An otherwise healthy 38 year old patient was brought in by her family with vomiting and mental status changes from her pain medications.
She had repair of a tibial plateau fracture performed four days earlier and was having a lot of pain. She didn’t like taking the Percocets that she was prescribed because they made her nauseous. She took one of them the day after her surgery and she was nauseous the rest of the day, so she vowed not to take any additional Percocets. However, her knee pain was worse that morning to the point that she couldn’t stand it any longer, so she took two Percocets … on an empty stomach, no less. A couple of hours later, she was acting strange and had vomited several times.
When she arrived, she was lethargic and retching. She was afebrile, but her blood pressure was 87/50, her pulse was 120, and her respirations were 28. With a fluid bolus and some Zofran, her vital signs improved and she felt better. The option to give her Narcan was discussed with the patient and family but with the improvement in her symptoms, they didn’t want to reverse the pain relief that the medications had given her.
The remainder of her exam went along with her history. She had been sleeping a lot after her surgery and hadn’t eaten much, so her mouth was a little dry. She was awake and drinking fluids in the ED. She was tachycardic, but her tachycardia was improving with fluids. Her abdominal exam was fairly normal – perhaps a little epigastric pain with palpation, but nothing concerning. Her knee was tender and a little swollen. The orthopedist came to the ED and evaluated her and stated that her knee looked good for her post-surgical status.
The patient was observed for an hour or so and felt better. She wanted to go home and sleep. Family agreed. Nurse manager was pushing to have the patient discharged so we could move more patients into the room from the waiting room. “Discharge?” was written on the tracking board under the patient’s name.
Anchors aweigh. Time for this ship to sail back to port.
A final re-exam of the patient showed a couple of abnormalities, though. Her pulse was still in the 100-110 range. Her blood pressure was now 108/50. But her respiratory rate was 28-32. Lungs clear on exam. Pulse ox 99%.
“Let’s just do a few labs and a chest x-ray. Check a d-dimer and do a blood gas. Her respiratory rate just doesn’t make sense.”
“After two hours NOW we’re deciding to do labs? She’s still in pain – that’s why she’s breathing fast.”
The nursing director rolled her eyes.
“Sorry, everyone. We’ll try to get her out of here as soon as we can.”
An hour later, the patient was being admitted to the ICU.
Anchoring occurs when we focus on an explanation for a patient’s symptoms too early in a workup. Even though the symptoms seem plausible, by “anchoring” on one explanation, we tend to discount other symptoms or findings that don’t fit in with the diagnosis we’ve made. By focusing in on one diagnosis, we can miss other diagnoses. Anchoring is sometimes difficult to overcome.
Some of the best ways to avoid making anchoring errors include:
- Re-evaluating the diagnosis in light of all the information available
- Considering whether treatments that would normally improve the diagnosed condition have actually improved the patient’s condition.
- Considering alternate explanations for data that conflicts with the presumptive diagnosis
- When in doubt, consider additional testing
- If still in doubt, consult a colleague. Four eyes are better than two.
So why was this patient admitted to the ICU?
Pulmonary embolism? Nope. A pulmonary embolism may have explained her fast respirations and tachycardia. She was at an increased risk of pulmonary embolism after recent surgery. But her d-dimer was normal. And why would she be vomiting and dehydrated from a PE?
Adverse effect to the Percocet? Would have explained the vomiting and some of the vital sign abnormalities. But why the rapid respiratory rate and the dry mouth?
The patient was admitted to the ICU because she needed to be on an insulin drip. Her glucose was 1100. Her pH was 6.8. She was in new-onset diabetic ketoacidosis.
When the labs started coming back, the doctor almost needed to be resuscitated. He was a few mouse clicks away from discharging a patient who probably would have died if she had went home without treatment for her underlying problem.
There but for the grace of God …
This and all posts about patients may be fictional, may be my experiences, may be submitted by readers for publication here, or may be any combination of the above. Factual statements may or may not be accurate. This was a reader-submitted story. If you would like to have a patient story published on WhiteCoat’s Call Room, please e-mail me.