Significance of blood glucose variability in the early phase for predicting survival in acutely ill patients with glucose intolerance

Introduction

Although high blood glucose (BG) causes high mortality in ICU patients, significance of BG variability is not elucidated. The purpose of the study was to clarify the significance of BG variability in the early phase for predicting survival and to suggest the importance of stabilizing BG variability for mortality reduction.

Methods

Strict BG control was performed by bedside-type artificial pancreas (AP), STG22. The number of the selected patients was 67, or patients with daily mean BG (BGm) below 200 mg/dl in early phase (3.2 ± 2.5 days after admission). Studied items: (1) mortality, (2) BG parameters (mg/dl): BGm, daily standard deviation of BG (BGsd), daily maximal and minimal BG (BGmax, BGmin), and daily BG difference (BGd:BGmax - BGmin), (3) accuracy (%) of the BG parameters for predicting survival (AS), (4) relationships among BG parameters, and (5) probable factors affecting BG variability (BGsd) including demographic data, SOFA score and MOF score (calculated using the criteria proposed by the Japanese Association for Critical Care Medicine, 1990).

Results

MQortality of the patients with BGsd below 14 mg/dl (group BGsd14b, n = 41) was significantly lower than that with BGsd above 14 mg/dl (group BGsd14a, n = 26) (17% vs 46%, P < 0.025). (2) AS (%) was as follows: BGm <196 mg/dl (75%), BGsd <17 mg/dl (73%), BGmax <225 mg/dl (72%), BGmin <172 mg/dl (72%), and BGd <8 0 mg/dl (70%). (3) Significant (P < 0.002) positive correlation was found between BGsd and BGd (r = 0.87), BGsd and BGmax (r = 0.79), and BGd and BGmax (r = 0.77). (4) There was a tendency that the ratio of the patients with MOF score above 3 was lower in group BGsd14b than that in group BGsd14a (7% vs 23%, P < 0.10), although there was no significant difference between the groups in BGm, demographic data, and SOFA score.

Conclusions

First, high BG variability in the early phase, which appeared to accompany severe illness and could be evaluated by BGsd, BGd, and BGmax, was considered to be one of the significant risk factors as well as high BG or elevated BGm. Stabilizing BG variability from the early phase was suggested to be important for mortality reduction. Second, AP was reliable for elucidating the significance of strict BG control, and was considered to be a valuable therapeutic artificial organ for the better outcome of the ICU patients through the strict BG control.