Dataset Information

Effects of blood glucose transcription mismatches on a computer-based intensive insulin therapy protocol.

ABSTRACT:

Purpose

Computerized clinical decision support systems (CDSS) for intensive insulin therapy (IIT) generate recommendations using blood glucose (BG) values manually transcribed from testing devices to computers, a potential source of error. We quantified the frequency and effect of blood glucose transcription mismatches on IIT protocol performance.

Methods

We examined 38 months of retrospective data for patients treated with CDSS IIT in two intensive care units at one teaching hospital. A manually transcribed BG value not equal to a corresponding device value was deemed mismatched. For mismatches we recalculated CDSS recommendations using device BG values. We compared matched and mismatched data in terms of CDSS alerts, blood glucose variability, and dosing.

Results

Of 189,499 CDSS IIT instances, 5.3% contained mismatched BG values. Mismatched data triggered 93 false alerts and failed to issue 170 alerts for nurses to notify physicians. Four of six BG variability measures differed between matched and mismatched data. Overall insulin dose was greater for matched than mismatched [matched 3.8 (1.6-6.0), median (interquartile range, IQR), versus 3.6 (1.6-5.7); p < 0.001], but recalculated and actual dose were similar. In mismatches preceding hypoglycemia, recalculated insulin dose was significantly lower than actual dose [recalculated 2.7 (0.4-5.0), median (IQR), versus 3.5 (1.4-5.6)]. In mismatches preceding hyperglycemia, recalculated insulin dose was significantly greater than actual dose [recalculated 4.7 (3.3-6.2), median (IQR), versus 3.3 (2.4-4.3); p < 0.001]. Administration of recalculated doses might have prevented blood glucose excursions.

Conclusions

Mismatched blood glucose values can influence CDSS IIT protocol performance.

SUBMITTER: Campion TR

PROVIDER: S-EPMC2916042 | biostudies-literature | 2010 Sep

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Effects of blood glucose transcription mismatches on a computer-based intensive insulin therapy protocol.

Campion Thomas R TR May Addison K AK Waitman Lemuel R LR Ozdas Asli A Gadd Cynthia S CS

Intensive care medicine 20100330 9

<h4>Purpose</h4>Computerized clinical decision support systems (CDSS) for intensive insulin therapy (IIT) generate recommendations using blood glucose (BG) values manually transcribed from testing devices to computers, a potential source of error. We quantified the frequency and effect of blood glucose transcription mismatches on IIT protocol performance.<h4>Methods</h4>We examined 38 months of retrospective data for patients treated with CDSS IIT in two intensive care units at one teaching hosp ...[more]

PMID: 20352190

Similar Datasets

Project description:BackgroundSustained hyperglycemia is a known risk factor for adverse outcomes in critically ill patients. The specific aim was to determine if a nurse initiated insulin infusion protocol (IIP) was effective in maintaining blood glucose values (BG) within a target goal of 100-150 mg/dL across different intensive care units (ICUs) and to describe glycemic control during the 48 hours after protocol discontinuation.MethodsA descriptive, retrospective review of 366 patients having 28,192 blood glucose values in three intensive care units, Surgical Trauma Intensive Care Unit (STICU), Medical (MICU) and Coronary Care Unit (CCU) in a quaternary care hospital was conducted. Patients were > 15 years of age, admitted to STICU (n = 162), MICU (n = 110) or CCU (n = 94) over 8 months; October 2003-June 2004 and who had an initial blood glucose level > 150 mg/dL. We summarized the effectiveness and safety of a nurse initiated IIP, and compared these endpoints among STICU, MICU and CCU patients.ResultsThe median blood glucose values (mg/dL) at initiation of insulin infusion protocol were lower in STICU (188; IQR, 162-217) than in MICU, (201; IQR, 170-268) and CCU (227; IQR, 178-313); p < 0.0001. Mean time to achieving a target glucose level (100-150 mg/dL) was similar between the three units: 4.6 hours in STICU, 4.7 hours in MICU and 4.9 hours in CCU (p = 0.27). Hypoglycemia (BG < 60 mg/dL) occurred in 7% of STICU, 5% of MICU, and 5% of CCU patients (p = 0.85). Protocol violations were uncommon in all three ICUs. Mean blood glucose 48 hours following IIP discontinuation was significantly different for each population: 142 mg/dL in STICU, 167 mg/dL in MICU, and 160 mg/dL in CCU (p < 0.0001).ConclusionThe safety and effectiveness of nurse initiated IIP was similar across different ICUs in our hospital. Marked variability in glucose control after the protocol discontinuation suggests the need for further research regarding glucose control in patients transitioning out of the ICU.

Project description:BackgroundThe absolute risk reduction (ARR) in cardiovascular events from therapy is generally assumed to be proportional to baseline risk-such that high-risk patients benefit most. Yet newer analyses have proposed using randomized trial data to develop models that estimate individual treatment effects. We tested 2 hypotheses: first, that models of individual treatment effects would reveal that benefit from intensive blood pressure therapy is proportional to baseline risk; and second, that a machine learning approach designed to predict heterogeneous treatment effects-the X-learner meta-algorithm-is equivalent to a conventional logistic regression approach.Methods and resultsWe compared conventional logistic regression to the X-learner approach for prediction of 3-year cardiovascular disease event risk reduction from intensive (target systolic blood pressure <120 mm Hg) versus standard (target <140 mm Hg) blood pressure treatment, using individual participant data from the SPRINT (Systolic Blood Pressure Intervention Trial; N=9361) and ACCORD BP (Action to Control Cardiovascular Risk in Diabetes Blood Pressure; N=4733) trials. Each model incorporated 17 covariates, an indicator for treatment arm, and interaction terms between covariates and treatment. Logistic regression had lower C statistic for benefit than the X-learner (0.51 [95% CI, 0.49-0.53] versus 0.60 [95% CI, 0.58-0.63], respectively). Following the logistic regression's recommendation for individualized therapy produced restricted mean time until cardiovascular disease event of 1065.47 days (95% CI, 1061.04-1069.35), while following the X-learner's recommendation improved mean time until cardiovascular disease event to 1068.71 days (95% CI, 1065.42-1072.08). Calibration was worse for logistic regression; it over-estimated ARR attributable to intensive treatment (slope between predicted and observed ARR of 0.73 [95% CI, 0.30-1.14] versus 1.06 [95% CI, 0.74-1.32] for the X-learner, compared with the ideal of 1). Predicted ARRs using logistic regression were generally proportional to baseline pretreatment cardiovascular risk, whereas the X-learner observed-correctly-that individual treatment effects were often not proportional to baseline risk.ConclusionsPredictions for individual treatment effects from trial data reveal that patients may experience ARRs not simply proportional to baseline cardiovascular disease risk. Machine learning methods may improve discrimination and calibration of individualized treatment effect estimates from clinical trial data.Clinical trial registrationURL: https://www.clinicaltrials.gov . Unique identifiers: NCT01206062; NCT00000620.

Project description:IntroductionA single centre has reported that implementation of an intensive insulin protocol, aiming for tight glycaemic control (blood glucose 4.4 to 6.1 mmol/l), resulted in significant reduction in mortality in longer stay medical and surgical critically ill patients. Our aim was to determine the degree to which tight glycaemic control can be maintained using an intensive insulin therapy protocol with computerized decision support and to identify factors that may be associated with the degree of control.MethodsAt a general adult 22-bed intensive care unit, we implemented an intensive insulin therapy protocol in mechanically ventilated patients, aiming for a target glucose range of 4.4 to 6.1 mmol/l. The protocol was integrated into the computerized information management system by way of a decision support program. The time spent in each predefined blood glucose band was estimated, assuming a linear trend between measurements.ResultsFifty consecutive patients were investigated, involving analysis of 7,209 blood glucose samples, over 9,214 hours. The target tight glycaemic control band (4.4 to 6.1 mmol/l) was achieved for a median of 23.1% of the time that patients were receiving intensive insulin therapy. Nearly half of the time (median 48.5%), blood glucose was within the band 6.2 to 7.99 mmol/l. Univariate analysis revealed that body mass index (BMI), Acute Physiology and Chronic Health Evaluation (APACHE) II score and previous diabetes each explained approximately 10% of the variability in tight glycaemic control. BMI and APACHE II score explained most (27%) of the variability in tight glycaemic control in the multivariate analysis, after adjusting for age and previous diabetes.ConclusionUse of the computerized decision supported intensive insulin therapy protocol did result in achievement of tight glycaemic control for a substantial percentage of each patient's stay, although it did deliver 'normoglycaemia' (4.4 to about 8 mmol/l) for nearly 75% of the time. Tight glycaemic control was difficult to achieve in critically ill patients using this protocol. More sophisticated methods such as continuous blood glucose monitoring with automated insulin and glucose infusion adjustment may be a more effective way to achieve tight glycaemic control. Glycaemia in patients with high BMI and APACHE II scores may be more difficult to control using intensive insulin therapy protocols. Trial registration number 05/Q0505/1.

Dataset Information

Effects of blood glucose transcription mismatches on a computer-based intensive insulin therapy protocol.

Purpose

Methods

Results

Conclusions

Publications

Effects of blood glucose transcription mismatches on a computer-based intensive insulin therapy protocol.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets