In adult males, one of the most common conditions requiring surgical intervention is an inguinal hernia. Inguinal hernia repair is one of the most frequently carried out surgeries, with over 20 million inguinal hernioplasties worldwide each year [1, 2]. In Brazil, approximately 300,000 surgeries for hernial repair are carried out each year [2, 3] and laparoscopic techniques are employed in about 30% of cases worldwide [1, 2], and in Brazil, approximately 0.6% of cases are carried out by the public health system [4].

The primary purpose of this study is to evaluate the learning curve for TEP hernioplasty, analysing parameters related to the mean operating time. As a secondary objective, the complications and the rate of early discharge related to the use of TEP hernioplasty will be assessed.

In 1992, Dulucq developed a totally extraperitoneal inguinal repair technique (TEP), successful at its first attempt [5, 6], which has been used, studied and improved, revolutionising the repair of inguinal hernia [7]. TEP repair takes into account the Lichtenstein technique (tension-free repair) and the pre-peritoneal approach advocated by Stoppa [6, 8].

Currently, the laparoscopic method of hernia repair has gained more attention than open surgery(Lichtenstein) [3, 9, 10]. This is due to a series of advantages already demonstrated by Putnis et al. including the following reduced nerve damage and hematoma formation, lower incidence of acute and chronic pain, reduced need for analgesics, lower risk of infection of the surgical site and quicker recovery, the possibility of bilateral repairs with less surgical trauma, improved aesthetic results, diagnosis and repair of femoral and obturator hernias [8, 10–12].

However, this method presents several disadvantages, such as longer surgery time to reach the plateau of the learning curve, higher risk of serious complications and more complex surgical techniques (knowledge of pelvic anatomy and field of surgery reduced) [3, 12–14], which can lead to more accidents and harm to patients during the learning process.

The performance of a surgeon in this process can be evaluated through learning curves. The concept of the learning curve was first described by German psychologist Hermann Ebbinghaus [15, 16]. Following the Bristol Royal Infirmary Inquiry in 1998, the learning curve has taken on vital importance in clinical practice. The inquiry centred around the investigation of the deaths of 30 infants submitted to cardiac surgery by insufficiently experienced surgeons [17, 18]. The concept of the learning curve predicts that a surgical procedure will become safer and more efficient the more a surgeon performs it, be it under observation or in simulation. This evolution can been seen graphically when inexperienced physicians are in the initial phase of this curve, reaching a level of improved individual performance. The clinical importance of this type of study is to determine the point in the learning curve where the surgeon is capable of carrying out a procedure efficiently, safely and independently [16, 18] within an adequate operation time.

The literature on the learning curve of TEP inguinal repair is scarce, and results do not permit a definitive conclusion, because the numbers vary from 20 to over 250 repetitions before the surgeon reaches a level of competency [5]. For example, Lim JW et al. suggest that this is reached after 60 repetitions [19]. According to the European Guidelines on Hernia Repair 2009, the first 30–50 surgeries performed have the largest impact on the curve and the final results did not differ between medical residents and fully qualified surgeons [20].

Methodology

A prospective study was conducted of patients undergoing TEP inguinal hernioplasty by one surgeon, between May 2009 and May 2014 in a specialised centre for abdominal hernias in Sao Paulo, Brazil. The surgeon had six years surgical experience of laparoscopy, and thereafter carried out training in advanced laparoscopic surgery prior to the period studied, participating in TEP and TAPP laparoscopic hernia correction only as first assistant.

The surgeon selected only patients with small hernias of low complexity, avoiding scrotal components. A polypropylene mesh was used (weight 120+/− 10 g/m2, pores 0.9 +/− 0.1 mm) and a size of 15 cm × 15 cm in all cases. All cases were operated on using the same technique [6].

Data were collected on age, sex, body mass index (BMI), operating time, early hospital discharge, hernia type and complications.

The level of operator competency was established by the operating time and frequency of complications.

The project was approved by the ethics committee. All patients in the study gave informed consent.

Statistical analysis

Quantitative variables are presented as mean, median, standard deviation, minimum, maximum and interquartile range. Qualitative variables are presented as absolute frequencies and percentages.

To compare groups, the Wilcoxon test was used. To compare variance, the F test was used.

Analysis was carried out using the KPSS test (Kwiatkowski–Phillips–Schmidt–Shin test) to verify whether the series is stationary to the trend. If it is not, the calculation is repeated without the first patient included and repeated until the variation is no longer statistically significant. From here, the model turns stationary, or the series develops around a mean with a consistent variance.

The level of significance adopted was 5%.

Results

The mean BMI was 25.83 (indicating an overweight group, according to the WHO) with a standard deviation of 3.34.

This comparison of groups was carried out to verify whether bilateral hernias or complications could be considered confounding factors for the stationary series.

Both bilaterality and complications can influence the operating time; so, these cases should be excluded from the calculation of the curve. Only patients with unilateral surgery and without complications were included.

Dividing the data into two groups: from the surgery on patient at time i. Therefore, there will be a group undergoing surgery prior to patient at time i and another undergoing surgery after this time. We can then verify whether the variance is constant over the whole period. Below is the graph of the results (Graphs 1, 2).

Fig. 1
figure 1

Graph 1 Operative time analysis along the learning curve using standard deviation. S.D. = Standard deviation

Full size image
Fig. 2
figure 2

Graph 2 Analysis of operating time throughout the learning curve using P-values

Full size image

We can observe that in the beginning, the standard deviation is 14.47. on the 65th patient, there is a reduction in the standard deviation to 11.77 and shortly thereafter to 9.43. This shows a statistical significance between the groups; between the first 65 cases and those after.

The values from the KPSS test for a null hypothesis of a trend-stationary series, with the lines being critical points with their statistical significances are shown in Graph 3.

Fig. 3
figure 3

Graph 3 Analysis of operating time over the learning curve using the KPSS test

Full size image

Some points achieve trend-stationary status early in the series, but not in a consistent manner. The series is more stationary after the 65th patient operated.

Analysis of groups of operated patients

After bilateral hernias were excluded from the analysis, the study included 137 patients undergoing a total of 157 TEP hernia repairs. In this study, the sample of patients was mostly non obese and the mean age was 51.56 years (Table 1). The majority were male (97.1%) and 57.6% (137 cases) of all procedures were unilateral hernioplasties (Table 2). Mean operating time of unilateral surgeries was 37.82 min with standard deviation of 17.88 (Table 3). Procedures that evolved with complications had shorter mean operating time (Table 4). With this sample of patients, we perceived that the operating time was practically the same, statistically speaking. The patients were chronologically divided into four groups. Groups 1, 2 and 3 were composed of 35 patients and group 4 of 32. There were no significant variations in clinical characteristics between the groups (Table 5). The mean operating time of the surgeries was 43.7 min (Group 1), 37.8 min (Group 2), 34.7 min (Group 3) and 34.8 min (Group 4). The mean operating time reached a plateau after 65 operations (Graph 3). Below is the comparison of mean operating time between groups (Graph 4). Complications were only reported in the first group: seroma (incarcerated hernia) (1 case), recurrence (1 case), scrotal hematoma (2 cases), umbilical and scrotal hematoma (1 case), hematuria without identifiable lesions on cystography and subcutaneous emphysema (1 case).

Table 1 Quantitative variables (BMI* and age)
Full size table
Table 2 Categorical variables (sex, bilateral hernias and complications)
Full size table
Table 3 Analysis of bilateral hernia repair related to operating time
Full size table
Table 4 Analysis of complications in relation to operating time
Full size table
Table 5 Clinical characteristics of the studied groups
Full size table
Fig. 4
figure 4

Graph 4 Comparison of mean operating time between groups

Full size image

Discussion

The laparoscopic procedure requires greater technical ability and deep tissue dissection; therefore, inexperienced surgeons require a longer time learning and the procedure presents a higher rate of complications, as well as high rates of recurrence. This difficulty is related to unfamiliarity with the pelvic anatomy and the limited field of operation [8, 13, 14].

However, beyond the difficulty of the surgery itself, there are some clinical factors of patients that can also complicate and in some cases contraindicate the use of TEP for inguinal hernias. These are: obesity (BMI > 35), patients under 18 years old, pregnancy, patients categorised as high surgical risk (ASA 3 or 4), presence of large inguinal–scrotal hernias or strangulated hernias, patients in anticoagulation treatment or with a history of surgery in the pre-peritoneal area [12, 13, 21].

The rate of complications and speed of learning TEP hernioplasty can be improved by establishing patient selection criteria, principally for surgeons still learning the procedure. For example, the approach for patients with a BMI > 25 is associated with a higher rate of complications and higher technical difficulty. In these patients, correction of inguinal hernias tends to be more complex because of the higher intra-abdominal pressure, and the reduced surgical field caused by a larger percentage of intra-abdominal fat. These patients are also more prone to bleeding. Despite the need for larger studies, this could be a method to positively impact the learning curve of TEP repair [21–23].

The learning curve, as well as being a tool to evaluate the competency of a surgeon [16], can also aid comparison of new techniques with existing ones, and can alter guidelines in certain situations. For example, the laparoscopic technique is recommended when there are bilateral inguinal hernias and recurrence following open surgery.

Furthermore, the mean surgery time will differ from surgeon to surgeon, and regardless of a surgeon’s time of arrival at the plateau of the learning curve. There is evidence of operating time, conversion rate and short term complication rate continuing to reduce, even after 400 cases [5, 24]. Surgeons carrying out the procedure frequently will have different learning curves to surgeons who operate on hernias only sporadically and medical residents, as well as different complication rates [25].

It is still a challenge to establish the point on the learning curve at which a surgeon will see a reduction in operating time and the rate of complications. This series of studies chronologically divided 137 patients who underwent TEP inguinal hernioplasty into four groups. 157 TEP inguinal hernioplasties were carried out, a much larger samples than many previous studies [5, 9], although it is still fewer than the 700 patients included in Choi et al. [19]. The operating time fell gradually, and the curve established itself after the 65th patient, while other studies suggest that this happens between 40 and 60 cases [9, 19]. The mean operating time after reaching the plateau of the learning curve was shorter than that of some other studies which showed results of 55.1 and 52.8 min [5, 9]. There are other studies, however, where the mean operating time reached 28 min, lower than the 35 min of surgery in this study [19].

Complications were observed only in Group 1. The four patients who were not discharged in under 12 h remained interned for clinical improvement because three had hematomas, and one hematuria without identifiable lesions on cystography. There was one case of recurrence, which was reoperated using the Lichtenstein technique. With the progression of the number of procedures carried out, there was a fall in mean operating time and, after Group 3, statistically, the plateau had been reached. This is associated with an improvement in surgical technique as the procedure is repeated.

It is worth reiterating that the number of cases with late discharge was not statistically significant for the study; however, when combined with the low rate of complications, this suggests that the procedures were therapeutically effective. Early discharge was considered as a patient returning home in less than 12 h after surgery. The time of hospital discharge less than 12 h post surgery for 97% of patients was lower than in other comparable studies [5, 9, 19].

It is worth noting that this study was carried out by one surgeon, and therefore the technical ability may vary between surgeons. The study was also carried out on one relatively homogenous population in the same centre. Multi-centre studies comparing a number of surgeons and socioeconomically heterogeneous populations should be carried out, bringing a more thorough and precise analysis of the evolution of the learning curve.

In our opinion, the establishment of goals, measurement of progress in each operation, the role of a proctor and the constant study of the technique are the most important factors in optimising the learning curve.

Conclusion

A reduction and then stabilisation of the mean operating time was identified after the 65th repetition of TEP inguinal hernioplasty. A reduction in the rate of complications was observed after the 35th patient operated upon. The majority (97%) of patients were discharged early, with a recovery time of less than 12 h.

Selecting patients with non-complex hernias is vital for the surgeon to learn the technique at the beginning of the learning curve.

Our findings on the relationship between obesity and complexity of surgery warrant further investigation to establish clinical guidelines.