Abstract
The annual Virtual Hospital Diabetes Meeting was hosted by Diabetes Technology Society on April 1 and April 2, 2022. This meeting brought together experts in diabetes technology to discuss various new developments in the field of managing diabetes in hospitalized patients. Meeting topics included (1) digital health and the hospital, (2) blood glucose targets, (3) software for inpatient diabetes, (4) surgery, (5) transitions, (6) coronavirus disease and diabetes in the hospital, (7) drugs for diabetes, (8) continuous glucose monitoring, (9) quality improvement, (10) diabetes care and educatinon, and (11) uniting people, process, and technology to achieve optimal glycemic management. This meeting covered new technology that will enable better care of people with diabetes if they are hospitalized.
Keywords: continuous glucose monitor, diabetes, electronic health record, hospital, insulin, technology
Introduction
On April 1 and April 2, 2022, Diabetes Technology Society (DTS) invited health care professionals (HCPs), industry representatives, researchers, academicians, and US federal regulatory officials to the annual Virtual Hospital Diabetes Meeting. Clinicians across multiple specialties participated in discussions about improving the management of diabetes care in the hospital using the newest technologies. This two-day meeting included eleven sessions, each of which focused on a current topic and related research in hospital diabetes care. Table 1 presents the meeting agenda with a list of session topics. This meeting report summarizes the key points of each speaker’s presentation.
Table 1.
Agenda of the Meeting with a List of Session Topics.
| Friday, April 1, 2022 |
| Session 1: Digital Health and the Hospital |
| Session 2: Blood Glucose Targets |
| Session 3: Software for Inpatient Diabetes |
| Session 4: Surgery |
| Session 5: Transitions |
| Saturday, April 2, 2022 |
| Session 6: COVID-19 and Diabetes in the Hospital |
| Session 7: Drugs for Diabetes |
| Session 8: Continuous Glucose Monitoring |
| Session 9: Quality Improvement |
| Session 10: Diabetes Care and Education |
| Session 11: Uniting People, Process, and Technology to Achieve Optimal Glycemic Management (Sponsored by Glytec) |
Session 1: Digital Health and the Hospital
Moderator
Signe Schmidt, MD, PhD
Steno Diabetes Center Copenhagen, Denmark
Machine Learning Glycemic Predictions for the Hospital Using Electronic Health Record Data
Nestoras Mathioudakis, MD, MHS
Johns Hopkins University, Baltimore, Maryland, USA
In recent years, machine learning algorithms using large electronic health record (EHR) data sets have been developed and validated for prediction of binary, categorical, or continuous glucose outcomes in hospitalized patients.
When comparing different inpatient-based machine learning algorithms for glycemic prediction, it is important to understand the study population, outcome and exposure variable definitions, prediction horizon, and validation methods used.
Various machine learning methods, including gradient boosting, random forest classification, and logistic regression, have demonstrated strong predictive accuracy for glycemic outcomes when derived from large cohorts and leveraging a large number of predictor variables. Further prospective studies are needed to deploy machine learning models into the EHR system and to evaluate their clinical impact.
Hyperglycemia and hypoglycemia in hospitalized patients are associated with increased morbidity, mortality, and health care expenditures. Inpatient glycemic management can be challenging due to the presence of multiple evolving factors that can influence glucose homeostasis, often in an unpredictable way. There is a growing interest in applying machine learning for the prediction of glycemic outcomes and improving clinical outcomes using EHRs. Machine learning can help in recognizing patterns in large number of patients and identifying predictors of glycemic disturbances at the individual level. One great advantage of machine learning is that it uses the large amount of data already available in the EHR without significant additional cost. Challenges of the approach include confounding by medical interventions, incomplete observations, and selection bias. While several machine learning algorithms have shown high predictive accuracy in the hospital setting,1 -8 most models have not yet been deployed in the EHR and evaluated prospectively with respect to clinical outcomes. The use of machine learning in healthcare is a dynamic process and algorithms need regular updates as new drugs are brought to market.
Order Sets and Other Strategies in the Electronic Health Record to Prevent Harm
Andrew Demidowich, MD
Johns Hopkins University, Baltimore, Maryland, USA
Choice Architecture principles, such as “Opt-in vs Opt-out,” Choice Defaults, Reminders, and Social Influence, can be easily applied to EHRs to “nudge” providers and hospital systems to improve their inpatient diabetes care.
Hiding choices in order sets for certain patients (e.g., low estimated glomerular filtration rate) can prevent harm.
Creating order sets which encompass all aspects of care (e.g., automatic hemoglobin A1c [HbA1c] ordering, carbohydrate restricted diet, endocrine consultation, diabetes education) enables practice standardization and reproducible excellence in care delivery.
The way choices are presented in the EHR influences provider behavior. The aim of choice architecture when applied in the EHR is to nudge—but not to force—HCPs to make the choice that maximizes benefits and minimizes risks and harms to patients. Consequently, by suitable design, the EHR can actively support the work of clinicians rather than just being a passive working tool. This may be particularly relevant for clinicians who treat patients with diseases that are not within their own specific area of expertise. Choice architecture applied to inpatient diabetes management service (IDMS) has been shown to increase compliance with treatment guidelines and reduce hypo- and hyperglycemia rates. Great potential lies in establishing a common library for EHR order sets and concepts for inpatient and outpatient diabetes care.
Customizing the Electronic Health Record for Diabetes
Kensaku Kawamoto, MD, PhD, MHS
University of Utah, Salt Lake City, Utah, USA
EHRs will generally require customization for optimal care of diabetes and other conditions.
“Traditional” EHR customization and optimization is a key strategy but has important limitations, including being restricted to what the EHR allows for customization, and potentially being difficult to share or scale.
Interoperable EHR add-on apps are a promising approach to enable next-generation features in EHRs.
EHRs are valuable and necessary working tools, but unfortunately, they are also the cause of clinician stress and burnout. One reason for this is that suboptimal system configuration results in inefficient workflows, and even simple orders can require a large number of clicks. Too much time is spent at the computer compared to time spent with the patient. Accordingly, there is a need for optimizing EHRs to improve working conditions and further to customize the systems to the specific needs of varying specialist fields, e.g., diabetes management. Customization can be achieved by modifying the EHR itself or alternatively by the use of add-on apps. A multistakeholder initiative was started in 2016 with the goal of improving patient care and the provider experience through standards-based, interoperable EHR apps that convert data to actionable insights. To date, more than 20 solutions have been developed and implemented.
Integration of Continuous Glucose Monitor Data into the Electronic Health Record
Juan Espinoza, MD, FAAP
Children’s Hospital Los Angeles, University of Southern California, Los Angeles, California, USA
HCPs need streamlined, integrated workflows to take advantage of the vast amount of patient data available.
Continuous glucose monitor (CGM) data integration is a complex, multistep process.
The Integration of CGM Data into the EHR (iCoDE) Project is a cross-sector collaboration intended to establish standards and best practices for CGM-EHR data integration.
Few health care organizations have managed to integrate CGM data directly into the EHR. A generalization of the CGM data pipeline is illustrated in Figure 1. Currently, most HCPs need to leave the EHR workflow to review CGM data in a different system separated from other critical patient information, such as test results and medication lists. The disruption of workflows causes errors and inefficiency with a negative impact on patient safety and user satisfaction. Furthermore, the lack of integration between platforms means that the full potential of CGM technology is not exploited because CGM data is less available for analysis than if it was accessible and searchable within the EHR. Integration of CGM data into EHRs is highly needed; however, it is complicated by the large variability among both data providers (CGM manufacturers) and data consumers (health care organizations). To facilitate the process, standards for CGM data classification and integration are currently being developed by the iCoDE consortium.
Figure 1.
Opportunities to adopt, adapt, or develop standards and best practices in the CGM data pipeline. Reproduced from Xu et al. 9 Abbreviations: CCD, Continuity of Care Documents; CDA, Clinical Document Architecture; CGM, continuous glucose monitor; CPT, Current Procedural Terminology; EHR, electronic health record; EMPI, Enterprise Master Patient Index; FHIR, Fast Healthcare Interoperability Resources; HIPAA, Health Insurance Portability and Accountability Act; HITRUST, Health Information Trust Alliance; HL7, Health Level 7; ICD-10, International Classification of Diseases 10th Revision; IEEE, Institute of Electrical and Electronics Engineers; LOINC, Logical Observation Identifiers Names and Codes; mHealth, mobile health; NIST CSF, National Institute of Standards and Technology Cybersecurity Framework; NPI, National Provider Identifier; OAuth, open authorization; OMOP, Observational Medical Outcomes Partnership; RxNORM, a normalized naming system for generic and branded drugs, and a tool for supporting semantic interoperation between drug terminologies and pharmacy knowledge base systems; SMART, Substitutable Medical Applications, Reusable Technologies; SNOMED, Systemized Nomenclature of Medicine; SOC2, System and Organization Controls type 2—Trust Services Criteria; UDI, Unique Device Identifier.
Session 2: Blood Glucose Targets
Moderator
Francisco J. Pasquel, MD, MPH
Emory University, Atlanta, Georgia, USA
Centers for Medicare and Medicaid Glycemic Management Measures: What They Will Do (and What They Won’t Do) to Improve Care Prediction Models for Adverse Glycemia
Gregory A. Maynard, MD, MSc, MHM
University of California, Davis, Sacramento, California, USA
Glucometrics are essential to assess trends in glycemic control, hypoglycemia rates, and quality of care and safety locally, to benchmark performance against other institutions, and to inform improvement efforts.
Standardization of metrics has been elusive, but widely accepted principles have been published and have been adapted by the Centers for Medicare and Medicaid (CMS) and the Centers for Disease Control and Prevention (CDC).
The CMS electronic clinical quality measures (eCQMs) will include measures of hypo- and hyperglycemia and will likely spur interest in inpatient glycemic control efforts. Future CDC’s National Health Safety Network (NHSN) metrics will offer more granularity, a wider variety of measures, risk stratification, and benchmarking capability.
Benefits of measuring glucometrics include the ability to assess baseline performance, track progress over time, compare performance of like units to each other, assess tradeoffs between hyperglycemia and hypoglycemia, benchmark one’s own hospital against other hospitals, and conduct real-time surveillance. The purpose of benchmarking in health care is to improve the quality of patient care, patient safety, and patient satisfaction by assessing performance against standards and best practices and then identifying opportunities for improvement. Society for Hospital Medicine has promoted benchmarking and provided the Glycemic Control Implementation Guide at no charge. The CMS has introduced eCQMs into their quality reporting programs to automate the capture, calculation, and reporting of quality measures (see Figure 2). Pay-for-reporting programs are characterized using financial incentives for providers that report data on certain predefined metrics. Calendar year 2023 will likely begin with a data collection period for the glycemic eCQMs. In fiscal year 2025, hospitals will be required to report performance in various eCQMs or face penalties. They will have a choice of metrics. In the future, failure to meet targets could result in decreased payment by CMS, as evolution to Pay for Performance occurs. CDC’s NHSN is a health care–associated tracking system. This program is currently working to establish a standard for submitting inpatient hypoglycemia data electronically to facilitate benchmarking of inpatient diabetes-related data, including measures of hypo- and hyperglycemia. The frequency of inpatient hypoglycemia will be a particularly important metric for documenting patient safety and quality.
Figure 2.
The eCQM strategy recommendations. Reproduced with permission from Centers for Medicare & Medicaid Services. 10 Abbreviations: API, application programming interface; CMS, Centers for Medicare and Medicaid; eCQM, electronic clinical quality measures; EHR, electronic health record; FHIR, Fast Healthcare Interoperability Resources.
Intensive Care Unit Targets for Glycemia
Jean-Charles Preiser, MD, PhD
Erasme University Hospital, Brussels, Belgium
Stress-related hyperglycemia resulting from decreased insulin sensitivity is a component of the stress response and could reflect a transiently adaptive mechanism.
The use of a single blood glucose target (BGT) to titrate insulin therapy in critically ill patients hospitalized in an intensive care unit (ICU) has been hard to achieve and found unable to provide a consistent benefit in large heterogeneous populations.
Individualized BGT defined according to the time elapsed from admission and a prior diabetic status appear as reasonable options to prevent the effects of hyper- and hypoglycemia.
The causes of inpatient stress hyperglycemia have patient-related, illness-related, and treatment-related causes. After any insult, insulin resistance occurs and is nonspecific. Stress hyperglycemia reflects the magnitude of insulin resistance. Regarding insulin resistance, this phenomenon is seen as a component of the stress response, a reflection of the ability to use the macronutrients by reflecting the balance of catabolism/anabolism and can be considered as adaptive survival during the acute phase of critical illness. A glycemic ratio of maximal observed glucose to admission glucose appears to be strongly and independently associated with mortality in critically ill patients. Optimal target glycemia depends on the stage of critical illness. There is a J-shaped curve between mean blood glucose (BG) and inpatient mortality with the optimal mean being around 80 to 100 mg/dL in patients without diabetes and around 100 to 130 mg/dL in patients with diabetes. The concept of targeting estimated average BG is therefore appealing but not easy to perform. In the controlling study, targeting an ICU patient’s preadmission usual glycemia using a dynamic sliding-scale insulin protocol did not demonstrate a survival benefit compared with maintaining glycemia below 180 mg/dL, but it increased the rate of hypoglycemia. 11
How to Reach Glycemic Targets in the Real World
Andjela T. Drincic, MD, FACP
University of Nebraska Medical Center, Omaha, Nebraska, USA
Key elements of glycemic control in the hospital include a glucose management program, hospital administrative (organizational) support, information technology (IT) support, clearly defined processes, external incentives, glucometrics (and insulinometrics), and innovative models of care to integrate these elements.
Inpatient glycemic management requires real-time and retrospective glucometrics.
Benchmarking, tracking progress over time, and using systems enhancement are helpful practices for achieving goals.
Most organizational roadmaps talk about basic organizational elements needed for success: people, process, passion metrics. Here is how we proceeded at the 728-bed University of Nebraska Medical Center. Our plan was to (1) optimize glucose therapy for all hospitalized patients with diabetes or hyperglycemia by monitoring of glucometric data, ongoing analysis, and implementation of quality improvement processes, and (2) promote and implement compliance with all accreditation and regulatory standards. We monitored glucose metrics and insulinometrics including active surveillance, retrospective review, and benchmarking analyses. Initially, we used homegrown metrics. Later, we switched to metrics recommended by the Society for Hospital Medicine. We used electronic surveillance of each patient dashboard to monitor for glycemic outcomes. Some of the problems (and remedies) we addressed via root cause analysis had included recurrent hypoglycemia (education of the nursing/provider staff to enhance communication), insulin drip problems (an intravenous insulin titration class), and hyperkalemia treatment–associated hypoglycemia (evaluation of the current order panel). Some examples of innovative models of care include Diabetes Resource Nurse Program (that was shown to decrease readmission rates) and Diabetes Pharmacy Stewardship Program (that contributed to improvement in all metrics, especially hypoglycemia). This type of approach can elevate a hospital to become a top performer according to Society for Hospital Medicine benchmarks.
Accuracy of Blood Glucose Monitoring Systems in the Hospital
James H. Nichols, PhD, DABCC, FAACC
Vanderbilt University Medical Center, Nashville, Tennessee, USA
Critically ill patients challenge the analytical performance of blood glucose monitors (BGMs), particularly capillary samples from patients with poor peripheral perfusion. Clinicians should consider arterial or venous samples instead of capillary finger sticks in patients with poor perfusion as these samples better reflect physiologic central circulation and can be analyzed by BGM.
The current generation of BGM systems do not meet the 95% US Food and Drug Administration (FDA) and Clinical Laboratory Improvement Amendments (CLIA) guidance criteria for accuracy in all samples collected from hospitalized patients (neonates, pediatrics, and adults) when compared with a laboratory method.
Although blood gas analyzers have been suggested as an alternative method to BGM in critically ill patients, there are fewer blood gas analyzers in the hospital setting, which limits access to rapid glucose testing. Also, the method is CLIA moderate complexity, which limits the staff who can perform testing. In our study, while blood gas glucose met International Organization for Standardization (ISO) 15197:2013 criteria for accuracy, the analyzers fell short of the FDA criteria, similar to current BGM systems on the market.
Critically ill patients with diabetes and its complications like diabetic ketoacidosis (DKA) can become very sick quickly, and early diagnosis and intervention are absolutely essential for a good outcome. This involves good clinical history, physical examination, and the appropriate assessment of laboratory values. Hospitalized patients present with extremes of physiology that challenge BGM performance. Critically ill patients are not just in the ICU. They can be in the emergency department, operating room, cardiac, burn units, and even general medical floors. Hematocrit, oxygen therapy, drugs (e.g., high doses of ascorbate, intravenous N-acetylcysteine, maltose in intravenous solutions, sugars like galactose), and metabolites (e.g., high triglycerides, sugars like galactose) can falsely increase or decrease glucose test results. Capillary samples are common for BGM testing, even in hospital settings where patients receive intensive medical interventions and therapy. However, collection of capillary blood from patients with poor peripheral circulation may not reflect central circulation, and arterial or venous blood might be physiologically better specimens when perfusion is impaired. Peripheral vascular disease can cause poor perfusion as well as conditions such as severe dehydration, DKA, hyperglycemic hyperosmolar nonketotic syndrome, hypotension, shock, congestive heart failure, and with the use of vasoconstricting medications. In a study of the accuracy of blood gas analyzers, we found agreement met ISO (>95%) criteria but did not meet the FDA 2016 criteria with 94.3% of results ≥75 mg/dL and 96.2% of results <75 mg/dL. 12 Blood glucose monitors cleared by the FDA do not necessarily perform to the accuracy requirements specified by the FDA after they are in use. 13 Despite not meeting all of the FDA accuracy guidance requirements, the first point-of-care (POC) capillary BGM for critically ill patients was cleared by the FDA in 2018. I believe that this decision will open the door for other POC BGMs to be cleared for this purpose in the future, even if they do not meet all the published FDA accuracy criteria. In a recent multicenter evaluation of a novel BGM, the product met the FDA accuracy requirements in most performance categories and demonstrated a low risk of potential insulin dosing errors. 14 In conclusion, accurate measurement of hospitalized critically ill patients is challenging, and for physiological and analytical reasons, no testing method is accurate to the FDA specifications all the time.
Individuals with diabetes have a higher lifetime likelihood for hospitalization and hospital readmission. A survey of individuals with diabetes following hospital discharge observed that 21% had difficulty obtaining diabetes medicines and supplies after discharge. Only 53% had received instructions on treating hypoglycemia and 81% had received written instructions regarding the insulin regimen. A Patient Comprehension Questionnaire administered after discharge noted that 37% of individuals with diabetes had misconceptions about their insulin regimen, their glucose monitoring, and hypoglycemia management. Individuals who scored lower on the questionnaire were more likely to be get readmitted to hospital. These findings support efforts to improve the hospital discharge process.
Session 3: Software for Inpatient Diabetes
Moderators
Andjela T. Drincic, MD, FACP
University of Nebraska Medical Center, Omaha, Nebraska, USA
Gerlies M. Treiber, MD
Medical University of Graz, Graz, Austria
Homegrown Insulin Dosing Software
Kristen M. Kulasa, MD
University of California, San Diego, La Jolla, California, USA
Studies support that computer-based titration of intravenous insulin provides higher staff satisfaction, better compliance with protocols, more time with glucose in target range, and less variability with more standardization.
Benefits of homegrown insulin calculators in the hospital include the ability to completely customize the algorithm and workflows, with no required external integration and low-cost scalability. Challenges include the need for significant dedicated IT resources, including the time required to build and maintain the calculator as well as analyst time and metrics to test and maintain the calculator.
Key components of implementation and operationalization of an insulin calculator include early establishment of a multidisciplinary team, identification of superusers, streamlined workflows, and early and repeated education.
In general, achieving blood targets between 140 and 180 mg/dL or more stringent goals (110-140 mg/dL) in hospital settings must be balanced with the risk of hypoglycemia. Homegrown computerized algorithms were created before commercial products became available to reach BGTs. At University of California San Diego Health, an insulin infusion computer calculator (IICC) was implemented and transitioned from a web-based platform directly into the electronic medication administration record (eMAR). The algorithm used in this IICC is an insulin-sensitive coefficient (ISC)-based algorithm using a BG value and rate of change to adjust the coefficient. Manual adjustments of the ISC were required with interruption in nutrition, prolonged hyperglycemia, and highly insulin-sensitive patients (type 1 diabetes, end-stage renal disease). The system is capable of achieving excellent glucometrics and works well for high-risk patients including obstetrics, renal failure, and those with hyperglycemic emergencies. The system also has a transition calculator for those getting off the insulin drip. Significant benefits of the homegrown system include a relatively low cost and the ability to customize it and adopt it across multiple satellite locations. Implementing the IICC algorithm requires careful planning, coordination with all stakeholders, educational support for rollout and maintenance, appointed superusers, and repeated education and training along the way. Adequate IT support is of paramount importance. In addition, system downtimes and upgrades can potentially impact any programming. At University of California, San Diego Health, an IICC that is safe and effective in a wide variety of clinical situations was successfully implemented and integrated directly into the eMAR of the EHR.
Commercial Insulin Dosing Software
Joseph A. Aloi, MD
Wake Forest Baptist Health, Winston-Salem, North Carolina, USA
Barriers to implementing a commercial insulin dosing software include the need for IT build, integrating external software into organization’s EHR, and presenting a compelling case for return on investment. The organization needs to balance the culture change required for implementation while building a cohesive working group and bridging the education from bedside caregiver to ordering providers.
Intravenous insulin dosing managed by commercial insulin doing software reduces hypoglycemia rates and ICU length of stay compared with metrics achieved with previously used institution algorithms.
Advantages of commercial basal-bolus insulin dosing software are yet to be determined.
The journey for inpatient glucose management in our institution started with the initiation of a glucose management committee, the development of a glucometrics dashboard, and the introduction of insulin dosing calculator. Barriers to implementation included little or no specialist diabetes training, culture where confidence is higher than the actual knowledge in inpatient glucose targets, physician’s preferences to avoid insulin drips, and cost considerations for licenses and staff training. Several computer-based algorithms aiming to direct the nursing staff adjusting the insulin infusion rate have become commercially available, namely, Glucommander, EndoTool, and GlucoStabilizer. The common feature among these electronic glucose management systems (eGMSs) is their ability to limit occurrences of hypoglycemia while achieving and maintaining patients at the target BG level. The internal review at Sentara Healthcare showed that after introduction of Glucommander intravenous protocols, the hypoglycemia rate dropped to <1% and time in range (TIR) 70-180 mg/dL increased to 83.50% in comparison with the previously used intravenous insulin protocols (see Figure 3). A 10-fold decrease in hypoglycemia rates (0.19%-0.32%) was reported by Atrium Health with EndoTool in comparison with standard practice. Rates of inpatient hypoglycemia events are considered an indicator of the quality of care provided by a hospital and eGMS support glycemic management initiatives. The system also helped decrease length of stay in DKA. One of the significant barriers in adoption is related to the cost of the system, limited not only to the licensing, but also to the inherent cost of time needed to educate the staff on proper use.
Figure 3.
The continuous glucose monitor (CGM) data is aggregated from Glytec’s de-identified data pool. Figure is courtesy of Joseph A. Aloi. Abbreviation: GM, Glucommander.
Commercial Insulin Dosing Software
Matthew F. Bouchonville, MD, CDCES
The University of New Mexico, Albuquerque, New Mexico, USA
Our institution uses a number of tools to achieve glycemic control in the hospital setting, including situational awareness dashboards, single-page views, and electronic glycemic control software.
Commercial electronic glycemic control software uses FDA-cleared proprietary algorithms for intravenous and subcutaneous insulin management.
Adaption, implementation, and maintenance of electronic glycemic control software requires continuous education.
The University of New Mexico’s journey for inpatient glucose management prior to adoption of eGMS included multiple qualitative initiatives including situational awareness dashboards and insulin single-page views. It has evolved with the transition to eGMS, embracing a process of continuous learning. Situational awareness dashboards allow us to quickly identify patients that might need an intervention for glycemic control. Insulin single-page view pulls all the information pertinent for optimal insulin dosing including details regarding nutrition, renal function, glucocorticoids, and so on. Commercial electronic glycemic control has a software with FDA-cleared proprietary algorithms for intravenous and subcutaneous insulin management and allows for individualized dosing recommendations. It can be integrated with the EHR and connected device systems. The company provides analytics and surveillance, and most of these have shown improvements in clinical and financial outcomes when compared with conventional means of glycemic control in the inpatient setting. Benefits of eGMS include improved glycemic metrics with reduced hypoglycemia and hyperglycemia rates, regulating a transition from intravenous to subcutaneous insulin, and attaining or sustaining glycemic goals. In addition, a standardization of practices, decrease of cognitive burden, and decrease of hospital stays and costs are related to eGMS. However, to minimize challenges with change to new systems, fear of automation, and increased nursing burden, key activities include continuous training and education of staff (MDs, nurses, pharmacists) and data analysis. It is important to recognize that there is no silver bullet when it comes to inpatient glucose control. The providers and nurses will need to understand what these systems can and cannot do and, when clinicians need to intervene, to override the automated insulin adjustment to make a necessary change in insulin dosing. Finally, it is worth remembering that intravenous insulin and subcutaneous insulin protocols or algorithms are very different, but, once implemented, it is important to consider the need for ongoing maintenance of these systems. The software cannot replace professionals that are needed to help monitor and mitigate the system.
Information Technology Issues with Building Home Grown Insulin Dosing Software
Robert J. Rushakoff, MD
University of California, San Francisco, San Francisco, California, USA
Patients receiving total parenteral nutrition (TPN) or enteral tube feedings (TFs) or nothing by mouth (NPO) may receive fixed doses of intermediate acting insulin or short-acting nutritional insulin plus correction insulin (with or without basal insulin). Achieving appropriate glucose control may be delayed as timely adjustments in insulin doses are rarely made to titrate to the needs of the individual patient.
We developed and programmed a self-adjusting subcutaneous insulin algorithm (SQIA) in the EHR for patients who are NPO, on TPN, or on TFs. The SQIA only requires a nurse to enter a patient’s current glucose value and then calculates the next insulin dose based on previous doses and current and previous glucoses. The advantage of this automated algorithm is that no new orders are required, even if insulin is added to TPN or tube feeding rate is changed.
Programming this algorithm into the EHR was far from simple and has taken a decade with the help of many nurses, physicians, pharmacists, and programmers.
Developing and implementing an automated SQIA for patients who are NPO or on TFs or TPN is an IT-tech adventure and was “a 30-year odyssey.” In the beginning, not enough computers were available in the ICUs. Therefore, a paper-based insulin algorithm was introduced using newer rapid insulins every 4 hours. A major challenge was the implementation of paper orders in the EHR system. With the integration of all calculators into the Medication Administration Record, more programming issues had to be solved: back interval timing, more than one glucose value, missed insulin dose, and a limited number of lines of code. Full implementation was possible in 2020 in synchronization with EHR updates, other major initiatives, and nurse education. The limitation of the automated titration protocol was noted early on, as there was no process prompting the nurse to adjust insulin titration scale for interruption in nutrition. In 2021, the optimization of the calculator involved adding the nursing attestation button prompting them to acknowledge nutrition status, allowing for calculation of the new dose. This has resulted in excellent glucometrics, low rate of hypoglycemia, and importantly, a decreased physician workload, since only one order for insulin triggering a calculator is needed (as opposed to daily changes in basal-nutritional insulin that are needed with traditional regimens). We are continuing with ongoing optimization of new workflows and dealing with new programming issues, like getting all the data accurately, TF documentation problems, and still limits on lines of programming.
Intellectual Property Considerations for Homegrown Insulin Dosing Software
Raymond R. Moser, MSEE, JD
Moser Taboada, Shrewsbury, New Jersey, USA
Intellectual property (IP) laws, such as trade secret, patent, and copyright, are important tools for developers of Homegrown Dosing Software to use to protect their software from unauthorized use.
Before using or releasing Homegrown Dosing Software, a developer must ensure the software does not infringe the IP of a third party or face severe financial consequences.
Before using Homegrown Dosing Software, users (doctors and/or patients) must consider the IP rights of the developer or face potential financial liability.
The technology has the following components: glucose monitor, insulin pump, and glucose control software. Dosing softwares include simple dose calculators, historically driven tracking, dosing computer, homegrown closed-loop pump control, and commercial insulin dosing softwares. The IP protection covering insulin dosing software has 3 forms: trade secrets, patent protection, and copyright protection. Trade secrets come up on a contractual basis, and their goal is to keep the confidential material secret, as illustrated by nondisclosure agreement. Patents are involved in some cases, mostly in commercial software. Copyright protection is the main form of IP protection for homegrown software. Copyright owners can give others the right to copy their work in form of a license. Copyrights are automatically created but the registration increases the protection of the copyright. IP stakeholders are developers (third party, patients, and HCPs) and users (patients, providers). Developers’ IP concerns include those regarding IP procurement (protecting and appropriately using their product covered by trade secrets, patents, and copyrights) and IP infringement (not infringing on other peoples’ protected IP to avoid liability). Users’ IP concerns relate to patents (look for indemnification from IP owner, license covers patents) and copyrights (the right to use the software). Open-source software is generally free of charge but must attribute licenses appropriately that have limited terms. For a developer or user with any concern about IP infringement, it is strongly recommended to consult an attorney.
Session 4: Surgery
Moderator
M. Cecilia Lansang, MD, MPH
Cleveland Clinic, Cleveland, Ohio, USA
Preoperative Assessment of Diabetes
Kathleen M. Dungan, MD, MPH
The Ohio State University, Columbus, Ohio, USA
Pre-anesthesia evaluation of patients with diabetes should include a thorough diabetes history, cardiovascular risk assessment, glucose-lowering medication use and behaviors, and an assessment of glucose control, including hypoglycemia risk.
There is insufficient evidence to recommend routine preoperative diabetes screening, but this could be considered in accordance with general population screening or for high-risk groups (joint replacement or vascular surgery).
There is insufficient evidence to determine an upper threshold of HbA1c for which elective surgery should be delayed though this may be considered using an individualized approach.
Postoperative complications can be seen more commonly in patients with uncontrolled diabetes, compared with those with controlled diabetes or no diabetes, and in persons with a diagnosis of diabetes versus no diabetes. A comparison of outcomes by diabetes status and level of control is presented in Figure 4. However, perioperative glucose also correlates with worse postoperative outcomes in persons without known diabetes compared with those with diabetes and may be more important for predicting complications than HbA1c. The question then arises as to whether it is HbA1c or perioperative glucose that is more relevant. Fructosamine can help guide the success of short-term treatment changes.
Figure 4.
Multifactorial causes of hospital-related hyperglycemia. Causal factors are specific to the patient, their illness, and their treatment. Hyperglycemia can exacerbate some illness-specific factors and increase the need for treatment-specific factors, leading to a vicious cycle by which hyperglycemia causes further hyperglycemia. Reproduced with permission from Dungan et al. 15 Abbreviation: HPA, hypothalamic-pituitary-adrenal axis.
In light of the above findings, societies recommend screening perioperatively for diabetes in high-risk groups (vascular or joint surgery), in accordance with the general population, or otherwise have not recommended to do so. In the same vein, some societies do not recommend an HbA1c test by which to delay surgery or recommend to use an individualized approach, whereas others have recommended to delay at an HbA1c if it is ≥8.5% or ≥9%.
Preoperative Management and Enhanced Recovery After Surgery
Marie E. McDonnell, MD
Brigham and Women’s Hospital, Boston, Massachusetts, USA
Use of Enhanced Recovery After Surgery (ERAS) protocols is growing and now widespread in medical centers around the world after several, mostly observational, studies showed benefit. The ERAS strategies to attenuate what is described as “the surgical stress response” include carbohydrate loading, which can lead to preoperative hyperglycemia.
The fundamental goal of ERAS carbohydrate drinks is to reduce the insulin resistance that has been demonstrated to occur in both starvation and rigorous physical stress. However, these mechanisms do not apply as clearly in individuals with diabetes. Based on the ERAS rationale, preoperative carbohydrate drinks should lead to better perioperative glycemic control, not worse, which is what the limited data available show.
Given an overall lack of prospective data showing benefit of preoperative carbohydrate drinks in people with diabetes, ERAS programs should be tailored so that people with diabetes may benefit from the overall intervention without the risk of perioperative hyperglycemia. Collaborations between ERAS proponents and endocrinologists and/or diabetes-focused clinicians are recommended to design protocols to minimize carbohydrate exposure to those at risk of hyperglycemia.
ERAS is based on the concept that surgery is an iatrogenic injury. Part of ERAS relates to nutrition, with the premise that perioperative nutrition screening and carbohydrate treatment can minimize muscle catabolism, reduce insulin resistance, and optimize recovery. If there is minimal risk of aspiration, then patients can take in solids for up to 8 hours before anesthesia and clear fluids up to 2 hours before anesthesia, with a preoperative drink of at least 50 g carbohydrates.
ERAS has been shown to reduce the length of hospital stay in a meta-analysis. However, studies on patients with diabetes do not clearly show this outcome, and administration of a carbohydrate drink can lead to higher levels of hyperglycemia. In addition, diabetes is a state of chronic insulin resistance, and acute hyperglycemia worsens the first phase of insulin response. There has been a call to put a moratorium on applying the individual ERAS element of carbohydrate drinks to persons with diabetes.
Managing Orthopedic Surgery Patients
Ruben Diaz, DNP, FNP-BC, BC-ADM, CDCES
Memorial Sloan Kettering Cancer Center, New York City, New York, USA
Measuring HbA1c prior to orthopedic surgery and serum fructosamine, both as an initial adjunctive marker and utilization with BG self-monitoring results, for patient selection and optimization in elective orthopedic surgery can mitigate risk.
Recognizing the impact of orthopedic surgery on patients with diabetes improve glycemic and metabolic control, and considering underlying comorbidities in the intraoperative care of this patient population can enhance outcomes.
Post orthopedic surgery for the patient with diabetes includes using insulin as the primary pharmacologic agent immediately postoperative. Dosing considerations should include steroid administration and fluctuating dietary patterns.
Patients with diabetes are disproportionately affected with orthopedic conditions—the microvascular and macrovascular changes in diabetes increase the risk of falls and fractures. Surgical-site infection rates are higher with hyperglycemia. At Hospital for Special Surgery, elective orthopedic surgery is postponed at an HbA1c of ≥8%. Insulin or sulfonylurea or meglitinides are started for optimization, followed by weekly calls for adjustment and lifestyle modification support based on self-monitoring BG results. Fructosamine can be helpful when an HbA1c might reveal inaccurate results, e.g., with anemia, blood transfusions, and the use of agents that induce hemolysis. BG is monitored every 1 to 2 hours intraoperatively, and intravenous insulin infusion, or subcutaneous insulin dosed every 2 hours, is used to maintain levels of 140-200 mg/dL. Prolonged pronation for posterior spine procedures might result in postoperative optic nerve damage; elevating the head above heart level can mitigate this risk. Postoperatively, they favor use of insulin for at least the first 24 to 48 hours with mixed use of other diabetes medications afterward.
Foot Surgery
David G. Armstrong, DPM, MD, PhD
University of Southern California, Los Angeles, California, USA
Every second, someone around the world develops a limb-threatening diabetic foot ulcer; 20% of these lead to hospitalization and another 20% lead to some level of amputation.
Technologies now allow the ability to monitor key aspects of tissue repair and healing.
Technologies predict re-ulceration and restenosis and potentially act to reduce acute-on-chronic events as well as increase ulcer-free, hospital-free, and activity-rich days.
In recent years, there have been fewer amputations in patients with diabetes, but there has been a higher incidence of diabetic foot ulcers. This is an acute event superimposed on a chronic condition, and the aim is to get these patients out of the hospital, back to their homes, and to their lives. The coronavirus disease (COVID-19) pandemic has paved the way toward a model of a wound center without walls, where this may even be the patient’s home. Wound care has also moved to theragnostics, not just diagnostics. There are now methods to detect bacterial load and local flow; analytes can be measured in intelligent textiles and dressings. To prevent recurrence, activity, balance, and pressure can be sensed, and activity can be dosed. Patients can send in photos of their wounds; intelligent bathmats and smart socks can measure foot temperature at hotspots. Injectable sensors can detect analytes such as glucose and oxygenation.
Session 5: Transitions
Moderator
Umesh Masharani, MB, BS
University of California, San Francisco, San Francisco, California, USA
Prediction Models for Adverse Glycemia
Mervyn Kyi, MB, BS, FRACP, PhD
Royal Melbourne Hospital, The University of Melbourne, Parkville, Victoria, Australia
In hospitalized patients, both hyperglycemia and hypoglycemia are associated with adverse pathophysiology and adverse clinical outcomes, hence the composite of both glycemic extremes could be considered as adverse glycemia.
A clinical prediction model using clinical factors available early in the admission course (admission glucose levels, HbA1c, glucose-lowering regimen, and glucocorticoid treatment) can predict development of persistent adverse glycemia during hospitalization.
Prediction tools for adverse glycemia could assist glycemic management teams, providing earlier and targeted management rather than responding to glycemic surveillance alone.
Currently, treating medical and surgical teams request a consult from the inpatient diabetes teams when individuals in the hospital have hyperglycemia or hypoglycemia. This is a reactive model of inpatient diabetes care. An alternative approach is a proactive model of inpatient diabetes care using the EHR to identify early in the hospital stay individuals with high and low glucose levels and allow diabetes care teams to intervene early. There is evidence that early intervention reduces hyperglycemia and hospital-acquired infections. This proactive model applies bedside management to all individuals with diabetes admitted to the hospital. A refinement of this proactive model is to target interventions only at individuals with diabetes who are at specific risk for hyperglycemia or hypoglycemia (dysglycemia). The Melbourne Clinical Prediction Tool is a validated instrument that uses admission glucose (<4 or >15 mmol/L, this equals <72 mg/dL or >180 mg/dL); sulfonylurea or insulin treatment prior to admission; HbA1c <7.1%, 7.1% to 8%, >8.1%; and glucocorticoid treatment to classify individuals with diabetes in the hospital into low, intermediate, and high risk for dysglycemia. It remains to be shown that the use of the tool improves glycemic and clinical outcomes.
Acute and Chronic Glucose Control in Critically Ill Patients with Diabetes: Impact of Prior Insulin Treatment
James S. Krinsley, MD, FCCM, FCCP
Stamford Hospital, Stamford, Connecticut, USA; Columbia University Vagelos College of Physicians and Surgeons, New York City, New York, USA
Preadmission glycemia, reflected by HbA1c, affects the interaction of glucose control during ICU admission and mortality. For the entire group of patients with HbA1c <6.5%, higher mean BG during ICU admission is strongly associated with increased risk of death for diabetes patients treated with oral agents before admission but decreased risk of death for diabetes patients treated with insulin. In contrast, for patients with HbA1c >8.0%, higher mean BG during ICU admission is strongly associated with decreased risk of death for both groups of patients.
There is a different relationship of glucose metrics during ICU admission to mortality when comparing diabetes patients treated with insulin as outpatients compared with diabetes patients treated with oral agents alone before hospitalization.
Current guidelines endorse a single BGT for all critically ill patients. These data support the need for randomized control trials (RCT) that test a strategy of individualized BGTs in the critically ill, based on preadmission characteristics.
The Leuven 1 study in the surgical ICU reported that starting insulin therapy in ICU patients with BG levels greater than 6.1 mmol/L (110 mg/dL) reduced mortality and morbidity. The Leuven 2 study of medical ICU patients reported that aggressive treatment of hyperglycemia reduced morbidity but not mortality. In both of these studies, only a small percentage of individuals had a diagnosis of diabetes at admission (13% and 17%, respectively). These findings of the Leuven studies, however, were not confirmed by other prospective studies. The NICE-SUGAR study recruited 6104 surgical and medical ICU patients with hyperglycemia (20% had diabetes). The tight control group achieved BG levels of 6.4 ± 1.0 mmol/L (115 mg/dL) and the conventional group 8 ± 1.3 mmol/L (144 mg/dL). There were more deaths (829 versus 751 deaths) in the tight glucose control group compared with the less tight glucose control group. A post hoc analysis of the Leuven studies suggested that the benefit of intensive insulin therapy was observed only in those individuals with hyperglycemia without a preexisting diagnosis of diabetes. There is also evidence from other studies that individuals with diabetes who have “poor” control prior to hospital admission do better with “loose” control, whereas those with “good” preadmission glucose control do better with “tight” control. Figure 5 includes retrospective data from 5567 medical and surgical ICU patients that demonstrate the relationship between mean ICU BG and mortality, stratified by HbA1c level. For patients with HbA1c <6.5%, higher mean ICU BG is associated with higher mortality, and for patients with HbA1c >8.0%, higher mean ICU BG is associated with lower mortality. In conclusion, comorbidities, insulin treatment, and HbA1c level prior to admission may help determine the glycemic targets in the critically ill individuals with diabetes.
Figure 5.
Relationship between mean blood glucose (BG) (mg/dL) and mortality, stratified by hemoglobin A1c (HbA1c) level. For patients with HbA1c less than 6.5%, higher mean BG is strongly associated with increased mortality. For patients with HbA1c greater than 8.0%, the opposite relationship is observed. Reproduced with permission from Krinsley et al. 16
Transitions to Discharge for Hospitalized Patients With Diabetes
Mary T. Korytkowski, MD
University of Pittsburgh, Pittsburgh, Pennsylvania, USA
The American Diabetes Association (ADA) recommends that patient understanding and knowledge of diabetes self-care be assessed at the time of hospital discharge, but it is not known how often this is done.
A Patient Comprehension Questionnaire administered within 48 hours following discharge to patients with diabetes hospitalized with noncritical illness identified a need for corrective information in 47 (37%) of 128 patients participating in one study. Major knowledge areas requiring clarification included insulin dosing, hypoglycemia detection and treatment, and glucose monitoring.
Patients who were more knowledgeable about their diabetes home management had a lower risk of hospital readmission at 30 days, supporting recommendations that patient understanding of home diabetes management be addressed prior to discharge.
Individuals with diabetes have a higher lifetime likelihood for hospitalization and hospital readmission. There are several factors affecting the risk of readmissions for hospitalized diabetes patients including low health literacy, ability to obtain prescribed medications, and poor understanding of their underlying illness and treatment. A survey of individuals with diabetes reported that 21% of respondents had difficulty obtaining diabetes medicines and supplies after discharge. Only 53% had received instructions on detecting or treating hypoglycemia and 81% had received written instructions regarding their home insulin regimen. A Patient Comprehension Questionnaire administered after discharge noted that 37% of individuals with diabetes had misconceptions about their insulin regimen, their glucose monitoring, and hypoglycemia management. Individuals who scored lower on the questionnaire were more likely to be get readmitted to hospital. These findings support efforts to improve the hospital discharge process.
Telehealth to Prevent Readmissions
Daniel J. Rubin, MD, MSc, FACE
Lewis Katz School of Medicine at Temple University, Philadelphia, Pennsylvania, USA
Telehealth probably does not reduce readmission risk as a stand-alone intervention for people with diabetes.
Telehealth may reduce readmission risk in the context of multicomponent interventions.
The effect of telehealth on HbA1c after hospital discharge is unclear.
Telehealth can provide effective diabetes management as measured by HbA1c, patient satisfaction, and diabetes-related distress. Readmission risk, however, is complex, and the evidence for readmission prevention is mixed and limited. In a systematic review, 5 of the 10 included studies reduced readmission risk within 30 days of discharge using peri-discharge support, education, care coordination, medication assessment, and/or home visits. The effective interventions, however, were not substantially different from the ineffective interventions. While telehealth probably does not reduce readmission risk as a stand-alone intervention, it may reduce risk in the context of multicomponent interventions. The effect of telehealth on HbA1c after discharge is unclear. By lowering acute care utilization, telehealth probably reduces postdischarge costs. Selecting higher risk patients with a tool like the Diabetes Early Re-admission Risk Indicator may optimize the benefits of intervention. Adequately powered RCTs are needed to further develop telehealth interventions and better understand their risks, benefits, reproducibility, and scalability.
Session 6: COVID-19 and Diabetes in the Hospital
Moderator
Roma Gianchandani, MD
Cedars Sinai Medical Center, Los Angeles, California, USA
Epidemiology Risks of Diabetes and Stress Hyperglycemia
Francisco J. Pasquel, MD, MPH
Emory University, Atlanta, Georgia, USA
Both diabetes and stress hyperglycemia are associated with severity of disease in patients with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection; a “glycemic gap” measure accounting for the degree of stress hyperglycemia in relation to underlying glycemia (as measured by HbA1c) may be a helpful marker of disease severity.
The relationships between diabetes/stress hyperglycemia and COVID-19 appear to be bidirectional. While diabetes may predispose patients to worse disease, there is suggestive evidence for a diabetogenic effect of SARS-CoV-2 on susceptible individuals.
The role of antidiabetic medications on disease severity in patients with COVID-19 is unclear, with confounding by indication limiting the interpretation of potential benefits of some drugs. Thus far, there is no clear evidence indicating a need to change prescriptions of antidiabetic agents in people with COVID-19.
The relationship between COVID-19 and diabetes is complex. Early in the pandemic, patients with diabetes and COVID-19 were noted to have a twofold to threefold higher disease severity with poor outcomes, especially when BG levels were more than 250 mg/dL for 2 to 3 days after admission. Stress hormones and inflammation were confounded by hospital nutrition, medications, and hyperglycemia treatment with insulin or oral agents. Outcomes for diabetes patients with COVID-19 are presented in Figure 6.
Figure 6.
Unadjusted in-hospital coronavirus disease 2019 (COVID-19) mortality rates, March 1 to May 11, 2020, by diabetes status. Error bars show 95% confidence intervals. Data for age groups 0 to 39 years and 40 to 49 years for type 1 diabetes and 0 to 39 years and 50 to 59 years for no diabetes have been excluded because of small numbers of events (one to four), to comply with data protection regulations. Reproduced with permission from Barron et al. 17
In COVID-19 infections, the metabolic phenotype of each patient contributed to severity of hyperglycemia. Those with prediabetes or poorly controlled diabetes had a higher incidence of severe hyperglycemia, DKA, and hyperosmolar state, especially if above 65 years, and an increased mortality, especially if above 65 years. Social determinants of health, glycemic control, and medications used prior to admission were other identified contributing factors.
Recent US Department of Veterans Affairs (VA) data suggest that patients with past COVID-19 infection compared with the general population are at a 40% higher risk of developing diabetes 12 months later. This translates to 14 patients per 1000 with magnified risk due to hospitalization, ICU stay, or predisposing factors for diabetes. Mechanisms for the greater susceptibility are varied and range from the virus’ affinity for the beta cell to persistent chronic inflammation. In addition, the more risk factors for diabetes one has, the higher their likelihood of developing diabetes after COVID-19.
DARE-19, an RCT randomizing patients to dapagliflozin with COVID-19, did show some improvement in outcomes with its use, although they were not statistically significant. The trial highlighted the safe use of dapagliflozin in a sick patient population as far as renal injury but did lead to more cases of DKA. A UK meta-analysis evaluating preinfection diabetes medication and mortality in 2.8 million patients highlighted the safety of metformin. Dipeptidyl peptidase-4 (DPP-4) inhibitors had a mixed consensus with earlier studies being protective. Being on insulin in the hospital was associated with poor outcomes, a relationship reflecting the severity of hyperglycemia. The data gave no clear indication to change the diabetes medications for COVID-19 infections.
In conclusion, diabetes and hyperglycemia were associated with poor outcomes in patients with COVID-19. There is emerging data on the risk of developing diabetes after COVID-19 infection, and these associations and relationships are still being worked out.
Adaptation of Inpatient Diabetes Mellitus Management During a Pandemic: From Research to Practice
Amisha Wallia, MD, MS
Northwestern University Feinberg School of Medicine, Chicago, Illinois, USA
Adaptations to inpatient and transitional diabetes care management (specifically using telemonitoring, CGM, and telemedicine) facilitated optimal care management during the pandemic.
It will be necessary to demonstrate optimal strategies for continued use of these adaptations for improving glycemic control in the hospital and transitional care setting and for delivering to special populations.
Utilization of novel technology to achieve optimal outcomes in high-risk populations (e.g., those with hyperglycemia, steroid-induced hyperglycemia, and diabetes) will be critical to the future of care.
Access to diabetes care and data was a challenge during the COVID-19 pandemic. Two major changes that addressed lack of data and access included CGM use in the hospital and telemedicine. Government regulation and policy changes granted Emergency Use Authorization for CGM in the hospital to help preserve personal protective equipment, reduce nursing contact, and save time. Translating this technology to the hospital required innovation and training of inpatient staff. Inpatient accuracy and outcomes data for CGM were initially from small studies but the size of study populations has expanded with COVID-19 use. No significant improvement was noted in glycemic control, but several reports reduced hypoglycemia. Work is still needed in this area including streamlined CGM implementation, data integration with the EHR, and ideally integration of insulin dosing with CGM data for hospitals.
The second change was in reimbursement for telemedicine/health. Endocrinology practices were at the forefront of virtual platform adoption. Both synchronous and asynchronous teleconsultation are shown to have (1) significant clinical value (reduction in HbA1c, systolic and diastolic blood pressure, fasting and postprandial glucose, and weight), (2) cost reduction, (3) convenience, and (4) a positive impact on quality-of-life measures. Glycemic control differed with the type of diabetes and in differed regions of the world. For type 1 diabetes, several studies reported improved control in Europe especially for patients using diabetes technology. For type 2 diabetes, there were conflicting reports with studies reporting weight gain and loss of control during the lockdown. There is a divide in use of diabetes technology in racial-ethnic cohorts, noted especially in type 1 diabetes. Therefore, opportunities for improving this are needed.
Unfortunately, the quick evolution of telemedicine into normal clinical care has added a new dimension of burden and burnout among providers. The workflow and processes need to be streamlined with focus on user-centered design and advocacy for provider support of this platform.
Mucormycosis
Viswanathan Mohan, MD, PhD, DSc
Dr. Mohan’s Diabetes Specialities Centre, Chennai, India
Mucormycosis is a severe, life-threatening fungal infection and its subtypes include (1) rhino-orbital-cerebral mucormycosis, (2) pulmonary, (3) cutaneous, (4) gastrointestinal, and (5) disseminated.
People with uncontrolled diabetes are at increased risk of mucormycosis and the latter can be prevented and managed by good control of diabetes.
The reason for the surge of mucormycosis in India during the second wave of COVID-19 is unclear but might in part relate to COVID-19 therapies (including indiscriminate use of glucocorticoids) that were used at the time.
Mucormycosis infections spiraled in the second wave of the COVID-19 pandemic in India with thousands of patients being affected. Surprisingly, this was not noted in any other country. Mucormycosis is a devastating opportunistic fungal infection, and the most prominent form seen was rhino-orbital-cerebral (nose, orbit, and the brain). Other forms noted include rhino-orbito-cerebral, nasal, rhino-palatal, and pan-sinusitis. Patients present with redness of nose and eyes, fever, headaches, vomiting, and altered mental status (which is an ominous finding). Examination can reveal eye inflammation, restricted eye movement, proptosis, black eschar on the nose and palate, and a cherry red spot on the fundus examination. The fungus can be isolated by placing a nasal or mouth swab in KOH preparation or sending for fungal culture. High-resolution computed tomographic (CT) scan can help demonstrate spread of infection. As an angio-invasive organism, mucormycosis can lead to tissue necrosis and cause disfigurement. It also has a high mortality rate. The factors that increased spread of mucormycotis during the Indian pandemic were multifold and included 85-90% of infected patients having diabetes, overuse of steroids, prolonged hospital stays (especially in the ICU), patients with transplant, human immunodeficiency virus infection and acquired immunodeficiency syndrome, and chemotherapy. Other causes were the reuse of oxygen tanks and nasal devices without complete sterilization.
Amphotericin B, the drug of choice, was in short supply because of the staggering number of cases. It has high nephrotoxicity, but the lipophilic variety has fewer adverse events, and that was also in short supply. Posaconazole is used for follow-up treatment. Several patients required surgical debridement and orbital enucleation. Referrals to otolaryngology, ophthalmology, and neurosurgery were staggering. BGM and BG control were identified as important contributors to success, and a large hyperglycemia management campaign was launched throughout the country, which helped curtail infection rates. Patients needed basal-bolus insulin especially when on steroids, and intravenous insulin was used for those with severe hyperglycemia. Luckily, within 2 months, the cases of mucormycosis rapidly came down to pre-COVID-19 levels.
Session 7: Drugs for Diabetes
Moderator
Viral N. Shah, MD
Barbara Davis Center for Diabetes, University of Colorado, Aurora, Colorado, USA
Noninsulin Agents (Besides Sodium-Glucose Cotransporter-2 Inhibitors and Checkpoint Inhibitors)
Guillermo E. Umpierrez, MD
Emory University, Atlanta, Georgia, USA
Individualized diabetes management is needed in non-ICU settings.
Despite guidelines recommendations, not all patients need basal or basal-bolus insulin regimens in the hospital setting.
Continuing oral agents plus correction with sliding-scale insulin are indicated in the absence of contraindications for most insulin naïve type 2 diabetes patients with mild/moderate hyperglycemia.
Although basal and bolus insulin therapy is considered the standard way for managing diabetes in hospital settings, not everybody would need insulin therapy. A basal-bolus regimen is associated with hypoglycemia risk and a potential for overtreatment in some patients. In a retrospective study at Emory University, about 24% of patients were treated with oral noninsulin agents and outcomes were similar or better with oral agents compared with insulin therapy. In a multicenter, open-label RCT, there was no difference in mean glucose, glucose between 70-140 mg/dL, and glucose >200 mg/dL between patients who received sitagliptin alone (n = 30), sitagliptin plus glargine, or basal-bolus regimen. These findings suggest that oral therapy with DPP-4 may be safely used in noncritical settings. 18 Studies have shown that use of sitagliptin or liraglutide after hospital discharge was associated with better HbA1c concentrations and lower rates of hypoglycemia.19,20 Treatment regimens for non-ICU hospitalized patients with type 2 diabetes are presented in Figure 7.
Figure 7.
Personalized treatment in non-intensive care unit (non-ICU) hospitalized patients with type 2 diabetes. Regimen complexity refers to the number and type of agents (oral agents, glucagon-like peptide-1 [GLP-1] receptor agonist, and insulin) used in the outpatient setting, with more complex regimens referring to those including multiple agents and/or insulin therapy. SSI refers to use of correctional sliding-scale insulin. Patients on multiple agents are likely to have worsening hyperglycemia if all preadmission agents are stopped and may respond better to basal + OAD or a basal-bolus approach. Reproduced with permission from Galindo et al. 21 Abbreviations: HbA1c, hemoglobin A1c; OAD, oral antidiabetic drug; SSI, sliding-scale insulin.
Sodium-Glucose Cotransporter-2 Inhibitors in the Inpatient Setting
Theocharis Koufakis, MD, PhD
AHEPA University Hospital, Thessaloniki, Greece
The rationale behind the use of sodium-glucose cotransporter 2 (SGLT-2) inhibitors in the inpatient setting is based on the low risk of hypoglycemia, the practical dosing scheme, and the potential to rapidly improve cardiorenal outcomes, particularly those related to heart failure.
Safety issues should be considered, including the risk of euglycemic DKA and volume depletion, especially when used concomitantly with high-dose diuretics.
The concept of administering these agents in the acute phase of cardiovascular episodes appears promising. Future RCTs and real-world data are anticipated to provide clear guidance for clinical practice.
SGLT-2 inhibitors are shown to ameliorate organ damage caused by heart failure, chronic kidney disease, and cardiovascular diseases, and use of these agents may lead to improved survival. However, they are associated with an increased risk for euglycemic DKA. 22 In a large RCT, sotagliflozin in hospitalized patients with diabetes and recently worsening heart failure resulted in a significant lower total number of deaths from cardiovascular causes. 23 There was no difference in the DKA incidence between the two groups. 23 Similarly, an RCT using empagliflozin in patients hospitalized for heart failure resulted in significant improved clinical outcomes at 90 days. 24 SGLT-2 inhibitors were found to be safe in non–critically ill hospitalized patients with diabetes and COVID-19. 25 These data suggest that SGLT-2 inhibitors may be used in non–critically ill settings in properly selected patients with diabetes.
Immune Checkpoint Inhibitors Diabetes Mellitus
Meng H. Tan, MD
University of Michigan, Ann Arbor, Michigan, USA
Immune checkpoint inhibitors (ICIs) have been used as immunotherapy drugs to treat many cancers with promising results but have many immune-related adverse events (irAEs), including the endocrine system. The ICI-associated autoimmune diabetes mellitus (ICI-ADM) is a rare (~1%) endocrine irAE.
The clinical onset of ICI-ADM is often acute (fulminant in some, especially in Asian individuals), and about two-thirds present with DKA. About 50% have comorbid irAE with thyroid irAE as the most common.
There are published Clinical Practice Guidelines to manage irAEs, including ICI-ADM. Education of patients and their primary care physicians of the presenting symptoms of ICI-ADM can facilitate early diagnosis.
During the past decade, ICIs such as cytotoxic T-lymphocyte-associated protein 4 (CTLA-4), and programmed cell death protein-1 (PD-1)/programmed cell death ligand-1 (PDL-1) have been used as immunotherapy drugs to treat many cancers with promising results. 26 However, these classes of drugs can cause many irAEs in many body systems, including the endocrine system. ICI-ADM is a rare (~1%) endocrine irAE. 27 Cancer patients treated with ICIs can have hyperglycemia. The differential diagnosis of this hyperglycemia includes aggravation of preexisting type 2 diabetes mellitus (DM), new-onset type 2 DM, steroid-induced DM, ICI-ADM, ICI-related pancreatitis, and ICI-related lipodystrophy.
The clinical onset of ICI-ADM is often acute (fulminant in some, especially in Asian individuals). About two-thirds present with DKA. Most have low C-peptide levels at onset. The median onset is ~9 weeks after ICI initiation; the onset can also occur after ICI is stopped. About 50% have comorbid irAE with thyroid irAE being the most common. 27 Not all ICIs have the same impact on ICI-ADM, as most cases are caused by anti-PD-1 and few are caused by anti-CTLA-4. Pancreatic enzymes are elevated in 51% of cases; pancreatic atrophy is seen on imaging.
Many national oncology and endocrine specialty societies have published Clinical Practice Guidelines to manage irAEs, including ICI-ADM. The approach is to grade the severity (mild, moderate, and severe) of the irAE, and then manage accordingly. “Cross-talk” between endocrinologists and oncologists is strongly encouraged. Education of patients and their primary care physicians of the presenting symptoms of ICI-ADM can facilitate early diagnosis.
Treatment of Diabetic Ketoacidosis and Hyperglycemic Hyperosmolar State
Ketan Dhatariya, MBBS, MSc, MD, MS, FRCP, PhD
Norfolk and Norwich University Hospitals, Norwich, Norfolk, England; Norwich Medical School, University of East Anglia, Norwich, England
The management of DKA seems straightforward, but several controversies still exist.
There are several differences between US guidelines and other guidelines for DKA and hyperglycemic hyperosmolar state (HHS), including the need to use the “D,” the “K” and the “A” to make a diagnosis and, for HHS, how resolution is defined.
US hyperglycemic emergency guidance needs to be updated because of technological advances that enable bedside ketone meters to help manage the condition and the recognition of the importance of euglycemic DKA.
HHS is a diabetic emergency that causes high morbidity and mortality. Its treatment differs in the United Kingdom (UK) and the United States (US). The diagnosis of DKA is based on disease severity in the US, which differs from the UK. 28 The diagnosis of HHS in the US is based on total rather than effective osmolality. Unlike the US, the UK has separate guidelines for DKA and HHS. Treatment of DKA and HHS also differs with respect to timing of fluid and insulin initiation. Bedside ketone measurement can be useful in prompt measurement of ketones to diagnose and manage DKA.29,30 There are several issues with the 2009 ADA consensus guidelines on hyperglycemic emergencies that need updating. For example, there is no clear acknowledgment on euglycemic DKA. 31 There is a need for the ADA guidelines to be updated.
Diabetes Drugs to Prevent Hospitalizations for Acute Cardiovascular Events
Vanita R. Aroda, MD
Brigham and Women’s Hospital, Boston, Massachusetts, USA
The lifetime risk of cardiovascular disease (CVD) in patients with type 2 diabetes is high and even higher with obesity, and the risk reflects a multitude of etiologies.
Glucagon-like peptide-1 (GLP-1) receptor agonists and SGLT-2 inhibitors are two unique classes of glucose-lowering agents that reduce the risk of major adverse cardiovascular events.
Among patients with type 2 diabetes with a high risk of atherosclerotic cardiovascular disease, established kidney disease, or heart failure, it is appropriate to include evidence-supported cardio/renal-protective agents in the treatment regimen to decrease the risk of major adverse cardiovascular events.
The lifetime risk of CVD among individuals with diabetes is high (55%-87%), and this relationship is further accentuated with increasing adiposity. 32 The burden of complications is increasing, driven by prevalent diabetes and the changing demographics of diabetes. In 2008, the FDA issued a guidance for the industry for evaluating cardiovascular risk in new antidiabetic therapies to treat type 2 diabetes. GLP-1 receptor agonist is shown to reduce major CVD events by 14%, cardiovascular deaths by 13%, and fatal or nonfatal myocardial infarction and stroke by 10% and 17%, respectively. 33 SGLT-2 inhibitors are shown to reduce major CVD events by 10%, cardiovascular deaths by 15%, and major reduction in hospitalization for heart failure by 32% 3. SGLT-2 inhibitors had no major effect on stroke. 34 GLP-1 receptor agonist and SGLT-2 inhibitors appear to have complementary, nonoverlapping physiologic benefits toward mitigating cardiorenal metabolic risk in patients with type 2 diabetes. This new evidence led to changes in ADA clinical guidance, and now, GLP-1 receptor agonist and SGLT-2 inhibitors are included as first tier drugs in patients with type 2 diabetes and established CVD, heart failure, and chronic kidney disease.
Session 8: Continuous Glucose Monitoring
Moderator
Elias K. Spanakis, MD
University of Maryland, Baltimore, Maryland, USA
Establishing a Continuous Glucose Monitor Program
Athena Philis-Tsimikas, MD
Scripps Whittier Diabetes Institute, San Diego, California, USA
Continuous glucose monitoring is useful in the hospital environment because uncontrolled BG levels contribute to higher morbidity, mortality, and health care costs. There are multiple contributors to poor inpatient glycemic control.
It is necessary to assemble a team to establish a CGM hospital program. Considerations for use of CGM in hospital include accuracy, efficacy, safety, patient satisfaction, efficiency, and cost. Team members should include an oversight leader, an on-site CGM support team, diabetes educators, and diabetes clinical nurse specialists.
Key drivers to gain approvals to establish a CGM hospital program require presentations to (1) lead hospital administrators and medical officers to review financial forecasts of costs and potential offsets, (2) justify placement of CGM in place of point of care (POC) for compliance and legal officers, and (3) educate physicians and nursing leaders to explain the methods of CGM use and benefits.
Continuous glucose monitoring can be an effective method of detecting hypoglycemia and hyperglycemia in the hospital. 35 Establishing an inpatient CGM program requires several steps. At Scripps, a CGM team is used, which includes a manager, an advanced practice nurse, and an on-site CGM placement team. This team receives a daily patient identification report, which includes high-risk patients with abnormal glucose values, and places CGM devices on these selected individuals. By using smart devices, glucose data is transmitted to a centralized location, where an advanced practice nurse reviews them remotely and provides recommendations for treating hyperglycemia and hypoglycemia. A bedside nurse also reviews CGM glucose values at the smart device and enters them into the EHR. Validation of CGM devices is performed with POC glucose testing at the time of insertion and daily, following the 20%/20 mg/dL guidelines. A centralized team monitors 24/7 for established hypoglycemia alarms and alerts the bedside nurse. Standardized glucose management algorithms for treating both hypoglycemia and hyperglycemia are used. 36 There are several key drivers to gain approval of CGM devices in the hospital. 37 Physicians and nurses need to be confident about the accuracy, efficacy, and safety of the CGMs. Capturing reductions in POC testing and staff time can convince hospital administrators and medical officers of the cost-effectiveness of CGM intervention, which may lead to improved outcomes and patient satisfaction.
Protocols and Hospital Administration
Eileen R. Faulds, PhD, MS, RN, FNP-BC, CDCES
The Ohio State University, Columbus, Ohio, USA
Health systems have demonstrated that hybrid protocols combining CGM with POC glucose testing can create a safe and effective bridge toward eventual development and approval of full nonadjunctive use of CGM systems, particularly for ICU settings.
Protocol description and evaluation within CGM research will aid future implementation. Clinical outcomes, implementation outcomes, and protocol fidelity should be reported to help health systems compare the efficacy of different protocols and understand their impact on adherence and safety.
Future inpatient CGM research should incorporate implementation outcomes and strategies, along with a robust protocol description, to help health systems as hybrid POC + CGM systems are used more widely.
Following COVID-19 declaration, many hospitals used hybrid protocols, especially in the ICU, which used POC glucose and CGM values combined for glucose evaluation.38 -43 Hybrid protocols have many benefits as they permit continued validation of CGM devices while facilitating reduced POC testing and mitigating any safety concerns. The Ohio State University hybrid protocol 38 requires initial validation with two POC-sensor pairs one hour apart using a 20/20 guideline. Once initial validation is achieved, the CGM is used nonadjunctively to monitor glucose and titrate intravenous insulin with ongoing POC testing, and validation performed every 6 hours. If the patient is transitioned to multiple daily injection insulin regimens while in the ICU, the CGM is used adjunctively. However, once the patient is transferred to the medical surgical floor or progressive care unit, the CGM may be used nonadjunctively without POC testing or validation.
Other hybrid protocols39 -43 reported a robust initial validation period, using a 20/20 rule for nonadjunctive use, and some of them used conservative low- and high-alarm thresholds. Ongoing validation differs among the different protocols, ranging from only a few measurements per day to hourly. Protocols differ also about the level of involvement of the health care workers.
Prevention of Hyperglycemia
Rodolfo J. Galindo, MD, FACE
Emory University, Atlanta, Georgia, USA
The current standard of care for glycemic monitoring in the hospital relies on capillary glucose testing before meals and before bedtime. While easy and widely implemented, this approach cannot detect nor prevent asymptomatic hypoglycemia and hyperglycemia events or nocturnal hypoglycemia.
The CGM provides a comprehensive assessment of glucose excursion for more than 24 hours, including asymptomatic or nocturnal events. Real-time CGM can provide patterns and glucose trends. The CGM has been shown to allow for early detection and prevention of hypoglycemic events.
Most of the current evidence on the use of CGM in the hospital is derived from the utilization and implementation of CGM during the COVID-19 pandemic. However, research on the validation of the accuracy and effectiveness of CGM in the hospital is still insufficient, particularly on its ability to improve hyperglycemia.
Over the last 100 years, there has been significant progress in glucose monitoring, from qualitative analysis of glucose in the urine to BG meters and more recently to CGM devices. 44 The POC glucose testing has many limitations because it provides little information about daily glucose control, 45 and the accuracy of many of the available BG meters is not ideal. 46 A comparison of hypoglycemia detection using POC BGM and CGM is presented in Figure 8. A recently published consensus statement shows the significant interest in using real-time continuous glucose monitors (RT-CGMs) in the hospital, 37 and systems like glucose telemetry can be used to achieve remote diabetes glucose management in the hospital. 47 The RT-CGM devices have been extensively used among patients with COVID-19 in the ICU setting.40,41,48 Few studies have examined whether RT-CGM devices can prevent hyperglycemia in the hospital setting. 36 In a study that included 110 type 2 diabetes patients admitted in the non-ICU setting, RT-CGM devices led to a reduction in mean CGM glucose values, compared with standard of care group, 219.51 (43.75) mg/dL versus 238.05 (45.26) mg/dL, P = .0311, and percent hyperglycemia >250 mg/dL, 27.00 [16.01-40.97] versus 32.96 [20.40-68.75] mg/dL, P = .0403. A statistically significant improvement in TIR 70-250 mg/dL was achieved (72.83 [59.03-83.57]% versus 63.95 [31.25-77.95]%, P = .0404) and also a nonstatistically significant increase in TIR 70-180 mg/dL (25.31 [11.78-42.97]% versus 19.89 [3.34-40.09]%, P = .1460). No differences in hypoglycemia or glucose variability were seen between the 2 groups.
Figure 8.
Hypoglycemia detection by POC (filled bars) glucose testing and a FreeStyle Libre Pro CGM (open bars) in 134 insulin-treated hospitalized patients. Blood glucose monitoring was performed before meals and at bedtime or as clinically required. Reproduced with permission from Galindo et al. 45 Abbreviations: CGM, continuous glucose monitor; POC, point of care.
Prevention of Hypoglycemia
Elias K. Spanakis, MD
University of Maryland, Baltimore, Maryland, USA
Hypoglycemia in the hospital is associated with adverse clinical outcomes.
The RT-CGM/glucose telemetry systems can be used as a safety net to prevent and decrease hypoglycemia in the hospital.
The RT-CGM devices will eventually be used in the hospital and will replace POC glucose testing.
Several studies have shown that CGM devices can have many benefits when used in the hospital. 49 Among them is reducing hypoglycemia, 50 which is associated with adverse clinical outcomes.51 -53 In an early pilot study, RT-CGM devices and glucose telemetry were effective in providing remote glucose monitoring, with the potential to prevent inpatient hypoglycemia. 47 In a larger study that included patients with insulin-treated type 2 diabetes who were at higher risk for inpatient hypoglycemia, RT-CGM/glucose telemetry, compared with POC glucose testing, was more effective in reducing hypoglycemia in the hospital setting. 50 Compared with POC glucose testing, glucose telemetry reduced hypoglycemic events <70 mg/dL per patient (0.67 versus 1.69, P = .024) and <54 mg/dL (0.08 versus 0.75 events/patient, P = .003). Time below range (TBR)% <70 mg/dL and TBR <54/mg/dL were also reduced (0.40% versus 1.88%, P = .002 for TBR <70 mg/dL and 0.05% versus 0.82%, P = .017 for TBR <54 mg/dL). In addition, this study showed that by using low CGM alarms at 85 mg/dL, which led nurses to proceed with hypoglycemia prevention actions, no worsening of hyperglycemia was caused. Additional future studies will evaluate whether RT-CGM devices can improve glucose control in the hospital (NCT 03877068) or whether RT-CGM devices can serve as a “safety net,” achieving tighter glycemic control (90-130 mg/dL) without increasing hypoglycemia (NCT 05135676).
Session 9: Quality Improvement
Moderator
Nestoras Mathioudakis, MD, MHS
Johns Hopkins University, Baltimore, Maryland, USA
Hospital Quality Metrics
Gregory H. Gilbert, MD
Mills-Peninsula Medical Center, Burlingame, California, USA
Sutter Health administration observes metrics regarding glucose control at 19 hospitals within their health system.
The incidence of both hypoglycemia and hyperglycemia is observed using multiple metrics derived from the Society of Hospital Medicine Glycemic Control Electronic Quality Improvement Programs (eQUIPS).
Quality is based on a rolling 12-month evaluation of selected metrics based on national benchmarks.
Glucometrics are tracked by health systems to gauge the safety and quality of glycemic control for hospitals and may be used for internal and external benchmarking. Sutter Health system in California tracks glycemic control measures among 19 hospitals. These metrics, which are stratified by ICU and non-ICU patients, include percent of days with a day-weighted mean BG ≥180 mg/dL (hyperglycemia), days with BG >299 mg/dL (severe hypoglycemia), and days with BG <70 mg/dL (hypoglycemia). Data is summarized on a rolling 12-month basis, calendar year to date, and current month, allowing executive leadership to evaluate overall health system performance of time, compare individual hospital performance, assess impact of quality improvement interventions, and identify hospitals in need of additional resources or training. The metrics are aligned with the Society of Hospital Medicine Glycemic Control reporting, which provides an opportunity to benchmark performance against other health systems. In addition, length of stay and readmission rates are reported among patients with and without diabetes.
A Hospitalist Diabetes Care Program
Mihail Zilbermint, MD, MBA, FACE
Johns Hopkins University School of Medicine, Baltimore, Maryland, USA
Community hospitals are struggling to improve diabetes care for hospitalized patients, while community-based endocrinologists are rarely coming to hospitals to see new consultations.
An endocrine hospitalist is an endocrinology-trained physician or advanced practice provider whose subspecialty interest is in the management of inpatients with diabetes and various endocrine disorders and who does not have major outpatient responsibilities.
Endocrine hospitalists may improve inpatient glycemia and help hospitals to prepare to report CMS hospital harm measures on severe hypoglycemia and hyperglycemia clinical quality measures.
Endocrine hospitalists are endocrinology-trained physicians or advanced practice providers who are focused exclusively on inpatient endocrine management. 54 Given the shortage of endocrinologists nationally, endocrine hospitalists bring value to hospitals by providing timely and expert endocrine consultative care, while also working to improve systems barriers that may result in suboptimal endocrine and diabetes care. In collaboration with pharmacists, nurses, dieticians, diabetes care and education specialists, and advanced practice providers, the endocrine hospitalist can enhance the quality of inpatient diabetes care by reducing rates of both hyperglycemia and hypoglycemia, decreasing length of stay, and reducing readmissions.55-57 A comparison of length of stay between a cohort of patients cared for by an IDMS team and a non-IDMS group is presented in Figure 9. These outcomes are achieved through quality improvement initiatives, development and implementation of hospital policies, discharge planning, nursing and medical staff education, collaboration with other subspecialties, and direct patient care (diabetes and endocrine consultations). At a community hospital within a large health system, involvement of endocrine hospitalists led to meaningful improvements in glucose outcomes, 58 27% decreased length of stay, and cost savings of ~$950,000 per year. 57 Reduction in hypoglycemia rates achieved through quality work overseen by an endocrine hospitalist was associated with annual cost savings of $100,000.56,59 Endocrine hospitalists can play an important leadership role in reviewing internal glucometric data, benchmarking against other hospitals, and oversight of reportable glycemic metrics.54,55,60 Endocrine hospitalists improve clinical outcomes and secure a return on investment for hospitals. 59
Figure 9.
Comparison of length of stay between a cohort of patients cared for by an IDMS team and a non-IDMS group. The mean length of stay in patients comanaged by the IDMS team decreased from 7.8 days to 5.7 days over time (27% reduction). There was no significant change in length of stay in the non-IDMS group. Modified from Mandel et al 57 under the terms of the Creative Commons Attribution-NonCommercial License (http://creativecommons.org/licenses/by-nc/4.0/). Abbreviations: IDMS, inpatient diabetes management system.
Pharmacy Programs to Improve Quality
Rima Bouajram, PharmD, BCCCP
University of California, San Francisco Medical Center, San Francisco, California, USA
Protocolized cross-disciplinary authorities and collaborations using technology to improve glycemic control and other patient outcomes may support proactive versus reactive glycemic control management.
EHR screening and alert tools may support more frequent glycemic control assessments and detect safety issues faster.
Multidisciplinary quality improvement efforts to identify institution-specific barriers to optimal glycemic control may support sustained improvement in patient outcomes.
A pharmacist-led collaborative initiative was implemented at the University of California, San Francisco Medical Center to address high rates of dysglycemia among patients treated with subcutaneous insulin in the ICU. The intervention consisted of pharmacist authority to comanage subcutaneous insulin regimens during weekdays for patients in the ICU. The pharmacist reviewed and proactively adjusted the insulin regimens in response to clinical changes or medication changes. The pharmacist performed just-in-time verbal communication of updates to multidisciplinary team members and a completed consult note in the EHR system. In addition to adjusting insulin based on historic glucose data, given the pharmacist’s review of medication orders prospectively prior to administration, the pharmacist often made proactive insulin dose adjustments to prevent dysglycemic events (e.g., reducing the insulin dose to prevent hypoglycemia during a steroid taper). In a pre-post analysis, this intervention resulted in significant reductions in hyperglycemia, severe hyperglycemia, hypoglycemia, and severe hypoglycemia. There were no severe hypoglycemic events noted in the postintervention period. In a retrospective analysis, patients on corticosteroids and COVID-19-positive patients were found to be at greater risk for dysglycemia. Dashboards have been developed within the EHR to flag patients with glucose levels out of range, and providers can use flags for patient lists to identify patients on the census with dysglycemia in the preceding 24 hours. Cross-disciplinary collaboration with physicians, advanced practice providers, and pharmacists is an effective strategy to address dysglycemia in the ICU.
Session 10: Diabetes Care and Education
Moderator
Jane Jeffrie Seley, DNP, MSN, MPH, GNP, BC-ADM, CDCES, FADCES
Weill Cornell Medicine, New York City, New York, USA
Interdisciplinary Inpatient Diabetes Education: Strategies that Succeed
Jane Jeffrie Seley, DNP, MSN, MPH, GNP, BC-ADM, CDCES, FADCES
Weill Cornell Medicine, New York City, New York, USA
Staff education designed for nurses, dietitians, pharmacists, and physicians is needed for inpatient clinicians to effectively teach patients how to safely and correctly provide diabetes survival skills prior to discharging the patient home.
Inpatient teleconsults for diabetes self-management education (DSME) with a diabetes care and education specialist require developing a standardized workflow with the care team and the availability of teaching resources.
Facilitating transitions from hospital to home should include education about diabetes survival skills, evaluation of the family’s ability to perform care, the patient’s ability to perform self-care, and a careful review of the EHR to evaluate home diabetes medication requirements.
Providing inpatient DSME is a proven strategy for preventing emergency room visits, 30-day readmissions, and lowering the HbA1c concentration. Unlike outpatient DSME, it should be limited to survival skills that promote a safe and effective short-term discharge regimen combined with a follow-up plan. 61 Teaching should include core topics such as medication taking, glucose monitoring, and prevention and treatment of hypoglycemia. 62 Teleconsults can offer just-in-time DSME by an off-site diabetes care and education specialist using a platform such as Cisco Jabber that provides a larger view of both the patient and the educator when teaching skills, such as insulin pens and glucose meters. These tele-consults are enhanced by staff training and distribution of survival skills handouts and teaching supplies to unit-based nurses who can then practice with patients. Prescriber education should focus on achieving and maintaining inpatient glycemic goals and reducing therapeutic inertia. A New York-Presbyterian/Weill Cornell campus pocket card delineates glycemic targets and provides clear instructions on insulin adjustments, thus simplifying and standardizing optimal dosing. 63
Patient-Owned Technology: To Wear or not to Wear
Rebecca Rick Longo, ACNP-BC, CDCES
Lahey Hospital and Medical Center, Burlington, Massachusetts, USA
Careful assessment of a patient’s clinical condition, ability to operate technology, and access to back up supplies is necessary for patients to safely continue their personal diabetes technology during an inpatient hospitalization.
Detailed hospital policies, checklists, and staff education are critical to safely allow patients to use their personal diabetes technology while they are hospitalized.
With appropriate patient selection and education, personal diabetes technology can be successfully continued during hospitalizations.
When patients wearing personal diabetes devices such as CGMs and automated insulin delivery systems enter the hospital, a decision should be made on a daily basis about whether or not the technology should be worn during hospitalization. 64 A CGM may be maintained for the patient’s own information while POC BGM testing is used for insulin dosing decisions and treatment of hypoglycemia. The decision to wear a patient’s own insulin pump requires an assessment by a clinician with specialized diabetes training such as a diabetes care and education specialist or endocrinologist to determine competency during the acute illness. Insulin pump therapy should not continue in patients unable or unwilling to participate in DSME or share pump management decisions with the care team. 65
Recommendations for how to handle an insulin pump on admission are presented in Figure 10. A functioning insulin pump and/or CGM as well as extra supplies including chargers and/or batteries must be brought by the patient/family to support these devices when they are to be worn in the hospital.
Figure 10.
Recommendations on the course of action for hospitalized patient with type 1 diabetes wearing insulin pumps. Reproduced from Mendez et al. 66 Abbreviations: CT, computed tomography; EGD, esophagogastroduodenoscopy; MRI, magnetic resonance imaging.
Preparation for Surgery: Preventing Cancelectomies
Shevy Serhofer, DNP, RN, CRNA
NAPA Anesthesia, Vassar Brothers Medical Center, Poughkeepsie, New York, USA
Procedure delays are a fundamental and costly problem in health care.
Glycemic-related procedure delays are preventable by checking relevant facility policies, coordinating decisions among team members, and having a plan that covers both best- and worst-case scenarios.
Various diabetes medications and diabetes technologies present specific challenges depending on the procedure and setting. Effective communication as well as clear and accessible guidelines prevent intraoperative dysglycemia and complications such as DKA and euglycemic DKA.
Procedure delays are a fundamental and costly problem in health care and can be categorized as organizational, patient-related, or medical. 67 As operating room costs are estimated to be approximately $150 per minute, 68 delays have significant financial implications, and they reduce hospital performance and patient satisfaction. Glycemic-related procedure delays are preventable by checking relevant facility policies, coordinating among team members, and having a plan in place that covers various tough case scenarios for hyperglycemia, hypoglycemia, and working with a patient who has their own diabetes technology. Diabetes medications, including oral antihyperglycemic agents, insulins, and noninsulin injectables, as well as insulin delivery devices and CGMs, present specific challenges, depending on the procedure and setting. Surgeries using lasers or electromagnetics and specialty areas, such as endoscopy, X-ray, CT scan, ultrasound, magnetic resonance imaging (MRI), and positron emission tomography (PET) scans, require effective communication as well as clear and accessible guidelines to prevent intraoperative dysglycemia and complications such as DKA and euglycemic DKA. 69
Managing Inpatient Nutrition for People with Diabetes: Challenges and Possible Solutions
Rachel Stahl, MS, RD, CDN, CDCES
Weill Cornell Medicine, New York City, New York, USA
Acute illness, medical conditions, and variability in appetite and intake affect mealtime glycemic management in the noncritical care setting.
An important strategy for matching the timing of meal tray delivery with BGM results and insulin dosing consists of three steps: (1) checking BG no more than 1 hour before meals, and less than 30 minutes is best; (2) giving prandial insulin 15 minutes before or after the start of the meal, and (3) providing patients with a consistent carbohydrate meal plan.
Survival skills for inpatient nutrition education includes teaching the Plate Method for balanced meal planning and carbohydrate counting with a focus on not just the quantity of foods but also the quality.
Proper coordination of BGM and administration of prandial insulin with meal delivery is key to support glycemic management. However, this coordination is complicated by such factors as delays in meal delivery and patients’ variability in nutritional intake. Key strategies for improving mealtime processes include educating staff that premeal BG checks should be done no more than 1 hour before meal trays are delivered, administering prandial insulin ±15 minutes of the first bite of the meal, providing consistent carbohydrate meal plans, and monitoring food intake.70,71 Setting up a notification system where food service staff can alert unit staff of delays in meal delivery and staff training on offering lower-carbohydrate food substitutions can help improve glycemic management. Survival skills for inpatient nutrition education should focus on the Plate Method for balanced meal planning, identifying carbohydrates counted, and encouraging high-fiber carbohydrate choices. Using scheduled BGM and meal trays as teachable moments can be a great way to encourage patient participation in DSME.
Session 11: Uniting People, Process, and Technology to Achieve Optimal Glycemic Management
Improving Outcomes for People with Diabetes
Jordan Messler, MD, SFHM, FACP
Glytec, Clearwater, Florida, USA
Currently delivered glycemic management often leaves room for errors because the impact of preventable hypoglycemia and untreated hyperglycemia is common, costly, and largely preventable.
To achieve quality improvement, it is necessary to overcome the implementation gap: what is known to what is done in practice.
A best-in-class glycemic management program requires a holistic strategy that assembles the right people and teams, provides them with clear protocols and standardized processes, and uses advanced technology to personalize care and improve patient safety.
Managing patients who are transitioning from intravenous insulin to subcutaneous insulin can be challenging; hence, it is essential to take a deliberate approach to transition.
Incorporating software like Glytec’s eGlycemic Management System supports a hospital’s glycemic management process by personalizing care, addressing the common challenges of intravenous and subcutaneous insulin, and providing the data and dashboards to measure and track performance.
Insulin dosing software based on CGM readings will play an important role in the future of glycemic management.
Hospitals are aware of the negative outcomes associated with preventable hypoglycemia and untreated hyperglycemia; however, there is an implementation gap between what is known and what can actually be done, largely because of lack of awareness, metrics, and standardization around glycemic management. Forming a multidisciplinary team and creating processes that are optimized for patient safety are the first steps toward achieving success. The next steps may include implementing a tool like Glytec’s eGlycemic Management System that combines technology and a support team to unite physicians, nurses, pharmacists, quality teams, hospital leadership, and others around a common goal: better, safer care. A case study review of how a hospital can use Glytec’s eGlycemic Management System to safely manage patients on intravenous insulin, transition insulin dosing from intravenous to subcutaneous, and then manage subcutaneous therapy demonstrated how technology can support the people and the processes by examples, such as real-time monitoring of patients, guidance toward best practice, and overcoming clinical inertia through daily adjustments of subcutaneous insulin. A comparison of time to target and incidence of hypoglycemia between computer-guided intravenous insulin therapy and subcutaneous insulin therapy is presented in Figure 11. Evolving technology requires forward thinking, and Glytec’s roadmap includes preparation for the future once CGMs are FDA cleared for inpatient use. With the new CMS metrics on severe hyperglycemia and hypoglycemia going into effect in 2023, achieving optimal glycemic management is a must have for hospitals. Building a case for change management takes time. It is important to take the steps now to gather the right team of people, implement standardized and measurable processes, and incorporate a technology solution that not only can support hospitals now but help prepare for the future.
Figure 11.
Comparison of time to target and incidence of hypoglycemia between computer-guided intravenous insulin therapy (green bars) and subcutaneous insulin therapy (blue bars). In total, 100 patients were retrospectively studied in each group. Reproduced with permission from Jordan Messler. Originally presented at the Virtual Hospital Diabetes Meeting. 72 Abbreviations: BG, blood glucose; ICU, intensive care unit.
Conclusions
The Virtual Hospital Diabetes Meeting presented the newest developments in the field of diabetes technology as well as the use of innovative technology to monitor and manage diabetes in the hospital. These presentations brought in medical, scientific, engineering, and regulatory perspectives to provide guidelines for future treatment of hospitalized people with diabetes.
Acknowledgments
For contributing key points to this report and reviewing the manuscript, we would like to thank: Joseph A. Aloi, MD; David G. Armstrong, DPM, MD, PhD; Vanita R. Aroda, MD; Rima Bouajram, PharmD, BCCCP; Matthew F. Bouchonville, MD, CDCES; Andrew Demidowich, MD; Ketan Dhatariya, MBBS, MSc, MD, MS, FRCP, PhD; Ruben Diaz, DNP, FNP-BC, BC-ADM, CDCES; Andjela T. Drincic, MD, FACP; Kathleen M. Dungan, MD, MPH; Juan Espinoza, MD, FAAP; Eileen R. Faulds, PhD, MS, RN, FNP-BC, CDCES; Rodolfo J. Galindo, MD, FACE; Gregory H. Gilbert, MD; Kensaku Kawamoto, MD, PhD, MHS; Mary T. Korytkowski, MD; Theocharis Koufakis, MD, PhD; James S. Krinsley, MD, FCCM, FCCP; Kristen M. Kulasa, MD; Mervyn Kyi, MB, BS, FRACP, PhD; Rebecca Rick Longo, ACNP-BC, CDCES; Nestoras Mathioudakis, MD, MHS; Gregory A. Maynard, MD, MSc, MHM; Marie E. McDonnell, MD; Jordan Messler, MD, SFHM, FACP; Viswanathan Mohan, MD, PhD, DSc; Raymond R. Moser MSEE, JD; James H. Nichols, PhD, DABCC, FAACC; Francisco J. Pasquel, MD, MPH; Athena Philis-Tsimikas, MD; Jean-Charles Preiser, MD, PhD; Daniel J. Rubin, MD, MSc, FACE; Robert J. Rushakoff, MD; Jane Jeffrie Seley, DNP, MSN, MPH, GNP, BC-ADM, CDCES, FADCES; Shevy Serhofer, DNP, RN, CRNA; Elias K. Spanakis, MD; Rachel Stahl, MS, RD, CDN, CDCES; Meng H. Tan, MD; Guillermo E. Umpierrez, MD; Amisha Wallia, MD, MS; Mihail Zilbermint, MD, MBA, FACE. We thank Annamarie Sucher-Jones for her expert editorial assistance.
Footnotes
Abbreviations: ADA, American Diabetes Association; AICD, pacemaker/automatic implantable cardioverter defibrillator; API, application programming interface; BG, blood glucose; BGM, blood glucose monitoring; BGT, blood glucose target; CCD, Continuity of Care Documents; CDA, Clinical Document Architecture; CDC, Centers for Disease Control and Prevention; CGM, continuous glucose monitor; CLIA, Clinical Laboratory Improvement Amendments; CMS, Centers for Medicare and Medicaid; COVID-19, coronavirus disease; CPT, Current Procedural Terminology; CT, computed tomography; CTLA-4, cytotoxic T-lymphocyte-associated protein 4; CVD, cerebral vascular disease; DKA, diabetic ketoacidosis; DM, diabetes mellitus; DPP-4, dipeptidyl peptidase-4; DSME, diabetes self-management education; DTS, Diabetes Technology Society; eCQMs, electronic clinical quality measures; EGD, esophagogastroduodenoscopy; eGMS, electronic glucose management systems; EHR, electronic health record; eMAR, electronic medication administration record; EMPI, Enterprise Master Patient Index; eQUIPS, Electronic Quality Improvement Programs; ERAS, Enhanced Recovery After Surgery; FDA, US Food and Drug Administration; FHIR, Fast Healthcare Interoperability Resources; GLP-1, glucagon-like peptide-1; GM, Glucommander; HbA1c, hemoglobin A1c; HCP, health care professional; HHS, hyperglycemic hyperosmolar state; HIPAA, Health Insurance Portability and Accountability Act; HITRUST, Health Information Trust Alliance; HL7, Health Level 7; HPA, hypothalamic-pituitary-adrenal axis; ICD-10, International Classification of Diseases, Tenth Revision; ICI, immune checkpoint inhibitor; ICI-ADM, ICI-associated autoimmune diabetes mellitus; iCoDE, integration of CGM data into the electronic health record; ICU, intensive care unit; IDMS, inpatient diabetes management service; IEEE, Institute of Electrical and Electronics Engineers; IICC, insulin infusion computer calculator; IP, intellectual property; irAE, immune-related adverse event; ISC, insulin-sensitive coefficient; ISO, International Organization for Standardization; IT, information technology; LOINC, Logical Observation Identifiers Names and Codes; mHealth, mobile health; MRI, magnetic resonance imaging; NHSN, National Health Safety Network; NIST CSF, National Institute of Standards and Technology Cybersecurity Framework; NPI, National Provider Identifier; NPO, nothing by mouth; OAD, oral antidiabetic drug; OAuth, open authorization; OMOP, Observational Medical Outcomes Partnership; PD-1, program cell death protein-1; PDL-1, program cell death ligand-1; PET, position emission tomography; POC, point of care; RCT, randomized controlled trial; RT-CGM, real-time continuous glucose monitor; RxNORM, a normalized naming system for generic and branded drugs, and a tool for supporting semantic interoperation between drug terminologies and pharmacy knowledge base systems; SARS-CoV-2, severe acute respiratory syndrome coronavirus 2; SGLT-2, sodium-glucose cotransporter-2; SMART, Substitutable Medical Applications, Reusable Technologies; SNOMED, Systemized Nomenclature of Medicine; SOC2, System and Organization Controls type 2—Trust Services Criteria; SQIA, subcutaneous insulin algorithm; SSI, sliding-scale insulin; TBR, time below range; TF, enteral tube feeding; TIR, time in range; TPN, total parenteral nutrition; UDI, Unique Device Identifier; UK, United Kingdom; US, United States; VA, US Department of Veterans Affairs.
Declaration of Conflicting Interests: The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: JP has received consultancy fees from Edwards/Dexcom, Medtronic, Optiscan, and Roche.
JJS has received remuneration for participation in one Advisory Board meeting for Dexcom.
GEU reports research funds to Emory University from Dexcom, Abbott, and Bayer.
AW reports research support from Novo Nordisk, UnitedHealth Group, and Eli Lilly.
MCL reports research funding from Dexcom.
UM reports research funding from Clementia Pharmaceutical and is an advisor for Ryse Health.
FJP has received unrestricted research support from Merck, Dexcom, and Insulet and consulting fees from Merck, Boehringer Ingelheim, Lilly, Medscape, and Dexcom.
VNS reports receiving research funding through University of Colorado from Dexcom Inc, Eli Lilly, NovoNorisk, Tandem Diabetes Care, and Insulet outside the submitted work. VNS’ employer, University of Colorado, received honoraria/consulting/speaking fees from Sanofi, Medscape, Lifescan, Dexcom, and Insulet.
EKS reports that this work was supported in part by a VA MERIT award from the US Department of Veterans Affairs Clinical Sciences Research and Development Service (1I01CX001825). He has received unrestricted research support from Dexcom (to the Baltimore VA Medical Center and to the University of Maryland) for the conduction of clinical trials.
DCK is a consultant to EOFlow, Fractyl Health, Integrity, Lifecare, Rockley Photonics, and Thirdwayv.
JH, AMY, KTN, NYX, RJR, ATD, RG, NNM, SS, AS, and GMT have nothing to disclose.
Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.
ORCID iDs: Jingtong Huang 
https://orcid.org/0000-0002-3119-9361
Andrea M. Yeung
https://orcid.org/0000-0002-5592-453X
Kevin T. Nguyen
https://orcid.org/0000-0001-9102-6537
Nicole Y. Xu
https://orcid.org/0000-0001-9353-8819
Jean-Charles Preiser
https://orcid.org/0000-0003-3163-0390
Robert J. Rushakoff
https://orcid.org/0000-0001-7502-9092
Jane Jeffrie Seley
https://orcid.org/0000-0003-1582-4320
Guillermo E. Umpierrez
https://orcid.org/0000-0002-3252-5026
Amisha Wallia
https://orcid.org/0000-0002-3183-4062
Andjela T. Drincic
https://orcid.org/0000-0001-8365-7662
Roma Gianchandani
https://orcid.org/0000-0002-3791-5891
M. Cecilia Lansang
https://orcid.org/0000-0002-7364-5550
Umesh Masharani
https://orcid.org/0000-0002-3269-804X
Nestoras Mathioudakis
https://orcid.org/0000-0002-0210-655X
Francisco J. Pasquel
https://orcid.org/0000-0002-3845-6703
Signe Schmidt
https://orcid.org/0000-0002-6968-6675
Viral N. Shah
https://orcid.org/0000-0002-3827-7107
Elias K. Spanakis
https://orcid.org/0000-0002-9352-7172
Andreas Stuhr
https://orcid.org/0000-0002-8087-2436
Gerlies M. Treiber
https://orcid.org/0000-0002-1658-5280
David C. Klonoff
https://orcid.org/0000-0001-6394-6862
References
- 1. Ruan Y, Bellot A, Moysova Z, et al. Predicting the risk of inpatient hypoglycemia with machine learning using electronic health records. Diabetes Care. 2020;43(7):1504-1511. doi: 10.2337/dc19-1743. [DOI] [PubMed] [Google Scholar]
- 2. Mathioudakis NN, Abusamaan MS, Shakarchi AF, et al. Development and validation of a machine learning model to predict near-term risk of iatrogenic hypoglycemia in hospitalized patients. JAMA Netw Open. 2021;4(1):e2030913. doi: 10.1001/jamanetworkopen.2020.30913. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Zale AD, Abusamaan MS, McGready J, Mathioudakis N. Development and validation of a machine learning model for classification of next glucose measurement in hospitalized patients. EClinicalMedicine. 2022;44:101290. doi: 10.1016/j.eclinm.2022.101290. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Fitzgerald O, Perez-Concha O, Gallego B, et al. Incorporating real-world evidence into the development of patient blood glucose prediction algorithms for the ICU. J Am Med Inform Assoc. 2021;28(8):1642-1650. doi: 10.1093/jamia/ocab060. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Mantena S, Arevalo AR, Maley JH, et al. Predicting hypoglycemia in critically ill patients using machine learning and electronic health records. J Clin Monit Comput. 2021. doi: 10.1007/s10877-021-00760-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Fralick M, Dai D, Pou-Prom C, Verma AA, Mamdani M. Using machine learning to predict severe hypoglycaemia in hospital. Diabetes Obes Metab. 2021;23(10):2311-2319. doi: 10.1111/dom.14472. [DOI] [PubMed] [Google Scholar]
- 7. Horton WB, Barros AJ, Andris RT, Clark MT, Moorman JR. Pathophysiologic signature of impending ICU hypoglycemia in bedside monitoring and electronic health record data: model development and external validation. Crit Care Med. 2022;50(3):e221-e230. doi: 10.1097/CCM.0000000000005171. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Witte H, Nakas CT, Bally L, Leichtle AB. Machine-learning prediction of hypo- and hyperglycemia from electronic health records: algorithm development and validation. JMIR Form Res. 2022. doi: 10.2196/36176. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Xu NY, Nguyen KT, DuBord AY, et al. The launch of the iCoDE standard project. J Diabetes Sci Technol. 2022:19322968221093662. doi: 10.1177/19322968221093662. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Services CfMM. Electronic Clinical Quality Measure (eCQM) Strategy Project Outcomes Report Version 1.0. 2020 [updated August 3, 2020]. https://ecqi.healthit.gov/sites/default/files/eCQM-Strategy-Outcomes-Report-August-2020-v1.pdf. Accessed June 8, 2022.
- 11. Bohé J, Abidi H, Brunot V, et al. Individualised versus conventional glucose control in critically-ill patients: the CONTROLING study—a randomized clinical trial. Intensive Care Med. 2021;47(11):1271-1283. doi: 10.1007/s00134-021-06526-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Liang Y, Wanderer J, Nichols JH, Klonoff D, Rice MJ. Blood gas analyzer accuracy of glucose measurements. Mayo Clin Proc. 2017;92(7):1030-1041. doi: 10.1016/j.mayocp.2017.03.009. [DOI] [PubMed] [Google Scholar]
- 13. Klonoff DC, Parkes JL, Kovatchev BP, et al. Investigation of the accuracy of 18 marketed blood glucose monitors. Diabetes Care. 2018;41(8):1681-1688. doi: 10.2337/dc17-1960. [DOI] [PubMed] [Google Scholar]
- 14. Nichols JH, Brandler ES, Fantz CR, et al. A multicenter evaluation of a point-of-care blood glucose meter system in critically ill patients. J Appl Lab Med. 2021;6(4):820-833. doi: 10.1093/jalm/jfab005. [DOI] [PubMed] [Google Scholar]
- 15. Dungan KM, Braithwaite SS, Preiser JC. Stress hyperglycaemia. Lancet. 2009;373(9677):1798-1807. doi: 10.1016/S0140-6736(09)60553-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Krinsley JS, Rule P, Pappy L, et al. The interaction of acute and chronic glycemia on the relationship of hyperglycemia, hypoglycemia, and glucose variability to mortality in the critically ill. Crit Care Med. 2020;48(12):1744-1751. doi: 10.1097/CCM.0000000000004599. [DOI] [PubMed] [Google Scholar]
- 17. Barron E, Bakhai C, Kar P, et al. Associations of type 1 and type 2 diabetes with COVID-19-related mortality in England: a whole-population study. Lancet Diabetes Endocrinol. 2020;8(10):813-822. doi: 10.1016/S2213-8587(20)30272-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Umpierrez GE, Gianchandani R, Smiley D, et al. Safety and efficacy of sitagliptin therapy for the inpatient management of general medicine and surgery patients with type 2 diabetes: a pilot, randomized, controlled study. Diabetes Care. 2013;36(11):3430-3435. doi: 10.2337/dc13-0277. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Umpierrez GE, Reyes D, Smiley D, et al. Hospital discharge algorithm based on admission HbA1c for the management of patients with type 2 diabetes. Diabetes Care. 2014;37(11):2934-2939. doi: 10.2337/dc14-0479. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Pasquel FJ, Urrutia MA, Cardona S, et al. Liraglutide hospital discharge trial: a randomized controlled trial comparing the safety and efficacy of liraglutide versus insulin glargine for the management of patients with type 2 diabetes after hospital discharge. Diabetes Obes Metab. 2021;23(6):1351-1360. doi: 10.1111/dom.14347. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Galindo RJ, Dhatariya K, Gomez-Peralta F, Umpierrez GE. Safety and efficacy of inpatient diabetes management with non-insulin agents: an overview of international practices. Curr Diab Rep. 2022;22(6):237-246. doi: 10.1007/s11892-022-01464-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Koufakis T, Mustafa OG, Ajjan RA, et al. The use of sodium-glucose co-transporter 2 inhibitors in the inpatient setting: is the risk worth taking. J Clin Pharm Ther. 2020;45(5):883-891. doi: 10.1111/jcpt.13107. [DOI] [PubMed] [Google Scholar]
- 23. Bhatt DL, Szarek M, Steg PG, et al. Sotagliflozin in patients with diabetes and recent worsening heart failure. N Engl J Med. 2021;384(2):117-128. doi: 10.1056/NEJMoa2030183. [DOI] [PubMed] [Google Scholar]
- 24. Voors AA, Angermann CE, Teerlink JR, et al. The SGLT2 inhibitor empagliflozin in patients hospitalized for acute heart failure: a multinational randomized trial. Nat Med. 2022;28(3):568-574. doi: 10.1038/s41591-021-01659-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Koufakis T, Metallidis S, Zebekakis P, Ajjan RA, Kotsa K. Sodium-glucose cotransporter 2 inhibitors in the era of COVID-19 pandemic: is the benefit to risk ratio still favorable. J Diabetes Sci Technol. 2020;14(4):745-747. doi: 10.1177/1932296820932155. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Maung TZ, Ergin HE, Javed M, Inga EE, Khan S. Immune checkpoint inhibitors in lung cancer: role of biomarkers and combination therapies. Cureus. 2020;12(5):e8095. doi: 10.7759/cureus.8095. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Lo Preiato V, Salvagni S, Ricci C, Ardizzoni A, Pagotto U, Pelusi C. Diabetes mellitus induced by immune checkpoint inhibitors: type 1 diabetes variant or new clinical entity? Review of the literature. Rev Endocr Metab Disord. 2021;22(2):337-349. doi: 10.1007/s11154-020-09618-w. [DOI] [PubMed] [Google Scholar]
- 28. Dhatariya KK, Vellanki P. Treatment of diabetic ketoacidosis (DKA)/hyperglycemic hyperosmolar state (HHS): novel advances in the management of hyperglycemic crises (UK versus USA). Curr Diab Rep. 2017;17(5):33. doi: 10.1007/s11892-017-0857-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Kilpatrick ES, Butler AE, Ostlundh L, Atkin SL, Sacks DB. Controversies around the measurement of blood ketones to diagnose and manage diabetic ketoacidosis. Diabetes Care. 2022;45(2):267-272. doi: 10.2337/dc21-2279. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Dhatariya KK; Joint British Diabetes Societies for Inpatient Care. The management of diabetic ketoacidosis in adults-an updated guideline from the Joint British Diabetes Society for Inpatient Care. Diabet Med. 2022;39(6):e14788. doi: 10.1111/dme.14788. [DOI] [PubMed] [Google Scholar]
- 31. Kitabchi AE, Umpierrez GE, Miles JM, Fisher JN. Hyperglycemic crises in adult patients with diabetes. Diabetes Care. 2009;32(7):1335-1343. doi: 10.2337/dc09-9032. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Fox CS, Pencina MJ, Wilson PW, Paynter NP, Vasan RS, D’Agostino RB, Sr. Lifetime risk of cardiovascular disease among individuals with and without diabetes stratified by obesity status in the Framingham heart study. Diabetes Care. 2008;31(8):1582-1584. doi: 10.2337/dc08-0025. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Sattar N, Lee MMY, Kristensen SL, et al. Cardiovascular, mortality, and kidney outcomes with GLP-1 receptor agonists in patients with type 2 diabetes: a systematic review and meta-analysis of randomised trials. Lancet Diabetes Endocrinol. 2021;9(10):653-662. doi: 10.1016/S2213-8587(21)00203-5. [DOI] [PubMed] [Google Scholar]
- 34. McGuire DK, Shih WJ, Cosentino F, et al. Association of SGLT2 inhibitors with cardiovascular and kidney outcomes in patients with type 2 diabetes: a meta-analysis. JAMA Cardiol. 2021;6(2):148-158. doi: 10.1001/jamacardio.2020.4511. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Wallia A, Umpierrez GE, Rushakoff RJ, et al. Consensus statement on inpatient use of continuous glucose monitoring. J Diabetes Sci Technol. 2017;11(5):1036-1044. doi: 10.1177/1932296817706151. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Fortmann AL, Spierling Bagsic SR, Talavera L, et al. Glucose as the fifth vital sign: a randomized controlled trial of continuous glucose monitoring in a non-ICU hospital setting. Diabetes Care. 2020;43(11):2873-2877. doi: 10.2337/dc20-1016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Galindo RJ, Umpierrez GE, Rushakoff RJ, et al. Continuous glucose monitors and automated insulin dosing systems in the hospital consensus guideline. J Diabetes Sci Technol. 2020;14(6):1035-1064. doi: 10.1177/1932296820954163. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Faulds ER, Jones L, McNett M, et al. Facilitators and barriers to nursing implementation of continuous glucose monitoring (CGM) in critically ill patients with COVID-19. Endocr Pract. 2021;27(4):354-361. doi: 10.1016/j.eprac.2021.01.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Chow KW, Kelly DJ, Rieff MC, et al. Outcomes and healthcare provider perceptions of real-time continuous glucose monitoring (rtCGM) in patients with diabetes and COVID-19 admitted to the ICU. J Diabetes Sci Technol. 2021;15(3):607-614. doi: 10.1177/1932296820985263. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Davis GM, Faulds E, Walker T, et al. Remote continuous glucose monitoring with a computerized insulin infusion protocol for critically ill patients in a COVID-19 medical ICU: proof of concept. Diabetes Care. 2021;44(4):1055-1058. doi: 10.2337/dc20-2085. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Agarwal S, Mathew J, Davis GM, et al. Continuous glucose monitoring in the intensive care unit during the COVID-19 pandemic. Diabetes Care. 2021;44(3):847-849. doi: 10.2337/dc20-2219. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Sadhu AR, Serrano IA, Xu J, et al. Continuous glucose monitoring in critically ill patients with COVID-19: results of an emergent pilot study. J Diabetes Sci Technol. 2020;14(6):1065-1073. doi: 10.1177/1932296820964264. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. Longo RR, Elias H, Khan M, Seley JJ. Use and accuracy of inpatient CGM during the COVID-19 pandemic: an observational study of general medicine and ICU patients. J Diabetes Sci Technol. 2021:19322968211008446. doi: 10.1177/19322968211008446. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. Galindo RJ, Aleppo G. Continuous glucose monitoring: the achievement of 100 years of innovation in diabetes technology. Diabetes Res Clin Pract. 2020;170:108502. doi: 10.1016/j.diabres.2020.108502. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45. Galindo RJ, Migdal AL, Davis GM, et al. Comparison of the FreeStyle Libre Pro flash continuous glucose monitoring (CGM) system and point-of-care capillary glucose testing in hospitalized patients with type 2 diabetes treated with basal-bolus insulin regimen. Diabetes Care. 2020;43(11):2730-2735. doi: 10.2337/dc19-2073. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46. Ekhlaspour L, Mondesir D, Lautsch N, et al. Comparative accuracy of 17 point-of-care glucose meters. J Diabetes Sci Technol. 2017;11(3):558-566. doi: 10.1177/1932296816672237. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. Spanakis EK, Levitt DL, Siddiqui T, et al. The effect of continuous glucose monitoring in preventing inpatient hypoglycemia in general wards: the glucose telemetry system. J Diabetes Sci Technol. 2018;12(1):20-25. doi: 10.1177/1932296817748964. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48. Perez-Guzman MC, Shang T, Zhang JY, Jornsay D, Klonoff DC. Continuous glucose monitoring in the hospital. Endocrinol Metab (Seoul). 2021;36(2):240-255. doi: 10.3803/EnM.2021.201. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49. Gothong C, Singh LG, Satyarengga M, Spanakis EK. Continuous glucose monitoring in the hospital: an update in the era of COVID-19. Curr Opin Endocrinol Diabetes Obes. 2022;29(1):1-9. doi: 10.1097/MED.0000000000000693. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50. Singh LG, Satyarengga M, Marcano I, et al. Reducing inpatient hypoglycemia in the general wards using real-time continuous glucose monitoring: the glucose telemetry system, a randomized clinical trial. Diabetes Care. 2020;43(11):2736-2743. doi: 10.2337/dc20-0840. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51. Umpierrez GE, Hellman R, Korytkowski MT, et al. Management of hyperglycemia in hospitalized patients in non-critical care setting: an endocrine society clinical practice guideline. J Clin Endocrinol Metab. 2012;97(1):16-38. doi: 10.1210/jc.2011-2098. [DOI] [PubMed] [Google Scholar]
- 52. Turchin A, Matheny ME, Shubina M, Scanlon JV, Greenwood B, Pendergrass ML. Hypoglycemia and clinical outcomes in patients with diabetes hospitalized in the general ward. Diabetes Care. 2009;32(7):1153-1157. doi: 10.2337/dc08-2127. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53. Spanakis EK, Umpierrez GE, Siddiqui T, et al. Association of glucose concentrations at hospital discharge with readmissions and mortality: a nationwide cohort study. J Clin Endocrinol Metab. 2019;104(9):3679-3691. doi: 10.1210/jc.2018-02575. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54. Zilbermint M. The endocrine hospitalist: enhancing the quality of diabetes care. J Diabetes Sci Technol. 2021;15(4):762-767. doi: 10.1177/19322968211007908. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55. Shelton C, Demidowich AP, Zilbermint M. Inpatient diabetes management during the COVID-19 crisis: experiences from two community hospitals. J Diabetes Sci Technol. 2020;14(4):780-782. doi: 10.1177/1932296820930268. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 56. Shelton C, Demidowich AP, Motevalli M, et al. Retrospective quality improvement study of insulin-induced hypoglycemia and implementation of hospital-wide initiatives. J Diabetes Sci Technol. 2021;15(4):733-740. doi: 10.1177/19322968211008513. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 57. Mandel SR, Langan S, Mathioudakis NN, et al. Retrospective study of inpatient diabetes management service, length of stay and 30-day readmission rate of patients with diabetes at a community hospital. J Community Hosp Intern Med Perspect. 2019;9(2):64-73. doi: 10.1080/20009666.2019.1593782. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 58. Demidowich AP, Batty K, Love T, et al. Effects of a dedicated inpatient diabetes management service on glycemic control in a community hospital setting. J Diabetes Sci Technol. 2021;15(3):546-552. doi: 10.1177/1932296821993198. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 59. Haque WZ, Demidowich AP, Sidhaye A, Golden SH, Zilbermint M. The financial impact of an inpatient diabetes management service. Curr Diab Rep. 2021;21(2):5. doi: 10.1007/s11892-020-01374-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 60. Sidhaye AR, Mathioudakis N, Bashura H, Sarkar S, Zilbermint M, Golden SH. Building a business case for inpatient diabetes management teams: lessons from our center. Endocr Pract. 2019;25(6):612-615. doi: 10.4158/EP-2018-0471. [DOI] [PubMed] [Google Scholar]
- 61. Magee MF, Baker KM, Bardsley JK, Wesley D, Smith KM. Diabetes to go-inpatient: pragmatic lessons learned from implementation of technology-enabled diabetes survival skills education within nursing unit workflow in an urban, tertiary care hospital. Jt Comm J Qual Patient Saf. 2021;47(2):107-119. doi: 10.1016/j.jcjq.2020.10.007. [DOI] [PubMed] [Google Scholar]
- 62. The Joint Commission. APPROVED: revisions to advanced certification requirements for inpatient diabetes care. Jt Comm Perspect. 2016;36(11):8-10. [PubMed] [Google Scholar]
- 63. Seley JJ, et al. Combating clinical inertia: Using an educational intervention to teach house staff to safely and effectively intensify insulin therapy in the hospital setting. From “Behavioral medicine, clinical nutrition, education, and exercise.” Diabetes. 2014; 63 (Supplement_1): A585. 10.2337/db14-2293-2343 [DOI] [Google Scholar]
- 64. Umpierrez GE, Klonoff DC. Diabetes technology update: use of insulin pumps and continuous glucose monitoring in the hospital. Diabetes Care. 2018;41(8):1579-1589. doi: 10.2337/dci18-0002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 65. Davis GM, Galindo RJ, Migdal AL, Umpierrez GE. Diabetes technology in the inpatient setting for management of hyperglycemia. Endocrinol Metab Clin North Am. 2020;49(1):79-93. doi: 10.1016/j.ecl.2019.11.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 66. Mendez CE, Umpierrez GE. Management of type 1 diabetes in the hospital setting. Curr Diab Rep. 2017;17(10):98. doi: 10.1007/s11892-017-0919-7. [DOI] [PubMed] [Google Scholar]
- 67. Caesar U, Karlsson J, Hansson E. Incidence and root causes of delays in emergency orthopaedic procedures: a single-centre experience of 36,017 consecutive cases over seven years. Patient Saf Surg. 2018;12:2. doi: 10.1186/s13037-018-0149-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 68. Gupta R, Tat Q, O’Brien J, et al. Utilization of lean project management principles and health informatics to reduce operating room delays in a vascular surgery practice. Am J Surg. 2022;223(1):176-181. doi: 10.1016/j.amjsurg.2021.07.040. [DOI] [PubMed] [Google Scholar]
- 69. Yeh T, Yeung M, Mendelsohn Curanaj FA. Managing patients with insulin pumps and continuous glucose monitors in the hospital: to wear or not to wear. Curr Diab Rep. 2021;21(2):7. doi: 10.1007/s11892-021-01375-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 70. Ryan DB, Swift CS. The mealtime challenge: nutrition and glycemic control in the hospital. Diabetes Spectr. 2014;27(3):163-168. doi: 10.2337/diaspect.27.3.163. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 71. Hughes L, Caragher M. Addressing hospital-acquired hypoglycemia. Am J Nurs. 2021;121(5):56-61. doi: 10.1097/01.NAJ.0000751116.58400.46. [DOI] [PubMed] [Google Scholar]
- 72. Rabinovich M, Hall A, Gayed R, Chester K, Crowe J, Messler J. Improved patient safety and glycemic outcomes in the ICU using computer-guided intravenous insulin therapy versus subcutaneous insulin therapy. Diabetes Technology Society (DTS) Virtual Poster Meeting. June 18, 2020. [Google Scholar]