Navigating Substance Use Disorder (SUD) Data Availability in APCDs
As concern about the growing incidence and impacts of substance use disorder (SUD) continue to increase, so has the demand for reliable and comprehensive data to guide policy and inform treatment strategies. To support this work, many researchers are looking to statewide all-payer claims databases (APCDs) – critical data resources that collect detailed information from health plans based on claims paid to healthcare providers and hospitals for services rendered.
Onpoint’s decades of experience working with APCDs gives us a unique perspective on the challenges inherent in analyzing SUD data from claims. This fall, Senior Data Analytics Manager Gina Robertson participated in an advance session of the annual conference of the National Association of Health Data Organizations (NAHDO) that highlighted key findings from our work and offered strategies for navigating SUD data availability within APCDs. Here’s a closer look at some of the insights that she shared.
The perception of SUD data availability in APCDs is not always the reality.
A significant barrier to SUD-related research is the broad perception by researchers, policymakers, and other approved users that SUD data in APCDs is always redacted, very limited, or simply unavailable. This perception – and the more complicated reality, discussed below – is mostly due to the fact that different health plans have different interpretations of federal privacy regulations like 42 CFR Part 2, which governs the confidentiality of SUD-related patient records, when reporting their data to an APCD.
While it is true that there are some limitations in using SUD data reported to APCDs, the data remains both widely available and highly valuable. By adopting approaches that optimize the available data, researchers can take advantage of the comprehensive data in an APCD to carry out important all-payer and all-population analyses related to SUD.
How Onpoint identified redacted SUD data in our APCD client data.
One of the first steps in using SUD data is understanding whether health plans are reporting or redacting – that is, withholding – SUD-related fields in their APCD data. Onpoint’s most recent evaluation of the quality and availability of SUD data in our clients’ APCDs began with this key assessment step, determining the redaction rates of common SUD-related fields.
Redactions usually occur in fields that have SUD-specific codes, including:
- Diagnosis codes (e.g., ICD) that pertain to a patient’s visit
- Procedure codes (e.g., CPT) that detail the specific services provided
- Provider specialty codes that identify a healthcare provider’s area(s) of focus
- National Drug Codes (NDCs) that specify the prescriptions used in a patient’s treatment
This analysis found that focusing on ICD diagnosis codes was the most effective approach for assessing the availability of SUD data within APCDs. These diagnosis codes are consistently reported across claims and offer a more comprehensive view than procedure, provider specialty, and drug codes, which appear frequently in claims but not with the same consistency.
Our cross-client evaluation found wide variation in SUD data availability across states and health plans.
Following our initial evaluation, Onpoint conducted surveys with health plans on behalf of our clients to better understand their specific redaction practices, which confirmed our decision to focus on diagnosis codes. We then benchmarked SUD data reporting rates and calculated the percentage of medical claim lines with a principal diagnosis related to SUD (using categories defined by the U.S. Substance Abuse and Mental Health Services Administration) to assess the overall impact of redactions on the APCD’s data completeness.
Among the key findings from our evaluation:
- There was wide variation in data availability across states. Variable reporting by health plans meant that some APCDs showed less than 20% availability of SUD data in certain markets (e.g., commercial, Medicaid, Medicare) while others had nearly complete availability in the same market – a variation that often reflected the practices of the largest insurers within each market.
- There was wide variation in redaction practices among health plans. Some plans, for example, excluded all SUD data except for nicotine-related diagnoses, while others reported only alcohol-related SUD data. Others, however, included nearly all data.
Strategies emerged for expanding data completeness and leveraging the use of SUD data.
Through this collaborative process with health plans and our clients, Onpoint’s analysts identified actionable steps that APCD administrators and researchers can take to help maximize the utility of SUD data, including:
- Engage directly with health plans. APCD administrators can engage in open dialogue with health plans to uncover opportunities to improve the completeness of the SUD data reported to an APCD. These discussions can help both parties better understand any concerns and collaborate on solutions that lead to more robust data sets.
- Leverage any SUD data available within the APCD. Even when the SUD data has been substantially redacted, valuable information may still be accessible in an APCD’s data sets. For instance, a health plan that excludes the most sensitive SUD categories might still provide comprehensive data on other categories such as nicotine- or alcohol-related treatment.
- Use external data sources to supplement what is missing. APCDs that offer integration services with external data sets – publicly available data, clinical data, or SUD program-specific data, for example – can help to fill critical gaps and enhance the overall impact of APCD-based research.
As more people are diagnosed with substance use disorder, the role of APCDs in supporting meaningful research will continue to grow. By understanding and addressing the limitations of SUD data within APCDs, researchers and policymakers can use the data to further inform and augment strategies that improve care and reduce harm.
Onpoint Health Data is committed to empowering our clients with the tools and insights that help them navigate these difficult data challenges. Together, we can turn data – even data with limitations – into actionable knowledge for better outcomes and the public good.
Contact us at information@onpointhealthdata.org to explore how we can help your organization address policy questions and data challenges in the health data landscape.