Title |
How to compare instrumental variable and conventional regression analyses using negative controls and bias plots
|
---|---|
Published in |
International Journal of Epidemiology, April 2017
|
DOI | 10.1093/ije/dyx014 |
Pubmed ID | |
Authors |
Neil M. Davies, Kyla H. Thomas, Amy E. Taylor, Gemma M.J. Taylor, Richard M. Martin, Marcus R. Munafò, Frank Windmeijer |
Abstract |
There is increasing interest in the use of instrumental variable analysis to overcome unmeasured confounding in observational pharmacoepidemiological studies. This is partly because instrumental variable analyses are potentially less biased than conventional regression analyses. However, instrumental variable analyses are less precise, and regulators and clinicians find it difficult to interpret conflicting evidence from instrumental variable compared with conventional regression analyses. In this paper, we describe three techniques to assess which approach (instrumental variable versus conventional regression analyses) is least biased. These techniques are negative control outcomes, negative control populations and tests of covariate balance. We illustrate these methods using an analysis of the effects of smoking cessation therapies (varenicline) prescribed in primary care. |
X Demographics
Geographical breakdown
Country | Count | As % |
---|---|---|
United Kingdom | 3 | 21% |
United States | 2 | 14% |
China | 1 | 7% |
Spain | 1 | 7% |
Japan | 1 | 7% |
Sweden | 1 | 7% |
Chile | 1 | 7% |
Australia | 1 | 7% |
Unknown | 3 | 21% |
Demographic breakdown
Type | Count | As % |
---|---|---|
Members of the public | 7 | 50% |
Scientists | 3 | 21% |
Practitioners (doctors, other healthcare professionals) | 3 | 21% |
Science communicators (journalists, bloggers, editors) | 1 | 7% |
Mendeley readers
Geographical breakdown
Country | Count | As % |
---|---|---|
Unknown | 83 | 100% |
Demographic breakdown
Readers by professional status | Count | As % |
---|---|---|
Researcher | 20 | 24% |
Student > Ph. D. Student | 15 | 18% |
Other | 6 | 7% |
Student > Postgraduate | 4 | 5% |
Student > Bachelor | 4 | 5% |
Other | 16 | 19% |
Unknown | 18 | 22% |
Readers by discipline | Count | As % |
---|---|---|
Medicine and Dentistry | 28 | 34% |
Social Sciences | 5 | 6% |
Pharmacology, Toxicology and Pharmaceutical Science | 5 | 6% |
Biochemistry, Genetics and Molecular Biology | 4 | 5% |
Nursing and Health Professions | 3 | 4% |
Other | 14 | 17% |
Unknown | 24 | 29% |