Radiology report form for pulmonary nodules encourages evidence-based practice

A few modifications to the radiology order form can serve an important function in encouraging physicians to adopt evidence-based practices when ordering CT to evaluate small pulmonary nodules, according to a study from Dartmouth-Hitchcock Medical Center.

A few modifications to the radiology order form can serve an important function in encouraging physicians to adopt evidence-based practices when ordering CT to evaluate small pulmonary nodules, according to a study from Dartmouth-Hitchcock Medical Center. 

Survey results, presented during a scientific session Monday at the 2010 RSNA meeting by Dr. William C. Black, director of chest radiology at Dartmouth-Hitchcock, indicate that referring physicians generally appreciate having access to information about the appropriate use of imaging to interrogate small pulmonary nodules and are influenced by such guidelines in clinical practice.

From experience, radiologists learn that pulmonary nodules of less than 1 cm are a common anatomic abnormality. Large screening studies have shown that incidental nodules appear in the chest radiographs of up to 50% of adults in the U.S. But Black noted that referring clinicians are still often unsure about the significance and proper management of such findings. Their order patterns for multislice CT and FDG PET/CT to measure and monitor small nodules frequently fall outside the American College of Radiology guidelines for such abnormalities. Overly aggressive responses often generate unnecessary imaging costs and radiation exposure.

To address this problem, Black and colleagues at Dartmouth-Hitchcock modified a radiology order form to include concise, evidence-based information describing the appropriate use of CT for incidentally detected pulmonary nodules. An Internet survey of 1075 of medical doctors and non-M.D. medical clinicians found they preferred the modified form to a standard imaging order form.

Seven of 23 questions in the survey focused on a hypothetical 65-year-old former smoker with a 7-mm pulmonary nodule detected incidentally during a chest CT that was negative for pulmonary embolism. About 41% of the clinicians answered the seven questions.

The standard form included information about only the nodule’s size and location. The modified form also included a probability estimate for malignancy (5% to 10% for the example, based on a validated prediction model) and information about guidelines from the Fleischner Society for nodule management.

A comparison of clinician response before and after they had read the modified report revealed its immediate effect on their clinical knowledge and the influence such information could have on their practice patterns.

About 27% thought the nodule was probably lung cancer after reading the report compared with 65% before reading it. About 3% said they would request a PET/CT for further evaluation after exposure to the guidelines, compared with 14% before reading the summary. About seven of 10 referring clinicians said they would have requested a follow-up CT in three to six months before examining the form, but only about three of 10 said they would still order imaging follow-up after exposure to the guidelines.

Overall, 90% of clinicians preferred the modified report, and 79% felt it made them more confident in their response to the incidental finding.

The 41% response rate was a limiting factor for the potential validity of the survey, Black said. Physician confidence in the recommended guidelines could have been affected by the lack of a definitive probability rule and guidelines for small pulmonary nodules, he said.

A modified version of the enhanced form will appear in the future at Dartmouth-Hitchcock as a feature of its new electronic medical records system, Black said.

“With that, we will try to integrate our reporting system with medical records so we can automate calculation of the probability of malignancy based on the prediction model and also automate the recommendations from the Fleischer material,” he said. “We want, ideally, to do this not just for pulmonary nodules but for all sorts of incidental findings.