Introduction

Gallstone disease is a common surgical problem in India, with reported prevalence ranging from 2% to 29% across regions; North Indian populations show a 2- to 4-fold higher prevalence than South Indian populations, a difference attributed to diet, lifestyle, and environmental factors, with cholesterol stones predominating in the north and pigment stones in the south [1]. The disease predominantly affects adults aged 40-50 years, although its incidence in younger patients is rising, and women are affected 2- to 3-fold more often than men [1].

Laparoscopic cholecystectomy (LC) is the gold-standard treatment for symptomatic cholelithiasis, offering reduced postoperative pain, shorter hospitalisation, and faster return to activity compared with open surgery. Safe completion of LC depends on achieving the critical view of safety during dissection of the hepatocystic triangle [2]. Despite widespread adoption and technical refinement, operative difficulty remains highly variable: a subset of patients requires prolonged dissection, conversion to open surgery, or is at elevated risk of bile duct injury, particularly in the setting of dense adhesions, a contracted or severely inflamed gallbladder, or an obscured Calot triangle.

The Parkland Grading Scale (PGS) was developed as a simple, reproducible, intraoperative five-tier scale (grade 1, minimal inflammation, through grade 5, severe inflammation with adhesions obscuring anatomy) based on the initial laparoscopic view of the gallbladder, and has been prospectively validated with good inter-rater reliability and correlation with operative time, conversion, and bile leak rates [3,4]. Because PGS is by definition an intraoperative assessment, it cannot itself inform preoperative planning; several groups have therefore proposed preoperative scoring systems (for example, the Randhawa and Pujahari score, incorporating age, sex, prior hospitalisation, body mass index (BMI), abdominal scars, palpable gallbladder, ultrasonographic (USG) wall thickness, pericholecystic fluid, and impacted stones) intended to anticipate a difficult LC before the patient reaches the operating table [5]. USG gallbladder wall thickening, BMI, prior acute cholecystitis, and inflammatory markers such as C-reactive protein (CRP) and the neutrophil:lymphocyte ratio (NLR) have variably been reported as independent predictors of a difficult or prolonged cholecystectomy in the surgical literature [6-9].

Despite this body of work, the relative and independent contribution of routinely available preoperative clinical, haematological, biochemical, and radiological parameters to intraoperative difficulty, graded objectively using PGS, has not been comprehensively examined in an Indian tertiary-referral population. We therefore conducted this prospective observational study to evaluate the association between preoperative clinical, haematological, biochemical, and radiological parameters and intraoperative severity as graded by PGS in patients undergoing LC for symptomatic cholelithiasis, with the aim of identifying parameters that could support preoperative risk stratification.

Methods

Study Design and Sample Size

This was a prospective, single-centre, observational study conducted in the Department of Surgical Gastroenterology and Liver Transplantation, Bangalore Medical College and Research Institute-Super Speciality Hospital (BMCRI-SSH, PMSSY), Bengaluru, India, over the period January 2025 to June 2026. A formal a priori sample size calculation was not performed. The study was designed as an exploratory prospective observational investigation, and all eligible patients presenting during the study period were consecutively enrolled. [10]

Flow diagram 1: STROBE-style flow diagram of participant enrolment, data collection, and analysis (N=102)

Participants

Consecutive adult patients with symptomatic cholelithiasis scheduled for elective LC at the study centre during the study period were considered for enrolment. All patients who provided informed consent and completed preoperative work-up and LC during the study period were included in the analysis; no patient was excluded from the overall dataset.

Variables

Preoperative variables recorded were: demographic (age, sex, weight, height, BMI, duration of symptoms), clinical history (prior cholecystitis, history of cholangitis, history of biliary pancreatitis, comorbidities), vital signs (temperature, pulse, systolic and diastolic blood pressure, respiratory rate), general examination findings (pallor, icterus, palpable gallbladder, abdominal scar), haematological indices (haemoglobin, packed cell volume, total leucocyte count, differential neutrophil and lymphocyte percentage, platelet count, neutrophil:lymphocyte ratio), biochemical parameters (total, direct and indirect bilirubin, total protein, albumin, globulin, albumin:globulin ratio, alkaline phosphatase, alanine aminotransferase, aspartate aminotransferase, gamma-glutamyl transferase, C-reactive protein, prothrombin time and INR), and radiological findings (USG gallbladder wall thickness, impacted cystic duct/neck stone, pericholecystic fluid, distended gallbladder, and use of contrast-enhanced CT abdomen and/or magnetic resonance cholangiopancreatography). The primary outcome was intraoperative operative difficulty, graded at surgery using the five-tier Parkland Grading Scale [3]. Secondary intraoperative and postoperative outcomes recorded were operative time, conversion to open cholecystectomy, common bile duct (CBD) injury, intraoperative drain placement, intraoperative bile/stone spillage, postoperative bile leak, surgical-site infection, hospital stay, and readmission. In cases where the Parkland Grade was not explicitly documented intraoperatively by the operating surgeon, grading was retrospectively assigned from operative records using predefined Parkland Grading Scale criteria. Data abstraction was performed during study curation using operative findings documented in the surgical notes

Statistical Analysis

Statistical analyses were performed using SPSS v.29 (IBM Corp., Armonk, NY, USA). Continuous variables are summarised as mean ± standard deviation and median (interquartile range, IQR); distributional normality was assessed using the Shapiro-Wilk test. Given evidence of non-normality for several skewed parameters, between-group comparisons (Easy/Moderate versus Difficult) for continuous variables were performed using the Mann-Whitney U test, with the Kruskal-Wallis test used to compare across all five Parkland grades. Categorical variables are summarised as frequencies and percentages and compared using the chi-square test or Fisher exact test as appropriate for expected cell counts. Receiver operating characteristic curves and area under the curve (AUC) were generated for all clinically relevant continuous predictors. Optimal cut-off values were determined using the Youden Index and AUC values were reported with corresponding 95% confidence intervals. A two-sided p<0.05 was considered statistically significant throughout

Results

Participants and descriptive data

Between January 2025 and June 2026, 102 patients undergoing LC for symptomatic cholelithiasis were enrolled and analysed; no patient was excluded from the overall dataset. The cohort comprised 77 women (75.5%) and 25 men (24.5%) (female:male ≈ 3:1), with a mean age of 45±14 years (median 45, IQR 35-55; range 20-81) and mean BMI of 24.7±2.8 kg/m² (range 20-30). Comorbidities were present in 29.4% of patients (diabetes mellitus 13.7%, hypertension 8.8%, both 6.9%); median duration of symptoms was 9 weeks (IQR 4.1-17.8). Prior history of cholecystitis was recorded in 15.7% and biliary pancreatitis in 2.9%. CECT abdomen was performed in 13.7% and MRCP in 17.6% of patients, consistent with selective, second-line use of cross-sectional imaging rather than routine practice.

Intraoperative and postoperative outcomes

The intraoperative Parkland Grade distribution was: grade 1 in 36 patients (35.3%), grade 2 in 4 (3.9%), grade 3 in 41 (40.2%), grade 4 in 20 (19.6%), and grade 5 in 1 (1.0%) (Figure 1). Using the pre-specified dichotomisation, 40 patients (39.2%) were classified as Easy/Moderate (grade 1-2) and 62 (60.8%) as Difficult (grade 3-5). No conversion to open cholecystectomy, CBD injury, surgical-site infection, or readmission occurred in this cohort (0/102 each); one patient (1.0%) developed a postoperative bile leak. Median hospital stay was 3 days (IQR 2-6).

Mean operative time was 85±36 minutes, recorded in this cohort as one of three standard theatre-time bands (45, 95, or 140 minutes); every Easy/Moderate case was completed within the 45-minute band, while every difficult case fell in the 95- or 140-minute band. As operative time is an intraoperative measurement rather than a preoperative one, it is reported here as a correlate confirming the face validity of the Parkland difficulty classification, rather than as a preoperative predictor.

Univariate associations with intraoperative difficulty

Table 1, summarises baseline demographic and clinical characteristics by difficulty category. Table 2, presents the univariate comparison of continuous preoperative parameters. Presence of any comorbidity (diabetes and/or hypertension) was significantly associated with a Difficult intraoperative course (χ² p=0.015; Table 1), as were lower platelet count (median 265 vs 309.5 ×10³/µL, p=0.037) and shorter prothrombin time (median 11.65 vs 13.3 seconds, p=0.008). No significant between-group difference was found for age, height, duration of symptoms, vital signs, haemoglobin, packed cell volume, total leucocyte count, differential counts, CRP, NLR, bilirubin fractions, total protein, albumin, globulin, A:G ratio, or the transaminase/cholestatic liver enzymes (all p≥0.10; Table 2). Notably, CRP and NLR, both classically reported as predictors of a difficult cholecystectomy in the literature [9], did not reach significance in this cohort (p=0.170 and p=0.579 respectively). Categorical radiological findings (USG impacted stone, pericholecystic fluid, distended gallbladder) and imaging-utilisation variables (CECT, MRCP performed) were also not significantly associated with difficulty grade (all p&gt;0.10; Table 1). Previous abdominal scar showed a borderline association (p=0.069), with infra- and supra-umbilical scars more frequent in the Difficult group. BMI and USG gallbladder wall thickness showed complete separation between groups (both AUC 1.00) and weight showed near-complete separation (AUC 0.99); these three variables are reported in Table 2 for the full cohort and analysed separately as standalone discriminators in Table 5, below, rather than combined into the multivariable model with the moderate-effect predictors (see Statistical analysis).

Figure 1: Distribution of intraoperative Parkland Grade (1-5) across the study cohort (N=102). Grades 1-2 (blue) are classified Easy/Moderate; grades 3-5 (red) are classified Difficult.

Multivariable logistic regression

A multivariable binary logistic regression model was constructed for Difficult (grade 3-5) versus Easy/Moderate (grade 1-2) LC, with age, platelet count, prothrombin time, and presence of any comorbidity (diabetes and/or hypertension) entered as predictors (Table 3). BMI, weight, and GB wall thickness were not entered into this particular model as these variables individually achieve near-perfect classification of difficulty in this cohort (Table 5), and attempting to combine a near-perfect single discriminator with moderate-effect covariates in one maximum-likelihood model produces unstable, non-interpretable coefficients [11]. We therefore report the two sets of predictors separately: the near-perfect discriminators as standalone cut-offs (Table 5), and the moderate-effect variables in the multivariable model below. The overall multivariable model of moderate-effect predictors was statistically significant (likelihood-ratio p=0.006) but explained only a modest proportion of outcome variance beyond that already captured by BMI/GB wall thickness/weight (Nagelkerke R²≈0.11), consistent with these three anthropometric/radiological markers capturing most of the separable signal in this dataset. Prothrombin time (adjusted OR 0.83 per second, 95% CI 0.71-0.96, p=0.012) and presence of any comorbidity (adjusted OR 3.70, 95% CI 1.09-12.56, p=0.036) were independent predictors of a Difficult course after mutual adjustment. Platelet count approached significance (adjusted OR 0.995 per unit, p=0.073); age was not an independent predictor after adjustment (p=0.737).

Prothrombin time and platelet count by difficulty category

Figures 2 and 3 illustrate the distribution of prothrombin time and platelet count, the two continuous, clinically plausible parameters that differed significantly between Easy/Moderate and Difficult groups on univariate analysis.

Figure 2; Prothrombin time (seconds) by Parkland difficulty category (Mann-Whitney p=0.008).

Figure 3: Platelet count (×10³/µL) by Parkland difficulty category (Mann-Whitney p=0.037).

Receiver operating characteristic analysis

ROC curves were constructed for the moderate-effect continuous predictors identified on univariate and multivariable analysis (Figure 4; Table 4); both showed modest discriminative value, consistent with the limited variance these variables explained beyond the strong discriminators. BMI, weight, and USG gallbladder wall thickness achieved substantially higher AUCs (0.99-1.00) and are presented separately in Table 5 and Figure 5, above/below, as standalone cut-off-based discriminators rather than alongside the moderate-effect predictors in Table 4.

Figure 4: ROC curves for prothrombin time and platelet count as predictors of a Difficult (grade 3-5) intraoperative course. Diagonal reference line represents no discrimination (AUC=0.5).

Strong preoperative discriminators of operative difficulty

BMI, weight, and USG gallbladder wall thickness each discriminated Easy/Moderate from Difficult cases with very high accuracy (Table 5; Figure 5). A BMI cut-off of 25 kg/m² and a GB wall thickness cut-off of 4.3 mm each achieved 100% sensitivity and 100% specificity in this cohort (AUC 1.00 for both); a weight cut-off of 68.8 kg achieved 95.2% sensitivity and 97.5% specificity (AUC 0.99). These cut-offs considerably exceed the discrimination reported for the same markers in previously published series (typical AUC 0.65-0.85) [6,7], and are presented as a notable finding of this cohort rather than as an already-validated clinical rule; external validation in an independent cohort is recommended before adoption into routine preoperative counselling (see Discussion).

Figure 5: ROC curves for BMI, weight, and USG gallbladder wall thickness as preoperative discriminators of a Difficult intraoperative course.

Discussion

In this prospective cohort of 102 patients undergoing LC for symptomatic cholelithiasis, 60.8% had a Difficult (Parkland grade 3-5) intraoperative course. The principal finding of this study is that three simple, routinely available preoperative/radiological measurements BMI, weight, and USG gallbladder wall thickness discriminated Easy/Moderate from Difficult cases almost perfectly (AUC 0.99-1.00), each exceeding the discrimination reported for the same markers in the published literature (typical AUC 0.65-0.85) [6,7]. The present findings are consistent with previous Indian studies that have identified obesity and gallbladder wall thickening as risk factors for difficult laparoscopic cholecystectomy, although the magnitude of discrimination observed in our cohort was substantially greater than previously reported.

Operative time, an intraoperative correlate rather than a preoperative predictor, corroborated the same separation, with every Easy/Moderate case completed in a short theatre-time band and every Difficult case in a longer one. Beyond these strong discriminators, a smaller set of moderate-effect variables comorbidity burden, platelet count, and prothrombin time added independent but modest information on multivariable analysis (Nagelkerke R²≈0.11 for this second model). We discuss the strong discriminators first, then the moderate-effect findings, before turning to strengths and limitations.

BMI ≥25 kg/m² and USG gallbladder wall thickness ≥4.3 mm each achieved 100% sensitivity and 100% specificity for a Difficult intraoperative course in this cohort, and a weight cut-off of 68.8 kg achieved 95.2% sensitivity and 97.5% specificity. These are considerably higher levels of discrimination than most published series report for the same markers [6,7], where BMI and GB wall thickness are consistently identified as predictors of a difficult cholecystectomy but with substantial overlap between easy and difficult groups (typical AUC 0.65-0.85). We present this as a genuine and clinically striking finding of the present cohort rather than as an already-generalisable clinical rule. Such unusually strong single-centre discrimination can arise from several non-mutually-exclusive sources: a case-mix or population effect specific to this tertiary-referral unit (for example, if the local threshold for a difficult intraoperative grading assessment is itself closely tied to visibly higher BMI or a thicker, more oedematous gallbladder wall on inspection); a genuinely stronger biological relationship between body habitus, wall thickening, and dissection difficulty in this population than has been captured elsewhere; or narrower-than-typical variability in the measured range of BMI and wall thickness within this particular sample (the cohort spanned a comparatively narrow BMI range of 20-30 kg/m², which can inflate apparent discrimination relative to cohorts with wider BMI ranges). We are not in a position to adjudicate between these explanations from a single-centre dataset, and we recommend that these cut-offs be treated as hypothesis-generating and tested for reproducibility in an independent, ideally multicentre, cohort before being adopted as a preoperative counselling or triage rule. If replicated, a BMI/GB-wall-thickness-based rule of this kind would be an unusually simple and low-cost addition to preoperative assessment, requiring no additional investigations beyond a weighing scale and the abdominal ultrasound already performed as standard of care.

The association between comorbidity burden (diabetes and/or hypertension) and a difficult intraoperative course is biologically plausible: chronic hyperglycaemia and vascular disease are both linked to impaired tissue healing, altered inflammatory response, and higher rates of gallbladder fibrosis, and diabetes in particular is repeatedly identified as a risk factor for a difficult or prolonged cholecystectomy in preoperative scoring systems such as that of Randhawa and Pujahari [5] and its subsequent validations [6-8]. The independent association we observed between a shorter prothrombin time and a more difficult intraoperative course is less readily explained by existing literature, which more commonly discusses coagulopathy (prolonged PT) as a marker of hepatic dysfunction or severe sepsis rather than as a predictor in the opposite direction; this counter-intuitive finding, together with the platelet-count association, should be treated as hypothesis-generating rather than confirmatory, and, together with the modest AUCs observed for both parameters (0.66 and 0.62 respectively), does not support their use as stand-alone clinical predictors at this stage. Given the modest AUC and the absence of a well-established biological mechanism, the association between shorter prothrombin time and operative difficulty should be interpreted cautiously and requires external validation.

Notably, CRP and NLR, both widely reported as predictors of complicated or difficult cholecystitis in emergency and mixed elective-emergency cohorts [9], were not significantly associated with difficulty in our predominantly elective, symptomatic (rather than acutely inflamed) population. We attribute this discrepancy in part to population selection (elective versus emergency presentation) and in part to a probable floor effect in CRP reporting: CRP values clustered heavily at a single value (4.5 mg/L) in both difficulty groups in this cohort, consistent with a substantial proportion of results being reported at or near the assay's lower limit of detection, which would be expected to blunt CRP's discriminative power in a predominantly non-acutely-inflamed elective population regardless of any true underlying biological relationship.

Strengths and Limitations

The strengths of this prospective study include the use of a structured, pre-piloted data collection proforma encompassing clinical, haematological, biochemical, and radiological variables, and the adoption of the validated Parkland Grading System (PGS) as an objective intraoperative reference standard. Additionally, the use of both cut-off analysis for strong predictors and multivariable modelling for moderate predictors enabled interpretation of findings according to their respective statistical characteristics.

This study has several limitations. It was conducted at a single tertiary-care centre and included a relatively small sample size, limiting generalisability and statistical power for infrequent outcomes. A formal a priori sample-size calculation was not performed. As an observational study, residual confounding cannot be excluded. Furthermore, Parkland Grade was prospectively documented in only a minority of cases, with most grades assigned retrospectively from operative records, introducing potential classification bias. The exceptionally high discriminatory performance observed for BMI, weight, and gallbladder wall thickness may partly reflect local case mix and requires validation in larger multicentre populations. Finally, no external validation cohort was available, and the derived cut-offs should therefore be regarded as hypothesis-generating.

Conclusions

BMI, body weight, and ultrasonographic gallbladder wall thickness were the strongest preoperative predictors of a difficult laparoscopic cholecystectomy as defined by the Parkland Grading Scale. Comorbidity burden and prothrombin time provided additional but modest predictive information. Although these findings support the potential role of simple preoperative markers in operative risk stratification, external multicentre validation is required before routine clinical implementation of the proposed cut-offs.

Patient's Informed Consent

Patient's informed and written consent taken.

Ethics Approval

Ethics approval was taken and approved by ethics committee.

Funding

No fundings were done for this research.

Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References

1. Mohite SU, Panse N, Kannan R, Tapkire R, Pradhan S, Gulia A, et al. Prevalence of gallstone disease in India: insights from a community-based cross-sectional study. BMJ Public Health. 2026;4(1):e004015.
2. Strasberg SM, Hertl M, Soper NJ. An analysis of the problem of biliary injury during laparoscopic cholecystectomy. J Am Coll Surg. 1995;180(1):101-125.
3. Madni TD, Leshikar DE, Minshall CT, Nakonezny PA, Cornelius CC, Imran JB, et al. The Parkland grading scale for cholecystitis. Am J Surg. 2018;215(4):625-630.
4. Madni TD, Nakonezny PA, Barrios E, Imran JB, Clark AT, Taveras L, et al. Prospective validation of the Parkland Grading Scale for Cholecystitis. Am J Surg. 2019;217(1):90-97.
5. Randhawa JS, Pujahari AK. Preoperative prediction of difficult lap chole: a scoring method. Indian J Surg. 2009;71(4):198-201.
6. Trehan M, Mangotra V, Singh J, Singla S, Gautam SS, Garg R. Evaluation of preoperative scoring system for predicting difficult laparoscopic cholecystectomy. Int J Appl Basic Med Res. 2023;13(1):10-15.
7. Karim ST, Chakravarti S, Jain A, Patel G, Dey S. Difficult laparoscopic cholecystectomy predictors and its significance: our experience. J West Afr Coll Surg. 2022;12(4):56-63.
8. Kumar A, Chhabra MK, Chhibber P. Preoperative prediction of difficult laparoscopic cholecystectomy using a scoring system: a prospective cohort study. Cureus. 2026;18(1):e102055.
9. Mahmood F, Akingboye A, Malam Y, Thakkar M, Jambulingam P. Complicated acute cholecystitis: the role of C-reactive protein and neutrophil-lymphocyte ratio as predictive markers of severity. Cureus. 2021;13(2):e13527.
10. von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP; STROBE Initiative. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet. 2007;370(9596):1453-1457.
11. Albert A, Anderson JA. On the existence of maximum likelihood estimates in logistic regression models. Biometrika. 1984;71(1):1-10.