Prior studies of the effect of wait time (time to diagnosis or treatment) on survival in patients with non-small-cell lung cancer (NSCLC) yielded results that were mixed and confounded by several sources of bias. We aimed to describe variation in wait times, identify predictors of longer wait times, and examine the effect of wait time on survival.
We performed a retrospective cohort study by reviewing the records of consecutive patients who were diagnosed with NSCLC between 1/1/02 and 12/31/03 at VA Palo Alto Health Care System. We used multivariable statistical methods to identify independent predictors of longer wait times, and to examine the effect of wait time on survival.
We identified 129 veterans with NSCLC (mean age 67 years, 98% men, 83% white), most with adenocarcinoma (51%) or squamous cell carcinoma (30%). A minority (18%) presented with a solitary pulmonary nodule (SPN). Median time from initial suspicion to treatment was 85 days (range 3 to 1355 days). Independent predictors of longer wait times (>90 days) were hospitalization within 7 days of initial suspicion (OR 6.0, 95% CI 2.2 to 16.2), tumor size >3 cm (OR 5.4, CI 2.1 to 14.1), chest x-ray findings other than the primary tumor (OR 2.6, CI 0.9 to 7.5), and any symptom (OR 2.5, CI 1.0 to 6.0). In multivariable analyses, wait time was not associated with survival (HR 1.5, CI 0.9 to 2.5). In patients with SPN, there was a trend toward improved survival when wait time was <90 days (figure).
Wait times for our patients with NSCLC were often longer than recommended. Patients with more severe presentations of NSCLC had shorter wait times, confounding the relationship between wait time and survival. Our results suggest the hypothesis that in patients with malignant SPN, survival may be better in patients treated within 90 days.
If confirmed, our results suggest that quality improvement efforts should focus on reducing wait times in patients with solitary nodules and others with potentially resectable NSCLC.
Michael Gould, None.