Abstract
Objective
Glycemic control in hospitalized patients with diabetes is crucial yet remains challenging. Hypoglycemia poses an increased risk of complications, prolonged stays, and mortality. Strategies such as basal-bolus-correction insulin can help balance glycemic control and reduce hypoglycemic events. This study evaluated the impact of a mandatory electronic medical record–based insulin order set across a multihospital system.
Methods
This quality improvement project was implemented at 10 hospitals in North Carolina, including urban, rural, and community sites. The intervention introduced a standardized insulin order set with multiple fail-safes addressing insulin-to-nutrition mismatch, basal and bolus coverage, and dose accuracy. The primary outcome was hypoglycemia days per 1000 patient-days in nonpregnant adults. Secondary outcomes included hyperglycemia days and the rate of hemoglobin A1c orders.
Results
Hypoglycemia days across all sites decreased from 13 to 10 per 1000 patient-days (24% reduction, P < .01), with 9 sites showing declines (7%-39%) and 1 site showing an increase. Hyperglycemia days decreased from 90 to 76 per 1000 patient-days (16% reduction, P < .01). Hemoglobin A1c ordering improved from 60% to 67% (12% increase, P < .01). The implementation of a system-wide electronic medical record–based insulin order set reduced hypoglycemia rates across 9 of 10 hospitals without adversely affecting hyperglycemia. The standardized protocol contributed to these improvements. Site-specific variations in outcomes suggest the need for tailored interventions in certain locations.
Conclusion
Implementation of a mandatory, standardized insulin order set with multiple fail-safes across a multihospital system effectively reduced hypoglycemic events, thus demonstrating their value in managing inpatient diabetes.
Key words: diabetes mellitus, type 2; diabetes mellitus, type 1; hyperglycemia; quality improvement
Highlights
-
•
Implementation of a mandatory, electronic medical record–based insulin order set across a 10-hospital system led to a 24% relative reduction in hypoglycemic days (P < .01)
-
•
The standardized order set incorporated evidence-based basal-bolus-correction insulin regimens, patient-specific dosing, and insulin-to-nutrition matching safeguards
-
•
Hyperglycemic days also declined by 16% (P < .01), demonstrating improved overall glycemic control without compromising safety
-
•
A system-wide increase in hemoglobin A1c ordering (from 60% to 67%, P < .01) reflects better engagement implementation of best practices
-
•
High uptake and sustained use were achieved through removal of alternative insulin ordering pathways, comprehensive education, and real-time feedback from a command center
Clinical Relevance
This study demonstrates the value of an electronic medical record–based insulin order set in reduction of hypoglycemic events across a large hospital system of 10 hospitals without compromising hyperglycemic control. This work is applicable to any institution performing quality improvement to reduce the rate of inpatient hypoglycemia.
Introduction
Among inpatients with diabetes, hypoglycemia is estimated to occur in 3.5% to 10.5% of hospital admissions.1 Inpatient hypoglycemia is associated with an increased risk for neurologic and cardiovascular complications, infections, kidney dysfunction, hospital length of stay and mortality.1, 2, 3, 4, 5, 6, 7, 8 Maintaining glycemic control in the hospital setting is challenging. A myriad of factors contribute to insulin induced hypoglycemia. Variations in severity of illness (septic shock, renal failure critical illness, heart failure, liver failure, and malignancy)1, medications, and mismatch between insulin and nutritional intake are common in acutely ill patients.9 Use of quality improvement (QI) projects, revised hospital glucose management policies, electronic medical record (EMR)–based insulin order sets, decision support tools, and electronic glycemic management systems to reduce in-hospital hypoglycemia have been shown to be valuable tools to reduce hypoglycemia.10, 11, 12, 13, 14, 15, 16, 17, 18 Notably, most of these studies are performed at a single institution.19 This leads to the question: can these changes be implemented across multiple hospitals with similar efficacy?
Insulin dosing and nutrition are the most common factors associated with hypogylecmia.20 Despite data showing that the use of basal-bolus insulin is safe and effective in hospitalized patients, dose decision making and consistent ordering of insulin can be challenging.21 Reductions in the frequency of hypoglycemia have been demonstrated when using weight-based insulin orders22 and through the use of decision support tools to assist in ordering of basal and prandial insulin for different nutritional statues (nothing by mouth, total parenteral nutrition, or tube feeds).16,19,23,24 Furthermore, the American Diabetes Association (ADA) recommends consistent carbohydrate intake for admitted patients, which has been shown to reduce hypoglycemic episodes compared with patient-controlled diets.25,26 Use of hospital standardized hypoglycemia protocols that empower nursing to be proactive to prevent hypoglycemia has also been demonstrated.27 Finally, the importance of education initiatives to help reduce hypoglycemia is also critical.28
The primary goal of this multihospital QI initiative was to reduce hypoglycemia across the institution. Secondary goals included maintaining rates of hyperglycemia and increasing hemoglobin A1c (HbA1c) capture rate. Key components included making the order set mandatory and focusing on multidisciplinary education for pharmacists, physicians, advanced practice providers and nurses.
Methods
System-Wide Initiative to Improve Glycemic Control
The project began as a centralized QI project across this health care system. The system included 1 large quaternary academic center, 1 urban academic-associated quaternary center, and 8 additional hospitals in North Carolina. Hospital sizes varied from a rural hospital of 25 beds to an academic hospital of 932 beds (mean, 320 beds), and the fraction of admitted patients who were on the Diabetes registry varied from 15% to 26% (mean, 21%). Prior to this initiative, the system’s hospitals had hypoglycemia rates above the available benchmark data and hyperglycemia rates generally within benchmarks. A committee of stakeholders (eg, physicians, advanced practice providers, and QI experts) from across the health care system met weekly to plan strategies to reduce hypoglycemia. It was hypothesized that implementation of a mandatory EMR-based insulin order set (including best practice guidelines from the ADA, along with education to nurses, pharmacists, and clinicians) would reduce hypoglycemia events measured in patients days without increasing hyperglycemia.18,22,26,29,30 Hypoglycemia was defined as a blood glucose level of <54 mg/dL on serum or point-of-care glucose reading. Hyperglycemia was defined as a blood glucose level of >300 mg/dL on serum or point-of-care glucose.
Development of a Standardized Insulin Order Set
Multiple entities already had functioning insulin order sets that had been devised primarily to address hyperglycemia. The committee review of hospital charts and practice patterns showed that a major cause of hypoglycemia was an insulin-to-nutrition mismatch and inappropriate ordering of correction-only insulin. Thus, a new larger team was formed of 25 people including physicians, nurse practitioners, physician assistants, pharmacists, nurses, diabetes educators, dieticians, laboratory personnel, compliance officers, information system development specialists, and QI staff to create an order set to address these 2 issues. This team met weekly and started to review each component of the insulin order set along with the best practice guidelines from the ADA, subject matter expert opinion, and literature review. Simultaneously, they engaged with leadership and providers to create associated educational materials and policies.
The first order set change focused on starting both basal and bolus insulin safely. Specific elements of the order set were designed to reduce the likelihood of hypoglycemia, for example, ordering basal insulin based on weight, starting a carbohydrate consistent diet, a link to endocrinology and diabetes education consults, built-in reductions in insulin doses when a patient’s dietary status changed, and insulin hold parameters for when glucose levels were below minimum targets. Dietary status was the initial decision point for insulin regimen, which led to specific orders for patients who were nothing by mouth or on enteral feeds or were eating by mouth. Correction insulin orders were based on weight with low-risk insulin sensitivity factors listed for a range of different patients’ body mass index. A guide for ordering clinicians was included. The order set also included tables that had both order instructions meshed with educational guidelines. These evidence-based recommendations were visible while clinicians are using the order set.
The insulin order set also facilitated capturing of an HbA1c by prompting ordering if testing had not been completed in the last 3 months.
Finally, the order set was designed to facilitate communication. Standardized hypoglycemia management orders were updated and synchronized with nursing policy. Glycemic reporting by the laboratory was also changed to reflect a system-wide policy rather than hospital-specific policies. Insulin administration instructions were made clear through the review process outlined in the following, which also facilitated and standardized communication between nurses and providers.
Order Set Consensus and Implementation
Once the committee generated the insulin order set, it was distributed to a diverse group of providers including advanced practice providers, pharmacists, nurses, and physicians at various levels of training (internal medicine subspecialties and surgical specialties) across the system for review. A formal Delphi survey approach was used to gather input and iteratively modify the order set (with repeated surveys) until there was an 80% agreement for each element.
After 80% approval rate for each component was attained, the order set was built into the EMR. Prior to going live, education was provided to the stakeholders via videos, pocket cards, flyers posted in clinical areas, web-based modules, and 26 formal presentations to various providers and leader groups. Provider-specific presentations were conducted with hospitalists, cardiologists, medicine residents, and surgical residents. These presentations reached all hospitals within the system as they were recorded and made available to anyone using the EMR.
Elements of the order set are shown in Supplementary Figures 1 to 4.
Basal and bolus insulin is ordered with specific instructions to guide insulin administration. Supplementary Figures 2 and 3 and 5 show the order set for patients eating and on tube feeds, respectively. The process was simplified with preselected nutritional regimens and commonly used options, accompanied by clear ordering instructions and built-in safeguards to reduce the risk of hypoglycemia (Supplementary Fig. 4). Standardized hypoglycemia management orders were updated to align with institutional nursing policy, and a standardized process was implemented to notify providers of abnormal glucose values.
When the new order set was initially created and made available for use (May 2022), it was an option for providers along with all previously available ways of ordering. After 3 months, all these alternative options for ordering insulin were removed from the EMR (August 2022), thus forcing insulin orders to be placed via a single method. Around this time, a “command center” was created that met daily to assess order set usage and to provide user support.
In the subsequent months, the team that created the order set met regularly to consider ongoing feedback from users, oversee modifications to the order set, and monitor hypoglycemia, hyperglycemia, and HbA1c ordering metrics. Multiple changes to the order set were implemented during the weeks/months after the order set implementation.
Patient Selection Criteria and Data Collection
Data were assessed on a system-wide level and per hospital. The target patients for this initiative included those who were nonpregnant adults (aged ≥18 years) admitted to the hospital for something other than diabetes but with a known diagnosis of diabetes (ie, active on the diabetes registry—defined by items such as having an International Classification of Diseases, 10th Revision, diagnosis code consistent with diabetes or an HbA1c level of ≥7.0% in the past 2 years). Patients admitted for acute complications/hyperglycemic emergencies (eg, diabetic ketoacidosis, hypoglycemia, and hyperosmolar hyperglycemic state) were excluded. Patients admitted for hypoglycemia were excluded because hypoglycemia was not caused by hospital insulin administration.
Uptake of the insulin order set was defined as use of the standardized insulin pathway in patients included in the analysis (ie, adults admitted to the hospital for something other than diabetes and on the diabetes registry). Order set utilization was monitored electronically through the Epic EMR. Although uptake was not a primary focus of this study, it was tracked to understand implementation context and monitor for unexpected insulin orders outside of the implemented order set.
The preorder set data were collected between July 1, 2021, to March 31, 2022, with the postorder set data collected from January 1, 2023, to September 30, 2023. The time between April 1, 2022, and December 31, 2022, was not considered in the analysis as that was during the transition time for user education and order set implementation. Given the delays in manuscript writing, additional data trends beyond what is included in the analysis are shown in the Supplementary Material.
Statistical Analysis
Sample size calculations were conducted for each of the outcomes using a t-sided test, with an α of 0.05 and a statistical power (1-β) of 80%. Clinical meaningful effect sizes were determined based on expected target changes of 5% reduction in hypoglycemia (estimated sample size of 690 patients), 7% reduction in hyperglycemia (540 patients), and 15% increase in HbA1c capture (115 patients). The total numbers of patients included in the preintervention arm and postintervention arm were 20 031 and 22 819 patients, respectively. Statistical analysis consisted of a 2-sided Χ2 test that was used to compare the rates of hypoglycemic days, hyperglycemic days, and HbA1c capture in the preintervention group versus intervention group using IBM SPSS statistical software package.31 Statistical significance was considered a P value of <.05.
The Office of Human Research Ethics review of this work determined that it did not constitute human subjects research as defined under federal regulations and did not require formal institutional review board approval due to a QI exemption.
Results
Use of the insulin order set is demonstrated in Figure 1. A system-wide analysis of pooled data across 10 hospitals demonstrated significant improvements in the rates of hypoglycemic days, hyperglycemic days, and HbA1c capture, as shown in Table 1 and Figures 2 and 3.
Fig. 1.
Rate of insulin order set utilization over time and before and after the intervention. Overall insulin order set utilization plateaued at approximately 80% system-wide. The remaining 20% of patients not using the order set included those on oral agents or self-administering insulin, those with clinical scenarios where the order set was not applicable, and cases where our definitions of active diabetes had subtle imperfections.
Table 1.
System-wide Pooled Data for the 10 Hospitals, Preintervention Versus Postintervention, for the 3 End Points Considered
| Metric | Preintervention July 2021 to March 2022(N = 20 031)a |
Postintervention January 2023 to September 2023 (N = 22 819)a |
Relative percent improvement | P value |
|---|---|---|---|---|
| Hypoglycemic days per 1000 patient-days | 12.7 | 9.6 | 24% | <.001 |
| Hyperglycemic days per 1000 patient-days | 90.2 | 75.6 | 16% | <.001 |
| HbA1C capture | 60 | 67 | 12% | <.001 |
Abbreviation: HbA1c = hemoglobin A1c.
Bolded P values indicate statistical significance, which is defined as a P of <.05.
Number of patients in studied groups.
Fig. 2.
Hypoglycemia event tracking∗. The numbers of hypoglycemic days per 1000 were 13 and 10 before and after the intervention, respectively (P < .01); ∗ Preintervention hypoglycemic days per 1000 patient-days (90), as per Table 2; postintervention hypoglycemic days per 1000 patient-days (10), as per Table 3.
Fig. 3.
Hyperglycemia event tracking∗. The numbers of hyperglycemic patient days per 1,000 were 90 and 76 before and after the intervention, respectively (P < .01); as per data in Table 3. ∗Preintervention hyperglycemic days per 1000 patient-days (90), as per Table 3; postintervention hyperglycemic days per 1000 patient-days (76), as per Table 3.
At the individual hospital level, hypoglycemic and hyperglycemic day data are shown in Tables 2 and 3, respectively. HbA1c capture percentage data are shown in Table 4. Hypoglycemic and hyperglycemic days were used rather than raw numbers to clean the data and avoid overreporting of either hyperglycemia or hypoglycemia.
Table 2.
Rate of Hypoglycemia Days, Preintervention Versus Postintervention
| Hospital | Preintervention Hypoglycemic days per 1000 patient-days |
Postintervention Hypoglycemic days per 1000 patient-days |
Relative percent decline in hypoglycemic days | P value |
|---|---|---|---|---|
| 1 | 11.0 | 7.5 | 32% | .02 |
| 2 | 5.3 | 3.6 | 32% | .52 |
| 3 | 15.3 | 9.4 | 39% | <.001 |
| 4 | 8.8 | 6.7 | 24% | .12 |
| 5 | 12.8 | 10.1 | 21% | .02 |
| 6 | 10.0 | 7.2 | 28% | .08 |
| 7 | 12.1 | 11.3 | 7% | .27 |
| 8 | 11.8 | 8.5 | 28% | <.001 |
| 9 | 7.1 | 14.4 | −103 | .03 |
| 10 | 19.4 | 12 | 38% | <.001 |
| All sites | 12.7 | 9.6 | 24% | <.001 |
Bolded P values indicate statistical significance, which is defined as a P of <.05.
Table 3.
Rate of Hyperglycemia Days, Preintervention Versus Postintervention
| Hospital | Preintervention Hyperglycemic days per 1000 patient-days |
Postintervention Hyperglycemic days per 1000 patient-days |
Relative percent decline in hyperglycemic days | P value |
|---|---|---|---|---|
| 1 | 151.2 | 101.8 | 33% | <.001 |
| 2 | 99.3 | 99.9 | −1% | .99 |
| 3 | 108.5 | 88.4 | 19% | <.001 |
| 4 | 95.3 | 64.3 | 33% | <.001 |
| 5 | 106.5 | 73.5 | 31% | <.001 |
| 6 | 113.4 | 65.2 | 43% | <.001 |
| 7 | 62.2 | 45.7 | 27% | <.001 |
| 8 | 79.4 | 89.6 | −13% | <.001 |
| 9 | 138.5 | 81.0 | 42% | <.001 |
| 10 | 105.2 | 88.7 | 16% | <.001 |
| All sites | 90.2 | 75.6 | 16% | <.001 |
Bolded P values indicate statistical significance, which is defined as a P of <.05.
Table 4.
Rate of Hemoglobin A1c Capture, Preintervention Versus Postintervention
| Hospital | Preintervention HbA1c encounter capture rate in the last 90 d |
Postintervention HbA1c encounter capture rate in the last 90 d |
Relative percent increase in HbA1c capture rate within the last 90 d | P value |
|---|---|---|---|---|
| 1 | 57% | 68% | 19% | <.001 |
| 2 | 68% | 74% | 9% | .17 |
| 3 | 64% | 71% | 11% | <.001 |
| 4 | 90% | 91% | 1% | .43 |
| 5 | 80% | 68% | −15% | <.001 |
| 6 | 48% | 60% | 25% | <.001 |
| 7 | 52% | 64% | 23% | <.001 |
| 8 | 56% | 65% | 16% | <.001 |
| 9 | 39% | 65% | 67% | <.001 |
| 10 | 59% | 64% | 8% | <.001 |
| All sites | 60% | 67% | 12% | <.001 |
Abbreviation: HbA1c = hemoglobin A1c.
Bolded P values indicate statistical significance, which is defined as a P of <.05.
Discussion
In this QI project, the use of a system-wide EMR-based insulin order set resulted in significant reductions in hypoglycemic rates across the health care system and in 9 of the 10 individual hospitals. Hypoglycemia reductions were thought to be multifactorial due to a combination of decision support tools including weight-based insulin dosing, dosing based on nutritional status, implementation of standardized nursing-driven hypoglycemia protocols, and extensive system-wide multiprovider educational efforts.
This approach included multiple changes to how insulin is administered across 10 hospitals in North Carolina. These changes, while having literature to support their individual value in hypoglycemia reduction, likely worked in tandem to increase the effectiveness in reducing rates of hypoglycemia. The removal of legacy orders from the EMR resulted in an 80% user uptake of the insulin order set, which through its many facets helped reduce hypoglycemia. Long-term monitoring data indicate that as of time of publication, the hypoglycemia reduction has been maintained (Supplementary Figs. 5 and 6). It is not possible to determine which aspects are most or least responsible for this change. This was a QI project and did not control multiple variables. However, it was also a real-world implementation. It is also not clear why 1 of the 10 sites had an increase in hypoglycemia. It is possible that its small sample size and low preintervention rates of hypoglycemia may have contributed. Additional site-specific adjustments to order sets may be needed based on education status, resources, and culture when implementing system-wide order sets in the future.
With regard to reductions in hypoglycemia, the findings are similar to those reported by others, where standard order sets and similar initiatives have been associated with 5% to 50% reductions in hypoglycemia rates in both retrospective and prospective cohort studies.16,18,22,28,29,32, 33, 34 Most of the published studies describe the utility of these order sets in the setting of single-hospital environments. Nevertheless, there are only a few multicenter studies that demonstrated short-term hypoglycemia reduction on a smaller scale than the current study.35,36 The current study remains novel in that it reports implementation across several hospitals (total of 10) in a health care system, with hospitals ranging from small rural hospitals to large quaternary academic centers. Reductions in hypoglycemic rates across 9 of the 10 hospitals and across the institution demonstrate the power of the multifaceted approach used in implementing the EMR-based insulin ordered set. Implementation of clinical decision support tools has been shown to have mixed results when it comes to hypoglycemia outcomes making the current results relevant.37
The secondary outcomes also demonstrated improvements—lower hyperglycemia and improved HbA1c capture. When implementing changes aimed to reduced hypoglycemia, reductions in hypoglycemia and hyperglycemia have been demonstrated.14,15,38 However, other studies with similar reductions in hypoglycemia have shown unchanged or worsened hyperglycemia data.13,16, 17, 18,22,39 Inpatient hyperglycemia, similar to hypoglycemia is also associated with poor outcomes including increased infections, hospital stay, disability after discharge, and mortality. The current study was able to demonstrate reductions in both hypoglycemia and hyperglycemia, which remains important for maximizing clinical outcomes.40 HbA1c capture within the past 90 days is recommended in the ADA standards of care for diabetes care in the hospital.26 Ease and prompting appropriate ordering of HbA1c via the insulin order set likely lead to an increase in HbA1c capture. HbA1c data at time of admission should be used to help determine insulin selection for admitted patients. It is possible that increased HbA1c capture could have assisted with reduction in rates of hypoglycemia and hyperglycemia observed in the study.
There are several limitations of this study. First and foremost, it was designed as a real-time QI study, and thus, each variable could not be evaluated independently of each other. Therefore, patient demographics were not included as part of the analysis. This limits the analysis ability to find confounders between groups. Second, the collection period for the data does not account for seasonal variations in admissions (influenza or COVID-19 seasons) that could impact both hypoglycemia and hyperglycemia rates. In future studies, making sure the preintervention and postintervention phases occur during similar times of year could help control for this factor. Finally, the definition used for inclusion in the diabetes registry (and, hence, to be included in the analysis) may be considered imperfect. Nevertheless, there are no perfect ways to do it and, therefore, remains a limitation.
Randomized, controlled studies are difficult to implement when multiple interventions are simultaneous and multiple diverse sites are being studied. Still, changes to the methods could help draw better conclusions. A randomized controlled trial could be considered for each individual intervention but would likely not contribute to reduction in hypoglycemia or hyperglycemia across multiple hospitals. Interventions across multiple different hospital systems would help demonstrate the value of a mandatory order set beyond a single health care entity as well. Long-term studies are needed to assess the sustained impact of order sets and their effects on other outcomes (eg, length of stay, reduction, mortality, and costs). Another aspect of hypoglycemia prevention not directly included in this study that should be investigated in future studies includes use of predictive hypoglycemic tools,41, 42, 43 use of continuous glucose monitoring devices in the inpatient setting,44, 45, 46 use of digital protocols,15,47 and insulin calculators.14 Another area of future exploration for further hypoglycemia reduction could be the use of noninsulin medications versus insulin in a large sample of hospitalized patients. Use of liraglutide in patients with COVID-19 admitted to psychiatric units has demonstrated good results through case reports.48, 49, 50, 51 Identification of the correct population and utilization of these glucose-dependent antihyperglycemics in the hospital setting should be an area of further investigation moving forward. As the Centers for Medicare and Medicaid Service start to analyze institutional inpatient hypoglycemia and hyperglycemia, further investigation into its prevention is vital.
Conclusion
Through the implantation of a QI-based mandatory EMR insulin order set, there was a statistically significant decrease in the rates of hypoglycemia, hyperglycemia, and HbA1c capture across this 10-hospital system. The EMR insulin order set allowed providers to more effectively order weight-based insulin (basal and prandial) catered to specific nutritional status, empowered nursing, and enabled the ordering of consistent carbohydrate diets. Hypoglycemia treatment protocols were also standardized. These changes were implemented with a large-scale education effort, all of which likely contributed to reductions in hypoglycemia. These results are positive and demonstrate a potential framework for other large multihospital institutions to implement order sets to help reduce rates of hypoglycemia. This study was also able to demonstrate an effective strategy for implementation of a large-scale QI project across a large system in general, which has its own value. As outcomes regarding inpatient hypoglycemia and hyperglycemia management continued to be evaluated both at the institutional level and at the federal level (Centers for Medicare and Medicaid Services), successful reductions in inpatient hypoglycemia without compromising hyperglycemia rates are more relevant than ever. Further research in the form of better controlled trials is needed to understand the magnitude of effect of each of the variables that contributed to the success of this study.
Disclosure
The authors have no conflicts of interest to disclose.
Acknowledgment
The authors thank the many Chief Medical Officers, physicians, nurses and other staff, across the health care system, and to the colleagues in Pharmacy, Information Systems, and Care Redesign, who helped to create, implement, facilitate and oversee the diabetes pathway. The following people served on the committees that created and modified the order sets: Julie Aucoin, Jacob Blanton, Lioubov Boulkina, Tonya Burnette, John Buse, Linda Butler, Ginger Calhoun, Michaela Carnino, Cheri Clay, Shelley Conner, Steve Cotten, Paul Couch, Jessica Cunningham, Michelle Curl, Toni Currin, Ben Dellva, Darren Dewalt, Ron Falk , Lisa Farrior, Daniel Gibbs, Jose Guillem, Jewook Ha, Lisa Hall, Chad Hall, Andrew Hannapel, Julie Harris, Jason Heilemann, Luke Heuts, Kathryn Higdon, Laura House, Eddie Hu, Pamela Hughes, Taren Hunt, John Ikonomidis, Thomas Ivester, Rhi Johannsen, Kristen Johns, Karen Johnson, Kristin Johnson, Shelda Johnson, Carla Jones, Nancy Kirby, Nichole Korpi-Steiner, Rob Lampman, Trinh Le, David Lowry, Catherine Madigan, Rodney McCaskill, Katherine McGinigle, Erica McKearney, David McSwain, Ryan Merchant, Ryan Mills, Juan Mira, Antoinette Nelson, Megan O'Toole, Angela Overman, Claire Paris, Jennifer Patterson, Jasetta Perkins, Mary Roach, Joseph Roberts, Jenna Salter, Xanthia Samaropoulos, Amy Shaheen, Greg Shull, Steven Skovran, Dawn Smith, Nicole Speight, Denise Talbott, Jennifer Tsomides, Brandon Tyndall, Meera Udayakumar, Kayla Waldron, Francine Walker, Amanda Whitman, Mauri Williams, and Tramaine Young.
Author Contributions
L.B.M. and M.S.J. contributed equally to this manuscript. D.D. and C.P. were responsible for formal analysis and data curation. D.D was responsible for visualization and writing original draft, review, and editing. A. Goley, A. Garg, B.A., T.R., L.M., N.L., C.K., and C.P. were responsible for conceptualization and writing original draft and review. A. Goley, T.R. L.M., N.L., C.K., and C.P. were responsible for methodology. C.P. was responsible for resource, software, and validation.
Declaration of Generative AI and AI-Assisted Technologies in the Writing Process
During the preparation of this work the authors used Claude.ai in order to edit grammar. After using this tool/service, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.
Supplementary Material
References
- 1.Brutsaert E., Carey M., Zonszein J. The clinical impact of inpatient hypoglycemia. J Diabetes Complications. 2014;28(4):565–572. doi: 10.1016/j.jdiacomp.2014.03.002. [DOI] [PubMed] [Google Scholar]
- 2.Garg R., Hurwitz S., Turchin A., Trivedi A. Hypoglycemia, with or without insulin therapy, is associated with increased mortality among hospitalized patients. Diabetes Care. 2013;36(5):1107–1110. doi: 10.2337/dc12-1296. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Turchin A., Matheny M.E., Shubina M., Scanlon J.V., Greenwood B., Pendergrass M.L. 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]
- 4.Lipska K.J., Ross J.S., Wang Y., et al. National trends in US hospital admissions for hyperglycemia and hypoglycemia among Medicare beneficiaries, 1999 to 2011. JAMA Intern Med. 2014;174(7):1116–1124. doi: 10.1001/jamainternmed.2014.1824. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Hsu P.F., Sung S.H., Cheng H.M., et al. Association of clinical symptomatic hypoglycemia with cardiovascular events and total mortality in type 2 diabetes: a nationwide population-based study. Diabetes Care. 2013;36(4):894–900. doi: 10.2337/dc12-0916. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.van den Berghe G., Wouters P., Weekers F., et al. Intensive insulin therapy in critically ill patients. N Engl J Med. 2001;345(19):1359–1367. doi: 10.1056/NEJMoa011300. [DOI] [PubMed] [Google Scholar]
- 7.Finfer S., Chittock D.R., Su S.Y., et al. Intensive versus conventional glucose control in critically ill patients. N Engl J Med. 2009;360(13):1283–1297. doi: 10.1056/NEJMoa0810625. [DOI] [PubMed] [Google Scholar]
- 8.Kyi M., Colman P.G., Wraight P.R., et al. Early intervention for diabetes in medical and surgical inpatients decreases hyperglycemia and hospital-acquired infections: a cluster randomized trial. Diabetes Care. 2019;42(5):832–840. doi: 10.2337/dc18-2342. [DOI] [PubMed] [Google Scholar]
- 9.Maynard G., Wesorick D.H., O'Malley C., Inzucchi S.E., Society of Hospital Medicine Glycemic Control Task Force Subcutaneous insulin order sets and protocols: effective design and implementation strategies. J Hosp Med. 2008;3(5 Suppl):29–41. doi: 10.1002/jhm.354. [DOI] [PubMed] [Google Scholar]
- 10.Kao S.L., Chen Y., Ning Y., et al. Evaluating the effectiveness of a multi-faceted inpatient diabetes management program among hospitalised patients with diabetes mellitus. Clin Diabetes Endocrinol. 2020;6(1):21. doi: 10.1186/s40842-020-00107-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.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]
- 12.Yan Z., Wane Z., Xiaohua L., et al. Application of nursing quality improvement to reduce the incidence of hypoglycemia during colonoscopy procedure in diabetes patients. Am J Nurs. 2020;9(3):149–154. doi: 10.11648/j.ajns.20200903.23. [DOI] [Google Scholar]
- 13.Cook C.A., Vakayil V., Pribyl K., et al. A pharmacist-driven glycemic control protocol to reduce the rate of severe hypoglycemia in high-risk patients. Hosp Pharm. 2022;57(1):45–51. doi: 10.1177/0018578720973891. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Mehta P.B., Kohn M.A., Rov-Ikpah E., et al. Novel automated self-adjusting subcutaneous insulin algorithm improves glycemic control and physician efficiency in hospitalized patients. J Diabetes Sci Technol. 2024;18(3):541–548. doi: 10.1177/19322968241232673. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Câmara de Souza A.B., Toyoshima M.T.K., Cukier P., et al. Electronic glycemic management system improved glycemic control and reduced complications in patients with diabetes undergoing coronary artery bypass surgery: a randomized controlled trial. J Diabetes Sci Technol. 2024 doi: 10.1177/19322968241268352. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Lane A., McKnight M., Samson K., Nohner M., Drincic A. Reducing inpatient hypoglycemia: a diversified approach to a complex problem. Endocr Pract. 2025;31(9):1110–1115. doi: 10.1016/j.eprac.2025.05.744. [DOI] [PubMed] [Google Scholar]
- 17.Dakroub B., DiScala S.L., Vartan C.M., Miller M.F., Silverman M., Chalavarya S. Management of critical glucose values in hospice. Am J Hosp Palliat Care. 2024;41(10):1157–1160. doi: 10.1177/10499091231223748. [DOI] [PubMed] [Google Scholar]
- 18.Sinha Gregory N., Seley J.J., Gerber L.M., Tang C., Brillon D. Decreased rates of hypoglycemia following implementation of a comprehensive computerized insulin order set and titration algorithm in the inpatient setting. Hosp Pract (1995) 2016;44(5):260–265. doi: 10.1080/21548331.2016.1250603. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Mathioudakis N., Jeun R., Godwin G., et al. Development and implementation of a subcutaneous insulin clinical decision support tool for hospitalized patients. J Diabetes Sci Technol. 2019;13(3):522–532. doi: 10.1177/1932296818798036. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Cruz P. Inpatient hypoglycemia: the challenge remains. J Diabetes Sci Technol. 2020;14(3):560–566. doi: 10.1177/1932296820918540. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Umpierrez G.E., Smiley D., Zisman A., et al. Randomized study of basal-bolus insulin therapy in the inpatient management of patients with type 2 diabetes (RABBIT 2 trial) Diabetes Care. 2007;30(9):2181–2186. doi: 10.2337/dc07-0295. [DOI] [PubMed] [Google Scholar]
- 22.Kravchenko M.I., Tate J.M., Clerc P.G., et al. Impact of structured insulin order sets on inpatient hypoglycemia and glycemic control. Endocr Pract. 2020;26(5):523–528. doi: 10.4158/ep-2019-0341. [DOI] [PubMed] [Google Scholar]
- 23.Maynard G., Kulasa K., Ramos P., et al. Impact of a hypoglycemia reduction bundle and a systems approach to inpatient glycemic management. Endocr Pract. 2015;21(4):355–367. doi: 10.4158/ep14367.Or. [DOI] [PubMed] [Google Scholar]
- 24.Maynard G., Lee J., Phillips G., Fink E., Renvall M. Improved inpatient use of basal insulin, reduced hypoglycemia, and improved glycemic control: effect of structured subcutaneous insulin orders and an insulin management algorithm. J Hosp Med. 2009;4(1):3–15. doi: 10.1002/jhm.391. [DOI] [PubMed] [Google Scholar]
- 25.Curll M., Dinardo M., Noschese M., Korytkowski M.T. Menu selection, glycaemic control and satisfaction with standard and patient-controlled consistent carbohydrate meal plans in hospitalised patients with diabetes. Qual Saf Health Care. 2010;19(4):355–359. doi: 10.1136/qshc.2008.027441. [DOI] [PubMed] [Google Scholar]
- 26.American Diabetes Association Professional Practice Committee 16. Diabetes care in the hospital: standards of care in diabetes-2025. Diabetes Care. 2025;48(1 Suppl 1):S321–S334. doi: 10.2337/dc25-S016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Golden S.H., Hager D., Gould L.J., Mathioudakis N., Pronovost P.J. A gap analysis needs assessment tool to drive a care delivery and research agenda for integration of care and sharing of best practices across a health system. Jt Comm J Qual Patient Saf. 2017;43(1):18–28. doi: 10.1016/j.jcjq.2016.10.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Shelton C., Demidowich A.P., 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]
- 29.Gerwer J.E., Bacani G., Juang P.S., Kulasa K. Electronic health record-based decision-making support in inpatient diabetes management. Curr Diab Rep. 2022;22(9):433–440. doi: 10.1007/s11892-022-01481-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Wilson L.M., Herzig S.J., Marcantonio E.R., et al. Management of diabetes and hyperglycemia in the hospital: a systematic review of clinical practice guidelines. Diabetes Care. 2025;48(4):655–664. doi: 10.2337/dc24-2510. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.IBM Corp. Released . IBM Corp; NY: 2020. IBM SPSS Statistics for Windows VA. [Google Scholar]
- 32.Kulasa K., Juang P. How low can you go? Reducing rates of hypoglycemia in the non-critical care hospital setting. Curr Diab Rep. 2017;17(9):74. doi: 10.1007/s11892-017-0902-3. [DOI] [PubMed] [Google Scholar]
- 33.Ilcewicz H.N., Hennessey E.K., Smith C.B. Evaluation of the impact of an inpatient hyperglycemia protocol on glycemic control. J Pharm Sci. 2019;22(1):85–92. doi: 10.18433/jpps30180. [DOI] [PubMed] [Google Scholar]
- 34.Moolchandani P., Patel S., Larsen T., Moriates C., Weinreb J., Everett E. Targeting insulin stacking to address overnight hypoglycaemia in hospitalised patients with diabetes. BMJ Open Qual. 2025;14(1) doi: 10.1136/bmjoq-2024-003178. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Mulla C.M., Lieb D.C., McFarland R., Aloi J.A. Tides of change: improving glucometrics in a large multihospital health care system. J Diabetes Sci Technol. 2015;9(3):602–608. doi: 10.1177/1932296814563953. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.McIver F.B., Mitchell C.A., Finn C.P., Kamp M.C. Standardising practices through form design and education improves insulin management. Aust Health Rev. 2009;33(3):434–441. doi: 10.1071/ah090434. [DOI] [PubMed] [Google Scholar]
- 37.Odom J., Goldstein R. Inpatient hypoglycemic rate reduction through the implementation of prescriber targeted decision support tools. Curr Diab Rep. 2025;25(1):15. doi: 10.1007/s11892-024-01571-1. [DOI] [PubMed] [Google Scholar]
- 38.Schnipper J.L., Liang C.L., Ndumele C.D., Pendergrass M.L. Effects of a computerized order set on the inpatient management of hyperglycemia: a cluster-randomized controlled trial. Endocr Pract. 2010;16(2):209–218. doi: 10.4158/ep09262.Or. [DOI] [PubMed] [Google Scholar]
- 39.Wong B., Mamdani M.M., Yu C.H. Computerized insulin order sets and glycemic control in hospitalized patients. Am J Med. 2017;130(3):366.e1–366.e6. doi: 10.1016/j.amjmed.2016.09.034. [DOI] [PubMed] [Google Scholar]
- 40.Management of diabetes and hyperglycemia in hospitalized patients, Dhatariya K., Umpierrez G.E. https://www.ncbi.nlm.nih.gov/books/NBK279093/ [PubMed]
- 41.Hu X., Xu W., Lin S., Zhang C., Ling C., Chen M. Development and validation of a hypoglycemia risk model for intensive insulin therapy in patients with type 2 diabetes. J Diabetes Res. 2020;2020 doi: 10.1155/2020/7292108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Mathioudakis N.N., Abusamaan M.S., Shakarchi A.F., 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) doi: 10.1001/jamanetworkopen.2020.30913. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Elbaz M., Nashashibi J., Kushnir S., Leibovici L. Predicting hypoglycemia in hospitalized patients with diabetes: a derivation and validation study. Diabetes Res Clin Pract. 2021;171 doi: 10.1016/j.diabres.2020.108611. [DOI] [PubMed] [Google Scholar]
- 44.Flint K.L., O'Connor M.Y., Sabean A., et al. The association of continuous glucose monitoring metrics with hospital-related clinical outcomes in nonintensive care units. Diabetes Technol Ther. 2025 doi: 10.1089/dia.2024.0628. [DOI] [PubMed] [Google Scholar]
- 45.Olsen M.T., Hansen K.B., Pedersen-Bjergaard U., Kristensen P.L. Time at high risk of hypoglycemia: validation of a consensus-based continuous glucose monitoring-metric for hospitalized patients. J Diabetes Sci Technol. 2025 doi: 10.1177/19322968251331600. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Cavalcante Lima Chagas G., Teixeira L., M R.C.C., et al. Use of continuous glucose monitoring and point-of-care glucose testing in hospitalized patients with diabetes mellitus in non-intensive care unit settings: a systematic review and meta-analysis of randomized controlled trials. Diabetes Res Clin Pract. 2025;220 doi: 10.1016/j.diabres.2024.111986. [DOI] [PubMed] [Google Scholar]
- 47.Cui L., Schroeder P.R., Sack P.A. Inpatient and outpatient technologies to assist in the management of insulin dosing. Clin Diabetes. 2020;38(5):462–473. doi: 10.2337/cd20-0054. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Galindo R.J., Dhatariya K., Gomez-Peralta F., Umpierrez G.E. 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]
- 49.Soto-Chávez M.J., Muñoz-Velandia O.M., Alzate-Granados J.P., Lombo C.E., Henao-Carrillo D.C., Gómez-Medina A.M. Effectiveness and safety of new oral and injectable agents for in-hospital management of type 2 diabetes in general wards: systematic review and meta-analysis. Diabetes Res Clin Pract. 2022;191 doi: 10.1016/j.diabres.2022.110019. [DOI] [PubMed] [Google Scholar]
- 50.Ferreira E.V.M., Oliveira R.K.F., Salomao R., et al. Early use of liraglutide for the treatment of acute COVID-19 infection: an open-label single-center phase II safety study with biomarker profiling. Infect Dis Rep. 2025;17(1):5. doi: 10.3390/idr17010005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Zhang L., Yu W.J., Zhu H., Li H.F., Qiao J. Successful treatment of hyperglycemia with liraglutide in a hospitalized 27-year-old patient with schizophrenia: a case report. World J Clin Cases. 2022;10(21):7495–7501. doi: 10.12998/wjcc.v10.i21.7495. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.