| Sumario: | Simple Summary: This study linked Japan's Center for Cancer Genomics and Advanced Therapeutics (C-CAT) registry with facility-level specialist records to examine first-line treatment duration in lung cancer. At facilities with two or more registry-listed Japanese Society of Medical Oncology specialists, the estimate pointed toward longer time to treatment failure, but the range of estimates after accounting for clustering within facilities included no association. A secondary analysis used availability of physicians listed by both the Japanese Society of Medical Oncology and the Japanese Respiratory Society as a lung-cancer-oriented indicator. Here, time to treatment failure reflects duration of a recorded treatment process; it does not show why treatment ended or whether a longer duration was clinically preferable. Specialist availability may serve as a facility-level indicator of oncology-care infrastructure. By demonstrating how national genomic-registry and workforce data can be linked, this study identifies the patient-level, treatment, and follow-up information needed in future national chemotherapy databases. Japan's Center for Cancer Genomics and Advanced Therapeutics (C-CAT) provides a national platform for studying genomic-medicine-era care. We conducted a nationwide registry/database study to examine whether facility-level Japanese Society of Medical Oncology (JSMO) specialist availability was associated with first-line time to treatment failure (TTF) in lung cancer. The primary exposure was treating-facility JSMO specialist count (0–1 vs. ≥2); dual JSMO/Japanese Respiratory Society availability was a secondary lung-cancer-oriented exposure. First-line TTF was defined from systemic therapy start to recorded first-line treatment end or death, whichever occurred first. The cohort included 5456 patients at 241 facilities: 1477 in the 0–1 group and 3979 in the ≥2 group. Kaplan–Meier estimated median first-line TTF was 4.2 versus 5.1 months (log-rank p < 0.001). In the full clinical adjustment model, the HR for ≥2 versus 0–1 specialists was 0.911 (facility-cluster robust 95% CI, 0.806–1.029; robust p = 0.133). The point estimate was in the direction of longer recorded treatment-process duration, but the robust confidence interval included no association. Linking C-CAT with facility-level workforce data demonstrates a health-services use of national genomic registry data and supports the development of chemotherapy-specific databases with patient-level, time-anchored clinical information.
|
|---|