David McIlroy1, Neil Banham2. 1. Hyperbaric medicine registrar, Hyperbaric Medicine Unit, Fremantle Hospital, at the time of the study. 154 B Samson Street, White Gum Valley, WA 6162, Australia, Phone: +61-(0)407-445-602, E-mail: davidmcilroy75@hotmail.com. 2. Medical Director, Hyperbaric Medicine Unit, Fremantle Hospital, Fremantle, WA, Australia.
Abstract
INTRODUCTION: Blood glucose is commonly measured in diabetic patients undergoing hyperbaric oxygen treatment (HBOT) from a 'finger-prick' capillary sample. Although this method is an accurate reflection of venous glucose under normal conditions it has not been validated under hyperbaric, hyperoxic conditions. METHODS: Four patients with diabetes mellitus undergoing HBOT had venous blood samples drawn simultaneously with routine capillary samples before, during and immediately after three of four HBOT sessions. The Bland-Altman method of assessing agreement between these two measures was used separately for the three time periods. RESULTS: The relationship between venous and finger-prick glucose at room air was altered significantly by HBOT. The bias (finger-prick minus venous measurements) was significantly less than zero during the HBOT session but not immediately after completion of the session. Owing to the small sample size, the limits of agreement straddled zero at all time points, although the lower limit was close to zero during treatment (finger measurement appeared to be higher than venous measurement on room air and lower than venous undergoing HBOT). CONCLUSION: Finger-prick capillary sampling may not be an accurate reflection of venous glucose during HBOT.
INTRODUCTION:Blood glucose is commonly measured in diabeticpatients undergoing hyperbaric oxygen treatment (HBOT) from a 'finger-prick' capillary sample. Although this method is an accurate reflection of venous glucose under normal conditions it has not been validated under hyperbaric, hyperoxic conditions. METHODS: Four patients with diabetes mellitus undergoing HBOT had venous blood samples drawn simultaneously with routine capillary samples before, during and immediately after three of four HBOT sessions. The Bland-Altman method of assessing agreement between these two measures was used separately for the three time periods. RESULTS: The relationship between venous and finger-prick glucose at room air was altered significantly by HBOT. The bias (finger-prick minus venous measurements) was significantly less than zero during the HBOT session but not immediately after completion of the session. Owing to the small sample size, the limits of agreement straddled zero at all time points, although the lower limit was close to zero during treatment (finger measurement appeared to be higher than venous measurement on room air and lower than venous undergoing HBOT). CONCLUSION: Finger-prick capillary sampling may not be an accurate reflection of venous glucose during HBOT.