Introduction

Obturator hernia (OH) is a relatively rare but quite important form of abdominal wall hernia. It has been first described by Pierre Roland Arnaud de Ronsil in 1724 and accounts for 0.07–1% of all hernias. It is also responsible for 0.4% of all patients with mechanical bowel obstruction [1,2,3,4,5]. OH occurs through the obturator foramen, which is formed by the rami of the ischium and the pubis, subsequently passing into the obturator canal which is 2–3 cm long and 1 cm wide, bordered by the free edge of the obturator membrane and the pubic bone and contains the obturator nerve and vessels [6].

Obturator hernia tends to occur more frequently in elderly, slender, multiparous women, with the vast majority of OH patients presenting with mechanical bowel obstruction. The associated clinical symptoms are mostly vague and non-specific with the pathognomonic sign for its diagnosis being the Howship–Romberg sign, which consists of ipsilateral inner thigh pain on internal rotation of the hip [1].

OH’s localization, its difficult palpation, and its non-specific clinical symptoms render its timely diagnosis extremely demanding, which, in most patients, is definitely achieved via an exploratory laparotomy or laparoscopy for acute bowel obstruction [7, 8]. These facts underline the paramount importance of timely diagnosis, especially in elderly people with multiple comorbidities, as any treatment delay may disproportionally increase the morbidity and mortality rates, with published mortality rates ranging between 11 and 50% [6, 9,10,11,12]. Fortunately, the wide spread and availability of computed tomography imaging allow for an early and precise diagnosis, favoring elective instead of emergent surgical treatment [7, 13].

This is the first systematic review investigating OH’s clinical presentation, radiological imaging, surgical treatment outcomes, perioperative morbidity, mortality, and recurrence rates.

Materials and methods

A combined automated and manual systematic database search in electronic search engines PubMed, Web of Science, Cochrane Library, and Google Scholar, for citations that included obturator hernias, using the keywords “obturator hernia” as well as “little old lady hernia” revealed over 800 studies. After an initial review of them, papers reporting only row data for each individual patient with surgically treated obturator hernias were included in our analysis. Publications of interest included randomized and non-randomized studies, case series, case reports, letters to the editor, and conference abstracts. Papers reporting only on obturator hernias discovered in the setting of a radiological examination or autopsy without commenting on their surgical treatment, studies in languages other than English, papers reporting on obturator hernias in children and adolescents, as well as studies published before 1990, were excluded from the analysis. Sub-studies of larger series by the same group were also excluded from our analysis to avoid data duplication.

Studies were individually screened by two independent reviewers and data were independently extracted using the same template, decided in advance. Discrepancies between the two reviewers were then examined, until a final consensus was reached. Data extracted from eligible studies included patient characteristics such as age, sex, body mass index (BMI), previous surgical history, clinical symptoms and signs on presentation, radiological evaluation modalities, surgical treatment, length of hospital stay (LOS), perioperative morbidity and mortality (defined as morbidity and mortality within 30 days from the surgical operation), as well as a possible obturator hernia recurrence during the follow-up period.

Statistical analysis was performed using the R language and environment for statistical computing (R Foundation for Statistical Computing, Vienna, Austria https://www.R-project.org). For continuous variables, Shapiro–Wilk test for normality was used and univariate analysis was performed using T test or the non-parametric Mann–Whitney U test, as appropriate. Categorical variables were examined using Chi-square test. Morbidity and mortality were analyzed by univariate logistic regression analysis. LOS was analyzed by linear regression with logistic transformation of the outcome. Survival analysis was performed using Kaplan–Meier log-rank test and univariate Cox proportional hazard models. Multivariate models were not attempted due to missing data. The level of statistical significance was set at 5% (α = 0.05).

Results

Study characteristics

A total of 74 studies [12, 14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86] were included in the present review (Flow Diagram). Sixty-one studies were reports of a single case and 13 studies [12, 32, 41, 43, 44, 48, 62, 64, 80, 81, 83,84,85] were case series, reporting on a total of 146 patients suffering from an obturator hernia. Forty studies originated from Asia, 17 from Europe, 7 from USA, 6 from Australia, 2 from Latin America, 1 from Canada and the remaining 1 from Africa. Nearly all included studies reported specific data regarding diagnosis, treatment strategy, and patients’ outcomes.

figure a

Patient characteristics

One hundred and forty-six patients with a total of 167 OH were included in our analysis, with 40.1% (n = 67 patients), 29.9% (n = 50 patients), and 25.2% (n = 21 patients) suffering from either a right, a left or bilateral OH, respectively. Apart from bilateral OH, a contralateral OH was intraoperatively identified in 5 patients (4.7%), further categorized as an occult contralateral OH. In 8 patients (4.8%), the OH localization was not reported. Mean patient age was 78.8 ± 13.4 years, and mean BMI was 17.9 ± 2.8 kg/m2, with females representing the vast majority of patients (97.9%). Additionally, 25.4% of the patients had undergone some kind of abdominal surgery in the past (Table 1). However, no statistically significant differences were observed, as for the mean patient age, mean BMI, patient’s gender as well as the past medical history of any abdominal surgical operation, between patients with either a right or a left OH (Table 2).

Table 1 Demographic data, obturator hernia localization as well as clinical, radiological, and operative and postoperative data for the entire cohort of the study
Full size table
Table 2 Comparison of left- and right-sided obturator hernias as for demographic, clinical, and operative and postoperative data
Full size table

The presence of an OH was associated with various non-specific symptoms and signs, which in a descending order of frequency included signs of bowel obstruction (93.8%), abdominal pain (84.8%), vomiting (75.7%), nausea (43.5%), and fever (12.8%). The Howship–Romberg sign was present in slightly more than half of the patients (56.2%) (Table 1). No statistically significant differences were demonstrated between these symptoms and signs and the localization of the OH (Table 2).

Following hospital admission and clinical examination, nearly all patients underwent a radiological evaluation, with the most common modality used being the computed tomography (CT) of the abdomen (84.2%), followed by plain abdominal radiographs (78.8%). An accurate preoperative diagnosis of OH was made in the majority of patients (82.1%) who underwent preoperative radiological evaluation (Table 1).

All patients (n = 146) were operated on, with 139 patients (95.2%) being operated emergently and the rest on an elective basis. An intraperitoneal surgical approach was performed in most cases (86.3%). Forty-two patients (28.8%) underwent a laparoscopic operation, out of which 9.5% required conversion to laparotomy. Nearly 40% of the patients required an enterectomy due to a non-viable bowel segment, caused by bowel strangulation (Table 1). Closure of the hernia defect was performed with the use of mesh in approximately half of the patients (n = 63 patients, 48.8%), whereas in the remaining patients, the hernia defect was closed with the use of stiches. Mesh repair was significantly more frequently implemented in patients without intestinal resection (n = 43 patients, 68.3%), compared with patients who underwent an intestinal resection due to bowel strangulation (n = 20 patients, 31.7%) (p = 0.036). An occult contralateral OH was intraoperatively identified in 5 patients (4.7%), without any significant difference between patients being operated on for a left or a right OH (Table 2). The mean LOS was 10.5 ± 9.7 days, whereas the perioperative morbidity and mortality rates were 26.7% and 11.6%, respectively. Recurrence of the surgically treated OH was observed in 2 cases (2.4%), both being right-sided hernias (Table 1). However, mean LOS, perioperative morbidity and mortality, as well as OH recurrence were irrespective of OH localization (Table 2).

Univariate logistic regression analysis

To investigate which parameters may influence the morbidity and mortality rates, we performed a univariate logistic regression analysis. Older patient age was demonstrated as a significant risk factor for postoperative morbidity (p = 0.007), without though showing any difference in the respective mortality rate. However, neither OH localization nor patient’s gender, BMI, or past medical history of an abdominal operation, were significantly correlated with the postoperative morbidity and mortality rate (Table 3).

Table 3 Univariate logistic regression analysis for factors associated with perioperative morbidity and mortality
Full size table

Regarding the presence of symptoms and signs as well as the preoperative imaging modality used, no statistically significant correlation was demonstrated between them and the postoperative morbidity and mortality rate, except for the preoperative use of abdominal CT scan, which was associated with a significantly decreased postoperative mortality rate (OR 0.29, p = 0.049), without affecting the morbidity rate though (Table 3).

Not surprisingly, the presence of a non-viable intestinal segment and the subsequent need for an enterectomy was significantly correlated with an increased morbidity (OR 2.76, p = 0.017) and mortality (OR 4.54, p = 0.016) rate. The laparoscopic treatment of an OH demonstrated a significantly decreased morbidity (OR 0.29, p = 0.022) and mortality rate (OR 0.84, p = 0.006), compared with the respective open repair. The same pattern was also observed as for the use of mesh for hernia defect closure, which was associated with a decreased morbidity rate (OR 0.35, p = 0.022), without affecting the mortality rate. Moreover, as it was expected, an increased morbidity rate was strongly correlated with an increased LOS (p < 0.001) and postoperative mortality rate (OR 1.55, p < 0.001), as well (Table 3).

Length of hospital stay was also independently analyzed as for various parameters which may be associated with. Older age (p < 0.001), presence of bowel obstruction (OR 3.08, p < 0.001), the use of plain abdominal radiographs for preoperative imaging (OR 2.26, p = 0.001), the need for bowel resection (OR 1.61, p = 0.001), emergent operation (OR 2.32, p = 0.01), as well as intraperitoneal surgical approach (OR 1.73, p = 0.018) were significant risk factors for an increased LOS. On the other hand, laparoscopic approach (OR 0.50, p < 0.001) and use of mesh for hernia defect closure (OR 0.66, p = 0.004) were associated with a decreased LOS (Table 4).

Table 4 Linear regression analysis for factors associated with length of hospital stay (LOS)
Full size table

Discussion

The distinct anatomical stages of OH formation have been well described in the literature, consisting of the first stage in which preperitoneal fat enters the obturator foramen forming a fat plug, followed by the second stage of hernia sac formation through peritoneal dimpling into the existing fat plug, with the hernia sac subsequently occupied by intraabdominal viscera herniating into it and causing symptoms and/or signs in the third stage [6, 87, 88].

Multiple predisposing factors have been associated with OH formation, including a wider pelvis causing a triangular-shaped obturator canal with an increased transverse diameter, multiparity, increased age usually in conjunction with emaciation, as well as chronic concomitant medical conditions which increase the intraabdominal pressure, such as chronic obstructive pulmonary disease, chronic constipation, ascites, and kyphoscoliosis [5]. The underlying mechanism predisposing elderly and emaciated people to OH formation is the extensive loss of preperitoneal adipose and lymphatic tissue, which normally overlies the obturator canal, thus creating a greater space around the obturator vessels and nerve, facilitating OH formation [44, 89]. Considering that most of the aforementioned predisposing factors affect elderly, emaciated, and multiparous women, it is easily conceivable why OH are frequently characterized as “little old lady’s hernias”.

As previously described, OH have been reported to be six-to-nine times more frequent in females than in males [7, 90, 91]. Further highlighting these data, our study demonstrated that an overwhelming predominance of females over males does exist in symptomatic OH development, as 97.9% of the patients included were females, absolutely justifying their characterization as “lady’s hernia”. Plausible explanations of this phenomenon may be the fact that most of the aforementioned predisposing factors occur in females than in males, as discussed earlier. One step further, the mean BMI of the present series of patients is far lower than the lower normal limit of 25 kg/m2 and even lower than 18.5 kg/m2 [92], classifying these patients as underweight. This is in accordance with multiple publications of Asian origin regarding patients with a lean body build [3, 10, 83, 93, 94], confirming the association between emaciation and OH emergence published in the literature.

Considering OH localization, it has been reported that OH are more frequently formed on the right side, as the presence of the sigmoid colon on the left side is postulated to be a protective factor against OH development [1, 3, 7, 10, 93, 94]. Furthermore, the incidence of an occult contralateral OH in the published literature varies widely, with published frequencies of 6% [95], 50% [62], and 63% [96], depending highly on the individual surgical approach chosen for peritoneal exploration and OH repair. Our study demonstrated that an occult contralateral OH was intraoperatively discovered in 4.7% of patients, a prevalence which is lower compared with all published series so far. A possible explanation of this phenomenon may be the fact that the accumulated experience in advanced laparoscopic surgery in recent years has led to a broader implementation of the totally extraperitoneal approach for OH repair, which, despite its several advantages in terms of postoperative morbidity rate, is associated with a limited ability of peritoneal exploration and, therefore, occult contralateral OH identification.

Obturator hernia is associated with non-specific symptoms and signs, which render its clinical diagnosis difficult and require a high level of suspicion, especially in the aforementioned group of patients with high-risk characteristics. The most common presenting symptom is acute mechanical bowel obstruction, which is usually consistent, but sometimes may be intermittent, if the hernia content is spontaneously reduced within the peritoneal cavity. Intermittent abdominal pain was observed in about one-third of OH patients in another study, highlighting that such an intermittent relief of bowel obstruction may be an important clue for its diagnosis [97]. One step further, a higher occurrence of Richter hernias (41–100%) protruding into the obturator canal and usually spontaneously reduced has been observed in such cases of intermittent intestinal obstruction [5, 6, 97]. Abdominal pain is the second most common presenting symptom of OH, whereas the Howship–Romberg sign, which is pathognomonic for OH, was found to be present in approximately half of our patients. This finding is also supported by the other studies [1, 2], leading to the conclusion that the absence of the Howship–Romberg sign cannot safely preclude OH presence. Moreover, the presence of that sign may be obscured by other patient’s comorbidities, such as spinal, knee and/or hip arthropathy as well as by the overwhelming severity of the concurrent abdominal pain, thus misguiding the involved clinician. It should be also mentioned that fever is not a common sign in OH patients, with only 10% of patients being febrile at their presentation.

The masquerading clinical appearance of OH requires the implication of sensitive and specific imaging studies for its prompt and timely diagnosis. A delayed diagnosis with subsequent delayed surgical treatment in OH cases with mechanical bowel obstruction may lead to intestinal strangulation, compromising intestinal viability [98, 99], thus adversely affecting morbidity and mortality rate as well as LOS [100]. Such a scenario of delayed diagnosis with its consequences may explain the significantly increased LOS in OH patients who were preoperatively submitted to plain abdominal radiographs as definitive imaging study. On the contrary, CT scan of the abdomen and pelvis has been demonstrated as the most sensitive and specific imaging modality for OH detection. Its usefulness was first reported in 1983 [11], with its accuracy being up to 85% [12], still outperforming all other imaging studies. In addition to the aforementioned diagnostic ability, the wider adoption of CT scan, always in conjunction with a detailed physical examination, may allow for a more effective preoperative diagnosis of occult hernias, especially occult contralateral OH, therefore enabling a simultaneous surgical correction and avoiding the risks of an undiagnosed OH. Despite its proven diagnostic efficacy, the association between CT results and patient prognosis was questioned in the literature [10, 94]. Although there are studies reporting that early surgical intervention is not sufficient by itself in reducing postoperative mortality rates [91], the results of the present study demonstrate that the accurate preoperative OH diagnosis by CT scan is associated with a significantly reduced postoperative mortality rate. This finding cannot be solely attributed to a timely surgical intervention, since an accurate preoperative OH diagnosis allows also for a more directed surgical operation, instead of an exploratory laparotomy. More studies with larger number of patients will undoubtedly shed more light on this topic.

Regardless of the operative approach chosen, the presence of strangulation necessitates resection of the affected bowel segment, with the previous reports having noticed that the presence of bowel strangulation negatively affects morbidity and mortality rates [100]. These data are in accordance with our results, which demonstrate that the performance of bowel resection is a statistically significant risk factor for increased postoperative morbidity and mortality rate as well as an increased LOS, further highlighting the importance of timely diagnosis and surgical intervention.

Several operative approaches and surgical techniques for OH reduction and hernia defect closure have been reported in the literature. Regarding operative approaches, intraperitoneal as well as totally extraperitoneal approaches have demonstrated successful outcomes in OH management. Intraperitoneal approach includes the traditional open laparotomy, usually via an infra-umbilical midline incision, or the laparoscopic transabdominal preperitoneal approach (TAPP) [62, 83, 101,102,103], whereas the extraperitoneal approach consists of either the trans-inguinal approach [64] or the TEP approach [90]. The major advantage of intraperitoneal approach is the ability to inspect the abdominal viscera as well as evaluate and treat a possible bowel strangulation, making it more suitable for emergent cases. Extraperitoneal approaches are favored in more elective OH cases, provided that the presence of a possible bowel necrosis can be ruled out with great accuracy preoperatively. In such cases, their advantages over intraperitoneal approaches, such as reduced rate of postoperative ileus, adhesion formation, and risk of injury to other abdominal viscera [104, 105] as well as the decreased LOS evidenced by the present study by far outweigh their limited ability of abdominal exploration. Therefore, the decision of performing either an intraperitoneal or an extraperitoneal approach is primarily based on the preoperative suspicion for bowel strangulation and the subsequent need for an enterectomy and secondarily on surgeon’s familiarity and experience with either approach.

Various techniques have been described for closure of the hernia defect following reduction of hernia contents within the abdominal cavity. Apart from the traditional primary peritoneal closure with the use of stitches, other techniques include reconstruction using a tissue flap made from aponeurosis and periosteum and, finally, the use of permanent prosthesis, such as a surgical mesh or plug, aiming to a defect closure without tension [4, 90, 102, 106,107,108]. Each technique has its own indications, advantages, and disadvantages. Primary peritoneal closure with stiches is the method of choice in small OH, in which primary closure may be achieved without jeopardizing hernia recurrence due to extensive tension as well as in cases of a contaminated surgical field due to intestinal resection [62, 81, 109, 110]. On the contrary, the presence of a concurrent groin hernia also favors the use of a large mesh for covering all hernia orifices at the same surgical operation. According to our results, the use of mesh for OH correction is inversely associated with the perioperative morbidity rate as well as LOS, a fact which may be attributed to the demonstrated more frequent application of mesh in non-contaminated cases, such as cases without need for bowel resection, which are associated with fewer postoperative complications than their contaminated counterparts.

Irrespective of the intraperitoneal or the extraperitoneal approach and the utilization or not of a mesh for hernia defect closure, the laparoscopic approach has several advantages over the open approach, which include decreased postoperative pain, early mobilization, and shorter LOS, as well as decreased postoperative morbidity rates [7, 111, 112]. The present study demonstrated that 42 patients underwent a laparoscopic OH repair, with less than 10% of them having been converted to a laparotomy because of technical difficulties, with the laparoscopic repair demonstrating a significantly reduced postoperative morbidity rate, compared with the respective open repair. Furthermore, the mean LOS was significantly decreased in the laparoscopic repair group of patients, compared with the open group. Therefore, based on the present outcomes as well as current literature, laparoscopic approach seems to be the optimum treatment strategy for OH patients, especially in centers with broad experience in advanced laparoscopic surgery.

The present study has certain limitations. First of all, it is a retrospective study, without good uniformity between the included studies, regarding the method of reporting patients’ data, accompanied by a significant percentage of missing data across these studies. Moreover, to date, there is paucity of data regarding the laparoscopic repair of the OH, especially the extraperitoneal approach (TEP). Nevertheless, this date already seems to demonstrate the significant advantages of the laparoscopic approach over the open repair, but the addition of more data leading to safer conclusions is definitely required.

In conclusion, despite their overall low incidence, OH tend to occur more frequently in elderly, slender, multiparous women, with the vast majority of patients presenting with mechanical bowel obstruction. They are not associated with specific symptoms and signs and are usually manifested with consistent abdominal pain. The pathognomonic Howship–Romberg sign is recognized in approximately half of the patients, implying that its absence cannot safely preclude OH presence. Timely and accurate diagnosis is the cornerstone of minimizing morbidity and mortality, and may be achieved by wide implementation of computed tomography, which remains the gold standard. Their operative repair is mandatory, since they are associated with high levels of bowel incarceration and strangulation, if left untreated, with most patients being operated on in an emergent setting. Their operative repair is accompanied by acceptable LOS as well as postoperative morbidity and mortality rates, with the laparoscopic approach demonstrating significant advantages over the open repair, thus seeming to be the optimum treatment strategy for these patients.