Dataset Information

Use of instrumental variable in prescription drug research with observational data: a systematic review.

ABSTRACT:

Objective

Instrumental variable (IV) analysis may offer a useful approach to the problem of unmeasured confounding in prescription drug research if the IV is: (1) strongly and unbiasedly associated to treatment assignment; and (2) uncorrelated with factors predicting the outcome (key assumptions).

Study design and methods

We conducted a systematic review of the use of IV methods in prescription drug research to identify the major types of IVs and the evidence for meeting IV assumptions. We searched MEDLINE, OVID, PsychoInfo, EconLit, and economic databases from 1961 to 2009.

Results

We identified 26 studies. Most (n=16) were published after 2007. We identified five types of IVs: regional variation (n=8), facility-prescribing patterns (n=5), physician preference (n=8), patient history/financial status (n=3), and calendar time (n=4). Evidence supporting the validity of IV was inconsistent. All studies addressed the first IV assumption; however, there was no standard for demonstrating that the IV sufficiently predicted treatment assignment. For the second assumption, 23 studies provided explicit argument that IV was uncorrelated with the outcome, and 16 supported argument with empirical evidence.

Conclusions

Use of IV methods is increasing in prescription drug research. However, we did not find evidence of a dominant IV. Future research should develop standards for reporting the validity and strength of IV according to key assumptions.

SUBMITTER: Chen Y

PROVIDER: S-EPMC3079803 | biostudies-literature | 2011 Jun

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Use of instrumental variable in prescription drug research with observational data: a systematic review.

Chen Yong Y Briesacher Becky A BA

Journal of clinical epidemiology 20101216 6

<h4>Objective</h4>Instrumental variable (IV) analysis may offer a useful approach to the problem of unmeasured confounding in prescription drug research if the IV is: (1) strongly and unbiasedly associated to treatment assignment; and (2) uncorrelated with factors predicting the outcome (key assumptions).<h4>Study design and methods</h4>We conducted a systematic review of the use of IV methods in prescription drug research to identify the major types of IVs and the evidence for meeting IV assump ...[more]

PMID: 21163621

Similar Datasets

Project description:BackgroundPrescription drug monitoring programs (PDMPs) are a key component of the president's Prescription Drug Abuse Prevention Plan to prevent opioid overdoses in the United States.PurposeTo examine whether PDMP implementation is associated with changes in nonfatal and fatal overdoses; identify features of programs differentially associated with those outcomes; and investigate any potential unintended consequences of the programs.Data sourcesEligible publications from MEDLINE, Current Contents Connect (Clarivate Analytics), Science Citation Index (Clarivate Analytics), Social Sciences Citation Index (Clarivate Analytics), and ProQuest Dissertations indexed through 27 December 2017 and additional studies from reference lists.Study selectionObservational studies (published in English) from U.S. states that examined an association between PDMP implementation and nonfatal or fatal overdoses.Data extraction2 investigators independently extracted data from and rated the risk of bias (ROB) of studies by using established criteria. Consensus determinations involving all investigators were used to grade strength of evidence for each intervention.Data synthesisOf 2661 records, 17 articles met the inclusion criteria. These articles examined PDMP implementation only (n = 8), program features only (n = 2), PDMP implementation and program features (n = 5), PDMP implementation with mandated provider review combined with pain clinic laws (n = 1), and PDMP robustness (n = 1). Evidence from 3 studies was insufficient to draw conclusions regarding an association between PDMP implementation and nonfatal overdoses. Low-strength evidence from 10 studies suggested a reduction in fatal overdoses with PDMP implementation. Program features associated with a decrease in overdose deaths included mandatory provider review, provider authorization to access PDMP data, frequency of reports, and monitoring of nonscheduled drugs. Three of 6 studies found an increase in heroin overdoses after PDMP implementation.LimitationFew studies, high ROB, and heterogeneous analytic methods and outcome measurement.ConclusionEvidence that PDMP implementation either increases or decreases nonfatal or fatal overdoses is largely insufficient, as is evidence regarding positive associations between specific administrative features and successful programs. Some evidence showed unintended consequences. Research is needed to identify a set of "best practices" and complementary initiatives to address these consequences.Primary funding sourceNational Institute on Drug Abuse and Bureau of Justice Assistance.

Dataset Information

Use of instrumental variable in prescription drug research with observational data: a systematic review.

Objective

Study design and methods

Results

Conclusions

Publications

Use of instrumental variable in prescription drug research with observational data: a systematic review.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets