Nondiabetes Hospitalization Is Wrong Time to Up Diabetes Meds

Marlene Busko

November 01, 2021

"Short-term hospitalization [for reasons other than diabetes] may not be the time to intervene in long-term diabetes management," researchers conclude.

They found that, in a national cohort of older almost entirely male veterans with non–insulin-treated type 2 diabetes who were hospitalized for non–diabetes-related common medical conditions, intensified diabetes treatment on hospital discharge was linked to an increased risk of severe hypoglycemia in the immediate postdischarge period.

However, diabetes treatment intensification – that is, receiving a prescription for a new or higher dose of diabetes medicine – was not associated with decreased risks of severe hyperglycemia or with improved glycemic (hemoglobin A1c) control at 30 days or 1 year, according to study results, published in JAMA Network Open.

"We didn't see a reduction in diabetes emergencies in more intensively treated patients," lead investigator Timothy S. Anderson, MD, said in an interview.

Also, importantly, there was a low rate of persistence with the new treatment. "Half of the patients were no longer taking these [intensified diabetes medicines] at 1 year, which tells me that context is key," he pointed out. "If a patient is in the hospital for diabetes [unlike the patients in this study], I think it makes a lot of sense to modify and adjust their regimen to try to help them right then and there."

The overall risk of severe hyperglycemia or severe hypoglycemia was pretty small in the overall cohort, Anderson noted, "but we do put people at risk of leaving the hospital and ending up back in the hospital with low blood sugar when we intensify medications, and there's not necessarily a good signal to suggest that it's all that urgent to change these medicines."

Instead, the "safer path" may be to make recommendations to the patient's outpatient physician and also inform the patient – for example, "We saw some concerns about your diabetes while you were in the hospital, and this is really something that should be looked at when you're recovered and feeling better from the rest of your health standpoint" – rather than making a diabetes medication change while the person is acutely ill or recovering from illness, said Anderson, from Beth Israel Deaconess Medical Center and Harvard Medical School, Boston.

The researchers also found an "unexpected" significant decrease in 30-day mortality in the patients with intensified diabetes treatment, which was probably because of confounding that was not accounted for, Anderson speculated, since clinical trials have consistently shown that benefits from diabetes medications take a longer time to show an effect.

"Important Study," but Lacked Newer Meds

This is an "important" study for primary care and in-hospital physicians that shows that "hospitalization is really not the time and the place" to intensify diabetes medication, Rozalina G. McCoy, MD, coauthor of an invited commentary, told this news organization in an interview.

"While overcoming treatment inertia is important, [it should be] done appropriately, so that we don't overtreat patients," McCoy, of the Mayo Clinic in Rochester, Minn., stressed.

The very low rate of persistence of taking intensified medications is a major finding, she agreed. Hospitalized patients "are not in their usual state of health, so if we make long-term treatment decisions based on their acute abnormal situation, that may not be appropriate."

However, patients with high A1c may benefit from a change at hospital discharge rather than when they see their primary care provider, with the caveat that they need close follow-up as an outpatient.

The study emphasizes the "need for longitudinal patient care rather than episodic patches," according to McCoy.

For example, a patient who is hospitalized for a chronic obstructive pulmonary disease or asthma exacerbation may be receiving steroids that cause high blood glucose levels but as soon as they're done with their steroid course, blood glucose will decrease, so the "need for close outpatient follow-up is very important."

One limitation of the current work is that an earlier study in the same population by the research group showed that 49% of patients whose treatment regimens were intensified had limited life expectancy or were at or below their A1c goal, so they would not have benefited from the stepped-up treatment, she noted.

Another limitation is that the findings cannot be generalized to women or younger patients, or to patients treated with glucagonlike peptide 1 (GLP-1)–receptor agonists or sodium-glucose cotransporter 2 (SGLT2) inhibitors.

The study patients were seen in the U.S. Veterans Health Administration health system when these newer agents were not used. Three-quarters of patients received intensified treatment with sulfonylurea or insulin, and only one patient received a new GLP-1–receptor agonist.

Ideally, McCoy said, patients should have been prescribed a GLP-1–receptor agonist if they had atherosclerotic cardiovascular disease or kidney disease, or an SGLT2 inhibitor if they had kidney disease or heart failure, which may have led to different results, and would need to be determined in further study.

Anderson agreed that "SGLT2 inhibitors and GLP1 agonists are broadly much safer than the older diabetes medicines, at least when it comes to risk of hypoglycemia, and may have more clear benefits in heart disease and mortality. So I would not want to extrapolate our findings to those new classes," he said. "A similar set of studies would need to be done."

Study Rationale and Findings

Hospitalized older adults with diabetes commonly have transiently elevated blood glucose levels that might lead clinicians to discharge them from hospital with a prescription for more intensive diabetes medications than they were on before they were hospitalized, but it is not clear if these diabetes medication changes would improve outcomes.

To investigate this, the researchers analyzed data from patients with diabetes who were 65 and older and hospitalized for common medical conditions in VHA hospitals during January 2011–September 2016, and then discharged to the community.

They excluded patients who were hospitalized for things that require immediate change in diabetes treatment and patients who were using insulin before their hospitalization (because instructions to modify insulin dosing frequently don't have a new prescription).

The researchers identified 28,198 adults with diabetes who were not on insulin and were hospitalized in the VHA health system for heart failure (18%), coronary artery disease (13%), chronic obstructive pulmonary disease (10%), pneumonia (9.6%), and urinary tract infection (7.5%), and less often and not in decreasing order, for acute coronary syndrome, arrhythmia, asthma, chest pain, conduction disorders, heart valve disorders, sepsis, skin infection, stroke, and transient ischemic attack.

Of these patients, 2,768 patients (9.8%) received diabetes medication intensification, and the researchers matched 2,648 of these patients with an equal number of patients who did not receive this treatment intensification.

The patients in each group had a mean age of 73 and 98.5% were male; 78% were White.

They had a mean A1c of 7.9%. Most were receiving sulfonylurea (43%) or metformin (39%), and few were receiving thiazolidinediones (4.1%), alpha-glucosidase inhibitors (2.7%), dipeptidyl peptidase 4 inhibitors (2.0%), or other types of diabetes drugs (0.1%).

Of the 2,768 patients who received intensified diabetes medication, most received a prescription for insulin (51%) or sulfonylurea (23%).

In the propensity-matched cohort, patients with intensified diabetes medication had a higher rate of severe hypoglycemia at 30 days (1% vs. 0.5%), which translated into a significant twofold higher risk (hazard ratio, 2.17).

The rates of severe hypoglycemia at 1 year were similar in both groups (3.1% and 2.9%).

The incidence of severe hyperglycemia was the same in both groups at 30 days (0.3%) and 1 year (1.3%).

In secondary outcomes, at 1 year, 48% of new oral diabetes medications and 39% of new insulin prescriptions were no longer being filled.

Overall, patients who were discharged with intensified diabetes medication were significantly less likely to die within 30 days than the other patients (1.3% vs. 2.4%; HR, 0.55).

However, this mortality benefit was found only in the subgroup of 2,524 patients who had uncontrolled diabetes when they were admitted to hospital (A1c >7.5%; mean A1c, 9.1%), and not in the propensity-matched subgroup of 2,672 patients who had controlled diabetes then (A1c up to 7.5%; mean A1c, 6.8%).

There was no significant difference in 1-year mortality in patients with versus without intensified treatment (15.8% vs. 17.8%).

There were also no significant between-group difference in rates of hospital readmission at 30 days (roughly 17%) or 1 year (roughly 51%).

The decreases in mean A1c from hospital discharge to 1 year later were also the same in both groups (going from 7.9% to 7.7%).

The study was funded by grants from the National Institute on Aging and the American College of Cardiology. Anderson has no relevant financial disclosures. McCoy reported receiving grants from the National Institute of Diabetes and Digestive and Kidney Diseases, AARP, and the Patient-Centered Outcomes Research Institute outside the submitted work. The disclosures of the other authors and the editorial coauthor are available with the article and commentary.

This article originally appeared on, part of the Medscape Professional Network.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.