Assessment of dextrose 50 bolus versus dextrose 10 infusion in the management of hyperkalemia in the emergency department

Journal Pre-proofs Assessment of Dextrose 50 Bolus versus Dextrose 10 Infusion in the Management of Hyperkalemia in the Emergency Department Irene Yang, Samantha Smalley, Tania Ahuja, Cristian Merchan, Silas W. Smith, John Papadopoulos PII: DOI: Reference:

S0735-6757(19)30615-1 https://doi.org/10.1016/j.ajem.2019.09.003 YAJEM 158513

To appear in:

American Journal of Emergency Medicine

Received Date: Revised Date: Accepted Date:

15 July 2019 24 August 2019 20 September 2019

Please cite this article as: I. Yang, S. Smalley, T. Ahuja, C. Merchan, S.W. Smith, J. Papadopoulos, Assessment of Dextrose 50 Bolus versus Dextrose 10 Infusion in the Management of Hyperkalemia in the Emergency Department, American Journal of Emergency Medicine (2019), doi: https://doi.org/10.1016/j.ajem.2019.09.003

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2019 Published by Elsevier Inc.

Assessment of Dextrose 50 Bolus versus Dextrose 10 Infusion in the Management of Hyperkalemia in the Emergency Department Irene Yang, PharmDa1; 550 First Ave, New York, NY 10016 [email protected] Samantha Smalley, PharmD, BCPSa; 550 First Ave, New York, NY 10016 [email protected] Tania Ahuja, PharmD, BCPSa; 550 First Ave, New York, NY 10016 [email protected] Cristian Merchan, PharmD, BCCCPa; 550 First Ave, New York, NY 10016 [email protected] Silas W. Smith, MDb; 550 First Ave, New York, NY 10016 [email protected] John Papadopoulos, BS, PharmD, FCCM, BCCCP, BCNSPa 550 First Ave, New York, NY 10016 [email protected] aDepartment

of Pharmacy, NYU Langone Health, 550 First Ave, New York, NY 10016, USA O. Perelman Department of Emergency Medicine, NYU Langone Health, 550 First Ave, New York, NY 10016, USA 1Current Address: Department of Pharmacy, Hackensack University Medical Center, 30 Prospect Avenue, Hackensack, NJ 07601, USA bRonald

Corresponding author Samantha Smalley, PharmD, BCPS; Emergency Medicine Clinical Pharmacotherapy Specialist; 545 1st Avenue, GBH-SC2-097, New York, NY 10016; Phone 212-263-3826; [email protected] Conflicts of interest statement: The author(s) declare no conflicts of interest with respect to the research, authorship and/or publication of this article Key words (3-6 words): Hyperkalemia, hypoglycemia, insulin, dextrose, drug shortage Funding: This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors Running Head: D50 vs. D10 to prevent hypoglycemia from IV insulin

Abstract Introduction Hypoglycemia is a common adverse effect when intravenous (IV) insulin is administered for hyperkalemia. A prolonged infusion of dextrose 10% (D10) may mitigate hypoglycemia compared to dextrose 50% (D50) bolus. Our objective was to evaluate whether D10 infusion is a safe and effective alternative to D50 bolus for hypoglycemia prevention in hyperkalemic patients receiving IV insulin.

Methods We conducted a retrospective review of patients ≥18 years who presented to the emergency department (ED) with hyperkalemia (K+ >5.5) and received IV insulin and D10 infusion or D50 bolus within 3 hours. The primary endpoint was incidence of hypoglycemia, defined as blood glucose (BG) ≤70mg/dL, in the 24 hours following IV insulin administration for hyperkalemia.

Results A total of 134 patients were included; 72 in the D50 group and 62 in the D10 group. There was no difference in incidence of hypoglycemia between the D50 and D10 groups (16 [22%] vs. 16 [26%], p=0.77). Symptomatic hypoglycemia, severe hypoglycemia, and hyperglycemia rates in the D50 and D10 groups were [5 (7%) vs. 2 (3%), p=0.45], [5 (7%) vs. 1 (2%), p=0.22], and [34 (47%) vs. 23 (37%), p=0.31] respectively. Low initial BG was a predictor for developing hypoglycemia.

Conclusions

In our study, D10 infusions appeared to be at least as effective as D50 bolus in preventing hypoglycemia in hyperkalemic patients receiving IV insulin. In context of ongoing D50 injection shortages, D10 infusions should be a therapeutic strategy in this patient population.

Abbreviations

IV: Intravenous D10: Dextrose 10% in water D50: Dextrose 50% in water BG: Blood glucose ED: Emergency Department K+: Potassium AHA: American Heart Association CPOE: Computerized SCr: Serum Creatinine CrCl: Creatinine Clearance REDCap: Research Electronic Data Capture SPSS: Statistical Package for Social Sciences IQR: Interquartile ranges

1. INTRODUCTION Hyperkalemia is a potentially life-threatening metabolic abnormality [1]. Insulin, which shifts K+ intracellularly, is administered in up to 64.4% of cases of severe hyperkalemia to temporarily reduce serum levels [2,3]. The reported incidence of hypoglycemia when insulin is administered in this setting is variable, ranging from 8.7 to 75% [4-6]. For severe hyperkalemia (K+ >6.5 mmol/L), the American Heart Association (AHA) recommends 10 units of IV regular insulin with 50 mL of D50 [7]. However, to our knowledge, studies have not been conducted to evaluate the effects of different dextrose concentrations and infusion durations on hypoglycemia rates in management of hyperkalemic patients. Hypoglycemia is a known adverse effect based on the pharmacological profiles of IV insulin and D50. The contrasting durations of effect between IV insulin and D50 bolus may contribute to hypoglycemia in hyperkalemic patients. The effects of IV insulin begin in approximately 10 to 15 minutes and terminate at a median of about 4 hours (range: 1 to 5 hours). However, blood glucose (BG) levels can return to baseline after 30 minutes following a D50 bolus [8 – 9]. In addition, peripheral administration of D50, a hyperosmolar medication, may be associated with phlebitis and extravasation injuries [9]. Use of a lower osmolarity agent, such as D10, may decrease these adverse effects. From a drug shortage perspective, D50 injections are commonly unavailable from wholesalers, forcing clinicians to rely on alternative or contingency hypoglycemic management strategies such as D10 infusions. Beginning April 2017, our health system was unable to obtain sufficient supply of D50 injections from our wholesaler supplier. Subsequently, we implemented an alternative use alert in our computerized physician order entry (CPOE; Epic, Epic Systems Corporation, Verona, Wisconsin) system, recommending D10 infusion over D50. Therefore, we sought to evaluate whether D10 infusion is a safe and effective alternative to D50 bolus for prevention of hypoglycemia in hyperkalemic patients treated with IV insulin.

2. METHODS 2.1 Study Design and Setting This was an institutional review board approved, retrospective review conducted at three urban emergency departments (ED) within a single health-system. Due to the retrospective nature of this review, informed consent was not required by the institutional review board. Chart reviews were conducted by a pharmacy resident and overseen by three clinical pharmacy specialists. The pharmacy resident was trained with our data collection tool prior to commencement of data abstraction; periodic monitoring for accuracy of data collection was completed by a clinical pharmacist. 2.2 Selection of Patients Electronic health records (EHRs) were used to identify patients who presented to the ED and received IV insulin with concomitant dextrose for the management of hyperkalemia, defined as a K+ >5.5. At our institution, there is no standardized protocol for the management of hyperkalemia. IV insulin is typically administered with 25 to 50 g of dextrose, as either a D50 bolus or D10 intermittent infusion. Inclusion criteria consisted of patients ≥18 years of age who received a D50 bolus or a D10 infusion lasting a minimum of 30 minutes within 3 hours of a noted K+ >5.5 mmol/L between August 20, 2016 and August 20, 2018. Our D50 injection shortage began in April 2017. Patients who received ≥25 g of dextrose for hypoglycemia prevention received all dextrose as one single dose without spaced dosing. Although providers were routed to select D10 in our CPOE during our D50 shortage, this alert could be overridden leading to more patients receiving D50 compared to D10 in our study period. Therefore, D50 patients were selected in a 1:3 ratio utilizing a random number generator to keep a balanced sample size. Exclusion criteria included receipt of a D50 bolus and D10 infusion simultaneously or a lack of at least 1 follow-up BG. Data collection included baseline characteristics,

administered doses of insulin and dextrose, medications with dextrose-containing diluents, and laboratory values including serum creatinine (SCr), K+ levels and BG levels. Creatinine clearance (CrCl) was calculated using the Cockcroft-Gault equation [10]. Data was managed with REDCap (version 8.10.18; Research Electronic Data Capture, Nashville, Tennessee) [11]. 2.3 Primary and Secondary Outcomes The primary endpoint was the incidence of hypoglycemia, defined as BG ≤70 mg/dL in the 24 hours following IV insulin administration for hyperkalemia. Secondary endpoints included the incidences of severe hypoglycemia, symptomatic hypoglycemia, hyperglycemia, use of rescue agents for the management of hypoglycemia, and extravasation. Severe hypoglycemia was defined as a BG ≤40 mg/dL. Symptomatic hypoglycemia was defined as documented symptoms in the EHR attributed to hypoglycemia and treated with a carbohydrate or rescue agent, regardless of BG level. Rescue agents were defined as glucose gel, glucose tablets, carbohydrates in the form of snacks or a meal, intramuscular glucagon, D10 infusion or D50 bolus with the intent to treat hypoglycemia. Hyperglycemia was defined as a BG ≥180 mg/dL. Adverse events included those adverse events documented in the EHR, such as extravasation, attributed to dextrose administration. 2.4 Statistical Analysis Statistical analysis was performed using SPSS (software for Windows, version 25.0; Statistical Package for Social Sciences, Chicago, Illinois) comparing patients who received D50 with those who received D10. Categorical variables were described as frequencies and proportions and compared using Chi-square test or Fisher’s exact test. Continuous variables were described as medians with interquartile ranges (IQR) and analyzed using Mann-Whitney U test. Timing of hypoglycemic episodes was compared using a paired-samples sign test. A twosided alpha of <0.05 was considered statistically significant. A univariate analysis was conducted to determine predictors of hypoglycemia. Baseline characteristics including renal

function, insulin doses and dextrose doses were included in the univariate analysis. A power analysis was not performed due to the lack of previous similar studies.

3. Results 3.1 Characteristics of Study Participants A total of 140 patients presenting to the ED with a K+ >5.5 mmol/L who received insulin and dextrose were screened for inclusion. Two were excluded for lack of at least 1 follow-up BG; 4 were excluded for initial simultaneous administration of D50 and D10. Therefore, 134 patients were included for statistical analysis (n = 72, D50; n = 62, D10; Figure 1). There were no differences in baseline characteristics (Table 1). Most patients in the D50 and D10 groups had renal dysfunction, with an elevated SCr (3.3 mg/dL [IQR 1.9, 8.0] vs. 3.1 mg/dL [IQR 1.7, 6.3], p=0.18), respectively. Prior to insulin administration, patients receiving D10 had lower median BG versus those receiving D50 (112 mg/dL [IQR 89, 133] vs. 123 mg/dL [IQR 100, 177], p=0.023). The median initial dextrose dose was 25 g in both groups (25 g [IQR 25, 50] vs. 25 g [IQR 25, 25], p=0.022), with a median D10 infusion rate of 250 mL/hr (IQR 250, 625) (Table 2).

3.2 Primary Outcome No difference was observed in the incidence of hypoglycemia between the D50 and D10 groups (16 [22%] vs. 16 [26%], p=0.769; Figure 2). The median incidence of hypoglycemic events per patient was numerically higher in the D50 group (2 [IQR 1, 3] vs. 1 [IQR 1, 2]; p=0.168). The median time to the first hypoglycemic event occurred earlier in the D50 group (3.8 h [IQR 1.3, 16] vs. 6.8 [IQR 1.6, 8.9]; p=0.873) (Table 3, Figure 3).

3.3 Secondary Outcomes Symptomatic hypoglycemia, severe hypoglycemia, and hyperglycemia rates between D50 and D10 were: 5 (7%) vs. 2 (3%), 5 (7%) vs. 1 (2%), and 34 (47%) vs. 23 (37%),

respectively. Incidence and timing of rescue agent administration were similar in the D50 and D10 groups ([15% vs. 16% patients, p=1], [3.8 vs. 4.0 h, p=0.515]) (Table 3).

3.4 Univariate Analysis for Predictors of Hypoglycemia In the univariate analysis, baseline BG ≤100 mg/dL was identified as a risk factor for developing hypoglycemia (OR 4.90, 95% CI 2.09-11.49). More patients who experienced hypoglycemia were ≥60 years old compared to those who did not (21 [64%] vs. 12 [12%]; p= 0.195). We did not see any differences in rates of hypoglycemia when examining baseline characteristics including weight, gender, diabetes status, renal function, initial dose of insulin or dextrose administered (Table 4). 4. Discussion In our study, we found the incidence of hypoglycemia for the management of hyperkalemia to be 24% at 24h following IV insulin administration. The reported incidence of hypoglycemia with IV insulin for the management of hyperkalemia in the ED predominantly ranges from 8.7% to 17% at 4 to 8h post insulin administration [4 – 6]. This may explain the high rates of hypoglycemia in our study as we may have captured delayed hypoglycemia events from IV insulin, which other studies did not. Our incidence of hypoglycemia within 8h of insulin administration was 13%, which is similar to the rates reported in literature [4 – 5, 12 – 13]. We aimed to identify patient characteristics for the risk of hypoglycemia. Similar to a recent study, we found low pre-treatment BG to be a predictor of hypoglycemia [13]. We also found age ≥60 years to be a predictor of hypoglycemia. Although other studies have found higher doses of insulin, no history of diabetes mellitus, female gender, abnormal renal function and low body weight to be predictive of hypoglycemia, we did not observe these findings in our limited sample size [4 – 5, 12, 14]. The ideal strategy to decrease hypoglycemic events in hyperkalemic patients receiving IV insulin has evaded clinicians to date. One study exploring dextrose dosing reported lower

rates of hypoglycemia 1h after IV insulin administration comparing 50 g versus 25 g, but no difference at 4h [15]. However, limited data exists on alternative methods of dextrose administration in the hyperkalemia setting. Other authors have retrospectively evaluated patients receiving continuous infusions of D10 500mL over 4h for the management of hyperkalemia and reported hypoglycemia in 6.1%, suggesting D10 to be a safe and effective modality [12]. However, all patients received 10 units of rapid acting insulin and no comparator arm was present. Other existing literature utilizing D10 infusions for hypoglycemia management is not specific to hyperkalemia and derived from the prehospital setting [16]. To our knowledge, our study is the first to compare a D10 infusion to a D50 bolus strategy for hypoglycemia prevention in patients receiving IV insulin for hyperkalemia. We aimed to compare these approaches due to the frequent drug shortages of D50 injections. The Joint Commission recommends identification of an appropriate alternative medication as a potential strategy to combat drug shortages [17]. Starting April 2017, due to continued shortages of D50, our institution began utilizing D10 infusions as an alternative agent for preventing hypoglycemia in the treatment of hyperkalemia with IV insulin. In addition to being a potential viable alternative to D50 injections, there are several theoretical advantages to utilizing D10 infusions. Besides prolonging the anti-hypoglycemic effects of dextrose, D10 infusion may be associated with decreased stimulation of additional insulin release compared to D50 bolus. When D50 is used specifically for the treatment of sulfonylurea-induced hypoglycemia, adverse insulin release may be induced as excess glucose suppresses gluconeogenesis and glycogenolysis [18]. While this mechanism may be beneficial for the management of hyperkalemia, it could potentially result in rebound hypoglycemia. Furthermore, there is a lower the risk of thrombophlebitis and extravasation with D10 as the osmolarity of D10 is 506 mOsm/L, as compared to D50 which has an osmolarity of 2,525 mOsm/L. Reported cases of D50 extravasation have been severe, with some patients requiring treatment with hyaluronidase and emergent fasciotomy. [19, 20].

5. Limitations There are several limitations applicable to our study. First, this was a retrospective chart review at a single-center on the basis of an unanticipated drug shortage. A power analysis was not performed due to the limited information available from previous studies. We had a small sample size and this may have limited the ability to find a difference in hypoglycemic rate. No institutional protocol for management of hyperkalemia or hypoglycemia in this context existed at the time of our study; therefore, prescriber insulin dosing and hypoglycemia treatment was not constrained and BG checks were not performed in a standardized fashion. Patients who received D10 during initial hyperkalemia management may have received D50 for hypoglycemia treatment (and vice-versa) within 24 hours. Lastly, report of symptomatic hypoglycemia was based on chart review and subject to accurate documentation.

6. Conclusion Hypoglycemia in the setting of insulin administration for hyperkalemia management remains problematic. We found D10 infusion to be an acceptable alternative to bolus D50 injection for hypoglycemia prevention. Although the optimum total glucose dose remains to be determined, D10 infusion may lead to fewer incidents of dysglycemia. D10 administration warrants consideration, especially in light any reported D50 injection shortages. The ideal D10 infusion rate and duration to minimize hypoglycemia should continue to be explored. Patients with low initial BG are at higher risk of hypoglycemia and require more aggressive glucose treatment and monitoring. References: [1.] Pierce DA, Russell G, Pirkle JL, Jr. Incidence of Hypoglycemia in Patients With Low eGFR Treated With Insulin and Dextrose for Hyperkalemia. Ann Pharmacother. 2015;49(12):1322-1326. [2.] Singer AJ, Thode HC, Jr., Peacock WF. A retrospective study of emergency department potassium disturbances: severity, treatment, and outcomes. Clin Exp Emerg Med. 2017;4(2):73-79.

[3.] Li T, Vijayan A. Insulin for the treatment of hyperkalemia: a double-edged sword? Clin Kidney J. 2014;7(3):239-241. [4.] Schafers S, Naunheim R, Vijayan A, Tobin G. Incidence of hypoglycemia following insulin-based acute stabilization of hyperkalemia treatment. J Hosp Med. 2012;7(3):239-242. [5.] Apel J, Reutrakul S, Baldwin D. Hypoglycemia in the treatment of hyperkalemia with insulin in patients with end-stage renal disease. Clin Kidney J. 2014;7(3):248-250. [6.] Estep P, Efird LE (2015) Evaluation of Hypoglycemia Incidence and Risk Factors in Patients Treated with IV Insulin Aspart for Hyperkalemia. Endocrinol Diabetes Res 1:1. [7.] Vanden Hoek TL, Morrison LJ, Shuster M, et al. Part 12: cardiac arrest in special situations: 2010 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Circulation. 2010;122(18 Suppl 3):S829-861. [8.] Hipszer B, Joseph J, Kam M. Pharmacokinetics of intravenous insulin delivery in humans with type 1 diabetes. Diabetes Technol Ther. 2005;7(1):83-93. [9.] Reynolds PM, MacLaren R, Mueller SW, Fish DN, Kiser TH. Management of extravasation injuries: a focused evaluation of noncytotoxic medications. Pharmacotherapy. 2014;34(6):617-632. [10.] Botev R, Mallie JP, Couchoud C, Schuck O, Fauvel JP, Wetzel JF, et al. Estimating glomerular filtration rate: Cockcroft-Gault and Modification of Diet in Renal Disease formulas compared to renal inulin clearance. Clin J Am Soc Nephrol. 2009;4(5):899-906. [11.] Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)--a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42(2):377-381. [12.] Coca A, Valencia AL, Bustamante J, Mendiluce A, Floege J. Hypoglycemia following intravenous insulin plus glucose for hyperkalemia in patients with impaired renal function. PLoS One. 2017;12(2):e0172961. [13.] Jacob BC, Peasah SK, Chan HL, Niculas D, Shogbon Nwaesei A. Hypoglycemia associated with insulin use during treatment of hyperkalemia among emergency department patients. Hosp Pharm 2019; 54(3): 197–202. [14.] Moussavi K, Fitter S, Gabrielson SW, Koyfman A, Long B. Management of Hyperkalemia With Insulin and Glucose: Pearls for the Emergency Clinician. J Emerg Med. 2019. doi: 10.1016/j.jemermed.2019.03.043

[15.] Farina N, Anderson C. Impact of dextrose dose on hypoglycemia development following treatment of hyperkalemia. Ther Adv Drug Saf. 2018;9(6):323-329. [16.] Moore C, Woollard M. Dextrose 10% or 50% in the treatment of hypoglycaemia out of hospital? A randomised controlled trial. Emerg Med J. 2005;22(7):512-515. [17.] The Joint Commission. Medication management (MM) (critical access hospitals/critical access hospitals). Medication selection and procurement—managing medication shortages. What are The Joint Commission requirements for managing medication shortages? 2017. https://www.jointcommission.org/standards_information/jcfaqdetails.aspx?StandardsFAQId=1530&Sta ndardsFAQChapterId=27&ProgramId=0&ChapterId=0&IsFeatured=False&IsNew=False&Keyword=. Accessed May 16, 2019. [18.] McLaughlin SA, Crandall CS, McKinney PE. Octreotide: an antidote for sulfonylurea-induced hypoglycemia. Ann Emerg Med. 2000;36(2):133-138. [19.] Lawson SL, Brady W, Mahmoud A. Identification of highly concentrated dextrose solution (50% dextrose) extravasation and treatment--a clinical report. Am J Emerg Med. 2013;31(5):886.e3-5. [20.] Chinn M, Colella MR. Prehospital Dextrose Extravasation Causing Forearm Compartment Syndrome: A Case Report. Prehosp Emerg Care. 2017;21(1):79-

Table 1 Baseline characteristics and past medical history D50 Bolus (n=72) Demographics Age (yr), median (IQR) 71 (61, 78) Male 46 (64) Weight (kg), median (IQR) 74 (61, 90) Height (cm), median (IQR) 168 (160, 173) Hospital admission 67 (93) Past Medical History Renal Function SCr at time of initial K+ (mg/dL) 3.3 (1.9, 8.0) CrCl (mL/min) 13 (8, 29) CrCl ≤30 mL/min 56 (78) ESRD 19 (26) Hemodialysis 14 (19) AKI* 48 (67) Diabetes 37 (51) Type I 1 (1) Type II 34 (47) Not stated 2 (3) Heart failure 25 (35) HFrEF 11 (15) HFpEF 14 (19) NPO at time of hyperkalemia 17 (24) Prior To Admission Medications Causing Hypoglycemia Diabetes Medications Medication route 13 (18) Oral 13 (18) IV 6 (8) Combination Oral and IV Medication classes 19 (26) Insulin 6 (8) Sulfonylurea 9 (13) Metformin 1 (1) Thiazolidinedione 11 (15) DPP4 Inhibitors 1 (1) GLP-1 Agonists 2(3) Meglinitides 0 (0) SGLT-2 Inhibitors 0 (0) Alpha-glucosidase inhibitors 0 (0) Pramlinitide Other medications which can cause hypoglycemia Pentamidine 1 (1) Sulfamethoxazole-Trimethoprim 1 (1)

D10 Infusion (n=62)

p-Value

67 (57, 76) 40 (65) 71 (62, 84) 167 (163, 174) 51 (83)

0.243 0.94 0.906 0.641 0.055

3.1 (1.7, 6.3) 21 (8, 41) 42 (68) 15 (24) 14 (23) 38 (61) 23 (37) 1 (2) 19 (31) 3 (5) 18 (29) 8 (13) 8 (13) 8 (13)

0.18 0.085 0.191 0.771 0.656 0.589 0.097 0.915 0.050 0.662 0.482 0.694 0.308 0.113

11 (18) 7 (11) 2 (3)

0.276 0.962 0.285

13 (21) 2 (3) 3 (5) 1 (2) 4 (7) 0 (0) 1 (2) 0 (0) 0 (0) 0 (0)

0.544 0.285 0.121 1 0.106 1 1

0 (0) 4 (6)

1 0.182

All data presented as (n, %) unless otherwise noted AKI defined as: increase in SCr by ≥ 0.3 mg/dL or ≥ 1.5 times baseline IQR: interquartile range; CrCl: creatinine clearance; ESRD: end stage renal disease; AKI: acute kidney injury; HFrEF: heart failure with reduced ejection fraction; HFpEF: heart failure with preserved ejection fraction; NPO: nothing by mouth; DPP4: dipeptidyl peptidase-4; GLP-1: glucose-dependent insulinotropic peptide; SGLT-2: sodium-glucose cotransporter-2

Table 2 IV insulin and IV dextrose doses and administrations Potassium (mmol/L) Initial K+ Repeat K+ after insulin treatment IV Insulin Regular Dose For initial hyperkalemia treatment (units) For initial hyperkalemia treatment (units/kg) Patients receiving repeat insulin for hyperkalemia treatment (n, %) For repeat hyperkalemia treatment (units) For repeat hyperkalemia treatment (units/kg) Time to repeat insulin administration (hours) Total insulin administered for hyperkalemia treatment in 24 hours (units) Total insulin administered for hyperkalemia treatment in 24 hours (units/kg) Dextrose Administration Dextrose administered during initial hyperkalemia treatment (g) Total dextrose administered during hyperkalemia treatment in 24 hours (g) Median D10 infusion rate (mL/hr)

D50 Bolus (n=72)

D10 Infusion (n=62)

p-Value

6.4 (6.2, 7.1) 5.5 (5.0, 6.4)

6.4 (6.1, 6.9) 5.6 (5.0, 6.1)

0.621 0.994

8 (5, 10)

9 (5, 10)

0.995

0.11 (0.07, 0.13)

0.10 (0.07, 0.14)

0.995

15 (21)

16 (26)

0.542

10 (7.5, 10)

9 (5, 10)

0.281

0.11 (0.08, 0.12)

0.11 (0.06, 0.14)

0.74

7.8 (5.1, 9.3)

6.5 (4.9, 13.2)

0.83

10 (5, 10)

10 (5, 10)

0.574

0.12 (0.07, 0.16)

0.12 (0.08, 0.18)

0.643

25 (25, 25)

25 (25, 50)

0.022

25 (25, 50)

50 (25, 50)

0.075

All data presented as median (IQR) unless otherwise specified IQR: interquartile range; kg: killogram

250 (250, 625)

Table 3 Blood glucose, incidence of hypoglycemia and safety BG prior to insulin administration BG (mg/dL) Baseline hypoglycemia* (n, %) Baseline BG ≤100 mg/dL Time from BG to insulin administration (hours) First BG post insulin administration BG (mg/dL) First BG ≤100 mg/dL, (n, %) Time to first BG check post insulin (hours) Primary Outcome Incidence of hypoglycemia (n, %) Time of first hypoglycemic event (hours) Hypoglycemia events per patient Patients with more than one hypoglycemia event (n, %) Secondary Outcomes Symptomatic hypoglycemia (n, %) Severe hypoglycemia (n, %) Hyperglycemia (n, %) Other adverse events, i.e., extravasation (n, %) Rescue items given (n, %)

D50 Bolus (n=72)

D10 Infusion (n=62)

p-Value

123 (100, 177) 2 (3) 18 (25) 0.8 (0.4, 1.6)

112 (89, 133) 2 (3) 26 (42) 1.3 ( 1.6, 2.0)

0.023 1 0.041 0.528

120 (104, 190) 17 (24) 1.3 (0.7, 3.2)

135 (96, 202) 18 (29) 1.6 (1.0, 2.4)

0.918 0.476 0.899

16 (22) 3.8 (1.3, 16)

16 (26) 6.8 (1.6, 8.9)

0.769 0.873

2 (1, 3) 10 (14)

1 (1, 2) 7 (11)

0.168 0.652

5 (7) 5 (7) 34 (47)

2 (3) 1 (2) 23 (37) 0 (0)

0.45 0.216 0.307

10 (16) D10 bolus (n=10) 0 (0) 5 (50) 1 (10) 3 (3) 25 (25, 25) 1 (10) 50 4.0 (2.1, 10.7) 1 (1, 1.3)

1

0 (0) 11 (15) D50 infusion (n=11) 1 (9) 2 (18) 0 (0) 7 (64) 50 (50, 50) 2 (18) 50 (50, 271) 3.8 (2.1, 13.8) 1 (1, 1)

Glucose tablets (n, %) Snack or meal (n, %) Glucagon (n, %) D50 bolus (n, %) g, median D10 infusion (n, %) g, median Time to initial rescue (hours) Number of interventions for each hypoglycemia encounter

1 0.249 0.463 0.339 1 0.515 0.606

*Hypoglycemia defined as BG ≤70 mg/dL All data presented as median (IQR) unless otherwise noted IQR: interquartile range; BG: blood glucose; g: grams

Table 4 Univariate predictors of hypoglycemia Hypoglycemia (n=33)

No hypoglycemia (n=101)

p-Value

Age ≥60 years

21 (64)

12 (12)

0.195

Female

11 (33)

37 (37)

0.731

Weight ≤60 kg

23 (70)

81 (80)

0.209

Baseline Characteristics

Hemodialysis

7 (21)

21 (21)

0.959

Diabetes Type I Type II Heart failure HFrEF HFpEF

14 (42) 1 (3) 11 (33) 9 (27) 6 (18) 2 (6) 6 (18)

46 (46) 1 (1) 42 (42) 34 (34) 13 (13) 20 (20) 19 (19)

0.754 0.401 0.400 0.495 0.448 0.064 0.936

17 (52)

18 (18)

0.001 OR 4.90, 95% CI 2.0911.49)

CrCl ≤ 30 mL/min

27 (82)

71 (71)

0.195

AKI

22 (67)

64 (64)

0.731

ESRD

10 (30)

24 (24)

0.453

20 (61)

53 (53)

0.415

21 (64)

67 (67)

0.777

6 (18)

25 (25)

0.437

25 (76)

68 (67)

0.362

NPO at time of hyperkalemia

BG ≤ 100 mg/dL

Renal function

Insulin Administration Initial insulin dose ≥0.1 units/kg Total insulin dose ≥0.1 units/kg Patients receiving repeat insulin doses Dextrose Administration Initial dextrose dose ≤25g

All data presented as (n, %) AKI defined as: increase in SCr by ≥ 0.3 mg/dL or ≥ 1.5 times baseline OR: odds ratio; CI: confidence interval; CrCl: creatinine clearance; AKI: acute kidney injury; BG: blood glucose; ESRD: end stage renal disease; HFpEF: heart failure with preserved ejection fraction; HFrEF: heart failure with reduced ejection fraction; kg: kilograms; mg: milligrams; NPO: nothing by mouth