How Can All-Payer Claims Databases Support Insurance Regulation?

By: Joel S. Ario | Kevin Casey McAvey

Editor’s Note: On March 24, 2018, Joel Ario, managing director at Manatt Health, and Kathy Hempstead, a senior advisor to the Robert Wood Johnson Foundation, delivered a presentation on All-Payer Claims Databases (APCDs) to the National Association of Insurance Commissioners (NAIC) Regulatory Framework Task Force. The presentation, summarized below, provided a detailed look at APCDs, including use cases and caveats. Sharing preliminary considerations, the presentation was the product of extensive research and nearly a dozen interviews with state APCD and insurance leaders. Click here to download the full presentation free.

The final presentation and report will be delivered at the next national NAIC meeting in August.


What Are APCDs?

All-Payer Claims Databases (APCDs) are centralized state data repositories for health insurance membership and claims records. States with APCDs typically require insurers operating in their markets to regularly submit medical, pharmacy and dental claim files for all members residing in and/or contracted in their state. Mature APCDs allow for cross-payer, marketwide enrollment, utilization and payment trend analyses and are an increasingly leveraged data resource by regulators, policymakers and researchers.

Maine was the first state to establish an APCD in 2003, and since that time, more than a dozen other states have either followed suit or are in the process of doing so. (Note: As shown on the map below, additional states have established voluntary “multi-payer” or partial APCDs.) 

According to results of Manatt Health’s first APCD Capacity Survey—co-administered with the National Association of Health Data Organizations—on average, APCDs include data for three-fifths of their states’ populations, with robust coverage of the private fully insured, Medicaid and even Medicare Advantage populations. This breadth and depth of coverage make APCDs a particularly attractive data resource for insurance departments.

Source: Manatt Health APCD Catalogue (accessed March 1, 2018)

Four APCD Use Cases for Insurance Regulators

Mature APCDs have the potential to inform insurance department policy and program goals across four areas:

  1. Price Transparency: APCDs can provide a wealth of dynamic price data to inform the decisions of regulators, policymakers and even consumers. New Hampshire, Maine and Maryland, for example, have used data from their respective APCDs to populate consumer websites, where prospective patients can compare estimated costs for procedures and services across providers. Frequently complementing insurer-developed, plan-specific patient cost calculators, state consumer websites are often part of broader strategies to promote consumer empowerment and market competition. Other states have leveraged their APCDs to produce statewide cost-driver reporting and to identify, investigate, and respond to unexplained price variations between payers, providers and services.
  2. Rate Review: APCDs include data that insurance regulators can use to enhance their understanding of insurance markets for rate review purposes. Mature APCDs have the potential to generate useful trend reports on most claims categories of interest to insurance regulators, highlighting member experience differences by geography, demographic characteristics and time period. Mature APCDs also can allow for impact monitoring of delivery system reforms, quality initiatives and other ad hoc analyses (e.g., frequency and severity of claims for 1332 waiver reinsurance program proposals). The Maryland Insurance Administration uses its state’s APCD, the MD Medical Care Database, for example, to check payer submissions and run deeper analyses. Oregon’s Department of Consumer and Business Services uses its APCD (APAC) data in its review of premiums for individual and small group health plans, while Massachusetts’ Center for Health Information and Analysis has worked in partnership with its Division of Insurance to reduce the statewide payer reporting burden by directly sourcing several insurance reports from its APCD (supporting “administrative simplification”).
  3. Network Adequacy: APCDs have the potential to be a significant resource in helping insurance regulators set and monitor network adequacy standards. The New Hampshire Insurance Department, for example, views its APCD, the New Hampshire Comprehensive Healthcare Information System (CHIS), as a prime resource to help it monitor the impact of its new network adequacy regulation (IR 2701), which measures adequacy by service category rather than provider type. Mature APCDs have the potential to help insurance regulators answer critical questions about payers’ networks, including: Which providers deliver what services? Where are certain services scarce? And how much do prices differ between in- and out-of-network services by plan?1 APCDs may also be able to help inform specific policy questions, such as the prevalence of out-of-network charges at in-network facilities (i.e., surprise billing).
  4. Responding to the Opioid Crisis: APCDs have helped states to monitor and develop strategies for combating public health crises, including the opioid epidemic. APCDs can provide information on prescribing patterns at regional and physician levels, as well as the availability of treatment (e.g., medication-assisted treatment (MAT), naloxone or residential programs). APCDs also allow states to look at an issue in the broader context of how substance use disorders change over time as drug availability changes. Massachusetts’ APCD served as the data-backbone of the state’s new “Promote Prevent” plan to promote mental, emotional and behavioral health. The Massachusetts APCD was linked to more than a dozen other health and social service data sets to provide policymakers with a comprehensive depiction of the Commonwealth’s most vulnerable populations and potential root causes (e.g., access) for their behavioral health issues. Virginia’s APCD was similarly used to identify trends in opioid prescription volume, refills and dispensing habits, while Colorado’s APCD was used to identify prescription fill anomalies by matching drug users with reported clinical conditions.

APCD Cautions

While APCDs have the potential to substantively inform numerous insurance agency programs and priorities, most APCDs nationally are not yet mature enough and/or do not have the staff capacity that would allow for simple report request fulfillment. When approaching interagency collaboration, it is essential for insurance leadership to keep the following considerations in mind:

  • Long-Term Partnership: Work with APCD agencies should be viewed as part of a long-term partnership, ideally fueled by small, shorter-term goals (and, hopefully, “wins”). Insurance and APCD leadership should work together to develop realistic goals and milestones jointly. (Also, keep in mind, the more APCDs are used, the more valuable they become: Use is the best data quality check, and it establishes foundational agency functions, processes and knowledge that will better support efficiencies in future collaborative efforts.)
  • Communications: APCD and insurance department staff may not always speak the same language. For example, when referring to “membership,” one agency’s staff may intend to convey “a member months average for state residents”; another, however, may interpret “a point-in-time count by state situs.” Developing clear, shared business specifications is foundational to any successful project partnership. Business specifications are the road map by which technical specifications (i.e., programming logic and code) will be developed.
  • Completeness, Timeliness and Accuracy: Though APCDs are significant data assets, they are not always the best data assets to answer every question. Depending upon the state and use case, APCDs may not have data that is relevant (e.g., clinical records), timely enough (e.g., claims lag) or complete enough (e.g., payer data submission anomalies) for use. Further, depending upon the “hosting agency” for a given APCD, various APCD data fields may have been emphasized for integrity. For example, when APCDs are hosted within Medicaid or health departments, critical foundational fields for insurance use—such as Situs or License Type—may not be as thoroughly tested and vetted as others, such as diagnosis codes (also important). Regardless of whether there is an imminent opportunity for APCD agency collaboration, insurance departments should be active and engaged stakeholders in their continuing development.
  • Staffing (and Funding): APCD agencies are financially lean, with limited staff. Clearly delineating project roles, timelines and committed resource needs for any collaboration is an important step, as is identifying where external counsel would be critical (i.e., don’t reinvent the wheel). Joint agency collaborations may also be able to identify and leverage additional state and federal funding, especially where new funds may be available to address public priorities (e.g., opioids) or where long-term value propositions can be presented (e.g., administrative simplification).

APCDs can be—and are already, for some state insurance departments—tremendous and dynamic data assets. Their value will only continue to increase as use improves data quality and uncovers new use cases.

1Where in-network APCD flag is available and tested