Consequences of dichotomization

Fedorov, Valerii V.; Mannino, Frank; Zhang, Rongmei

doi:10.1002/pst.331

Cited by 247 publications

(184 citation statements)

References 20 publications

Supporting

Mentioning

178

Contrasting

Unclassified

Order By: Relevance

“…Several studies 16,17 have demonstrated that 100 continuous observations are statistically equivalent to at least 157 dichotomized observations. Selvin 18 derives a formula to calculate the efficiency loss due to categorizing a continuous variable.…”

Section: Problems With Dichotomizationmentioning

confidence: 99%

Analysis by Categorizing or Dichotomizing Continuous Variables Is Inadvisable: An Example from the Natural History of Unruptured Aneurysms

Naggara¹,

Raymond²,

Guilbert³

et al. 2011

AJNR Am J Neuroradiol

243

153

View full text Add to dashboard Cite

SUMMARY:In medical research analyses, continuous variables are often converted into categoric variables by grouping values into Ն2 categories. The simplicity achieved by creating Ն2 artificial groups has a cost: Grouping may create rather than avoid problems. In particular, dichotomization leads to a considerable loss of power and incomplete correction for confounding factors. The use of data-derived "optimal" cut-points can lead to serious bias and should at least be tested on independent observations to assess their validity. Both problems are illustrated by the way the results of a registry on unruptured intracranial aneurysms are commonly used. Extreme caution should restrict the application of such results to clinical decision-making. Categorization of continuous data, especially dichotomization, is unnecessary for statistical analysis. Continuous explanatory variables should be left alone in statistical models.ABBREVIATIONS: ACA ϭ anterior cerebral artery; CHUM ϭ Centre hospitalier de l'Université de Montré al; ICA ϭ internal carotid artery; ISUIA ϭ International Study of Unruptured Intracranial Aneurysms; MCA ϭ middle cerebral artery; Pcirc ϭ posterior circulation; PcomA ϭ posterior communicating artery; SAH ϭ subarachnoid hemorrhage; UIA ϭ unruptured intracranial aneurysms U IAs are common (approximately 2% of the adult population), but they most often remain silent until a rupture occurs (incidence, 2-20/10,000/year).1 No one is sure what to do with them, but with the increasing accessibility of noninvasive imaging of the brain, the problem is growing rapidly. 2A common and yet controversial approach to decisionmaking is to compare the natural history of the disease and the risks of treatment.3,4 One prominent risk factor for rupture of UIAs is size. In 1998, a landmark study on this subject, the ISUIA, estimated from retrospectively obtained data that the risk of rupture of aneurysms smaller than 10 mm was extremely low.5 Subsequent guidelines published in 2000 discouraged the treatment of aneurysms smaller than that size. 6In a 2003 study, the same group, confronted with different results when data were collected prospectively, claimed that aneurysms Ͻ7 mm in a special subgroup of patients (defined by the absence of a history of rupture of another lesion, having an aneurysm located in the anterior circulation, and selected for observation) were at zero risk of rupture, but only when some carotid aneurysms were excluded (PcomA aneurysms). 7Despite these specifications, a threshold of 7 mm is now used by many as a normative criterion for clinical decisions 8 or in cost-effectiveness analyses. 9Size of UIAs can serve to illustrate the problems associated with the categorization of continuous variables, in particular dichotomization. Our aim is to consider how continuous variables should be treated and analyzed when we suspect that risk increases or decreases in proportion to the variable in question. We address the following questions: What are the advantages and disadvantages of categorization? If we de...

show abstract

Section: Problems With Dichotomizationmentioning

confidence: 99%

Analysis by Categorizing or Dichotomizing Continuous Variables Is Inadvisable: An Example from the Natural History of Unruptured Aneurysms

Naggara¹,

Raymond²,

Guilbert³

et al. 2011

AJNR Am J Neuroradiol

243

153

View full text Add to dashboard Cite

show abstract

“…Dichotomising variables reduces power, can bias estimates, and can increase false positives. [19][20][21][22][23][24][25][26][27][28][29][30][31][32][33] To illustrate this point, Figure 2 shows data for 50 samples that have a correlation of 0.4. These are naturally analysed with a Pearson correlation or linear regression (solid line), both of which give p = 0.002 for the association.…”

Section: Don't Dichotomise or Bin Continuous Variablesmentioning

confidence: 99%

Four simple ways to increase power without increasing the sample size

Lazic

2017

Preprint

View full text Add to dashboard Cite

Underpowered experiments have three problems: the probability of a false positive result is higher, true effects are harder to detect, and the true effects that are detected tend to have inflated effect sizes. Many biology experiments are underpowered and recent calls to change the traditional 0.05 significance threshold to a more stringent value of 0.005 will further reduce the power of the average experiment. Increasing power by increasing the sample size is often the only option considered, but more samples increases costs, makes the experiment harder to conduct, and is contrary to the 3Rs principles for animal research. We show how the design of an experiment and some analytical decisions can have a surprisingly large effect on power.

show abstract

“…There has also been interest in promoting the analysis of continuous rather than dichotomous measures in clinical trials and preclinical research. 38 In AKI, power analysis would be on the basis of the anticipated changes in actual creatinine values rather than setting an artificial dichotomy for what is deemed to be a clinically important effect (such as a 50% reduction in eGFR). This would increase the statistical power of a study without requiring larger numbers.…”

Section: Common Concerns About Preclinical Study Design and Reportingmentioning

confidence: 99%

“…However, alternative approaches for statistical analysis on the basis of the use of continuous versus dichotomous end point variables (such as creatinine) could be used to mitigate the effect of these requirements of research costs. 38 Other problems will be even more difficult to address, and potential solutions will be less obvious. For example, calling for the publication of negative results is unquestionably good for the science community at large but consumes an individual scientist's time and resources with limited academic reward.…”

Section: Recommendations Implementation and Challengesmentioning

confidence: 99%

Bridging Translation by Improving Preclinical Study Design in AKI

Caestecker

Humphreys

Liu

et al. 2015

Journal of the American Society of Nephrology

View full text Add to dashboard Cite

Despite extensive research, no therapeutic interventions have been shown to prevent AKI, accelerate recovery of AKI, or reduce progression of AKI to CKD in patients. This failure in translation has led investigators to speculate that the animal models being used do not predict therapeutic responses in humans. Although this issue continues to be debated, an important concern that has not been addressed is whether improvements in preclinical study design can be identified that might also increase the likelihood of translating basic AKI research into clinical practice using the current models. In this review, we have taken an evidence-based approach to identify common weaknesses in study design and reporting in preclinical AKI research that may contribute to the poor translatability of the findings. We focused on use of N-acetylcysteine or sodium bicarbonate for the prevention of contrastinduced AKI and use of erythropoietin for the prevention of AKI, two therapeutic approaches that have been extensively studied in clinical trials. On the basis of our findings, we identified five areas for improvement in preclinical study design and reporting. These suggested and preliminary guidelines may help improve the quality of preclinical research for AKI drug development.

show abstract

Consequences of dichotomization

Cited by 247 publications

References 20 publications

Analysis by Categorizing or Dichotomizing Continuous Variables Is Inadvisable: An Example from the Natural History of Unruptured Aneurysms

Analysis by Categorizing or Dichotomizing Continuous Variables Is Inadvisable: An Example from the Natural History of Unruptured Aneurysms

Four simple ways to increase power without increasing the sample size

Bridging Translation by Improving Preclinical Study Design in AKI

Contact Info

Product

Resources

About