Background: About ~50% of hospitals/emergency departments (ED) do not provide psychiatric services1 despite mental health related ED visits being common. Patients who present to the ED for psychiatric concerns often end up “boarding” in the ED, meaning that evaluation is complete, but the patient has not been transferred to the appropriate level of care. It is well established that psychiatric boarding can negatively affect patients, staff, and health systems. One possible root cause for limited access to ED psychiatric services could be reimbursement. Differential reimbursement between outpatient psychiatric and medical/surgical services is also well-articulated; however, less is known about this topic in the ED, and there has been little published regarding ED-based psychiatric patient billing patterns. Notably, in a small 2019 survey of ED physicians, there was significant billing variability for psychiatric boarders; 35% billed for observation, 31% did not bill for observation, and 35% were unsure if they billed for observation.2 This may indicate that facilities are under-billing for observation services.
Objective: The purpose of this study was to describe billing, and by proxy practice patterns, in non-boarders and boarders accessing emergent psychiatric services.
Methods: This retrospective cohort study analyzed commercially insured administrative claims data with the most recent 2 years of data (2021, 2022) from the Merative™ Marketscan® Commercial Database.3 The study cohort included adults with a minimum of one ED-associated claim, a primary mental health/substance use disorder diagnosis (F-code), and a non-null procedure code. The primary outcome was frequency of pre-defined CPT/HCPCs code frequency (67 codes). Patients could have more than one ED encounter. Approved under Stanford University IRB PHS 40974.
Results: There were 112,985 unique enrollees with 359,504 associated claims. Fifty-five pre-selected CPT/HCPCs codes were present in the sample. Most claims had a service encounter of < 1 day (n = 344,901; non-boarder) as compared to encounters >= 1 day (n = 14,603; boarder). For non-boarders, top codes were 99284 (32%; n = 112,025), 99285 (25% n = 85,832), and 99283 (22%; n = 77,173). For boarders, top codes were 99285 (34%; n = 5,017), 99284 (21%; n = 3,011), and G0378 (17%; n = 2,472). ED-related medical evaluation codes were the most common for non-boarders (86%, n = 294,933) and boarders (68%, n = 9,919). Psychiatry-related codes occurred in 8% of non-boarder claims (n = 28,733) and in 15% of boarder claims (n = 2,145). Observation codes were in 6% of non-boarder claims (n = 18,989) and 17% of boarder claims (n = 2,493). Finally, crisis-related billing codes were infrequent <1% for non-boarders (n = 2,246) and boarders (n = 46), respectively.
Conclusion: Observation billing codes were associated with 17% of ED boarder claims, potentially indicating underused codes, and thus, that these ED visits are being under-reimbursed. Additionally, the lack of psychiatric-related claims, and infrequent use of crisis-codes, across the sample suggests that access to psychiatric and crisis-related services in the ED setting may be worse than prior estimates. Moreover, the use of boarding codes in non-boarder claims may signify that these patients could be considered boarders in future analyses.
Learning Objectives:
Describe some of the potential contributing factors related to billing heterogeneity in high-acuity settings for patients with psychiatric and substance use disorders.
Articulate that observation billing codes are likely underused and may result in under-reimbursement for this patient population.
Understand that the infrequent billing of psychiatric services may indicate substantively lower rates of access to psychiatric services in the ED than previously described.
