What is the impact of incidental findings (IFs) on diagnostic imaging examinations on hospital inpatients admitted for chest pain? A study from Johns Hopkins University School of Medicine in Baltimore investigated this question by conducting an analysis of 376 inpatients. The analysis, published in the Journal of Hospital Medicine, showed that the patients’ length of stay increased by 26%. Applied Radiology talked with lead author Venak P. Gundareddy, MD, to discuss how clinical practices at Johns Hopkins were affected by the findings of the study.
Radiologists, referring physicians, and hospital administrators are grappling with this issue, especially with utilization of healthcare resources and their associated costs. Steady improvements and capabilities of imaging technologies are increasing the odds of seeing and identifying IFs. In this electronic age, access to a patient’s medical records to corroborate the significance of these findings is just a computer click away. The identification of radiologic IFs may increase workloads for clinicians and other healthcare professionals as well as having cost implications if further investigated, and ethical and legal implications of not pursued.
In this context, the authors felt it was important to understand the impact at Johns Hopkins Bayview Medical Center with respect to patients admitted for chest pain. They used data from across-sectional study of all patients hospitalized with chest pain. The main focus of this study was to investigate healthcare disparities and resource utilization in patients with or without a concomitant diagnosis of cocaine use.
Of 1,811 consecutive admissions, 376 patients under age 65 were identified as appropriate for the study. They accounted for 453 admissions during the 24-month study.
Of these, 197 patients — 52% — had 364 new radiologic IFs. Only 27 (7.4%) had findings of major clinical significance. These included three malignant or premalignant lesions. 154 patients (42%) had IFs of moderate clinical significance, predominantly pulmonary nodules. Half of the study cohort had IFs of minor clinical significance, the most common being atelectasis and spinal degenerative changes.
While the median length of stay for patients with and without radiographic IFs was two days, the authors stated that after adjustment for confounders, identification of clinically significant IFs during admission resulted in a longer length of stay and an additional three diagnostic tests on average. The authors determined that these patients had more extra-cardiac workups. Interestingly, patients with COPD or a history of malignancy did not have any more IFs than patients without this history.
Dr. Gundareddy and his colleagues recommend that incidental findings be reviewed in the context of their clinical significance, and further testing should be deferred to the outpatient setting. “The advantages of outpatient testing are several. The patient’s physicians can evaluate these in the context of the patient’s known history and plan a work-up accordingly. This will help ensure that any testing is not being unnecessarily ordered or duplicated. Also, the patient will in all likelihood be discharged earlier than if they had additional inpatient testing. This decreases exposure to hospital-acquired infections,” he said.
Dr. Gundareddy believes radiologists should report all incidental findings and proactively recommend follow-up and surveillance imaging.
“In my personal experience, radiologists do call me when they find an interesting finding to obtain more details about the patient’s history and presentations,” he said. “Here at Johns Hopkins, radiology reports which include a comment of recommended follow-up investigation and a timeline in which to do these tests are treated with the importance they deserve.”
Noting that increasing age is positively associated with incidental findings, Dr. Gundareddy commented that physicians should adhere to the “Choose Wisely” philosophy of being prudent when ordering diagnostic tests . “When we do not ‘choose wisely’ and order radiology imaging when the pre-test probability is low, we do not add any value to patient care but rather increase the burden of health resource utilization,” he said. He cited the example of ordering a CT Chest pulmonary embolism (PE) protocol when the Wells Score probability is low for PE.
“Our study shows that the overwhelming majority of IFs identified from imaging patients with chest pain are of minor to moderate clinical significance. Although some of the incidental findings are major and require immediate clinical action, in the majority of cases, the burden of incidental findings will increase health care resource utilization and provide minimal benefit to the patient,” Dr. Gundareddy concluded.
Incidental findings raise hospital patient length of stay by 26%. Appl Radiol.