In many locations throughout the Trella Health Marketscape, there are "total" counts that don't equal what seems to be the counts of all parts. Here is a general rule for using Marketscape metrics with some specific examples.
Most metrics are calculated uniquely to satisfy specific parameters defined by the connection between the column and the row. For example: how many Medicare patients were admitted to post-acute facility A in Quarter 1? To calculate this metric, we look at every claim that fits both criteria to get a count of patients for that cell in the table.
What we do to calculate the count for that one cell is then repeated in the next row for post-acute facility B. It is possible that some of the same patients will have claims for this different facility in a different row. The patient is counted in each location.
In general, whenever you see the word Total anywhere in the product it is best to think of the metric as a uniquely calculated number and not the sum of other metrics.
Quarterly Patient Counts vs. Annual Patient Counts
Question - In this case, we expect the counts in the four quarters to add up to the One Year Total column. But, when we add the numbers for the four quarters, the total is 19,869 Medicare Patients, not the 16,151 in the Total column. Why?
Explanation - The reason is that all of the counts in these columns represent counts of distinct patients. If a patient was admitted into post-acute care in more than one quarter, each admission would be counted in the relative quarter, but only once in the "Total" column.
Observation - The size of the difference between the totaled columns and the One Year Total is interesting. All columns reflect one patient population. The difference between the two totals is indicative of a provider who is seeing a larger percentage of patients multiple times. If this percentage of visits outstrips other providers in the same specialty, that could suggest potential discovery through digging deeper.
For counts compiled over a time period, it is better to not think of the first column as a total, which implies that you are combining or adding the four columns. It is better to think of the metrics provided as an annual metric and quarterly metrics.
Claims don't add up
These claim type metrics can be found on the Explore page for physicians.
Question - The patient counts that are tied to each claim count look like divisions of the total. Why don't the three claim types add up to the Physician HHA Patients? In this example, the three claim types add up to 929, not 490.
Although this example is taken from Marketscape for Home Health, the same principle applies to the similar claims in Marketscape for Hospice and Marketscape for SNF.
Explanation - Although the counts in each of the claim type columns does represent distinct patients; for a given physician/patient pair, some claims could represent more than one claim type. For example, a claim that was the last one prior to admission that was within 90 days of admission would be counted in both columns; Patient 3 Months Prior and Last Claimed Patients. A single claim could therefore be counted in more than one column, but that patient would only be counted once under Physician HHA Patients.
Columns don't add up
Explanation - The definition for this column helps - Count of patients discharged from the facility in each row who had claims filed for Home Health at the Home Health Agency on this page within 30 days of discharge. In this case there are two ways duplicates end up in the data. The first is there are multiple facility stays within 30 days prior to the start of the HH stay, and each of these stays is counted wherever they apply. The second is the same patient has multiple HH stays, and each of these stays is counted. The number at the bottom removes the duplicates.
The counts for each listed facility is a count of all D/C to post-acute events during the reporting period. The number at the bottom is distinct patients treated during that same period. Two views of the same events.