LEVERAGING ELECTRONIC MEDICAL RECORD-BASED ANALYTICS TO OPTIMIZE THE DESIGN OF AN OBSERVATIONAL PROSPECTIVE STUDY OF ASTHMA AND CHRONIC OBSTRUCTIVE PULMONARY DISEASE
Author(s)
Wang X1, Gerhardsson M2, Sveréus A2, Hiller J3, Wong D3, Nasuti P3, Jenner H3, Mckenney A3, Kamauu AW4, Flickinger KM5
1AstraZeneca, Gaithersburg, MD, USA, 2AstraZeneca, Gothenburg, Sweden, 3IMS Health, London, UK, 4Anolinx LLC, Salt Lake City, UT, USA, 5PAREXEL, Chester Springs, PA, USA
OBJECTIVES: AstraZeneca is developing a global 3-year observational longitudinal study (NOVELTY) on patients with a diagnosis of asthma and/or Chronic Obstructive Pulmonary Disease (COPD) to describe patient characteristics, treatment patterns and the burden of illness over time and to identify phenotypes/endotypes associated with differential outcomes. The objectives for this Electronic Medical Record (EMR) analysis were to understand patient characteristics within current clinical care, to optimize the study design, and to reduce burden on sites and patients. METHODS: We conducted an analysis of 11 country-specific EMR data sources (USA, Canada, United Kingdom, France, Germany, Italy, Spain, Sweden, Japan, China and Australia) to estimate the number of patients per country and by disease severity, and to validate routine clinical measurements capture. Patients were identified using diagnosis codes during 1 year inclusion period and in additional ±1 year of patient history from the index date. The disease severity was defined by an algorithm derived from clinical guidelines. Descriptive methods were applied to analyze the coverage and completeness of variables relevant to the study. Completeness was determined as % of patient cohort with ≥1 data entry in their EMR record for that variable within the patient history timeframe. RESULTS: A variation of asthma (1,400-58,000) and COPD (500-68,000) patients per country or per network in USA were identified and 5-15% of patients had both diagnoses. The distribution of severity levels varied in different countries and 20-55% were unclassified. Demographics, labs and comorbidities were well captured in contrast to other variables like spirometry, Patient Reported Outcome and Fractional Exhaled Nitric Oxide. CONCLUSIONS: This EMR-based analysis provides valuable insight of current clinical practices and patient profiles, promotes the patient center study design and enables novel data collection approach. The analysis suggests which variables are routinely collected and which variables will require special care and training to collect within the study.
Conference/Value in Health Info
2016-05, ISPOR 2016, Washington DC, USA
Value in Health, Vol. 19, No. 3 (May 2016)
Code
PRM57
Topic
Real World Data & Information Systems
Topic Subcategory
Reproducibility & Replicability
Disease
Respiratory-Related Disorders