(P04-006-23) Rate and Accuracy of Coded Malnutrition Diagnosis Among Hospitalized Adult Patients Following Implementation of a Malnutrition Quality Improvement Initiative Program

Objectives: This study compared coded malnutrition diagnosis usage among hospitalized adult patients before and after a Malnutrition Quality Improvement Initiative (MQii) program began. Another component investigated patient outcomes and clinician agreement on type/severity of documented malnutrition following program implementation.

Methods: This retrospective electronic medical record (EMR)-based study primarily involved adult patients admitted to an acute care hospital between January and June 2017, who were identified by a registered dietitian (RD) as meeting ≥ 2 of 6 Academy of Nutrition and Dietetics/ American Society for Parenteral and Enteral Nutrition criteria for malnutrition. Data collected included type of malnutrition noted by RD and physician (MD), timeliness of care, length of stay (LOS), and 30-day readmission. Descriptive statistics were used to evaluate clinician agreement on malnutrition severity, compliance with quality measures, and patient outcomes. Additionally, separate de-identified data sets were obtained showing the overall frequencies of coded malnutrition diagnosis among all patients admitted during the 2017 study period, as well as over a 6-month period one year earlier (January to June 2016), for comparison of code usage before and after program initiation.

Results: Use of inappropriate codes (nutritional marasmus) and those with unspecified severity decreased after the MQii program began. For 191 patients with RD-identified malnutrition, 83% had a concurrent MD-validated diagnosis; however, agreement in severity only occurred about half of the time. In 74% of cases with discrepancies, the coded diagnosis was less severe than noted by the RD. Average LOS was 8.3 days. Prolonged LOS ≥ 7 days occurred among 45.6% of patients. The 30-day readmission rate was 21.5%.

Conclusions: High quality care was demonstrated by decreased use of inappropriate and non-specific malnutrition codes after MQii program initiation, good protocol compliance for timely care, and high rates of MD validation of RD-identified malnutrition. However, areas that should be targeted for further improvement include clinician agreement on malnutrition severity, as well as increased resource allocation to the coding department to enable effective data capture and performance analysis.

Funding Sources: Department of Dietetics and Nutrition