Self-reported open cholecystectomy training, practice, and confidence in the laparoscopic era: An international convenience-sample survey

Abstract

BACKGROUND

Open cholecystectomy has become an infrequently performed rescue operation that is required precisely when laparoscopic dissection is unsafe, when the critical view of safety cannot be obtained, or when bile duct injury is suspected. The training that supports this operation in the laparoscopic era is poorly characterized at the surgeon level, and it is not known how often standardized assessment or simulation-based training for emergency conversion is actually delivered.

AIM

To describe self-reported operative exposure, structured training, intraoperative practice and confidence regarding open cholecystectomy among surgeons responding to an internationally disseminated survey.

METHODS

Anonymous cross-sectional online survey using an investigator-designed 24-item questionnaire disseminated by nonprobability convenience sampling through professional messaging groups of general, trauma and emergency, and hepatopancreatobiliary surgeons and through the channels of a national surgical society between July 2025 and February 2026; no sampling frame was defined and no response rate could be calculated. Eligible respondents were general surgery residents, fellows and attending surgeons. Analyses were descriptive and item-level; primary estimates are reported with Wilson 95% confidence intervals (CI), and no inferential comparisons were performed. Cumulative open cholecystectomy volume is additionally presented stratified by years of practice and by current role. The internal structure of the instrument was examined in a separate psychometric evaluation reported as Supplementary material. Confidence items were recorded on 5-point ordinal scales with explicit verbal anchors and are reported as complete response distributions. The instrument did not include an item on country or region of practice.

RESULTS

All 391 submitted questionnaires were complete and non-duplicate and were analyzed (270 attendings, 105 residents, 16 fellows); 68.3% (95%CI: 63.5-72.7) reported more than 100 laparoscopic cholecystectomies as primary surgeon, compared with 37.6% (95%CI: 32.9-42.5) for open cholecystectomy; 29.7% (95%CI: 25.4-34.4) reported 10 or fewer open cases. Direct observation by a supervisor was reported by 83.1% (95%CI: 79.1-86.5), but formal standardized assessment with the Objective Structured Assessment of Technical Skills or the Global Operative Assessment of Laparoscopic Skills was reported by only 20.7% (95%CI: 17.0-25.0), and simulation or hands-on training addressing emergency conversion by 16.4% (95%CI: 13.0-20.4). Intraoperative cholangiography was used selectively by 69.8% and indocyanine green fluorescence was never used by 77.5%. Self-reported confidence was concentrated in the highest category for the cystic pedicle (55.5%) and the critical view of safety (53.7%), but dispersed across all categories for common bile duct exploration and for identification and repair of bile duct injury, for which 32.5% (95%CI: 28.0-37.3) selected the two lowest categories. The instrument did not support a unidimensional measurement model, so all results are reported item by item.

CONCLUSION

Among these self-selected respondents, open cholecystectomy exposure, standardized assessment and conversion-specific simulation were heterogeneous and frequently limited, and self-reported confidence fell sharply for the biliary scenarios in which open expertise is most consequential. These findings concern self-reported confidence rather than objective competence and do not describe a defined surgical workforce.

Key Words: Cholecystectomy; Bile duct injury; Surgical education; Clinical competence; Simulation training; Cross-sectional studies

Core Tip: Among 391 self-selected respondents to an internationally disseminated survey, laparoscopic and open cholecystectomy exposure diverged sharply: Two thirds reported more than 100 laparoscopic cases as primary surgeon, but only about one third reported the same for the open operation and nearly one in three reported ten or fewer. Only one in five had ever undergone standardized skills assessment and one in six had received simulation training for emergency conversion. Self-reported confidence remained high for routine steps but fell for common bile duct exploration and for repair of bile duct injury, the scenarios in which open biliary competence actually matters. Because the sample was self-selected and confidence is not competence, these findings are descriptive and hypothesis-generating.

INTRODUCTION

Laparoscopic cholecystectomy has been the standard treatment for symptomatic gallbladder disease for more than three decades, with well-documented advantages over the open operation in postoperative pain, length of stay, recovery and cosmesis[1,2]. The proportion of cholecystectomies performed through an open approach has consequently fallen in most high- and middle-income countries, while the total volume of biliary procedures has continued to rise[3].

Open cholecystectomy nevertheless remains part of contemporary practice. Conversion continues to be used when laparoscopic dissection is unsafe or cannot be completed, particularly in severe inflammation, aberrant biliary anatomy, uncontrolled bleeding or suspected bile duct injury[4,5]. The Tokyo Guidelines and the multi-society safe cholecystectomy guideline both frame conversion as a safety strategy rather than a technical failure[6,7]. The operation has therefore changed category: It is no longer a routine alternative but a rescue procedure invoked at the precise moment when the risk of biliary injury is highest.

This shift has consequences for training. Surgeons trained after the diffusion of laparoscopy accumulate substantial minimally invasive experience while having fewer opportunities to perform open biliary surgery[8,9], and reduced exposure has been linked to concern about the recognition and management of iatrogenic biliary injury[10]. A landmark single-institution analysis showed that the mean number of open cholecystectomies performed by graduating chief residents fell from 70.4 in the prelaparoscopic era to 3.6 in the most recent decade[11], and national case-log analyses have shown that growth in resident biliary volume is almost entirely attributable to the laparoscopic operation[12,13]. Recent expert guidance reframes conversion itself as a complex procedure requiring specific expertise rather than a simple bailout[14].

In parallel, standardized assessment instruments such as the Objective Structured Assessment of Technical Skills (OSATS) and the Global Operative Assessment of Laparoscopic Skills (GOALS), together with simulation-based curricula, have been introduced to support competency-based training[15,16]. How often these strategies are actually delivered for open biliary surgery, and how they coexist with intraoperative adjuncts such as intraoperative cholangiography and indocyanine green (ICG) fluorescence[17,18], has not been described at the level of the individual surgeon across training systems.

The present study aimed to describe, among respondents to an internationally disseminated survey of general surgery residents, fellows and attending surgeons: (1) Cumulative operative exposure to laparoscopic, robotic and open cholecystectomy; (2) Exposure to formal, structured and simulation-based training; (3) Self-reported use of intraoperative adjuncts and formal training in common bile duct exploration (CBDE); and (4) Self-reported confidence in key steps of open cholecystectomy and in the management of intraoperative bile duct injury. All analyses were planned as descriptive.

MATERIALS AND METHODS

Study design and ethical considerations

This was an anonymous, cross-sectional online survey. The study was designed and reported in accordance with the Checklist for Reporting of Survey Studies and, where applicable, with the relevant items of the STROBE statement for cross-sectional studies[19,20]. The protocol was reviewed by the Institutional Review Board of the University of Maryland, Baltimore, and determined to be exempt under 45 CFR 46.104(d), category 2 (approval No. HP-00118360). Participation was voluntary and anonymous, no personally identifiable information or patient data were collected, and electronic informed consent was obtained from all respondents.

Participants and recruitment

Eligible participants were general surgery residents, surgical fellows and attending (consultant) surgeons practicing general surgery, minimally invasive surgery, hepatopancreatobiliary surgery, or trauma and acute care surgery. There were no geographic restrictions. Recruitment took place between July 16, 2025 and February 27, 2026, using nonprobability convenience sampling. An open Google Forms link was distributed through two types of channel: Professional WhatsApp groups of general surgeons, of trauma and emergency surgeons and of hepatopancreatobiliary surgeons, and the communication channels of the Brazilian Society of Integrated Trauma Care. No individual invitations were sent, no incentives were offered and participation was voluntary and self-initiated. Because the link could be forwarded freely, the number of surgeons reached could not be determined, no sampling frame could be defined, and a response rate could not be calculated.

The instrument did not include an item on country or region of practice, and the approved protocol specified that no country-specific data would be reported. The geographic distribution of respondents therefore cannot be described quantitatively, and the study is presented as an international convenience sample rather than as a sample of any defined surgical workforce. This constraint applies to every result reported below.

Survey administration, duplicate prevention and data cleaning

The questionnaire was hosted on Google Forms (Google LLC, Mountain View, CA, United States) and was administered in English. The electronic consent statement appeared on the landing page and all substantive items were configured as required, so that submitted questionnaires contain no item-level missing data except for the three role-specific items described below. The form did not collect e-mail addresses and did not require respondents to sign in to an account, so repeat submission could not be prevented by a platform setting; the exported dataset was therefore screened before analysis for: (1) Incomplete submissions; (2) Duplicate records, defined as identical responses submitted in close succession; (3) Respondents outside the eligible roles; and (4) Implausible or internally inconsistent response patterns. No record met any of these criteria, no record was excluded and no imputation was performed. Two records shared an identical response pattern but were submitted 23 days apart and were retained as independent responses. Participant flow is reported in the Results.

Survey instrument

The instrument was a self-administered, investigator-designed 24-item questionnaire covering five domains: (1) Demographics and training level; (2) Cumulative operative exposure to laparoscopic, robotic and open cholecystectomy, as primary surgeon and as assistant; (3) Exposure to formal training, direct observation, standardized skills assessment (OSATS, GOALS) and simulation-based training for emergency conversion; (4) Intraoperative practices, including intraoperative cholangiography, ICG fluorescence and formal training in CBDE; and (5) Self-reported confidence in key intraoperative steps, recorded on a 5-point ordinal scale[15,16]. Confidence anchors ranged from not at all confident (1) to very confident (5) for dissection and control of the cystic pedicle, for obtaining the critical view of safety and for identification and repair of bile duct injury; from not at all confident (1) to extremely confident (5) for surgical exploration of the common bile duct; and from not useful (1) to extremely useful (5) for the perceived usefulness of the training formats received. The questionnaire was administered through the platform described above and is provided in full as Supplementary material. The item on standardized skills assessment asked whether the respondent had ever been formally assessed with a validated structured instrument, and OSATS and GOALS were offered as the two exemplars most likely to be recognized across training systems: OSATS as a procedure-agnostic instrument applicable to open operating, and GOALS as the laparoscopic-specific instrument most familiar in gallbladder surgery. Neither instrument is proposed here as a validated measure of open cholecystectomy performance. The construct of interest was exposure to structured assessment of any kind during biliary training, so combining the two represents the most permissive available definition of formal assessment.

The questionnaire was developed by the first author from a single clinical observation: Contemporary algorithms for the difficult gallbladder recommend conversion to open cholecystectomy as the safe exit, yet not every surgeon trained in the laparoscopic era is equipped to perform that operation. The domains and items were written to describe that gap, namely cumulative open exposure, structured training and assessment, use of intraoperative adjuncts, and self-rated confidence in the specific steps on which a conversion depends. The draft was not reviewed by an external expert panel, was not pretested in a pilot sample, and was not subjected to cognitive interviewing; no content validity index was calculated and no item relevance ratings were obtained. Content validity was therefore not established by any formal procedure. The questionnaire is therefore regarded as a descriptive instrument rather than as a validated measurement scale, and every result is reported at the level of the individual item.

Two items were written as conditional questions but were presented without automated branching, so that all respondents could answer them irrespective of the preceding answer: The item on years of practice after completion of training, directed at attending surgeons, and the item on the modality of formal CBDE training, directed at respondents reporting such training. Both are reported with their denominators stated explicitly, and where the conditional subgroup is the meaningful one the remaining responses are shown separately rather than discarded. Neither item offered a not-applicable option, so residents and fellows may have reported years of training rather than years of independent practice, and the modality reported by respondents with formal CBDE training should also be regarded as approximate; the role composition of respondents to the years-of-practice item is therefore reported in the Results. The same absence of branching is evident in the two other role-specific items: The resident year-of-training item was answered by 106 respondents, of whom 105 were residents, and the fellow year-of-training item was answered by 35 respondents, of whom 9 were fellows. In addition, the response options for years of practice (1-5, 5-10, 10-15, 15-20 and more than 20 years) overlapped at their limits, so a respondent with exactly 5, 10, 15 or 20 years could have selected either adjacent category. Responses were analyzed as presented, without reallocation, and the resulting strata are treated as approximate.

Statistical analysis

Analyses were descriptive. Categorical variables are summarized as absolute and relative frequencies. The denominator applied to each item is stated in the corresponding table, and non-response or non-applicable answers are reported as a separate category rather than excluded silently. To apply confidence intervals (CIs) consistently, the primary descriptive estimates were defined as the proportions of respondents reporting more than 100 laparoscopic, more than 100 open and 10 or fewer open cholecystectomies as primary surgeon; direct supervisor observation; formal standardized skills assessment; simulation training for emergency conversion; formal CBDE training; conversion experience; selective use of intraoperative cholangiography; never using ICG fluorescence; and the highest and the two lowest categories of each of the four confidence items. For every primary estimate, a two-sided 95%CI was calculated over the full sample (n = 391) with the Wilson score method; the CI is given in the text at first mention of each estimate. Secondary item-level frequencies in Tables 1, 2, 3, and 4 are reported without CI. No sample size was calculated a priori, because the study was descriptive and the size of the accessible population was unknown. Precision was instead assessed after data collection: With 391 respondents, the half-width of a 95%CI for a proportion is at most approximately 5.0 percentage points, which was considered adequate for descriptive purposes, whereas subgroup estimates are considerably less precise (maximum half-width approximately 9.6 points for the 105 residents and 24.5 points for the 16 fellows) and are therefore presented without CI and without comparison. The estimand was the distribution of self-reported exposure, training practices and confidence among respondents to this survey, and not among general surgeons at large; accordingly, no inferential analyses, between-group comparisons or regression models were performed and no P values are reported. Training era, current role, subspecialty and institutional volume are structurally confounded in a convenience sample and cannot be adjusted for in the absence of a sampling frame, so cumulative open cholecystectomy volume is presented descriptively by years of practice and by current role without formal comparison. Confidence items are reported as complete response distributions; no means, sums or composite scores were derived. Analyses were conducted in R version 4.5.2 (R Foundation for Statistical Computing, Vienna, Austria) with independent statistical consultancy support; the statistical methods and results were reviewed by a biomedical statistician (Patricia da Silva Fucuta, PhD, Fukuta - Estatística e Metodologia Científica, São José do Rio Preto, São Paulo, Brazil).

Table 1 Self-reported cumulative cholecystectomy volume by approach and role (n = 391).

Cholecystectomies performed	0-10	11-30	31-50	51-100	> 100
Laparoscopic (primary surgeon)	36 (9.2)	34 (8.7)	25 (6.4)	29 (7.4)	267 (68.3)
Robotic (primary surgeon)	369 (94.4)	10 (2.6)	4 (1.0)	4 (1.0)	4 (1.0)
Open (primary surgeon)	116 (29.7)	51 (13.0)	36 (9.2)	41 (10.5)	147 (37.6)
Open (as assistant)	97 (24.8)	64 (16.4)	47 (12.0)	39 (10.0)	144 (36.8)

Values are n (%). All rows sum to 391 respondents; there were no missing responses for these items.

Table 2 Training, supervision, formal assessment and training-format exposure (n = 391).

Item	n/N	%
Open cases during residency reported as sufficient for independent performance	248/391	63.4
Directly observed and informally evaluated by a supervisor	325/391	83.1
Formal standardized skills assessment (OSATS or GOALS)	81/391	20.7
Simulation or hands-on training focused on emergency conversion	64/391	16.4
Has ever converted a laparoscopic cholecystectomy to open	322/391	82.4
Formal training in common bile duct exploration	238/391	60.9
Formally evaluated in open surgical skills as part of the curriculum	226/391	57.8
Training formats ever used (multiple response)
Operating-room mentorship by a senior surgeon	307/391	78.5
Video-based training	98/391	25.1
Live animal laboratory	53/391	13.6
Cadaver laboratory	37/391	9.5
Dry laboratory	23/391	5.9
Modality of CBDE training, among the 238 reporting formal training
Open only	81/238	34.0
Laparoscopic only	28/238	11.8
Both open and laparoscopic	129/238	54.2
Modality item as answered by the 153 reporting no formal training
Open only	49/153	32.0
Laparoscopic only	36/153	23.5
Both open and laparoscopic	68/153	44.4

The modality item was written as conditional on reporting formal common bile duct exploration training but was presented without automated branching and was answered by all 391 respondents; both subgroups are shown so that the denominators are explicit. CBDE: Common bile duct exploration; GOALS: Global Operative Assessment of Laparoscopic Skills; OSATS: Objective Structured Assessment of Technical Skills.

Table 3 Self-reported confidence in key intraoperative steps of open cholecystectomy and perceived usefulness of training formats (n = 391).

Item	1 (lowest)	2	3	4	5 (highest)
Dissection and control of the cystic artery and duct	30 (7.7)	21 (5.4)	40 (10.2)	83 (21.2)	217 (55.5)
Obtaining the critical view of safety	26 (6.6)	24 (6.1)	50 (12.8)	81 (20.7)	210 (53.7)
Surgical exploration of the common bile duct	69 (17.6)	45 (11.5)	78 (19.9)	88 (22.5)	111 (28.4)
Intraoperative identification and repair of bile duct injury	62 (15.9)	65 (16.6)	89 (22.8)	87 (22.3)	88 (22.5)
Perceived usefulness of training formats received	24 (6.1)	26 (6.6)	66 (16.9)	101 (25.8)	174 (44.5)

Values are n (%). Ordinal categories were not treated as interval data; no means or composite scores were computed. Anchors as presented to respondents: 1 = not at all confident and 5 = very confident for the cystic pedicle, critical view of safety and bile duct injury items; 1 = not at all confident and 5 = extremely confident for common bile duct exploration; 1 = not useful and 5 = extremely useful for the perceived usefulness of training formats.

Table 4 Cumulative open cholecystectomy volume as primary surgeon, stratified by years of practice and by current role (n = 391).

Stratum	0-10	11-30	31-50	51-100	> 100
By years of practice
1-5 years (n = 60)	26 (43.3)	11 (18.3)	8 (13.3)	6 (10.0)	9 (15.0)
5-10 years (n = 46)	9 (19.6)	9 (19.6)	8 (17.4)	12 (26.1)	8 (17.4)
10-15 years (n = 43)	5 (11.6)	9 (20.9)	4 (9.3)	6 (14.0)	19 (44.2)
15-20 years (n = 32)	1 (3.1)	3 (9.4)	4 (12.5)	4 (12.5)	20 (62.5)
More than 20 years (n = 104)	3 (2.9)	2 (1.9)	3 (2.9)	11 (10.6)	85 (81.7)
Not answered (n = 106)	72 (67.9)	17 (16.0)	9 (8.5)	2 (1.9)	6 (5.7)
By current role
Residents (n = 105)	71 (67.6)	17 (16.2)	7 (6.7)	3 (2.9)	7 (6.7)
Fellows (n = 16)	3 (18.8)	3 (18.8)	2 (12.5)	1 (6.3)	7 (43.8)
Attendings (n = 270)	42 (15.6)	31 (11.5)	27 (10.0)	37 (13.7)	133 (49.3)

Values are n (%) within each stratum. No statistical comparison was performed. The not-answered stratum corresponds to respondents who did not answer the item on years of practice, of whom 92 were residents, 9 fellows and 5 attendings. Strata are approximate because the years-of-practice options overlap at their limits.

Because the questionnaire was newly developed, its internal structure was examined in a separate exploratory and confirmatory factor analysis. That evaluation is reported in full as Supplementary material and was not used to derive, weight or adjust any of the descriptive results presented here; its implications for the interpretation of self-reported confidence are addressed in the Discussion. The core findings of that evaluation are summarized in the Results; the full analysis, including item-level factor loadings and fit statistics, is provided as Supplementary material.

RESULTS

Participant flow

Google Forms records only submitted questionnaires, so the number of times the link was opened, the number of questionnaires started and the number abandoned before submission could not be determined. Between July 16, 2025 and February 27, 2026, 391 questionnaires were submitted. All were complete for the required items, none was identified as a duplicate record and none was excluded as ineligible, so the analytic sample comprised all 391 submitted questionnaires. Two questionnaires with an identical response pattern, submitted 23 days apart, were retained as independent responses.

Respondent characteristics

The analytic sample comprised 391 respondents (Table 5). Attending surgeons comprised the largest group (n = 270, 69.1%), followed by residents (n = 105, 26.9%) and fellows (n = 16, 4.1%). Among attendings, general surgeons predominated (n = 137, 50.7% of attendings), followed by trauma and acute care surgeons (n = 55, 20.4%), minimally invasive surgeons (n = 48, 17.8%) and hepatopancreatobiliary surgeons (n = 30, 11.1%). Most residents were in the middle years of training (postgraduate year 3, 43.8%; postgraduate year 2, 30.5%). The item on years of practice was answered by 285 respondents (72.9%, 95%CI: 68.3-77.1), a figure exceeding the 270 who identified themselves as attendings because the item was presented without branching. These 285 respondents comprised 265 attendings, 7 fellows and 13 residents, and 5 attendings did not answer the item; 20 respondents to this item (7.0%) were therefore not attendings. The 106 non-responders comprised 92 residents (86.8%), 9 fellows and 5 attendings, which confirms that non-responders were predominantly, but not exclusively, residents. Over the full sample, 104 respondents (26.6%) reported more than 20 years of practice and 60 (15.3%) five years or fewer. Hepatopancreatobiliary surgeons, whose open biliary exposure would be expected to be atypical of general surgical practice, constituted a small minority of the sample (30 attendings and 5 fellows, 35 of 391 respondents, 9.0%).

Table 5 Respondent characteristics (n = 391).

Characteristic	n	%
Current role
General surgery resident	105	26.9
Fellow - hepatopancreatobiliary surgery	5	1.3
Fellow - minimally invasive surgery	7	1.8
Fellow - trauma, critical and acute care surgery	4	1.0
Attending general surgeon	137	35.0
Attending hepatopancreatobiliary surgeon	30	7.7
Attending minimally invasive surgeon	48	12.3
Attending trauma and acute care surgeon	55	14.1
Postgraduate year (residents, n = 105)
PGY-1	14	13.3
PGY-2	32	30.5
PGY-3	46	43.8
PGY-4	1	1.0
PGY-5	12	11.4
Years of practice after completion of training
1-5 years	60	15.3
5-10 years	46	11.8
10-15 years	43	11.0
15-20 years	32	8.2
More than 20 years	104	26.6
Not answered	106	27.1

Percentages for current role and years of practice are calculated over the full sample (n = 391); percentages for postgraduate year are calculated over the 105 residents. The response options for years of practice are reproduced exactly as presented to respondents and overlap at 5, 10, 15 and 20 years, which is a limitation of the instrument. The item was directed at attending surgeons but, having been presented without automated branching, was answered by 285 respondents (265 attendings, 7 fellows, 13 residents); the 106 non-responders comprised 92 residents, 9 fellows and 5 attendings. Because the options overlap and the item was not restricted to attendings, these strata are approximate. PGY: Postgraduate year.

Operative volume

Self-reported cumulative volume by approach is shown in Table 1. As primary surgeon, 267 respondents (68.3%, 95%CI: 63.5-72.7) reported more than 100 laparoscopic cholecystectomies. For open cholecystectomy as primary surgeon, 147 respondents (37.6%, 95%CI: 32.9-42.5) reported more than 100 procedures, whereas 116 (29.7%, 95%CI: 25.4-34.4) reported 10 or fewer and 51 (13.0%) reported 11 to 30. A similar distribution was observed for open cholecystectomy as assistant. Robotic experience was uncommon: 369 respondents (94.4%) reported 10 or fewer robotic cholecystectomies.

Training, assessment and exposure

Responses concerning training, supervision and standardized assessment are shown in Table 2. Although 325 respondents (83.1%, 95%CI: 79.1-86.5) reported having been directly observed and informally evaluated by a supervisor during an open cholecystectomy, only 81 (20.7%, 95%CI: 17.0-25.0) reported formal standardized skills assessment with OSATS or GOALS for this procedure. Simulation or hands-on training specifically addressing emergency conversion was reported by 64 respondents (16.4%, 95%CI: 13.0-20.4). Overall, 248 respondents (63.4%) judged the number of open cases during residency sufficient for independent performance and 226 (57.8%) reported formal evaluation of open surgical skills as part of the program curriculum.

Operating-room mentorship by a senior surgeon was the most frequently reported training format (n = 307, 78.5%); video-based training (25.1%), live animal laboratory (13.6%), cadaver laboratory (9.5%) and dry laboratory (5.9%) were each reported by a minority. Formal training in CBDE was reported by 238 respondents (60.9%, 95%CI: 55.9-65.6); among these, the modality was open only in 81 (34.0%), laparoscopic only in 28 (11.8%) and both in 129 (54.2%). Because the modality item was presented without branching it was also answered by the 153 respondents who reported no formal CBDE training; those responses are excluded from the percentages above and shown separately in Table 2. Conversion of a laparoscopic cholecystectomy to an open procedure had been performed at least once by 322 respondents (82.4%, 95%CI: 78.3-85.8).

Intraoperative adjuncts

Intraoperative cholangiography was used routinely by 44 respondents (11.3%), selectively by 273 (69.8%, 95%CI: 65.1-74.2) and never by 74 (18.9%). ICG fluorescence was used routinely by 15 respondents (3.8%), selectively by 73 (18.7%) and never by 303 (77.5%, 95%CI: 73.1-81.4).

Self-reported confidence

Confidence distributions are summarized in Table 3 and Figure 1. For dissection and control of the cystic artery and duct, 217 respondents (55.5%, 95%CI: 50.5-60.3) selected the highest category and 51 (13.0%, 95%CI: 10.1-16.7) the two lowest combined; for obtaining the critical view of safety, the corresponding figures were 210 (53.7%, 95%CI: 48.8-58.6) and 50 (12.8%, 95%CI: 9.8-16.5). Responses were considerably more dispersed for the more complex scenarios. For surgical exploration of the common bile duct, 111 respondents (28.4%, 95%CI: 24.1-33.1) selected the highest category and 114 (29.2%, 95%CI: 24.9-33.8) the two lowest combined. For intraoperative identification and repair of bile duct injury, responses were distributed across all five categories, with 88 respondents (22.5%, 95%CI: 18.6-26.9) in the highest category and 127 (32.5%, 95%CI: 28.0-37.3) in the two lowest combined.

Figure 1 Self-reported confidence in key intraoperative steps of open cholecystectomy (n = 391). Values are percentages of all respondents on a 5-point ordinal scale. Anchors as presented to respondents: 1 = not at all confident and 5 = very confident for the cystic pedicle, the critical view of safety and the bile duct injury item; 1 = not at all confident and 5 = extremely confident for common bile duct exploration.

Exposure by years of practice and current role

Cumulative open cholecystectomy volume as primary surgeon, stratified by years of practice and by current role, is presented in Table 4 and Figure 2. The proportion reporting more than 100 open cholecystectomies rose across successive strata of practice duration, from 15.0% among respondents with 1 to 5 years of practice to 81.7% among those with more than 20 years, while the proportion reporting 10 or fewer open cases fell from 43.3% to 2.9% across the same strata. Residents accounted for most low-volume responses: 67.6% of residents reported 10 or fewer open cholecystectomies, compared with 15.6% of attendings. No comparisons were performed and no statistical tests were applied to these strata. Because the years-of-practice categories overlap at their limits and include some non-attending respondents, these strata are approximate.

Figure 2 Cumulative open cholecystectomy volume as primary surgeon, stratified by years of practice (n = 285). Values are percentages within each stratum. The 106 respondents who did not answer the item on years of practice, 92 of whom were residents, are not shown. The years-of-practice categories are reproduced as presented to respondents and overlap at their limits. No statistical comparison was performed.

Internal structure of the instrument

In the exploratory step, parallel analysis nominally indicated six factors, but the first empirical eigenvalue substantially exceeded the second (7.47 vs 0.88) and accounted for 44% of the total variance, so a unidimensional model was tested; four items fell below the 0.30 retention threshold and were excluded, and 13 items were carried forward. In the confirmatory model (unweighted least squares mean- and variance-adjusted estimator, n = 391), standard and scaled fit indices indicated adequate to excellent fit (comparative fit index: 0.996 and 0.979; root mean square error of approximation: 0.039, 90%CI: 0.025-0.053, and 0.052, 90%CI: 0.039-0.064; standardized root mean square residual: 0.054), whereas the robust indices, which correct for the non-normality of ordinal survey data and are the appropriate benchmark here, did not support the model (comparative fit index: 0.813; Tucker-Lewis index: 0.776; root mean square error of approximation: 0.190, 90%CI: 0.171-0.209). The dichotomous items on formal standardized assessment, simulation training and use of intraoperative adjuncts did not load meaningfully on the general factor, which was dominated by the confidence ratings and by cumulative operative volume. Because the instrument did not support a valid unidimensional measurement model, no composite or summed score was computed at any stage, and all items are reported as descriptive frequencies only. Because content validity had not been established by a formal procedure, this evaluation addresses internal structure only and does not constitute validation of the instrument. The implications for the interpretation of self-reported confidence are considered in the Discussion.

DISCUSSION

This survey provides a contemporary, surgeon-level description of self-reported exposure, training, intraoperative practice and confidence related to open cholecystectomy among 391 self-selected residents, fellows and attending surgeons. Because respondents were not drawn from a defined population and confidence was self-rated, the four observations below describe this sample; they are not estimates of competence or of any surgical workforce.

First, the gap between laparoscopic and open exposure was substantial. Approximately two thirds of respondents reported more than 100 laparoscopic cholecystectomies as primary surgeon, but fewer than 40% reported the same for the open operation and nearly 30% reported ten or fewer open cases. This is consistent with case-log analyses documenting a progressive decline in open biliary exposure since the diffusion of laparoscopy[8,11-13], and it matters because the open operation is now recommended precisely for the difficult gallbladder with severe inflammation, fibrosis or distorted hilar anatomy[4-7].

Second, formal standardized assessment appeared uncommon among respondents. Most respondents had been directly observed by a supervisor, but only one in five had ever undergone assessment with a validated instrument such as OSATS or GOALS. Direct observation without a structured scoring framework does not reliably generate reproducible feedback, and competency-based frameworks increasingly assume the use of validated tools[15,16].

Third, simulation or hands-on training explicitly addressing emergency conversion was reported by only 16.4% of respondents. Conversion is the scenario in which familiarity with open biliary anatomy, tactile dissection and open suturing is most critical, and in which delay or poor execution has been associated with major bile duct injury[5,6,10]. In a cohort of senior Irish trainees, 92% reported no independent experience with open cholecystectomy and 65% had performed two or fewer cases with trainer assistance, yet 69% reported feeling somewhat or very comfortable converting to open surgery, a discrepancy those authors attributed to a Dunning-Kruger effect[21,22]. Contemporary expert guidance similarly reframes conversion as a complex procedure that warrants consultation with a surgeon experienced in open hepatobiliary surgery, and emphasizes safer laparoscopic alternatives such as subtotal cholecystectomy when the critical view of safety cannot be obtained[14].

Fourth, self-reported confidence differed across items according to the complexity of the task. Confidence for the routine steps clustered in the upper range of the scale, whereas confidence for CBDE and for identification and repair of bile duct injury was distributed across all five categories, with roughly a third of respondents in the two lowest categories for the latter. Because these are the scenarios in which open biliary expertise is most consequential, this dispersion deserves attention from program directors[10]. An Australasian survey identified exposure to more than 20 open cholecystectomies as a threshold above which trainees more frequently reported confidence in performing an independent elective open cholecystectomy (87.8% vs 57.3%) and in converting from laparoscopic to open surgery (87.8% vs 58.7%)[23].

Self-reported confidence must nevertheless be read with caution, and the psychometric evaluation of the instrument reinforces this point. Examination of the internal structure of the questionnaire (Supplementary material) did not support a valid unidimensional measurement model: Robust fit indices were inadequate, the dichotomous items on formal assessment, simulation training and intraoperative adjuncts failed to load on the general factor, and the surviving factor was dominated by confidence ratings and operative volume. Because the model was not supported, the general factor should not be interpreted as a measured construct; at most, the pattern suggests that confidence ratings tracked accumulated operative experience more closely than any marker of structured training, and in no case can they be read as technical proficiency. Two explanations are compatible and not mutually exclusive: Those items ask whether an event occurred rather than what the respondent can do, which constrains behavioral specificity; and the experiences they describe were nearly universally absent in this sample, with 79.3% never assessed with a standardized instrument and 83.6% never exposed to conversion-specific simulation, producing a floor effect. For this reason, no composite score was computed, the items are reported here only as descriptive frequencies, and instruments in this field should be built from behaviorally anchored items rather than from confidence ratings[22].

The reported use of ICG fluorescence was low (routine in 3.8%, any use in 22.5%) despite evidence that fluorescence cholangiography may support safer identification of biliary anatomy in difficult cases[24]. The predominantly selective use of intraoperative cholangiography (69.8%) aligns with guidance advocating tailored rather than uniform use[25], but national data indicate that the absolute number of cholangiograms performed by graduating residents declined between 2012 and 2023 despite rising cholecystectomy volumes, with the experience concentrated in the chief resident year[26] - an adjunct for biliary mapping being learned ever later in training.

These patterns must be read together with the structure of the sample. Respondents trained before, during and after the diffusion of laparoscopy are represented in the same dataset, so training era is inseparable from role, seniority and subspecialty. The gradient in Table 4 is exactly what a decline in open exposure across training eras would produce, and also exactly what the simple accumulation of cases over a career would produce; the cross-sectional design cannot distinguish the two, and the stratification is presented for description only. For the same reason, the expression laparoscopic era designates here the period in which the survey was conducted, not an exposure that has been isolated analytically.

A further boundary condition concerns the health system in which respondents practice. The premise of this study, that open cholecystectomy has become a rescue operation, holds for settings in which laparoscopy is universally available, but not for many low- and middle-income and rural environments, where the open operation remains a routine first-line procedure performed for reasons of equipment, cost and logistics rather than of intraoperative difficulty. Surgeons working in those settings would be expected to report substantially greater open exposure and a different confidence profile, and their presence in an undefined proportion within this convenience sample is one plausible contributor to the heterogeneity observed. Because country and region of practice were not collected, the contribution of practice context cannot be quantified, and the results should be read as describing these respondents rather than as characterizing any national or regional surgical workforce.

Implications for education and practice

The following implications are offered as hypotheses for testing rather than as recommendations derived from comparative evidence. Training programs may benefit from structured open cholecystectomy modules combining operating-room mentorship with validated assessment instruments and with simulation designed specifically for emergency conversion and bile duct injury management. Cadaveric, animal or high-fidelity simulation may partially compensate for reduced clinical volume. A structured competency-based laparoscopic cholecystectomy curriculum has been shown to improve resident operative performance and reduce skill variability[27], which supports developing an analogous pathway for open technique. For practicing surgeons, continuing professional development focused on open biliary anatomy, CBDE and the early recognition of iatrogenic injury may help preserve a skill set that remains indispensable when laparoscopy is unsafe. Prospective evaluation with pre- and post-intervention objective assessment is the natural next step.

Limitations

Several limitations must be acknowledged. First, the design was cross-sectional and used nonprobability convenience sampling through professional messaging groups and the channels of a national surgical society; respondents may differ substantially from the broader population of surgeons, no sampling frame was defined, and a response rate could not be calculated. Self-selection is likely to have favored surgeons with an interest in surgical education or biliary surgery and those connected to the dissemination networks, in directions that cannot be quantified. Because recruitment relied on professional messaging groups and on the channels of a Brazilian surgical society, respondents from Brazil and from Latin America are likely to be over-represented, although the absence of a country item means this cannot be verified. The findings describe the respondents and should not be generalized to surgeons in any country or health system. Second, country and region of practice, hospital type (for example academic or community, public or private), and institutional cholecystectomy volume were not collected, so neither geographic representativeness nor the influence of practice context can be assessed. Third, all variables were self-reported and are subject to recall and social desirability bias; self-reported confidence is not a measure of technical competence, and no case-log verification or objective performance data were available for corroboration. Fourth, the questionnaire was investigator-designed, its content validity was not established by a formal procedure, and its internal structure did not support a unidimensional measurement model, as detailed above; this is why results are reported strictly as item-level descriptive frequencies and why comparative analyses were intentionally not performed. Fifth, two items written as conditional were presented without automated branching and were therefore answered by respondents to whom they did not apply, and the response options for years of practice overlapped at their category limits; these affect descriptive items only, may have caused some misclassification in the years-of-practice strata, are reported transparently with explicit denominators and role composition, and have been corrected in a revised instrument now undergoing formal validation. Sixth, the item on standardized skills assessment combined OSATS, which is applicable to open operating, with GOALS, which is validated for laparoscopic performance only; the item therefore captures exposure to structured assessment during biliary training rather than assessment of open technique specifically, and cannot distinguish between the two. Because that definition is deliberately permissive, the proportion of 20.7% should be read as an upper bound. Finally, the survey platform does not record link openings or abandoned attempts, so the completion rate among surgeons who opened the questionnaire is unknown, and no platform setting was available to prevent repeat submission, although screening identified no duplicate record.

CONCLUSION

Among surgeons responding to this internationally disseminated survey, exposure to open cholecystectomy, standardized assessment and simulation-based training for emergency conversion were heterogeneous and, for a meaningful proportion, limited, while self-reported confidence fell markedly for CBDE and for the management of bile duct injury. Because the sample was recruited by convenience and without a defined sampling frame, these findings describe self-selected respondents and their self-reported confidence, not objective competence or any defined surgical workforce, and they apply most directly to health systems in which the open operation has become a bailout rather than a routine first-line procedure. They nonetheless identify a coherent target for curricular development in an operation that has become a rescue procedure, and they indicate that future instruments in this field should be built from behaviorally anchored items rather than from confidence ratings.