|
|
||||||||
Electronic Letters to:
|
Electronic letters published:
|
|
|||
|
Anne Helene Garde, Researcher National Institute of Occupational Health
Send letter to journal:
ahg{at}ami.dk Anne Helene Garde
|
In this article, the values for CVid, CVim, CViy and CVg were underestimated by a factor of approximately 2.5, because of a calculation error. Correspondingly, the CVa/CViy values were overestimated and the relative prediction intervals were too narrow. The correct CVid (not estimated for total cholesterol and HbA1c and therefore indicated by - for these two analytes) , CVim, CViy and CVg are: -, 15%, 13% and 12% for total cholesterol, 14%, 16%, 20% and 49% for DHEA-S, -, 3.0%, 7.0% and 7.5% for HbA1c/total hemoglobin, 5.7%, 16%, 14% and 30% for IgA, 55%, 40%, 58% and 63% for prolactin, and 19%, 64%, 55% and 68% for free testosterone. The correct CVa/CViy and relative prediction intervals are: 0.13 and (43%, 234%) for total-cholesterol, 0.26 and (27%, 368%) for DHEA- S, 0.26 and (63%, 158%) for HbA1c/total hemoglobin, 0.07 and (42%, 240%) for IgA, 0.05 and (3%, 3053%) for prolactin, and 0.13 and (4%, 2609%) for free testosterone. Please note this correction when referencing this paper. The authors apologize for any inconvenience this error may have caused. |
|||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH |