Abstract
Introduction: Prehospital first responders historically have used an IV bolus of 50 mL of 50% dextrose solution (D50) for the treatment of hypoglycemia in the field. A local Emergency Medical Services (EMS) system recently approved a hypoglycemia treatment protocol of IV 10% dextrose solution (D10) due to occasional shortages and higher cost of D50. We use the experience of this EMS system to report the feasibility, safety, and efficacy of this approach. Methods: Over the course of 104 weeks, paramedics treated 1,323 hypoglycemic patients with D10 and recorded patient demographics and clinical outcomes. Of these, 1,157 (87.5%) patients were treated with 100 mL of D10 initially upon EMS arrival, and full data on response to treatment was available on 871 (75%) of these 1,157. We captured the 871 patients’ capillary glucose response to initial infusion of 100 mL of D10 and fit a linear regression line between elapsed time and difference between initial and repeat glucose values. We also explored the need for repeat glucose infusions as well as feasibility, and safety. Results: The study cohort included 469 men and 402 women with a median age of 66. The median initial field blood glucose was 37 mg/dL, while the subsequent blood glucose had a median of 91 mg/dL. The median time to second glucose testing was eight minutes after beginning the 100mL D10 infusion. Of 871 patients, 200 (23.0%) required an additional dose of IV D10 solution due to persistent or recurrent hypoglycemia and seven (0.8%) patients required a third dose. There were no reported deaths or other adverse events related to D10 administration for hypoglycemia. Linear regression analysis of elapsed time and difference between initial and repeat glucose values showed near-zero correlation. Conclusions: The results of one local EMS system over a 104-week period demonstrate the feasibility, safety, and efficacy of using 100 mL of D10 as an alternative to D50. D50 may also have theoretical risks including extravasation injury, direct toxic effects of hypertonic dextrose, and potential neurotoxic effects of hyperglycemia. Additionally, our data suggest that there may be little or no short-term decrease in blood glucose results after D10 administration.
Key words: :
Introduction
Medical personnel, including prehospital first responders, have historically treated patients with hypoglycemia with 50% dextrose solution (D50). D50 is available in prefilled syringes containing 25 grams of glucose in 50 mL of water. While protocols vary, it is often given as a bolus after initial patient assessment, glucometer confirmation of hypoglycemia, confirmation of a patient's inability to take oral glucose, and obtaining intravenous access.
There are several reasons that treating hypoglycemia with a bolus of D50 may not be optimal. D50 is more viscous than other intravenous fluids and may require two hands to administer. This prevents paramedics from doing other tasks during its administration such as electrocardiograms or neurological assessments. D50 has the potential risk of extravasation injury, potential supratherapeutic dosing of dextrose resulting in hyperglycemia, and direct toxic effects of hypertonic dextrose itself.Citation1–3 Although no studies have linked the use of D50 in the prehospital setting to adverse outcomes, using an alternative concentration in a greater total volume of solution might help to lessen some of the potential risks of D50.
Beginning January 1, 2013 a local EMS system began using 100 mL of intravenous 10% dextrose solution (D10) as its intravenous agent to treat hypoglycemia. It collected clinical and demographic data as part of ongoing quality control measures examining this protocol change. Prior to the protocol change, D50 had been the standard intravenous dextrose agent in this system. The change was initiated due to frequent shortages of D50 and the lower cost of D10, as well as the potential theoretical benefit of not using D50. The purpose of this study was to analyze the quality improvement data collected on hypoglycemic patients and received intravenous D10 as their initial treatment during the first 2 years after this protocol change and to consider these results in the context of the medical literature.
Methods
This was a prospective observational cohort study based on quality improvement data. The EMS system protocol modification began January 1, 2013, and data collection continued through December 31, 2014, offering 24 months of observation. Prehospital personnel captured routine clinical data during patient encounters. We analyzed patients who (i) were older than age 18, (ii) displayed symptoms of hypoglycemia, (iii) whose mental status did not allow administration of oral dextrose, (iv) whose blood sugar was below 70 mg/dL, and (v) in whom intravenous access was unavailable.
Data were obtained through the electronic patient care report (PCR). Emergency department courses and hospital outcomes were not studied. Negative outcomes were defined as deterioration of the clinical status while in the care of the prehospital personnel or if there were complaints reported to the local EMS agency by receiving hospitals.
The chief outcomes were the blood glucose responses to initial D10 infusion of 100 mL and to repeat infusions if necessary. The point-of-care blood glucose machines used by prehospital personnel cannot assess glucose values lower than 20–25 mg/dL and instead indicate “Low.” For this analysis, a value of 20 mg/dL was used when comparing glucose values for values of “Low.” Upon visual inspection of the data distribution, we fit a linear regression line to the correlation of elapsed time and change of glucose value on EMS arrival and upon subsequent recheck after the initial bolus of 100 mL of D10. Prehospital personnel were also to note any adverse outcomes or events relating to dextrose administration.
Data analysis was performed using Microsoft Excel (Version 12.3.6, Microsoft Corporation, Redmond, WA, USA). The Institutional Review Board of the Alameda Health System - Highland Hospital approved the study as exempt from review as a report of quality improvement initiatives.
Results
Of 1,323 patients who were administered D10 by EMS personnel in the 104 weeks of data collection, 1,157 met inclusion criteria. The remaining 166 patients were excluded by the following criteria: oral dextrose initially by paramedics, received intramuscular glucagon, or delayed treatment (Figure ). Of the 1,157 patients who met criteria, 180 patients received out-of-protocol IV dextrose, for example in aliquots greater or less than 100 mL. Another 106 patients had incomplete data on response to D10, most commonly because of hospital arrival before repeat glucose value could be obtained. The remaining 871 patients received initial treatment with 100 mL of IV D10 and had repeat glucose measurements recorded. Age and sex data were available for all patients. The study cohort of 871 patients included 469 men and 402 women (Table ). The median initial field blood glucose was 37 mg/dL (IQR 29–46), with subsequent blood glucose median of 91 mg/dL (IQR 68–121). Elapsed time after D10 administration before recheck was not uniform with a median time to recheck of 8 minutes (IQR 5–12). Of 871 patients, 200 (23%) received an additional dose of intravenous D10 solution in the field due to persistent or recurrent hypoglycemia and seven patients required a third dose.
Figure 1. Patient selection.
Table 1. Patient demographics and results in study cohort
Of the 1,323 patients administered D10 during the study period, the 452 patients excluded from the study cohort for the aforementioned reasons were similar demographically to the study cohort. The median initial blood glucose was the same at 37 mg/dL and the median age was also 66. There were slightly more women at 229 (51%) in the excluded group compared to the cohort.
Of the 871 patients analyzed, median change in glucose after D10 administration was 55 mg/dL (IQR 32-80) (Figure ). Seven patients had a drop in blood glucose after D10 administration, all of 10 mg/dL or less except for one patient with a drop of 19 mg/dL who had an insulin pump infusing that was not removed by EMS personnel during D10 infusion. The remaining 864 patients had an elevation of their blood glucose on recheck (maximum rise to 325 mg/dL). Linear regression analysis of elapsed time and difference between initial and repeat glucose values showed a slope of 0.38 (95% CI –0.10, 0.87), with an R-squared value of 0.00281. (Figure .) The R-squared value of the regression was 0.00281. The analysis of variable distribution is shown in Figure .
Figure 2. Change in blood glucose after 100 mL of D10. Each point on this graph represents an individual patient (n = 871). The y-axis of this graph represents the absolute change in blood glucose in mg/dL between the initial value obtained by prehospital personnel and the subsequent value obtained after initial D10 infusion. A positive value indicates that the patient's glucose increased and a negative value indicates that the patient's glucose decreased. The x-axis represents time in minutes that transpired between the initial value obtained by EMS personnel and the subsequent value after D10 infusion for the patient. Each point represents an absolute change in blood glucose during a given time. The solid line in the figure represents a line of best fit that was applied to these data points using linear regression. A slope of 0.38 (95% CI –0.10, 0.87) suggests that there may be little or no short-term decay in blood glucose values after D10 administration. The dashed lines show the 95% confidence intervals of the regression line.
There were no reported adverse events related to dextrose infusion. Six patients who received intravenous D10 were pronounced dead in the field during the period of study. On investigator review, all patients had altered level of arousal or were in cardiac arrest prior to arrival of EMS personnel and their deaths were deemed to be unrelated to dextrose administration.
Discussion
Our results suggest that utilizing D10 as the primary intravenous treatment for hypoglycemia is efficacious, with all but three patients demonstrating an increase in their blood glucose after D10 administration. Although time to return to normal mental status was not measured, previous trial data shows correlation between increasing serum glucose and return to normal mental status.Citation4–6 Of the study patients, only 200 out of 871 (23%) received a second bolus of D10 in the field; however, some patients who only received one dose of D10 may have received additional dextrose upon hospital arrival and comparison to D50 is unknown. No concerns or complaints were noted by prehospital personnel in terms of comparing D10 use to D50. Some providers noted D10 was easier to infuse likely due to its lower viscosity. The data suggest that D10 administration is safe, with no adverse events related to administration noted. Since we do not have data for a comparable population of patients who received D50 for hypoglycemia, we cannot conclude, however, that D10 is as safe as or even better than D50.
The linear regression fit line of repeat blood glucose levels against elapsed time for first administration of D10 in the patient cohort suggests minimal or no short-term decay in blood glucose values after D10 administration; however, goodness of fit was poor as demonstrated by the low R-squared value. Visual inspection suggested a normal distribution of the residuals of the observed change in blood glucose values. Prior analysis of D50 infusion in volunteers shows that D50 increases serum glucose over the first 5 minutes (with a mean of 244 mg/dL); however, by 30 minutes, glucose levels approach pre-treatment levels consistently.Citation7 All of the glucose recheck times in this study were before 30 minutes, allowing that the lack of short-term decay of glucose levels might not have continued if longer-term data was available. Additionally, D10 may elicit a less profound insulin response than D50 and may paradoxically have a slower decay. Reported data are unable to test this hypothesis.
Despite the traditional use of D50, there is a minimal amount of data to support it as the standard of care. There is only one trial comparing D50 to D10 for the treatment of hypoglycemia.Citation4 Paramedics from a large United Kingdom ambulance service treated 51 adult hypoglycemic patients with altered mental status. Patients were randomized to receive 5 gram (50 mL) intravenous aliquots of D10 or 5 gram (10 mL) intravenous aliquots of D50 to a maximum dose of 25 grams. The study used a main outcome of time to regain normal consciousness and showed no statistical difference between the protocols. The authors concluded that their D10 protocol was safe and effective. The patients receiving D10 had lower post-treatment dextrose levels and had received less total dextrose. Our group performed a pilot study of 18 weeks and 164 patients, showing good feasibility, safety, and a minimal short term blood glucose decay.Citation8
From a physiologic standpoint, the maximum rate of glucose oxidation is about 4 to 7 mg/kg/min in humans.Citation9,10 Higher glucose loads will lead to an increase in lipogenesis and in energy required to excrete the excess CO2.Citation11 For a 80-kg person, the maximum oxidation rate is approximately 0.56 gm per min. If D10, instead of D50, is administered for the treatment of hypoglycemia, this concentration may already overwhelm the ability of the body to uptake glucose.
Glucose control in critically ill patients may be difficult. Few studies on glucose control exist in the prehospital setting, evidence on the ideal glucose levels in the inpatient setting remains controversial.Citation3 Hypoglycemia has established neurologic adverse effects in the critically illCitation12; however, both in human and animal studies, hyperglycemia has been shown to negatively affect neurologic outcomes as well.Citation1,3,13 In fact, even the timing of glucose assessment may have negative effects on outcomes.Citation14 Extrapolating inpatient data to the prehospital setting is problematic for multiple reasons. There are relatively few therapeutic interventions and fewer monitoring resources and available in the prehospital setting, the generally short duration of time that patients are under the care of prehospital providers, and the relatively low acuity of patients encountered compared to the ICU.Citation15
Figure 3. Residual distribution.
The utility of D50 for the treatment of hypoglycemia has shown safety, efficacy, and feasibility in the limited trial data available.Citation4,15,16,14 Several authors, however, have suggested concerns regarding the potential harm caused by D50 even though there is a lack of direct evidence.Citation2,16,17 In theory, smaller amounts of glucose utilization might assuage concerns regarding hyperglycemia risks and a more dilute solution might lower the risks of extravasation injury. The only study comparing D50 to D10 demonstrated no difference in time to return to GCS of 15 (the primary outcome). However, there were no safety differences identified.Citation4
Limitations
The limitations in this case series are similar to the limitations of any uncontrolled observational cohort study. Data was not complete. Of the initial 1171 patients, 180 (15.5%) received out of protocol D10 (higher or lower aliquots) and were excluded from the analysis. In addition, of the 977 patients treated per the described D10 protocol, a further 106 patients (9.2%) had incomplete data and were excluded from the analysis. Economic analysis is not reported and follow-up for patients after their treatment in the prehospital setting was not available to the authors. The regression analysis does not adjust for unmeasured covariates. While there were no reported adverse effects related to D10 administration and all patients eventually became euglycemic, it is not clear if there were significant delays to euglycemia in the reported group compared to D50 as a matched group was not studied. Finally, for those patients who did require a second or even rarely, a third dose of D10, it was not possible to determine if there were long term poor outcomes because of the EMS protocol change. This study represents the results of one EMS system during a protocol change and subsequent QI follow-up but may not be generalizable to other systems.
Conclusions
The experience of one local EMS system over a 2-year period demonstrates that D10 is a safe, effective, and feasible alternative to D50 in the acute prehospital management of hypoglycemia.
References
- Baker L, Rattan J, Bruno A. Management of hyperglycemia in acute ischemic stroke. Curr Treat Options Neurol. 2011;13(6):616–28.
- Browning RG, Olson DW, Stueven HA, Mateer JR. 50% dextrose: antidote or toxin? Ann Emerg Med. 1990;19(6):683–7.
- Fahy BG, Sheehy AM, Coursin DB. Glucose control in the intensive care unit. Crit Care Med. 2009;37(5):1769–76.
- Moore C, Wollard M. Dextrose 10% or 50% in the treatment of hypoglycaemia out of hospital? A randomized controlled trial. Emerg Med J. 2005;22(7):512–5.
- Collier A, Steedman DJ, Patrick AW, et al. Comparison of intravenous glucagon and dextrose in treatment of severe hypoglycemia in an accident and emergency department. Diabetes Care. 1987;10(6):712–5.
- Patrick AW, Collier A, Hepburn DA, Steedman DJ, Clarke BF, Robertson C. Comparison of intramuscular glucagon and intravenous dextrose in the treatment of hypoglycaemic coma in an accident and emergency department. Arch Emerg Med. 1990;7(2):73–7.
- Balantine JR, Gaeta TJ, Kessler D, Bagiella E, Lee T. Effect of 50 milliliters of 50% dextrose in water administration on blood sugar of euglycemic volunteers. Acad Emerg Med. 1998;5(7):691–4.
- Kiefer M, Hern H, Alter H., Barger J. Dextrose 10% in the treatment of out-of-hospital hypoglycemia. Prehosp Disaster Med. 2014;Apr;29(2):190–4.
- Berdanier CD. Handbook of Nutrition and Food,2013. CRC Press, 2002, p 978.
- Kim H, Son E, Kim J, Choi K, Kim C, Shin W, Suh AO. Association of hyperglycemia and markers of hepatic dysfunction with dextrose infusion rates in Korean patients receiving total parenteral nutrition. Am J Health-Syst Pharm. 2003;60:1760–6.
- Burke JF, Wolfe RR, Mullany CJ, Mathews DE, Bier DM. Glucose requirements following burn injury. Ann Surg. 1979;190(3)274–83.
- Duning T, Opiyo N, English M. Hypoglycemia aggravates critical illness-induced neurocognitive dysfunction. Diabetes Care. 2010;33(3):639–44.
- Nakakimura K, Fleischer JE, Drummond JC, et al. Glucose administration before cardiac arrest worsening neurologic outcome in cats. Anesthesiology. 1990;72(6):1005–11.
- Beskind DL, Rhodes SM, Stolz U, et al. When should you test for and treat hypoglycemia in prehospital seizure patients? Prehosp Emerg Care. 2014 Jul-Sep;18(3):433–41.
- Cain E, Ackroyd-Stolarz S, Alexiadis P, Murray D. Prehospital hypoglycemia: the safety of not transporting treated patients. Prehosp Emerg Care. 2003;7(4):458–65.
- Hoffman JR, Schriger DL, Votey SR, Luo JS. The empiric use of hypertonic dextrose in patients with altered mental status: a reappraisal. Annals Emerg Med. 1992;21(1):20–4.
- Hoffman RS, Goldfrank LR. The poisoned patient with altered consciousness: controversies in the use of a ‘coma cocktail’. JAMA. 1995;274(7):562–9.