Diabetes mellitus: hyperosmolar state: increased risk with acute infection or new-onset diabetes.

Clinical bottom line (level 3b)

  1. Patients with diabetes were more likely to be admitted to hospital in a diabetic hyperosmolar state if they had the following predisposing factors:
    • female
    • no previous history
    • presence of acute infection
Watchel et al: Archives of Internal Medicine 1987; 147: 499-501
Expires August 2004

The study

Case-control study with objective outcomes, adjusted for confounding factors, not validated in an independent set of patients.

Setting: two university hospitals, USA; 1974-84

270 patients (aged mean 69 years, 64% female) diabetes
Cases: 135 patients (71% female, mean age 69): non-ketotic hyperosmolar coma, hyperosmolar dehydration or diabetic ketoacidosis
Controls: 135 patients (53% female, mean age 69): diabetes


Multivariate logistic regression performed on predisposing factors.

Outcomes studied:
  • predisposing factors: concurrent disease (e.g. heart failure, alcoholism, dementia, previous stroke), medication (e.g. thiazides, furosemide, beta-blockers, steroids), insulin therapy, sex

The evidence

  • independent predisposing factors for diabetic hyperosmolar state:
    • female
    • no history of diabetes
    • presence of acute infection

Comments

  1. No odds ratios given for predisposing factors.
  2. No information was give for the reasons for admission of the controls: a better control group may have been patients with DKA without hyperosmolality.

Citation

  1. Watchel TJ, Silliman RA, Lamberton P, et al: predisposing factors for the diabetic hyperosmolar state. Archives of Internal Medicine 1987; 147: 499-501
Search Terms: hyperosmol* and diabet* in Cochrane
Contributor: Chris Ball and Musab Hayatli, August 2000
Reviewer:

Clinical Question.
Patient diabetes
Intervention or Exposure predisposing factors
Outcome admitted in a hyperosmolar state