Pharmacogenetics Testing for Personalized Medication: Investigation of the Coverage Policies in Arizona

Abstract

Specific Aims: To determine which Arizona health insurance companies provide coverage for pharmacogenetic testing, and what those specific criteria are for coverage. In addition, providing these specific criteria of pharmacogenetic testing that will be valuable to physicians and patients in our continual improvement in personalized medicine.

Methods: Retrospective study review of insurance companies in Arizona from a previous study to determine if changes of genetic coverage were made during April 1st, 2016 to February 25th, 2018. Changes to genetic coverage as well as genetic counseling, BRCA 1 and 2, tumor marker screening, cytochrome P450 (CYP P450), prior authorization requirements, and co-payments of each Arizona plan were conducted and reviewed. Data was obtained through a questionnaire that focused on the above criteria and was assessed through phone calls with customer service representatives of each participating Arizona insurance plan. Data then was recorded based on: 0 – “Unknown,” 1 – “Yes”, 2 – “No.” A Chi square test was used in assessment of the data obtained.

Main Results: The subjects involved in this study included 36 insurance plans, however one of the plans did change and no longer existed. From the findings it showed that only three portions were statistically significant from our previous data which were: if it required a copay (p = 0.008), if they covered CYP P450 polymorphism screenings (p = 0.0078), and if they covered tumor marker screenings (p = 0.0076).

Conclusions: Comparing our pre-data (two years ago) to our new collected data, the same number of insurance companies covered genetic testing. From the questionnaire that was used to collect our information, only three areas had statistically significant changes when comparing the pre-post data. This included copay requirement (P=0.0080), covering CYP P450 polymorphism screenings (p = 0.0078) and covering tumor marker screenings (p = 0.0076). When comparing the overall data, it showed that there was a statistically significant difference between our pre-post data (p = 0.009).