Increased Body Mass Index is Associated With Increased Cost for Primary Total Knee Arthroplasty Exclusive of...

Surgeons participating in alternative payment models may encounter financial disincentives in caring for high-cost patients. While prior studies have shown a relationship between body mass index (BMI) and cost in total knee arthroplasty (TKA), this question has not been examined using recent data in a nationally-representative dataset. We sought to use recent data to assess the relationship between BMI and cost in TKA.

We queried the 2016-2022 PINC AI Healthcare Dataset for primary TKAs identified by the Current Procedural Terminology code 27447. Baseline patient demographics, characteristics, and outcomes were gathered. Patient costs included index and 90-day period costs starting from the surgical encounter. A multivariate generalized linear model estimated patient costs across eight BMI categories (WHO BMI categories, with BMI >40 grouped into BMI 40-45, BMI 45-50, and BMI >50), and a linear regression model assessed the effect of BMI on costs. Both models controlled for age, sex, payer, race, ethnicity, and Elixhauser Comorbidity Index.

A total of 24,064 primary TKAs were included. Patient characteristics were age (mean 68 years, SE 0.06), sex (62.3% female), payer (63.8% Medicare), race (88.6% white), ethnicity (3.1% Hispanic), and Elixhauser Comorbidity Index (mean 2.07, SE 0.01). The mean BMI was 32.0 (SE 0.1), with mean index and 90-day costs of $14,051 ± $32.6 and $17,377 ± 107.3, respectively. Predicted patient index and 90-day costs were both highest in the BMI >50.0 category at $15,555 and $20,716, respectively. Predicted index and 90-day costs were lowest in the 25.0-29.9 ($13,765) and <18.5 ($16,403) categories, respectively. The multivariate analysis showed that patient index and 90-day costs increased by $23.20 (P < .001) and $48.10 (P < .01), respectively, per one-point increase in BMI.

Higher patient BMI is associated with increased index and 90-day episodic costs for primary TKA. Stakeholders may consider incorporating this information in bundled payment models and other value-based reimbursement models.