Skip to main content

Table 2 Aim 1 healthcare system database characteristics and approaches

From: The IDeaS initiative: pilot study to assess the impact of rare diseases on patients and healthcare systems

HCS data source and analytics tools/approach

Number of patients

Geographic area and insurance representation

NCATS

Database queried using DEVEXI software, a commercially available HIPAA compliant web-based longitudinal health research platform that links de-identified, medical, and dental claims data

 ~ 9 million patients in entire data warehouse, for which 4.3 million were within 5-year time period (June 5, 2007 to June 6, 2012)* used to estimate disease percentage

State of Florida, ~ 64% live in metropolitan area of at least 1 million people

Public 100% (Florida Medicaid/Medicare)

Eversana

Patient cohorts identified using medical claims data from IBM®Marketscan® Research Database

Inclusive of 195 million patients, in time period 2006–2020, for which only patients first diagnosed with one of the 14 RD and having at least one medical or pharmaceutical claim in from Jan 1, 2013 to Dec 31, 2017 were used to estimate disease percentage.#

Entire U.S

Commercial 82%, Public 17%

Sanford

Custom query of Sanford Epic EHR and associated databases

Inclusive of 1.6 million patients within 5-year time period Jan 1, 2013 to Dec 31, 2017 were used to estimate disease percentage.@

North Dakota, South Dakota, Western Minnesota, Northwest Iowa, Northeast Montana

Commercial 42%, Public 58%

OHSU

Custom queries of OHSU Epic Clarity database and Research Data Warehouse

3.69 million in entire warehouse, ~ 1 million of whom had encounters within 5-year time period Jan 1, 2013 to Dec 31, 2017 and were used to estimate disease percentage

Oregon: 79.07%,

Washington: 11.41%

Unknown: 5.57%

All others: 3.95%

*Commercial 46.46%, Public 52.86%, Self-Pay 0.09%, Workers Comp 0.57% *Percentage of visits excluding visits with no listed payer

  1. NCATS National Center for Advancing Translational Sciences, OHSU Oregon Health and Science University, EHR electronic health records
  2. *Most recent 5-year time period data available (ICD-9 coding only because all data prior to 2015)
  3. #Prevalence was calculated first by identifying the number of patients first diagnosed with the disease during or prior to that year and having at least one medical or pharmaceutical claim during that year. The percentages of patients with the disease in each year was then calculated by dividing the corresponding number of patients in the database with at least one medical claim during that same year, and finally the percentage of patients with each of the diseases during the 5-year period of 2013–2017 was calculated as the average of the corresponding percentages over the 5 years
  4. @The query was further refined by performing a step-wise analysis using increasingly stringent parameters associated with the RD of interest, along with additional data integrity checks, when appropriate (such as ensuring that ICD9 and ICD10 codes matched before and after the transition in 2015), to arrive at the final patient count
Back to article page

Orphanet Journal of Rare Diseases

ISSN: 1750-1172