• Derek Weycker, James Hackett, John S Edelsberg, Gerry Oster, and Andrew G Glass.
• Policy Analysis Inc., Brookline, MA 02445, USA. dweycker@pai2.com
• Ann Pharmacother. 2006 Mar 1; 40 (3): 402-7.

BackgroundIn clinical trials in patients receiving myelosuppressive chemotherapy, 10-11 days of prophylaxis with filgrastim has been found to reduce the incidence of febrile neutropenia. In clinical practice, however, many patients receive shorter courses of therapy, even though the effectiveness of this regimen is unknown.ObjectiveTo examine the relationship between duration of filgrastim prophylaxis and risk of hospitalization in patients receiving chemotherapy for non-Hodgkin's lymphoma (NHL), breast cancer, or lung cancer.MethodsUsing a large, automated, US healthcare claims database, we identified all adults who received chemotherapy for NHL, breast cancer, or lung cancer between 1998 and 2002. For these patients, we identified their first course of chemotherapy and each unique cycle within that course. We then focused attention on all patient cycles in which filgrastim was administered on or before cycle day 5 (filgrastim prophylaxis). Pooling all such cycles, we examined the relationship between duration of filgrastim prophylaxis and risk of hospitalization for neutropenia or infection and risk of hospitalization for any reason, using generalized estimating equations.ResultsMean +/- SD duration of filgrastim prophylaxis was 6.5 +/- 3.1 days across 332 cycles for 133 NHL patients, 6.1 +/- 2.9 days across 482 cycles for 205 breast cancer patients, and 4.3 +/- 3.1 days across 522 cycles for 260 lung cancer patients. In multivariate analyses, risk of hospitalization for neutropenia or infection was found to decline with each additional day of filgrastim prophylaxis for patients with NHL (OR 0.81; p = 0.003), breast cancer (OR 0.77; p = 0.001), and lung cancer (OR 0.91; p = 0.084). Risk reductions with each additional day of prophylaxis ranged from 15% to 19% for patients with NHL, 17% to 23% for those with breast cancer, and 8% to 9% for those with lung cancer. Similar reductions in risk were noted for all-cause hospitalization.ConclusionsAmong patients with NHL, breast cancer, or lung cancer, shorter courses of filgrastim prophylaxis may increase the risk of hospitalization.

37 keywords...

hide…