Validity of diagnostic coding for undernutrition in hospitals

Area of Science:

• Clinical Nutrition
• Health Informatics
• Hospital Administration

Background:

• Accurate diagnosis coding is crucial for patient care and health system management.
• Undernutrition is a significant health concern in hospitalized patients.
• The International Classification of Diseases, 10th Revision (ICD-10) is widely used for diagnostic coding.

Purpose of the Study:

• To evaluate the accuracy of ICD-10 diagnostic codes for undernutrition in Danish hospitals.
• To assess the concordance between ICD-10 codes and the Nutritional Risk Screening 2002 (NRS-2002) guidelines.
• To determine the positive predictive value (PPV) of undernutrition diagnoses in hospital discharge records.

Main Methods:

• A random sample of hospitalized patients with an undernutrition diagnosis (2002-2011) from the North Denmark Region was reviewed.
• Medical records were analyzed to determine the PPV of the undernutrition diagnosis.
• The study stratified analyses by calendar period, hospital type, patient demographics, and specialty.

Main Results:

• The positive predictive value for any confirmed undernutrition was 70.9% (95% CI: 63.5-77.6).
• The PPV for definite (screening-confirmed) undernutrition was only 11.0% (95% CI: 6.8-16.7).
• Only 26.2% of patients with an undernutrition code had been screened using the NRS-2002.

Conclusions:

• There is modest agreement between ICD-10 codes for undernutrition and the NRS-2002 standard in Danish hospitals.
• Hospital discharge registry data for undernutrition should be interpreted with caution.
• Improved diagnostic practices and adherence to screening guidelines are recommended.