To Improve Patient Outcomes, We Must Measure Them

“An eighty-two-year-old man developed fainting episodes. His tests showed severe blockages in three coronary arteries and both carotid arteries. The doctors recommended doing a cardiac-bypass operation as soon as possible, followed by surgery to open up one of his carotid arteries.”

A detailed account of his story is written in The New Yorker article titled Overkill by Dr. Atul Gawande.  The man is the father of Dr. Gawande’s friend.  He describes how the blockages weren’t causing the man’s fainting episodes or any other impairments to his life. The operation would not make him feel better. Instead, “success” to the doctors meant reducing the future risk of a stroke.

The man had a stroke during the cardiac surgery. A week later, he recovered his ability to talk, although much of what he said didn’t make sense. But he had at least survived. “We’re going to put this one in the win column,” Dr. Gawande’s friend recalls the surgeon saying.

We rarely measure patient outcomes

Whether patient outcomes are defined by the surgeon, “reducing stroke risk”, or by the patient, “improving quality of life”, we must measure patient outcomes in order to improve them.  We collect a myriad of reimbursement related quality measures, yet measuring overall patient outcomes is rare. In fact, for most expensive treatments or surgeries, it is less than rare because it is none.

A study by Dr. Martin Makary and his colleagues published in the Journal for Health Care Quality describes this major shortcomings in the way patients are tracked after treatments. In an interview with National Public Radio, Dr. Makary stated, “99 percent of people in America, when they go in to have surgery, the outcome is not measured. Nobody’s keeping track. So I’m amazed at how one-fifth of the economy, functions with so little measure of its performance.”

Fewer than 10% of people living in the United States ever participate in a clinical trial.  As much as 80% of patients are excluded prior to clinical trials because they may have an additional complicating condition. Yet, 31.5% of people have multiple chronic conditions and only 20.2% have a single chronic condition. We test for negative medication to medication interactions. Yet we don’t test combining more than 2 medications when 20% of the people in the United States are taking 5 or more medications. It is no wonder 5% of the people in the United States are spending 50% of the healthcare dollars, they are likely to have multiple chronic conditions and take 5 or more medications. The complex patients that need patient outcomes measures the most are often excluded from clinical research.

A recent study showed that one out of every 12 visitors to the Emergency Room return within 3 days. Another study showed Medicare spent $8.5 billion on low-value care such as putting coronary-artery stents in patients with stable cardiac disease. While ER visits and the low-value care may achieve their quality measures, we are unable to tell when they are effective without measuring patient outcomes.

Where we could start measuring Patient Outcomes

For a small cost, we can measure patient outcomes. We could simply add 5 questions to the intake process for patients in Medicare’s new alternative payment models (Medicare Shared Savings Program (MSSP) and Bundled Payments for Care Improvement (BPCI)). For the first time, we could get patient outcomes insight for surgeries such as the joint replacements which Medicare spent $6.6B on for 450,000 patients in 2013.

BPCI – Patient Outcomes

  • Elective Surgeries – survey before surgery
  • Ambulatory Care Sensitive Admissions – survey during patient stay

MSSP – Patient Outcomes

  • Complex Patients (multiple chronic conditions) – survey during Annual Wellness Visit
  • Elective Surgeries – survey when surgery is recommended. This will help compare patients that have surgery vs. those that do not.

What we could ask

We could ask Patient Well-Being questions at the time described above using a scale of 1-5.

  1. Rate how you are feeling physically overall. Pain, nausea, health concerns
  2. Rate how you are feeling mentally overall. Mood, sentiment, anxiety, depression, stress
  3. Rate how capable you are with your functional abilities. Walking, talking, working, activities of daily living
  4. Rate the level of impact the physical, mental or functional abilities are having on your quality of life. Activities, economic, purpose, competence, social connections
  5. Rate your overall health.

The patient could be sent a voluntary survey follow-up via mail or email at 3, 6, 9, and 12 months after the initial survey. If the initial survey was related to a surgery or a hospital admission, the following additional questions could be included:

  1. Rate the overall effectiveness of the surgery (or hospital admissions).
  2. Rate the overall performance of your surgery (or hospital admission) providers?
  3. Would you recommend the surgery (or hospital treatment) you received based on the conditions you had prior to the surgery (or hospital treatment)?

What could we learn?

We would certainly learn something we don’t already know. It would enable physicians and patients to access a treasure trove of patient outcomes data to help them decide the right treatments and strategies.  It would help physicians and scientific researchers begin the rapid iteration cycle of improving healthcare treatments by understanding patient outcomes.

For the father of Dr. Gawande’s friend, the patient outcomes may have been mixed. The stroke risk was reduced, yet so was his quality of life. It is insight that both physicians and patients would love to know.

Related Posts