Benchmarking is a management tool used in many industries to answer the never-ending question: “How does my operation compare against those of my peer group?” In simple terms, benchmarking is defined as the sharing of performance data among a group of non-competing businesses in the same industry. The group uses the data to identify which participant has the best performance and then this becomes the benchmark against which everyone is measured in the future. The group then studies the benchmark participant's processes to identify those methods that are producing this above-average performance with the goal of quickly migrating these practices to each member of the group. The result should be that each participant will be performing similar to or better than the “benchmark participant.”
On the surface it sounds simple—but there are some important hurdles that need to be jumped before the benefits of benchmarking begin to pay off. These include:
Getting enough benchmarking participants to ensure that a sufficiently wide variety of business practices will be compared. You don't want either a very small group or a homogeneous group that all run their operations the same way.
Selecting and tracking the right measurement criteria so that both the symptoms and the root cause of the good performance can be clearly identified.
Ensuring that all participants will have easy access to each other's business practices once the benchmark member(s) have been identified. Many benchmarking efforts have failed because the “good performers” suddenly became uncooperative when they discover they are “better” than the rest of the participants.
In 2004, I first started applying benchmarking techniques to the therapy departments of skilled nursing facilities (SNFs) as a consultant to a small Midwestern contract therapy provider. I quickly determined that the therapy profession and nursing homes, in general, had not yet begun using proven techniques I had learned 20 years earlier as a manager at General Electric. After a few months as a consultant, I joined the firm as chief operating officer and began implementing a rudimentary benchmarking effort that consisted of a monthly collection and analysis of 10 key operational data elements from 12 different teams of therapists. Since I was new to the industry and was not a clinician (I actually am an ex-rocket scientist, with a degree in mechanical and aerospace engineering), I was conservative and erred on the side of too much data, just to be sure we were capturing the real drivers of therapy performance.
Because higher Medicare Part A (Med-A) RUG levels correlate to both better quality of care for residents and higher reimbursement for our clients' facilities, we decided to benchmark this metric first. Our hope was to identify those therapy teams with the best RUG levels, study their practices, document their unique procedures, and then cross train the other departments with these procedures, with a win-win outcome for both patients and facilities.
Our first step was to isolate and remove the RUG scores of short-term rehab patients (primarily those who had just undergone joint replacements) because their therapy goals were very straightforward—it is common knowledge that these folks are at Ultra High treatment levels from the day of admission until the day they leave. By filtering out these high-end rehab patients, we were then able to concentrate our efforts on the larger and more challenging population, namely those long-term care residents who find themselves in therapy with Medicare Part A benefits after an unplanned visit to the hospital.
After six months of collecting, filtering, and comparing data on this larger group of patients, we confirmed our suspicion that it is a very homogeneous group with an almost predictable distribution of common ailments. We also discovered that across 12 SNFs in four different states, the percentage of the long-term care population with Medicare Part A benefits did not vary from building to building by more than a few percentage points, regardless of the time of year.
We found, however, that one of our therapy teams had consistently higher RUG scores than the other 11 for this very similar patient population. Not only was this one team consistently producing RUG scores that were 35% Ultra High (versus a 10% average for all others), they also had a consistent 10% longer length of stay (LOS) on therapy caseload. As we drilled down deeper and compared clinical data from this benchmark team against the other teams, we saw similar approaches with physical and occupational therapy, but very different approaches with speech therapy.
It quickly became apparent that the two speech-language pathologists (SLPs) on this team had created a set of protocols that dramatically improved the health of the patients while also improving the RUG scores for the SNF. They had observed repeatedly that a brief hospital visit by an elderly person often has a traumatic effect, with the symptoms being oral motor deficits and/or an increase in cognitive communication problems. They developed a set of screening techniques to establish precise baselines for these conditions for all residents, as well as expanded goals and more precise ways to measure progress against these goals.