A 60 year-old woman with chronic mesenteric ischemia was admitted for symptomatic inferior mesenteric artery (IMA) stenosis. She underwent IMA stent placement with a postoperative course notable for abdominal pain and nausea. Two weeks later, she was transferred to a medicine service for uncontrolled hypertension. Overnight, the patient had sudden-onset nausea and vomiting with a distended abdomen and tenderness to palpation but no peritoneal signs. Abdominal x-ray (KUB) was interpreted by the overnight internal medicine physician as dilated loops of small bowel consistent with ileus. The provider was called to a Code Blue so did not call radiology to confirm, which due to limited staffing does not read x-rays overnight without verbal request.
The next morning, the patient developed hematemesis and hypotension and was transferred to the ICU. Radiology read of the overnight KUB noted portal venous gas concerning for bowel ischemia. CTA revealed a >20mm IMA dissection distal to the stent with infarction of the stomach, small bowel, cecum, liver, and spleen. Given her extensive disease, the patient was not a surgical candidate. She died with family present.
Late presentation of IMA dissection is a rare complication of stent placement that may be worsened by severe hypertension. In this case, its delayed diagnosis was a medical error comprising both active and latent errors. Active errors included the inaccurate overnight abdominal x-ray interpretation, lack of familiarity with post-stent complications, and “slip” of not calling for a radiology read due to competing responsibilities. Latent errors included limited overnight radiology staffing and no automatic overnight x-ray read.
This case highlights inherent challenges when hospitalists increasingly co-manage surgical patients and overnight providers of all departments cover a large number of patients. In this clinical environment, it is essential to put systems in place that prevent cognitive “slips” from contributing to adverse patient outcomes. The current overnight system for radiology x-ray review requires a clinician to call radiology for a “wet-read”. This additional step allowed for a “slip” by the ordering provider who failed to call for this interpretation when distracted by other clinical duties.
When caring for complex patients overnight, it is vital to have standardized processes for coordinating care communication between radiology and front-line clinicians. In this case, a protocol for automatic overnight x-ray read by radiology may have prevented the delayed diagnosis of a late fatal dissection. Furthermore, a uniform expectation of communicating potential postoperative sequelae among co-managing services may help improve recognition of atypical complications unfamiliar to most hospitalists.