SAN DIEGO – Head CTs for patients with in-hospital delirium. Ammonia tests to check for hepatic encephalopathy in chronic liver disease. Renal ultrasounds for acute kidney injury.
Those are three low value tests highlighted in hospitalist Dr. Leonard Feldman’s latest iteration of his lecture series “Things We Do for No Reason.”
Dr. Leonard Feldman
Dr. Feldman, associate professor of internal medicine and pediatrics at Johns Hopkins University, Baltimore, has presented his list of usually unnecessary hospitalist practices for five years at the Society of Hospital Medicine’s annual meetings. With three new ones explained during the 2016 meeting, there are now 19 on the list and more to come, he said.
“So far, I’ve picked things that are relatively low-hanging fruit, things for which there’s good evidence we shouldn’t be doing and if you saw the evidence, you’d say ‘that’s right, we shouldn’t,’” he said.
Dr. Feldman’s intent is to help clinicians stop certain “learned behaviors,” tests and procedures which research and experience now show “are not helping people, sometimes harm people, and often result in a cascade” of further unnecessary tests and care.
The conference presentations have been so popular, the Journal of Hospital Medicine in October 2015 started a “Things We Do for No Reason” series.
Here are the three most recent tests hospitalists should avoid:
Ammonia levels for chronic liver disease
Dr. Feldman said doctors were taught in medical school that ammonia levels rise in patients with cirrhosis and when they rise too high, the patient may develop hepatic encephalopathy. They also learned that if levels are normal, the patient should not have hepatic encephalopathy.
But a number of studies have found “neither of those is true,” he said. What’s possibly worse is that “you close your mind to other possible diagnoses way too early.” Nevertheless, the practice at many hospitals is to perform multiple tests to trend those levels.”
“I had a patient who had an ammonia test sent the other day while in the emergency room, and it was elevated,” Dr. Feldman recalled in a recent phone interview. “The patient got admitted, but when we re-tested, it wasn’t.”
Part of the problem is that blood samples are often incorrectly processed. “When you draw the blood, you have to put it on ice and it needs to get to the lab very quickly. And I think we do neither of those things on a regular basis,” he said. Also, if the patient has a tourniquet or is clenching a fist, use of muscle creates ammonia.
Dr. Feldman said that at a hospital like Johns Hopkins in Baltimore, where there are high rates of hepatitis C, there might be 50 patients with chronic liver disease, or 20% of patients on medicine service. It’s not the cost of the blood test that he’s worried about because that’s probably minimal. Rather, it’s the test’s downstream provocation of more unnecessary care “and missed opportunities to intervene with a treatable diagnosis.”
In general, he said, “for patients with chronic liver disease, we shouldn’t be checking ammonia.”
Head CTs for inpatients with new onset delirium
Performing a costly head CT scan on a patient who presents in the emergency department with delirium is appropriate. But for low-risk patients who develop delirium inside the hospital without a clear reason, such as a fall or focal neurologic symptoms suggesting a stroke, a head CT is probably not necessary, Dr. Feldman said.
“But we have this knee-jerk reaction, this reflex, that when a patient becomes delirious, we probably should run a head CT on them,” he added.
Dr. Feldman acknowledged that the frequency of head CTs on inpatients with delirium has been hard to tease out.
“But all the studies indicate that patients who develop delirium while in the hospital, without any sort of risk factor, are very unlikely to have pathology found on a head CT,” he said, noting that the cause of their delirium is likely something else, like dehydration, an infection, disruption of sleep, urinary retention, or medication effect.
Of course, if patients aren’t getting better without the CT, order the CT, he said. “Even if the patient has no risk factor, there’s still a 3% chance of having an abnormality like a tumor or stroke.”
Renal ultrasound for patients with new acute kidney injury
To determine if an acute kidney injury is caused by a treatable obstruction, such as a large prostate causing urinary retention, doctors often first order a renal ultrasound, a test that can cost $300, and must be read by a radiologist.