Currently Viewing:
The American Journal of Managed Care December 2019
Clinical Characteristics and Treatment Patterns Among US Patients With HIV
Julie L. Priest, MSPH; Tanya Burton, PhD; Cori Blauer-Peterson, MPH; Kate Andrade, MPH; and Alan Oglesby, MPH
Partnering in Postacute Darkness? CMS Has Data That Will Help
Terry E. Hill, MD
From the Editorial Board: Anne K. Gauthier, MS
Anne K. Gauthier, MS
Preventive/Office Visit Patient Knowledge and Their Insurance Information Gathering Perceptions
Evan K. Perrault, PhD; Katie J. Schmitz, BA; Grace M. Hildenbrand, MA; and Seth P. McCullock, MA
Currently Reading
Cost-Sharing Payments for Out-of-Network Care in Commercially Insured Adults
Wendy Yi Xu, PhD; Bryan E. Dowd, PhD; Macarius M. Donneyong, PhD; Yiting Li, PhD; and Sheldon M. Retchin, MD, MSPH
Catheter Management After Benign Transurethral Prostate Surgery: RAND/UCLA Appropriateness Criteria
Ted A. Skolarus, MD, MPH; Casey A. Dauw, MD; Karen E. Fowler, MPH; Jason D. Mann, MSA; Steven J. Bernstein, MD, MPH; and Jennifer Meddings, MD, MS
A Claims-Based Algorithm to Reduce Relapse and Cost in Schizophrenia
Heidi C. Waters, PhD, MBA; Charles Ruetsch, PhD; and Joseph Tkacz, MS
Cost Burden of Hepatitis C Virus Treatment in Commercially Insured Patients
Christine Y. Lu, PhD; Dennis Ross-Degnan, ScD; Fang Zhang, PhD; Robert LeCates, MA; Caitlin Lupton, MSc; Michael Sherman, MD; and Anita Wagner, PharmD
Delivery System Performance as Financial Risk Varies
Joseph P. Newhouse, PhD; Mary Price, MA; John Hsu, MD, MBA; Bruce Landon, MD, MBA; and J. Michael McWilliams, MD, PhD
Association of Care Management Intensity With Healthcare Utilization in an All-Condition Care Management Program
Hussain S. Lalani, MD; Patti L. Ephraim, MPH; Arielle Apfel, MPH; Hsin-Chieh Yeh, PhD; Nowella Durkin; Lindsay Andon, MSPH; Linda Dunbar, PhD; Lawrence J. Appel, MD; and Felicia Hill-Briggs, PhD; for the Johns Hopkins Community Health Partnership
Outcome Measures for Oncology Alternative Payment Models: Practical Considerations and Recommendations
Jakub P. Hlávka, PhD; Pei-Jung Lin, PhD; and Peter J. Neumann, ScD

Cost-Sharing Payments for Out-of-Network Care in Commercially Insured Adults

Wendy Yi Xu, PhD; Bryan E. Dowd, PhD; Macarius M. Donneyong, PhD; Yiting Li, PhD; and Sheldon M. Retchin, MD, MSPH
This study of claims among adults covered by employer-sponsored plans revealed substantial variations in out-of-network cost-sharing payments. The growth of cost sharing for nonemergent hospitalizations is concerning.
Outcome Measures

We first studied annual use of covered OON care and annual total OOP cost sharing for OON care, including co-payments, coinsurance, and deductibles for any care paid as OON benefits. This included inpatient hospitalizations, outpatient services, and covered prescription drugs filled in OON pharmacies. Spending was aggregated to per-person per-year and was adjusted to 2017 dollars.

Further, we categorized medical services from OON providers based on emergency status and site of service. Because prescription drug fills from OON pharmacies did not fit in any of these categories, they were excluded from this analysis. The categories were (1) nonemergent outpatient visits, (2) visits to EDs that did not lead to a hospitalization, (3) emergent hospital admissions in conjunction with an ED visit, and (4) nonemergent or elective hospitalizations. The OOP cost-sharing spending for OON medical services was also examined, conditioning on OON care utilization in each setting as described above.

Adjustment Covariates

Following the algorithm of the Hierarchical Condition Categories (HCC) risk adjustment model designed for the commercial population, a risk score was assigned to each enrollee.18 The score reflects health conditions associated with expenditure levels in a year and took into consideration enrollee age, sex, and diagnostic conditions in each year. Higher risk scores indicate more complex healthcare needs and potential for higher spending (International Classification of Diseases, Tenth Revision codes were adopted in October 2015, but the HCC scores in our sample were similar before and after the transition).

Health plan characteristics were reflected by plan design types, including health maintenance organizations (HMOs) and exclusive provider organizations, in which enrollees choose from a list of providers for nonemergent care; preferred provider organizations (PPOs) and point-of-service plans, in which enrollees are offered lower cost-sharing levels to use a list of providers; high-deductible/consumer-driven health plans (HDHPs), which include high deductible requirements; and comprehensive plans without network limitations.


A logistic regression model was constructed to estimate the probability of OON care in a year. Among those with OON care, a generalized linear regression model (GLM) using log link was used to estimate the OON cost sharing, given various factors that potentially impact OOP spending for OON care. Further, we estimated the probability of having OON medical care based on the ED status and care settings. Similarly, a GLM was used to estimate cost sharing for OON care in each case.

All models considered health risk scores, plan characteristics, rural residence, state-fixed effects, and year-fixed effects on OOP payments for OON care. Age and sex were accounted for in the algorithm constructing the HCC risk scores and thus were not separately listed as covariates in regression models. Lastly, because access to network providers may differ between rural and urban areas, rural residency was defined as enrollees living in nonmetropolitan areas.

To reflect the national population of ESI enrollees, our analysis included sampling weights constructed based on the Public Use Microdata Sample of the American Community Survey.19,20 In addition, robust clustered standard errors by unique enrollees were computed to reflect that the same enrollees may be observed multiple times.

Several additional analyses were performed. First, because the employee sample could have fluctuated during the study interval, we repeated the analyses for a subsample of beneficiaries who were continuously enrolled across the entire 6-year period. Second, we examined the trend of in-network cost sharing to determine whether the trend differed from that of OON care. Finally, because some employers may have increased or decreased benefits across years, we constructed a model that allowed insurance benefit design to change over time within the same plan type. Detailed model specifications are in eAppendix B.

Copyright AJMC 2006-2020 Clinical Care Targeted Communications Group, LLC. All Rights Reserved.
Welcome the the new and improved, the premier managed market network. Tell us about yourself so that we can serve you better.
Sign Up