160 Inappropriate Follow-Up of Abdominopelvic Imaging Findings Indeterminate or Suspicious for Malignancy in Emergency Department and Admitted Patients

Abstract

To evaluate the effects of social determinants of health and health system factors as well as communication on the follow-up of abdominopelvic imaging findings indeterminate or suspicious for malignancy in ED and admitted patients. Our radiology department uses standardized codes to categorize abdominopelvic imaging findings as benign, indeterminate, or suspicious for malignancy on all ultrasound, CT and MRI reports. We assessed ED and inpatients with imaging findings indeterminate or suspicious for malignancy over 7 randomly selected weeks between October 2015 and January 2017. Social determinants of health (age, sex, race, primary language, median household income by zip code, insurance status at the time of index imaging, and physical and psycho-cognitive comorbidities), health system factors (primary care provider [PCP] location, type of follow-up visits within our system, and prior as well as subsequent encounters within our health system), and communication of the incidental finding to the provider in the radiology report impression or to the patient in the discharge instructions were gathered from the electronic health record (EHR). Potentially inappropriate follow-up was defined as no completed relevant pathology, imaging, or clinical follow-up within our health system or other hospitals in our health information exchange (HIE), and no documented rationale for lack of follow-up. Univariate analysis evaluated the association between inappropriate follow-up and the included covariates. Among 210 patients with 235 abdominal imaging findings indeterminate or suspicious for malignancy, 37% of patients (78/210) and 35% of imaging findings (83/235) received potentially inappropriate follow-up; 31% (24/78) of patients without appropriate follow-up had no subsequent encounters within our health system. Potentially inappropriate follow-up was higher among patients with no prior health-system visits (p=.000), no known PCP (p=.000), PCP outside our health system (p=.001), commercial insurance (p = .001), no psycho-cognitive comorbidities (p = .009), and radiologist categorization of a finding as indeterminate for malignancy as opposed to suspicious (p=.000). Communication of the incidental finding to the provider in the radiology report impression or to the patient in the discharge instructions was not correlated with appropriate follow-up (p = .658 and p=.255). Nearly 4 out of 10 patients with imaging findings indeterminate or suspicious for malignancy detected during an ED visit or inpatient admission are at risk for inappropriate follow-up. Lack of documented PCP, PCP outside of health system, commercial insurance, and no documented psycho-cognitive comorbidities were correlated with inappropriate lack of follow-up within our system or other hospitals in our HIE. These findings suggest that better protocols are necessary to ensure adequate follow-up for incidental findings in patients with fragmented access to health care. Communication of imaging findings indeterminate or suspicious for malignancy to providers through the radiology report impression or directly to patients is insufficient to ensure appropriate follow-up.