Table 9.
Seventy-eight Proposed Recommendations for the Guideline Voted on by the Panel.
| Continuation of home continuous glucose monitors after hospitalization | |
| Clinical Practice: Strong Recommendations | |
| • HCPs should consult with an inpatient diabetes team if available, when continuing or initiating a CGM or AID system. • HCPs should avoid relying on CGM data for glycemic management decisions in patients with severe hypoglycemia or hyperglycemia (ie, BG < 40 mg/dL or > 500 mg/dL). • HCPs should avoid using CGMs for management of (1) diabetic ketoacidosis until glucose is in the CGM measurement range, and then CGMs should be used adjunctively or (2) situations with rapidly changing glucose levels and fluid/electrolyte shifts. • HCPs should avoid continuing or initiating CGMs to patients with skin infections near the sensor site or placing sensors in areas with significant edema as well as patients treated with vasoactive agents or poor tissue perfusion. • HCPs should use a CGM checklist for elective procedures during the preoperative visits to ensure proper documentation of devices and real-time data reporting. • HCPs should advise pregnant women to continue the use of a CGM during a hospitalization to identify glucose trends and prevent hypo- or hyperglycemia. • HCPs should instruct patients to bring supplies with them to the hospital for the duration of any preplanned admission or elective procedures. • HCPs should check capillary BG or serum BG concentrations after procedures for noncritically ill patients and venous/arterial blood for critically ill patients to ensure the patient’s CGM is functioning properly. • HCPs should use trend arrows and rate of change to help prevent extreme glycemic excursions and (when a CGM is used adjunctively) to help determine when a BG test is required. • HCPs should set alarm thresholds for inpatient glycemic targets, such as predicting hypoglycemia (typically BG < 80-85 mg/dL) or predicting hyperglycemia. • Nursing should document CGM and/or CSII system information in the EHR for all admissions or elective procedures. | |
| Research: Strong Recommendations | |
| • Researchers need to provide more data to support definitive recommendations on improved outcomes for continuation of home/ambulatory CGM use after hospitalization. • Researchers need to conduct studies on the roles of CGM and POC BG testing and identify the optimal features of telemetry to inform nursing staff about actionable CGM patterns. • Researchers need to perform further studies to assess the accuracy of CGMs during pregnancy, labor and delivery, and the peripartum period. • Researchers need to study the impact of lag time on glucose measurements (ie, situations with rapid changes in the glucose concentration) in the hospital. | |
| Hospital Policies: Strong Recommendations | |
| • Hospitals need to develop standard CGM data reports and workflows. • Hospitals need to implement policies for testing capillary BGs and calibrating CGMs if the CGM requires calibration. • Hospitals need to develop a system for automatic staff notification for CGM alarms that predict impending or current hypoglycemia or hyperglycemia. • Hospitals need to develop specific guidelines for using CGMs and AID systems for their affiliated nursing homes and skilled nursing facilities. | |
| Initiation of continuous glucose monitors after hospitalization | |
| Clinical practice: strong recommendation | Clinical practice: mild recommendation |
| • HCPs should consider prescribing CGMs to reduce the need for frequent nurse contact for POC glucose testing and the use of PPE for patients on isolation with highly contagious infectious diseases (eg, COVID-19). | • HCPs should avoid initiating CGMs in patients with severe hypoglycemia or hyperglycemia (ie, BG < 40 mg/dL or > 500 mg/dL) or during periods of rapid glucose fluctuations. |
| Research: Strong Recommendations | |
| • Researchers need to provide data to support initiation of CGMs for improving patient-centered outcomes. • Researchers need to provide data on hospital outcomes when initiating CGMs in the hospital, including improved glycemic outcomes, detection and/or reduction of hypoglycemia and hyperglycemia, reduction of ICU LOS, and cost-effectiveness. • Researchers need to conduct studies on long-term benefits for initiating CGMs in the hospital after discharging patients with newly diagnosed diabetes or recurrence of diabetic ketoacidosis or other complications of diabetes. • Manufacturers need to develop educational tools for patients, hospital staff, and HCPs. | |
| Hospital Policies: Strong Recommendations | |
| • Hospitals need to develop plans, including process maps, protocols, staff educational resources, and order sets for prescribing CGM use during hospitalizations before implementing a CGM. • Hospitals need to provide educational tools for patients, nurses, house staff, and attending physicians when a patient in the hospital starts on a CGM. | |
| Continuation of automated insulin dosing systems in the hospital | |
| Clinical practice: strong recommendations | Clinical practice: mild recommendation |
| • HCPs should prescribe AID systems only for appropriate candidates, who will need to have adequate knowledge and skills for using AID systems. • HCPs should reassess a decision periodically to transition use of outpatient AID systems into the hospital in order to ensure that AID system continue to represent the best treatment option for each patient. • HCPs should prepare an alternative plan for diabetes management in case it becomes inappropriate for a patient to continue using an AID system in the hospital. • HCPs should discontinue AID systems in critically ill hospitalized patients (such as those with hypovolemia or sepsis). • HCPs should recognize glycemic patterns due to CGM compression, which can cause false low readings. |
• HCPs should avoid initiating an AID system during a hospitalization. |
| Research: Strong Recommendations | |
| • Researchers need to conduct studies about whether continuing AID systems in the hospital is beneficial to improve glycemic management or clinical outcomes. • Researchers need to provide data on hospital outcomes when using AID systems in the hospital, including improved glycemic outcomes, detection and/or reduction of hypoglycemia, reduction of ICU LOS, and cost-effectiveness. • Manufacturers need to research whether all types of CGMs and AID systems can be used during radiological/imaging studies or diathermy. | |
| Hospital Policies: Strong Recommendations | |
| • Hospitals need to develop institution-specific protocols and order sets for the proper use of AID systems during a hospitalization. • Hospitals need to require that patients using AID systems bring with them sufficient supplies for these devices during a hospitalization. • Hospitals need to develop protocols for using AID systems during elective procedures and surgeries. | |
| Recommendation Not Reaching Consensus | |
| • HCPs should switch AID systems from “auto” mode to “manual” mode when a patient is admitted to the hospital wearing an AID system. | |
| Logistics and hands-on care of hospitalized patients using continuous glucose monitors and automated insulin dosing systems | |
| Clinical Practice: Strong Recommendations | |
| • HCPs should inquire about and document the medication and supplement history of patients who use CGMs to determine whether there are any agents that can interfere with glucose measurements. • HCPs should ensure that off-label use of CGMs and AID systems is consistent with medical practice and appropriate precautions have been taken to protect patients. • Nursing should document hands-on training of CGM use and AID system therapy through a technology certification program. • Nursing should confirm that the patient is appropriate to continue using a CGM or an AID system and also review the agreement and hospital policy with the patient. • Nursing should inspect the insertion site every shift with attention to skin integrity and signs of erythema or infection, and should document site changes. • Nursing should know device basics, institutional policies, HCPs’ roles, and whom to contact if questions arise. • Nursing should administer a patient competency assessment or survey to assess patient ability to safely assist with managing a CGM or an AID system. • Nursing should set expectations and clarify that there will be a need to continue checking POC capillary glucose even when using a CGM. • Nursing should measure POC BG concentrations to confirm or supplement CGM readings (usually a minimum of four times daily: before each of three meals and at bedtime if patients are eating, or every six hours if patients are fasting) as well as at patient request; however, the CGM glucose, trend arrows, and rate of change may be used to help determine if and when a BG test is required. | |
| Research: Strong Recommendations | |
| • Researchers need to conduct further studies on the best logistics and hands-on care for patients using CGMs and AID systems to achieve the best outcomes. • Manufacturers need to research interoperable components for AID systems that are compatible with hospital EHRs. | |
| Logistics and hands-on care of hospitalized patients using continuous glucose monitors and automated insulin dosing systems | |
| Hospital Policies: Strong Recommendations | |
| • Hospitals need to provide interpreter services to translate CGM and AID system agreements. • Hospitals need to state in their policy and patient agreement documents that treatment decisions will be based on hospital-calibrated BGM readings (or laboratory readings) and not on CGM readings, barring a need to isolate a patient with a severe and highly contagious infection. • Hospitals need to maintain their CGM and AID system policy and patient agreement documents in easily accessible electronic files stored in the EHR order set for CGMs and AID systems. • Hospitals need to develop policies for when to discontinue or temporarily suspend the use of CGMs and AID systems. • Hospitals need to survey their HCPs, nursing, and patients to improve outcomes and satisfaction. | |
| Data management of continuous glucose monitors and automated insulin dosing systems in the hospital | |
| Clinical practice: strong recommendation | Clinical practice: mild recommendation |
| • HCPs should develop a set of core data elements and definitions for CGM data for inclusion in common data models and the EHR. | • Nursing should contact an HCP immediately when CGM results cross critical value thresholds set by the institution. |
| Research: strong recommendations | Research: mild recommendation |
| • Researchers need to conduct further studies on the best data management practices of CGMs and AID systems. • Researchers need to develop and validate robust glucose telemetry systems for both ICU and non-ICU populations. • Researchers need to develop a diabetes technology maturity model that helps institutions understand the requirements to successfully integrate diabetes-related data and technology. • Researchers need to develop, disseminate, and validate CGM- and AID system-specific PRO measures to improve patient care. • Manufacturers need to research methods for quality control for CGMs and AID systems, which is critical as part of inpatient use of CGMs and AID systems. • Manufacturers need to research optimally expanded device labeling in order to overcome clinical inertia and align practice with regulatory policy. • Manufacturers need to research systems for integration of CGM data following initial upload into the cloud (eg, the Eversense CGM) subsequently into the EHR. • Manufacturers need to research secure communication systems for protecting data from wireless wearables, telemedicine systems, and Bring-Your-Own-Device portable computers used by HCPs (also known as “data in motion”). |
• Researchers need to develop computerized insulin decision support system that will integrate with CGMs. |
| Hospital Policies: Strong Recommendations | |
| • Hospitals need to develop appropriate security protocols, dedicated data storage, visualization tools, and adequate cyber insurance coverage (also known as “data at rest”). • Hospitals need to integrate AID system data into the EHR system for nursing and HCPs to have easy access to this information. • Hospitals need to determine the number of laboratory or POC BG tests that must be performed while patients are using CGMs or AID systems in the hospital. • Hospitals need to adopt the Unique Device Identifier system for healthcare facilities to track devices in the EHR. • Hospitals need to identify CGM data reports in the patient’s EHR to distinguish them from laboratory glucose results. • Hospitals need to present clear criteria to clinicians to identify data that will require intervention. • Hospitals need to implement CGM- and AID system-specific PROs to improve patient care. • Hospitals need to develop a universal platform for their EHRs that can be used by all CGMs to present core data elements, summary glucometrics, consistent formats, and uniform interfaces across all CGM products. • Hospitals need to arrange for CGM results to be automatically uploaded into the EHR. • Hospitals need to manage CGM data with the same safety and security measures as all other PHI. • Hospitals need to develop policies for CGM and AID system use with atypical scenarios outside of diabetes, when glucose monitoring is valuable. | |
AID, automated insulin dosing; BG, blood glucose; CGM, continuous glucose monitoring; CSII, continuous subcutaneous insulin infusion; EHR, electronic health record; HCP, health care practitioner; ICU, intensive care unit; LOS, length of stay; POC, point of care; PPE, personal protective equipment; PRO, patient-reported outcome.