A 52 year-old woman lies in extremis in ICU following a gastric perforation discovered after contrast is put through an NG tube. So what? These things happen. Yep, they sure do. Here’s how this particular one happened.
The woman had an Upper GI four months earlier showing a paraesophageal hernia, where the esophagus went through. The report mentioned the hernia, it didn’t mention its prediposition to gastric volvulus. While one can’t mention every possibility in a report, it might have been useful to mention this particular fact.
Four months later, the woman presented with abdominal pain. Her initial CT of the abdomen and pelvis mentioned the hernia and an ovarian cyst, but no comment was made about gastric distention or the type of hiatal hernia. Unfortunately, the prior study wasn’t re-evaluated during the reading. That was unfortunate, but many studies today contain a thousand images; indeed, a radiologist may encounter 100,000 images a day. An NG tube was passed, and a second scan, with contrast, showed the perforation – really well. Fatigue, the volume of images, hospital and referring physician demand for quick reads, compensation for number of studies (not images) viewed, make errors more possible. Reviewing past studies is not compensated, so there is less of a tendency to do so. What do clinicians do if they receive a huge chart when a new patient arrives, inconveniently booked into a follow up slot? Compensation is based on a numbers game; what game is played dictates what gets done well, what gets done, and what doesn’t. Having been on both sides of the medical fence, I can easily spot a distracted, harried and hurried physician. All three of these are a setup for cognitive errors, the single biggest type of mistake a physician can make.
The patient developed peritonitis. Perhaps if fewer CT scans were ordered, it would be easier to routinely evaluate prior studies as part of the reading process. Once having practiced neurology, I believe, and the literature supports, a person with intact cortical function, no neck pain, no tenderness to palpation and no neurological deficit doesn’t need a cervical spine CT after an injury. Whole body scans are often done when clinical judgment would suffice. Besides being a radiation issue, it is a time issue affecting emergency department throughput (we patients call it waiting), a money issue, because these studies are expensive, and a quality of care issue. After my bicycle accident, I had several studies, but nobody took off my shirt to look at the road rash on my back. Nobody palpated my entire body, since severe pain in one place may mask a significant injury elsewhere. Those additions take perhaps thirty seconds.
Worse, if little clinical history is provided, it affects the type of study and the radiologist’s approach. Differentiating PE from dissection makes a big difference in timing of the scan after contrast. “Chest Pain” is not helpful to the radiologist. Yes, you are in a hurry. I was too. But I always put clinical information on my neuroimaging requests. The radiologists appreciated it. I got better reports. It helped my patients.
This woman survived. The medical community ought to learn from this, rather than copping out and saying “these things happen,” “nobody’s perfect,” or “who made you the quality expert?” “Nobody’s perfect” doesn’t cut it if a person dies from something preventable, either in or out of medicine. You don’t hear the civilian or military aviation community say that. They learn from the mistakes, and they publicize them. Read a few sometime, and you would be surprised how much we could take away from their field, rather than the mantra, “We’re doctors. We’re different.”
Rheumatic fever, polio and gas gangrene used to happen; 30 years ago auto accidents killed twice as many people per capita, anesthesia deaths were once far more common. Sean Elliott and Alonzo Mourning would have died from uremia the way actress Jean Harlow did.
Excessive workload increases the likelihood of a radiologist’s not reviewing past studies and not dictating, “Paraesophageal hernias can lead to gastric volvulus.” What does workload do to you in your field? There ought to be a way that physicians can do a decent job, make decent money, learn from their mistakes and those of others, have a life and not fear lawsuits. Here are a few thoughts:
It’s time we had community standards for common, high risk procedures that lend themselves to standardization: hyperalimentation, ventilator management, pre-op antibiotic delivery and central lines. We can standardize and still respect individual differences. It’s time disciplines who function together, like emergency medicine and radiology, work together. It’s time to have legislation mandating a free from discovery error reporting system which I proposed in 2001 and which failed the legislature in 2004 and 2005; the hospital association worked to kill it. It’s time to have liability reform so physicians aren’t treated like criminals when they err. Maybe if we did the first two and supported the third, we could get the fourth. Along the way, we might additionally work towards complete medical coverage for childhood up to at least age 18 – with real-time data on outcomes and costs. We might start having a better system. Maybe we could expand age 18 to all.
And perhaps see fewer cases of preventable peritonitis.
Boreal B[l]og 
Leave a Reply