Year 1 Data
Population-based modeling prior to program implementation predicted 1027 nodule patients in Year 1, of which 125 would require diagnostic intervention. Approximately 6000 CT scans were reviewed annually at the medical center during the study timeframe (screening and incidental). In Year 1, 1792 patients with lung nodules requiring further evaluation were identified. After interventional pulmonology review, 665 patients were entered into the nodule database (indicated with an “A” in Figure 1), of which 607 were managed by the nodule program and required either biopsy or follow-up. Fifty-eight patients were followed in the database but managed by their primary care physician or an outside pulmonologist. The remaining 1127 were ruled out after CT scan review, due to lack of significant findings or nodule stability for ≥2 years in previous scans.
A total of 182 patients underwent a diagnostic intervention (indicated with a “B” in Figure 1). These included classically defined lung nodules (density on imaging <3 cm) and mass lesions and patients with concomitant airway lesions, adenopathy, pleural effusion, etc. Sixty percent of the referred nodules were incidental, 17% were symptomatic, and 12% were identified through screening (Figure 2A). During the phase-in period, 63 nodule referrals were received. The program grew by quarter in Year 1, with 129 referrals in the first quarter of full program implementation, followed by 139, 151, and 188 in the subsequent three quarters.
Most diagnostic procedures were ENB (46%) or linear endobronchial ultrasound-guided biopsy (42%), typically conducting during the same procedure, with a relatively even distribution after the first quarter (Figure 2B). In Year 1, only 7 patients (7/182; 3.8%) required a second separate procedure to obtain a diagnosis, complete staging, and/or aid in fiducial marker placement.
Of the 182 cases undergoing diagnostic intervention, 121 received a cancer diagnosis (119 primary and 2 metastatic [pancreatic and renal]). This represents 66.5% (121/182) of nodules undergoing diagnostic intervention and 18.2% (121/665) of all nodules in Year 1. While referrals increased by quarter, the number of diagnostic procedures (47 vs. 44) and cancer diagnoses (35 vs. 34) remained stable between the first and last quarters of Year 1.
Early-stage cancer was detected in 35.5% (26.4% Stage 1, 9.1% Stage 2) while late-stage cancer was detected in 64.4% (26.4% Stage 3, 38.0% Stage 4; Figure 2C). This compares favorably to 2012 Medicare baseline data of 23% early-stage cancer [17]. Treatment modalities are shown in Figure 2D.
The remaining 452 patients without diagnostic intervention were placed into “watchful waiting”. After follow-up imaging, 9/452 (2%) required diagnostic intervention of which 5 (5/452; 1.1%) had cancer, 3 had negative biopsies and remain in follow-up with no progression, and 1 had a pseudomonas infection that responded to treatment. A total of 443 patients remain in monitoring with no progression noted on follow-up imaging.
Program Data Through Two Years
Cumulative data for the first two years of the program is shown in Figure 3. The average days to treatment decreased steadily in Year 1, from 41 days in the phase-in period to 36 days in the first quarter and 28 days in last quarter. An increase to 42 and 43 days was observed in the fourth quarter of 2017 through the first quarter of 2018, dropping to 30 days in the second quarter, and rising again to 52 days in the third quarter of Year 2 (Figure 3A). Potential reasons for these observations are detailed in the Discussion.
The number of referrals increased throughout the first 5 quarters and then levelled off slightly during Year 2 (Figure 3B). The average proportion of incidental nodules increased from 60% in Year 1 to 65% in Year 2. Among the 745 patients entered into the database in Year 2, 169 had diagnostic interventions after the initial scan. Four patients required a second procedure to complete staging or obtain a diagnosis. The remaining 576 patients were placed into watchful waiting. With follow-up scans, 5/576 (0.9%) required diagnostic intervention, of which 4/5 had cancer and 1/5 had a negative biopsy. A total of 572 remain in monitoring.
The relative proportion of early-stage lung cancer detection was higher than observed prior to the implementation of the lung nodule program (Figure 3C). The diagnosis occurred at Stage 1-2 in an average of 23% of patients in the year 2012 (before program implementation), compared to 36% in Year 1 and 38% in Year 2. However, most lung cancer diagnoses remained late-stage; potential reasons are discussed below.
Screening Program Data
The number of new screening scans increased steadily each quarter, with 255 in Year 1 and 549 in Year 2. Patients with screening-detected nodules that required more than yearly CT follow-up were moved to the active database (29% in Year 1 and 20% in Year 2). Among first-year patients, 71% had follow-up scanning within 18 months of the initial scan.
Financial Data
The total revenue generated by the nodule program in Year 1 was $2,193,994.81 (Table 1). Two-thirds of that revenue was generated by video-assisted thoracoscopic surgery, radiation oncology, and medical oncology treatment procedures, approximately 20% by imaging, and 15% by bronchoscopic or transthoracic biopsy. Revenue increased steadily through 2 years after program implementation, with $363,504 generated in the first quarter, increasing to a cumulative total of $3,812,461 at 18 months (Figure 3D).