Nonclinical Factors Predict Selection of Initial Observation for Renal Cell Carcinoma

To determine the rate of observation utilization over time and to identify factors influencing its use.

Using the National Cancer Data Base, we studied observation utilization in patients diagnosed with localized renal cell carcinoma from 2003 to 2010. Relationships between temporal, demographic, provider, and clinical factors and the likelihood of observation were evaluated using multivariate logistic regression.

Of 109,410 analyzed patients, 7047 (6.4%) underwent observation with stable use over time (range, 6.1% to 6.8%). Patient and disease factors were the strongest predictors of observation. Specifically, the odds of biopsy were 1.8-11 times higher for elderly or comorbid patients and 1.6-8.4 times higher for small (clinical T1a), biopsied, or bilateral tumors (P <.01 for all). Racial and socioeconomic factors also significantly predicted observation usage. In particular, observation rates were higher among poor, African American, and uninsured or socially insured patients, with these groups having 1.2-3.5 times higher odds of observation (P <.01). Patients receiving care at community, low-volume, or nearby hospitals were also significantly more likely to undergo observation (P <.01).

Despite the continued rise in the incidence of incidental renal masses, initial observation use has remained stable. In accordance with treatment guidelines, observation is preferentially utilized in elderly and comorbid patients. However, nonclinical factors also predict observation use, suggesting that utilization may be influenced by racial and socioeconomic disparities in health care quality.