Using Exploratory Visualization in the Analysis of Medical Product Safety in Observational Healthcare Data
Increasing the knowledge about the safety of medical products remains a top priority throughout the pharmaceutical development life cycle and particularly after regulatory approval when the products are used in real-world populations. The increasing availability and use of observational healthcare data, such as administrative claims and electronic health records, provides opportunity for generating better information to support therapeutic decision-making. Analysis of observational databases is quite different from randomized clinical trials, and offers unique opportunities for exploratory visualization to complement traditional epidemiologic investigations. The Observational Medical Outcomes Partnership was established to conduct methodological research on the appropriate use of observational data to identify and evaluate the effects of medical products in the real world. Several visualization tools were developed throughout the process, which demonstrate the value in combining standardized analytics with interactive graphics. These tools include a patient profile to study longitudinal patterns within clinical observations for a given person; a cohort profile to evaluate collections of patients; a treemap to assess disease prevalence across a database; a trellis scatterplot to investigate subgroup differences in drug utilization; a heatmap to support evaluation of high-dimensional confounding adjustment; and a multi-method forest plot to enable sensitivity analyses of estimated effects of drug and outcomes. This chapter illustrates these visualizations through a case study exploring the relationship between ACE inhibitor exposure and the subsequent health event angioedema.
- Bennett CL, Nebeker JR, Yarnold PR et al (2007) Evaluation of serious adverse drug reactions: a proactive pharmacovigilance program (RADAR) vs safety activities conducted by the Food and Drug Administration and pharmaceutical manufacturers. Arch Intern Med 167(10):1041–1049PubMedCrossRefGoogle Scholar
- Chobanian A (2009) The joint national committee on prevention detection and evaluation of high blood pressure. US Department of Health and Human Services NHLBI, Bethesda, MDGoogle Scholar
- Chou R, Helfand M, Carson S (1995) Drug class review on angiotensin converting enzyme inhibitors. Final report. Portland, OR: Oregon Health & Science UniversityGoogle Scholar
- DuMouchel W (1999) Bayesian data mining in large frequency tables, with an application to the FDA spontaneous reporting system. Am Statist 53(3):177–189Google Scholar
- Norén G, Bate A, Hopstadius J, Star K, Edwards I (2008) Temporal pattern discovery for trends and transient effects: its application to patient records. Paper presented at: Proceeding of the 14th ACM SIGKDD international conference on knowledge discovery and data mining, Las Vegas, Nevada, USAGoogle Scholar
- Norris S, Weinstein J, Peterson K, Thakurta S (2010) Drug class review: direct renin inhibitors, angiotensin converting enzyme inhibitors, and angiotensin II receptor blockers. Accessed on Oct 8, 2012, http://derp.ohsu.edu/about/fi nal-document-display.cfm Accessed Oct 8, 2012.
- Public Law 110–85 (2007) Food and Drug Administration Amendments Act of 2007Google Scholar
- Racoosin J (2009) FDA’s sentinel initiative—a national strategy for monitoring medical product safety. 2nd Drug Information Association (DIA) Conference on Signal Detection and Data Mining. New York, NYGoogle Scholar
- Rosenbaum P (2002) Observational studies. Springer, New YorkGoogle Scholar
- Schneeweiss S (2009) On guidelines for comparative effectiveness research using nonrandomized studies in secondary data sources. Value Health 10 Sep 2009Google Scholar
- Strom B (2005) Pharmacoepidemiology, 4th edn. Wiley, ChichesterGoogle Scholar
- US Department of Health and Human Services, Food and Drug Administration (1999) Managing the risks from medical product use: creating a risk management framework US Department of Health and Human Services, Food and Drug Administration, May 1999Google Scholar
- Zorych I, Madigan D, Ryan P, Bate A (2011) Disproportionality methods for pharmacovigilance in longitudinal observational databases. Stat Methods Med Res (in print)Google Scholar