Comparing buildings is one of the main reasons multi-facility organizations build dashboards. It is also one of the easiest ways to mislead. A 150-bed building will almost always have more overtime hours than a 50-bed building, simply because it has more staff. Comparing raw totals would suggest a problem where none exists.
Fair comparison means adjusting for the differences that are not the building's fault, so that what remains reflects actual performance and practice. This is called normalizing, and it is simpler than it sounds.
Normalize for size
The most basic adjustment is to express measures relative to size.
- Staffing hours per resident day instead of total hours
- Overtime as a share of total hours instead of total overtime hours
- Calls per bed or per resident instead of total calls
- A/R relative to monthly revenue, as in days in A/R, rather than total balances
These measures let a large and a small building appear on the same scale.
Account for resident mix
Buildings differ in the needs of the residents they serve. A building with a large short-stay rehabilitation population has different workflows, admissions volume, and billing patterns from one that primarily serves long-stay residents. When comparing, consider grouping buildings with similar profiles or at least interpreting results in light of the differences.
Consider the local market
Labor markets, referral landscapes, and payer environments vary by community. A building in a tight labor market may have higher agency use for reasons outside its control. Include context notes, and compare buildings against peers in similar markets where possible.
Compare against each building's own history
One of the fairest comparisons is a building against itself over time. Is it improving or declining relative to its own baseline? Pair this with shared targets, so leaders can see both trajectory and position. A building that has made a big improvement but is still above a shared target deserves recognition, not criticism.
Watch small numbers
Small buildings produce noisier statistics. A single event can swing a rate noticeably. When comparing, show counts alongside rates, use multi-month averages, and be cautious about drawing conclusions from one period.
Use ranges, not only rankings
Rankings imply that the building at the bottom is failing and the building at the top is excellent, even when the difference is trivial. Ranges and bands, such as "within expected range," "worth a look," and "needs attention," tell leaders more and cause less defensiveness. Define the bands using your organization's data and judgment.
Keep definitions identical
Normalization is pointless if buildings calculate measures differently. Document each definition, including the numerator, the denominator, the time window, and what is excluded. Revisit them when systems or processes change.
A hypothetical example
Imagine a hypothetical group where one building seems to have the highest overtime hours in the region. Normalized as a share of total hours, it is in the middle of the range. Another building, with modest total overtime, turns out to have the highest share, concentrated in a few employees. Without normalizing, leaders would have focused on the wrong building.
Use comparisons to learn, not to punish
The best use of comparison is to ask what the strongest performers do differently and whether it can be shared. If comparison becomes a scoreboard that leads to blame, administrators will start managing the numbers instead of the building. Invite curiosity.
Helpful questions
- What does this building do that others could adopt?
- What constraints does this building face that others do not?
- What support would help the building at the low end of the range?
Mistakes to avoid
- Comparing raw totals. Always adjust for size.
- Ignoring context. Mix and market matter.
- Overinterpreting single periods. Use trends.
- Letting definitions drift. Consistency is the foundation.
- Turning comparison into ranking for its own sake. Use it to guide support.
Start with a few normalized measures
Pick five or six measures, normalize them carefully, and display them with trend and context. Add more once leaders trust the basics.
Let the system do the arithmetic
Normalization is not difficult mathematically, but doing it manually across many buildings and many measures is tedious and error-prone. Automating it ensures that the numbers leaders see are consistent.
CarePulse Analytics builds normalized multi-facility views so comparisons between buildings are fair and clear. If you would like to see what that might look like for your group, a demo is a good place to start.