Computer system and method for generating healthcare risk indices using medication compliance information
View Patent ↗A healthcare risk index is generated using a patient or individual's pharmacy claims. The index may be used to explain and predict variation in pharmacy-related costs and variation in total healthcare costs or utilization. In particular, the index is generated by first examining the individual's pharmacy claims to identify any chronic conditions possessed by the individual. Similarly, the individual's pharmacy claims are examined to identify any compliance medications prescribed to the individual. The chronic condition information is used to generate a chronic condition score by summing regression coefficients for each chronic condition possessed by the individual. Likewise, the compliance medication information is used to generate a compliance medication score by summing products of regression coefficients for each compliance medication prescribed to the individual with associated medication supply weights. From there, a modified chronic condition score is generated by multiplying the chronic condition score by an overall chronic condition regression coefficient. The modified chronic condition score may then be further modified by subtracting a no-claims weight from the chronic condition score in cases where the individual has no pharmacy claims. Finally, the risk index may be determined by summing the modified chronic condition score and the compliance medication score.
1. A method comprising
accessing, on a processor, a plurality of pharmacy claims associated with an individual;
generating, on a processor, a raw risk index indicative of at least one of said individual's medical costs, acute medical conditions, and variation in medical costs, by using the plurality of pharmacy claims;
processing, on the processor, the plurality of pharmacy claims to identify a plurality of compliance medications prescribed to the individual;
associating, on the processor, a regression coefficient with each compliance medication of the plurality of compliance medications;
determining, on the processor, compliance of the individual with the plurality of compliance medications to generate a plurality of compliance medication scores a compliance medication score of the plurality of compliance medication scores being associated with a compliance medication prescribed to the individual;
multiplying, on the processor, each of the plurality of compliance medication scores with an associated regression coefficient of the plurality of coefficients, the plurality of regression coefficients including the regression coefficient; and
modifying, on the processor, the raw risk index based on multiplying each of the plurality of compliance medication scores with the associated regression coefficient to generate the risk index.
2. The method of claim 1 , wherein the compliance medication score is determined by summing a total supply of the compliance medication prescribed to the individual and multiplying by the regression coefficient.
3. The method of claim 2 , wherein the regression coefficient is generated from a log transform.
4. The method of claim 2 , wherein the regression coefficient is generated from a square-root transform.
5. The method of claim 1 , wherein the compliance medication score represents compliance of the individual with the compliance medication prescribed to the individual.
6. A non-transitory computer readable medium, which when executed by one or more processors, cause the one or more processors to perform the following operations:
access a plurality of pharmacy claims associated with an individual;
generate a raw risk index indicative of at least one of said individual's medical costs, acute medical conditions, and variation in medical costs, by using the plurality of pharmacy claims;
process the plurality of pharmacy claims to identify a plurality of compliance medications prescribed to the individual;
associate a regression coefficient with each compliance medication of the plurality of compliance medications;
determine compliance of the individual with the plurality of compliance medications to generate a plurality of compliance medication scores, a compliance medication score of the plurality of compliance medication scores being associated with a compliance medication prescribed to the individual;
multiply each of the plurality of compliance medication scores with an associated regression coefficient of the plurality of coefficients, the plurality of regression coefficients including the regression coefficient; and
modify the raw risk index based on multiplying each of the plurality of compliance medication scores with the associated regression coefficient to generate the risk index.
7. The non-transitory computer readable medium of claim 6 , wherein the compliance medication score is determined by summing a total supply of the compliance medication prescribed to the individual and multiplying by the regression coefficient.
8. The non-transitory computer readable medium of claim 7 , wherein the regression coefficient is generated from a log transform.
9. The non-transitory computer readable medium of claim 7 , wherein the regression coefficient is generated from a square-root transform.
10. The non-transitory computer readable medium of claim 6 , wherein the compliance medication score represents compliance of the individual with the compliance medication prescribed to the individual.
11. A system comprising:
a computing device including a processor and a memory coupled to the processor to:
access a plurality of pharmacy claims associated with an individual,
generate a raw risk index indicative of at least one of said individual's medical costs, acute medical conditions, and variation in medical costs, by using the plurality of pharmacy claims,
process the plurality of pharmacy claims to identify a plurality of compliance medications prescribed to the individual,
associate a regression coefficient with each compliance medication of the plurality of compliance medications,
determine compliance of the individual with the plurality of compliance medications to generate a plurality of compliance medication scores, a compliance medication score of the plurality of compliance medication scores being associated with a compliance medication prescribed to the individual,
multiply each of the plurality of compliance medication scores with an associated regression coefficient of the plurality of coefficients, the plurality of regression coefficients including the regression coefficient, and
modify the raw risk index based on multiplying each of the plurality of compliance medication scores with the associated regression coefficient to generate the risk index.