Kotapati S, Mitra D, Davis KL, Iloeje U. Predictors of treatment choice in high-risk and metastatic melanoma: evidence from linked electronic medical records and administrative claims data. Poster presented at the 2008 ISPOR 13th Annual International Meeting; May 7, 2008. [abstract] Value Health. 2008 May; 11(3):A82-3.

OBJECTIVE: Evaluate predictors of four major therapeutic choices (surgery, radiation, chemotherapy, immunotherapy) in high-risk (stage IIB/C, III) and metastatic (stage IV) melanoma.

METHODS: Data were acquired from Convergence CT, a company that links longitudinal electronic medical records and claims data from large physician practices, clinics, ambulatory centers, and hospitals in the US. Subjects with ≥1 diagnosis of malignant melanoma (ICD-9 172.xx, 173.xx, V10.82) from July 1, 2003 November 30, 2006 and pathology-confirmed disease stage of IIB/C, III, or IV were selected. Additional stage IV patients were identified based on evidence of a subsequent ICD-9 code (197.xx, 198.xx) for secondary metastases. Post-diagnosis prevalence of the key treatments was analyzed descriptively. Logistic regression was used to assess predictors of therapeutic choice.

RESULTS: A total of 268 subjects were identified. Stage distribution was: IIB/C (18%); III (21%); IV (61%). 58% were ≥ 65 years of age and 62% were male. Surgery was the predominant treatment in stage IIB/C and III (received by >80% of subjects), but was seen in only 38% of stage IV patients. Across all stages, radiation, chemotherapy, and immunotherapy were less common (23%, 27%, and 10%, respectively). Being elderly [odds ratio = 2.19; 95% CI = (1.10–4.35)] and having stage IV disease [7.31 (2.38–22.39)] was associated with a significantly increased likelihood of receiving no active treatment. Older age (65+), higher co-comorbidity burden, and having stage IV disease were associated with a decreased probability of surgery [0.55 (0.30–0.99), 0.92 (0.86–0.99), 0.08 (0.03–0.22), respectively]. Receiving radiation was reduced by older age, but increased by having stage IV disease [2.38 (0.91–6.22)]. Significant predictors of chemotherapy were stage IV disease [2.65 (1.01–6.93)] and higher co-morbidity burden [1.08 (1.01–1.17)]. Finally, increasing age substantially reduced the likelihood of receiving immunotherapy [0.24 (0.10–0.60)].

CONCLUSION: Factors influencing practice patterns and treatment choice in a population with high risk or metastatic melanoma. Across therapeutic choices, age and disease stage were the significant predictors.