Table 1.

Summary of Studies on the Treatment of PN-Associated Hyperglycemia

Article	Population	Design	Protocol	Key Outcomes	Conclusion
Jakoby and Nannapaneni, 2012 (38)	Patients with or without diabetes in the ICU or non-ICU setting (n = 48)	Prospective cohort study comparing outcomes with historical control subjects	TDD for prandial insulin was determined based on I/D ratio of 1:20 for those without diabetes and 1:5–10 for those with diabetes and those on steroid therapy. Group 1: 66% of prandial insulin TDD was administered as RHI in the PN bag and 33% as NPH at intervals of 6–8 hours; additional weight-based NPH dose (0.15 units/kg/day if admission blood glucose was <200 mg/dL or 0.25 units/kg/day if admission blood glucose was >200 mg/dL) was added to the prandial NPH insulin for basal coverage in patients with diabetes or on steroids (n = 22). Group 2: historical control group was given RISS, RHI added to the PN bag, or basal (NPH/glargine) insulin (n = 26).	Group 1 had better mean blood glucose (138 ± 37 vs. 159 ± 46 mg/dL, P <0.0001) and spent more time in range (60 vs. 35% of time with blood glucose 80–140 mg/dL, P <0.0001) compared with Group 2. Hypoglycemia was infrequent in both groups but higher in Group 1.	Protocol-directed daily insulin dosing linking insulin to dextrose yielded better glycemic control than relying on dosing strategies based on supplemental/sliding-scale insulin alone.
Hakeam et al., 2017 (36)	Non-ICU patients with diabetes who underwent cardiac surgery (n = 67)	Prospective, randomized, open-label study	Insulin TDD was calculated based on the previous day RISS requirement and given as follows: Group 1: 80% as glargine insulin along with RISS every 6 hours (n = 35) Group 2: 80% as RHI in the PN bag (n = 32) Subsequent dose adjustments were made based on RISS requirement on the previous day and blood glucose level.	Blood glucose control (<180 mg/dL) was achieved in 52.9% of glargine group and 47.76% of group receiving RHI in the PN bag (P = 0.06). Mean blood glucose was similar in both the groups on days 5–9, but patients receiving RHI in the PN bag reached the target blood glucose level sooner. Patients in the glargine group had a higher percentage of blood glucose >180 mg/dL on day 5 compared with those receiving RHI in the PN bag (22.39 vs. 5.97%, P = 0.0059). Fewer blood glucose levels >234 mg/dL were noted in the group receiving RHI in the PN bag. Six hypoglycemic events were noted: two with glargine (5.7%) and four with RHI in the PN bag (11.4%; P >0.1).	Both glargine and RHI in PN are effective for controlling PN-induced hyperglycemia in patients with diabetes. Adding RHI to the PN bag reaches the glucose goal faster with fewer hyperglycemic episodes.
Ramos et al., 2018 (33)	Surgical patients with or without diabetes in the non-ICU setting (n = 80)	Retrospective record-based review	Group 1: weight-based glargine insulin (0.4 units/kg for those with diabetes [n = 41] and 0.3 units/kg for those without diabetes [n = 39]) along with correction lispro insulin every 6 hours with 10–20% increase or decrease in dose every day to achieve blood glucose <180 mg/dL Group 2: none	50% of the patients achieved target blood glucose of <180 mg/dL. Hypoglycemia (blood glucose <70 mg/dL) was seen in 22.5% of the study population.	Subcutaneous basal-plus-correction insulin can be used to achieve glycemic control, but frequent dose adjustments should be made. Hypoglycemia was higher because of unplanned interruption of TPN and lack of communication with the endocrinology team.
Fatati et al., 2018 (34)	Patients with or without diabetes in the ICU or non-ICU setting (n = 26)	Retrospective record-based review	Group 1: 13 patients with or without diabetes were followed for 7 days; TDD was calculated based on I/D ratio of 1:10; 50% of TDD was given as degludec insulin and uptitrated accordingly with the remainder given as RHI along with correction doses for blood glucose >250 mg/dL. RHI dose was reduced as degludec insulin is increased. Group 2: none	Mean blood glucose on day 1 versus day 7: In patients without diabetes: 151 ± 47.3 versus 157 ± 66.7 mg/dL In patient with diabetes: 210 ± 66.5 versus 192 ± 48.6 mg/dL In patients with diabetes, blood glucose was within target (<180 mg/dL) for 4 days and higher during the last 3 days.	Insulin degludec has been shown to maintain stable metabolic control. Difficulty in achieving targets in patients with diabetes was possibly the result of suboptimal application of the insulin titration protocol.
Olveira et al., 2020 (37)	Patients with diabetes in the non-ICU setting (n = 161)	Prospective RCT	TDD was estimated based on weight (0.2–0.5 units/kg). Group 1: 100% TDD given as RHI added to the PN bag as basal and nutrition component (n = 80) Group 2: TDD divided into 50% as RHI added to the PN bag (nutrition component) and 50% as basal insulin glargine (n = 81) RISS was given every 6 hours in both groups, and two-thirds of the total correction was added daily to the previous regimen in both groups.	Mean blood glucose was similar in both groups (165.3 ± 35.4 mg/dL in the RHI-only group vs. 172.5 ± 43.6 mg/dL in the RHI-plus-glargine group; P = 0.25). Mean blood glucose was lower in the RHI-plus-glargine group within 2 days after PN discontinuation (160.3 ± 45.1 mg/dL in the RHI-only group vs. 141.7 ± 43.8 mg/dL in the RHI-plus-glargine group; P = 0.024). The RHI-plus-glargine group had a significantly higher rate of nonsevere hypoglycemia (blood glucose <70 mg/dL) (11.2% in the RHI-only group vs. 27.2% in the RHI-plus-glargine group; P = 0.016).	Both groups showed similar glycemic control, although Group 2 (RHI-plus-glargine) had better metabolic control after PN was interrupted.

RISS, regular insulin sliding scale.