7
ORIGINAL ARTICLE Learning curve for robotic-assisted laparoscopic rectal cancer surgery Rosa M. Jiménez-Rodríguez & José Manuel Díaz-Pavón & Fernando de la Portilla de Juan & Emilio Prendes-Sillero & Hisnard Cadet Dussort & Javier Padillo Accepted: 21 November 2012 # Springer-Verlag Berlin Heidelberg 2012 Abstract Introduction One of the main uses of robotic assisted ab- dominal surgery is the mesorectal excision in patients with rectal cancer. The aim of the present study is to analyse the learning curve for robotic assisted laparoscopic resection of rectal cancer. Patients and methods We included in our study 43 consec- utive rectal cancer resections (16 females and 27 males) performed from January 2008 through December 2010. Mean age of patients was 66 ± 9.0 years. Surgical procedures included both abdomino-perineal and anterior resections. We analysed the following parameters: demographic data of the patients included in the study, intra- and postoperative data, time taking to set up the robot for operations (set-up or docking time), operative time, intra- and postoperative com- plications, conversion rates and pathological specimen fea- tures. The learning curve was analysed using cumulative sum (CUSUM) methodology. Results The procedures understudied included seven abdomino-perineal resections and 36 anterior resections. In our series of patients, mean robotic set-up time was 62.9± 24.6 min, and the mean operative time was 197.4 ± 44.3 min. Once we applied CUSUM methodology, we obtained two graphs for CUSUM values (operating time and success), both of them showing three well-differentiated phases: phase 1 (the initial 911 cases), phase 2 (the middle 12 cases) and phase 3 (the remaining 2022 cases). Phase 1 represents initial learning; phase 2 plateau represents in- creased competence in the use of the robotic system, and finally, phase 3 represents the period of highest skill or mastery with a reduction in docking time (p 0 0.000), but a slight increase in operative time (p 0 0.007). Conclusion The CUSUM curve shows three phases in the learning and use of robotic assisted rectal cancer surgery which correspond to the phases of initial learning of the technique, consolidation and higher expertise or mastery. The data obtained suggest that the estimated learning curve for robotic assisted rectal cancer surgery is achieved after 2123 cases. Keywords Rectal cancer . Robotic assisted surgery . Learning curve Introduction Rectal cancer is a common entity which represents about one third of large intestine neoplasms with an incidence of 1525/100,000 people in European countries [1]. The spe- cial characteristics of this entity require the intervention of multidisciplinary teams in its management, and the out- comes are closely associated with the technique employed and the skill of the surgeon. Some studies have been published recently on the advan- tages of laparoscopic management of rectal cancer [26]. However, the learning curve for this technique is long, and it entails a gradual process. In fact, a great number of authors claim that minimally invasive surgery for the management of rectal cancer must be carried out by skilled surgeons who master the technique [7]. They also mention the necessity of long training periods [8], even longer than in the case of other procedures as cholecystectomy [9], being 6080 cases R. M. Jiménez-Rodríguez (*) : J. M. Díaz-Pavón : F. de la Portilla de Juan : E. Prendes-Sillero : H. C. Dussort : J. Padillo Department of General Surgery, University Hospital Virgen del Rocío, C/Jándula 2, Bl-3, P-3, 4º2, 41013 Sevilla, Spain e-mail: [email protected] Int J Colorectal Dis DOI 10.1007/s00384-012-1620-6

Learning curve for robotic-assisted laparoscopic rectal cancer surgery

  • Upload
    javier

  • View
    215

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Learning curve for robotic-assisted laparoscopic rectal cancer surgery

ORIGINAL ARTICLE

Learning curve for robotic-assisted laparoscopic rectalcancer surgery

Rosa M. Jiménez-Rodríguez & José Manuel Díaz-Pavón &

Fernando de la Portilla de Juan & Emilio Prendes-Sillero &

Hisnard Cadet Dussort & Javier Padillo

Accepted: 21 November 2012# Springer-Verlag Berlin Heidelberg 2012

AbstractIntroduction One of the main uses of robotic assisted ab-dominal surgery is the mesorectal excision in patients withrectal cancer. The aim of the present study is to analyse thelearning curve for robotic assisted laparoscopic resection ofrectal cancer.Patients and methods We included in our study 43 consec-utive rectal cancer resections (16 females and 27 males)performed from January 2008 through December 2010.Mean age of patients was 66±9.0 years. Surgical proceduresincluded both abdomino-perineal and anterior resections.We analysed the following parameters: demographic dataof the patients included in the study, intra- and postoperativedata, time taking to set up the robot for operations (set-up ordocking time), operative time, intra- and postoperative com-plications, conversion rates and pathological specimen fea-tures. The learning curve was analysed using cumulativesum (CUSUM) methodology.Results The procedures understudied included sevenabdomino-perineal resections and 36 anterior resections. Inour series of patients, mean robotic set-up time was 62.9±24.6 min, and the mean operative time was 197.4±44.3 min.Once we applied CUSUM methodology, we obtained twographs for CUSUM values (operating time and success),both of them showing three well-differentiated phases:phase 1 (the initial 9–11 cases), phase 2 (the middle 12cases) and phase 3 (the remaining 20–22 cases). Phase 1

represents initial learning; phase 2 plateau represents in-creased competence in the use of the robotic system, andfinally, phase 3 represents the period of highest skill ormastery with a reduction in docking time (p00.000), but aslight increase in operative time (p00.007).Conclusion The CUSUM curve shows three phases in thelearning and use of robotic assisted rectal cancer surgery whichcorrespond to the phases of initial learning of the technique,consolidation and higher expertise or mastery. The dataobtained suggest that the estimated learning curve for roboticassisted rectal cancer surgery is achieved after 21–23 cases.

Keywords Rectal cancer . Robotic assisted surgery .

Learning curve

Introduction

Rectal cancer is a common entity which represents aboutone third of large intestine neoplasms with an incidence of15–25/100,000 people in European countries [1]. The spe-cial characteristics of this entity require the intervention ofmultidisciplinary teams in its management, and the out-comes are closely associated with the technique employedand the skill of the surgeon.

Some studies have been published recently on the advan-tages of laparoscopic management of rectal cancer [2–6].However, the learning curve for this technique is long, and itentails a gradual process. In fact, a great number of authorsclaim that minimally invasive surgery for the managementof rectal cancer must be carried out by skilled surgeons whomaster the technique [7]. They also mention the necessity oflong training periods [8], even longer than in the case ofother procedures as cholecystectomy [9], being 60–80 cases

R. M. Jiménez-Rodríguez (*) : J. M. Díaz-Pavón :F. de la Portilla de Juan : E. Prendes-Sillero :H. C. Dussort :J. PadilloDepartment of General Surgery, University Hospital Virgen delRocío, C/Jándula 2, Bl-3, P-3, 4º2,41013 Sevilla, Spaine-mail: [email protected]

Int J Colorectal DisDOI 10.1007/s00384-012-1620-6

Page 2: Learning curve for robotic-assisted laparoscopic rectal cancer surgery

[10], the estimated number of necessary cases per sur-geon to reach the learning curve for laparoscopic resec-tion of rectal cancer.

Robotic assisted surgery, which enriches the possibilitiesof conventional laparoscopic surgery improving vision andthe freedom of movement of instruments, emerges as atechnique aiming at overcoming the limitations posed byrectal cancer and other surgical fields of difficult access, inorder to obtain better outcomes.

The clinical results after surgery and in the short-termfollow-up period obtained by means of this sort of approachhave been described as extrapolated to those obtained bymeans of laparoscopic techniques [11–13]. Yet, the estimat-ed number of necessary procedures to achieve the learningperiod in colorectal pathology could be smaller than in thecase of conventional laparoscopy, considering the advan-tages offered by the robotic system. Nevertheless, onlytwo studies have reported the analysis of the learning curvefor colorectal surgery, and only one of them focuses partic-ularly on rectal cancer surgery [14, 15]. The aim of thepresent study is to evaluate the learning curve for roboticassisted resection of rectal cancer using the cumulative sum(CUSUM) method.

Material and methods

Forty-three consecutive patients diagnosed with rectal can-cer were included in the study. The demographic data in-cluded the age of each patient, sex, BMI, classification ofpatients according to the American Society of Anesthesiol-ogists (ASA) and Charlson comorbidity Index [16].

Intraoperative parameters included robotic set-up time foreach patient, operative time, total operative time, need fortransfusion, intraoperative complications and conversionrates. As regards the postoperative period of patients, weconsidered postoperative complications, need for reinter-vention and postoperative hospital stay.

Operative time was defined as the period of time elapsingfrom pneumoperitoneum to the closure of incisions. There-fore, we have also considered the time required to positionthe robot to perform each of the procedures.

The time required to set up the robotic system and toprepare the patient up to the onset of surgical time was definedas robotic set-up time/docking time. Total operative timeresults from the addition of surgical time and preparation time.

All procedures were carried out by means of a systematicapproach which included ligation of the inferior mesentericartery and vein with mediolateral dissection and routineidentification of the ureter and hypogastric plexus by roboticapproach. Whenever required, the splenic angle was takendown using only robotic techniques. Total or tumour-specific mesorectal excision was performed depending on

the location of the tumour. In the case of anterior resections,anastomosis was carried out intracorporeally with CEEA 28or 31 mm (Covidien, Norwalk, CT, US), as required.

Patients classified as T1 or T2 and N0 did not receiveneoadjuvant therapy. The rest received radio- and chemo-therapy according to the individualised study of each caseby the colorectal cancer committee of our institution.

Patients were chronologically arranged from the earliestto the latest date of surgery. The value of CUSUMoperating

time (OT) was applied to every patient [8, 15, 17, 18] bymeans of the following equation:

CUSUMOT value ¼ C iþ 1ð Þ þ Xi�mean Xið Þwhere Xi represents the operative time employed with eachpatient; C represents each standard deviation of mean time,and i is the case number. The CUSUMOT of the first casewas the difference between the OT for the first case and themean OT for all the cases (IOT). The CUSUMOT of thesecond case was the CUSUMOT of the previous case addedto the difference between the OT of the second case and lOT.The same procedure was repeated for each of the patientsexcept for the last one which was calculated as 0. As none ofthe patients died, it was not necessary to calculate the risk-adjusted CUSUM value for the purposes of the study.

In order to calculate the CUSUM curve to measure “suc-cess” (CUSUMSU), success was defined as any surgicalprocedure which did not show any of the following param-eters or events: conversion to open surgery, intraoperativecomplication, postoperative complication nor mortality. Weapplied the following equation:

CUSUMSU value ðnÞ ¼ Σ xi � Tð Þ þ �1ð Þxipi

where xi represents the presence or absence of the eventunder study; T stands for the observed probability of theevent, in this case, the probability of success, which, in thepresent study, accounts for 65.11 %, and pi is the individualprobability of success in our series once we apply to everyvalue a logistic regression model which meets the require-ments for success. All the patients gave their informedconsent to take part in the study.

Statistical analysis

The learning curve was analysed using the CUSUM method[8, 15]. This graph represents the running total of differ-ences between the individual data points and the mean of alldata points. This methodology was used for all cases, takinginto account the operative time employed in each of themand the probability of success in each intervention comparedto the general probability of the series. Continuous variablesare depicted as mean ± standard deviation; categorical var-iables are depicted as percentage.

Int J Colorectal Dis

Page 3: Learning curve for robotic-assisted laparoscopic rectal cancer surgery

The statistical analysis was carried out using SPSS (IBMSPSS Statistics). Kruskal–Wallis test was used to performinterphase comparisons. Comparisons between phases 1 and2 combined (learning curve) and phase 3 (competence) werecarried out using Student’s t test in the case of independentsamples and Mann–Whitney U test, depending on the nor-mality values of samples (alpha 0.05). Differences werestatistically significant whenever p<0.05.

Results

From January 2008 through December 2010, 43 consecutivepatients (16 women and 27 men) underwent surgery due torectal cancer. Mean age of patients was 66±9.0 years, andtheir mean BMI 27.4±3.9 kg/cm2. The interventions wereperformed by the same team of three surgeons with a longexperience in laparoscopic surgery who had previously re-ceived experimental training to carry out robotic assistedsurgery (J.M.D.P., E.P.S., J.M.H.C.).

The techniques carried out were 36 anterior resectionsand seven abdomino-perineal resections. Demographic andpreoperative characteristics of patients as well as tumourlocation (Fig. 1) and surgical procedures are shown inTable 1. It also includes information as regards mean dock-ing time (62.9±24.6 min; interval 30–130 min) and opera-tive time (197.4±44.3 min; interval 90–330 min), whichaccounts for 24 and 76 % of total operative time.

Intraoperative complications occurred in three patients(6.9 %) in the form of haemorrhage, rectal stump openingand staple failure. Postoperative complications occurred inseven patients (16.2 %), namely, six leakages (13.9 %) andone prolonged paralytic ileus, with a mean total hospital stayof 12.27±10.29 days.

Most resected tumours were classified as T3 (44.1 %),and six of the patients (13.9 %) included in the seriesshowed complete pathological response after analysis ofthe resected piece. The majority of the resected samplesdid not show nodal involvement (62.7 %), and none ofthe patients showed distant metastasis at the time of theintervention.

Once operative times were arranged, we calculatedCUSUMOT values for each of the patients to obtain the graphof the learning curve (Fig. 2). We established two cut-offpoints in the areas of the curve where we observed a changedue to the increase or decrease in operative time. Thus, wecould differentiate three phases in the graph: phase 1 (theinitial nine cases) in which the cut-off point is the point atwhich the operative time decreases until it stabilises andwhich corresponds to the part of the graph resembling adecreasing slope. Phase 2 (the middle 12 cases) correspondsto a plateau phase starting at the end of phase 1 and ending ina cut-off point where graph values begin to increase. Finally,phase 3 (the remaining 21 cases) shows a new increase in xvalues. Interphase comparison as regards the characteristicsof patients or postoperative results did not yield statisticallysignificant differences. This means that the three phasesshowed homogeneity in relation to age, BMI, ASA, comor-bidity index or postoperative complications, hospital stay ortype of procedure or operative time required for the proce-dure. The comparative analysis of the characteristics ofpatients and intra- and postoperative parameters betweenthe phases identified in the CUSUMOT is shown in Table 2.

If we compare phases 1 and 2 combined (initial and plateauphases) with phase 3, we do not detect differences as regardsdemographic characteristics of the patients and intraoperative(type of surgical procedure and intraoperative complications)or postoperative outcomes (complications, hospital stay, etc.).However, we do observe significant differences in relation toset-up time, which was shorter in phase 3 and operative time,which was longer in phase 3 in comparison with phases 1+2(p00.000 and p00.007, respectively).

Once the CUSUMOTcurve was analysed, we calculated theCUSUMSU curve for each of the parameters included in ourstudy (Fig. 3) which showed two cut-off points coincidingwith the CUSUMOTcurve; the location of these critical pointscorresponds to patients 11 and 23, which means an increase inthe learning phase in two patients and the stabilisation of theconsolidation phase in 12 patients. Table 2 includes the statis-tical comparison of the variables considered in each of thephases showing only significant differences as regards roboticset-up time (p00.000).

Fig. 1 Graph representing thelocation of the tumours in eachof the cases arranged inchronological order

Int J Colorectal Dis

Page 4: Learning curve for robotic-assisted laparoscopic rectal cancer surgery

Discussion

As far as robotic assisted surgery is concerned, there arereports analysing the learning curves corresponding tourological procedures such as radical prostatectomy ornephrectomy and to gynaecological procedures such ashysterectomy. However, in the field of surgery, there areonly two reports on the learning process of roboticassisted management of colorectal surgery [14, 15], andonly one of them deals specifically with rectal cancerlearning curve.

CUSUM is a method employed to analyse learningcurves, which has been applied on several occasions tocolorectal surgery [8, 15]. Bokhari et al. [15] use thisstatistical method to analyse the learning curve for robot-ic assisted rectosigma and rectum surgery in benign andmalignant tumours. Other authors [8] have also used thesame method to evaluate the acquisition of skills in thelaparoscopic approach to colorectal diseases [8, 17, 18].

In our study, we have applied the CUSUM method toanalyse the learning curve for full robotic assisted rectalcancer resection. CUSUM method is a control chart tocalculate cumulative sums. Its main advantages are its inde-pendence from sample size, its effectiveness to detect smallshifts in the system and its ability to allow for a continuousanalysis in time and for a rapid evaluation of data. There-fore, CUSUM learning curves are used as indicators ofsatisfactory outcome in relation to the acquisition of clinicalskills. They are also used to evaluate the quality of theapproaches and to measure in a continuous way the con-sequences any sort of change introduced in the techniquemight have.

With the exception of the report by Bokhari et al. [15], sofar, the analysis of the learning curve for robotic assistedsurgery has been carried out using a quartile system [19, 20].Bell et al. [20] reviewed operative times in a series of 100

Table 1 Demographic characteristics of patients, procedures, intraoper-ative characteristics, postoperative results and TNM postoperative stage

Parameters

Patientcharacteristics

Age (years) 66.06±9.0

Comorbidity Index 3.02±0.9

BMI (kg/m2) 27.46±3.9

ASA 2.4±0.5

Sex (%)

Female 16 (37.2)

Male 27 (62.8)

High tumours (n) 10 (23.2)

Medium tumours (n) 19 (44.1)

Low tumours (n) 14 (32.5)

Procedures Anterior resections (%) N036 (83.7)

Abdominoperineal resections (%) N07 (16.2)

Intraoperativecharacteristics

Docking time (min) 62.90±24.6

Operative time (min) 197.44±44.3

Intraoperative complications (%) N03 (6.9)

Postoperativeresults

Length of stay (days) 12.27±10.2

Postoperative complications (%) N07 (16.2)

Postoperativestaging

T (%) T103 (6.9)

T2011 (25.5)

T3019 (44.1)

T402 (4.6)

CR06 (13.9)

TIS02 (4.6)

N (%) N0027 (62.7)

N109 (20.9)

N207 (16.2)

M M000

M100

Continuous variables are depicted as mean ± standard deviation; cate-gorical variables are depicted as percentage, n (%)

Fig. 2 CUSUMOT graphrepresenting the invertedparabola where we observe theassociation between each case(in chronological order) and theoperative time employed. Thethree phases/stages described inthe text are highlighted in thegraph

Int J Colorectal Dis

Page 5: Learning curve for robotic-assisted laparoscopic rectal cancer surgery

consecutive hysterectomies, grouping them in quintilesof 20 patients each. These authors observe that bothcomplications and operative time decrease throughoutthe study period, although this decrease is more remark-able in the first quintile. Out et al. [21, 22], on theother hand, report their results on a series of 100 radicalprostatectomies, dividing the patients into three groupsand comparing the results obtained in each group.Improvements in operative time occur after the first 30cases in their personal experience.

Akmal et al. [14] analyse robotic assisted rectal cancersurgery conducted by the same surgeon dividing the inter-ventions into two groups comprising the same number ofpatients. They compare the demographic, operative data andmorbidity between the two groups. Unlike other studies[15], Akmal et al. do not find differences in the parametersunder study. Nevertheless, they conclude that further studiesshould be carried out in this sense as the robotic approach tototal mesorectal excision may reduce the learning curve forthis technique.

Once the two CUSUM equations (CUSUMOT andCUSUMSU) are applied, we obtain two curves: theCUSUMOT curve showing the operating time throughoutthe 43 cases ordered chronologically and the CUSUMSU

curve which shows the probability of success of theintervention (absence of conversion or complications,etc.), which should increase as the team of surgeonsgradually master the technique.

Our study, resembling that by Bokhari [15] for bothbenign and malignant colorectal pathology, shows graphicswith three well-differentiated phases: a first phase whichrepresents the initial learning stage of surgeons and showsa rapid decrease of operative time, the second phase show-ing stabilisation of operative time which together with theprevious phase can be defined as the learning curve and,finally, the third phase comprising the most difficult cases.During the first phase, the tumours are located in the upper

Table 2 Comparison of procedures carried out, docking and operative times, intra- and postoperative complications and mean hospital staydistributed by phases according to CUSUMOT and CUSUMSU

Phase 1 Phase 2 Phase 3 Comparative phases 1 and2 vs 3

CUSUMOT CUSUMSU CUSUMOT CUSUMSU CUSUMOT CUSUMSU CUSUMOT CUSUMSU

Age (years) 63.3±2.2 62.63±8.4 66.4±3.2 66.33±10.3 67±1.8 67.80±8.3 NS NS

BMI (kg/m2) 26.8±0.9 26.48±2.6 26.9±1.2 26.85±4.4 28±0.8 28.37±4.1 NS NS

ASA 2.4±0.1 2.45±0.5 2.5±0.1 2.5±0.5 2.4±0.1 2.45±0.6 NS NS

Gender (F/M) 4/5 5/6 3/9 3/9 9/13 8/12 NS NS

CharlsonComorbidity Index

3.0±0.8 2.90±1.0 3.16±1.3 3.16±1.1 2.95±0.7 3.00±0.7 NS NS

AR (%) 8 (18.6) 8 (18.6) 8 (18.6) 10 (23.2) 20 (46.5) 18 (41.8) NS NS

APR (%) 1 (2.3) 3 (4.6) 4 (9.2) 2 (4.6) 2 (4.6) 2 (4.6) NS NS

DT (min) 86.1±20.8 87.72±19.1 78.3±23.3 70.00±24.8 45.0±7.5 45.0±7.9 <0.001 <0.001

OT (min) 160.0±30 166.36±36.4 201.6±40.6 209.16±40.5 210.4±4 207.5±44.1 0.007 NS

TN (%) 0 0 (0) 1 (2.3) 1 (2.3) 0 0 (0) NS NS

IC (%) 0 (0) 0 (0) 2 (4.6) 3 (4.6) 1 (2.3) 0 (2.3) NS NS

CR (%) 1 (2.3) 1 (2.3) 2 (4.6) 2 (4.6) 3 (6.9) 3 (6.9) NS NS

LOS (days) 13.2 12.09±8.1 8.7 8.91±6.5 13.8 14.25±11.1 NS NS

PC (%) 2 (4.6) 2 (4.6) 1 (2.3) 3 (6.9) 4 (9.2) 1 (2.3) NS NS

Continuous variables are depicted as mean ± standard deviation; categorical variables are depicted as percentage, n (%)

AR anterior resection, APR abdomino-perineal resection, DT docking time, OT operative time, TN transfusion necessity, IC intraoperativecomplications, CR conversion rates, LOS length of hospital stay, PC postoperative complications, NS not significant

Fig. 3 CUSUMSU graph representing success cases in our study. Thethree phases/stages described in the text are highlighted in the graph

Int J Colorectal Dis

Page 6: Learning curve for robotic-assisted laparoscopic rectal cancer surgery

region, whereas in the third phase, they are frequently lo-cated in the middle and inferior regions, which complicatesthe procedures and justifies the increase in operative timeobserved in the third phase, despite the skills acquired.

The inclusion of more complex patients with rectaltumours located in a lower region is carried out in line withthe growing technological adaptation of surgeons to the ro-botic system (icons, technical directions or placement of tro-cars) and to the team of surgeons itself (verbal and non-verbalcommunication between the surgeon and the surgical assistantor the previous laparoscopic experience of the team).

Tekkis et al. [8] in their analysis of the learning curve forboth right and left colon cancers laparoscopy do not detect adecrease of complications, despite the skills acquired. Theexplanation, according to these authors, is the inclusion ofmore complex cases in the last phase. Neither have we beenable to confirm any decrease in complications nor in read-mission rates as surgeons’ skills improve. Nevertheless, set-up time, which does not depend on the characteristics ofpatients but on the nursery and anaesthesia staff, decreasesdrastically and progressively in the third phase in compari-son to the previous two phases.

In general terms, the CUSUMmethod establishes that theprocedures of laparoscopic colorectal cancer resection nec-essary to achieve the learning curve are 55 for the rightcolon and 62 for the left one [8]. In the case of rectal cancersurgery, the estimated number of procedures is 60–80 [10].Yet, there are no reports on the estimated number of neces-sary procedures to reach the learning curve for roboticassisted surgery of malignant colorectal tumours.

According to our study and basing the analysis of thedifferent phases of the parameters on operative time, weobserve a clear decrease in set-up time and a sustainedoperative time from the first 21–23 interventions, de-spite the increasing difficulty of the cases included inour study. During these learning phases, both the intra-operative (conversion, complications, etc.) and the post-operative parameters (complications, mean hospital stay,etc.) remain within a range similar to that reported byother authors for robotic assisted rectal cancer laparos-copy and laparotomy [23–25].

Conclusions

We conclude that the learning curve for robotic assistedrectal cancer surgery may be divided in an initial phase ofskills acquisition, a second phase of consolidation of thetechnique and a third phase when the surgeon masters thetechnique and deals with more complex cases. According toour study, the estimated learning phase for rectal cancer isachieved after 21–23 procedures.

References

1. Herold A, Lehur P, Matzel K, O’conell P (2008) Coloproctology(European Manual of Medicine), 1st edn. Springer, Berlin andHeidelberg GmbH & Co. K ed

2. Miyajima N, Fukunaga M, Hasegawa H, Tanaka J, Okuda J,Watanabe M (2009) Results of a multicenter study of 1,057 casesof rectal cancer treated by laparoscopic surgery. Japan Society ofLaparoscopic Colorectal Surgery. Surg Endosc 23:113–118

3. Liakakos T, Kopanakis K, Schizas D (2010) Favorable outcomes withlaparoscopic surgery for rectal cancer. Surg Endosc 24:1217–1218

4. Staudacher C, Vignali A (2010) Laparoscopic surgery for rectalcancer: the state of the art. World J Gastrointest Surg 2:275–282

5. Guillou PJ, Quirke P, Thorpe H, Walker J, Jayne DG, Smith AM,Heath RM, Brown JM (2005) MRC CLASICC trial group. Short-term endpoints of conventional versus laparoscopic-assisted sur-gery in patients with colorectal cancer (MRC CLASICC trial):multicenter, randomized controlled trial. Lancet 365:1718–1726

6. Van de Paas MHGM, Te Velde EA, Cuesta MA, Bonjer HJ (2011)Favorable outcomes with laparoscopic surgery for rectal cancer.Surg Endosc 25(6):2060–2061. doi:10.1007/s00464-010-1457-4

7. Park IJ, Choi GS, Lim KH, Kang BM, Jun SH (2009)Multidimensional analysis of the learning curve for laparoscopicresection in rectal cancer. J Gastrointest Surg 13:275–281

8. Tekkis PP, Senagore AJ, Delaney CP, Fazio VW (2005) Evaluationof the learning curve in laparoscopic colorectal surgery: compari-son of right-sided and left-sided resections. Ann Surg 242:83–91

9. Schlachta CM, Mamazza J, Seshadri PA, Cadeddu M, Gregoire R,Poulin EC (2001) Defining a learning curve for laparoscopiccolorectal resections. Dis Colon Rectum 44:217–222

10. Son GM, Kim JG, Lee JC, Suh YJ, Cho HM, Lee YS, Lee IK,Chun CS (2010) Multidimensional analysis of the learning curvefor laparoscopic rectal cancer surgery. J Laparoendosc Adv SurgTech A 20:609–617

11. Pigazzi A, Ellenhorn JD, Ballantyne GH, Paz IB (2006) Robotic-assisted laparoscopic low anterior resection with total mesorectalexcision for rectal cancer. Surg Endosc 20:1521–1525

12. Jiménez Rodríguez RM, Díaz Pavón JM, De la Portilla De Juan F,Prendes Sillero E, Cadet Dussort JMH, Padillo J (2011)Prospective, randomized, short-term outcome study: robotic-assisted laparoscopic surgery versus conventional laparosocpicsurgery in colorectal carcinoma resection. Cir Esp 89:432–438

13. Kwak JM, Kim SH, Kim J, Son DN, Baek SJ, Cho JS (2011)Robotic vs laparoscopic resection of rectal cancer: short-term out-comes of a case–control study. Dis Colon Rectum 54:151–156

14. Akmal Y, Baek JH, McKenzie S, Garcia-Aguilar J (2012) Robot-assisted total mesorrectal excisión: is there a learning curve? SurgEndosc. doi:10.1007/s00464-012-2216-5

15. Bokhari MB, Patel CB, Ramos-Valadez DI, Ragupathi M, HaasEM (2011) Learning curve for robotic-assisted laparoscopic colo-rectal surgery. Surg Endosc 25:855–860

16. Charlson ME, Pompei P, Ales KL, McKenzie CR (1987) A newmethod of classifying prognostic comorbidity in longitudinal stud-ies: development and validation. J Chronic Dis 40:373–383

17. Bege T, Lelong B, Esterni B, Turrini O, Guiramand J, Francon D,Mokart D, Houvenaeghel G, Giovannini M, Delpero JR (2010)The learning curve for the laparoscopic approach to conservativemesorrectal excision for rectal cancer. Lessons drawn from a singleinstitution’s experience. Ann Surg 251:249–253

18. Son G, Kim J, Lee J, Suh Y, Cho H, Lee Y, Lee I, Chun C (2010)Multidimensional analysis of the learning curve for laparoscopicrectal cancer surgery. J Laparoendosc Adv Surg Tech 20:609–617

19. Lim PC, Kang E, Park do H (2011) A comparative detail analysisof the learning curve and surgical outcome for robotic hysterecto-my with lymphadenectomy versus laparoscopic hysterectomy with

Int J Colorectal Dis

Page 7: Learning curve for robotic-assisted laparoscopic rectal cancer surgery

lymphadenectomy in treatment of endometrial cancer: a case-matched controlled study of the first one hundred twenty twopatients. Gynecol Oncol 120:413–418

20. Bell MC, Torgerson JL, Kreaden U (2009) The first 100 da Vincihysterectomies: an analysis of the learning curve for a singlesurgeon. S D Med 62(91):93–95

21. Ou YC, Yang CR, Wang J, Cheng CL, Patel VR (2011)Learning curve of robotic-assisted radical prostatectomy with60 initial cases by a single surgeon. Asian J Surg 34:74–80

22. Ou YC, Yang CR, Wang J, Cheng CL, Patel VR (2010)Robotic-assisted laparoscopic radical prostatectomy: learningcurve of first 100 cases. Int J Urol 17:635–640

23. Tekkis PP, Senagore AJ, Delaney CP (2005) Conversion rates inlaparoscopic colorectal surgery: a predictive model with 1253patients. Surg Endosc 19:47–54

24. Marusch F, Gastinger I, Schneider C et al (2001) Importance ofconversion for results obtained with laparoscopic colorectal sur-gery. Dis Colon Rectum 44:207–214

25. Belizon A, Sardinha CT, Sher ME (2006) Converted laparoscopiccolectomy: what are the consequences? Surg Endosc 20:947–951

Robotic assisted rectal cancer surgery offers multiple advantages forsurgeons, and it seems to yield the same clinical outcomes as regardsthe short-time follow-up of patients compared to conventional laparos-copy. The aim of the present study is to determine whether the learningcurve for this surgical technique can be extrapolated to the well-knownconventional laparoscopic technique or whether it enhances laparo-scopic surgical performance offering further advantages.

Int J Colorectal Dis