SHORT-TERM OUTCOMES FOLLOWING HARTMANN’S PROCEDURE: A MULTICENTRE TEN-YEAR STUDY WITH RISK STRATIFICATION AND PREDICTIVE MODELLING dr.aliyasen1@gmail.com

 

SHORT-TERM OUTCOMES FOLLOWING HARTMANN’S PROCEDURE: A MULTICENTRE TEN-YEAR STUDY WITH RISK STRATIFICATION AND PREDICTIVE MODELLING

1

United Kingdom

Oral only

Hartmann’s procedure carries high morbidity and mortality, yet contemporary multicentre data and practical prognostic tools remain limited. This study assessed outcomes across four centres and developed a simplified risk stratification score benchmarked against the NELA score.

A retrospective cohort study included adults undergoing emergency Hartmann’s procedures (2015–2024) across four hospitals in the UK. Preoperative and intraoperative variables were analysed using multivariable logistic regression to identify predictors of 30-day mortality and major complications (Clavien–Dindo ≥ III). Model performance was assessed using the area under the ROC curve (AUC), and a points-based HARM (Hartmann’s Assessment of Risk and Mortality) score was derived from independent predictors.

Of 623 patients (median age 69.0 years; 51.3% female), 30-day mortality was 10.1% and major complications occurred in 18.8%. Independent predictors of mortality included age (P=0.002), ASA grade (P=0.004), BMI (P=0.012), NELA score (P=0.009), and faecal contamination (P=0.013). The final multivariable model predicting mortality achieved an AUC of 0.874. The simplified HARM score, incorporating age (0–2 points), ASA (0–2), and contamination level (0–2), yielded an AUC of 0.813, outperforming NELA (AUC = 0.805) (Figure 1). Additionally, Diabetes Mellitus (P=0.026), respiratory comorbidities (0.018), ASA grade (P=0.005), and NELA score (P=0.012) were associated with major complications (Clavien–Dindo ≥ III).

This multicentre study confirms high early mortality and major complications following Hartmann’s procedure. The HARM score is a pragmatic bedside tool that enables accurate risk stratification using routinely available data. It may support clinical decision-making and targeted perioperative counselling. Prospective validation is recommended.