LAS VEGAS — Continued use of personal insulin pump therapy during hospitalization for adult patients with diabetes was associated with significantly better glucose control and shorter length of stay than pump discontinuation, new research showed. 

The data suggest that standardized protocols should be developed for inpatient insulin pump use, study author Alaa Almallouhi, MD, a second-year internal medicine resident at Naples Comprehensive Health, Naples, Florida, told Medscape Medical News. 

“Inpatient endocrinology consult is not available at all hospitals, and not all hospitalists have the experience to manage insulin pumps. But if there were a standardized hospital protocol, the regular hospitalist would know what to do. I think we need to work on developing these, the same as we have for [diabetic ketoacidosis], and many other things we deal with in the hospital,” said Almallouhi, who presented the study at the American Association of Clinical Endocrinology Annual Meeting. 

This in line with the recommendation of the American Diabetes Association’s Standards of Care–2026: “Continue use of insulin pump including automated insulin delivery in people with diabetes who are hospitalized when clinically appropriate. This is contingent upon availability of necessary supplies, resources, training, ongoing competency assessments, and implementation of institutional diabetes technology protocols.”

Nonetheless, there is still wide variability across hospitals and health systems regarding inpatient management of patients who use these devices, Almallouhi reported. Pumps are often discontinued due to safety concerns, she noted. 

Better Glycemic Control, Shorter Hospital Stay

The retrospective study included 59 insulin pump-wearing patients hospitalized at Naples Comprehensive Healthcare System between November 2022 and November 2025. Of those, the insulin pump was discontinued in 35 (60%) and continued in 24 (40%). Those who continued using their pumps also kept using their continuous glucose monitor (CGM); some patients who discontinued the pump kept their CGM devices on but some did not, Almallouhi told Medscape. 

There were no statistically significant baseline differences between the two groups in age (67.9 years in those who continued on the pump, 65.5 years for those discontinued), sex (58.3% vs. 45.7% men, respectively), BMI (28.8 vs. 27.8), proportion who had type 2 diabetes (58.3% vs. 51.4%), or the percentage with infections at admission (16.7% vs. 25.7%).

Among those whose pumps were discontinued, reasons for the discontinuation were not documented in 74.2%. Pump failure was the reason for 20% and diabetic ketoacidosis for 5.7%. 

“This suggests that in many cases, discontinuation may reflect workflow or documentation gaps rather than a recorded clinical contraindication,” Almallouhi noted. 

Time in range (100-180 mg/dL) was significantly greater in the pump continuation group, at 58.1%, vs 39% in the discontinuation group (P = .008), and mean inpatient glucose was significantly lower in the pump continuation group, at 164.5 mg/dL, vs 193.9 mg/dL in the discontinuation group (P = .030). Time in hypoglycemia (< 70 mg/dL) in the pump continuation group was 0.84% vs 2.92% in the discontinuation group (P = .094). 

Length of hospital stay was significantly shorter with pump continuation, 2.6 vs 4.4 days, P = .012. “That’s nearly a 2-day reduction in hospitalization, suggesting that better glycemic stability may have translated into more efficient inpatient care,” she said. 

Pump continuation was also associated with fewer adverse clinical events, including hypoglycemia (4.3% in the pump continuation group vs 16.7% in pump discontinued group), 30-day readmission (0 vs 13.9%), and 30-day emergency department visit (0 vs 13.9%), although the study was underpowered to detect significant differences in those events. 

The data “support safe continuation in selected patients,” Almallouhi concluded. 

Asked what criteria she uses to determine eligible patients for pump continuation, Almallouhi replied, “You want to make sure the patient is mentally okay, and that they can self-manage their pump. They should not be taking medication that can cause sleepiness or alter their mental status.” 

‘A Lot of Education’ Needed for Staff

Asked to comment, session moderator Viral N. Shah, MD, professor of medicine and director of diabetes clinical research at the Center for Diabetes and Metabolic Diseases, Indiana University School of Medicine, Indianapolis, told Medscape Medical News that allowing patients to wear their personal automated insulin delivery systems — an insulin pump and CGM that communicate to semi-automate insulin delivery — could produce better outcomes “provided that there is a good understanding between the patient, the healthcare provider, and the staff.” 

He pointed out that one limitation of the study is the use of chart reviews of electronic health records data. “The chart review depends on how physicians are writing, so if someone is just in a hurry and forgets to write that the patient is on a pump, it is missed,” he explained. 

He also noted that even with protocols, great care would be necessary to minimize the many human errors that could occur, such as a patient wearing a pump mistakenly receiving injected insulin, leading to hypoglycemia. 

Moreover, staff would need to be educated about all the different systems currently on the market. 

“The pump is not just a pump,” said Shah. “It’s an automated insulin delivery system. There are five on the [US] market, and each one works differently. So I think the hospitalist, the inpatient staff, and service people need a lot of education about this. It’s going to take some time.”

Almallouhi reported no relevant financial relationships. 

Shah’s institute has received research funding from Lilly, Enable Biosciences, Zucara Therapeutics, Cystic Fibrosis Foundation, and Breakthrough T1D. He personally has received honoraria from Sanofi, Novo Nordisk, Lilly, Dexcom, Insulet, Tandem Diabetes Care, Medtronic, Sequel Med Tech, Abbott Diabetes, Roche, Biomea Fusion, and T1D Scout for advising, consulting, and/or speaking. 

Miriam E. Tucker is a freelance journalist based in the Washington, DC area. She is a regular contributor to Medscape, with other work appearing in the Washington Post, NPR’s Shots blog, and Diatribe. She is on X (formerly Twitter) @MiriamETucker and BlueSky @miriametucker.bsky.social