Learning curve in the arthroscopic Latarjet procedure: An analysis of the first 171 cases

Abstract

Background

Arthroscopic Latarjet procedure is accepted as effective and safe. However, it is more likely to be challenging in execution. Therefore, the analysis of the learning curve remains to be worth examination.

Objective

To examine the learning curve in the arthroscopic Latarjet procedure through an analysis of the correlation between the average operative time, complications rate, and the total number of performed surgeries and execution frequency.

Materials and method

A total of 171 patients who underwent arthroscopic Latarjet surgery between 2013 and 2020 were included. Clinical findings were recorded postoperatively. We fixed the operative time with account taken of the execution frequency and the number of intra- and postoperative complications. All procedures were completed by one surgeon.

Results

By statistical analysis, a significant association between the operative time and the number of surgical interventions was found (p < 0.05). After 120 procedures, the average operative time demonstrated a constant reduction and reached 62.8 min (p < 0.05). We detected the relation between the average operative time and surgery frequency. The significant correlation between the number of intraoperative complications and the number of procedures performed manifests after the 20th surgery (p < 0.05). In total, 13 episodes of complications were recorded (7.6 %). 9 cases of intraoperative complications (5.3 %) conversed into the open Latarjet procedure. 4 episodes of postoperative complications (2.3 %) were documented, among them 2 hematomas which were treated conservatively, and 2 cases of recurrent shoulder instability. There were no neurological, neurovascular, and infection complications seen in our cohort.

Discussion

It took more than 120 arthroscopic procedures to reach the stable and reproducible indices in operative time, and more than 20 surgeries to achieve a significant reduction in the intraoperative complications rate. There was no correlation between the postoperative complication rate and the number of performed procedures. The overall level of complications did not exceed the level mentioned in the literature. The high surgical activity could be considered as a favorable factor to increase the efficiency of the learning process.

Conclusion

Gaining clinical experience with the passing of the learning stages impacts significantly the rate of intraoperative complications and operative time. The arthroscopic Latarjet procedure is safe and characterized with good clinical results even within the initial stages of the learning curve.

1. Background

To choose reasonably the surgical treatment of anterior shoulder instability means the assessment of its clinical results and reproducibility. The learning curve illustrates clearly the potential of the approach to be reproduced. In surgical practice, the learning curve is described as the number of procedures needed to achieve stable outcomes, whilst the significant and valid reduction in operative time^1,2 and total complications rate is recognized as the marker.

In a general way, the learning curve consists of 4 phases, as mentioned by A.N. Hopper et al.: (1) the relatively rapid improvements in performing at the first parts of the curve, then (2) the level of excellence gained with each surgery is reducing, (3) phase, mentioned as competency in performance, within the additional experience improves outcomes in a minor way, (4) reduction of the performance indices, associated with the age.³ Reaching the 3rd phase is supposed as surgical skills are reinforced, reflecting the clinical performance, competency and excellence.

The analysis of the learning curve has crucial clinical and system input. It is connected with risks to the patient, and this thesis is emphasized by the General medical council of Great Britain in Bristol.⁴ The considerations regarding the learning curve are meaningful by surgeon training development.⁵ Finally, under the circumstances of progressively rising health care costs, the learning curve is becoming a criterion to assess the economical efficiency and optimal recourses allocation in the hospital.⁶

In the literature, the learning curve in arthroscopic Latarjet procedure is a topic of careful consideration. The arthroscopic approach is one of the trends in surgical treatment concepts in anterior shoulder instability. The arthroscopic Latarjet procedure combines the advantages of both open and minimally invasive approaches, namely:

• •less surgical aggression,
• •as a result of optimized visualization - more accurate graft positioning, which is the basis for the reconstruction of anatomical interaction between humeral head and glenoid,
• •lower risk of postoperative contracture and blood supply preservation due to the limited soft tissue-damaging,
• •faster functional recovery,
• •options to treat the collateral injuries at the same step, simultaneously,
• •lower early pain syndrome,
• •cosmetic benefits.7, 8, 9, 10

However, the limitation of performing these procedures routinely is the technical challenge.11, 12, 13 Consequently, the learning curve in arthroscopic Latarjet is the object to examine.

The key conclusion of review from S. Ekhtiari et al. is that 20–40 arthroscopic Latarjet procedures were needed to perform to get the competencies, conveyed in operative time reduction.¹⁴ This point is consistent with the conclusions from other researchers.15, 16, 17, 18, 19 In this context, E. M. Valsamis et al. claimed the high surgical activity as a positive factor to master the operation approach.¹⁵ The evidence regarding the average operative time varies slightly. According to G. Cunningham et al., the average time of arthroscopic Latarjet procedure is 146 ± 51 min.¹⁸ Whilst S. Ekhtiari et al. pointed the average time of 138.7 min (103–183 min) within the first phases of the learning curve, reducing to 108.8 min (76–139 min) on later stages of learning.¹⁴ According to R. Сastricini et al., the very first 15 procedures took 132 min on average to perform, thereafter declined to 99 min.¹⁹ B. Kordasiewicz et al. estimated the average time in 112.7 min.¹⁷ The findings from B. Kordasiewicz et al. and G. Cunningham et al. showed the relatively high postoperative complications rate and the high number of conversions into open Latarjet procedure during the first phases of the learning curve.^17,18

This paper aims to analyze the individual learning curve in arthroscopic Latarjet procedure and to assess the correlation between the operative time, the complication rate with the total number of the performed surgeries, and time intervals between them.

So, based on the consideration of the foregoing premises we could hypnotize that 30 procedures are required to achieve the significant stable reduction in the average operative time, which is estimated in 117 min, characterizing the reproducible and stable average operative time (calculated as arithmetic average point according to the literature data). The second hypothesis is that the reduction of intra- and postoperative complications will be expected. Similarly, the time gap between the surgeries is expected to impact the speed and efficiency of the learning process, which could be observed in the average operative time.

2. Materials and methods

In this paper, we examine the data of 171 arthroscopic Latarjet procedures from 2013 till 2020, which were performed in the Nikiforov Russian Center of Emergency and Radiation Medicine (St Petersburg, Russia). All the surgeries included were performed by the same operator with the same instrumentation.

The inclusion criteria were: patients elder than 18 years, with recurrent shoulder instability, with clinical indications to arthroscopic Latarjet procedure. All the patients were informed properly and agreed to be enrolled in this study before the date of the surgery.

We collected the following data: demographics, clinical card, surgical details. We assessed the number of adverse events and the operation time (stated in the surgical report and the anesthesia records), retrospectively. The operative time was defined as the period from the incision to the suturing. Based on the calculated average operative time depending on the chronological numerical order we plot the graph curve, which demonstrated the development in the operative duration by the phases of the learning process. To examine the effect of gained clinical experience on the average operative time and complications rate, the surgeries were divided into 9 groups, each of them included 20 procedures separated by chronological order. The time gap between the surgeries within each group was assessed as well. We analyzed the adverse events by analyzing the intraoperative complications which caused the conversion into the open Latarjet procedure, and postoperative complications, which were specified as infection, neurological and vascular complications, and recurrent shoulder instability. The episodes of recurrent shoulder instability were registered postoperatively by the checkup examination or telephone interview.

The statistical processing was performed by the software Excel (Microsoft) and Statistica. The quantitative measures were described by the mean value, standard mean square error, and the maximal and the minimal value. We used Lilliefors' test and the Shapiro-Wilk test for defining the normal distribution. The quantitative measures were compared using the Fisher variance test, Tukey's honestly significant difference test, and the Newman-Keuls multiple comparisons for post hoc comparisons. By analyzing multi-way tables we used Pearson's chi-squared test and the two-tailed Fisher's exact test. The correlation between the quantitative measures with nonnormal distribution was analyzed by Spearman's rank correlation coefficient. The significance threshold was 0.05.

The arthroscopic Latarjet procedure was performed according to L. Lafosse et al., with minor author's modifications.⁸

3. Results

171 arthroscopic Latarjet procedures were analyzed. The mean age of the patients was 31.4 years + -0.7 (18–61 years), 94.2 % (161 patients) were male. The average operative time within the study was 89.6 min (+-2.9). A significant association between the operative time and the number of surgical interventions was found (p < 0.05). After the 120th procedure, the average operative time demonstrated a constant reduction and reached 62.8 min + -2.2 (p < 0.05). Fig. 1 illustrates the results of the learning curve study in the context of the operative time concerning the total number of surgeries and reflects the skill acquisition tempo. The conversion cases are highlighted by color-coding.

Fig. 1.

Dynamics of the operative time concerning the total number of the arthroscopic Latarjet procedures.

The average, the maximal, and the minimal operative time within the groups, selected by 20 surgeries and ordered in chronological order, are presented in Table 1. So, the average operative time in the Group 1 is 159 min (+-8.7), the maximal time reaches 230 min, the minimal time is 90 min. At the same time, the average operative time in the Group 8 is 59 min (+-2.5), the maximal time peaks 80 min, and the minimal operative time drops to 35 min. The lowering of the operative time by comparing the results between Group 1 and Group 8 is 100 min, or 63 % (59 vs 159 min).

Table 1.

The average operative time of the arthroscopic Latarjet procedure in the groups by 20 episodes, ordered in the chronological order.

Group, №	Numerical order of surgeries	Number of surgeries	Average operative time (min)	Standard mean square error	Minimal operative time (min)	Maximum operative time (min)
1	0–20	20	158,8	8,67	90	230
2	21–40	20	110,5	7,82	65	225
3	41–60	20	91,5	5,11	50	140
4	61–80	20	85,0	5,09	45	145
5	81–100	20	82,3	5,42	60	140
6	101–120	20	79,0	3,13	60	110
7	121–140	20	64,0	3,45	40	95
8	141–160	20	59,3	2,49	35	80
9	161–171	11	65,0	6,61	30	100
Results		171	89,6	2,90	30	230

Based on the analysis results of the Newman-Keuls multiple comparisons for post hoc comparisons, we could assume, that Group 1 and Group 2 (1–40 procedures) may be characterized as skills shaping phase, and the sharp changes in the operative time at the start of the learning process are seen. Group 3 (41–60 surgeries) may be described as the transitional period. Groups 4, 5, 6 (61–120 cases) exhibit diminishing returns in the experience gaining. In Group 7 (after the 120th surgery) we could see the plateau phase, within the additional experience improves outcomes in a minor way (р<0.05). The data is presented in Table 2.

Table 2.

The Newman-Keuls multiple comparisons, p-value.

	{1} M = 158.75	{2} M = 110.50	{3} M = 91.500	{4} M = 85.000	{5} M = 82.250	{6} M = 79.000	{7} M = 64.000	{8} M = 59.250	{9} M = 65.000
1{1}		0,000	0,000	0,000	0,000	0,000	0,000	0,000	0,000
2 {2}	0,000		0,019	0,005	0,003	0,001	0,000	0,000	0,000
3{3}	0,000	0,019		0,424	0,491	0,415	0,009	0,001	0,010
4 {4}	0,000	0,005	0,424		0,735	0,741	0,073	0,019	0,066
5{5}	0,000	0,003	0,491	0,735		0,689	0,111	0,038	0,085
6 {6}	0,000	0,001	0,415	0,741	0,689		0,155	0,072	0,085
7{7}	0,000	0,000	0,009	0,073	0,111	0,155		0,559	0,902
8{8}	0,000	0,000	0,001	0,019	0,038	0,072	0,559		0,759
9{9}	0,000	0,000	0,010	0,066	0,085	0,085	0,902	0,759

р-value <0.05.

Fig. 2 shows the development in the average operative time according to the learning curve concept of A.N. Hopper et al.³ According to our analysis, we mark the transitional period of the learning process, which specifies more precisely the characteristics of the individual learning curve in our case.

Fig. 2.

The development in the average operative time according to the learning curve phases.

It seems to be meaningful the two-fold lowering of the operative time if comparing two reference stages – the initial learning phase and the plateau phase. It is 71.9 min, or 53.4 % (134 min vs 62.8 correspondingly). In Table 3 we present the average, the maximal, and the minimum intervals between the surgeries, ordered by the groups. So, the average frequency rate in this study is 15.9 days (+-2.8), the maximal interval is 371 days, the minimum interval is 0 days.

Table 3.

The time interval between the arthroscopic Latarjet procedures in the groups by 20 episodes, ordered in the chronological order.

Group, №	Numerical order of surgeries	Number of surgeries	Average interval between operations (days)	Standard mean square error	Minimal interval between operations (days)	Maximal interval between operations (days)
1	0–20	20	64,1	20,8	0	371
2	21–40	20	15,6	4,4	1	74
3	41–60	20	8,2	1,9	0	33
4	61–80	20	7,4	1,8	0	30
5	81–100	20	5,6	1,4	0	19
6	101–120	20	11,2	2,0	0	36
7	121–140	20	10,1	2,8	0	49
8	141–160	20	12,6	5,9	1	119
9	161–171	11	6,5	2,0	1	21
Results		171	15,9	2,8	0	371

The analysis by the Spearman's rank correlation coefficient (rs = 0.455) gives evidence of the influencing variable factor of regular surgical performance, its equitability on the increase of the operative time in Group 1 and Group 2 only (1–40 procedures). During the first 40 surgeries, the average time gap is 39.2 days (+- 10.99), which is 2.5 times more than such value of 8.8 days (+-1.15) in the latest Groups (41–171 surgeries), where the Spearman's rank correlation coefficient, describing the correlation between the operative time and the time interval, is non-significant (rs = 0), the linear regression is non-significant, too. The statistical data is given in Fig. 3, Fig. 4. We exclude one outlying observation (time gap was 371 days), which was happened in Group 1–2.

Fig. 3.

The correlation between the operative time and the time interval between the procedures in Group 1–2

Days after the previous surgery: time of operation, min: r = −0.1013; p = 0.2498.

Fig. 4.

The correlation between the operative time and the time interval between the procedures in Group 3-9.

We revealed 9 episodes, or 5.3 %, of the intraoperative adverse events, which led to the conversion in the open Latarjet procedure. A significant difference in the number of conversions during the initial learning phase (1–20 surgeries) versus the latest stages is evident (р<0.05). All the conversions are caused by the split coracoid process happened intraoperatively.

The follow-up period is 84 months; the total number of postoperative complications is 4, or 2.3 % of total performed procedures in this study. Among them, we fix 2 postoperative hematomas (or 1.2 %) during the initial stages of learning (in Groups 1–2), which were treated conservatively, and 2 cases of recurrent shoulder instability (or 1.2 %) in Groups 6–7. There were no neurological, vascular, and infection complications seen in our cohort. As a result, 9 cases of intra- and postoperative complications, or 69.2 % of the total number of complications, are registered during the initial stages of the learning process, in Groups 1–2. The overall number of complications is 13, or 7.6 % of total performed procedures in this study. 2 revisions, or 1.2 %, were performed. The structure of recurrent shoulder instability, postoperative complications, and conversions are demonstrated in Table 4.

Table 4.

The structure of the intra- and postoperative complications and conversions into the open Latarjet procedure.

Group №	Numerical order of surgeries	Number of conversions into open Laratjet procedure, by the groups, n	Number of conversions into open Laratjet procedure, % of all surgeries	Distribution of conversions into open Laratjet procedure, % of total conversions	Number of postoperative complications, by the groups, n	Number of postoperative complications, % of all surgeries	Distribution of postoperative complications, % of total postoperative complications	Number of recurrent instability, by the groups, n	Number of recurrent instability, % of all surgeries	Distribution of recurrent instability, % of total recurrent instability	Total number of conversions, postoperative complications, recurrent instability, n	Total number of conversions, postoperative complications, recurrent instability, % of all surgeries	Distribution of conversions, postoperative complications, recurrent instability, % of total adverse events
1	0–20	7	4,1 %	77,8 %	1	0,6 %	50,0 %	0	0,0 %	0,0 %	8	4,7 %	61,5 %
2	21–40	0	0,0 %	0,0 %	1	0,6 %	50,0 %	0	0,0 %	0,0 %	1	0,6 %	7,7 %
3	41–60	0	0,0 %	0,0 %	0	0,0 %	0,0 %	0	0,0 %	0,0 %	0	0,0 %	0,0 %
4	61–80	1	0,6 %	11,1 %	0	0,0 %	0,0 %	0	0,0 %	0,0 %	1	0,6 %	7,7 %
5	81–100	0	0,0 %	0,0 %	0	0,0 %	0,0 %	0	0,0 %	0,0 %	0	0,0 %	0,0 %
6	101–120	1	0,6 %	11,1 %	0	0,0 %	0,0 %	1	0,6 %	50,0 %	2	1,2 %	15,4 %
7	121–140	0	0,0 %	0,0 %	0	0,0 %	0,0 %	1	0,6 %	50,0 %	1	0,6 %	7,7 %
8	141–160	0	0,0 %	0,0 %	0	0,0 %	0,0 %	0	0,0 %	0,0 %	0	0,0 %	0,0 %
9	161–171	0	0,0 %	0,0 %	0	0,0 %	0,0 %	0	0,0 %	0,0 %	0	0,0 %	0,0 %
	Total	9	5,3 %		2	1,2 %		2	1,2 %		13	7,6 %

Fig. 5 shows the development of the individual learning curve in the context of the crucial learning phases – the initial stage and the plateau phase - by the meaningful variables, which are the average operative time with regard to performance frequency and the number of conversions.

Fig. 5.

The development of the average operative time and the number of conversions by the learning phases.

Generally, the learning curve illustrates the correlation between the number of the performed surgeries and significant clinical and organizational values of operative time and complications rate. In our study, the hypothesis about the correlation between the operative time, intraoperative complications rate, and the passing of the learning phases is confirmed. This finding is consistent with S. Ekhtiari et al.: the difference in the operative time during the initial learning stage and the latest ones are recognized as significant in all papers, included in the systematic review.¹⁴ This statement is mentioned by Kordasiewicz et al. as well.¹⁷

We found no prove the hypothesis, that 30 procedures are needed to perform to achieve a significant stable reduction in the average operative time. In our paper, we declare that 120 procedures are required to achieve the optimal stable average operative time. We emphasize achieving the plateau stage, which better marks the surgeon's competence in the reproducible and stable clinical results and risk correction of the adverse events, in comparison to the other learning stages, and more especially to the initial phase. Therefore, we suppose that it is particularly relevant to examine the learning curve on the larger sample to assess reliably the actual requirement in the number of the performed procedures, demanded for the greatest possible clinical safety and efficiency.

It may support and justify the decision to put the arthroscopic Latarjet into clinical practice with consideration to the surgical activity of both the hospital and the surgeon. This corresponds with E. M. Valsamis et al. judging the high regular surgical activity is a positive predictor of relatively better clinical results, including the tempo in mastering the operation approach.¹⁵ We suppose that the regularity rate in surgical performance affects the average operative time as well. In our study, we found a correlation between the average operative time and time intervals. However, this suggestion should be verified by additional examinations.

In this paper, the average operative time after mastering the procedure is lower than in the research papers from G. Cunningham et al.,¹⁸ S. Ekhtiari et al.,¹⁴ B. Kordasiewicz et al.,¹⁷ and R. Castricini et al.¹⁹: 63 min in our cohort versus 146, 109, 113 and 99 min accordingly. So, the hypnosis that the average operative time is 117 min, characterizing the reproducible and stable clinical results, have not been confirmed in our study. The average operative time concerning the total procedures enrolled is 90 min, which is lower than mentioned in the aforementioned literature. This operative time reduction is provided mainly by our modifications in the original surgical technique by L. Lafosse. We diminish the number of arthroscopic portals, eliminate the fixation of the coracoid process by a flexible guide before osteotomy, and change the sequencing of subscapular muscle split forming and coracoids process osteotomy, coracoid process screw fixation without the preliminary reaming of scapula articular process.

The common positive factors favoring the reduction in the average operative time are enhanced cooperation in the surgical team, improvements in preoperative preparation of equipment and instruments, which provide the number and sequencing of manipulations, hence optimize the time spending.

We observe the dynamics of intraoperative complications causing the conversion into open Latarjet procedure during the learning process comparable to findings mentioned in the literature. In our study, the number of intraoperative complications reduces significantly after 20 performed procedures (р<0.05). Thus, we reach the learning phase marked with a stable intraoperative level of efficiency and safety earlier comparing to indices of the average operative time. Remarkably, 20 procedures are demanded to reduce significantly the number of intraoperative adverse events, associated with conversion into open Latarjet procedure, whereas 3 times more surgeries are necessary to perform to accomplish the learning phase in the context of the average operative time. In our study, the arthroscopic Latarjet procedure learning curve may be described with 120 procedures needed to reach the optimal results in the average operative time, and with 20 surgeries demanded to reduce the conversion rate.

In our study, the postoperative complication rate (1.2 %) and recurrent shoulder instability rate (1.2 %) did not exceed the value, mentioned in the systematic review from N.S. Horner et al. - 8 % and 1.9 % respectively.²⁰ We found no significant correlation between the performed surgeries and the postoperative complication rate. To some extent, our results are consistent with findings from mentioned literature. Thus, B. Kordasiewicz et al. point to the relatively higher rate of intra- and postoperative complication rate during the initial learning phases of the arthroscopic Latarjet procedure, which reduces with passing the later learning stages.¹⁷ The authors demonstrate that the number of the intraoperative complications is 5 episodes, or 17 %, in the group of the first 30 surgeries (p = 0,024), in the second group there are no intraoperative adverse events, whilst there are 3 complication cases (10 %) in the 3rd group. All 3 recurrent instability cases (3.3 %) are found in group I (p = 0,033). The revision surgery is required in 10 % of cases.¹⁷ In the study of G. Cunningham et al. the higher conversion rate is found during the initial learning phase comparing to later stages.¹⁸

In our study, the total complication number is lower, not exceeding 8 %. We may interpret the absence of a significant correlation between the complications number with the learning process stage in the other research papers by the sample size.^14,19

Worth further discussion is the issue of the minimally required follow-up period. Therefore, we suppose the long-term results are worth further consideration in the context of learning curve analysis. To our mind, the degree of impact of the following issues regarding the arthroscopic Latarjet surgical technique is the topic of concern: standard and congruent configuration of the articular arch, anterior or posterior reaming of scapula articular process, fixation with buttons or screws, simultaneous treatment of rotator cuff injuries, long head of the biceps pathology or treatment of labrum and capsule tissue injuries.

4. Conclusion

The analysis of the arthroscopic Latarjet learning curve demonstrates a significant reduction in the operative time and in the number of intraoperative complications with an increased number of the performed surgeries and with the lower time interval between the procedures. The arthroscopic Latarjet procedure is safe and clinically justified even within the initial stages of the learning curve, whilst the favorable prognosis is the high surgical activity.

Ethical statement for solid state ionics

Hereby, I Aleksandr Vetoshkin consciously assure that for the manuscript/insert title/the following is fulfilled:

• 1)This material is the authors' own original work, which has not been previously published elsewhere.
• 2)The paper is not currently being considered for publication elsewhere.
• 3)The paper reflects the authors' own research and analysis in a truthful and complete manner.
• 4)The paper properly credits the meaningful contributions of co-authors and co-researchers.
• 5)The results are appropriately placed in the context of prior and existing research.
• 6)All sources used are properly disclosed (correct citation). Literally copying of text must be indicated as such by using quotation marks and giving proper reference.
• 7)All authors have been personally and actively involved in substantial work leading to the paper, and will take public responsibility for its content.

The violation of the Ethical Statement rules may result in severe consequences.

To verify originality, your article may be checked by the originality detection software iThenticate. See also http://www.elsevier.com/editors/plagdetect.

I agree with the above statements and declare that this submission follows the policies of Solid State Ionics as outlined in the Guide for Authors and in the Ethical Statement.

Funding statement

No funding was received for this work.

Patient's consent to publication

This is to state that I give my full permission for the publication, reproduction, broadcast and other use of photographs, recordings and other audio-visual material of myself (including of my face) and textual material (case histories) in all editions of the above-named product and in any other publication (including books, journals, CD-ROMs, online and internet), as well as in any advertising or promotional material for such product or publications.

I declare, in consequence of granting this permission, that I have no claim on ground of breach of confidence or any other ground in any legal system against — (author's/developer's name) — and its agents, publishers, successors and assigns in respect of such use of the photograph(s) and textual material (case histories).

I hereby agree to release and discharge (author's/developer's name), and any editors or other contributors and their agents, publishers, successors and assigns from any and all claims, demands or causes of action that I may now have or may hereafter have for libel, defamation, invasion of privacy, copyright or moral rights or violation of any other rights arising out of or relating to any use of my image or case history.