ARTICLE IN PRESS

Can J Diabetes xxx (2017) 1–0

Contents lists available at ScienceDirect

Canadian Journal of Diabetes

journal homepage:
w w w. c a n a d i a n j o u r n a l o f d i a b e t e s . c o m

Original Research

Emergency Department Management of Diabetic Ketoacidosis

and Hyperosmolar Hyperglycemic State: National Survey of
Attitudes and Practice
Alexandra L. Hamelin HBSc a,*, Justin W. Yan MD, MSc b,c, Ian G. Stiell MD, MSc a,d
a The University of Ottawa, Ottawa, Ontario, Canada
b
The Division of Emergency Medicine, Department of Medicine, London Health Sciences Centre, London, Ontario, Canada
c Schulich School of Medicine and Dentistry, Western University, London, Ontario, Canada
d
The Ottawa Hospital Research Institute, Ottawa, Ontario, Canada

a r t i c l e i n f o a b s t r a c t

Article history: Objectives: In 2013, the Association, now Diabetes Canada, published national clinical practice guidelines
Received 15 March 2017
for the effective management of diabetic ketoacidosis and hyperosmolar hyperglycemic states in adults. We
Received in revised form
sought to determine emergency physician compliance rates and attitudes toward these guidelines and to
11 May 2017
Accepted 15 May 2017 identify potential barriers to their use in Canadian emergency departments.
Methods: An online survey consisting of questions related to the awareness and use of the Canadian Dia-
betes Association 2013 Clinical Practice Guidelines for the Prevention and Management of Diabetes in Canada
Keywords:
diabetic ketoacidosis was distributed to 500 randomly selected members of the Canadian Association of Emergency Physi-
hyperglycemia cians. Also included in the survey were 3 clinical vignettes to assess adherence rates to the guidelines.
emergency physicians Results: The survey response rate was 62.2% (311 of 500). The majority of physicians reported the guide-
hyperosmolar hyperglycemic state lines to be useful (83.6%); 54.6% of respondents were familiar with the guidelines, and 54.7% claimed to
guidelines use them in clinical practice. The most frequently reported barrier to guideline implementation was a lack
of education (56.0%). The clinical vignettes demonstrated respondent variability in fluid administration and
sodium bicarbonate administration, as well as some variability in insulin and potassium administration.
Conclusions: Although Canadian emergency physicians were generally supportive of the guidelines, many
were unaware that these guidelines existed, and barriers to their implementation were reported. These
results suggest the need to improve knowledge translation strategies across Canadian emergency depart-
ments to standardize management of diabetic ketoacidosis and hyperosmolar hyperglycemic states and
support the highest quality of patient care, as well as to ensure that future guidelines include manage-
ment strategies applicable to the emergency department setting.
© 2017 Canadian Diabetes Association.

r é s u m é
Mots clés :
Objectifs : En 2013, l’Association canadienne du diabète dorénavant appelée Diabète Canada, publiait les
acidocétose diabétique
hyperglycémie
lignes directrices nationales de pratique clinique pour une prise en charge efficace de l’acidocétose diabétique
médecins d’urgence et du syndrome d’hyperglycémie hyperosmolaire chez les adultes. Nous avons cherché à déterminer les
syndrome d’hyperglycémie hyperosmolaire taux de conformité et les attitudes des médecins d’urgence à l’égard des lignes directrices, et à identifier
lignes directrices les obstacles potentiels à leur utilisation dans les services des urgences du Canada.
Méthodes : Cinq cents membres de l’Association canadienne des médecins d’urgence choisis de manière
aléatoire ont reçu un sondage en ligne comportant des questions au sujet de leur connaissance et de leur
utilisation des lignes directrices Lignes directrices de pratique clinique 2013 de l’Association canadienne du
diabète pour la prévention et la prise en charge du diabète au Canada. Trois vignettes cliniques ont également
été incluses au sondage pour évaluer le respect des lignes directrices.

* Address for correspondence: Alexandra L. Hamelin, HBSc, Clinical Epidemiology Unit, Room F662, The Ottawa Hospital, 1053 Carling Avenue, Ottawa, Ontario K1Y 4E9,
Canada.
E-mail address: ahame086@uottawa.ca

1499-2671 © 2017 Canadian Diabetes Association.

The Canadian Diabetes Association is the registered owner of the name Diabetes Canada.
http://dx.doi.org/10.1016/j.jcjd.2017.05.005
 ARTICLE IN PRESS
2 A.L. Hamelin et al. / Can J Diabetes xxx (2017) 1–0

Résultats : Le taux de réponse au sondage était de 62,2 % (311 sur 500). La majorité des médecins rapportaient
l’utilité des lignes directrices (83,6 %) ; 54,6 % des répondants connaissaient les lignes directrices, et 54,7 %
prétendaient les utiliser dans la pratique clinique. Le manque de formation constituait l’obstacle à la mise
en œuvre des lignes directrices le plus fréquemment rapporté (56,0 %). Les vignettes cliniques démontraient
la variabilité entre les répondants dans l’administration des fluides et l’administration du bicarbonate
de sodium, ainsi qu’une certaine variabilité dans l’administration de l’insuline et du potassium.
Conclusions : Bien que les médecins d’urgence du Canada se soient généralement montrés favorables aux
lignes directrices, beaucoup ignoraient leur existence et rapportaient les obstacles à leur mise en œuvre.
Ces résultats montrent qu’il est nécessaire d’évaluer les stratégies d’application des connaissances dans
tous les services d’urgence du Canada pour uniformiser la prise en charge de l’acidocétose diabétique et
du syndrome d’hyperglycémie hyperosmolaire, et d’appuyer des soins aux patients de la plus grande qualité,
et de veiller à ce que les futures lignes directrices comportent des stratégies de prise en charge applicables
aux services des urgences.
© 2017 Canadian Diabetes Association.

Introduction Methods

Diabetes mellitus is an increasingly prevalent chronic disease
affecting an estimated 2 million Canadians, representing approxi-
mately 6.7% of the population (1). Diabetic ketoacidosis (DKA)
and hyperosmolar hyperglycemic state (HHS) are common, acute
metabolic complications of decompensated diabetes mellitus
characterized by an absolute or relative insulin deficiency. Common
precipitants to these hyperglycemic events include medication non-
compliance, infection such as pneumonia and urinary tract infec-
tion, and myocardial ischemia (2). If not effectively managed, these
2 conditions can lead to significant morbidity and mortality. The
estimated mortality rate for DKA ranges from 1% to 5%, whereas the
rate for HHS ranges from 5% to 20% (3). Many patients who have
experienced DKA or an HHS, after being discharged from the hos-
pital, have poor outcomes or return to the emergency department
(ED) and require admission (4).
Because it is the emergency physician (EP) who initially stabi-
lizes and treats these patients, it is important that appropriate
evidence-based management guidelines be followed to ensure a
standard of care for all patients in an attempt to prevent these
adverse outcomes. Multiple studies have demonstrated improved
patient outcomes after the implementation of a standardized DKA
protocol, such as decreased hospital lengths of stay and decreased
times to ketone clearance and correction of the anion gap (5,6). In
Canada, clinical practice guidelines for these hyperglycemic emer-
gencies were published in 2013 by the Canadian Diabetes Associa-
tion (CDA) using an expert committee representing the most current
evidence-based data for health-care professionals to ensure a stan-
dardization of care and improved clinical outcomes (7). This includes
intravenous fluid resuscitation, electrolyte replacement, adminis-
tration of insulin, correction of acidemia with bicarbonate when
appropriate, and identification and treatment of any precipitating
causes (7). However, despite the presence of established treat-
ment guidelines for DKA and HHS, there continues to be an issue
of significant practice variation and compliance rates among Cana-
dian EPs with treatment targets not being met, including variable
fluid, insulin and potassium administration, as well as premature
cessation of insulin infusions and reopening of the anion gap (8,9).
To date, physician attitudes toward these guidelines and their use
in the ED have not been well described in the literature.
The primary objectives of this study were to determine Cana-
dian EPs’ familiarity with and use of the 2013 CDA guidelines and
to identify any physician-perceived barriers to their implementa-
tion in Canadian EDs. We also sought to assess the practice vari-
ability among EPs in comparison with the guidelines. Secondary
objectives were to determine physician attitudes toward patients
who repeatedly present to the ED with hyperglycemic emergen-
cies, as well any additional factors that influence disposition
decisions.
Study design and population
In this cross-sectional study, a self-administered online survey
was distributed to 500 randomly selected members of the Cana-
dian Association of Emergency Physicians (CAEP) who had granted
the organization permission to contact them for prospective research.
Members who were not physicians or not currently practising clini-
cal emergency medicine were excluded. On our behalf, CAEP ran-
domly distributed the survey using Survey Monkey (http://
www.surveymonkey.com) on day 1 along with an e-mail message
informing the participants about the goals of the study. Two
follow-up e-mail messages were distributed on days 7 and 14.
Consent to participate in the study was implied by completion of
the survey. Approval was obtained from the Ottawa Hospital’s Health
Science Network Research Ethics Board. The survey was adminis-
tered between July and September of 2015.
Survey instrument
The content of our survey was pilot tested with 5 Canadian EPs
from 2 tertiary academic centres to ensure face and content valid-
ity. In its final version, the survey consisted of 23 questions includ-
ing demographic and work characteristics of ED physicians, with
16 of the questions relating to physician management of DKA and
HHS in the ED (Figures 1 and 2). A flowchart of the 2013 CDA guide-
lines was included in the survey (Figure 1), and participants were
asked to rate their familiarity with, use of and overall impression
of the guidelines by using a 7-point Likert scale. Other questions
attempted to identify physician-perceived barriers to guideline
implementation. Included in the survey were 3 clinical vignettes
(Figure 2) with different presentations of DKA and HHS and options
for selecting investigations and treatment options to determine
whether responses were consistent with guideline recommenda-
tions. We also sought to determine which disposition factors respon-
dents believe to be important when discharging patients from the
ED. In addition, we addressed physician attitudes toward patients
with repeat visits to the ED with hyperglycemic emergencies.
Outcome measures
The primary outcome measures were physician-reported famil-
iarity, use and usefulness of the 2013 CDA guidelines as rated with
a 7-point Likert scale, as well as any physician-perceived barriers
to guideline implementation in the ED and the current practice vari-
ability among physicians in comparison with the guidelines. Sec-
ondary outcomes included physician attitudes toward patients who
 ARTICLE IN PRESS
A.L. Hamelin et al. / Can J Diabetes xxx (2017) 1–0 3

Figure 1. 2013 Canadian Diabetes Association (CDA) Clinical Practice Guidelines. beta-OHB, beta-hydroxybutyrate; DKA, diabetic ketoacidosis; D5W, 5% dextrose in water;
D10W, 10% dextrose in water; ECFV, extracellular fluid volume; IV, intravenous; K, potassium; KCl, potassium chloride; NaCl, sodium chloride; NaHCO3, sodium bicarbonate.

repeatedly present to the ED with hyperglycemic emergencies and Results

any additional factors that influence disposition decisions.
Data analysis
Data entry and analysis were performed using SPSS version 21
statistical package (IBM Corporation, Chicago, Illinois, United States).
Descriptive statistics were summarized using proportions and means
as appropriate for the data. The associations between respondent
attitudes toward the 2013 CDA guidelines (familiarity, use and
reported usefulness) and respondent demographic information were
assessed by the appropriate univariate analyses; 2-sided p values
less than 0.05 were considered to be statistically significant.
Of the 500 CAEP members surveyed, 311 completed the survey,
resulting in a response rate of 62.2%. Respondent characteristics are
presented in Table 1.
Table 2 provides a summary of the physician responses to the
3 clinical vignettes, with selections consistent with the 2013 CDA
guidelines highlighted. Most physicians selected appropriate inves-
tigations to determine a precipitating cause in accordance with
the guidelines. The highest variance was noted in fluid, potassium
and bicarbonate administration. Although potassium administra-
tion was indicated in each vignette, respondent compliance ranged
from 65.5% to 91.7%. In the 1 vignette in which sodium bicarbonate
 ARTICLE IN PRESS
4 A.L. Hamelin et al. / Can J Diabetes xxx (2017) 1–0

Figure 2. Clinical vignettes. BP, blood pressure; ECG, electrocardiogram; ED, emergency department; HR, heart rate; IV, intravenous; KCl, potassium chloride; NaHCO3, sodium
bicarbonate; PCO2, partial pressure of carbon dioxide; RR, respiratory rate; SC, subcutaneous.
 ARTICLE IN PRESS
A.L. Hamelin et al. / Can J Diabetes xxx (2017) 1–0 5

Table 1 Table 3
Demographic and work characteristics of responding emergency physicians (N=311) 2013 Canadian Diabetes Association Practice Guidelines: Physician familiarity, use,
overall impression and identified barriers to implementation
Characteristic Number of
respondents (%) Element Number of
respondents (%)
Gender (n=310)
Male 194 (62.6) Familiarity (n=311)
Emergency medicine credentials (n=310) Very familiar 16 (5.1)
FRCP (EM) 80 (25.8) Familiar 56 (18.0)
CCFP (EM) 143 (46.1) Somewhat familiar 98 (31.5)
CCFP 35 (11.3) Neutral 21 (6.8)
Pediatrics 2 (0.6) Somewhat unfamiliar 54 (17.4)
PEM 4 (1.3) Unfamiliar 39 (12.5)
Resident 34 (11.0) Not at all familiar 27 (8.7)
Other 12 (3.9) Use (n=311)
Hospital Setting (n=311) Always 15 (4.8)
Major academic center 157 (50.5) Often 93 (30.0)
Community teaching hospital 133 (42.8) Sometimes 62 (19.9)
Non-teaching hospital 21 (6.8) Neutral 28 (9.0)
Province of current practice (n=309) Infrequently 38 (12.2)
Alberta 38 (12.3) Very Infrequently 22 (7.1)
British Columbia 46 (14.9) Never 53 (17.0)
Manitoba 14 (4.5) Overall impression (n=310)
New Brunswick 5 (1.6) Very useful 38 (12.3)
Newfoundland 6 (1.9) Useful 146 (47.1)
Nova Scotia 20 (6.5) Somewhat useful 75 (24.2)
Northwest Territories 3 (1.0) Neutral 34 (11.0)
Nunavut 0 (0) Somewhat not useful 13 (4.2)
Ontario 134 (43.3) Not very useful 3 (1.0)
Prince Edward Island 1 (0.3) Not at all useful 1 (0.3)
Quebec 28 (9.1) Barrier (n=242)
Saskatchewan 14 (4.5) Education issues 135 (55.8)
Yukon Territory 0 (0) Staffing/human resource issues 62 (25.6)
Characteristic Mean (SD) Poor policy adherence 60 (24.8)
Number of years in practice (n=286) 14.4 (10.2) Lack of time/disruption of flow 60 (24.8)
Hours per week of patient care in the ED (n=307) 28.7 (11.4) Low incidence/unfamiliarity with DKA/HHS 32 (13.2)
Percent clinical practice (n=311) 86.0 (20.3) Lack of access to medications 6 (2.5)
Other 56 (23.1)
CCFP, Canadian College of Family Physicians; CCFP (EM), Canadian College of Family
Physicians—Emergency Medicine; ED, emergency department; FRCP (EM), Fellow DKA, diabetic ketoacidosis; HHS, hyperglycemic hyperosmolar state.
of the Royal College of Physicians—Emergency Medicine; PEM, Pediatric Emer-
gency Medicine Fellowship; SD, standard deviation.

administration is recommended according to the guidelines, it was
selected by half of respondents. In addition, there was also some
variation in insulin administration. About 20% of respondents
selected an insulin bolus, when an insulin infusion is indicated in
the guidelines.
Physicians’ familiarity with and use and impression of the guide-
lines, in addition to physician-perceived barriers to their imple-
mentation, are shown in Table 3. The majority of physicians reported
the guidelines to be useful (83.6%); however, only 54.6% of respon-
dents were familiar with the guidelines, and only 54.7% claimed to
use them in clinical practice. The most frequently reported barrier
to guideline implementation was a lack of education (56.0%).
Respondents familiar with the guidelines had fewer mean number
of years in clinical practice compared with those not familiar with
the guidelines (9.5±9.0 vs. 14.1±10.6 years, p<0.01).
Table 4 describes physician attitudes toward patients who present
to the ED repeatedly for treatment of hyperglycemic emergencies.
Nearly half (44.2%) of respondents found these patients to be prob-
lematic, with the most common perceived cause of return visits
being patient noncompliance (90.3%).
Table 5 provides physician responses in rating the importance
of specific aspects of patient disposition from the ED. The disposition

Table 2
Clinical vignettes: Number (%) of responding emergency physicians selecting investigations and treatment options consistent with the 2013 Canadian Diabetes Association
(CDA) Guidelines

Investigation Vignette 1 (n=310) Vignette 2 (n=307) Vignette 3 (n=308)

Liver enzymes 125 (37.9) 201 (61.8) 186 (57.1)

Cardiac enzymes 31 (9.4) 314 (96.6)* 237 (72.7)
Urine dip and microscopy 314 (95.2)* 272 (83.7) 292 (89.6)
Urine culture 291 (88.2)* 189 (58.2) 210 (64.4)
Blood culture 291 (88.2) 185 (56.9) 182 (55.8)
Chest x-ray 228 (69.1) 310 (95.4)* 253 (77.6)
ECG 216 (65.5) 321 (98.8)* 292 (89.6)
Treatment Vignette1 (n=310) Vignette2 (n=307) Vignette3 (n=308)
1–2 L crystalloid rapid bolus 228 (87.3) 266 (81.8) 185 (56.7)
250–500 mL/h crystalloid infusion 188 (57)* 189 (58.2)* 237 (72.7)*
0.1 U/kg SC or IV insulin 71 (21.5) 69 (21.2) 65 (19.9)
0.1 U/kg/h insulin infusion 293 (88.8)* 247 (76)* 231 (70.9)*
20–40 mmol/L IV potassium chloride 267 (80.9)* 213 (65.5)* 299 (91.7)*
1–2 ampoules IV bicarbonate 9 (2.7) 168 (51.7)* 11 (3.4)
ECG, electrocardiogram; IV, intravenous; SC, subcutaneous.
* Selections consistent with 2013 CDA Guidelines.
 ARTICLE IN PRESS
6 A.L. Hamelin et al. / Can J Diabetes xxx (2017) 1–0

Table 4 or modify treatment with an oral hypoglycemic agent (49.5%) or

Repeat emergency department visits by patients with hyperglycemic emergen- an insulin regimen (45.0%).
cies: Physician concern, perceived contributing factors and reported ideal changes
to optimize patient management

Concern Number of Discussion

(n=310) respondents (%)

Problematic 137 (44.2) Interpretation of results

Neutral 58 (18.7)
Not problematic 115 (37.1)
The results of this large national survey demonstrate that EP
Cause Number of
awareness and use of the current 2013 CDA guidelines for DKA and
(n=309) respondents (%)
HHS was low across Canada. However, many physicians believed
Patient noncompliance 279 (90.3)
that the guidelines were, in fact, useful. Because a lack of educa-
Patient mental health 212 (68.6)
Poor economic status of patient 183 (59.2) tion was the most frequently reported barrier to implementation
Lack of coordination and follow up 144 (46.6) of the guidelines, this suggests the need to improve knowledge trans-
with family physician lation strategies across Canadian EDs to standardize management
Lack of patient diabetes education 137 (44.3)
of DKA and HHS and support the highest quality of patient care.
Lack of access to specialized diabetes care 99 (32.0)
Other 23 (7.4) When the clinical vignettes were used, it was found that there were
low rates of potassium administration and variability in sodium
Change Number of
(n=303) respondents (%)
bicarbonate and insulin administration, all factors that can have a
considerable clinical impact on the outcome for these patients (10).
Improved coordination and follow up 242 (77.8)
Although the use of sodium bicarbonate remains controversial,
with family physician
Increased diabetes education for patients 222 (71.4) serum potassium should be closely monitored throughout the man-
Increased availability to diabetes specialists 144 (46.3) agement of DKA and HHS, because untreated hypokalemia may lead
Improved access to proper medications 113 (36.3) to arrhythmias and cardiac arrest (10). This variability could poten-
Increased education for physicians and nurses 77 (24.8)
tially be explained by the fact that the CDA expert committee is com-
Improved staffing/human resources 62 (19.9)
Improved policies and procedures 53 (17.0) posed mainly of internists, specifically endocrinologists, whose
Other 29 (9.3) patient populations and practice patterns vary from EPs in the need
for management of hyperglycemic emergencies (11). Therefore com-
pliance with guidelines may not only involve simply achieving
Table 5 awareness of their existence, but also improving guidelines so that
Physician-identified emergency department disposition factors for patients with
they are clinically relevant and sensible to EPs.
hyperglycemic emergencies
It has previously been demonstrated that many patients pre-
Disposition factors Number of senting to the ED with hyperglycemic emergencies make return visits
respondents (%)
to the ED (4). It therefore comes as no surprise that half of physi-
Importance of early follow up with another cians in our survey reported that they find repeat patients prob-
physician (n=310) lematic, with the most frequently perceived cause of return visits
Important 290 (93.5)
Neutral 11 (3.5)
being patient noncompliance. In addition, half of respondents also
Not important 9 (2.9) perceived that a lack of follow up with a family physician is prob-
Ability to arrange early follow up with another lematic, which is potentially another major cause of repeat visits
physician (n=311) to the ED. In these circumstances, the ED would perhaps provide
Easy 167 (53.7)
Neutral 42 (13.5)
the only opportunity to optimize medical therapy. In our study, fewer
Difficult 102 (32.8) respondents were likely to start or modify treatment with an oral
Likelihood of referral to a specialist (n=311) hypoglycemic agent or an insulin regimen and were more likely to
Likely 232 (75.0) refer patients to a specialist. It has been previously demonstrated
Neutral 26 (8.4)
Unlikely 53 (17.0)
that the introduction of a standardized DKA protocol at a tertiary
Likelihood of starting or modifying treatment with care centre results in a significant increase in referrals to an endo-
an oral hypoglycemic agent (n=311) crinologist or a diabetes nurse specialist (12). Although patient care
Likely 154 (49.5) provided by an endocrinologist has been shown to decrease hos-
Neutral 44 (14.1)
Unlikely 113 (36.3)
pital lengths of stay, it is unclear which patients require a referral
Likelihood of starting or modifying an insulin and admission to the hospital (11). Currently, the CDA guidelines
regimen (n=311) do not provide disposition guidance on such decisions, which might
Likely 140 (45.0) be useful for EPs in an attempt to improve patient outcomes while
Neutral 42 (13.5)
Unlikely 129 (41.5)
ensuring appropriate use of hospital resources.
Factors influencing decision to refer to specialist (n=309)
Poorly controlled diabetes despite therapy 289 (93.5) Previous studies
Unavailability of family physician 280 (90.6)
Increased frequency of hyperglycemic episodes 270 (87.4)
Increased severity of hyperglycemic episodes 246 (79.6)
Although no previous study has explored physician attitudes
Other 113 (36.3) toward DKA or HHS management guidelines, other studies have
attempted to explore the practice variability that currently exists.
An observational study by Sola et al. investigated DKA manage-
ment of both patients in the ED and admitted patients over a period
factor physicians reported as being the most important was early of 6 years. In this study, there was an average delay in fluid admin-
follow up with another physician (93.5%). However, only slightly istration of 102 minutes and an average delay in insulin administration
more than half (53.7%) of respondents believed their ability to of 144 minutes after arrival in the ED, with an under-replacement
arrange follow up was easy. Most respondents were likely to refer of fluids and potassium therapy as compared with published stan-
patients to a specialist (75.0%). Fewer physicians were likely to start dard guidelines (13). Similarly, a retrospective study by Devalia, in
 ARTICLE IN PRESS
A.L. Hamelin et al. / Can J Diabetes xxx (2017) 1–0
which management of acute DKA in the first 4 hours was assessed,
demonstrated good awareness of an established treatment proto-
col; however, it was not always followed with suboptimal fluid man-
agement and electrolyte monitoring (14). Similar studies have been
conducted in the United Kingdom on inpatient management of hyper-
glycemic emergencies, demonstrating poor adherence rates in terms
of fluid management and metabolic monitoring; however, these con-
clusions are difficult to generalize to patients in the ED setting (15,16).
The use of bicarbonate remains controversial. A systemic review
by Chua et al, which included 3 randomized controlled trials of bicar-
bonate therapy versus no bicarbonate therapy, demonstrated a tran-
sient reduction in metabolic acidosis in the initial 2 hours but no
evidence of increased glycemic control or clinical efficacy (17). Ret-
rospective studies of bicarbonate use were also reviewed and showed
an increased risk of cerebral edema, particularly in children (17).
Even though it becomes difficult to generalize the risks of bicar-
bonate therapy in adults in comparison with children, the evi-
dence to date does not support the use of bicarbonate therapy in
view of the possible clinical and physiologic harm and the lack of
clinical or sustained physiologic benefits. Further research is needed
to clarify the role bicarbonate administration plays in severe DKA,
specifically in adults, to guide future treatment protocols.
Strengths and limitations
Although we achieved an overall response rate of 62.2%, which
is very reasonable for a physician survey, the results may not be gen-
eralizable to community EDs. Respondents working in commu-
nity EDs were underrepresented in our study, and the CAEP database
may not be representative of all Canadian physicians working in EDs.
Also, given that we surveyed physicians who are voluntary members
of an emergency medicine database, this may have resulted in selec-
tion of physicians whose work is more strongly influenced by current
research. Another key limitation, as with any survey, is that our data
are based on self-reporting and may not reflect actual practice,
because respondent bias may be present. Finally, a copy of the guide-
lines was included in the survey before the questions related to the
clinical vignettes were presented. It is therefore possible that respon-
dents were more likely to use the guidelines in their answers com-
pared with clinical practice.
Clinical implications
The general acceptance of the 2013 CDA guidelines in this study
gives us confidence that EPs are ready for a protocol that will help
guide management of DKA and HHS in the ED. Several factors may
be contributing to the low compliance rates with guidelines, includ-
ing low levels of protocol awareness and questionable applicabil-
ity of the existing protocol in the ED setting. One possible solution
is for clinical diabetes specialists to work collaboratively with EPs
to develop future practice guidelines so that they are clinically sen-
sible in the ED. Another consideration is that EDs should consider
developing pre-printed order sets for managing patients with hyper-
glycemic emergencies. One previous study demonstrated that the
implementation of a standardized pre-printed protocol for DKA man-
agement significantly improved adherence rates with high satis-
faction rates from treating physicians, nurses and patients (12).
Finally, the current guidelines do not include any disposition guid-
ance for EPs, which is a topic that future guidelines could include
to help EPs better utilize hospital resources.
Research implications
The use of protocols for the management of hyperglycemic emer-
gencies can improve clinical outcomes; however, further research
7
should focus on determining the factors related to physician aware-
ness and use of management protocols to enhance knowledge trans-
lation and generate strategies to disseminate and implement future
guidelines for acute diabetic complications. Further research is also
needed to risk stratify patients to determine whether an acute hyper-
glycemic emergency requires either outpatient follow up with a
family physician or a diabetes specialist or hospitalization, which
in turn would provide disposition guidance for EPs. Lastly, the use
of bicarbonate in the treatment of DKA and HHS remains a topic
of controversy, with even less evidence available for its use in adult
populations. Further research is needed to clarify the role of bicar-
bonate administration in DKA and HHS to develop future practice
guidelines.
Conclusions
Canadian EPs generally find the CDA guidelines helpful; however,
many are unaware that these guidelines exist, and barriers to their
implementation have been reported. In addition to development
of knowledge translation strategies to improve the use of evidence-
based practices, the guidelines need to be created and updated in
a collaborative manner by diabetes specialists and EPs to ensure that
acute management strategies are applicable to the ED setting. This
will ensure adherence to best practices, which in turn will improve
the outcomes of patients with hyperglycemic emergencies.
Acknowledgments
The authors thank Angela Marcantonio and Catherine Clement
for their assistance and support for this study. This research received
no specific grant from any funding agency, commercial or not-for-
profit sector.
Author Contributions
All named authors made significant contributions to the study
and drafted and approved the final version of the manuscript.
References
1. Government of Canada. Diabetes, by age group and sex (percent). Statistic Canada;
2016. http://www.statcan.gc.ca/tables-tableaux/sum-som/l01/cst01/health53b
-eng.htm. Accessed November 3, 2016.
2. Maletkovic J, Drexler A. Diabetic ketoacidosis and hyperglycemic hyperosmolar
state. Endocrinol Metabol Clin N Am 2013;42:677–95.
3. Kitabchi AE, Umpierrez GE, Miles JM, Fisher JN. Hyperglycemic crises in adult
patients with diabetes. Diabetes Care 2009;32:1335–43.
4. Yan JW, Gushulak KM, Columbus MP, et al. Sentinel visits in emergency depart-
ment patients with diabetes mellitus as a warning sign for hyperglycemic emer-
gencies. CJEM 2017 (in press).
5. Bull SV, Douglas IS, Foster M, Albert RK. Mandatory protocol for treating adult
patients with diabetic ketoacidosis decreases intensive care unit and hospital
lengths of stay: Results of a nonrandomized trial. Crit Care Med 2007;35:41–6.
6. Waller SL, Delaney S, Strachan MWJ. Does an integrated care pathway enhance
the management of diabetic ketoacidosis? Diabet Med 2007;24:359–63.
7. Goguen J, Gilbert J. Canadian diabetes association clinical practice guidelines
expert committee. Hyperglycemic emergencies in adults. Can J Diabetes
2013;37:S72–6.
8. Murthy R, Assal A, Rushton M, et al. Not so sweet? Management of diabetic keto-
acidosis at a Canadian academic centre. Can J Diabetes 2014;38:S51–2.
9. Golds G, Arnason T. Diabetic ketoacidosis management at a Canadian tertiary
care centre as compared to the Canadian diabetes association guidelines. Can
J Diabetes 2016;40:S50.
10. Gosmanov AR, Gosmanova EO, Dillard-Cannon E. Management of adult dia-
betic ketoacidosis. Diabetes Metab Syndr Obes 2014;7:255–64.
11. Levetan CS, Passaro MD, Jablonski KA, Ratner RE. Effect of physician specialty
on outcomes in diabetic ketoacidosis. Diabetes Care 1999;22:1790–5.
 ARTICLE IN PRESS
8 A.L. Hamelin et al. / Can J Diabetes xxx (2017) 1–0
12. Mohamed M, De Sousa F, Boyle L, et al. Management of DKA and The Ottawa
Hospital. Presented at The Ottawa Hospital Patient Safety Conference, Ottawa,
Ontario, Canada, October 2016.
13. Sola E, Garzón S, García-Torres S, et al. Management of diabetic ketoacidosis in
a teaching hospital. Acta Diabetol 2006;43:127–30.
14. Devalia B. Adherence to protocol during the acute management of diabetic keto-
acidosis: Would specialist involvement lead to better outcomes? Int J Clin Pract
2010;64:1580–2.
15. Crasto W, Htike Z, Turner L, Higgins K. Management of diabetic ketoacidosis fol-
lowing implementation of the JBDS guidelines: Where are we and where should
be go? Br J Diabetes Vasc Dis 2015;15:11–16.
16. Dhatariya KK, Nunney I, Higgins K, et al. National survey of the management
of diabetic ketoacidosis (DKA) in the UK in 2014. Diabet Med 2016;33:252–
60.
17. Chua HR, Schneider A, Bellomo R. Bicarbonate in diabetic ketoacidosis—a sys-
tematic review. Ann Intensive Care 2011;1:1–12.