Data Management And Sharing System of China National Clinical Research Center for Neurological Diseases
Letter by Gu and Liu Regarding Article, "HbA1c (Glycated Hemoglobin) Levels and Clinical Outcome Post-Mechanical Thrombectomy in Patients With Large Vessel Occlusion"
First author：Gu Hong-Qiu
Corresponding author：Gu Hong-Qiu
essential_information
- Person in charge：Gu Hong-Qiu
- Data Sources： OTHERS
- Number of papers：OTHERS20190778
- Published time：2019-05-14
2019-05-14
Published articles
Journal：Stroke
Publication type: Letter To The Editor
IF(IF): 7.914
PMID: 31084318 Link To Pubmed Full text Download
Background And Purpose
Method
In their study, Choi et al1 stated that the number of patients with favorable outcomes was significantly lower in the HbA1c>6.5% group than in the HbA1c≤6.5% group. A shift analysis from the Cochran-Mantel-Haenszel test also revealed a statistically sig- nificant shift in the distributions of mRS scores between HbA1c groups (P=0.005).
Conclusion
It is important to note that the P value for the shift analysis should come from the line of row mean scores differ, which treats the column variable as an ordinal variable. The P value in the General Association line is used in the x2 test for dichotomized analyses, which treats the column variable as a categorical variable.
Main results
However, we found that the P value is 0.0003 in the Cochran-Mantel-Haenszel test for shift analysis using the data provided in this study, not 0.005 in the χ2 test for dichoto- mized analysis as was reported. In addition, the P values in the χ2 test for favorable outcomes and excellent outcomes are 0.043 and 0.0198, respectively, not 0.006 and 0.024 as was reported in the main text. Furthermore, the adjusted odds ratio for mortality in Table 3 in the original article were all significantly <1, indicat- ing that increased HbA1c, admission serum glucose, and fasting blood glucose on day 2 are protective factors of mortality. This is counterintuitive and confusing for readers without additional comments, interpretation or statistical exploration; for example, we could use restricted cubic splines or fractional polynomials to better understand the nonlinear relationship between these glu- cose parameters and mortality.