7
Recommendation 5
➤ ES suggests continuation of personal CGM in the inpatient setting
with or without ADIP therapy rather than discontinuation. (2|⊕
)
Remarks:
▶ This should be done via a hybrid approach in which CGM use is combined with
periodic POC-BG testing to validate the accuracy of CGM.
▶ Inpatient CGM use is not currently approved by the U.S. Food and Drug
Administration (FDA) but currently has enforcement discretion. It has been used
in hospitals recently with Emergency Use Authorization during the COVID-19
pandemic.
Recommendation 6
➤ ES recommends that inpatient glycemic surveillance and management
programs leveraging EHR data be used for inpatients at risk for
hypoglycemia. (1|⊕
)
Remarks:
▶ The panel defined leveraging EHR data as specific hospital staff using glycemic data
collected within the EHR ( from all admitted patients) to identify those at risk for
and those having hypoglycemic and hyperglycemic episodes to develop mechanisms
for managing and mitigating these adverse outcomes. Standard care is lack of such a
program.
▶ The EHR data leveraged includes patterns of glycemia with proactive alerts for high
and for low trends, so that hypoglycemia and severe hyperglycemia can be identified
in a systematic fashion. Staff can then intervene upon these trends (e.g. adjusting
insulin infusion rates, etc) to avoid unwanted outcomes (repeat hypoglycemia,
glycemic variability, etc).
Recommendation 7
➤ ES suggests long-acting insulin analogs be used rather than human
NPH insulin for adult and pediatric outpatients on basal insulin
therapy who are at high risk for hypoglycemia. (2|⊕
)
Remarks:
▶ Patients who are at high risk for hypoglycemia are defined as those with a history
of severe hypoglycemia (requiring assistance to manage), IAH, and/or medical
conditions that predispose them to severe hypoglycemia including renal and hepatic
dysfunction.
▶ The panel placed high value on reducing severe hypoglycemia and found moderate-
certainty evidence for severe hypoglycemia reduction as an outcome in those using
long-acting analog insulins versus NPH insulin. However, the panel acknowledges
that most studies of long-acting analog insulins do not assess for significant adverse
effects (including cardiovascular outcomes) and that many studies were designed to
demonstrate non-inferiority of analog insulin compared with human NPH insulin.
