Abstract

Basal bolus insulin therapy (BB) is recommended for glucose control in noncritically ill inpatients based on two small RABBIT2 trails. However, BB is complex and has limited data in large scale “real world” settings. This is a retrospective analysis of 4,558 noncritical, medical-surgical admissions from 1/2013 to 9/2015 in patients ≥18 years who received SC but not IV insulin for ≥ 75% of the stay. Insulin therapy was identified within the first 2 days as BB (long + rapid, n=2319), sliding scale (SS, rapid only, n=1855) or Basal Only (BO, long only, n=384). A propensity score analysis adjusted for clinical, demographic and physician variables affecting choice of therapy. Point of care glucose (mg/dl) was used to define patient days as hypo- (any <70), hyper- (mean >180) or euglycemic (no <70 AND mean ≤180). Negative binomial regression analyzed glucose outcomes (number of hypo-, hyper- and euglycemic days, mean glucose); linear regression modeled log of length of stay (LOS); logistic regression modeled odds of 30 and 60 day readmission. All analyses were conducted overall and by medical or surgical DRG. BB had worse glucose control but less LOS in medical DRG. BO had better hyper-, euglycemic and mean glucose. BB and BO had more hypo- days. This suggests that BO may be superior and simpler than BB for hyperglycemia but medical patients may benefit from BB for reasons beyond glucose control. Insulin groupsOVERALL (n= 4,558)MEDICAL DRG (n =3,211)SURGICAL DRG (n= 1,347)Length of stay*BB vs SS0.97 (0.141)0.94 (0.003)1.02 (0.540)BO vs SS1.01 (0.711)0.99 (0.741)1.03 (0.665)Hyperglycemic days*BB vs SS0.99 (0.700)0.97 (0.416)1.00 (0.922)BO vs SS0.69 (<0.001)0.69 (<0.001)0.69 (<0.001)Hypoglycemic days*BB vs SS2.45 (<0.001)2.29 (<0.001)2.78 (<0.001)BO vs SS2.71 (<0.001)2.37 (<0.001)3..40 (<0.001)Euglycemic days*BB vs SS0.91 (0.008)0.85 (<0.001)1.01 (0.823)BO vs SS1.15 (0.012)1.11 (0.151)1.20 (0.063)Mean glucose*BB vs SS1.01 (0.415)1.69 (0.266)-1.18 (0.589)BO vs SS-17.27 (<0.001)-17.95 (<0.001)-15.91 (<0.001)1Results for 30 and 60 day readmission are not significantly different across the groups and are not presented in the table;*Results above shown as: Estimate (p-value) Disclosure A.R. Sadhu: None. B. Patham: None. A. Vadhariya: None. M.L. Johnson: None.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call