HIV and ID Observations

Sometimes, you can predict a bad outcome. Examples:

• Proposing marriage after an awkward first date — and doing so over gas station nachos.
• Moving to a Cambridge apartment with no off-street parking, then buying a Tesla Cybertruck.
• Trying to recruit for ID fellowships from a group of cosmetic dermatologists.

But predicting what happens in clinical medicine? Not so easy. Which is why the clairvoyance expected by certain health insurance companies baffles the mind — they seem to believe we can diagnose, prognosticate, and determine outcomes with the omniscience of the Oracle of Delphi.

Take this recent gem. I’m sharing it here not because it’s unusual, but because the absurdity deserves a moment in the spotlight.

(Part of a series.)

Here’s the scenario (some details changed to protect privacy):

The patient, a 64-year-old man, went to our emergency room with fatigue, acute kidney injury, and a hemoglobin drop. He’d recently undergone gastric sleeve surgery, making the clinical status more uncertain than usual — plus a background of diabetes and high blood pressure as medical comorbidities. Given the symptoms and risk factors, he was admitted to medicine for hydration, monitoring, and endoscopy. He got better. We all celebrated. Cue the credits.

But then… the sequel. (Spoiler alert: It’s a horror film.)

A few days later, I received an email saying that the insurance company had denied inpatient level of care — in plain English, they didn’t want to pay. Would I have time to do a “peer-to-peer” discussion to try and reverse the decision?

They might as well have asked me to call an airline to rebook a cancelled flight during a massive Nor’easter, that’s how much I was looking forward to this task. But given how justified the admission was, and my trying to be a team player to defend good clinical practice in the face of our Private Insurance Overlords, I set up some time to talk with my “peer.”

I use quotation marks because while I’m sure she was, technically, a healthcare professional, her role in this drama felt more like prosecutor than peer.

She had some of the hospital data. Not all. Enough to cherry-pick to support their refusal to pay, but not enough to understand the full context of the case since, of course, she had never seen, spoken with, or evaluated the patient.

She asked me a series of questions, some of which were about information she already possessed, as if hoping I’d contradict myself like a suspect in a police procedural. (“So you’re saying the patient had a drop in his hemoglobin during the hospitalization? Interesting, doctor… very interesting. I see here it remained 7.5–8.3 during his stay. Do you consider that a drop?”)

I explained, again, the patient’s presentation. The drop in hemoglobin from his baseline of 10.5. The post-bariatric surgery. The concerning acute kidney injury in someone with diabetes, hypertension, and obesity. You know — the reasons why he was admitted. 

But then came the decision, delivered with the cool finality of a game show host eliminating a contestant. Because the patient had no hemodynamic instability during his stay, and no ongoing bleeding, the hospitalization was deemed… unnecessary.

Denied.

“I cannot overturn the decision,” she said, as if quoting some higher order of evidence from randomized clinical trials rather than a faceless algorithmic edict she no doubt had up on her screen as she was talking with me.

I took a deep breath.

Then I asked her to imagine herself as the patient — sitting in the ER, post-recent surgery, with those symptoms and those lab results. Or better yet, as the clinician doing the initial assessment, deciding whether to admit or to send him home.

Would she have discharged this man? Would her judgment have changed if she weren’t now on the payroll of Giant Healthcare Insurance Company? Had she, like so many burned-out clinicians, left clinical medicine because of pointless, time-wasting demands like this conversation — only to end up perpetuating the same dysfunction from the other side?

No answer. Silence on the other end. Then, she repeated,  “Thank you, Dr. Sax for your perspective. I cannot overturn the decision.”

Because of course, the outcome — the good outcome — was only apparent after the fact. One reason to admit people is when we don’t know if they’ll do well.

So yes, the patient got better. No, he was not critically ill. But that’s not evidence the admission was unnecessary; that’s evidence the admission went about as well as could be hoped. Isn’t that what we all want?

Unfortunately, our healthcare system now seems to reward retrospective omniscience more than clinical judgment. “If only you had known he wouldn’t bleed again!” they say. Right. And if only I had known to buy Nvidia stock when it first went public in 1999.

I’ll stop now — time to call my airline because my flight has been canceled due to an unexpected mid-summer blizzard. Should be more fun than this call.