These days, improving the quality of care isn’t enough – it’s also necessary to demonstrate the value of the improvement and that it was worth the investment needed to bring about the change.
Let’s be clear here: when I refer to “value” in healthcare I’m not thinking of the expression “value = quality / cost.” While that definition can be helpful when trying to compare value (achieving the same level of quality at a lower cost equates to better value, as does improving quality while holding costs constant) I’m referring to a more general definition, one that acknowledges that value is subjective and can vary over time and by perspective. Reducing avoidable readmissions has value for both payors (reduced reimbursement costs) and patients (improved quality of life), but the type and magnitude of that value is different for each party. Additionally, one can imagine instances when the link between quality and value has very little to do with cost: very poor quality has little value no matter how cheap it is, while high quality care can be very valuable even if it is extremely expensive.
So, in this context of trying to measure the value of quality improvement, where do we often run into trouble? There are several common obstacles one may encounter.
Obstacle #1: We’re not certain about the validity or reliability of our quality measure.
The ability to demonstrate changes in quality are much easier if what we choose to measure is a valid reflection of the underlying quality we’re targeting and can reliably identify and detect when that underlying quality is different or has changed. (See my article on validity and reliability here for more detail.) If our validity and/or reliability is suspect, then it’s possible that even if we improve the underlying quality we may not see significant improvement in the quality measure (have any of you ever felt like you made significant improvements to hospital quality only to see that 30-day readmissions didn’t budge?). In this case, when we struggle to demonstrate that quality has improved, we will also struggle to show that significant value was attained, since value is dependent on quality.
Obstacle #2: The benefits we see are not measurable. . .
If you have a valid and reliable quality measure, you should be able to detect and demonstrate improvements in quality. However, that doesn’t guarantee that you will be able to measure or quantify all of the benefits associated with that improvement. For example, reducing falls in a long-term care facility may produce a variety of benefits, including those related to the facility’s “brand” (how comfortable family members feel about having loved ones there, how likely they may be to recommend it to someone else, etc.). While improving their brand is certainly valuable to the facility…it may not be immediately clear how to measure it. Even if it is possible to measure it, doing so may be out of the scope of the project or simply not feasible given available resources. If you are not able to measure it (for whatever reason), you certainly cannot assign value to it!
Obstacle #3: . . .or are not monetizable. . .
Some benefits may be easily and directly translated into dollar amounts that can be quantified. Reducing readmissions, for example, can equate to to specific cost-avoidance amount for a payer due to fewer hospitalizations that require reimbursement. In contrast, the improvement in quality of life patients experience by not being in the hospital may be difficult to equate to specific, quantifiable dollar amounts. If the benefits included in your analysis are all of the latter type, you may get frustrated with the lack of specificity you are able to produce when estimating the true, financial benefits of the improvement in quality.
Obstacle #4: . . .or are not directly attributable to the quality improvement activities.
Even when quality measures are valid and reliable so that we are confident that improvements in measure performance signals true improvements in quality, we still may have trouble attributing certain benefits directly to the quality improvement activities. Reducing physician burnout can improve physician satisfaction, productivity, and even provider-patient communication. If physicians are happier with their work and in their position, they may also be less likely to leave, which could reduce turnover – a measurable and often monetizable benefit. However, depending on the situation, it may be difficult to ascertain what portion of any observed drop in turnover rate is directly attributable to the activities that were intended to reduce burnout. If you don’t feel comfortable assigning 100% of the observed change to the intervention, you are then left with determining what portion can be attributed to it…maybe without much guidance or data to help you decide.
How to avoid these obstacles
The key to avoiding these obstacles is to think about how you will measure value at the same time you are deciding how you will measure quality. Too often, thoughts regarding how to assess value are left until after the quality evaluation plan is completed, or worse yet, after the intervention is completed. At that point, it’s often the case that those wanting to estimate the associated value are left with a handful of measures, none of which are measurable, monetizable, and attributable to the intervention.