Fig 1 - uploaded by Roberto S Rosales
Content may be subject to copyright.
Source publication
Statistics, the science of numerical evaluation, helps in determining the real value of a hand surgical intervention. Clinical research in hand surgery cannot improve without considering the application of the most appropriate statistical procedures. The purpose of the present paper is to approach the basics of data analysis using a database of car...
Similar publications
Malaysia Ministry of Education has put critical thinking skills elements into one of
the six aspirations that pupils need to achieve. This was a turning point that
teaching methods should be transformed. Therefore, innovation of teaching
methods needs to be done. One of the strategies suggested is adventure-based
learning (ABL). The main objective...
Background
Reproductive, maternal, newborn, child health, and nutrition (RMNCH&N) data is an indispensable tool for program and policy decisions in low- and middle-income countries. However, being equipped with evidence doesn’t necessarily translate to program and policy changes. This study aimed to characterize data visualization interpretation ca...
Malaysia Ministry of Education has put critical thinking skills elements into one of the six aspirations that pupils need to achieve. This was a turning point that teaching methods should be transformed. Therefore, innovation of teaching methods needs to be done. One of the strategies suggested is adventure-based learning (ABL). The main objective...
Citations
... When the model was adjusted for confounders (age, aetiology and previous surgery), achieving a discrimination capacity of 95%, the odds ratios for SRUIS increased with Darrach and S-K, taking Bowers procedure as reference, although statistical significance was not reached. The estimated probabilities of the unadjusted model (Table 3) coincided with the percentages of SRUIS after the surgical procedure (Table 1) because the chi-squared test is a particular form of logistic regression analysis (Rosales and Atroshi, 2018). However, they completely differed from the estimated probabilities based on the confounderadjusted model. ...
We conducted an ambispective cohort study to assess the association between symptomatic radioulnar impingement syndrome (SRUIS) and distal radioulnar joint (DRUJ) salvage surgery to examine the influence of confounders on the final effect. The outcome variable was the incidence of SRUIS and the exposure variable was the surgical procedure. Seventy-two patients with median age of 48 years (IQR 25–78) were examined using bivariate and logistic regression multivariate analyses, and confounders were analysed in 15 multivariate models. Overall, SRUIS occurred in 21 patients (29%). Bivariate analysis showed a significant association between SRUIS and type of surgical procedure, observed in 71% after Sauvé–Kapandji, 50% after Bowers and 15% after Darrach procedure. When adjusted for age, aetiology and previous surgery, the significant association disappeared. Confounding is an important factor when accounting for SRUIS after DRUJ salvage surgery. The risk of SRUIS did not depend on the procedure, but rather on patient’s age, aetiology and previous surgery.
Level of evidence: II
... Observe in the reference model the hierarchical principle, in black the covariate X1, if the interaction 'X * X1' is included in the model it is recommended that the variable X1 should be included as confounder. binomial regression models); outcome variable of interest is time until an event occurs as in survival analysis (Cox and Fine-Gray regression models) (Rosales and Atroshi, 2018). An example of modelling confounding in survival analysis is shown in the article by Corella et al. (2021). ...
Confounding and effect modification are two key epidemiological concepts in clinical research. They influence the association between an exposure and a disease in causation or aetiology studies, or between an intervention and an outcome in treatment effect studies. Although they are different concepts, distinguishing between confounding and effect modification is not always easy. Moreover, a covariate may be a confounder, an effect modifier, or both at the same time.
The purpose of this article is to introduce these epidemiological concepts that may not be familiar in the field of hand surgery and explain how they should be handled in clinical research, focusing on multivariate regression analysis.
... Some studies have used the classic Pearson or Spearman correlation coefficients for measuring agreement between quantitative variables (Wilcke et al., 2009). However, the correlation analysis is not a correct strategy because the classic correlation coefficients assess the degree of linear association between two measures, but they do not assess the actual agreement (Bland and Altman, 1981;Domenech, 2017;Rosales and Atroshi, 2018). The ICC measures the agreement between two quantitative (continuous) variables (e.g. ...
This article presents the methodological requirements for clinical examination and patient-reported outcomes measurements. The assessment of any measurement for clinical research in hand surgery is difficult. A method of measuring a criterion could be 100% reliable but 100% invalid. Bias may be present in our assessment if we do not take into account the methodological requirements related to reliability, validity, and responsiveness of our measures. Reliability refers to intra-observer agreement, inter-observer agreement, or agreement between two methods of assessment, and, for patient-reported measures, internal consistency and test–retest reliability. Validity is the capability of a clinical method to measure what it proposes to measure. Assessing validity involves comparing a measure with one or more other measures, and, if possible, with a reference standard criterion. Responsiveness is the ability to detect important clinical change. The Consensus-based Standards for the Selection of Health Measurement Instruments provides the standards required for design and recommended statistical analyses of patient-reported outcome measures.
... Missing data were observed in 7% of the observations, corresponding to 17 subjects with one missing value in the variable 'UseIntrinPP', and two individuals with missing value in the variable 'YearsHS'. In spite of that the missing values, incidence was acceptable (<10% observations); they were estimated by mean imputation with a reduction of the standard deviation (SD) of 0.38% for 'YearsHS' and 3.29% for 'UseIntrinPP', while the means were not affected (Rosales and Atroshi, 2018). ...
... Besides, the significant association between the 'UseIntrinPP and 'Age' as a quantitative variable (b = 0.429, p = 0.014) showed a percentage of change in the effect (b = 0.508) of 16% when the association between 'UseIntrinPP' and 'Age' was adjusted by 'YearsHS'. This implied that the 'YearsHS' was a confounder or confounding variable (Rosales and Atroshi, 2018). ...
... Observe in the reference model the hierarchical principle, in black the covariate X1, if the interaction 'X * X1' is included in the model it is recommended that the variable X1 should be included as confounder. binomial regression models); outcome variable of interest is time until an event occurs as in survival analysis (Cox and Fine-Gray regression models) (Rosales and Atroshi, 2018). An example of modelling confounding in survival analysis is shown in the article by Corella et al. (2021). ...
