Shifting the Mindset to Outcomes over Inputs
The market is full of gadgets and gizmos to help monitor health.
Rarely have such devices in the general population been shown to improve health outcomes. Measuring and gathering data alone does not improve health.
An ad I saw recently for a home cardiac monitor shared a testimonial that, “every morning I check my heart rhythm to see if I’m ok, and it just gives me peace of mind.” My experience indicates that more often, worried but asymptomatic patients are “alerted” by their smartwatch, and then upon further testing show not to have any problem. Other examples abound: home monitoring for SIDS in infants, continuous glucose monitors for people with non-insulin dependent diabetes. In short: people like data. We gather it on the odd chance it will help reveal some problem we didn’t know we had or reassure us that we are healthy. Such tests can be costly, lack clarity on how to interpret an abnormal, or spurn further testing that actually increases the risk of harm. Over the years, I’ve received many requests from patients for a “whole blood panel” of tests; “just check everything.” Patients are accustomed to yearly phlebotomizing that provokes minimal change in their health or behavior. In response, I spend time explaining that much routine testing (outside of evidenced-based screening recommendations) without symptoms, medicines, or chronic conditions to monitor may not help and rather, may harm, by finding abnormalities that send us chasing. I ask, “What are you hoping this test result will change for you?” Similar examples abound, including commercial screening programs which include carotid artery ultrasound screening (which is recommended AGAINST by the US Preventive Services Task Force). Unsuspecting patients are caught paying for inputs rather than outcomes. Review of a website of one such company shows no outcomes data. Could they show, “X number of strokes prevented” or “you will have a X% less chance of stroke because you were tested”? Many of the patients I’ve seen screened already have risk factors for arterial disease. We’ve even discussed their risk factors and are actively treating. The test did not change anything in their health.
Sadly, many of the “quality improvement” projects I’ve embarked upon focused on inputs rather than outputs. For example, I spent considerable time focused on safe opioid prescribing to ensure our provider’s prescribing complied with CDC guidelines. We neglected to ask and measure if our patients were achieving better function and fewer adverse effects from their chronic prescription opioids. Quarterly reports from the state show us how many opioid pills we prescribed compared to our peers, measuring how well we are performing as physicians but neglecting to measure what is most important: how are our patients?
It provokes the question, what is a health outcome? A health outcome is a measurable result, typically categorized as duration of life or quality of life (a functional capability or a reduction in suffering). In contrast, inputs are processes, tests, and treatments that we presume will affect the outcome. Inputs are necessary, but should be dynamic, changing as we measure outcomes. Ineffective inputs are categorized at best, as part of the learning process, and at worst, as waste. Consumers and healthcare providers, systems and payers alike have become oriented primarily towards inputs rather than outcomes, paying for services and processes. Do more (or less) in hopes it changes the result.
How can we reorient care around outcomes?
Ask the patient about what matters to them. It could be a functional capability, a reduction of symptoms, or feeling less overwhelmed with their health. The answer will sometimes surprise you but can be a thoughtful segway to focusing care on outcomes over inputs. When the patient says, “I just want to know X,” I next ask, “what are you most worried about?” This gives the patient opportunity to express important information that helps guide me in education to the patient and order the appropriate diagnostics and treatment. Alternatively, sometimes achieving a health outcome of importance can mean not doing something. For example, it’s acceptable for a patient to decline a recommended cancer screening test when they would not accept the treatment even if it was positive.
After evaluating a patient, focus the conversation on testing and treatment that has evidence to deliver outcomes and then measure to see if it does. Although it is easy to rage against insurance formularies and prior authorizations, the use of evidence-based care guidelines is not all bad. There certainly can be malalignment in either direction: insurance covers an exam that is not indicated (think screening PSA in an 80 year old) or the payer may be putting up barriers to tests and treatments that are indicated. When the outcome isn’t achieved, we should look back at the input and adjust.
Recognize that health outcomes measures are different than process measures. When thinking about “quality metrics,” exchange (but don’t add!) process measures for health outcome measures. Health outcomes need to be measured at the level of the patient, not just the population. Many physicians and systems are overwhelmed with metrics, but they are measuring the wrong things. You can start with home-grown questions, use PROMs (patient reported outcome measures) or incorporate formal data sets from organizations like ICHOM (International Consortium for Health Outcome Measures). Some outcomes are harder to measure than others, particularly those with a longer lead time. However, many process measures become moot when outcomes are measured and positively achieved.
We assume that if the process box is checked, the test is ordered, or the medicine prescribed, that it should produce better outcomes. However, measuring the process (the input) is different than measuring the outcome itself. We must change our mindset to reorient around outcomes to achieve healthcare reform – lower costs and higher quality. We must take the time to answer the question: Does what we are doing make a difference in the individual patient’s health and how do we know?
