2 mo (range ≤ 1 to 58) from the date of surgery. For the primary (short-term) outcome, there was pain improvement in 15 patients (35%) (complete resolution in 3 and partial resolution in 12 patients), pain improvement then return of the pain in 18 patients (42%), and pain unchanged in 9 patients (21%) and worse in 1 patient (2%). The average time to return of the pain was 8.7±-9.8 mo (range ≤ 1 to 34) from the date of surgery, and triggers were trauma (n = 3), sports (n = 2), pregnancy (n = 2), bikini wax (n = 1), and unknown (n = 10).
The predictor variables are listed in . Neither a concurrent surgical procedure (such as excision of endometriosis, whether classified as symptomatic or incidental), nor the presence of an ipsilateral patent processus vaginalis, was associated with the primary outcome (). The only predictor variable significantly associated with the primary outcome was a positive Carnett's test in the ipsilateral abdominal lower quadrant, with a positive Carnett's test associated with improvement of the pain at the last postoperative visit (Spearman's rho = 0.34, P = .024). Of the 11 patients with a positive Carnett's test informative for the primary outcome, improvement occurred in 8 patients (73%) (complete resolution in 2 and partial resolution in 6 patients), improvement then return of the pain in 2 patients, no change in 1 patient, and worsening in no patients. Of the 32 patients with a negative Carnett's test, improvement occurred in 7 patients (22%) (complete resolution in 1 and partial resolution in 6 patients), improvement then return of the pain in 16 patients, no change in 8 patients, and worsening in 1 patient. None of the other predictor variables in had an association with the primary outcome.
In addition, there was possible evidence of selection bias, as patients who had a longer time to the last postoperative visit had a trend towards more pain for the primary outcome (Spearman's rho=-.31, P = .045). However, when a linear regression model was constructed with a positive Carnett's test and time to last postoperative visit as predictor variables for the primary outcome, the time to last postoperative visit fell out of the model (P = .14) with only the positive Carnett's test remaining significant (P = .024).
There were no intraoperative complications, and the postoperative complications were mild and uncommon (10%): hospitalization for postoperative pain (n = 2), bladder infection (n = 1), endometritis (n = 1), and “slow recovery” (n = 1). Eight patients required reoperation for pain (17%), which included repeat laparoscopy for mesh removal (n = 3), open groin exploration by a general surgeon (n = 2), hysterectomy (n = 1), hysterectomy and ipsilateral salpingo-oophorectomy (n = 1), and unknown (n = 1). One patient requested mesh removal after a motor vehicle accident resulted in return of the pain, and no inguinal abnormality was noted during repeat laparoscopy. A second patient requested mesh removal after return of the pain secondary to trauma, and again no inguinal abnormality was noted. A third patient requested mesh removal after the pain returned (unknown cause), and the inguinal canal looked slightly inflamed and thickened with pathology showing mild chronic and foreign body inflammation. None of these patients had significant improvements in their pain after mesh removal (and one patient requested repeat inguinal exploration and mesh placement), although follow-up was limited. An additional 2 patients were referred to a general surgeon and underwent open groin exploration. One of these patients had experienced no improvement after laparoscopic inguinal exploration and mesh placement, while the other patient had experienced improvement then return of the pain. After open groin exploration, both patients had an initial improvement, then return of the same pain within 3 mo.
For the secondary (long-term) outcomes, 13 patients (27%) returned the questionnaire. The time between the surgery and the date of the questionnaire was 73.2 ± 30.6 mo (range = 23 to 102). There was no evidence of selection bias, as these 13 patients had a similar distribution for the primary outcome (improvement in 4, improvement then return of the pain in 5, no change in 3, and worse in 0) compared to the rest of the sample (P = .54), and similar initial (preoperative) SF-36 subscale scores compared to the rest of the sample (P = .39 to 0.98). In these 13 patients who returned the follow-up questionnaire, pain was now described as being improved in 9 patients (69%), unchanged in 4 patients (31%), and worse in 0 patients. Three SF-36 subscales improved from the initial questionnaire (preoperative) to the follow-up questionnaire (postoperative): physical functioning (P = .032), social functioning (P = .036), and pain (P = .035) (). The SF-36 question about health change also improved (P = .003) ().
Comparison of SF-36 Subscale Scores