| Re: Standardizing glucose measurements across surveys [message #20034 is a reply to message #20026] |
Mon, 21 September 2020 08:01   |
Bridgette-DHS
Messages: 3199
Registered: February 2013
|
Senior Member |
|
|
Following is a response by Senior DHS Staff, Gulnara Semenov and Fred Arnold:
In NFHS-4 random blood glucose was taken as it was not possible to obtain a sample of respondents who had truly 'fasted', even though time since last meal (smb51) or time since last drink (smb52) information was collected. The dataset includes a recoded variable for glucose level (smb70) and the cut-off levels used to determine high glucose levels are 141-160mg/dl high glucose and >160mg/d very high glucose. This is regardless of whether or not the person was taking medication for diabetes. If you're interested in those taking medication, you could tabulate the percentage with normal blood glucose levels (<=140mg/dl) and taking medication. Please note that the high glucose values from the random blood glucose are not sufficient to 'diagnose' diabetes. For more information on diagnostic criteria for diabetes, please read this WHO report: https://www.who.int/publications/i/item/classification-of-di abetes-mellitus
In 2016-2017 Haiti DHS, The HemoCue® HbA1c 501 was used to measure glycated hemoglobin (HbA1c) in capillary blood. Blood glucose status based on glycated hemoglobin (HbA1c) testing is expressed in percentage (%). An HbA1c of 6.5% is used as a cut off (WHO 2011). This is described in details in the final report, page 8, 322, 337, Table 18.17 https://dhsprogram.com/pubs/pdf/FR326/FR326.pdf
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
