a b s t r a c t

Purpose: To characterize hypoglycemia management and identify characteristics associated with refractory (need for additional treatment following initial management) and recurrent (adequate initial treatment followed by blood glucose [BG] <=50 mg/dL) hypoglycemia.

Methods: Retrospective review of adult emergency department (ED) patients who presented to a large academic medical center with hypoglycemia (BG <= 50 mg/dL) between January 2011 and July 2015. Data collection focused on BG measurements and Treatment practices. Data are reported using descriptive statistics, Wilcoxon rank sum, and ?² analysis as appropriate.

Results: Two hundred forty-four patients were included (mean age, 59 +- 18.7 years; weight, 85 +- 24.3 kg). Pa- tients arriving via prehospital care (n = 124) were assessed faster in the ED (median, 25 minutes; interquartile range [IQR], 10-40 minutes) compared with ambulatory arrival (median, 43 minutes; IQR, 17-95 minutes; P =

.0018). There were 174 patients with a BG <= 50 mg/dL in the ED. Of those, 108 (62.1%) were treated with intra- vascular bolus dextrose/intramuscular glucagon and 21 patients (12%) did not receive any treatment or food. The overall median time to treatment after identification of hypoglycemia was 12 minutes (IQR, 6-27.8 minutes); treatment was administered faster after bedside Point-of-care testing assessment compared with when serum samples resulted (11 [IQR, 6-23.5] minutes vs 25 [IQR, 10.75-42.5] minutes, respectively; P = .015). The overall time to repeat BG was obtained 22 (IQR, 8-44) minutes after bolus treatment, but this interval increased with subsequent measurements. Refractory or recurrent hypoglycemia occurred in 30.3% of patients. Mean initial BG was lower in the subset of patients who developed Refractory hypoglycemia compared with those who did not (35.1 +- 9.8 vs 37.6 +- 10.2 mg/dL, P = .079), although not statistically significant. Patients with recurrent hy- poglycemia were also less likely to receive dextrose containing intravenous fluids compared with those without recurrent hypoglycemia (P = .028). Infection was the only associated characteristic with refractory or recurrent hypoglycemia (P = .021).

Conclusions: Overall, 12% of patients did not receive treatment for hypoglycemia in the ED with a BG <= 50 mg/dL. Time to treatment after identification was faster when identified by care testing vs serum sample result. Time to repeat BG in the ED was relatively quick, but did increase over time. About one-third of patients had refractory or recurrent hypoglycemia and infection was associated with this occurrence. Lack of dextrose containing intrave- nous fluid was associated with the incidence of recurrent hypoglycemia.

(C) 2016

Introduction

Over the last 20 years, the number of Americans living with diabetes has tripled, affecting an estimated 20.9 million people in 2011 alone [1]. Although most emergency department (ED) clinicians

? Source of support: None.

?? Prior presentations: American College of Clinical Pharmacy Virtual Poster Session, pre- sented 18th May 2016.

* Corresponding author at: Department of Pharmacy, University of Rochester Medical Center, 601 Elmwood Ave., Box 638, Rochester, NY 14642. Tel.: +1 585 275 6147.

E-mail address: [email protected] (N.M. Acquisto).

spend a great deal of time managing hyperglycemia, the potentially fatal consequences of hypoglycemia cannot be overlooked. The estimated cost for an ED visit and inpatient admission secondary to hypoglycemia is over $1300 and $17 000, respectively [2]. The annual number of ED visits for hypoglycemia has remained relatively unchanged between 2006 and 2009, suggesting that better management is warranted [1].

In cases of Severe hypoglycemia, dextrose and glucagon have an essential role. Dextrose is the drug of choice for management. It is available as D50W, 25 g/50 mL prepackaged syringes, making it easy to administer. Other formulations such as D10W and D5W may also be used for initial treatment; however, the volume to administer is greater and will take longer to infuse. In practice, a typical adult dose is 25 g [3]. In clinical practice, a 25-g dose of dextrose seems to raise serum blood glucose (BG) for approximately 30 to 60 minutes, but

http://dx.doi.org/10.1016/j.ajem.2016.10.009

0735-6757/(C) 2016

may depend on the degree of hypoglycemia and other patient specific factors. As a result, multiple doses or administration of dextrose containing intravenous fluids (IVFs) may be warranted. Weight-based dextrose doses are commonly used in pediatric patients (0.5-1 g/kg) [4]. When considering this weight-based dose, it may seem that a single 25-g dose of dextrose for all adult patients may not be enough to resolve hypoglycemia. Glucagon is the preferred agent when there is no intravascular (IV) access and is given as a 1 mg intramuscular [5]. However, the onset of action is delayed for approximately 15 minutes, making glucagon less than ideal.

Despite routine initial management, some patients are prone to developing additional hypoglycemic episodes, an event that can be overlooked in a busy ED setting. Prolonged hypoglycemia can increase neuronal cell death possibly leading to worsened outcomes [6]. Several retrospective studies have identified possible patient characteristics associated with severe hypoglycemic episodes, potentially placing these patients at risk for rebound hypoglycemia [7-10]. The elderly population appears most vulnerable to this phenomenon, perhaps due to worsening renal function leading to the inability to eliminate diabetic medications in addition to polypharmacy and Drug interactions [1,7,8,11]. Medications such as insulin and sulfonylureas are most frequently linked to hypoglycemia incidence [7-9]. The duration of diabetic disease has been identified as a major risk factor and is thought to play a role in insulin resistance and higher medication requirements [7,9,12,13]. Nonmodifiable patient characteristics such as sex, with women being most affected, and race, where African Americans and Hispanics appear to be more susceptible to hypoglycemia, have been identified as risk factors for hypoglycemia [8,10].

To our knowledge, there are no publications describing the current

management of hypoglycemia in the ED regarding overall treatment and subsequent BG measurements. Therefore, the purpose of this study is to evaluate the management of hypoglycemia in the ED, as well as to identify patient characteristics associated with refractory (need for additional treatment after initial management to achieve BG >= 80 mg/dL) or recurrent (resolved hypoglycemia followed by subse- quent hypoglycemia BG <= 50 mg/dL) hypoglycemia in the ED.

Methods

Study design

This was a retrospective evaluation of patients who presented to the ED with hypoglycemia. The institution’s Research Subjects Review Board approved the conduct of this study.

Setting and population

This study was conducted at an 850-bed academic medical center between January 1, 2011, and July 24, 2015. All adult patients with International Classification of Diseases, Ninth Revision Diagnosis codes for hypoglycemia or a documented BG <= 50 mg/dL during the first 6 hours of ED admission were included. This threshold was chosen to ensure that patients with significant hypoglycemia requiring treatment were included.

Patients were excluded if they developed hypoglycemia after initial presentation and assessment in the ED or if they were included in the study at a previoUS time point.

Study protocol

A review of the electronic medical record was conducted by 2 inves- tigators using a standardized data collection form. Patient information collected included age, sex, weight, race, arrival by prehospital services, comorbid conditions (including diabetes mellitus, chronic kidney disease or acute kidney injury, and coronary artery disease), outpatient medications that may be associated with hypoglycemia (antidiabetic

medications [sulfonylureas, meglitinides, thiazolinediones, ?-glucosidase, SGLT2 inhibitors, bile acid sequestrants, DDP-4 inhibitors, GLP agonists, amylin analogues, insulin], angiotensin-converting enzyme inhibitors, ?-blockers, quinolones, salicylates, disopyramide, and chloroquine), alcohol consumption (collected either as a quantifiable blood alcohol concentration or an International Classification of Diseases, Ninth Revision code reflective of alcohol use), and presence of concurrent infection [7,14-16]. Kidney function was assessed using the Cockcroft-Gault formula, using the first available serum creatinine or plasma creatinine for the specific patient encounter. Chronic kidney disease was defined as creatinine clearance less than 30 mL/min and acute kidney injury defined as a increase of at least 50% in baseline serum creatinine [17]. Active infection was identified by administration of systemic antibiotics during ED admission or recently treated (<=14 days) as an outpatient per medication history. Patients were not excluded if infection was ruled out later by the treatment team.

To assess the primary outcome, data from both the prehospital (if available) and ED setting were evaluated. Medications used to treat hypoglycemia (ie, glucagon or dextrose) including the dose, route, and time of administration were recorded. Similarly, all BG measurements including point-of-care testing or serum collection and time drawn and resulted were obtained. Additional information collected included maintenance therapy of dextrose-containing IVFs or food/juice.

Data analysis

Using a conservative estimate of the confidence interval, we chose to target 385 patients to provide us with a 95% Confidence level and confidence interval of +-5%. Descriptive statistics were used to present current practices in hypoglycemia treatment including dose, time to treatment, time to subsequent BG evaluations, and utilization of maintenance therapy. Identification of characteristics associated with refractory or recurrent hypoglycemia was evaluated using Wilcoxon rank sum analysis for continuous variables and ?² analysis for categorical variables.

Results

A total of 244 patients with hypoglycemia on admission to the ED (including those treated for hypoglycemia by prehospital emergency services) were evaluated. Mean age was 59 +- 18.7 years, 49% were women, mean weight was 85 +- 24.3 kg, and most patients were white (55.3%). Seventy-three percent of patients had a preexisting diagnosis of diabetes in addition to 44.3% having a history of chronic kid- ney disease. A total of 34.4% presented with an active infection and 76.6% were taking an outpatient medication associated with hypoglycemia. Complete demographic information is available in Table 1.

Prehospital care

One hundred twenty-four patients (51%) arrived via emergency medical services (EMS). In this prehospital environment, the average initial BG was 43 +- 31 mg/dL. Of the 95 patients (76%) with a BG <= 50 mg/dL, 88 (92.6%) were treated with bolus IV dextrose (n = 64), oral dextrose (n = 10), or IM glucagon (n = 20). Of these, 73 (83%) of patients treated received an additional BG measurement in the field to evaluate therapy effectiveness. All patients arriving by EMS did get a repeat BG in the ED. Time to this measurement after ED arrival was obtained in a median 25 (interquartile range [IQR], 10-40) minutes. prehospital management is summarized in Table 2.

Emergency department care

Patients who arrived via ambulatory means had an initial ED BG of 45 +- 25 mg/dL. Initial BG measurement after ED arrival was obtained significantly later in these patients (median, 43 [IQR, 17-95] minutes)

Table 1

Patient demographics

Characteristic
Age (y), mean +- SD	59 +- 18.7
Male, no. (%)	120 (48.9)
Weight (kg), mean +- SD	85 +- 24.3
Race

White, no. (%) 89 (55.3)

Black, no. (%) 67 (41.6)

Comorbidities

Diabetes mellitus, no. (%) 178 (72.9)

Chronic kidney disease, no. (%) 108 (44.3)

Infection, no. (%) 84 (34.4)

Coronary artery disease, no. (%) 74 (30.3)

Alcohol use disorder, no. (%) 34 (13.9)

Table 3

Overall ED treatment characteristics for patients with BG <= 50 mg/dL

Patients with BG <= 50 mg/dL in ED, no. (%)	174 (71.3)
Patients treated with IV bolus dextrose, no. (%)	108 (62.1)
Patients treated with dextrose-IVF only, no. (%)	3 (2)
Patients treated with food only, no. (%)	40 (23)
Patients treated with dextrose-IVF + food only, no. (%)	2 (1)
Patients with no documented treatment no. (%)	21 (12)

and fourth (n = 3) measurements, the time to repeat BG increases as the number of repeat measurements after treatment increases. The ob- served median time to obtaining a repeat BG measurement increased from a median of 20 (IQR, 6.5-39) minutes after the first dose of bolus

Receiving outpatient medication associated with hypoglycemia, no. (%)

187 (76.6)

dextrose to a median of 41 (IQR, 24-42.5) minutes after the fourth dose of bolus dextrose (Figure).

Insulin, no. (%) 147 (60.2)

?-Blocker, no. (%) 100 (41)

Salicylate, no. (%) 85 (34.8)

Angiotensin-converting enzyme inhibitor, no. (%) 67 (27.4)

Oral antidiabetic medication, no. (%) 47 (19.3)

Quinolone antibiotic, no. (%) 32 (13.1)

Transported via EMS, no. (%) 124 (51)

compared with those that arrived via EMS (P = .0018). Of the overall 244 patients who presented to the ED, 174 (71.3%) of patients had at least one BG measurement in the ED of 50 mg/dL or less; therefore, we were able to evaluate the ED management of hypoglycemia in this subset of patients. A total of 186 bolus doses of IV dextrose (D50W) were administered in 108 (62.1%) ED patients. Median dose was 25 g (range, 12.5-75 g), equating to a mean (SD) weight-based dose of 0.33 (0.12) g/kg. Two patients received glucagon 1 mg IM in the ED for treatment. There were 66 (38%) patients with identified hypoglycemia in the ED who did not receive bolus IV dextrose; however, 45 patients were given food (n = 40), dextrose containing IVF (n = 3), or both (n = 2). Ultimately, 21 (12%) of patients did not receive treatment for hypoglycemia identified in the ED. Of the entire patient population (arrival by EMS or ambulatory) that was treated with IV bolus dextrose (n = 172), most, 70%, were initiated on maintenance therapy once they arrived in the ED after initial bolus dextrose management with food and/or dextrose containing IVF. Complete hypoglycemia management in the ED information is in Table 3.

Treatment and reassessment timeframe analysis

There were 142 measured BG values <=50 mg/dL obtained in the ED in 108 patients in the ED who received treatment (IV bolus dextrose, dextrose-containing IVF, or food). Of those treated with IV bolus dex- trose, the median time from result to treatment was 12 (IQR, 6-27.8) minutes. In evaluating measurements obtained by POCT (n = 122) and serum only (n = 20), the median time to IV bolus dextrose was 11 (IQR, 6-23.5) minutes and 25 (IQR, 10.75-42.5) minutes from the time the value resulted to treatment, respectively (P = .015). An additional component of hypoglycemia management is reevaluation of BGs after treatment, with some patients having up to 4 different measurements in the ED. See Table 4 for a summary of treatment and

Refractory and recurrent hypoglycemia

Of 244 total patients, 74 (30.3%) experienced refractory and/or recurrent hypoglycemia. Refractory hypoglycemia occurred in 60 patients (24.5%). Mean initial BG (EMS or ED) was similar in refractory patients compared with those without this group (35.1 +- 9.8 vs 37.6 +- 10.2 mg/dL, P = .079). The initial dextrose dose for both groups was similar (median, 25 vs 25 g; mean weight-based dose, 0.31 +- 0.14 vs 0.33 +- 0.14 g/kg, respectively; P = .22). Recurrent hypoglycemia was observed in 29 patients (11.5%). Dextrose containing IVF was started after initial IV bolus dextrose in only 2 patients in the recurrent group compared with 31 of those without recurrent hypoglycemia (P = .028). A total of 13 patients in the recurrent group were given food; however, the timing of this in relation to IV bolus dextrose and the recurrent event is unknown. Active infection was the only patient characteristic associated with refractory or recurrent hypoglycemia (hazard ratio, 1.44; 95% confidence interval, 1.06-1.96).

Discussion

Hypoglycemia is a common presentation in the ED. Identification of these patients is often by clinical or objective assessment. Management may be difficult because there are no established treatment guidelines, multiple treatment methods may be needed (IV bolus dextrose, gluca- gon, dextrose containing IVF, food/juice), and frequent reassessments are required. We sought to evaluate the management of hypoglycemia in the ED and identify patient characteristics associated with refractory or recurrent hypoglycemia.

At our facility, we identified that half of the patients arriving via EMS waited at least 25 minutes to have an initial BG measured in the ED. It is possible that these patients may not have appeared as acutely symptomatic considering 92% of hypoglycemic patients were treated before arriving to the ED and we did not evaluate total transport time to determine if there were significant delays in the time from

Table 4

Assessment of hypoglycemia in the ED

Initial BG for all patients (mg/dL), mean +- SD 59.9 +- 46.5

reassessment timeframes. The overall median time to repeat BG was

22 (IQR, 8-44) minutes after IV bolus dextrose. When these are separated into first (n = 122), second (n = 42), third (n = 9),

ED arrival via EMS to 1st ED point-of-care BG (min), median (IQR)

ED arrival via ambulatory means to 1st ED point-of-care BG (min), median (IQR)

25 (10-40)?

43 (17-95)

BG <= 50 mg/dL to treatment (min), median (IQR) 12 (6-27.8)

Via POCT only (n = 122) (min), median (IQR) 11 (6-23.5)??

Table 2

Hypoglycemia management in the prehospital setting

Via serum measurements only (n = 20) (min), median (IQR)

25 (10.75-42.5)

Treatment to Repeat BG (min), median (IQR) 22 (8-44)

Initial prehospital BG (mg/dL), mean +- SD	43 +- 31
Patients with BG <= 50 mg/dL treated with dextrose/glucagon, no. (%)	88 (92.6)
Patients with repeat BG after therapy, no. (%)	73 (83)
ED arrival to 1st ED point-of-care BG (min), median (IQR)	25 (10-40)

* Patients arriving via prehospital care were assessed faster in the emergency department compared with ambulatory arrival, P = .0018.

?? Patients were treated faster after hypoglycemia was identified with a POCT measurement compared with identification after resultant values from serum samples, P = .015.

45

40

35

30

Median Time

to Repeat BG 25

After Therapy, 20

min (IQR)

15

10

5

0

Reassessment of IV Bolus Dextrose Therapy

1 (n = 122) 2 (n = 42) 3 (n = 9)

28 (19 – 35)

25 (10.5 – 54)

20 (6.5 – 39)

Number, IV Bolus Dextrose to Repeat BG

41 (24 – 42.5)

4 (n = 9)

Figure. Reassessment of IV bolus dextrose therapy.

prehospital treatment to ED BG. We also did not investigate if these pa- tients were seen faster by ED providers compared with ambulatory pa- tients. The time to initial ED BG for patients arriving via ambulatory means was obtained significantly longer than EMS patients, this was an expected finding because these patients may not have presented with obvious symptoms of hypoglycemia and this may have been found as an incidental finding.

Twelve percent of patients with hypoglycemia in the ED were not treated. Upon closer investigation, none of these 21 patients were brought to the hospital for a chief concern of hypoglycemia; only 2 had prehospital BG analysis collected, both of which were greater than 50 mg/dL in the prehospital setting. There was no dextrose or glucagon administered by EMS providers. Upon arrival to the ED, these 21 pa- tients waited 3-fold longer to receive their initial BG (median, 92.5 [IQR, 54.5-152.3] minutes) compared with the patients who were treat- ed (median, 31 [IQR, 13-69] minutes, P b .05). Nine of the 21 patients had a repeat BG analysis and only 6 showed their hypoglycemia to re- solve. Regardless of the chief concern, it appears that these patients did not receive the treatment they required.

To treat hypoglycemia, the condition must first be identified. We de- termined that treatment is significantly delayed when hypoglycemia is identified on a serum blood sample. At our institution, a BG <= 40 mg/dL is considered a critical value and called to the ED communication nurse. Because we used a hypoglycemia definition of BG <= 50 mg/dL, there are patients in the 41-50 mg/dL range that are not considered a critical value. This may attribute to the prolonged time from result to treatment in the patients with serum BG measurements compared with those with low POCT levels. Interestingly, in our study, 17 of the 21 untreated patients had a BG in this range, which suggests that a sizable portion of patients with significant hypoglycemia are not being treated.

In terms of attentiveness, it appears that the ED staff is expedient in treating hypoglycemia after its recognition as evidenced by the decreased time from BG to treatment between the first and fourth BGs. However, patients can only be treated if they are reassessed appropriately. The time from treatment to repeat BG increased as time progressed. This finding is likely because the need to collect multiple repeat BG measurements for a single patient may diminish in priority.

Refractory or recurrent hypoglycemia occurred in approximately one-third of all patients. On average, each patient received a single 25-g dose of IV bolus dextrose for the first treatment administered in the ED. This is a typical dose to begin Rescue therapy for hypoglycemia as it can be delivered in a single administration of 50 mL dextrose 50% in water at the bedside.

Although our study did not show a significantly lower initial ED BG, there exists a possibility that our small population may caused type II error and it may be worth revisiting an initial starting dose of only 25 g in adult patients with significant hypoglycemia. Potentially, consid- er 50 g IV bolus dextrose in patients with a reported BG <= 35 mg/dL. We also found that the incidence of recurrent hypoglycemia was associated with the lack of dextrose containing IVF after initial treatment. Consid- ering that both groups were given food, our study provides significant evidence illustrating the need to start continuous IV dextrose- containing fluids after bolus treatment. Our results emphasize the im- portance of not just treating with an initial bolus, but also accessing the need for maintenance therapy until the hypoglycemia event can be further investigated and prevented from reoccurring. Regarding pa- tient characteristics associated with refractory or recurrent hypoglyce- mia, we expected to find more statistically significant associations than active infection alone considering that more than three-quarters of our patients were currently taking medications known to cause hypo- glycemia. Our small sample size may have contributed to type II error. We also did not take into account certain considerations for each patient such as overall length of antidiabetic pharmacotherapy or total number of antidiabetic medications; we feel that future research should explore if such additional patient-specific therapy factors may predispose patients to refractory or recurrent hypoglycemia.

The present study contributes new data to the current body of evi- dence as there are few studies similar to ours. Ginde et al [10] performed a retrospective review of hypoglycemia patients presenting to 3 EDs. Associations with sulfonylurea and insulin use were identified as characteristics of those at risk for hypoglycemia. This differs from our secondary end point in that we sought to evaluate characteristics only associated with refractory or recurrent hypoglycemia. This study also evaluated discharge instructions, characteristics associated with hospitalization, and treatment in the ED; however, dose assessments, evaluation of repeat BG measurements, and assessment of refractory or recurrent hypoglycemia were not performed. Similarly, Rowe et al

[18] performed a multicenter medical record review of Diabetic patients presenting with hypoglycemia for 2 years in Canada evaluating discharge instructions, diagnostic testing, and treatment in the ED. Again, this study did not evaluate treatment doses, repeat BG measure- ments, or the incidence of refractory or recurrent hypoglycemia. We also included all hypoglycemia patients regardless of a diabetes diagnosis because initial hypoglycemia management should not differ in the ED.

The only other study, to our knowledge, that examined recurrent hypoglycemia followed up 233 diabetic patients presenting with severe hypoglycemia in the ED for 48 hours [7]. A total of 31.8% of patients

were categorized as having recurrent hypoglycemia, defined as the Serum glucose level lower than 70 mg/dL. Similar to our study, no single outpatient antidiabetic medication including insulin was associated with recurrence, infection and was, however, found to be a significant risk factor. An association between hypoglycemia and coronary artery disease or ?-blocker use was found in this study; however; we were not able to find an association, possibly due to type II error. Our study only examined the initial 6-hour treatment course in the ED and the definitions of recurrent hypoglycemia differ, which may explain why our rate of recurrent hypoglycemia was less, 11.5%. However, we did evaluate the incidence of not only recurrent, but also refractory hypo- glycemia and found a similar rate of 30.3%.

Limitations

Our study has several limitations to consider. First, because it was a retrospective medical record review at a single center, our study may not be generalizable to all other institutions. We also did not perform interrater reliability testing between data abstractors, but we did use a standardized data collection sheet and code book so that the data were abstracted in the same fashion. Unfortunately, we did not meet our predefined sample size, rendering some investigations underpow- ered. However, we also found several significant results. For patients who were given only food, it became apparent that treatment efficacy would not be possible to ascertain because the timing of food adminis- tration is not logged in consistent fashion between nurses. Finally, our classification of hypoglycemia is less than the generally recognized 70-mg/dL definition, but we believe this value to be appropriate consid- ering that most patients in this range require rescue therapy.

Conclusion

Overall, 12% of patients did not receive treatment for hypoglycemia in the ED with a BG <= 50 mg/dL. Time to treatment after identification was faster when identified by POCT vs serum sample result. Time to re- peat BG in the ED was relatively quick, but did increase over time. About one-third of patients had refractory or recurrent hypoglycemia and in- fection was associated with this occurrence. Lack of dextrose containing IVF was associated with the incidence of recurrent hypoglycemia.

References

1. CDC.gov [Internet]. Centers for Disease Control and Prevention. Emergency department visits [updated 2014 October 15, cited 2016 July 6]. Available from: http://www.cdc.gov/diabetes/statistics/hypoglycemia/fig1.htm/.
2. Quilliam BJ, Simeone JC, Ozbay AB, Kogut SJ. The incidence and costs of hypoglyce- mia in type 2 diabetes. Am J Manag Care 2011;17(10):673-80.
3. Lexi-comp, Inc [Internet]. Wolters Kluwer Clinical Drug Information, Inc [cited 2016 June 10]. Available from: http://online.lexi.com/action/home.
4. Kleinman ME, Chameides L, Schexnayder SM, et al. Part 14: pediatric advanced life support: 2010 American Heart Association guidelines for cardiopulmonary resuscitation and emergency cardiovascular care. Circulation 2010;122(Suppl. 3): S876-908.
5. McNaughton CD, Self WH, Slovis C. Diabetes in the emergency department: acute care of diabetes patients. Clin Diabetes 2011;29(2):51-9.
6. Cryer PE. Hypoglycemia, functional brain failure, and brain death. J Clin Invest 2007; 117(4):868-70.
7. Lin YY, Hsu CW, Sheu WH, Chu SJ, Wu CP, Tsai SH. Risk factors for recurrent hypoglycemia in hospitalized diabetic patients admitted for severe hypoglycemia. Yonsei Med J 2010;51(3):367-74.
8. Shorr RI, Ray WA, Daugherty JR, Griffin MR. Incidence and risk factors for serious hypoglycemia in older persons using insulin or sulfonylureas. Arch Intern Med 1997;157:1681-6.
9. Leese GP, Wang J, Broomhall J, Kelly P, Marsden A, Morrison W, et al. Frequency of severe hypoglycemia requiring emergency treatment in type 1 and type 2 diabetes. Diabetes Care 2003;26(4):1176-80.
10. Ginde AA, Espinola JA, Cmargo CA. Trends and disparities in U.S. emergency department visits for hypoglycemia, 1993-2005. Diabetes Care 2008; 31(3):511-3.
11. Morales J, Schneider D. Hypoglycemia. Am J Med 2014;127:S17-24.
12. Shafiee G, Mohajeri-Tehrani M, Pajouhi M, Larijani B. The importance of hypoglyce- mia in diabetic patients. J Diabetes Metab Disord 2012;11(1):17.
13. Murad MH, Coto-Yglesias F, Wang AT, Sheidaee N, Mullan RJ, Elamin MB, et al. Clinical review: drug-induced hypoglycemia: a systematic review. J Clin Endocrinol Metab 2009;94(3):741-5.
14. Vue MH, Setter SM. Drug-induced glucose alterations part 1: drug-induced hypoglycemia. Diabetes Spectr 2011;24(3):171-7.
15. Diabetes.org [Internet]. Alexandria: American Diabetes Association. What are my options? [updated 15 march 3; cited 2016 July 6]. Available from: http:// www.diabetes.org/living-with-diabetes/treatment-and-care/medication/oral-medi- cations/what-are-my-options.html?referrer=https://www.google.com/.
16. Diabetes.org [Internet]. Alexandria: American Diabetes Association. Insulin and other injectable medications [updated 2015 June 29, cited 2016 July 6]. Available from: http://www.diabetes.org/living-with-diabetes/treatment-and-care/medication/ insulin/.
17. Palevsky PM, Liu KD, Brophy PD, Chawla LS, Parikh CR, Thakar CV, et al. KDOQI US commentary on the 2012 KDIGO clinical practice guideline for acute kidney injury. Am J Kidney Dis 2013;61(5):649-72.
18. Rowe BH, Singh M, Villa-Roel C, Leiter LA, Hramiak I, Edmonds ML. Acute management and outcomes of patients with diabetes mellitus presenting to Canadian emergency departments with hypoglycemia. Can J Diabetes 2001;39:55-64.