Effects of computerized decision support systems on blood glucose regulation in critically ill surgical patients

Digital interventions to improve safety and quality of inpatient diabetes management: A systematic review

IMPORTANCE

Diabetes is common amongst hospitalised patients and contributes to increased length of stay and poorer outcomes. Digital transformation, particularly the implementation of electronic medical records (EMRs), is rapidly occurring across the healthcare sector and provides an opportunity to improve the safety and quality of inpatient diabetes care. Alongside this revolution has been a considerable and ongoing evolution of digital interventions to optimise care of inpatients with diabetes including optimisation of EMRs, digital clinical decision support systems (CDSS) and solutions utilising data visibility to allow targeted patient review.

OBJECTIVE

To systematically appraise the recent literature to determine which digitally-enabled interventions including EMR, CDSS and data visibility solutions improve the safety and quality of non-critical care inpatient diabetes management.

METHODS

Pubmed, Embase and Cochrane databases were searched for suitable articles. Selected articles underwent quality assessment and analysis with results grouped by intervention type.

RESULTS

1202 articles were identified with 42 meeting inclusion criteria. Four key interventions were identified; computerised physician order entry (n = 4), clinician decision support systems (n = 21), EMR driven active case finding (data visibility solutions) and targeted patient review (n = 10) and multicomponent system interventions (n = 7). Studies reported on glucometric outcomes, evidence-based medication ordering including medication errors, and patient and user outcomes. An improvement in glucometric measures particularly mean blood glucose and proportion of target range blood glucose levels and rates of evidence-based insulin prescribing were consistently demonstrated.

CONCLUSION

Digitally-enabled interventions utilised to improve quality and safety of inpatient diabetes care were heterogenous in design. The majority of studies across all intervention types reported positive effects for evidence-based prescribing and glucometric outcomes. There was less evidence for digital interventions reducing diabetes medication administration errors or impacting patient outcomes (length of stay).

Review of Intravenous and Subcutaneous Electronic Glucose Management Systems for Inpatient Glycemic Control

PURPOSE OF REVIEW

The goal of this review is to summarize current literature on electronic glucose management systems (eGMS) and discuss their benefits and disadvantages in the inpatient setting.

RECENT FINDINGS

We review different versions of commercially available eGMS: Glucommander™ (Glytec, Greenville, SC), EndoTool^R (MD Scientific LLC, Charlotte, NC), GlucoStabilizer™ (Medical Decision Network, Charlottesville, VA), GlucoCare™ (Pronia Medical Systems, KY), and discuss advantages such as reducing rates of hypoglycemia, hyperglycemia, and glycemic variability. In addition, eCGMs offer a uniform standard of care and may improve workflows across institutions as well reduce barriers. Despite ample literature on intravenous (IV) versions of eGMS, there is little published research on subcutaneous (SQ) insulin guidance. Although use of eGMS requires extensive training and institution-wide adoption, time spent on diabetes management is better facilitated by their use.

Individualizing Glycemic Control in the Critically Ill

Hyperglycemia is a common phenomenon in critically ill patients, even in those without diabetes. Two landmark studies established the benefits of tight glucose control (blood glucose target 80-110 mg/dL) in surgical and medical patients. Since then, literature has consistently demonstrated that both hyperglycemia and hypoglycemia are independently associated with increased morbidity and mortality in a variety of critically ill patients. However, tight glycemic control has subsequently come into question due to risks of hypoglycemia and increased mortality. More recently, strategies targeting euglycemia (blood glucose ≤180 mg/dL) have been associated with improved outcomes, although the risk of hypoglycemia remains. More complex targets (ie, glycemic variability and time within target glucose range) and the impact of individual patient characteristics (ie, diabetic status and prehospital glucose control) have more recently been shown to influence the relationship between glycemic control and outcomes in critically ill patients. Although our understanding has increased, the optimal glycemic target is still unclear and glucose management strategies may require adjustment for individual patient characteristics. As glucose management increases in complexity, we realize that traditional means of using meters and strips and paper insulin titration algorithms are potential limitations to our success. To achieve these complex goals for glycemic control, the use of continuous or near-continuous glucose monitoring combined with computerized insulin titration algorithms may be required. The purpose of this review is to discuss the evidence surrounding the various domains of glycemic control and the emerging data supporting the need for individualized glucose targets in critically ill patients.

A randomized study on the usefulness of an electronic outpatient hypoglycemia risk calculator for clinicians of patients with diabetes in a safety-net institution

Objective: Hypoglycemia (HG) occurs in up to 60% of patients with diabetes mellitus (DM) each year. We assessed a HG alert tool in an electronic health record system, and determined its effect on clinical practice and outcomes.Methods: The tool applied a statistical model, yielding patient-specific information about HG risk. We randomized outpatient primary-care providers (PCPs) to see or not see the alerts. Patients were assigned to study group according to the first PCP seen during four months. We assessed prescriptions, testing, and HG. Variables were compared by multinomial, logistic, or linear model. ClinicalTrials.gov ID: NCT04177147 (registered on 22 November 2019).Results: Patients (N = 3350) visited 123 intervention PCPs; 3395 patients visited 220 control PCPs. Intervention PCPs were shown 18,645 alerts (mean of 152 per PCP). Patients' mean age was 55 years, with 61% female, 49% black, and 49% Medicaid recipients. Mean baseline A1c and body mass index were similar between groups. During follow-up, the number of A1c and glucose tests, and number of new, refilled, changed, or discontinued insulin prescriptions, were highest for patients with highest risk. Per 100 patients on average, the intervention group had fewer sulfonylurea refills (6 vs. 8; p < .05) and outpatient encounters (470 vs. 502; p < .05), though the change in encounters was not significant. Frequency of HG events was unchanged.Conclusions: Informing PCPs about risk of HG led to fewer sulfonylurea refills and visits. Longer-term studies are needed to assess potential for long-term benefits.

Glucose Management Technologies for the Critically Ill

Hyperglycemia is common in the intensive care unit (ICU) both in patients with and without a previous diagnosis of diabetes. The optimal glucose range in the ICU population is still a matter of debate. Given the risk of hypoglycemia associated with intensive insulin therapy, current recommendations include treating hyperglycemia after two consecutive glucose >180 mg/dL with target levels of 140-180 mg/dL for most patients. The optimal method of sampling glucose and delivery of insulin in critically ill patients remains elusive. While point of care glucose meters are not consistently accurate and have to be used with caution, continuous glucose monitoring (CGM) is not standard of care, nor is it generally recommended for inpatient use. Intravenous insulin therapy using paper or electronic protocols remains the preferred approach for critically ill patients. The advent of new technologies, such as electronic glucose management, CGM, and closed-loop systems, promises to improve inpatient glycemic control in the critically ill with lower rates of hypoglycemia.

Application of an artificial neural network model for early outcome prediction of gamma knife radiosurgery in patients with trigeminal neuralgia and determining the relative importance of risk factors

OBJECTIVES

Stereotactic radiosurgery (SRS) is a minimally invasive modality for the treatment of trigeminal neuralgia (TN). Outcome prediction of this modality is very important for proper case selection. The aim of this study was to create artificial neural networks (ANN) to predict the clinical outcomes after gamma knife radiosurgery (GKRS) in patients with TN, based on preoperative clinical factors.

PATIENTS AND METHODS

We used the clinical findings of 155 patients who were underwent GKRS (from March 2000 to march 2015) at Iran Gamma Knife center, Teheran, Iran. Univariate analysis was performed for a long list of risk factors, and those with P-Value < 0.2 were used to create back-propagation ANN models to predict pain reduction and hypoesthesia after GKRS. Pain reduction was defined as BNI score 3a or lower and hypoesthesia was defined as BNI score 3 or 4.

RESULTS

Typical trigeminal neuralgia (TTN) (P-Value = 0.018) and age>65 (P-Value = 0.040) were significantly associated with successful pain reduction and three other variables including radiation dosage >85 (P-Value = 0.098), negative history of diabetes mellitus (P-Value = 0.133) and depression (P-Value = 0.190). On the other hand, radio dosage>85 (P-Value = 0.008) was significantly associated with hypoesthesia, other related risk factors (with p-Value<0.2), were history of multiple sclerosis (P-Value = 0.106), pain duration more than 10 years before GKRS (P-Value = 0.115), history of depression (P-Value = 0.139), history of percutaneous ablative procedures (P-Value = 0.148) and history of diabetes mellitus (P-Value = 0.169).ANN models could predict pain reduction and hypoesthesia with the accuracy of 84.5% and 91.5% respectively. By mutual elimination of each factor in this model we could also evaluate the contribution of each factor in the predictive performance of ANN.

CONCLUSIONS

The findings show that artificial neural networks can predict post operative outcomes in patients who underwent GKRS with a high level of accuracy. Also the contribution of each factor in the prediction of outcomes can be determined using the trained network.

Evaluating the Implementation of the EndoTool Glycemic Control Software System

PURPOSE

The purpose of this study was to compare achievement of glycemic control on insulin drips before and after the implementation of EndoTool, a glucose management software system used in a community hospital setting.

METHODS

A retrospective chart review was performed of patients on an insulin drip who were managed before and after implementation of the EndoTool software. Fifty patients were selected for each group. Statistical analyses were run to compare metrics gathered between groups.

RESULTS

Patients in the standard care group were on an insulin drip for an average of 23.9 hours compared to 20.9 hours in the EndoTool group (P = 0.38). Hypoglycemia occurred at an average rate of 0.036 events per patient in the standard group and 0.007 events per patient in the EndoTool group (P = 0.17). The average rate of hyperglycemia was 0.358 events per patient in the standard group and 0.283 events per patient in the EndoTool group (P = 0.25). The average time to achieve the blood glucose target was 2.78 and 3.67 hours in the standard and EndoTool groups, respectively (P = 0.27). Total patient values were within target range 45.2% of the time in the standard care group and 47.3% of the time in the EndoTool group (P = 0.71).

CONCLUSION

Analysis of the implementation of EndoTool in the community hospital setting found no statistically significant differences between groups, although rates of hypo- and hyperglycemia showed a trend toward improved safety in the EndoTool group. These results could be attributed to the conservative parameters the hospital set in the initial phase of EndoTool implementation.

Risk of Hypoglycemia During Insulin Infusion Directed by Paper Protocol Versus Electronic Glycemic Management System in Critically Ill Patients at a Large Academic Medical Center

BACKGROUND

Insulin infusions are commonly utilized to control hyperglycemia in critically ill patients and decrease hyperglycemia associated complications. Safety concerns have been raised in trials evaluating methods of glycemic control regarding the incidence of hypoglycemia and its relationship to increased mortality. Electronic glycemic management systems (eGMS) may result in less variable blood glucose (BG) control and less hypoglycemia. This study aimed to compare BG control, time in target BG range, and the rate of hypoglycemia when critically ill patients were managed with an insulin infusion guided by paper-based protocol (PBP) versus eGMS.

METHODS

This retrospective review compared critically ill patients ≥ 18 years old that received insulin infusion from March to May 2015 (PBP group) and October to January 2017 (eGMS group). The primary outcome was the incidence of hypoglycemia. Secondary outcomes included frequency and severity of hypoglycemia, duration in glycemic target, length of insulin therapy, as well as ICU and hospital length of stay.

RESULTS

Fifty-four patients were evaluated, 27 in each group. Percentage of days with BG <70 mg/dL was significantly reduced after eGMS implementation (21.5% v 1.3%, P < .0001) including the frequency of severe hypoglycemia (BG < 40 mg/dL) (5.4% v 0.01%, P < .0001). Patients in the eGMS group spent a greater amount of time in target BG range (31.5% v 63.7%, P < .0001).

CONCLUSIONS

An eGMS has the potential to address many of the unmet needs of an optimal glycemic control strategy, minimizing hypoglycemia, and glycemic variability in a heterogeneous critically ill population.

USE OF A COMPUTER-GUIDED GLUCOSE MANAGEMENT SYSTEM TO IMPROVE GLYCEMIC CONTROL AND ADDRESS NATIONAL QUALITY MEASURES: A 7-YEAR, RETROSPECTIVE OBSERVATIONAL STUDY AT A TERTIARY CARE TEACHING HOSPITAL

OBJECTIVE

Inpatient hyperglycemia, hypoglycemia, and glucose variability are associated with increased mortality. The use of an electronic glucose management system (eGMS) to guide intravenous (IV) insulin infusion has been found to significantly improve blood glucose (BG) control. This retrospective observational study evaluated the 7-year (January 2009-December 2015) impact of the EndoTool^® eGMS in intensive and intermediate units at Vidant Medical Center, a 900-bed tertiary teaching hospital.

METHODS

Patients assigned to eGMS had indications for IV insulin infusion, including uncontrolled diabetes, stress hyperglycemia, and/or postoperative BG levels >140 mg/dL. This study evaluated time required to achieve BG control (<180 mg/dL; <140 mg/dL for cardiovascular surgery patients); hypoglycemia incidence (<70 and <40 mg/dL); glucose variability (assessed by SD and coefficient of variation percentage [CV%]); excursions (BG levels >180 mg/dL after control attained); and the impact of eGMS on hospital-acquired condition (HAC)-8 rates.

RESULTS

Data were available for all treated patients (492,078 BG readings from 16,850 patients). With eGMS, BG levels were brought to target within 1.5 to 2.3 hours (4.5 to 4.8 hours for cardiovascular patients). Minimal hypoglycemia was observed (BG values <70 mg/dL, 0.93%; <40 mg/dL, 0.03%), and analysis of variance of BG values <70 mg/dL showed significant reductions over time in hypoglycemia frequency, from 1.04% in 2009 to 0.46% in 2015 (P<.0001). The CV% per patient visit was 26.5 (±12.9)%, and 4% of patients experienced glucose excursions (defined as BG levels >180 mg/dL once control was attained). HAC-8 rates were reduced from 0.083 per 1,000 patients (2008) to 0.032 per 1,000 patients (2011).

CONCLUSION

The use of eGMS resulted in rapid, effective control of inpatient BG levels, including significantly reduced hypoglycemia rates.

ABBREVIATIONS

BG = blood glucose CMS = Centers for Medicare and Medicaid Services CV = coefficient of variation CV% = coefficient of variation percentage eGMS = electronic glucose management system GV = glycemic variability HAC = Hospital-Acquired Condition ICU = intensive care unit IU = intermediate unit IV = intravenous LOS = length of stay VMC = Vidant Medical Center.

Conceptual Framework for Developing a Diabetes Information Network

Objective:

To provide a conceptual framework for managing diabetic patient care, and creating an information network for clinical research.

Background:

A wide range of information technology (IT) based interventions such as distance learning, diabetes registries, personal or electronic health record systems, clinical information systems, and clinical decision support systems have so far been used in supporting diabetic care. Previous studies demonstrated that IT could improve diabetes care at its different aspects. There is however no comprehensive conceptual framework that defines how different IT applications can support diverse aspects of this care. Therefore, a conceptual framework that combines different IT solutions into a wide information network for improving care processes and for research purposes is widely lacking. In this study we describe the theoretical underpin of a big project aiming at building a wide diabetic information network namely DIANET.

Research design and methods:

A literature review and a survey of national programs and existing regulations for diabetes management was conducted in order to define different aspects of diabetic care that should be supported by IT solutions. Both qualitative and quantitative research methods were used in this study. In addition to the results of a previous systematic literature review, two brainstorming and three expert panel sessions were conducted to identify requirements of a comprehensive information technology solution. Based on these inputs, the requirements for creating a diabetes information network were identified and used to create a questionnaire based on 9-point Likert scale. The questionnaire was finalized after removing some items based on calculated content validity ratio and content validity index coefficients. Cronbach’s alpha reliability coefficient was also calculated (αTotal= 0.98, P<0.05, CI=0.95). The final questionnaire was containing 45 items. It was sent to 13 clinicians at two diabetes clinics of endocrine and metabolism research institute in order to assess the necessity level of the requirements for diabetes information network conceptual framework. The questionnaires were returned by 10 clinicians. Each requirement item was labeled as essential, semi-essential, or non-essential based on the mean of its scores.

Results:

All requirement items were identified as essential or semi-essential. Thus, all of them were used to build the conceptual framework. The requirements were allocated into 11 groups each one representing a module in the conceptual framework. Each module was described separately.

Conclusion:

We proposed a conceptual framework for supporting diabetes care and research. Integrating different and heterogeneous clinical information systems of healthcare facilities and creating a comprehensive diabetics data warehouse for research purposes, would be possible by using the DIANET framework.

Predicting outcomes in patients with perforated gastroduodenal ulcers: artificial neural network modelling indicates a highly complex disease

Purpose

Mortality prediction models for patients with perforated peptic ulcer (PPU) have not yielded consistent or highly accurate results. Given the complex nature of this disease, which has many non-linear associations with outcomes, we explored artificial neural networks (ANNs) to predict the complex interactions between the risk factors of PPU and death among patients with this condition.

Methods

ANN modelling using a standard feed-forward, back-propagation neural network with three layers (i.e., an input layer, a hidden layer and an output layer) was used to predict the 30-day mortality of consecutive patients from a population-based cohort undergoing surgery for PPU. A receiver-operating characteristic (ROC) analysis was used to assess model accuracy.

Results

Of the 172 patients, 168 had their data included in the model; the data of 117 (70 %) were used for the training set, and the data of 51 (39 %) were used for the test set. The accuracy, as evaluated by area under the ROC curve (AUC), was best for an inclusive, multifactorial ANN model (AUC 0.90, 95 % CIs 0.85–0.95; p < 0.001). This model outperformed standard predictive scores, including Boey and PULP. The importance of each variable decreased as the number of factors included in the ANN model increased.

Conclusions

The prediction of death was most accurate when using an ANN model with several univariate influences on the outcome. This finding demonstrates that PPU is a highly complex disease for which clinical prognoses are likely difficult. The incorporation of computerised learning systems might enhance clinical judgments to improve decision making and outcome prediction.

Stress-induced hyperglycemia: is it harmful following trauma?

There is an established association between the presence of SIH and worse morbidity and mortality after trauma. However, given the limitations of existing data, no definitive statements can be made as to whether aggressive treatment of hyperglycemia actually benefits outcome. Although early studies seemed to show a clear benefit in surgical ICU patients, subsequent studies have not duplicated these results. In addition, severe hypoglycemic episodes associated with glycemic control protocols have provided further concern, because they have been associated with higher rates of mortality. These disparate outcomes in prospective, randomized trials have not allowed definitive conclusions to be drawn regarding the exact glucose levels that should be maintained. Regardless, some postinjury control of glucose levels is likely necessary. Without data to support the practice, tight glycemic control keeping glucose levels below 110 mg/dL is likely not necessary and probably detrimental to patient outcome. It seems that a more moderate level of glycemic control, aimed at providing stabilization of glucose levels while reducing hyperglycemic and hypoglycemic events, is being practiced in most institutions. Performance of prospective, randomized trials in the trauma population along with further advancement and refinement of techniques to more precisely reduce glucose variability will further clarify the level of glucose control associated with improved outcomes.

Systematic review of clinical decision support interventions with potential for inpatient cost reduction

Background

Healthcare costs are increasing rapidly and at an unsustainable rate in many countries, and inpatient hospitalizations are a significant driver of these costs. Clinical decision support (CDS) represents a promising approach to not only improve care but to reduce costs in the inpatient setting. The purpose of this study was to systematically review trials of CDS interventions with the potential to reduce inpatient costs, so as to identify promising interventions for more widespread implementation and to inform future research in this area.

Methods

To identify relevant studies, MEDLINE was searched up to July 2013. CDS intervention studies with the potential to reduce inpatient healthcare costs were identified through titles and abstracts, and full text articles were reviewed to make a final determination on inclusion. Relevant characteristics of the studies were extracted and summarized.

Results

Following a screening of 7,663 articles, 78 manuscripts were included. 78.2% of studies were controlled before-after studies, and 15.4% were randomized controlled trials. 53.8% of the studies were focused on pharmacotherapy. The majority of manuscripts were published during or after 2008. 70.5% of the studies resulted in statistically and clinically significant improvements in an explicit financial measure or a proxy financial measure. Only 12.8% of the studies directly measured the financial impact of an intervention, whereas the financial impact was inferred in the remainder of studies. Data on cost effectiveness was available for only one study.

Conclusions

Significantly more research is required on the impact of clinical decision support on inpatient costs. In particular, there is a remarkable gap in the availability of cost effectiveness studies required by policy makers and decision makers in healthcare systems.