The Role of Financial Toxicity Components on Long-Term Trauma Outcomes: What Matters Most?
*Saba Ilkhani MD MPH1, *Tanujit Dey phD2, *Ben Grobman 2, *Wardah Rafaqat 5, *Sabrina Sanchez 6, *John Hwabejire 5, *Kavitha Ranganathan 3, *John W. Scott 4, *Juan p P. Herrera-Escobar 1, Ali Salim MD1, *Geoffrey Anderson MD MPH1
1Trauma, Burn, and Surgical Critical Care, Brigham and Women's hospital, Boston, MA; 2Harvard Medical School, Boston, MA; 3Division of Plastic and Reconstructive Surgery, Brigham and Women's hospital, BOSTON, MA; 4University of Washington, Seattle, WA; 5Division of Trauma, Emergency Surgery & Surgical Critical, Massachusetts General Hospital, Boston, MA; 6Boston Medical center, Boston, MA
Introduction: Defining financial toxicity (FT) and determining the best metric to measure financial complications of healthcare conditions is a complex challenge in the field of health services research.Our objective was to assess the predictive power of the variables that we used for identifying FT after traumatic injuries
Study design: We interviewed adult trauma patients with an Injury Severity Score of ? 9 who had received care at one of three level-1 trauma hospitals 6-12 months after their injury. Then, we used five exposure variables, including the following: 1) No care: did not receive care due to the cost in the previous year 2) Non-medical financial problems: difficulties in paying monthly bills, household costs, and credit card minimums after hospitalization.3) Government assistance: if the patient had recently requested or qualified for it, 4) A decrease in income compared to pre-injury levels.; and 5) OOP/Income: The proportion of the annual median income allocated to out-of-pocket expenses (OOP). Long-term outcomes were assessed using PROMIS physical and mental health scores. Best subset regression followed by a machine learning tool called Random Forest was conducted to find the best combination and quantify the predictive value of the variables.
Results: = A total of 1,217 patients were enrolled, among whom 6.9% reported foregoing care due to lack of funds, 17.4% reported applying for or qualifying for governmental assistance after injury, 22.6% experienced income loss following injury, 21.7% faced non-medical financial problems. The most significant variable associated with the physical health outcome was government assistance (p-value=0.008). The size three model, which included government assistance, income loss, and OOP/ Income, was the most robust overall model related to physical health. On the other hand, the most significant associated variable for mental health was non-medical financial problems (p-value<0.001). And the best overall model associated with mental health was the model with all five variables involved. Non-medical financial problems and governmental assistance were ranked as the best predictors for mental and physical health respectively.
Conclusion: In summary, the five financial questions we adopted were associated with overall poor health both (a) individually and (b) cumulatively. These findings represent initial steps towards targeted screening interventions, enhancing our comprehension of these intricate relationships and their influence on comprehensive health.
Back to 2024 Abstracts