Skip to main content
. 2015 Jun 9;3:e1005. doi: 10.7717/peerj.1005

Table 1. Multivariable linear regression model to predict retested blood glucose (mg/dl) after initiation of insulin infusion.

Covariatesa Coefficient Standard error Lower limit of 95% CI Upper limit of 95% CI p
Glucose−0.5 34.33 8.81 17.07 51.59 <0.001
Glucose-1.96 94.00 1.60 90.85 97.14 <0.001
(Insulin rate-2.85) −1.06 0.33 −1.70 −0.42 <0.001
(Time interval)2-8.13 −0.002 0.0006 −0.004 −0.001 <0.001
Time interval-0.35 −18.90 2.30 −23.41 −14.39 <0.001
(Dextrose rate)0.5 22.02 8.58 5.20 38.84 0.01
PO volume −0.02 0.01 −0.03 0.00 0.06
TPN volume 0.07 0.03 0.02 0.12 0.01
Lactate (mmol/l) 0.87 0.23 0.41 1.32 0.00
History of congestive heart failure 2.64 1.18 0.33 4.95 0.03
History of renal failure −3.13 1.87 −6.78 0.53 0.09
History of liver disease 4.50 2.03 0.52 8.47 0.03
(Glucose−0.5) × (Insulin rate-2.85) 10.92 4.34 2.41 19.43 0.01
(Glucose-1.96) × (Insulin rate-2.85) 1.60 0.67 0.28 2.91 0.02
Constant 186.52 0.98 184.61 188.44 <0.001

Notes.

Number of obs =4,593, F(14, 4578) = 1787.14, Prob >F = 0.0000, R-squared = 0.8453, Adj R-squared = 0.8449, Root MSE = 30.569.

PO
by mouth, orally (from the Latin “per os,” by mouth)
TPN
total parental nutrition
a

Some covariates were centered and transformed with fractional polynomials.