Treatment of continuous data as categoric variables in

Priore, Giuseppe Del; Zandieh, Peyman; Lee, Men Jean

doi:10.1016/s0029-7844(96)00504-2

Cited by 35 publications

(5 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To build prognostic models, most studies have included continuous predictors as categorized forms with their arbitrary optimal cut-off points. Although there has been constant criticism of this approach (15)(16)(17), categorization is widespread in clinical studies (18). One of the perceived advantages of categorization is that it is easy to apply to clinical practice that determines the diagnostic or therapeutic procedures.…”

Section: Introductionmentioning

confidence: 99%

Risk Stratification Using Multivariable Fractional Polynomials in Diffuse Large B-Cell Lymphoma

et al. 2020

View full text Add to dashboard Cite

The risk stratification of diffuse large B-cell lymphoma (DLBCL) is crucial. The International Prognostic Index, the most commonly used and the traditional risk stratification system, is composed of fixed and artificially dichotomized attributes. We aimed to develop a novel prognostic model that allows the incorporation of up-to-date attributes comprehensively without information loss. We analyzed 204 patients with primary DLBCL who were uniformly treated with R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone) from 2007 to 2012 at Asan Medical Center. Using the multivariable fractional polynomial (MFP) method and bootstrap resampling, we selected the variables of significance and the best fitted functional form in fractional polynomials. Age, serum β2-microglobulin, serum lactate dehydrogenase, and BCL2 expression were selected as significant variables in predicting overall survival (OS), while age was excluded in predicting 2-years event-free survival. The prognostic score calculated by the MFP model effectively classifies patients into four risk groups with 5-years OS of 89.91% (low risk), 81.21% (low-intermediate risk), 66.40% (high-intermediate risk), and 37.89% (high risk). We suggest a new prognostic model that is simple and flexible. By using the MFP method, we can incorporate various clinicopathologic factors into a risk stratification system without arbitrary dichotomization.

show abstract

Section: Introductionmentioning

confidence: 99%

Risk Stratification Using Multivariable Fractional Polynomials in Diffuse Large B-Cell Lymphoma

et al. 2020

View full text Add to dashboard Cite

show abstract

“…In assuming normality and using standard linear regression, investigators are restricted to how factors shift the mean birthweight and are blind to whether a factor more strongly impacts infants at the lower (SGA) or upper (LGA) tail of the birthweight [18]. Alternatively, logistic regression on a SGA/LGA birthweight is often used [19] for the ease of clinical interpretation; however by dichotomizing the outcome of interest (e.g. LGA) the impact on the full distribution is ignored and results are not intuitive for individuals falling near to but at opposite sides of the cutoff point.…”

Section: Introductionmentioning

confidence: 99%

Quantifying the Impact of Gestational Diabetes Mellitus, Maternal Weight and Race on Birthweight via Quantile Regression

et al. 2013

View full text Add to dashboard Cite

BackgroundQuantile regression, a robust semi-parametric approach, was used to examine the impact of gestational diabetes mellitus (GDM) across birthweight quantiles with a focus on maternal prepregnancy body mass index (BMI) and gestational weight gain (GWG).MethodsUsing linked birth certificate, inpatient hospital and prenatal claims data we examined live singleton births to non-Hispanic white (NHW, 135,119) and non-Hispanic black (NHB, 76,675) women in South Carolina who delivered 28–44 weeks gestation in 2004–2008.ResultsAt a maternal BMI of 30 kg/m2 at the 90th quantile of birthweight, exposure to GDM was associated with birthweights 84 grams (95% CI 57, 112) higher in NHW and 132 grams (95% CI: 104, 161) higher in NHB. Results at the 50th quantile were 34 grams (95% CI: 17, 51) and 78 grams (95% CI: 56, 100), respectively. At a maternal GWG of 13.5 kg at the 90th quantile of birthweight, exposure to GDM was associated with birthweights 83 grams (95% CI: 57, 109) higher in NHW and 135 grams (95% CI: 103, 167) higher in NHB. Results at the 50th quantile were 55 grams (95% CI: 40, 71) and 69 grams (95% CI: 46, 92), respectively.SummaryOur findings indicate that GDM, maternal prepregnancy BMI and GWG increase birthweight more in NHW and NHB infants who are already at the greatest risk of macrosomia or being large for gestational age (LGA), that is those at the 90th rather than the median of the birthweight distribution.

show abstract

“…In addition, in the pursuit of simplicity, some of the current scoring systems handle the BPM in a categorical manner when they are, in fact, generated as continuous variables. It was reported that dichotomisation of continuous variables in a multiple regression procedure may be associated with considerable loss of statistical power and introduction of bias[33,34]. However, we should acknowledge that considering an indicator as a continuous variable does not necessarily improve the prediction accuracy.…”

Section: Discussionmentioning

confidence: 92%

Blood parameters score predicts long-term outcomes in stage II-III gastric cancer patients

Tang

Wang

et al. 2019

WJG

View full text Add to dashboard Cite

BACKGROUNDIncreasing numbers of laboratory blood parameters (BPM) have been reported to greatly affect the long-term outcomes of gastric cancer (GC) patients. However, the existing prognostic models do not comprehensively analyze these predictors.AIMTo construct a new prognostic tool, based on all the prognostic BPM, to achieve more accurate prognosis prediction for GC.METHODSWe retrospectively assessed 850 consecutive patients who underwent curative resection for stage II-III GC from January 2010 to April 2013. The patients were classified into developing (n = 567) and validation (n = 283) cohorts using computer-generated random numbers. A scoring system, namely BPM score, was then constructed using least absolute shrinkage and selection operator (LASSO) Cox regression model in the developing cohort, and validated in the validation cohort. A nomogram consisting of BPM score and tumor-lymph node-metastasis (TNM) stage was further created. The discrimination and calibration of the nomogram were evaluated via Harrell’s C-statistic and the Hosmer-Lemeshow test.RESULTSUsing the LASSO model, we established the BPM score based on five BPM: Albumin, lymphocyte-to-monocyte ratio, neutrophil-to-lymphocyte ratio, carcinoembryonic antigen, and carbohydrate antigen 19-9. The BPM scores were divided into high- and low-BPM groups based on a cut-off value of -0.93. High-BPM patients were significantly older and had more advanced, larger tumors. In the developing cohort, significant differences were found in 5-year overall survival (OS) and 5-year disease-specific survival between the high-BPM and low-BPM patients. Similar results were found in the validation group. Multivariable analysis showed that the BPM score was an independent predictor of OS. High-BPM patients had a poorer 5-year OS for each subgroup. Furthermore, a nomogram that combined the BPM score and TNM stage had significantly better prognostic value compared with TNM stage alone.CONCLUSIONThe BPM score provides more accurate prognosis prediction in stage II-III GC patients and is an effective complement to the TNM staging system.

show abstract

Treatment of continuous data as categoric variables in

Cited by 35 publications

References 6 publications

Risk Stratification Using Multivariable Fractional Polynomials in Diffuse Large B-Cell Lymphoma

Risk Stratification Using Multivariable Fractional Polynomials in Diffuse Large B-Cell Lymphoma

Quantifying the Impact of Gestational Diabetes Mellitus, Maternal Weight and Race on Birthweight via Quantile Regression

Blood parameters score predicts long-term outcomes in stage II-III gastric cancer patients

Contact Info

Product

Resources

About