Cited by 198Advanced Causal Inference TechniquesMeta-analysis and systematic reviewsStatistical Methods in Epidemiology

The DOI matches its Crossref record; 29 of the 30 references it cites were found in a registry or an index.

Commentary: How to Report Instrumental Variable Analyses (Suggestions Welcome)

Swanson, Sonja A., Hernán, Miguel A.·2013·Epidemiology

Open this paper in CoChat — free account

Abstract

Instrumental variable (IV) methods are becoming mainstream in comparative effectiveness research,1 but IV methods are radically different from traditional epidemiologic methods. The goal of IV methods is to eliminate confounding without ever measuring the confounders. This apparent miracle can be achieved only when four conditions are met (see below).2,3 Here we suggest a checklist for investigators who use IV methods. Like others before,4 we hope this step-by-step guide will improve the reporting of IV estimates and increase the transparency of the underlying assumptions. Our discussion focuses on reports of causal effects of medical interventions and is informed by two papers by Davies et al5,6 that appear in this issue of EPIDEMIOLOGY: an application that estimates the causal effects of cyclooxygenase-2 (COX-2) selective versus nonselective nonsteroidal anti-inflammatory drugs (NSAIDs), and a literature review of IV papers that describes their use, and perhaps misuse, in epidemiology. We supplement this review with additional information from our own review of IV analyses of observational studies with a relatively well-defined medical intervention. Details of our review can be found in the online supplement (https://links.lww.com/EDE/A664). IV methods require a variable—the “instrument”—that meets the three so-called instrumental conditions: (1) the instrument is associated with the treatment, (2) the instrument does not affect the outcome except through treatment (also known as the exclusion restriction assumption), and (3) the instrument does not share any causes with the outcome. An example of such an instrument is the randomization indicator in double-blind randomized experiments.2 Davies et al6 summarize instruments that have been proposed in epidemiologic studies, including the physician preference type they use themselves.5 Unfortunately, no variable can be proved to be an instrument in observational studies because only condition (1) can be empirically verified. We outline some steps for reporting IV analyses based on the variables proposed as instruments (Figure) and discuss how these steps have been reported in previous studies (Table). A detailed specification of how IV methods should be implemented is beyond the scope of this commentary.FIGURE: Flowchart for reporting of IV analyses with a single, non–time-varying, dichotomous instrument.TABLE: Proportions Reporting According to Suggested Guidelines in a Review of IV Publications Assessing Effects of Medical Interventions (n = 81)The first step in our reporting flowchart is to empirically verify condition (1). When the association between the proposed instrument and the treatment is weak, the proposed “weak” instrument may amplify biases due to small violations of conditions (2) or (3), producing very biased effect estimates.7,8 Alternatively, if the proposed instrument is very strong, it may be more likely to violate conditions (2) or (3); in the extreme, a perfect correlation between the proposed instrument and the observational treatment implies the proposed instrument is associated with the same set of (possibly unmeasured) confounders as the treatment.9 Most prior studies have evaluated the strength of their proposed instrument via risk differences, odds ratios, or partial R2 or F statistics like Davies et al5 did and several authors have recommended.6,8 (In contrast, the use of R2 or F statistics as effect measures is subject to many study-specific artifacts and should generally be avoided.10,11) Unlike condition (1), conditions (2) and (3) are not empirically verifiable: they are assumptions. Therefore, the next step is to use subject-matter knowledge to build a case for why the proposed instrument may be reasonably assumed to meet each of these two conditions; this is similar to how one would use subject-matter knowledge to justify the no unmeasured confounding assumption of traditional methods. When dichotomizing a polytomous or continuous treatment, condition (2) may hold for the original treatment, but not for its dichotomous version. Investigators can provide some support for condition (3) by showing the proposed instrument is not associated with measured confounders.6 Although any associations found with measured covariates could be adjusted for by modeling (see below), associations with measured confounders may raise concerns over associations with unmeasured confounders. Davies et al5 provide arguments to support conditions (2) and (3) when physician’s preference is proposed as an instrument; other authors have provided counterarguments.2 In our review, we notice a tendency for conditions (2) and (3) to be treated as a single condition. Although these conditions are statistically similar, it is important to consider them separately in order to incorporate subject-matter knowledge in discussions about their validity and in decisions on adjustment. For example, Davies et al5 found some evidence that the proposed instrument is ass

Integrity signals

Registry

Crossref · 10.1097/ede.0b013e31828d0590

Cites

29 of 29 references verified0 citation errors · 0 not checkable

Statistics

Recomputing p-values needs the paper’s full text.

Claims the paper makes

This paper’s full text isn’t openly available, so its claims and self-check weren’t run. The registry and citation checks still apply.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Identification of Causal Effects Using Instrumental Variables
Angrist, J. D., Imbens, G. W., & Rubin, D. B. (1996) · Journal of the American Statistical Association · doi:10.1080/01621459.1996.10476902 · Cited by 4,230
Identification and estimation of local average treatment effects
Imbens, G. W., & Angrist, J. D. (1994) · Econometrica · doi:10.2307/2951620 · Cited by 4,101
Problems with Instrumental Variables Estimation When the Correlation Between the Instruments and the Endogeneous Explanatory Variable is Weak
Bound, J., Jaeger, D. A., & Baker, R. (1995) · Journal of the American Statistical Association · doi:10.1080/01621459.1995.10476536 · Cited by 3,898
Causal Diagrams for Epidemiologic Research
Greenland, S., Pearl, J., & Robins, J. M. (1999) · Epidemiology · doi:10.1097/00001648-199901000-00008 · Cited by 3,728
Air Pollution and Hospital Admissions for Heart Disease in Eight U.S. Counties
Schwartz, J. D. (1999) · Epidemiology · doi:10.1097/00001648-199901000-00005 · Cited by 3,045
Identification and Estimation of Local Average Treatment Effects
Angrist, J. D., & Imbens, G. W. (1995) · National Bureau of Economic Research · doi:10.3386/t0118 · Cited by 2,915
Maternal Pesticide Exposure from Multiple Sources and Selected Congenital Anomalies
Shaw, G. M., Wasserman, C. R., O’Malley, C. D., Nelson, V. E., & Jackson, R. J. (1999) · Epidemiology · doi:10.1097/00001648-199901000-00011 · Cited by 1,783
Causal knowledge as a prerequisite for confounding evaluation: an application to birth defects epidemiology
Hernan, M. A. (2002) · American Journal of Epidemiology · doi:10.1093/aje/155.2.176 · Cited by 1,445
Quantifying Biases in Causal Models: Classical Confounding vs Collider-Stratification Bias
Greenland, S. (2003) · Epidemiology · doi:10.1097/01.ede.0000042804.12056.6c · Cited by 810
Bounds on Treatment Effects from Studies with Imperfect Compliance
Balke, A., & Pearl, J. (1997) · Journal of the American Statistical Association · doi:10.1080/01621459.1997.10474074 · Cited by 564

Show all 30 references

Research works printed but not linked: 3

Bound, 1995, J Am Statist Assoc, 90, 443
Robins, 1996, J Am Statist Assoc, 91, 456
Greenland, 2003, Epidemiology, 14, 300

Sorted by citations

Interpreting findings from Mendelian randomization using the MR-Egger method.
Burgess, S., & Thompson, S. G. (2017) · European Journal of Epidemiology · doi:10.1007/s10654-017-0255-x · Cited by 4,571
Reading Mendelian randomisation studies: a guide, glossary, and checklist for clinicians
Davies, N. M., Holmes, M. V., & Davey Smith, G. (2018) · BMJ · doi:10.1136/bmj.k601 · Cited by 4,404
Strengthening the Reporting of Observational Studies in Epidemiology Using Mendelian Randomization: The STROBE-MR Statement
Skrivankova, V. W., Richmond, R. C., Woolf, B. A. R., Yarmolinsky, J., Davies, N. M., Swanson, S. A., VanderWeele, T. J., Higgins, J. P. T., Timpson, N. J., Dimou, N., et al. (2021) · JAMA · doi:10.1001/jama.2021.18236 · Cited by 3,994
Self-citation
Mendelian Randomization as an Approach to Assess Causality Using Observational Data
Sekula, P., Del Greco M, F., Pattaro, C., & Köttgen, A. (2016) · Journal of the American Society of Nephrology · doi:10.1681/asn.2016010098 · Cited by 2,351
Sensitivity Analyses for Robust Causal Inference from Mendelian Randomization Analyses with Multiple Genetic Variants
Burgess, S., Bowden, J., Fall, T., Ingelsson, E., & Thompson, S. G. (2017) · Epidemiology · doi:10.1097/ede.0000000000000559 · Cited by 2,017
A review of instrumental variable estimators for Mendelian randomization
Burgess, S., Small, D. S., & Thompson, S. G. (2015) · Statistical Methods in Medical Research · doi:10.1177/0962280215597579 · Cited by 1,769
“If the monotonicity assumption is not satisfied, then the causal effect of the exposure on the outcome can be estimated consistently if it is constant for all individuals across the population.(29) A weaker version of this assumption, that of no modification of the causal effect across levels of the instrument within…”
Background
Combining information on multiple instrumental variables in Mendelian randomization: comparison of allele score and summarized data methods
Burgess, S., Dudbridge, F., & Thompson, S. G. (2016) · Statistics in Medicine · doi:10.1002/sim.6835 · Cited by 1,153
“These strong assumptions are not necessary for the estimation of a causal effect; alternative assumptions, such as monotonicity of the IV–risk factor association or no additive effect modification of the causal effect across levels of the instrument at different values of the risk factor, are able to identify a causal…”
Background
The reporting of studies conducted using observational routinely collected health data statement for pharmacoepidemiology (RECORD-PE)
Langan, S., Schmidt, S. A. J., Wing, K., Ehrenstein, V., Nicholls, S. G., Filion, K. B., Klungel, O. H., Petersen, I., Sørensen, H. T., Dixon, W. G., et al. (2018) · BMJ · doi:10.1136/bmj.k3532 · Cited by 673
Mendelian randomization studies: a review of the approaches used and the quality of reporting
Boef, A. G. C., Dekkers, O. M., & le Cessie, S. (2015) · International Journal of Epidemiology · doi:10.1093/ije/dyv071 · Cited by 610
“Reporting guidelines for IV analyses recommend the use of the partial F-statistic as a measure of the strength of the association between the IV and the exposure.(7,8) It encompasses information on the strength of the instrument and on the number of observations in the analysis.”
Method
Natural Experiments: An Overview of Methods, Approaches, and Contributions to Public Health Intervention Research
Craig, P., Katikireddi, S. V., Leyland, A., & Popham, F. (2017) · Annual Review of Public Health · doi:10.1146/annurev-publhealth-031816-044327 · Cited by 570
“IV methods have not yet been widely used to evaluate public health interventions because it can be difficult to find suitable instruments and to demonstrate convincingly, using theory or data, that they meet the second and third conditions above (35, 71).”
Method
Actionable druggable genome-wide Mendelian randomization identifies repurposing opportunities for COVID-19
Gaziano, L., Giambartolomei, C., Pereira, A. C., Gaulton, A., Posner, D. C., Swanson, S. A., Ho, Y.-L. A., Iyengar, S. K., Kosik, N., Vujković, M., et al. (2021) · Nature Medicine · doi:10.1038/s41591-021-01310-z · Cited by 231
Self-citation
A systematic review of the cost and cost effectiveness of using standard oral nutritional supplements in the hospital setting
Elia, M., Normand, C. E. M., Norman, K., & Laviano, A. (2015) · Clinical Nutrition · doi:10.1016/j.clnu.2015.05.010 · Cited by 218
Nature as a Trialist?: Deconstructing the Analogy Between Mendelian Randomization and Randomized Trials
Swanson, S. A., Tiemeier, H., Ikram, M. A., & Hernán, M. A. (2017) · Epidemiology · doi:10.1097/ede.0000000000000699 · Cited by 202
Self-citation
Understanding the Assumptions Underlying Instrumental Variable Analyses: a Brief Review of Falsification Strategies and Related Tools
Labrecque, J., & Swanson, S. A. (2018) · Current Epidemiology Reports · doi:10.1007/s40471-018-0152-1 · Cited by 190
Self-citation
Adjusted Analyses in Studies Addressing Therapy and Harm
Agoritsas, T., Merglen, A., Shah, N. D., O'Donnell, M. J., & Guyatt, G. H. (2017) · JAMA · doi:10.1001/jama.2016.20029 · Cited by 186
An introduction to instrumental variable assumptions, validation and estimation
Lousdal, M. L. (2018) · Emerging Themes in Epidemiology · doi:10.1186/s12982-018-0069-7 · Cited by 173
“Moreover, they show how much “information” that needs to be provided by a fourth assumption in order to obtain a point estimate [24].”
Background
Potential Bias of Instrumental Variable Analyses for Observational Comparative Effectiveness Research
Garabedian, L. F., Chu, P., Toh, S., Zaslavsky, A. M., & Soumerai, S. B. (2014) · Annals of Internal Medicine · doi:10.7326/m13-1887 · Cited by 133
Empirical anti-MRSA vs standard antibiotic therapy and risk of 30-day mortality in patients hospitalized for pneumonia.
Jones, B. E., Ying, J., Stevens, V., Haroldsen, C., He, T., Nevers, M., Christensen, M. A., Nelson, R. E., Stoddard, G. J., Sauer, B. C., et al. (2020) · JAMA Internal Medicine · doi:10.1001/jamainternmed.2019.7495 · Cited by 122
Instrumental variable approaches to identifying the causal effect of educational attainment on dementia risk
Nguyen, T. T. X., Tchetgen, E. J. T., Kawachi, I. I., Gilman, S. E., Walter, S., Liu, S. Y., Manly, J. J., & Glymour, M. M. (2015) · Annals of Epidemiology · doi:10.1016/j.annepidem.2015.10.006 · Cited by 116
Relaxing parametric assumptions for non-linear Mendelian randomization using a doubly-ranked stratification method
Tian, H., Mason, A. M., Liu, C., & Burgess, S. (2023) · PLoS Genetics · doi:10.1371/journal.pgen.1010823 · Cited by 104
“There are various versions of the homogeneity assumption; the simplest version of the assumption states that the effect of the instrument on the exposure is constant for all individuals in the population [19, 20].”
Background
Commentary: Mendelian randomization-inspired causal inference in the absence of genetic data
Zuccolo, L., & Holmes, M. V. (2016) · International Journal of Epidemiology · doi:10.1093/ije/dyw327 · Cited by 102
“Presenting such data is strongly recommended as optimal reporting of IV and MR analyses(12, 13), as is also routinely done in the reporting of RCTs (the conventional ‘Table 1’ to demonstrate that confounders are equally balanced across intervention and comparator arms), as a basic check of the assumption that an instr…”
Method
Obstetric Facility Quality and Newborn Mortality in Malawi: A Cross-Sectional Study
Leslie, H. H., Fink, G., Nsona, H., & Kruk, M. E. (2016) · PLoS Medicine · doi:10.1371/journal.pmed.1002151 · Cited by 97
Selection Bias When Estimating Average Treatment Effects Using One-sample Instrumental Variable Analysis
Hughes, R. A., Davies, N. M., Davey Smith, G., & Tilling, K. (2019) · Epidemiology · doi:10.1097/ede.0000000000000972 · Cited by 89
Falsification Testing of Instrumental Variables Methods for Comparative Effectiveness Research
Pizer, S. D. (2015) · Health Services Research · doi:10.1111/1475-6773.12355 · Cited by 93
MR-Clust: clustering of genetic variants in Mendelian randomization with similar causal estimates
Foley, C. N., Mason, A. M., Kirk, P., & Burgess, S. (2020) · Bioinformatics · doi:10.1093/bioinformatics/btaa778 · Cited by 81
“We note that these parametric assumptions are sufficient, but not necessary for the estimation of the average causal effect; weaker assumptions have been proposed (Swanson and Hernán, 2013).”
Background
Confidence Intervals for Causal Effects with Invalid Instruments by Using Two-Stage Hard Thresholding with Voting
Guo, Z., Kang, H., Cai, T., & Small, D. S. (2018) · Journal of the Royal Statistical Society Series B (Statistical Methodology) · doi:10.1111/rssb.12275 · Cited by 87
“The work by Windmeijer et al. (2016) is the methodologically most similar to our method in that it also estimates V∗ and uses the estimate of V∗ to obtain oracle-optimal point estimate and confidence interval of β∗ like we do in Section 3.5. To estimate V∗, Windmeijer et al. (2016) utilizes the adaptive Lasso with a m…”
Influential
Toward a Clearer Portrayal of Confounding Bias in Instrumental Variable Applications
Jackson, J. W., & Swanson, S. A. (2015) · Epidemiology · doi:10.1097/ede.0000000000000287 · Cited by 74
“thus, the bias resulting from confounding of the instrument is the product of (1) the difference in mean outcome across levels of U conditional on X; (2) the prevalence difference in covariate U across the instrument Z, divided by (3) the prevalence difference of treatment X across instrument Z.”
BackgroundMethodInfluentialSelf-citation
The challenging interpretation of instrumental variable estimates under monotonicity
Swanson, S. A., & Hernán, M. A. (2018) · International Journal of Epidemiology · doi:10.1093/ije/dyx038 · Cited by 71
Self-citation
Recommendations for Statistical Reporting in Cardiovascular Medicine: A Special Report From the American Heart Association
Althouse, A. D., Below, J. E., Claggett, B. L., Cox, N. J., de Lemos, J. A., Deo, R. C., Duval, S., Hachamovitch, R., Kaul, S., Keith, S. W., et al. (2021) · Circulation · doi:10.1161/circulationaha.121.055393 · Cited by 74
Selecting on Treatment: A Pervasive Form of Bias in Instrumental Variable Analyses
Swanson, S. A., Robins, J. M., Miller, M., & Hernán, M. Á. (2015) · American Journal of Epidemiology · doi:10.1093/aje/kwu284 · Cited by 69
Self-citation
Global Energy Transitions and Political Systems
Lee, J., & Yang, J. (2019) · Renewable and Sustainable Energy Reviews · doi:10.1016/j.rser.2019.109370 · Cited by 69
Sample size importantly limits the usefulness of instrumental variable methods, depending on instrument strength and level of confounding
Boef, A. G. C., Dekkers, O. M., Vandenbroucke, J. P., & le Cessie, S. (2014) · Journal of Clinical Epidemiology · doi:10.1016/j.jclinepi.2014.05.019 · Cited by 66
Instrumental Variable Analyses and Selection Bias
Canan, C. E., Lesko, C. R., & Lau, B. M. (2017) · Epidemiology · doi:10.1097/ede.0000000000000639 · Cited by 65
Adverse childhood experiences and later life adult obesity and smoking in the United States
Rehkopf, D. H., Headen, I. E., Hubbard, A., Deardorff, J., Kesavan, Y., Cohen, A. K., Patil, D., Ritchie, L. D., & Abrams, B. F. (2016) · Annals of Epidemiology · doi:10.1016/j.annepidem.2016.06.003 · Cited by 64
“The instrumental variable approach has the potential to address two fundamental limitations of the traditional regression approach: unmeasured confounding and measurement error of the exposure variable, but has four fundamental characteristics that must be met for valid inference [31].”
Background
Assessing causal relationships using genetic proxies for exposures: an introduction to Mendelian randomization
Katikireddi, S. V., Green, M. J., Taylor, A. E., Smith, G. D., & Munafo, M. R. (2017) · Addiction · doi:10.1111/add.14038 · Cited by 63
A Gentle Introduction to Instrumental Variables
Widding‐Havneraas, T., & Zachrisson, H. D. (2022) · Journal of Clinical Epidemiology · doi:10.1016/j.jclinepi.2022.06.022 · Cited by 62
Robust instrumental variable methods using multiple candidate instruments with application to Mendelian randomizationPreprint
Burgess, S., Bowden, J., Dudbridge, F., & Thompson, S. G. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1606.03729 · Cited by 77
“These assumptions are not necessary to identify a causal effect; weaker assumptions can be made [46] (such as monotonicity of the causal effect [47] or a weaker version of the homogeneity assumption for the causal effect [48, 49]).”
Background
Genetically predicted body mass index and Alzheimer's disease–related phenotypes in three large samples: Mendelian randomization analyses
Mukherjee, S., Walter, S., Kauwe, J. S. K., Saykin, A. J., Bennett, D. A., Larson, E. B., Crane, P. K., Glymour, M. M., Adult Changes in Thought Study Investigators, & Religious Orders Study/Memory and Aging Project Investigators (2015) · Alzheimer s & Dementia · doi:10.1016/j.jalz.2015.05.015 · Cited by 58
Investigating the transparency of reporting in two-sample summary data Mendelian randomization studies using the MR-Base platform
Woolf, B. A. R., Cara, N. H. D., Moreno-Stokoe, C., Skrivankova, V. W., Drax, K., Higgins, J. P. T., Hemani, G., Munafo, M. R., Smith, G. D., Yarmolinsky, J., et al. (2022) · International Journal of Epidemiology · doi:10.1093/ije/dyac074 · Cited by 51
Avoiding collider bias in Mendelian randomization when performing stratified analyses
Coscia, C., Gill, D., Benítez, R., Pérez, T., Malats, N., & Burgess, S. (2022) · European Journal of Epidemiology · doi:10.1007/s10654-022-00879-0 · Cited by 52
“1 3 the IV estimate can be interpreted as a local average causal effect; under homogeneity, it can be interpreted as an average causal effect [5].”
Background
Impact of Molecular Diagnostics for Tuberculosis on Patient-Important Outcomes: A Systematic Review of Study Methodologies
Schumacher, S. G., Sohn, H., Qin, Z. Z., Gore, G. C., Davis, J. L., Denkinger, C. M., & Pai, M. (2016) · PLoS ONE · doi:10.1371/journal.pone.0151073 · Cited by 51
Integrating summarized data from multiple genetic variants in Mendelian randomization: bias and coverage properties of inverse-variance weighted methodsPreprint
Burgess, S., & Bowden, J. (2015) · arXiv (Cornell University) · doi:10.48550/arxiv.1512.04486 · Cited by 59
“Although these assumptions are not necessary to identify a causal parameter (weaker assumptions have been proposed (Swanson and Hernán, 2013)), alternative assumptions mean that the causal parameters identified by different instrumental variables are likely to be different.”
Background
Definition and Evaluation of the Monotonicity Condition for Preference-based Instruments
Swanson, S. A., Miller, M., Robins, J. M., & Hernán, M. Á. (2015) · Epidemiology · doi:10.1097/ede.0000000000000279 · Cited by 50
“Moreover, such a conclusion rests on several strong assumptions, eg,: (1) that our patients could be seen with equal probability by all survey respondents, (2) that preference is dichotomous and measured without error, (3) that the prescribing decisions of the physicians who responded to the survey adequately represen…”
MethodResultSelf-citation
Think Globally, Act Globally: An Epidemiologist’s Perspective on Instrumental Variable EstimationPreprint
Swanson, S. A., Hernán, M. Á., & Hernán, M. Á. (2016) · Cited by 51
Self-citation
Epidemiology, genetic epidemiology and Mendelian randomisation: more need than ever to attend to detail
Sheehan, N. A., & Didelez, V. (2019) · Human Genetics · doi:10.1007/s00439-019-02027-3 · Cited by 47
“Generally poor reporting of MR studies has been commented on elsewhere in the literature and several suggestions have been made for improvement (Davies et al. 2015; Glymour et al. 2012; Swanson and Hernán 2013; VanderWeele et al. 2014; von Hinke et al. 2016).”
BackgroundResultInfluential
Dissonant Conclusions When Testing the Validity of an Instrumental Variable
Yang, F., Zubizarreta, J. R., Small, D. S., Lorch, S. A., & Rosenbaum, P. R. (2014) · The American Statistician · doi:10.1080/00031305.2014.962764 · Cited by 45
“For several recent discussions of instrumental variables in health outcomes, see Baiocchi et al. (2010), Brookhart and Schneeweiss (2007), Cheng, Qin, and Zhang (2011), Swanson and Hernán (2013), and Tan (2006).”
Background
Mendelian randomization studies in the elderly
Boef, A. G. C., le Cessie, S., & Dekkers, O. M. (2015) · Epidemiology · doi:10.1097/ede.0000000000000243 · Cited by 39
Exploring the causal effect of maternal pregnancy adiposity on offspring adiposity: Mendelian randomisation using polygenic risk scores
Bond, T. A., Richmond, R. C., Karhunen, V., Cuéllar-Partida, G., Borges, M. C., Zuber, V., Alves, A. C., Mason, D., Yang, T., Gunter, M. J., et al. (2022) · BMC Medicine · doi:10.1186/s12916-021-02216-w · Cited by 42
Cohort Multiple Randomised Controlled Trials (cmRCT) design: efficient but biased? A simulation study to evaluate the feasibility of the Cluster cmRCT design
2, O. B. O. G. W. P., Pate, A., Candlish, J., Sperrin, M., & van Staa, T. P. (2016) · BMC Medical Research Methodology · doi:10.1186/s12874-016-0208-1 · Cited by 41
Preference-based instrumental variables in health research rely on important and underreported assumptions: a systematic review
Widding‐Havneraas, T., Chaulagain, A., Lyhmann, I., Zachrisson, H. D., Elwert, F., Markussen, S., McDaid, D., & Mykletun, A. (2021) · Journal of Clinical Epidemiology · doi:10.1016/j.jclinepi.2021.06.006 · Cited by 40
Causal null hypotheses of sustained treatment strategies: What can be tested with an instrumental variable?
Swanson, S. A., Labrecque, J. A., & Hernán, M. Á. (2018) · European Journal of Epidemiology · doi:10.1007/s10654-018-0396-6 · Cited by 41
“This is perhaps due to investigators recognizing the additional assumptions needed for these effect estimates to be valid are biologically implausible, let alone the additional issues with interpretation of these effect estimates when the treatment strategies are sustained over time [2, 7, 8, 11].”
BackgroundSelf-citation
Effects of implementing universal and rapid HIV treatment on initiation of antiretroviral therapy and retention in care in Zambia: a natural experiment using regression discontinuity
Mody, A., Sikazwe, I. T., Namwase, A. S., Mwanza, M. W., Savory, T., Mwila, A. C., Mulenga, L., Herce, M. E., Mweebo, K., Somwe, P., et al. (2021) · The Lancet HIV · doi:10.1016/s2352-3018(21)00186-7 · Cited by 37
Think Globally, Act Globally: An Epidemiologist’s Perspective on Instrumental Variable Estimation
Swanson, S. A., & Hernán, M. Á. (2014) · Statistical Science · doi:10.1214/14-sts491 · Cited by 37
“Consider one of the most commonly proposed instruments in epidemiologic research, physician or facility prescribing preference (Swanson and Hernán, 2013).”
BackgroundInfluentialSelf-citation
Methodological considerations in assessing the effectiveness of antidepressant medication continuation during pregnancy using administrative data
Swanson, S. A., Hernández–Dı́az, S., Palmsten, K. K., Mogun, H., Olfson, M., & Huybrechts, K. F. (2015) · Pharmacoepidemiology and Drug Safety · doi:10.1002/pds.3798 · Cited by 37
Self-citation
Comparative effectiveness of buprenorphine-naloxone versus methadone for treatment of opioid use disorder: a population-based observational study protocol in British Columbia, Canada
Piske, M., Thomson, T. J., Krebs, E., Hongdilokkul, N., Bruneau, J., Greenland, S., Gustafson, P., Karim, M. E., McCandless, L. C., Maclure, M., et al. (2020) · BMJ Open · doi:10.1136/bmjopen-2019-036102 · Cited by 36
“74 We will also consider alternative empirical approaches for assessing conditions 2 and 3, consistent with recommendations of Glymour et al. 75 The use of IVs is controversial, in part because conditions (2) and (3) listed above are not explicitly testable for unmeasured confounders.”
BackgroundMethodInfluential
Full matching approach to instrumental variables estimation with application to the effect of malaria on stunting
Kang, H., Kreuels, B., May, J., & Small, D. S. (2016) · The Annals of Applied Statistics · doi:10.1214/15-aoas894 · Cited by 33
“…of an exposure on the outcome when there is unmeasured confounding, provided that a valid instrument is available [Angrist, Imbens and Rubin (1996), Baiocchi, Cheng and Small (2014), Brookhart and Schneeweiss (2007), Cheng, Qin and Zhang (2009), Hernán and Robins (2006), Swanson and Hernán (2013)].”
BackgroundMethod
Cross-sectional studies: strengths, limitations, and methodological considerations
Baumeister, S., Kocher, T., Papapanou, P. N., Holtfreter, B., & Demmer, R. T. (2026) · Journal of Periodontal Research · doi:10.1111/jre.70063 · Cited by 38
Does education protect against depression? Evidence from the Young Finns Study using Mendelian randomization
Viinikainen, J., Bryson, A., Böckerman, P., Elovainio, M. J., Pitkänen, N., Pulkki-Råbäck, L., Lehtimäki, T. J., Raitakari, O. T., & Pehkonen, J. (2018) · Preventive Medicine · doi:10.1016/j.ypmed.2018.08.026 · Cited by 34
“Fourth, the local average treatment effect (LATE) identified by IV estimation captures the average effect of education on depressive symptoms among compliers (Swanson and Hernán, 2013) i.”
Background
Estimating the real-world effects of expanding antiretroviral treatment eligibility: Evidence from a regression discontinuity analysis in Zambia
Mody, A., Sikazwe, I. T., Czaicki, N. L., Mwanza, M. W., Savory, T., Sikombe, K., Beres, L. K., Somwe, P., Roy, M., Pry, J. M., et al. (2018) · PLoS Medicine · doi:10.1371/journal.pmed.1002574 · Cited by 30
“Under the aforementioned conditions, this method estimates the causal effect of ART initiation on retention in care in those patients who initiated ART in response to the change in guidelines—commonly referred to as the local average treatment effect (LATE) [25,32,33].”
BackgroundMethod
The Causal Effect of Tracing by Peer Health Workers on Return to Clinic Among Patients Who Were Lost to Follow-up From Antiretroviral Therapy in Eastern Africa: A “Natural Experiment” Arising From Surveillance of Lost Patients
Bershteyn, A., Odeny, T., Lyamuya, R., Nakiwogga-Muwanga, A., Diero, L. O., Bwana, M., Braitstein, P. K. A., Somi, G. R., Kambugu, A. D., Bukusi, E. A., et al. (2017) · Clinical Infectious Diseases · doi:10.1093/cid/cix191 · Cited by 29
Are All Biases Missing Data Problems?
Howe, C. J., Cain, L. E., & Hogan, J. W. (2015) · Current Epidemiology Reports · doi:10.1007/s40471-015-0050-8 · Cited by 28
Reporting and guidelines for mendelian randomization analysis: A systematic review of oncological studies
Lor, G. C. Y., Risch, H. A., Fung, W. T., Au Yeung, S. L., Wong, I. O. L., Zheng, W., & Pang, H. (2019) · Cancer Epidemiology · doi:10.1016/j.canep.2019.101577 · Cited by 27
Estimating the Effect of School Water, Sanitation, and Hygiene Improvements on Pupil Health Outcomes
Garn, J. V., Brumback, B., Drews‐Botsch, C., Lash, T. L., Kramer, M. R., & Freeman, M. C. (2016) · Epidemiology · doi:10.1097/ede.0000000000000522 · Cited by 26
“% with diarrheab (SD) 9% (28) 4% (21) 8% (27) 4% (19) 9% (29)”
Instrumental Variable Analyses in Pharmacoepidemiology: What Target Trials Do We Emulate?
Swanson, S. A. (2017) · Current Epidemiology Reports · doi:10.1007/s40471-017-0120-1 · Cited by 25
Self-citation
Introduction to Mendelian randomization
Yeung, S. L. A., Luo, S., Iwagami, M., & Goto, A. (2024) · Annals of Clinical Epidemiology · doi:10.37737/ace.25004 · Cited by 26
An Instrumental Variables Design for the Effect of Emergency General Surgery
Keele, L. J., Sharoky, C. E., Sellers, M. M., Wirtalla, C. J., & Kelz, R. R. (2018) · Epidemiologic Methods · doi:10.1515/em-2017-0012 · Cited by 22
“Following the recommendation in Swanson and Hernán (2013), we report nonparametric bounds for the IV estimate (Balke and Pearl 1997).”
Method
Transparency in real‐world evidence (RWE) studies to build confidence for decision‐making: Reporting RWE research in diabetes
Patorno, E., Schneeweiß, S., & Wang, S. V. (2020) · Diabetes Obesity and Metabolism · doi:10.1111/dom.13918 · Cited by 22
“39,40 For all RWE studies conducted using RWD, the statistical software programs, packages or platforms used in cohort extraction and analysis should be reported (Table 2D).”
Estimation of time-varying causal effects with multivariable Mendelian randomization: some cautionary notes
Tian, H., & Burgess, S. (2022) · International Journal of Epidemiology · doi:10.1093/ije/dyac240 · Cited by 21
Mendelian randomisation approaches to the study of prenatal exposures: A systematic review
Diemer, E. W., Labrecque, J. A., Neumann, A., Tiemeier, H., & Swanson, S. A. (2021) · Paediatric and Perinatal Epidemiology · doi:10.1111/ppe.12691 · Cited by 21
“As guidelines for MR suggest that the key conditions need to be assessed on a case-by-case basis relative to the study design and research question, 12,17-20 and prenatal MR studies present several unique challenges relative to other types of MR designs, 21,22 it is important to understand how…”
Self-citation
Effect of Pharmacological Treatment of Attention-Deficit/Hyperactivity Disorder on Criminality
Widding‐Havneraas, T., Zachrisson, H. D., Markussen, S., Elwert, F., Lyhmann, I., Chaulagain, A., Bjelland, I., Halmøy, A., Rypdal, K., & Mykletun, A. (2023) · Journal of the American Academy of Child & Adolescent Psychiatry · doi:10.1016/j.jaac.2023.05.025 · Cited by 24
Is smoking heaviness causally associated with alcohol use? A Mendelian randomization study in four European cohorts
Taylor, M. R., Rode, L., Bjø‌rngaard, J. H., Taylor, A. E., Bojesen, S. E., Åsvold, B. O., Gabrielsen, M. E., Lewis, G., Nordestgaard, B. G., Romundstad, P., et al. (2018) · International Journal of Epidemiology · doi:10.1093/ije/dyy027 · Cited by 22
“The Instrumental Variable Checklist(51) (Supplementary Table 3, available as Supplementary Data at IJE online) was used for MR analysis and the instrumental variable flow chart considered throughout analyses and reporting.(52)”
The price of admission: does moving to a low-poverty neighborhood increase discriminatory experiences and influence mental health?
Osypuk, T. L., Schmidt, N. M., Kehm, R. D., Tchetgen, E. J. T., & Glymour, M. M. (2018) · Social Psychiatry and Psychiatric Epidemiology · doi:10.1007/s00127-018-1592-0 · Cited by 21
Evaluating a Key Instrumental Variable Assumption Using Randomization Tests
Branson, Z., & Keele, L. J. (2020) · American Journal of Epidemiology · doi:10.1093/aje/kwaa089 · Cited by 19
“Alternatively, bias in covariates from an IV can be compared to bias from an intention-to-treat analysis.4 Jackson and Swanson13 outlined a method of graphical display for easy bias comparisons, and Davies et al.14,15 suggest using formal statistical tests to compare the bias associated with the IV…”
Method
Comparative Effectiveness of Oxaliplatin Versus 5-flourouricil in Older Adults
Mack, C. D., Brookhart, M. A., Glynn, R. J., Meyer, A. M., Carpenter, W. R., Sandler, R. S., & Stürmer‎, T. (2015) · Epidemiology · doi:10.1097/ede.0000000000000355 · Cited by 20
Application of the Instrumental Inequalities to a Mendelian Randomization Study With Multiple Proposed Instruments
Diemer, E. W., Labrecque, J., Tiemeier, H., & Swanson, S. A. (2020) · Epidemiology · doi:10.1097/ede.0000000000001126 · Cited by 19
Self-citation
Guidelines for estimating causal effects in pragmatic randomized trialsPreprint
Murray, E. J., Swanson, S. A., & Hernán, M. Á. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1911.06030 · Cited by 27
Self-citation
Lack of Identification in Semiparametric Instrumental Variable Models With Binary Outcomes
Burgess, S., Granell, R., Palmer, T., Sterne, J. A. C., & Didelez, V. (2014) · American Journal of Epidemiology · doi:10.1093/aje/kwu107 · Cited by 18
“If investigators discover lack of identification in an applied context, they should consider not providing a point estimate for the causal effect and simply reporting the test for association between the IV and the outcome to establish a causal effect of the exposure on the outcome (which should be reported in any cas…”
Background
Reporting methodological issues of the mendelian randomization studies in health and medical research: a systematic review
Islam, S. N., Ahammed, T., Anjum, A., Albalawi, O., & Uddin, M. J. (2022) · BMC Medical Research Methodology · doi:10.1186/s12874-022-01504-0 · Cited by 15
“For example, the homogeneity assumption (4 h) for estimating the Average Treatment Effect (ATE) and the monotonicity assumption (4 m) for estimating the Local Average Treatment Effect (LATE) is needed to be theoretically justifiable [5, 18, 31].”
BackgroundMethodInfluential
Quantitative Falsification of Instrumental Variables Assumption Using Balance Measures
Ali, M. S., Uddin, M. J., Groenwold, R. H. H., Pestman, W. R., Belitser, S. V., Hoes, A. W., de Boer, A., Roes, K. C. B., & Klungel, O. H. (2014) · Epidemiology · doi:10.1097/ede.0000000000000152 · Cited by 16
The Impacts of Tourism Stays on Residents’ Self-Reported Health: A Pan-European Analysis on the Role of Age and Urbanization Level
Bornioli, A., Vermeulen, S., van Haaren, J., Valente, R., & Mingardo, G. (2022) · Sustainability · doi:10.3390/su14031157 · Cited by 16
“The method employs an “instrument” variable that fulfils three conditions [37,38]: (1) it is associated with the regressor variable (tourism pressure); (2) it does not affect the dependent variable, except through the independent variable; and (3) it does not have causes shared with the dependent variable.”
BackgroundMethod
Commentary
Swanson, S. A. (2016) · Epidemiology · doi:10.1097/ede.0000000000000558 · Cited by 16
Self-citation
Stronger Instruments and Refined Covariate Balance in an Observational Study of the Effectiveness of Prompt Admission to Intensive Care Units
Keele, L. J., Harris, S. K., Pimentel, S. D., & Grieve, R. (2018) · Journal of the Royal Statistical Society Series A (Statistics in Society) · doi:10.1111/rssa.12437 · Cited by 15
Predicting the Direction of Causal Effect Based on an Instrumental Variable Analysis: A Cautionary Tale
Burgess, S., & Small, D. S. (2016) · Journal of Causal Inference · doi:10.1515/jci-2015-0024 · Cited by 16
The effect of tracer contact on return to care among adult, “lost to follow‐up” patients living with HIV in Zambia: an instrumental variable analysis
Beres, L. K., Mody, A., Sikombe, K., Nicholas, L. H., Schwartz, S., Eshun‐Wilson, I., Somwe, P., Simbeza, S., Pry, J. M., Kaumba, P. C., et al. (2021) · Journal of the International AIDS Society · doi:10.1002/jia2.25853 · Cited by 14
Graphical Diagnosis of Confounding Bias in Instrumental Variable Analysis
Zhao, Q., & Small, D. S. (2018) · Epidemiology · doi:10.1097/ede.0000000000000822 · Cited by 14
Quality of Care in Chronic Kidney Disease and Incidence of End-stage Renal Disease in Older Patients
Fukuma, S., Ikenoue, T., Shimizu, S., Norton, E. C., Saran, R., Yanagita, M., Kato, G., Nakayama, T., & Fukuhara, S. (2020) · Medical Care · doi:10.1097/mlr.0000000000001301 · Cited by 13
Survival After Surgical Ablation of Atrial Fibrillation During Coronary Artery Bypass in Medicare Beneficiaries
Schaffer, J. M., Kluis, A., Squiers, J. J., George, T. J., Shih, E., Banwait, J. K., Mack, M. J., & DiMaio, J. M. (2025) · The Annals of Thoracic Surgery · doi:10.1016/j.athoracsur.2025.03.044 · Cited by 14
Genetically defined elevated homocysteine levels do not result in widespread changes of DNA methylation in leukocytes
Mandaviya, P. R., Joehanes, R., Aïssi, D., Kühnel, B., Marioni, R. E., Trương, V., Stolk, L., Beekman, M., Bonder, M. J., Franke, L., et al. (2017) · PLoS ONE · doi:10.1371/journal.pone.0182472 · Cited by 13
“Although assumptions one and two are satisfied in our case [5], the GRS model might violate assumption three [37, 38].”
Background
Evaluating different physician's prescribing preference based instrumental variables in two primary care databases: a study of inhaled long‐acting beta2‐agonist use and the risk of myocardial infarction
Uddin, M. J., Groenwold, R. H. H., de Boer, A., Afonso, A. S. M., Primatesta, P., Becker, C., Belitser, S. V., Hoes, A. W., Roes, K. C. B., & Klungel, O. H. (2016) · Pharmacoepidemiology and Drug Safety · doi:10.1002/pds.3860 · Cited by 13
Comparative effectiveness research methodology using secondary data: A starting user’s guide
Sun, M., & Lipsitz, S. R. (2017) · Urologic Oncology Seminars and Original Investigations · doi:10.1016/j.urolonc.2017.10.011 · Cited by 13
Methodological Challenges When Studying Distance to Care as an Exposure in Health Research
Caniglia, E. C., Zash, R., Swanson, S. A., Wirth, K. E., Diseko, M., Mayondi, G. K., Lockman, S., Mmalane, M. O., Makhema, J. M., Dryden‐Peterson, S., et al. (2019) · American Journal of Epidemiology · doi:10.1093/aje/kwz121 · Cited by 13
Self-citation
Determinants of selection into buprenorphine/naloxone among people initiating opioid agonist treatment in British Columbia
Homayra, F., Hongdilokkul, N., Piske, M., Pearce, L. A., Zhou, H., Min, J. E., Krebs, E., & Nosyk, B. (2019) · Drug and Alcohol Dependence · doi:10.1016/j.drugalcdep.2019.107798 · Cited by 13
Framing Causal Questions in Life Course Epidemiology
Stavola, B. L. D., Herle, M., & Pickles, A. R. (2021) · Annual Review of Statistics and Its Application · doi:10.1146/annurev-statistics-040120-024748 · Cited by 12
Does Transfer to Intensive Care Units Reduce Mortality? A Comparison of an Instrumental Variables Design to Risk Adjustment
Keele, L. J., Harris, S. K., & Grieve, R. (2019) · Medical Care · doi:10.1097/mlr.0000000000001093 · Cited by 12
Calendar time as an instrumental variable in assessing the risk of heart failure with antihyperglycemic drugs
Gokhale, M., Buse, J. B., Mack, C. D., Funk, M. J., Lund, J. L., Simpson, R. J., & Stürmer‎, T. (2018) · Pharmacoepidemiology and Drug Safety · doi:10.1002/pds.4578 · Cited by 12
“It is often conservatively suggested to avoid IV methods when the assumptions are likely to be violated because relatively minor violations of the assumptions may lead to large biases.(14) This can be particularly problematic when there is no valid comparison analytic method.”
Using Counterfactual Worlds to Triangulate Evidence in the Real World
Labrecque, J. A., & Swanson, S. A. (2023) · Current Epidemiology Reports · doi:10.1007/s40471-023-00340-0 · Cited by 10
Self-citation
Satisfaction with pandemic management and compliance with public health measures: Evidence from a German household survey on the COVID-19 crisis
Jaschke, P., Keita, S., Vallizadeh, E., & Kühne, S. (2023) · PLoS ONE · doi:10.1371/journal.pone.0281893 · Cited by 12
Physician’s Preference-based Instrumental Variable Analysis
Boef, A. G. C., van Paassen, J., Arbous, M. S., Middelkoop, A., Vandenbroucke, J. P., le Cessie, S., & Dekkers, O. M. (2014) · Epidemiology · doi:10.1097/ede.0000000000000151 · Cited by 11
“Heart failure 18 (17) 9 (11) 19 (21) 14 (15) 9 (10)”
Effect of pharmacological treatment of attention-deficit/hyperactivity disorder on later psychiatric comorbidity: a population-based prospective long-term study
Lyhmann, I., Widding‐Havneraas, T., Bjelland, I., Markussen, S., Elwert, F., Chaulagain, A., Mykletun, A., & Halmøy, A. (2024) · BMJ Mental Health · doi:10.1136/bmjment-2024-301003 · Cited by 10
“17 22–24 SEs in all models were clustered at the clinic level, and we report 95% CIs.”
Background
Stent revascularization versus bypass surgery for peripheral artery disease in type 2 diabetic patients – an instrumental variable analysis
Chang, C., Lin, J., Hsu, J., Wu, L., & Lai, M. (2016) · Scientific Reports · doi:10.1038/srep37177 · Cited by 10

Authors: Swanson, S. A., & Hernán, M. A.
Affiliations: Harvard University (US)
Corresponding author: Swanson, Sonja A.
ORCID: Swanson, Sonja A.
Venue: Epidemiology
Publisher: Ovid Technologies (Wolters Kluwer Health)
Published: 2013-05
Volume: 24
Issue: 3
Pages: 370-374
ISSN: 1044-3983 (ISSN-L)
DOI: 10.1097/ede.0b013e31828d0590
OpenAlex: W2330290021
Cited by: 197
References: 30
Citations by year: 2026: 8 · 2025: 15 · 2024: 15 · 2023: 12 · 2022: 23 · 2021: 14 · 2020: 9 · 2019: 18 · 2018: 16 · 2017: 15 · 2016: 19 · 2015: 21 · 2014: 10 · 2013: 1
FWCI: 14.82
Type: journal article
Language: en
Open access: closed
Locations: Epidemiology (publishedVersion)
PubMed (publishedVersion)
Topics: Physical Sciences › Mathematics › Statistics and Probability › Advanced Causal Inference Techniques
Social Sciences › Decision Sciences › Statistics, Probability and Uncertainty › Meta-analysis and systematic reviews
Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods in Epidemiology
Keywords: instrumental variables, reporting guidelines, confounding, exclusion restriction, weak instruments, reporting checklist, observational studies, medical interventions, COX-2 inhibitors, nonsteroidal anti-inflammatory drugs, causal effect estimation
MeSH terms: Data Interpretation, Statistical, Female, Gastrointestinal Diseases, Humans, Male, Myocardial Infarction, Models, Statistical, Epidemiologic Research Design, Causality, Confounding Factors, Epidemiologic, Cyclooxygenase 2 Inhibitors, Research Report

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Commentary: How to Report Instrumental Variable Analyses (Suggestions Welcome)

Abstract

Overview

Integrity signals

Verification

Claims the paper makes

Claims other papers make about this one

Citations

Cites

Cited by

Related

Metadata

Record

About these checks