Developing a Nomogram for Assessing Nutritional Status in Surgical Patients at Kenyatta National Hospital

Abstract

Background: Poor nutritional status in the background of a disease process is associated with a prolonged hospital stay, increased economic burden, and higher morbidity and mortality. Robust nutritional assessment tools exist which have varying accuracy and cost implications. The use of a combination of red cell distribution width, total lymphocyte count and body mass index (CORL-BMI), could be a reasonable alternative in a resource-limited center for assessment of nutritional status. Objective: This study aimed to develop a tool (CORL-BMI) that can be used in the nutritional assessment of surgical patients at Kenyatta National Hospital (KNH). Methodology: This was a hospital-based cross-sectional analytic study that was carried out in the surgical wards and theatres (minor and main theatres) of KNH from March 2020 to June 2020. Patients above the age of eighteen years, who met the inclusion criteria and consented, were conveniently recruited. The data collection tool captured; social demographic, clinical information like weight, height and laboratory indices like TLC, CBC and pre albumin levels of participants. Patient characteristics that were categorical were presented as frequencies and percentages, while those that were continous were presented as means with standard deviations. The association between RDW-SD, TLC, and BMI with pre-albumin levels, as well as the association between CORL-BMI and pre-albumin levels was performed using Pearson Correlation. The sensitivity and specifity of CORL-BMI was performed with the use of ROC curve. The prevalence of malnutrtion was determined as a proportion of those with Pre-Albumin levels of less than 18 over the total sample size of the study, and reported as a percentage. The nomogram for assessing risk of malnutrition was generated with the use of R Statistical Software version 3.6.1 after running a logistic regression model to predict the probability of malnutrition. All statistical tests were considered significant where p<0.05. Results: The correlation of CORL-BMI with pre-albumin is moderate. The value of the area under the curve (AUC) has archived statistical significance with p-value < 0.05 with favorable sensitivity and specificity, 74.1% and 21.4% respectively. Conclusion: This study has demonstrated that utility of a combination of laboratory parameters (TLC and RDW-SD) and physical examination findings (BMI) to generate a nomogram (CORL-BMI) could be used in assessment of nutritional status in resource-limited settings. Significance: This study introduces a novel alternative tool (CORL-BMI) that could be used for assessing nutritional status in surgical patients that is cheaper and easy to use especially in resource-limited settings.