Quick and Easy Is Not without Cost: Implications of Poorly Performing Nutrition Screening Tools in Hip Fracture
Abstract
Objectives
To evaluate the performance of commonly applied nutrition screening tools and measures and to consider the potential costs of undiagnosed malnutrition in a case-based reimbursement funding environment.
Design
A diagnostic accuracy study to compare a variety of nutrition screening techniques against primary, secondary, and comparative measures of nutritional status.
Setting
Public metropolitan hospital orthogeriatric unit.
Participants
Individuals with acute hip fracture admitted to the orthogeriatric unit; 150 prospective, consecutively admitted individuals were considered, with eight exclusions, yielding a sample size of 142 participants.
Measurements
Screens included the Mini Nutritional Assessment—Short Form, Malnutrition Screening Tool, Malnutrition Universal Screening Tool, Nutrition Risk Screen 2002, anthropometric measures, and albumin. Malnutrition was diagnosed using International Statistical Classification of Diseases and Health Related Problems, Tenth Edition, Australian Modification (ICD-10-AM) criteria. Healthcare coders costed malnutrition-related diagnostic related groups and cost-weight changes for individual participants.
Results
Malnutrition prevalence was 48%. Screening tools had only slight to moderate agreement with ICD-10-AM diagnosis of malnutrition, and none of the screening tools tested met the a priori requirement of 80% sensitivity and 60% specificity. The estimated cost effect of poor screening tool sensitivity on a 16-bed hip fracture unit ranged from AUS$46,506 to AUS$228,896 per year.
Conclusion
Poor screening tool sensitivity leads to undiagnosed malnutrition; tools that are quick and easy to apply are not without cost. Routine nutrition assessment should replace nutrition risk screening in hip fracture settings with a high prevalence of malnutrition reliant on case-mix funding. Further pragmatic studies are urgently required to determine whether findings apply to other elderly inpatient populations with endemic malnutrition, comorbidities, and cognitive impairment.