How automated insurance authorization affects transparency, provider burden, and payer policies and processes: The true impact of prior authorization.

5 Background: Insurance prior authorization (PA) is pervasive in modern oncology treatment, with both Medical and Radiation Oncology ranked among the top 5 in a survey of PA burdens across all medical specialties. The cumbersome process delays patient care and increases provider friction, with a major contributor being the perceived lack of transparency regarding PA decisions. Challenges arise because policies are often located behind portal logins, are inaccessible within the clinical workflow, or are too broad to capture patient nuances. We hypothesized that integrating automated insurance authorization software into the clinical workflow would enhance transparency to payer policies. Additionally, we investigated whether these policies could predict whether authorizations would be issued. Methods: We tracked radiation medicine treatment PA requests at a multisite academic healthcare network. Cases were tracked utilizing a standard clinical EMR. We concurrently implemented InsightRT (Siris Medical, Division of TurningPoint Healthcare), an EMR-integrated software system, to automate authorization documentation by capturing and transmitting patient and treatment characteristics. The online portal informed whether a prescribed treatment met the payer policy and tracked the approval process. A pilot subset of physicians was transitioned to using InsightRT. Results: A total of 1418 non-Medicare/Medicaid PA requests were tracked, with 1235 using the traditional clinical workflow and 183 utilizing InsightRT. The software system indicated 162 (88.5%) were “in policy” and only 2 prescriptions were eventually denied. Of the 21 (11.5%) prescriptions that were "Out of Policy" as indicated by the software, only 1 was denied. Sensitivity and specificity of Insight RT policy determination was 98.8% and 97.6% respectively. Overall, there were 43 denials across a total of 1418 cases with payer review only downgrading or changing 4 cases out of 1418, or 0.3%. The other 39 denials were overturned, with 23 (59%) by peer-to-peer (P2P), and 16 (41%) by documentation review. Conclusions: Automated insurance prior authorization provides increased transparency to the provider with high sensitivity and specificity. However, improvements to payer policies and processes are needed to decrease unnecessary burden and delays to care, as the true impact of PA is limited to 0.3% in this cohort. All cases from 12/23 - 5/24. All Cases from non Medicare/Medicaid 1418 Cases processed in InsightRT 183 Cases processed in EMR and not InsightRT 1235 Denials 43 Overturned denials that went to P2P 23 Overturned denials that only needed clinical documentation 16 Cases with confirmed P2P that were downgraded 4 The True Impact of PA - % of all cases changed due to PA 0.30%