A methodology for identifying behavioral health advanced practice registered nurses in administrative claims. (18th April 2022)

• Record Type: Journal Article
• Authors: Richard, Jessica V.; Huskamp, Haiden A.; Barnett, Michael L.; Busch, Alisa B.; Mehrotra, Ateev
• Is Part Of: Health services research. Volume 57:Number 4(2022)
• Page Start: 973
• Page End: 978
• Publication Date: 2022-04-18
• Subjects: administrative claims -- advanced practice providers -- advanced practice registered nurses -- Medicare -- mental health -- methods; Medical care -- Periodicals; Medical care -- Evaluation -- Periodicals; Hospital care -- Periodicals; Health services administration -- Periodicals; 362
• DOI: 10.1111/1475-6773.13974
• Languages: English
• ISSNs: 0017-9124

Abstract

Abstract: Objective (Study Question): Advanced practice registered nurses (APRNs) play an increased role in mental illness treatment. Health services research that uses claims to study mental health is often limited because behavioral health nurse practitioners (i.e., APRNs who specialize in mental illness, also known as psychiatric mental health APRNs) cannot be easily identified in claims data. We describe two methodologies to identify behavioral health APRNs in administrative claims. Data Sources/Study Setting (w/ Hospital/Institution Setting Anonymized): We use 2010–2018 claims from the traditional Medicare fee‐for‐service program along with 2010–2019 commercial claims and Medicare Advantage data from the OptumLabs Data Warehouse (OLDW). Self‐reported specialty data from the National Plan & Provider Enumeration System (NPPES) were used for validation. Study Design: For each APRN, we calculated the percentage of visit diagnoses and of prescriptions in each database that were for mental health and classified those with ≥80% as behavioral health APRNs. We validated our definition with NPPES self‐reported specialty for Medicare data. Data Collection/Extraction Methods: Not applicable. Principal Findings: Among APRNs with 10+ visits, 10, 978 (8.1%) in Medicare and 9829 (11.7%) in commercial claims data met our visit‐based criteria as behavioral health APRNs. Among APRNs with 10+ prescriptions, 8160 (6.2%) in Medicare and 16, 538 (9.0%) in commercial claims data met our prescription‐based criteria as behavioral health APRNs. Among the APRNs who self‐reported they were behavioral health APRNs, 92.8% and 90.5% met our visit‐based and prescription‐based criteria, respectively. Conclusions: We present and validate two methods of identifying behavioral health APRNs in claims that can be used by other researchers.