Subgroup analysis and statistical power

Martins, Wellington P.; Zanardi, José Vitor Cabral

doi:10.1016/j.ejogrb.2011.07.046

Cited by 2 publications

(2 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…ES was used to define the discriminative capacity, and was computed as the difference between the mean of the two groups mentioned above, divided by the pooled standard deviation. The pooled standard deviation was estimated from the corrected standard errors and the weighted number of individuals in the groups [ 39 ]. General guidelines define an effect size of 0.2 as small, 0.5 as moderate, and 0.8 as large [ 40 ].…”

Section: Methodsmentioning

confidence: 99%

Validation and comparison of EuroQoL-5 dimension (EQ-5D) and Short Form-6 dimension (SF-6D) among stable angina patients

Han

Zhao

et al. 2014

Health Qual Life Outcomes

View full text Add to dashboard Cite

ObjectivesSeveral preference-based health-related quality of life (HRQoL) instruments have been published and widely used in different populations. However no consensus has emerged regarding the most appropriate instrument in therapeutic area of stable angina. This study compared and validated the psychometric properties of two generic preference-based instruments, the EQ-5D and SF-6D, among Chinese stable angina patients.MethodsConvergent validity of the EQ-5D and SF-6D was examined with eight a priori hypotheses from stable angina patients in conjunction with Seattle Angina Questionnaire (SAQ). Responsiveness was compared using the effect size (ES), relative efficiency (RE) and receiver operating characteristic (ROC) curves. Agreement between the EQ-5D and SF-6D was tested using intra-class correlation coefficient (ICC) and Bland-Altman plot. Factors affecting utility difference were explored with multiple linear regression analysis.ResultsIn 411 patients (mean age 68.08 ± 11.35), mean utility scores (SD) were 0.78 (0.15) for the EQ-5D and 0.68 (0.12) for the SF-6D. Validity was demonstrated by the moderate to strong correlation coefficients (Range: 0.368-0.594, P< 0.001) for five of the eight hypotheses in both the EQ-5D and SF-6D. There were no serious floor effects for the EQ-5D and SF-6D, but ceiling effects for the EQ-5D were large. The areas under ROC of them all exceeded 0.5 (0.660-0.814, P< 0.001). The SF-6D showed a better discriminative capacity (ES: 0.573 to 1.179) between groups with different stable-angina-specific health status than the EQ-5D (ES: 0.426 to 1.126). RE suggested that the SF-6D (RE: 44.8 to 177.8%) was more efficient than the EQ-5D except for physical function. Poor agreement between them was observed with ICC (0.448, P< 0.001) and Bland-Altman plot analysis. Multiple liner regression showed that clinical variables significantly (P< 0.05) influenced differences in utility scores between the EQ-5D and SF-6D.ConclusionsBoth EQ-5D and SF-6D are valid and sensitive preference-based HRQoL instruments in Chinese stable angina patients. The SF-6D may be a more effective tool with lower ceiling effect and greater sensitivity. Further study is needed to compare other properties, such as reliability and longitudinal response.

show abstract

Section: Methodsmentioning

confidence: 99%

Validation and comparison of EuroQoL-5 dimension (EQ-5D) and Short Form-6 dimension (SF-6D) among stable angina patients

Han

Zhao

et al. 2014

Health Qual Life Outcomes

View full text Add to dashboard Cite

show abstract

“…First, there are many signals that induce EGFR positivity in IHC, not just EGFR amplification [27][28][29]. Second, if the candidate biomarker is tested in the cohort including a subgroup in which the biomarker has no impact on predicting prognosis, the biomarker may come out as ineffective by reducing the statistical effect size [30]. This means that although EGFR positivity may play a role as a prognostic biomarker in a small subgroup, its role as a biomarker with statistical significance could be hidden due to the heterogeneity of the study group.…”

Section: Discussionmentioning

confidence: 99%

Cyclin D1 Serves as a Poor Prognostic Biomarker in Stage I Gastric Cancer

Lee

et al. 2022

CIMB

View full text Add to dashboard Cite

TNM stage still serves as the best prognostic marker in gastric cancer (GC). The next step is to find prognostic biomarkers that detect subgroups with different prognoses in the same TNM stage. In this study, the expression levels of epidermal growth factor receptor (EGFR) and cyclin D1 were assessed in 96 tissue samples, including non-tumorous tissue, adenoma, and carcinoma. Then, the prognostic impact of EGFR and cyclin D1 was retrospectively investigated in 316 patients who underwent R0 resection for GC. EGFR positivity increased as gastric tissue became malignant, and cyclin D1 positivity was increased in all the tumorous tissues. However, there was no survival difference caused by the EGFR positivity, while the cyclin D1-postive group had worse overall survival (OS) than the cyclin D1-negative group in stage I GC (10-year survival rate (10-YSR): 62.8% vs. 86.5%, p = 0.010). In subgroup analyses for the propensity score-matched (PSM) cohort, there were also significant differences in the OS according to the cyclin D1 positivity in stage I GC but not in stage II and III GC. Upon multivariate analysis, cyclin D1 positivity was an independent prognostic factor in stage I GC. In conclusion, cyclin D1 may be a useful biomarker for predicting prognosis in stage I GC.

show abstract

Subgroup analysis and statistical power

Cited by 2 publications

References 4 publications

Validation and comparison of EuroQoL-5 dimension (EQ-5D) and Short Form-6 dimension (SF-6D) among stable angina patients

Validation and comparison of EuroQoL-5 dimension (EQ-5D) and Short Form-6 dimension (SF-6D) among stable angina patients

Cyclin D1 Serves as a Poor Prognostic Biomarker in Stage I Gastric Cancer

Contact Info

Product

Resources

About