When Nothing is Something

Dr. Alexander is a valued colleague of mine at The University of Chicago. He is a co-editor of the 5^th edition of our textbook, Symptom to Diagnosis: An Evidence-Based Guide, due out early next year. When he came into my office last week and told me about this case, announcing that “sometimes, the best thing you can do for a patient is nothing,” I immediately asked him to write it up for Sensible Medicine. I am so glad that he did.

Adam Cifu

In the outpatient general medicine clinic, I live in a world of uncertainty. If I have a clinical question for which a diagnostic test may help me care for a patient, I am only sometimes able to get a result the same day. An X-ray or a simple blood test? A same-day result is possible. CT scan or MRI? A week, if my patient and I are lucky.

This is not the case when I treat patients in the hospital, where patient rooms are a hot commodity and decreasing length of stay seems to be the most important aspect of a hospitalization. In this environment, test results come back not in days or weeks, but hours. My colleagues, who, unlike me, spend most of their time in the inpatient setting, become accustomed to this fast-paced environment and become unaccustomed to stewing in the uncertainty of a clinical scenario. But what happens when that test comes back and uncertainty remains? Now, both external pressure (from the hospital) and internal pressure (from a learned intolerance of uncertainty) lead the best clinicians to implement care plans that can hopefully ‘fix’ the problem quickly.

These are the clinical scenarios in which I, primarily an outpatient primary care doctor, can provide the most value as a supervising clinician on the inpatient general medicine service.

We were caring for Ms. D, a vibrant 94-year-old woman. Before being hospitalized, Ms. D was living independently, maintaining a dynamic social life. Five days prior to admission, she passed out at home and awoke on the floor in considerable pain. She was unable to call for help and spent another two days on the floor before paramedics came and took her to the emergency room.

When Ms. D arrived at the hospital, she was confused and dehydrated, having been unable to get food or water. She was admitted to the medical intensive care unit because her sodium level was very low -- hypovolemic hyponatremia. Hyponatremia requires careful management as overly rapid correction can cause serious neurologic complications. For three days in the intensive care unit, clinicians monitored her sodium level, made small adjustments to fluid rates, and restricted her water intake. Despite their best efforts and making the most accurate calculations possible, her sodium level never responded as anticipated. Ultimately, when her sodium level stabilized and her mental status was back to her robust baseline, she was transferred to the general medicine floor. My team was tasked with ‘fixing’ her sodium.

After hearing this presentation from my intern, my team proposed a plan similar to what the intensive care unit clinicians had been doing: making various calculations, continuing to restrict water intake for the patient, and consulting with various specialists. Nothing was necessarily wrong with this plan. We teach medical students and residents to make these calculations in these scenarios. Yet, over the last ten years in the clinic, I am constantly reminded of the amazing resilience of the human body. For 94 years, up until 5 days ago, despite not having her sodium checked every four hours, Ms. D lived a pretty remarkable life. Therefore, I proposed the seemingly most preposterous plan imaginable.

I proposed that we do nothing.

Now, in fairness to my intensive care colleagues, I assumed Ms. D’s management after she had regained her mentation. I also knew that her sodium had been stable (though still only 120 mEq/L) for the preceding 36 hours. I also knew that her sodium was normal (140 mEq/L) a few months earlier and that she had not had any recent changes to her health or medications.

In this context, doing nothing seemed appropriate.

“But should we continue to limit her water intake?” Nope. Let her drink whatever she wants.

“But how much should we tell her to drink?” Give her a big cup of water – tell her to drink it when she is thirsty.

“But what should we do with the IV fluids?” Just stop them. She can eat and drink just fine, let’s have her do that.

Silence from the team. Rapture from the patient.

The next morning, Ms. D’s sodium had increased 3 points to 123 mEq/L; the following day, it was 126 mEq/L. The team was astounded. She and I had a lovely chat about her late husband and about how she was happy to see her two daughters, who had flown in when they learned their mother was hospitalized. We talked with Ms. D about the importance of physical therapy and how some rehabilitation before going home would probably be the safest option for her. She agreed.

In an environment where quick decision-making is seen as a gift and prompt treatment is considered paramount, doing nothing becomes the hardest thing for clinicians to do.

That is all the more reason to relish being right when doing nothing turns out to be the best thing for the patient.

Jason Alexander is a general internist and associate professor of medicine at the University of Chicago.

Photo Credit: Timo Volz