Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2020 Dec 31;3(4):536-544.
doi: 10.1093/jamiaopen/ooaa048. eCollection 2020 Dec.

Framework for identifying drug repurposing candidates from observational healthcare data

Michal Ozery-Flato  1 , Yaara Goldschmidt  2 , Oded Shaham  2 , Sivan Ravid  1 , Chen Yanover  2
Affiliations

Framework for identifying drug repurposing candidates from observational healthcare data

Michal Ozery-Flato et al. JAMIA Open. .

Abstract

Objective: Observational medical databases, such as electronic health records and insurance claims, track the healthcare trajectory of millions of individuals. These databases provide real-world longitudinal information on large cohorts of patients and their medication prescription history. We present an easy-to-customize framework that systematically analyzes such databases to identify new indications for on-market prescription drugs.

Materials and methods: Our framework provides an interface for defining study design parameters and extracting patient cohorts, disease-related outcomes, and potential confounders in observational databases. It then applies causal inference methodology to emulate hundreds of randomized controlled trials (RCTs) for prescribed drugs, while adjusting for confounding and selection biases. After correcting for multiple testing, it outputs the estimated effects and their statistical significance in each database.

Results: We demonstrate the utility of the framework in a case study of Parkinson's disease (PD) and evaluate the effect of 259 drugs on various PD progression measures in two observational medical databases, covering more than 150 million patients. The results of these emulated trials reveal remarkable agreement between the two databases for the most promising candidates.

Discussion: Estimating drug effects from observational data is challenging due to data biases and noise. To tackle this challenge, we integrate causal inference methodology with domain knowledge and compare the estimated effects in two separate databases.

Conclusion: Our framework enables systematic search for drug repurposing candidates by emulating RCTs using observational data. The high level of agreement between separate databases strongly supports the identified effects.

Keywords: Parkinson’s disease; causal inference; comparative effectiveness research; drug repositioning; electronic health records.

PubMed Disclaimer

Figures

Figure 1.
Figure 1.
An illustration of the per-patient key dates in the study design of emulated randomized controlled trials. Each row corresponds to a certain type of medical event. Rectangles indicate diagnosis (Dx) events; ovals indicate prescription (Rx) events; event type is specified in the first (ie, leftmost) event in each row and then abbreviated (eg, “SD” in top row is the abbreviation for “Studied Drug”).
Figure 2.
Figure 2.
An overview of our framework’s emulation pipeline and the underlying components. The central module is the Randomized Controlled Trial (RCT) Emulator, which orchestrates the entire process. First, the Tested Drugs Extractor identifies a list of repurposing candidates, based on the user-provided Drug Criteria. For each such candidate, using the input Study Design parameters, the Treatment & Control Cohorts Extractor assigns patients to the respective cohorts. The Confounders & Outcomes Extractor computes a baseline and follow-up attributes for patients in both cohorts. The Drug Repurposing Engine then instantiates an RCT Emulator for each candidate, which estimates its effect on disease outcomes in the treatment versus control cohorts, adjusting for the extracted confounders and using methods implemented in the Causal Inference Library.
Figure 3.
Figure 3.
A comparison of estimated effects: balancing weights vs. outcome prediction. Each chart shows a different setting of the trial with respect to the outcome (dementia, fall, and psychosis; rows), control cohort (random and ATC-L2; left and right columns), and database (Explorys and MarketScan; alternate columns). Each point corresponds to a drug whose estimated effect was significant at FDR 5% by at least one of the two compared methods. The red line is the fitted least squares regression line; blue line indicates y = x.
Figure 4.
Figure 4.
Estimated causal effects: Explorys vs MarketScan. Each point corresponds to a drug estimated to have a significant effect in both databases. Marker type represents the combination of trial outcome and control cohort; points in the first and third quadrant indicate harmful and beneficial effects, respectively. Red line is the least squares regression line, blue line is y = x.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Langedijk J, Mantel-Teeuwisse AK, Slijkerman DS, et al.Drug repositioning and repurposing: terminology and definitions in literature. Drug Discov Today 2015; 20 (8): 1027–34. - PubMed
    1. Ashburn TT, Thor KB.. Drug repositioning: identifying and developing new uses for existing drugs. Nat Rev Drug Discov 2004; 3 (8): 673–83. - PubMed
    1. Pushpakom S, Iorio F, Eyers PA, et al.Drug repurposing: progress, challenges and recommendations. Nat Rev Drug Discov October 2018; 18 (1): 41–58. - PubMed
    1. Hurle MR, Yang L, Xie Q, et al.Computational drug repositioning: from data to therapeutics. Clin Pharmacol Ther 2013; 93 (4): 335–41. - PubMed
    1. Li J, Zheng S, Chen B, et al.A survey of current trends in computational drug repositioning. Brief Bioinform 2016; 17 (1): 2–12. - PMC - PubMed

LinkOut - more resources