Skip to main content
Search
Search
Advanced Search
Skip main navigation
Open access
Research Article
3 April 2024

Assessing the performance of physician's prescribing preference as an instrumental variable in comparative effectiveness research with moderate and small sample sizes: a simulation study

Authors: Lisong Zhang https://orcid.org/0000-0002-0289-2933 [email protected], Jim Lewsey https://orcid.org/0000-0002-3811-8165, and David A McAllister https://orcid.org/0000-0003-3550-1764Author Info & Affiliations
Volume 13, Number 5
https://doi.org/10.57264/cer-2023-0044
PDFFormats

Abstract

Aim: This simulation study is to assess the utility of physician's prescribing preference (PPP) as an instrumental variable for moderate and smaller sample sizes. Materials & methods: We designed a simulation study to imitate a comparative effectiveness research under different sample sizes. We compare the performance of instrumental variable (IV) and non-IV approaches using two-stage least squares (2SLS) and ordinary least squares (OLS) methods, respectively. Further, we test the performance of different forms of proxies for PPP as an IV. Results: The percent bias of 2SLS is around approximately 20%, while the percent bias of OLS is close to 60%. The sample size is not associated with the level of bias for the PPP IV approach. Conclusion: Irrespective of sample size, the PPP IV approach leads to less biased estimates of treatment effectiveness than OLS adjusting for known confounding only. Particularly for smaller sample sizes, we recommend constructing PPP from long prescribing histories to improve statistical power.

Supplementary Material

File (supplementary material.docx)

References

1.
Brookhart MA, Schneeweiss S. Preference-based instrumental variable methods for the estimation of treatment effects: assessing validity and interpreting results. Int. J. Biostat. 3(1), Article 14 (2007).
PubMed
Google Scholar
2.
Kuo YF, Montie JE, Shahinian VB. Reducing bias in the assessment of treatment effectiveness: androgen deprivation therapy for prostate cancer. Med. Care 50(5), 374–380 (2012).
PubMed
Google Scholar
3.
Davies NM, Taylor AE, Taylor GM et al. Varenicline versus nicotine replacement therapy for long-term smoking cessation: an observational study using the Clinical Practice Research Datalink. Health Technol. Assess. 24(9), 1–46 (2020).
Google Scholar
4.
Kollhorst B, Abrahamowicz M, Pigeot I. The proportion of all previous patients was a potential instrument for patients' actual prescriptions of nonsteroidal anti-inflammatory drugs. J. Clin. Epidemiol. 69, 96–106 (2016).
PubMed
Google Scholar
5.
Taylor GMJ, Taylor AE, Thomas KH et al. The effectiveness of varenicline versus nicotine replacement therapy on long-term smoking cessation in primary care: a prospective cohort study of electronic medical records. Int. J. Epidemiol. 46(6), 1948–1957 (2017).
PubMed
Google Scholar
6.
Nelson RE, Nebeker JR, Hayden C, Reimer L, Kone K, LaFleur J. Comparing adherence to two different HIV antiretroviral regimens: an instrumental variable analysis. AIDS Behav. 17(1), 160–167 (2013).
PubMed
Google Scholar
7.
Boef AG, Dekkers OM, Vandenbroucke JP, le Cessie S. Sample size importantly limits the usefulness of instrumental variable methods, depending on instrument strength and level of confounding. J. Clin. Epidemiol. 67(11), 1258–1264 (2014).
PubMed
Google Scholar
8.
Zhang Z, Uddin MJ, Cheng J, Huang T. Instrumental variable analysis in the presence of unmeasured confounding. Ann. Transl. Med. 6(10), 182 (2018).
PubMed
Google Scholar
9.
Ionescu-Ittu R, Delaney JA, Abrahamowicz M. Bias-variance trade-off in pharmacoepidemiological studies using physician-preference-based instrumental variables: a simulation study. Pharmacoepidemiol. Drug Saf. 18(7), 562–571 (2009).
PubMed
Google Scholar
10.
Davies NM, Gunnell D, Thomas KH, Metcalfe C, Windmeijer F, Martin RM. Physicians' prescribing preferences were a potential instrument for patients' actual prescriptions of antidepressants. J. Clin. Epidemiol. 66(12), 1386–1396 (2013).
PubMed
Google Scholar
11.
Martens EP, Pestman WR, de Boer A, Belitser SV, Klungel OH. Instrumental variables: application and limitations. Epidemiology 17(3), 260–267 (2006).
PubMed
Google Scholar
12.
Ionescu-Ittu R, Abrahamowicz M, Pilote L. Treatment effect estimates varied depending on the definition of the provider prescribing preference-based instrumental variables. J. Clin. Epidemiol. 65(2), 155–162 (2012).
PubMed
Google Scholar
13.
Baiocchi M, Cheng J, Small DS. Instrumental variable methods for causal inference. Stat. Med. 33(13), 2297–2340 (2014).
PubMed
Google Scholar
14.
Tchetgen Tchetgen EJ, Walter S, Vansteelandt S, Martinussen T, Glymour M. Instrumental variable estimation in a survival context. Epidemiology 26(3), 402–410 (2015).
PubMed
Google Scholar
|