Impact of measurement error on predicting population-based inpatient glucose control

Instrument measurement error (ME) may affect ability of damped trend analysis to forecast inpatient glycemic control. A statistical approach was developed to introduce ME into damped trend analysis algorithm. Point-of-care blood glucose device data were extracted from the laboratory system. Forecasts were generated from various inpatient subgroups and time intervals. ME produced differences in damped trend model during the forecast learning cycle. However, forecast trajectory stayed identical regardless of ME in 85% (119/140) of studied scenarios. Forecasts did not change with greater ME. ME inherent in the point-of-care blood glucose device had little effect on trajectory of damped trend exponential forecasts and apparently would not influence decision making in inpatient glycemic control algorithms. High blood glucose (sugar) levels can lead to complications for hospitalized patients, including more surgical infections or longer hospital stays. The ability to forecast glucose control through trend analysis could identify problems sooner and allow earlier care to keep levels in the recommended range. However, measurement error (ME) is inherent in the glucometer used to check point-of-care glucose values and could limit the usefulness of forecasting methods. This study examined how ME affects forecasting. It showed little effect on glucose forecasts and showed potential robustness of trend analysis in assessment of inpatient glucose control.