Study objectives: The physicians who initially evaluate patients with non-small cell lung cancer (NSCLC) strongly impact the course of therapy. Their beliefs in treatment and prognosis may contribute to practices of variable quality and appropriateness. We sought to better describe beliefs among pulmonologists and thoracic surgeons who were selected for guiding early therapy and referrals in patients with NSCLC.
Design: Mail questionnaire focusing on survival estimates, treatment perceptions, and referral patterns.
Participants: Twelve hundred pulmonologists and 800 thoracic surgeons who were clinically active members of the American College of Chest Physicians.
Measurements and results: Response rates of 50% for pulmonologists and 52% for thoracic surgeons were obtained after two mailings. Five-year survival estimates for patients with resected stage I NSCLC revealed that 30% of respondents overestimated survival rates and 18% underestimated survival rates. The underestimation of survival rate was found among more respondents who are practicing pulmonology than thoracic surgery (22% vs 10% [corrected], respectively), who were trained before 1980 than after 1980 (29% vs 10% [corrected], respectively), and who were seeing < 10 lung cancer patients annually than those who were seeing > 25 (31% vs 0.14%, respectively). Beliefs in the survival benefit of adjuvant chemotherapy or of radiation in stage I-IIIA disease divided respondents within both specialties. Chemotherapy plus radiation vs radiation alone in unresectable stage IIIA-B NSCLC was viewed as benefiting survival less often by physicians seeing < 10 lung cancer patients annually rather than > 25 (57% vs 77% [corrected], respectively) and by physicians underestimating rather than correctly estimating survival in early-stage disease (58% vs 72% [corrected], respectively). Chemotherapy was believed to confer survival benefits in patients with stage IV disease by one third of respondents.
Conclusions: Certain physician characteristics, particularly the length of time since training and NSCLC patient volume, are associated with beliefs not conclusively supported in the medical literature or with opinions inconsistent within and between specialties.