Validity of the Adapted Waterlow Score as a Tool in Predicting Adverse Outcomes in Acute Pancreatitis When Compared With the Ranson Score and Serum CRP Levels

Introduction Acute pancreatitis (AP) is one of the most common emergency room presentations, with an annual worldwide incidence of 150-420 cases per million and a 30-50% mortality rate. Unfortunately, predicting prognosis in AP patients at an early admission phase is difficult. The Waterlow score is a semi-quantitative composite score devised to stratify the risk of developing pressure sores in in-patient settings. Components such as age, gender, nutritional status, body mass index, mobility, smoking status, comorbidities, surgery, use of medications, neurological status, and continence, all combine to generate a single numerical figure between 2 and 64 to overcome the difficulty of the composite scoring systems such as the Ranson score, Glassgow score, Physiological and Operative Severity Score for the enUmeration of Mortality and morbidity (POSSUM) and its modification (Portsmouth or P-POSSUM), Simplified Acute Physiology Score (SAPS 2), Acute Physiology and Chronic Health Evaluation II (APACHE II) score, and Bedside index for severity in acute pancreatitis (BISAP). Aim This study aimed to validate the Waterlow score as a surrogate marker in predicting in-patient mortality, length of hospital stay (LOHS), intensive care unit (ICU) admissions, and AP complications compared to the Ranson score and serum CRP levels. Method A prospective case series analysis was conducted from August 2018 to November 2019 at a tertiary care center in central Gujarat. This study included 51 newly diagnosed patients with AP or acute on chronic pancreatitis on accrual. Out of 51 patients, four were excluded from our study, as three were discharged against medical advice (DAMA) and one other absconded from the ward premises. The study group comprised 47 patients diagnosed with AP, out of which 40 patients were male (85.1%) and seven were females (14.9%). The mean age of the population was 37.5 ± 13.0 years. Microsoft Excel calculated the standard deviation and mean for data entry. All statistical analyses were made with MedCalc software version 19.1.3 (MedCalc Software Ltd., Ostend, Belgium). Result When the Waterlow score, Ranson score, and serum C-reactive protein (CRP) levels were plotted against inpatient mortality, the area under the curve (AUC) was found to be 0.85, 0.80, and 0.98, respectively. The plot of CRP lay closest to the vertical axis. On plotting the receiver operator characteristic (ROC) curve of all three markers of AP severity, the area under the graph was calculated to be 0.81 (Waterlow score), 0.81 (Ranson score), and 0.82 (serum CRP), respectively. Nearly all three markers performed similarly. Waterlow score offers a high sensitivity but poor specificity in sorting outpatients at risk of adverse outcomes. It was observed that the Waterlow score could significantly predict mortality, ICU admissions, and LOHS in this study population. In contrast, the Ranson score and serum CRP levels could not indicate LOHS. A risk assessment tool based upon a subset of Waterlow parameters with additional variables may perform better than a tool based solely on Waterlow risk factors. Conclusion This study showed no correlation between the Waterlow score, the rate of local complications (p = 0.9513), and the incidences of surgical interventions (p = 0.9915). Therefore, a risk assessment tool based upon a subset of Waterlow parameters with additional variables may perform better than a tool based solely on Waterlow risk factors.