The Oregon Health Care Quality Corporation (Q Corp) is an independent, nonprofit organization dedicated to improving the quality and affordability of health care in Oregon by leading community collaborations and producing unbiased information. Q Corp works with diverse stakeholders – including consumers, providers, employers, policymakers and health insurers – to improve the health of all Oregonians.
In partnership with its network of stakeholders, Q Corp has developed the most comprehensive administrative claims database in Oregon. Since 2008, Q Corp has been aggregating claims data from multiple payers to produce quality and utilization reports for consumers, providers, health plans, policymakers and employers. In 2013 eight commercial plans, two Medicaid managed care plans and Oregon Healthy Authority Divisions of Medical Assistance Programs (Medicaid fee-for-service) contributed administrative medical and pharmacy claims data comprising more than 400 million records. The Q Corp claims database includes information for over three million unique members since 2006. Aggregated data also included claims from select Medicare Advantage plans. Q Corp’s claims database in 2013 included over 75 percent of the state’s commercial insured population, 71 percent of the Medicaid population and over 38 percent of the Medicare population.
Q Corp and CLF staff discussed the possible health outcome and prevention information the Q Corp claims database could contribute to the Regional Equity Atlas 2.0. Based on the priorities that were identified through CLF’s stakeholder engagement process, they determined that chronic disease incidence rates and use of preventive care services would provide the most useful information. For chronic disease incidence rates, Q Corp’s database offered the strongest data for diabetes, asthma and cardiovascular disease. Q Corp expanded the HEDIS specifications for the asthma measure to result in more useful information when mapped geographically (see the data specifications for more information). For preventive care, Q Corp recommended using data on the prevalence of well-child visits and use of the emergency department for conditions that could have been treated in primary care.
In order to calculate incidence rates by geography, it is necessary to aggregate patient-level data by street address. Street address data are included in the files Q Corp receives from many of its data suppliers, but Q Corp had never used that information. Q Corp reached out to the data suppliers to gain permission for the use of this information to calculate incidence rates. One data supplier who does not provide street addresses to Q Corp but has access to those data in its internal records used the same data specifications to calculate the rates by geography and sent them to Q Corp to combine with the other rates.
Q Corp understands the critical importance of data security and maintains extensive policies and procedures to ensure that protocols are followed by all staff, data users and vendors. Q Corp and its data services vendor, Milliman, have data use agreements with each data supplier and follow HIPAA standards and protocols for data access and transfer.
In order to maintain confidentiality, data were aggregated by Milliman into census tracts and neighborhoods. For each indicator, Milliman presented the data as a rate, with the numerator representing the number of records that meet the criteria for the indicator in that geography (e.g. patients with diabetes, patients with well-child visits, etc.) and the denominator representing the total number of eligible records in that geography. The proportions of commercial, Medicare, and Medicaid records in both the numerator and denominator were also calculated, but were not mapped in the Equity Atlas 2.0.
If the number of records in a specific geographic area failed to meet a minimum sample size threshold, Milliman did not include that geographic area in its report. The sample size threshold that was used was based on the recommendations outlined by the CDC in their National Center for Health Statistics Staff Manual on Confidentiality (PDF). Based on this threshold, data from geographies where the numerator was less than five people, or the difference between the denominator and numerator was less than five people, were suppressed. As an example, if a geography had 50 patients in the denominator, then it was reportable so long as the numerator was between 5 and 45. Additionally a denominator threshold of 25 was applied to ensure robust reported rates.
Some patient records did not have complete patient street addresses. If the street address could not be accurately located within the aggregated geography, those patient records were not mapped.
While Q Corp’s database offers the most comprehensive patient-level data available in a mappable form for the Portland metro region, it also has limitations that are important to keep in mind when interpreting the Equity Atlas 2.0 maps. The maps only include data for patients who were continuously enrolled in a health plan that participates in Q Corp’s measurement and reporting initiative or Medicaid fee-for-service during the measurement year, with no more than one gap of up to 45 days. The maps do not include information on uninsured patients, patients who pay for their own health care services, Medicare fee-for- service patients, or patients served by a plan or Medicaid provider that does not supply data to Q Corp.
| Oregon Health Care Quality Corporation Data Specifications for Regional Equity Atlas 2.0 Data (PDF) |