When reading about ROI or how to perform ROI for healthcare quality improvement, you will read about certain benefits as being “direct” vs “indirect” or “tangible” vs “intangible.” Conceptually, it is not difficult to understand what those might be: reductions in utilization or adverse events may be direct and tangible, while improvements in staff job satisfaction or patients’ peace of mind may be indirect and intangible. However, there are often times when the lines are blurred at it is necessary to be a little more specific about whether we can reasonably include a benefit in the analysis. For example, we can try to measure satisfaction or peace of mind…so does that mean we can use it? Maybe not, especially if we can’t figure out how to convert its change into a dollar value.
A better distinction
I believe a better way to classify benefits is to consider whether they are all of: measurable, monetizable, and attributable (MMA) to the quality improvement activities. Let’s look at these individually:
Measurable vs Unmeasurable
When we say that something is measurable, we are simply saying that we can define and specify it, and there exists (or will exist, after we collect them) data we can use. Whether or not this measurement and these data are adequate to serve our purposes (i.e., are valid and reliable) is a separate issue. At this point, we are simply interested in identifying whether it is something we can, at a minimum, attempt to measure. In all likelihood, there will be benefits that are within the scope of the intervention and that are important but may simply lack a working definition or data to assess them. For example, improving patient safety may improve a facility’s “brand,” but that may not be measurable either due to a lack of a specific definition or the means or data to measure it.
Monetizable vs non-monetizable
Just because a benefit is measurable does not mean it can easily be translated into monetary terms for use in an ROI analysis. The most common examples of monetizable benefits include things like: changes in utilization, resource use, time, and adverse events (which lead to utilization). Those benefits which are often more difficult or impossible to monetize include benefits such as satisfaction, individual effort and buy-in, quality of life, and changes in capacity or efficiency. These are often measurable by some standard or with some tool (e.g., a survey to assess satisfaction or quality of life), but equating them to monetary values can be challenging, or may be more closely scrutinized by those evaluating the analysis.
Directly attributable vs not
Finally, just because a benefit is measurable and monetizable does not necessarily mean that it is appropriate to use in the calculation of ROI. To be included, it should be directly attributable to the intervention, either in whole or in part. There may be multiple reasons why a benefit may not be directly attributable to the intervention. One reason is the presence of confounding variables or the notion that multiple aspects can affect what is being measure. Multiple simultaneous interventions make it difficult to know how much each one contributed to any observed improvement, and several factors may influence the likelihood of readmission making it challenging to determine how much of an improvement is due to any one particular aspect.
Attribution can also be a challenge in light of existing trends in whatever is being measured. Detractors of the effectiveness of national programs to reduce mortality in Medicare patients, for example, have pointed out that mortality rates were already declining prior to the programs’ initiations, and suggest that it is difficult to know how much of the observed decline was directly due to the program and how much would have been observed simply as a continuation of the already existing trend.
Another situation where it can be difficult to attribute a benefit to the intervention is if there are numerous intermediate steps between the activities involved in the intervention and those that are both measurable and monetizable. For example, an intervention that reduces physician burnout can improve the well-being of physicians, which can improve physician-patient interactions and thus patient satisfaction, which in turn can improve HCAHPS scores that are tied to reimbursement. However, it may be difficult to ascertain how much of the improvement in HCAHPS scores is directly due to the steps taken to reduce physician burnout because at each intermediate step there may be other reasons or factors affecting each result along the way.
Putting them together
Often the distinction between measurable and not, monetizable and not, or attributable and not is difficult to identify. When performing ROI analysis, therefore, potential benefits should be explored and examined up front to see what may be all three of measurable, monetizable, and attributable, and which may struggle in at least one. Then, adjustments can be made to ensure that in addition to appropriate quality measures, relevant value measurements will be taken as well to assess the financial return of the project.