Not long ago, one of my elderly patients (who gave me permission to tell this story) began requiring blood transfusions for a condition known as ischemic colitis. Usually, it occurs because there are blockages present in the arteries that supply blood to the colon and, starved of blood, the inner lining of the colon becomes inflamed. But her arteries, it turns out, are normal. So we term her ischemic colitis microscopic because we consider the blockages to be in the tiny arterioles. Unfortunately, in her, it’s entirely unclear why these tiny arterioles should be blocked at all. But at whatever level the blockage of blood flow occurs, the predominant symptom is the same: continuous bleeding from the colon.

If she were younger and healthier, we could surgically remove her colon (though, ironically, if she were younger and healthier, she wouldn’t likely have ischemic colitis in the first place). Unfortunately, given her age and frailty, neither her family nor I believe she would survive such an operation. So, periodically, when she becomes severely anemic and feels poorly we’ve been transfusing her.

It turns out to be extremely difficult to arrange this, however. Not many centers are equipped to transfuse patients as outpatients, and those that are have long waiting lists. So I’ve been sending her to a local ER. While this is arguably not the best use of medical resources, given that the only other option is to admit her to the hospital, it’s the strategy that makes the most sense. And it’s worked.

Until last time, that is. As is my habit, I called the ER before I sent her to alert the attending physician that she was coming. I explained why I was sending her and asked him if he could transfuse four units rather than the usual two. My rationale was simple: I wanted to prolong the interval between transfusions.

His response, while to some degree understandable, wasn’t what I’d hoped for: “We’ll evaluate her when she arrives.” He then hung up rather brusquely.

I communicated my misgivings about my interaction with the ER attending to the family, telling them that if there was any trouble they should call me and I’d speak with the ER attending again. I didn’t hear back from them, so I called the next day. They had indeed transfused her, my patient’s daughter told me, though she didn’t know how many units.

But they’d also told her they wanted to do a colonoscopy and run a number of other tests. When my patient’s daughter asked them why they wanted to do these tests when they’d already been done, the results had already been discussed and a strategy had already been agreed upon with me, their answer was that they thought something might have changed.

I’d already spent many hours in discussion with the patient and her family regarding how aggressive we should be with her medically. Given that she not only has microscopic ischemic colitis but also dementia, that we can’t cure either condition and that both conditions have a dismal prognosis, the patient, her family and I have all agreed that our aim isn’t to prolong her life but rather to keep her as functional and comfortable as possible.

For this reason, our posture has been to strictly limit the number of interventions we allow her to be exposed to. The reason is simple: everything we do in medicine carries risk, and it’s hard to justify taking any risk in a patient who isn’t likely to gain significant benefit from almost any intervention we could think to offer, as is the case with my patient. Unfortunately, thinking through the nuances of risk and benefit carefully in a detailed manner takes time and isn’t something many physicians do, especially when they don’t know the patient well.

This explains, in my view, what happened to my patient. A gastroenterologist quickly decided she needed another colonoscopy without thinking through what the findings were likely to be, what interventions we could offer her, and what risks those interventions posed. He didn’t call me, the patient’s primary care physician, to discuss it either. If he had, I’d have told him what I told the family: “Do nothing to her besides give her blood.”

Unfortunately, this scenario is all too common in medicine today. Care is frequently fragmented because doctors don’t communicate well with patients and their families or with other doctors. As a result, healthcare resource over-utilization continues to flourish. A colonoscopy would have added nothing to our understanding of my patient’s condition, wouldn’t likely have shown anything new that we could or would treat, would have suggested no new options for the condition we already knew she had and would have put her at unnecessary risk for a complication.

The lesson to me as her primary care physician was clear: communicate, communicate, communicate. I must communicate with everyone involved in the care of my patients — not just with patients themselves and their families but also with every healthcare professional they encounter. Outcomes often hinge on the tiniest of details, and as a primary care physician, it’s my responsibility to know those details well enough to protect my patient from harm at the hands of the healthcare system itself. Not every patient has a daughter as savvy and as dedicated as my patient does, one who could (and did) throw herself in the path of a number of doctors who, though well intentioned, were sloppy in their thinking and their communication. What patients need besides a family member to act as a strong patient advocate is a good primary care physician: one with enough time to shepherd patients through the health care system with enough attention to ensure they receive care that hits the sweet spot — not too much, not too little. For as certain as too little care — that is, too little access to care — is harmful, in many circumstances, so is too much.

Alex Lickerman is an internal medicine physician who blogs at Happiness in this World.  He is the author of The Undefeated Mind: On the Science of Constructing an Indestructible Self.

Image credit: Shutterstock.com