Using a Modified Delphi Panel to Estimate Health Service Utilization for Patients with Advanced and Non-Advanced Systemic Light Chain Amyloidosis

Purpose: Patients with diagnosed with systemic light chain (AL) amyloidosis at advanced Mayo stages have greater morbidity and mortality than those diagnosed at non-advanced stages. Estimating service use by severity is difficult because Mayo stage is not available in many secondary databases. We used an expert panel to estimate healthcare utilization among advanced and non-advanced AL amyloidosis patients.

Patients and methods: Using the RAND/UCLA modified Delphi method, expert panelists completed 180 healthcare utilization estimates, consisting of inpatient and outpatient visits, testing, chemotherapy, and procedures by disease severity and organ involvement during two treatment phases (the 1 year after starting first line [1L] therapy and 1 year following treatment [post-1L]). Estimates were also provided for post-1L by hematologic treatment response (complete or very good partial response [CR/VGPR], partial, no response or relapse [PR/NR/R]). Areas of disagreement were discussed during a meeting, after which ratings were completed a second time.

Results: During 1L therapy, 55% of advanced patients had ≥1 hospitalization and 38% had ≥2 admissions. Rates of hematopoietic stem cell transplant (HSCT) in advanced patients were 5%, while pacemaker or implantable cardioverter defibrillator (ICD) placement were 15%. During post-1L therapy, rates of hospitalization in advanced patients remained high (≥1 hospitalization: 20-43%, ≥2 hospitalizations: 10-20%), and up to 10% of advanced patients had a HSCT. Ten percent of these patients underwent pacemaker/ICD placement.

Conclusion: Experts estimated advanced patients, who would not be good candidates for HSCT, would have high rates of hospitalization (traditionally the most expensive type of healthcare utilization) and other health service use. The development of new treatment options that can facilitate organ recovery and improve function may lead to decreased utilization.