3046 — Evaluation of Potentially Malignant Pulmonary Nodules in the VA System: Resource Use and Adherence to Guideline Recommendations
Wiener RS, CHQOER, Bedford VAMC, MA; Gould MK, Kaiser Permanente, Southern California; Slatore CG, Portland VAMC; Kneeland T, Dartmouth-Hitchcock Medical Center; Hickey E, CHQOER, Bedford VAMC, MA; Welch HG, Schwartz LM, and Woloshin S, VA Outcomes Group, White River Junction VAMC, VT;
Each year thousands of Veterans are diagnosed with a potentially malignant pulmonary nodule; nodule evaluation may create a substantial burden on the VA system. We sought to develop an algorithm to identify a sample of Veterans with newly diagnosed pulmonary nodules and to track resource utilization attributable to nodule evaluation.
We created VISTA-based algorithms to identify patients with a pulmonary nodule <3 cm at the White River Junction (WRJ) and Boston VAMCs by searching chest imaging reports for terms like “nodule” or “nodular opacity.” We developed a structured data abstraction tool and used it in a randomly selected subset of 292 Veterans to quantify resource utilization related to nodule evaluation in the subsequent 3 years. We terminated data abstraction early if the nodule disappeared or was determined to be malignant. We categorized management as adherent to guidelines if it was consistent with the Fleischner Society’s recommendations for nodule evaluation.
The algorithms identified 493 WRJ Veterans and 1875 Boston Veterans with potential pulmonary nodules on chest imaging studies from 2006, with a positive predictive value of 78% at WRJ and 72% at Boston. Among the 292 Veterans whose charts were reviewed in detail, 59% had subcentimeter nodules. 7% of these 292 Veterans turned out to have lung cancer; another 31% had a nodule that resolved. Evaluation incurred substantial resource use, including 821 imaging studies (193 x-ray, 597 CT, 31 PET), 122 consults, 67 invasive procedures, and 20 hospitalizations directly related to nodule evaluation. Management deviated from guidelines in 40% (more aggressive 16%, including 9% who underwent prolonged surveillance >3 years; less aggressive 24%, including 3% with no evaluation). Factors contributing to deviation from guidelines included inappropriate radiologist recommendations (17%, p <0.001) and patient refusal (3%, p <0.001).
Nodules are commonly detected among Veterans, their evaluation places a burden on the VA system, and VA practice in evaluating pulmonary nodules is often inconsistent with guideline recommendations.
Simple algorithms can identify Veterans with pulmonary nodules to facilitate research in a larger VA cohort. Our study suggests that nodule evaluation is an area that can be targeted for quality improvement within the VA.