The results from the National Readmission Database also show that deaths are higher among those who have 30-day readmissions, and that hypertension, chronic kidney disease (CKD), anemia, and female sex were specific risk factors for readmission. On the other hand, obesity and hyperlipidemia were associated with lower readmission rates.
“The 20% [readmission] rate was quite surprising and quite high by any metrics,” said lead investigator Hafeez Shaka, MD, during a virtual press briefing. “Efforts should be channeled towards identifying these predictors in hospitalized patients as well as proper discharge planning to decrease the burden of readmissions,” he added.
And, he told Medscape Medical News in an interview, “Clinicians should start discharge planning from the time of admission as there has been a trend towards shorter hospital stays for these patients.”
“More should be done to ensure patient education especially about DKA and its precipitants,” stressed Shaka, resident physician at John H. Stroger Jr. Hospital of Cook County, Chicago, Illinois.
“Joint decision making should be employed to promote adherence to an insulin regimen that works rather than what is thought to be ideal. Closer post-hospital follow up might be necessary.”
Asked to comment, session moderator Vijay Shivaswamy, MBBS, of the Division of Endocrinology at the University of Nebraska Medical Center, Omaha, said he wasn’t surprised by the one in five readmission rate.
“It is alarming, absolutely. I always tell my patients that I hope you don’t miss any insulin, but if you’re going to miss something, don’t miss the long-acting. That’s going to keep you out of the hospital.”
He added that although this is a “great study” that used a large database — it included more than 90,000 adults hospitalized with DKA in 2017 — the lack of information on factors such as adherence, socioeconomic status, and other issues probably limit the clinical utility of the findings without further study, noting, “I think it’s more exploratory than actionable.”
Those Readmitted Were Twice as Likely to Die
DKA, which can occur following insufficient insulin or from infection, is a dangerous accumulation of ketones in the blood due to excess glucose. When severe, DKA can require hospitalization for fluid replacement and insulin therapy.
“DKA is a feared complication of type 1 diabetes because it can lead to diabetic coma and death,” noted Shaka, who presented the findings March 21 at virtual ENDO 2021.
Shaka and colleagues analyzed data for 91,625 adults aged 18 years and older with type 1 diabetes admitted to US hospitals for DKA between January 1 and November 30, 2017.
Of 91,401 who were discharged alive, 20.2%, or 18,553 individuals, were readmitted within 30 days. For most, DKA was the reason for readmission.
Compared with the index DKA admission, those who were readmitted within 30 days were twice as likely to die, with a risk ratio of 2.06 (P < .001). Length of stay for readmission was an average of 1 day longer (P < .001), at an increased average cost of $8217 (P < .001).
Independent predictors of readmission included female sex (hazard ratio [HR], 1.14), discharge against medical advice (HR, 1.54), hypertension (HR, 1.28), chronic kidney disease (HR, 1.13), and anemia (HR, 1.42); all were significant (P < .001).
However, lower readmission rates were seen among individuals with obesity (HR, 0.70; P < .001) and hyperlipidemia (HR, 0.92; P = .007).
During the briefing, Shaka speculated that although previous data have shown a survival advantage for people with obesity in hospital settings, those mechanisms probably don’t apply to patients with type 1 diabetes hospitalized with DKA.
More likely, he said, is that in type 1 diabetes both obesity and hyperlipidemia may be more common in individuals who take more insulin to maintain glycemic control, while those with less stable glucose levels tend to be underweight or within normal weight.
“Obesity may be a marker for getting more insulin…That’s a hypothesis. It remains to be tested to see if there might be another pathophysiologic basis for it,” Shaka commented.
More Questions Than Answers…
And as for why females may be more likely than males to be readmitted for DKA, Shaka said: “Females tend to be more health-seeking…They could present sooner if they feel they’re going into DKA.” He also noted that the female predominance persisted even after adjustment for age and comorbidities, and that this finding is consistent with previous studies.
“An interesting hypothesis would be to assess if there is a difference in outcomes during hospitalizations based on sex,” he said.
Reiterating that quite a few important factors weren’t included in the data, such as individual components of hyperlipidemia, CKD stage, and, notably, socioeconomic status, Shivaswamy cautioned: “It leaves you with more questions than answers.”
Nonetheless, Shaka said, “More can always be done for the patient with type 1 diabetes either while in the hospital or on discharge. This would save lives as well as costs.”
Shaka has reported no relevant financial relationships. Shivaswamy is site principal investigator for trials funded by Novo, Lilly, and Kowa Pharmaceuticals.
ENDO 2021. Abstract OR09. Presented March 21, 2021.