Using an Annualization Factor on the AB34 Cumulative Data

We collect one year’s worth of historical data on every member in the program to compare to their post-enrollment statistics. However, most members have not been in the program for exactly one year – They may have been in the program anywhere from 1 day up to (currently) over 4 years. To complicate things even more, on a program level this is changing with every new enrollment and disenrollment and with every passing day. Because of this, we need to “annualize” the post-enrollment data to compensate for the average length of enrollment for all clients currently enrolled in the program, so that the data grids compare one year pre-enrollment to one year post-enrollment (apples to apples instead of apples to oranges).

1. In order to do this, you need the average tenure of the currently enrolled clients in the program. Consider the following (hypothetical) program of 10 members:
Member
Length of Stay in Program
1
398
2
243
3
25
4
579
5
2
6
634
7
132
8
234
9
89
10
254
Total
2590


The total number of days that the 10 currently enrolled members have spent in the program is 2,590. Since there are 10 members in the program, divide 2,590 by 10 to arrive at an average tenure of 259 days per member. This is the average tenure for the currently enrolled AB 34 members.

2. Next, you divide one year (365 days) by the average tenure to get the annualization factor:

365
_______________ = 1.41 annualization factor
259 average tenure days

3. Finally, you apply the annualization factor to the post-enrollment cumulative data:

Non-annualized Number of Hospital Days Since Enrollment: 923

Multiply this by the annualization factor

Annualized Number of Hospital Days Since Enrollment: 1,301

Thus, if a program has a tenure of LESS than a year (as in this example), their post enrollment numbers will be GREATER than their raw numbers. If a program has a tenure of MORE than a year, then their post enrollment numbers will be SMALLER than their raw numbers. (And, of course, if the average tenure was EXACTLY 1 year, then the annualized numbers and the raw numbers would be equal).

Back