Q 71What is the most appropriate medical intervention for adults with type 1 diabetes with severe hypoglycaemia?

Author / Title / Reference / YrPatrick AW, Collier A, Hepburn DA, Steedman DJ, Clarke BF, Robertson C 1990 Comparison of intramuscular glucagon and intravenous dextrose in the treatment of hypoglycaemic coma in accident and emergency department. Archives of Emergency Medicine 7:73–77
N=29 insulin treated diabetic people presenting consecutively to the Accident and Emergency department Scotland
Research DesignRandomised Controlled Trial
AimTo compare intramuscular glucagon and intravenous dextrose in the treatment of hypoglycaemic coma in an accident and emergency department
PopulationPeople had hypoglycaemia diagnosed on a capillary blood specimen with a Reflolux II Meter
InterventionN=15 Intramuscular glucagon (1 mg i.m.)
ComparisonN=14 Intravenous dextrose (50 ml 50% i.v.)
OutcomeRestoration of normal conscious level and average duration of hypoglycaemic coma
CharacteristicsMean age: 47/48 years in glucagon/dextrose treated group respectively
ResultsInitial plasma glucose, glycated haemoglobin and estimated duration of coma were not significantly different prior to treatment.
Average duration of coma: 120 min (range 60–240 min) in glucagon treated people, and 120 min (range 20–480 min) in dextrose treated people
Significantly slower recovery to normal conscious levels was seen in the glucagon-treated group compared to the dextrose treated group: 9 min (range 5–30 min) vs. 3 min (range 2–15 min) respectively; p less than 0.01
Two glucagons treated people required administration of additional i.v. dextrose after failure to show signs of clinical recovery within 15 min of treatment.
No correlation was seen between time taken to recovery of consciousness and either initial plasma glucose concentration or duration of hypoglycaemia
All but one patient in each group reported either partial or total loss of awareness of the onset of hypoglycaemia.
Hierarchy of Evidence Grading1b
CommentsPeople randomly allocated to treatment
Drugs administered to right thigh
Venous blood withdrawn prior to treatment. Measurements taken for plasma glucose, alcohol and glycated haemoglobin.
Additional samples taken at 5, 10, 15 and 30 min after initiation of treatment for plasma glucose estimation
Time taken to recovery of normal conscious level noted, additional 12.5 g of dextrose i.v. administered in absence of recovery after 15 mins
Duration of hypoglycaemia prior to treatment established by retrospective questioning of relatives or family.
Precipitating causes and details of insulin and other drug therapy, as well as possible precipitating causes or preceding loss of awareness of the onset of hypoglycaemia were also obtained from families and friends
No participants were receiving treatment with beta blockers or other drugs which might alter hypoglycaemic awareness of affect the counter regulatory response.
Precipitating factors of hypoglycaemia unknown, three people had detectable levels of alcohol in their plasma, maximum level recorded = 3 mmol/L
Regular treatment of participants: twice daily mixture of soluble, and isophane or lente insulins (n=19), or once daily lente or protamine zinc insulins, either with or without additional soluble insulin before breakfast (n=10)
Experimental protocol not described in detail, leading to possible bias, no details of blinding outlined
No clinical definition of outcome: ‘normal conscious levels’ not defined
No details of statistical methods
Duration of coma, precipitating factors, and other therapy collected retrospectively, possible recollection bias. Only rough estimation of coma duration was possible in several cases.
Reference / Citation172
Author / Title / Reference / YrCollier A, Steedman DJ, Patrick AW, Nimmo GR, Matthews SM, Macintyre CA, Little K, Clarke B 1987 Comparison of intravenous glucagon and dextrose in treatment of severe hypoglycaemia in an accident and emergency department. Diabetes Care 712–715
N=49 consecutive insulin treated diabetic subjects with hypoglycaemic coma referred to the accident and emergency department Scotland
Research DesignRandomised Controlled Trial
AimTo compare intravenous glucagon and dextrose in treatment of severe hypoglycaemia in an accident and emergency department
PopulationPeople referred to the accident emergency department for hypoglycaemic coma
Interventionn=25 i.v. glucagon (1 mg)
Comparisonn=24 50 ml 50% dextrose (25g)
OutcomeTime taken to return to a normal level of consciousness, adverse effects during treatment.
CharacteristicsMean age: glucagon group = 39.4±17.1; dextrose group = 40.2±14.2 years
Duration of diabetes: glucagon group = 14.0 (4–47.2); dextrose group = 13.0 (2–32.0)
ResultsPeople in the two groups were comparable in terms of initial blood glucose, prevailing glycaemic control, age, duration of diabetes and duration of hypoglycaemia (1.3 (range 0.3–4.0) vs. 1.5 (0.3–9.0) h)
A significant difference was seen in the subsequent glycaemic profiles between the two treatment groups (no raw data provided)
Time to regain normal consciousness
The glucagon treated group was slower to achieve normal conscious level compared to the dextrose treated people. Median time: 6.5 min (range 2–16 min) vs. 4 min (range 1–15 min), respectively (p less than 0.001)
All people returned to a normal level of consciousness within 30 min of admission.
Plasma glucose levels at admission of duration or hypoglycaemia did not correlate with the time to recovery of normal consciousness in either groups of people.
Side effects
Eight people in the glucagon group vomited compared with 9 in the dextrose group
Thirteen people in the glucagon group reported headaches on return to normal consciousness compared with 12 in the dextrose group.
Other factors
Two people in both groups required additional administration of (12.5 g i.v.) dextrose
Precipitating factors for hypoglycaemia were difficult to elicit: no obvious cause was found in 30% of people, 12 people reported missing a meal, 10 were considered to have an irresponsible attitude toward diabetes control, three people had blood alcohol levels greater than 100 mg/dL and 1 patient became hypoglycaemic due to marked fluctuations in insulin requirements due to pregnancy.
One patient received s.c. glucagon administration prior to admission (plasma glucose on arrival was 56 mg/dL compared to mean 18 mg/dL across the groups and consciousness level was a borderline grade 3)
Hierarchy of Evidence Grading1b
CommentsConsciousness assessed and graded as: 0: normal orientation in time and place; 1: drowsy; 2: maximal response to minimal stimuli; 3: minimal response to maximal stimuli; and 4: unresponsive to painful stimuli
People randomly allocated to emergency treatment
Blood was taken for plasma glucose estimation at 5, 10, 15 and 30 min
Return to consciousness was evaluated by the same observer in all participants
Additional i.v. dextrose (12.5 g) was administered to people failing to achieve grade 3 consciousness at 15 and 30 min
Following recovery from hypoglycaemia, all people were questioned about headache symptoms and it was noted whether they had vomited after administration of treatment.
Details of age, duration of diabetes, insulin dosage, most likely precipitating cause of hypoglycaemia and estimation of duration of the hypoglycaemic episode were established through questionnaires completed by participants or family and friends
Patients reported the usual cause of hypoglycaemia, and any steps taken to try to abort the episode
One patient with Addison’ s disease was excluded from the statistical analysis
No power analysis given at outset, no details given of why the participants were selected to take part in this study
No details are provided of any blinding of this study
Authors state that significant differences were seem in glycaemic profiles between the two groups following treatment, although no raw data are provided for this.
Length of hypoglycaemic episode and precipitating factors are retrospectively collected from participants and present friends and relatives following return to consciousness, leading to possible recall bias.
Exclusion criteria prior to inclusion in the study are not explicit. One patient who entered the study was pregnant one had received prior s.c. glucagon administration leading to exceptionally high plasma glucose levels at admission compared to the mean. No details are given of the treatment regimen received by these people. Due to the small sample size the results of these people may affect the outcome of the study.
One patient with Addison’ s disease was excluded from the statistical analysis of this study
Reference / Citation173

From: Appendix D, Evidence tables

Cover of Type 1 Diabetes in Adults
Type 1 Diabetes in Adults: National Clinical Guideline for Diagnosis and Management in Primary and Secondary Care.
NICE Clinical Guidelines, No. 15.1.
National Collaborating Centre for Chronic Conditions (UK).
Copyright © 2004, Royal College of Physicians of London.

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.