Predicting
LARS
with
POLARS
: Results of a multicentre study
Michail Klimovskij, Richard Fortunato, Charlotte Florance, Guoda Blazyte, Austeja Cepeniene, Justas Zilinskas, Manal Mubarak Alquaimi, Narimantas Evaldas Samalavicius, Alharbi A. Abdullah, Sami Benli, Tahsin Çolak, Ugur Sungurtekin, Utku Ozgen, Vitaliy Groshilin, Julio del Castillo‐Diego, Lidia Cristóbal‐Poch, Audrius Dulskas Abstract
Background
Rectum‐preserving colorectal surgery, with low or ultra‐low anastomosis, became the standard of care with advances in laparoscopic and now robotic surgery. As a downside, it is not uncommon for function to be affected, causing low anterior resection syndrome (LARS), which is a common postoperative challenge.
Aim
To determine the correlation of using the predictive preoperative POLARS scores with actual postoperative LARS scores for patients undergoing rectal resections with low pelvic anastomosis.
Methods
A multicentre, prospective study evaluating LARS was conducted with patients undergoing rectal resection with restoration of bowel continuity. Defaecatory function was assessed preoperatively using the POLARS scoring questionnaire and 6 months after the surgery using the LARS scoring questionnaire. The scores were then compared to determine any correlations.
Results
A total of 310 patients from 13 centres were enrolled, of whom 95 had complete paired POLARS and postoperative LARS score data. POLARS predicted minor LARS in 69 patients (72.6%), whereas 25 patients (26.3%) reported minor LARS postoperatively. Major LARS was predicted by POLARS in 21 patients (22.1%), whereas 24 patients (25.3%) self‐reported major LARS postoperatively. Only 7 of these 24 major‐LARS cases (29.2%) were correctly categorized as major LARS by POLARS. For detecting major LARS, POLARS showed sensitivity of 29.2%, specificity of 80.3%, positive predictive value of 33.3% and negative predictive value of 77.0%. Spearman correlation between predicted and observed scores was weak but statistically significant (rho = 0.21, p = 0.042). Bland–Altman analysis demonstrated a mean bias of +7.99 points with 95% limits of agreement from −19.23 to 35.20, reflecting systematic overprediction.
Conclusions
POLARS showed weak positive correlation with postoperative bowel function and demonstrated low sensitivity but moderate specificity for major LARS. It systematically overestimated LARS severity in many patients. While POLARS may help identify some individuals at elevated risk, it is insufficient as a stand‐alone tool for accurate prediction of postoperative LARS severity and should be combined with clinical judgement and patient‐reported follow‐up.