|| Checking for direct PDF access through Ovid
To determine if caretakers of young children with IDDM could consistently reproduce small incremental measurements of insulin (U100).Fifteen caretakers of children with IDDM were asked to deliver repeated small doses of insulin, including doses separated by only 0.25 U of insulin. A sensitive gravimetric technique was used to determine the error in measurement of these low doses of insulin. Statistical analysis was used to evaluate accuracy and internal consistency of each caretaker at each dose.The means plus/minus SD at each dose level were as follows: 2.75 plus/minus 0.13 U at 2.5 U, 3.19 plus/minus 0.13 U at 3.0 U, 3.55 plus/minus 0.13 U at 3.25 U, and 3.70 plus/minus 0.11 U at 3.5 U. All doses were biased toward overadministration. There was a statistically significant difference in the dose delivered when the target doses were varied by only 0.25 U. The average differences and standard errors between 2.5 U and 3.0 U, 3.0 U and 3.25 U, and 3.25 U and 3.5 U were 0.44 plus/minus 0.20 U, 0.36 plus/minus 0.018 U, and 0.15 plus/minus 0.017 U, respectively.Participants were not accurate in measuring small insulin doses, consistently overdrawing insulin by an average of 0.22 U. Caretakers are reasonably internally consistent with a given dose, since participants were able to measure statistically significant differences in 0.25 U dose changes. The error in insulin measurement does not vary with the intended dose level. Caretakers in the same family deliver insulin doses as variable from each other as they are from the population as a whole; however, when two or more individuals are responsible for one insulin dose in a child with IDDM, they have a combined variability that is approximately 40 percent greater than a single individual's variability.