Calculation methods

Throughout Clinical Analytics, you have the option to select the calculation method for your measures. We have defined the calculation methods for you here to help with you decision.

Definitions

Measure values for your internal data can be calculated and displayed in a few different ways.

Value

This is typically the sum of all values in the set.

Composite

When you see a value labeled as a composite, this is typically a fraction or rate with a numerator and denominator.

If you are using an Rate measure (like Mortality Rate, for example), your measure appears with several columns, including Encounters (the total number of discharges) and Deaths (the observed number of encounters discharged as Expired). The Composite column is the rate fraction: Deaths/Encounters.

If you are using an O/E measure (like Mortality O/E, for example), your measure columns also include an Expected column. The Expected value is the benchmark rate multiplied by the Encounters. The Composite value is then the Observed Deaths divided by the Expected Deaths. This value is an index value with no units and should be interpreted such that an index value over 1 indicates more observed deaths than expected while an index value less than 1 indicates fewer observed deaths than expected.

Avg: Arithmetic Mean

This is the typical average you are likely most familiar with. All of the values of the set are added together, then the sum is divided by how many values are in the set.

Geometric Mean

The geometric mean is calculated by multiplying together all of the values in the data set, then taking the nth root of the product, where n is the number of values in the set.

Benchmark

This is the risk-adjusted equivalent value for that measure from the benchmark data. See Benchmark calculations for more information.

Opportunity

The opportunity for a measure is an indication of improvement potential. This is essentially the difference between your facility's measure value and the risk-adjusted benchmark value from the benchmark data set. Green cells indicate the facility is performing better than the benchmark while red cells indicate an area of improvement for the facility. See Opportunity for more information.

Percentile

Quantiles are cut-points in the data that divide the range of values into equal probabilities. Clinical Analytics Software chooses to cut the data into 5% intervals (1% chunks are called percentiles). All values in the database for the measure are ordered from least (0th percentile) to greatest (100th percentile), then the system sets the 5% quantile cutoffs. When you request a particular quantile, the value you see is greater than the chosen percent of the data.