Review

Risk factors and prevention strategies for diabetic

ketoacidosis in people with established type 1 diabetes
Dominic Ehrmann, Bernhard Kulzer, Timm Roos, Thomas Haak, Mohammed Al-Khatib, Norbert Hermanns

Lancet Diabetes Endocrinol Diabetic ketoacidosis (DKA) is a serious acute complication of type 1 diabetes, which is receiving more attention given
2020; 8: 436–46 the increased DKA risk associated with SGLT inhibitors. Sociodemographic and modifiable risk factors were identified
Research Institute Diabetes with strong evidence for an increased risk of DKA, including socioeconomic disadvantage, adolescent age
Academy Mergentheim, Bad
(13–25 years), female sex, high HbA1c, previous DKA, and psychiatric comorbidities (eg, eating disorders and
Mergentheim, Germany
(D Ehrmann PhD, depression). Possible prevention strategies, which include the identification of people at risk based on non-modifiable
Prof B Kulzer PhD, T Roos MSc, sociodemographic risk factors, are proposed. As a second risk mitigation strategy, structured diabetes self-
Prof N Hermanns PhD); management education that addresses modifiable risk factors can be used. Evidence has found that structured
Department of Clinical
education leads to reduced DKA rates. Knowledge of these risk factors and potent risk mitigation strategies are
Psychology and Psychotherapy,
University of Bamberg, important to identify subgroups of people with an elevated DKA risk. This knowledge should also be used when
Bamberg, Germany adjunct therapy options with an increased DKA risk are considered. Prevention of DKA in people with type 1 diabetes
(D Ehrmann, Prof B Kulzer, is an important clinical task, which should also be addressed when SGLT inhibitors are part of therapy.
Prof N Hermanns); Diabetes
Clinic Mergentheim,
Bad Mergentheim, Germany Introduction death from diabetic coma or DKA was associated with
(Prof B Kulzer, Diabetic ketoacidosis (DKA) is a potentially life- the largest percentage of the estimated loss in life
Prof Thomas Haak MD, threatening complication caused by an insulin deficiency expectancy (29·4% in men and 21·7% in women).20
Prof N Hermanns);
and HealthPlus Diabetes &
and is characterised by metabolic acidosis, hyper­ Recurrent episodes of DKA in particular are associated
Endocrinology Centre, ketonaemia, and hyperglycaemia, for which diagnostic with an increase in mortality of up to a 23·3% risk of
Abu Dhabi, United Arab criteria differ (>200 mg/dL, >11·1 mmol/L, >250 mg/dL, death in people who have had more than five episodes of
Emirates (M Al-Khatib MD) or >13·9 mmol/L).1–4 The symptoms of DKA develop over DKA compared with 5·2% in people with one episode.21
Correspondence to: several hours and include nausea, vomiting, polyuria, A new aspect of type 1 diabetes is adjunct treatment
Prof Norbert Hermanns,
and excessive thirst. DKA is the presenting manifestation with SGLT inhibitors that regulate glucose concentrations
Research Institute Diabetes
Academy Mergentheim, of diabetes in approximately 30% of people with type 1 by increasing urinary glucose excretion. This treatment
Bad Mergentheim 97980, diabetes, with higher numbers in children (up to 55%5) can lead to DKA without marked hyperglycaemia. Before
Germany than adults (up to 6%6).1,7–10 In people with established the use of SGLT inhibitors, the percentage of patients
hermanns@fidam.de
type 1 diabetes, a systematic review11 showed that, overall, with DKA with an admission glucose concentration of
the incidence of DKA is as high as 5·6 events per less than 200 mg/dL (11·1 mmol/L) was 3·0% of all
100 person-years, with a prevalence of up to 12·8 per patients with DKA; however, this percentage is expected
100 people, depending on the clinical setting, region of to rise with the increased use of SGLT inhibitors.22 In
the world, state of development of a country, and level of addition, there is evidence that SGLT inhibitors lead to
income inequality.1,10–12 an increase in DKA episodes in people with type 1
After a decline in DKA-related hospital admissions in diabetes.23–26 Therefore, new guidelines and protocols
the USA from 2000 to 2009, the hospital admission rate have been published to mitigate this increased DKA risk
for people with diabetes increased by 54·9% from with the use of SGLT-2 and dual inhibitors.25,27–29 Given
2009 (19·5%) to 2014 (30·2%).13 Notably, in young people the off-label use30,31 and the approval of SGLT inhibitors
with type 1 diabetes (≤20 years of age), the mean hospital as adjunct therapy in type 1 diabetes in Europe and Japan,
admission rate for DKA was 3·3 times higher than the the frequency of DKA will most likely increase.26
mean admission rate for severe hypoglycaemia (4·81 per Because of the unchanged high prevalence and
100 person-years vs 1·45 per 100 person-years).14 These sequelae, the prevention of DKA remains a challenge for
hospital admission rates lead to a high economic burden the care of people with type 1 diabetes. Although the
of DKA.15 In the UK, cost estimations up to £2064 per pathophysiology of DKA, the precipitating causes, and
DKA episode were reported.16 In the USA, mean DKA- the acute management have been expertly reviewed,1,2,32
related hospital charges amount to between US$20 428 knowledge about prevention strategies for people with
and $26 566,17,18 resulting in aggregate charges for established type 1 diabetes remains scarce.32 Thus, we
diabetes with DKA of $5·1 billion in 2014.17 concentrate on three aspects in our Review. As a first step
The overall mortality of DKA has remained an in the prevention of DKA, we propose that the timely
important problem for more than 30 years,19 with the US identification of people with an increased risk of DKA
Centers for Disease Control and Prevention estimating is necessary, and we summarise evidence on socio­
the case fatality rate at 0·4%.13 Data from the Scottish demographic, non-modifiable, and modifiable risk
nationwide register showed that in people with type 1 factors. As a second step, modifiable risk factors should
diabetes younger than 50 years, nearly 16% of all deaths be reduced, and we review evidence for the effectiveness
were caused by diabetic coma or DKA.20 Furthermore, of diabetes self-management education and other

436 www.thelancet.com/diabetes-endocrinology Vol 8 May 2020


measures. For the third step, we apply this knowledge to
the therapy with SGLT inhibitors and review the
relevance of these preventive strategies in mitigating the
DKA risk associated with the use of SGLT inhibitors.
Identification of people at risk
A substantial proportion of people with type 1 diabetes
are repeatedly affected by DKA, which suggests the
existence of risk groups for DKA.33,34 From the reviewed
literature, we describe potentially non-modifiable and
modifiable risk factors for recurrent DKA.
Non-modifiable risk factors
The DKA risk begins to increase after early childhood
(>5 years) and plateaus in the age group between 13 years
and 25 years; thereafter, the DKA risk decreases with
increasing age.14,35–41 Support for this inverted U-shaped
association comes from Everett and Mathioudakis,42 who
showed that in the paediatric sample, higher age
(≥10–19 years) was associated with a higher recurrence
rate of DKA compared with younger age (2–9 years; one
to three readmissions, OR 1·47, 95% CI 1·15–1·87),
whereas in the adult sample, older age (≥40 years) was a
protective factor for recurrent DKA compared with
younger adult age (20–39 years; OR 0·85, 95% CI
0·81–0·88). Psychosocial develop­ments in the course of
puberty, such as striving for autonomy, the transfer of
responsibilities for diabetes care from parents to the
adolescent, or the transition from paediatric care to adult
care, could be factors that contribute to this risk
distribution.43
There is ample evidence that socioeconomic dis­
advantages are a major risk factor for DKA. Several
indicators of socioeconomic disadvantages such as low
socioeconomic status (SES), area-level deprivation (a
geographically based measure of SES), low income,
homelessness, and health insurance status were risk
factors.11,21,33,35,36,38,41,42,44–53 One systematic review concluded
that in all but one of 22 reviewed studies, lower SES and
higher area-level deprivation were associated with an
increased risk of DKA.44 Previous data showed that
even in a moderately-deprived area, the risk of DKA
readmissions was 4·2 times (for 1–3 readmissions) to
7·8 times (for >3 readmissions) that in the least deprived
area in the paediatric sample, compared with 1·2 times
(for 1–3 readmissions) to 1·3 times (for >3 readmissions)
the risk in the adult sample.42 Having private insurance
was consistently associated with the lowest risk of
recurrent DKA readmissions in the adult sample, with a
risk reduction compared with those without private
insurance ranging from 24% to 70%.35,42 In a representative
adult sample from the USA, the lowest income quartile
had a 50% increased risk of multiple episodes of DKA
(≥4 readmissions for DKA: OR 1·46, 95% CI 1·30–1·64)
compared with the highest quartile.35 The mediating
factors for this association are not fully understood,35 but
little access to health care,44 low general education, and
www.thelancet.com/diabetes-endocrinology Vol 8 May 2020
Review
low health literacy are potent contributors. In summary,
socioeconomic factors play a huge role in the recurrence
of DKA and seem to be more pronounced in children
and youth.42
Female sex was a risk factor for DKA-related hospital
admissions14,36,38,54 and recurrent episodes of DKA in many
studies, in various countries and cultures.33,35,41,55 A study
using the US National Readmission Database analysed
181 284 DKA admissions and showed that female sex
increased the risk of recurrent DKA admissions by up to
40% compared with male sex.35 A Chinese study observed
a 2-times increased risk of recurrent DKA in women and
girls (OR 2·12, 95% CI 1·50–3·04).33 In the adult sample
of the German Diabetes Prospective Follow-up Registry
(DPV) registry consisting of 46 966 adults with type 1
diabetes, no significant difference in DKA-related
hospital admissions between women and men (2·59 vs
2·46 per 100 patient-years, p>0·05) were detected.37
However, in the paediatric DPV sample, DKA was more
frequent in girls than in boys (5·35 vs 4·34 per 100 patient-
years, p=0·002).14 This pattern was corroborated by
Everett and Mathioudakis.35,42 In summary, recurrent
DKA seems to be more frequent in girls and adolescent
women than in their male counterparts. Several possible
reasons for this finding have been discussed such as
deliberate omissions of insulin injections because of
body image issues,49 worse glycaemic control, and (sub­
clinical) mental health issues (eg, distress, depression,
eating disorders56,57), which are more common in women
than in men.42
During pregnancy, recognition of DKA can be
particularly difficult, since DKA tends to occur at lower
glucose concentrations and somatic warning symptoms
(eg, nausea, vomiting) lose their specificity for indicating
DKA.58
In a large international comparison with 49 859 children
and adolescents (<18 years) in the USA, Germany and
Austria, and England and Wales, ethnic minorities had a
1·27-times greater risk of DKA (OR 1·27, 99% CI
1·11–1·44) compared with non-ethnic minority groups,
with the highest risk for ethnic minorities in the USA
(OR 1·78, 99% CI 1·45–2·18).54 In Germany, migration
background was consistently associated with higher DKA
risk.14,37,40 However, it is difficult to separate ethnicity and
migration background and their covariates (eg, poverty,
absence of health insurance, low educational status, and
health literacy) from the effect of SES.
Modifiable risk factors
There is evidence in children,33,41,59 adolescents,33,34 and
adults33,34,60 that previous DKA increases the risk of future
episodes of DKA, which indicates the existence of specific
risk groups for DKA. Similarly, the studies by Yan and
colleagues34 and Hurtado and colleagues61 found that
high percentages (33·8–40·8%) of people admitted to
hospital because of hyperglycaemia or DKA were
readmitted within 30 days for the same reason. An
437
 Review
analysis of the Dose Adjustment for Normal Eating
(DAFNE) research database found that adults with at
least one DKA episode in the 12 months before
participation in the DAFNE course had a nearly 9·5-times
higher risk of DKA during the following 12 months
compared with those without a DKA episode in the
12 months before DAFNE (OR 9·35, 95% CI
2·75–31·75).62 In children, DKA at the time of the
diagnosis of type 1 diabetes increased the risk of
readmittance to hospital for DKA compared with those
without DKA at diagnosis.41
There is convincing evidence that elevated HbA1c
concentrations are a risk factor for recurrent DKA in
studies including children,40,49,54 adolescents,14,40,45,50,63–65
adults,37,60,63,64,66 and mixed-age samples.21,33,36,51,55 Higher
HbA1c was exponentially related to higher DKA rates for
all ages.14,37 Across three multinational registries and all
age ranges, those with HbA1c between 7·5% and 8·9%
(58–74 mmol/mol) had a 2·4-times greater risk of DKA
(OR 2·40, 95% CI 1·99–2·90), and even those with HbA1c
at or higher than 9·0% (75 mmol/mol) had an 8·0-times
greater risk of DKA (OR 8·04, 95% CI 6·72–9·62), both
compared with an optimum glycaemic control of less
than 7·5% (<58 mmol/mol).14,37,54 Elevated HbA1c concen­
trations might be indicative of general problems with
diabetes management (eg, non-adherence67). Since HbA1c
values are regularly monitored in type 1 diabetes, it is a
comparatively easily accessible risk marker for DKA.
Studies analysing the reasons for recurrent DKA
admission consistently found that in up to 75% of
patients, poor adherence to insulin therapy (ie, missed
insulin doses) was the immediate contributing
factor.1,6,46,55,67–69 Sayed and colleagues47 suggested that not
only non-adherence to insulin therapy but also more
general aspects of diabetes self-management, such as an
absence of records in a diabetes diary, were associated
with an increased DKA risk. Therefore, it is reasonable to
regularly assess more general problems with self-
management to identify people at risk of DKA.
There is consistent evidence in children, adolescents,
and adults that the presence of a psychiatric disorder,66,70,71
particularly an eating disorder,59,72 depressive disorder,46,73
or schizophrenia,74 is a risk factor for DKA. The analysis
of Del Degan and colleagues67 showed that the diagnosis
of a psychiatric disorder is a possible independent risk
factor for DKA (OR 2·72, 95% CI 0·94–7·89, p=0·06),
even when adjusted for non-adherence, although the
association was not statistically significant. In addition,
treat­
ment with antipsychotic or antidepressant medi­
cation21,75 is associated with increased DKA risk, but it is
unclear whether this is because of a direct side-effect of
the medication or because of the disorder per se.76
Clinical and subclinical mental health issues are more
common in people with diabetes than in people without
diabetes,56,57 and are frequently associated with risk
factors such as poor adherence77 and poor glycaemic
control.78,79 In summary, subclinical mental health issues
438
and psychiatric comorbidities contribute to an elevated
DKA risk. Thus, regular screening should be imple­
mented in clinical practice.57
Acute infection is one of the most frequent precipitating
risk factors of DKA,46,55,80 and demands immediate
attention and treatment.1 In addition, surgery, trauma,
dehydration, and the resulting activation of the glucose
counter-regulatory system can also increase the
probability of DKA, if insulin is not adjusted adequately.1
Somatic comorbidities, such as diabetic nephropathy,60
gastroparesis,55 and epilepsy,81 are also mentioned as risk
factors. Furthermore, careful consideration should also
be given when medicating oncological diseases with
immune checkpoint inhibitors because they were found
to increase the risk of type 1 diabetes and DKA.82
A European study analysing 3250 people with type 1
diabetes from 16 European countries found no
association between alcohol consumption and DKA,83
nor did the study by Cooper and colleagues,63 which
found similar rates of alcohol intake or abuse in those
being readmitted for DKA and those with only one
admission with DKA over 5 years. In other studies,
alcohol abuse46 and heavy drinking60,84 were associated
with higher DKA rates; thus these factors should be
addressed in the prevention of DKA.
Cocaine use was associated with a 4·4-times increased
DKA risk compared with non-users.68 A previous study
also showed that people who used cannabis had a
2·5-times greater risk of DKA.85 In addition, drug abuse
seems to be associated with impaired general health-
related behaviour, given that all-cause 30-day readmission
after a DKA event was elevated in patients who used
drugs.61
The characteristics of diabetes care appear to also be
associated with DKA risk.37,86 Registry data showed that
people cared for in centres that are less experienced or
less specialised in type 1 therapy (<50 people with type 1
diabetes) had a nearly 42% increased risk of DKA.37,86
This difference seems especially relevant for patients
who use a pump, because Hoshina and colleagues86
found significantly higher DKA rates among adults who
used a pump in smaller clinics than in adults on
multiple daily insulin injection (MDI) therapy. Children
with type 1 diabetes who had a DKA event were less
likely to attend a meeting with an endocrinologist
within 120 days before the DKA.87 Furthermore, good
quality of care, such as regular visits to the
endocrinologist and continuous care, seems to be a
protective factor for DKA, as lower attendance and poor
contact with diabetic health-care teams were associated
with recurrent DKA.47,55,64 Fragmentation of care
(ie, admission to hospital for recurrent DKA at more
than one hospital) was also an important risk factor that
was associated with an 88% higher probability of having
multiple DKA events.88 This evidence suggests that the
diabetological expertise of clinics and the use of the
health-care system are protective factors for DKA.
www.thelancet.com/diabetes-endocrinology Vol 8 May 2020

A meta-analysis published in 1997 showed an increased
risk of DKA in patients who use a pump.89 However,
newer registry data such as the type 1 diabetes-exchange
registry (continuous subcutaneous insulin infusion
[CSII]: 2% with DKA vs MDI: 4% with DKA; p=0·00239) or
the German DPV registry (CSII: 2·21 per 100 patient-
years vs MDI: 3·12 per 100 patient-years37) indicate no
increased risk of DKA in patients who use a pump.70 This
development might be because of evolving pump
technology (eg, alarms for malfunctions or higher dose
accuracy), greater understanding of infusion set
problems (eg, cannula dislodgement, kinking, occlusion),
and subsequent intensive education and training of
patients who use a pump about the importance of the
insulin infusion set.70,90
Review of risk factors
The review of risk factors showed that recurrent DKA is
especially common in late childhood, adolescence, and
young adulthood and is more likely to occur in girls or
young women. Social disadvantages and migration
background are further risk factors that are difficult or
impossible to change. Previous DKA, elevated HbA1c,
non-adherence, mental health issues, somatic comor­
bidities, drug or alcohol abuse, and low quality of diabetes
care are principally modifiable risk factors, which should
be addressed by preventive measures. However, not all
risk factors were analysed simul­taneously, which makes
it difficult to quantify the independent risk associated
with each of these factors. Empirical evidence from
prospective register studies is needed to find out the
relative weight of each factor. These factors could then be
combined to form a risk score, allowing a better
quantification of individual DKA risk.
In figure 1, we present a conceptual model, in which an
initial risk mitigation strategy might be the identification
of subgroups of people with diabetes with an elevated
risk of DKA by assessing sociodemographic, non-
modifiable, and modifiable risk factors. Knowing which
people are at an increased risk of DKA could be helpful to
direct specific preventive measures to these risk groups.
Furthermore, the non-modifiable risk factors, such as
younger age, female sex, migration background, and SES
factors, can be of clinical importance for the prevention
of DKA in addition to their capacity to identify people at
risk. They might be used to tailor prevention strategies to
achieve approaches that are age, sex, socially, and
culturally appropriate. However, the effectiveness of such
tailored strategies should be further evaluated.
Preventive strategies to reduce DKA risk
A second step for DKA prevention is directly addressing
modifiable risk factors via diabetes self-management
education, and medical and psychological interventions
(figure 1).
A promising strategy to address several modifiable risk
factors simultaneously is diabetes self-management
www.thelancet.com/diabetes-endocrinology Vol 8 May 2020
Review
Preventive strategy 1: identification of
people at risk
Identification Identification
Non-modifiable risk Modifiable risk factors
factors • Previous DKA
• Low socioeconomic • Elevated HbA1c
status • Non-adherence
DKA risk
• Adolescent age • Alcohol or drug abuse
• Female sex • Quality of care
• Ethnicity or migration • Poor mental health
• Somatic comorbidities
Tailoring Preventive strategy 2: reduction of risk Addressing
factors
• Diabetes self-management education
• Medical and psychological interventions
• Telemedical and digital approaches
Risk enhancing Risk reducing
Figure 1: Conceptual model of preventive strategies for DKA
Preventive strategy I: non-modifiable and modifiable risk factors associated with
an increased risk for DKA (red arrows) can be used to identify people with an
increased DKA risk (green arrows). Preventive strategy II: preventive actions
should be directed to modifiable risk factors to reduce their negative effect on
DKA risk (addressing). This can be done by diabetes self-management education,
medical and psychological interventions, and telemedical and digital
approaches. Furthermore, preventive strategies can directly address DKA risk.
Tailoring: non-modifiable risk factors can be used to tailor preventive strategies
to achieve interventions that are age, sex, socially, and culturally appropriate.
DKA=diabetic ketoacidosis.
education, which has been effective in lowering HbA1c,91,92
enhancing treatment adherence,90,93 improving mental
health,90,94,95 and enhancing abilities to handle techno­
logical devices such as insulin pumps or properly monitor
glucose continuously.96,97 Diabetes self-management
education is also a powerful tool to transfer skills and
knowledge to prevent DKA, and is crucial for the proper
management of unexplained hyperglycaemia and ketosis
or ketonuria at home (sick day management).
In our literature search on the preventive effect of
diabetes self-management education on DKA risk, we
identified 25 studies, of which 11 reported either the
number of people who had DKA or diabetic ketosis, or
the event rate of DKA (table 1). We first looked at data
from seven studies that provided the number of people
with at least one event of DKA or ketosis.62,90,93,95,96,101,102,104
As seen in table 2, participation in a self-management
education programme led to a substantial reduction of
people with any ketosis-associated event (DKA or ketosis)
in four of the seven studies, with ORs ranging from
0·16 to 0·44. This effect of self-management education
can be seen for DKA alone and in one study on diabetic
ketosis alone, but seems to be even more pronounced for
DKA. In a second step, we looked at six studies that
provided the number of DKA events.62,93,98,99,100,103
Participation in a structured diabetes education
programme led to a substantial risk reduction for DKA in
five of the six studies, with incidence rate ratios ranging
from 0·13 to 0·47 (table 3), further emphasising the
effectiveness of diabetes education in reducing DKA risk.
439
 Review

Mean age, years Follow-up Study design Education Conduct of DKA-related outcome DKA-related effect
(SD) period programme education
Muehlhauser 26 (± 10) 12 months Randomised DTTP Inpatient setting; Hospitalisation for DKA from Reduction of the total number of DKA
et al (1987)93 controlled trial 5 days patient records events from 16 to 2; reduction of the
number of participants with at least one
DKA event from 13 to 2
Tankova 34·4 (± 11·2) 12 months Pre-post DTTP Inpatient setting; Hospitalisation for DKA from Reduction of DKA incidence from
et al (2001)98 5 days patient records and interview 0·30 to 0·14 per patient-year
Lemozy-Cadroy 34·6 (± 11·7) 12 months Pre-post Not specified Inpatient setting; Hospitalisation for DKA from Reduction of DKA incidence from
et al (2002)99 5 days interview 0·13 to 0·02 per patient-year
Samann 38 (± 14) 12 months Pre-post DTTP Inpatient setting; Hospitalisation for DKA from Reduction of DKA incidence in those
et al (2006)100 5 days interview with two or more events before
education from 3·3 to 0·6 per
patient-year; reduction of DKA incidence
in those with one or no DKA event
before education from 0·06 to 0·03 per
patient-year
Hermanns 45·1 (± 13·5) 6 months Randomised PRIMAS Outpatient setting; Diabetic ketosis from patient No effect on the number of participants
et al (2013)95 controlled trial 12 lessons (6 weeks) records and case report forms with at least one diabetic ketosis event
Elliott 41·0 (± 13·6) 12 months Pre-post DAFNE Outpatient setting Physician-diagnosed episode RR before vs after DAFNE: 0·39 (95% CI
et al (2014)62 (adaptation of requiring admission to hospital 0·23–0·65); reduction of DKA incidence
the DTTP) from medical records from 0·07 to 0·03 per patient-year;
reduction in the number of participants
with at least one DKA event from
41 to 18; reduction in the number of
participants with multiple DKA events
from 11 to 3
Speight 47 (± 14) 12 months Pre-post Australian 5 days Hospitalisation for DKA from RR reduction after participation of 0·70
et al (2016)101 DAFNE OzDAFNE database (99% CI 0·02–0·91); reduction in the
number of participants with at least one
DKA event from 20 to 6
Ehrmann 43·6 (± 13·6) 6 months Pre-post PRIMAS Outpatient setting; Diabetic ketosis from patient Reduction in the number of participants
et al (2016)102 12 lessons records and case report forms with at least one diabetic ketosis event
(3–12 weeks) from 11 to 5
Ehrmann 42·8 (± 14·2) 6 months Randomised INPUT Outpatient setting; Diabetic ketosis from patient Reduction in the number of participants
et al (2018)90 controlled trial 12 lessons records and case report forms with at least one diabetic ketosis event
(6–12 weeks) from 13 to 9
Humayun 41·3 (± 14·9) 12 months Pre-post Bournemouth Outpatient setting; DKA from patient questionnaires Non-significant reduction of DKA
et al (2018)103 Type 1 Intensive 4 weeks incidence from 0·06 to 0·03 per
Education patient-year
(adaptation of
the DTTP)
Bergis 47·2 (± 14·1) 6 months Pre-post INPUT Outpatient setting; Diabetic ketosis from patient Reduction in the number of participants
et al (2013)96 12 lessons records and case report forms with at least one diabetic ketosis event
(3–12 weeks) from 21 to 8
DAFNE=Dose Adjustment For Normal Eating. DKA=diabetic ketoacidosis. DTTP=Diabetes Treatment and Teaching Programme. INPUT=Insulin Pump Therapy programme. PRIMAS=Programme for diabetes
education and treatment for a self-determined living with type 1 diabetes. RR=relative risk.

Table 1: Overview of studies including data on reduction of DKA or diabetic ketosis by structured diabetes education

However, it should be kept in mind that only a small
amount of the evidence comes from randomised
controlled trials (RCTs) compared with pre-post trials
(table 1).
When diabetes education seems insufficient, more
intensified, multidisciplinary approaches such as
psychological interventions and individual coaching,
behavioural family systems therapy, and patient
management can be used to address DKA risk factors or
DKA directly.32,105–108 Of course, medical interventions
aiming at precipitating causes and somatic risk factors of
DKA (eg, infections, optimisation of insulin therapy) or
psychiatric comorbidities, are needed to prevent DKA.1
There is observational evidence that continuous glucose
monitoring can reduce DKA rates in a paediatric sample.109
Telemedical and digital approaches can provide
psychosocial support and contribute to the prevention
of acute complications such as DKA.110 Particularly
relevant is the provision of a 24-h emergency call service
that offers medical advice when symptoms of DKA are
present or blood glucose or ketone concentrations are
high. There is some evidence that approaches that
included such a service are effective in preventing
DKA.108
For the prevention of DKA, interventions that address
the modifiable risk factors for DKA or target management

440 www.thelancet.com/diabetes-endocrinology Vol 8 May 2020

 Review

of ketosis or DKA directly are needed. Such preventive

Education Control OR (95% CI) DKA or diabetic
strategies should be tailored to the non-modifiable risk ketosis
factors to match the needs of these risk groups to increase
n N n N
their effectiveness in preventing DKA. This approach
also includes specific interventions, such as peer support Muehlhauser et al 2 89 13 93 0·16 (0·04–0·71) DKA
(1987)93
and community-based approaches that actively seek
Elliott et al (2014)62 18 864 41 864 0·44 (0·25–0·76) DKA
populations at high risk, with little access to health care.110
Speight et al (2016)99 6 506 20 506 0·30 (0·12–0·75) DKA
Further research, also coming from RCTs, is clearly
Hermanns et al (2013)95 4 75 4 75 1·00 (0·25–4·00) Diabetic ketosis
needed to further investigate these preventive strategies
Ehrmann et al (2016)100 5 212 11 212 0·45 (0·16–1·31) Diabetic ketosis
in the identified risk groups.
Ehrmann et al (2018)90 9 133 13 129 0·67 (0·29–1·57) Diabetic ketosis

Therapy with SGLT inhibitors Bergis et al (2019)96 8 168 21 168 0·38 (0·17–0·86) Diabetic ketosis

SGLT inhibitors (SGLT2 and dual SGLT1 and SGLT2 DKA=diabetic ketoacidosis. n=number of people with at least one event of DKA or diabetic ketosis. N=total number of
people.
inhibitors) have become an adjunct therapy option for
people with type 1 diabetes in Europe and in Japan. Table 2: Effect of diabetes education on the number of people with DKA and diabetic ketosis events
In addition, SGLT inhibitors are used off-label.30,31 These
inhibitors have been shown to have positive effects on
cardiovascular endpoints, mortality, and renal disease
Education Control Incidence rate
in people with type 2 diabetes,111,112 and reductions in ratio (95% CI)
HbA1c without increasing the risk of hypoglycaemia and
Events Person-months Events Person-months
weight loss between 2 kg and 4 kg have been
Muehlhauser et al (1987)93 2 1068 16 1116 0·13 (0·03–0·57)
observed.113,114 Since many people with type 1 diabetes
are affected by metabolic syndrome115 and do not meet Tankova et al (2001)101 28 2412 60 2412 0·47 (0·30–0·73)

glycaemic targets set by guidelines,39 the addition of Lemozy-Cadroy et al (2002)103 2 840 10 912 0·22 (0·05–0·99)

SGLT inhibitors in people with type 1 diabetes seems
promising.28 Currently, there are eight published RCTs
in people with type 1 diabetes showing, depending on
the class and dose of SGLT inhibitors, HbA1c reductions
between 0·25 percentage points and 0·54 percentage
points, and an average weight reduction of 3·5 kg.26
However, in these RCTs, a quadrupled risk of DKA with
SGLT inhibitor therapy (4·1 vs 1·0 events per 100 person-
years) was found;25,26 but this finding differs for lower
versus higher doses (appendix). This result translated
to an overall number needed to harm of approximately
32 people given SGLT inhibitors (appendix). However,
there are substantial differences between the four
investigated substances (canagliflozin, dapagliflozin,
sotagliflozin, and empagliflozin), indicated by the wide
range of the number needed to harm, from 6 to
69 people with type 1 diabetes, depending on the
specific substance and dose (appendix). In addition,
comparing DKA rates across these RCTs is difficult,
because a different definition of DKA was used in all
the trials, as outlined by Dhatariya2 and Taylor and
colleagues.26 Finally, it should be noted that as a general
guideline, SGLT inhibitor therapy should not be given
to patients who consume excessive alcohol.28
Notably, in the placebo (1·0 events per 100 person-
years) and SGLT inhibitor group (4·1 events per
100 person-years), the DKA rate was lower than the
incidence seen in real-world registry data (5·1 events per
100 person-years).11 This finding might be because of a
selection bias usually present in all RCTs and effective
risk mitigation strategies initiated by the study protocols
(eg, education about DKA risk, provision of free ketone
meters and ketone strips, training of study centres) in
Samann et al (2006)102 340 114 996 878 114 996 0·39 (0·34–0·44)
Elliott et al (2014)62 22 10 368 57 10 368 0·39 (0·24–0·63)
Humayun et al (2018)104 4 1704 9 1704 0·44 (0·14–1·44)
DKA=diabetic ketoacidosis.
Table 3: Effect of diabetes education on the incidence of DKA events
both groups. However, in real-world settings, these
effects are more likely to be absent or less strictly
enforced. Thus, DKA rates in real-world settings could be See Online for appendix
greater than observed in the RCTs.
Identification of DKA risk factors in patients who use
SGLT inhibitors
The key risk factor for DKA in patients who use SGLT
inhibitors is the loss of marked hyperglycaemia as a
signal for developing ketosis or DKA,28,116 which can result
in a delayed recognition of ketosis or DKA. In the RCTs,
some additional risk factors for DKA in patients who use
SGLT inhibitors emerged (figure 2). Among them are
large reductions in insulin doses (especially basal
insulin), diets with reduced carbohydrate intake,
dehydration, and low BMI (<25 kg/m²).25,28,117,118 Notably,
DKA rates were significantly higher in patients using an
insulin pump with SGLT inhibitors,28,117 whereas in
general, no elevated DKA rates have been observed in
patients who use an insulin pump without SGLTs.37,39,70,119
An interruption of insulin supply might sooner lead to an
insulin deficiency in patients who use a pump with SGLT
inhibitors. Illness and infections were also associated
with a more rapid development of DKA in patients who
used SGLT inhibitors.28 For a further overview of SGLT
inhibitor-specific risk factors for DKA, we refer to the

www.thelancet.com/diabetes-endocrinology Vol 8 May 2020 441

 Review

Given the metabolic benefits of SGLT inhibitors on the

Preventive strategy 1: eligibility criteria one hand and the markedly enhanced DKA risk on the
for patients who could potentially use
SGLT inhibitors other, Taylor and colleagues26 and Wolfsdorf and Ratner116
• 18 years of age suggest an evaluation of the individual risk-to-benefit
• Reasonable diabetes self-management
• No drug or alcohol abuse
ratio before the use of SGLT inhibitors in the treatment
• Access to high-quality care of type 1 diabetes.
Informing • No acute illness Informing
eligibility eligibility
In general, it seems plausible that the abovementioned
• No pregnancy
non-modifiable and modifiable risk factors depicted in
figure 1 can also be applied to establish the DKA risk in
General risk factors SGLT inhibitor-specific patients who use SGLT inhibitors.117,118,120 Together with
• Low socioeconomic risk factors
status • Large reductions in
SGLT inhibitor-specific risk factors, these general risk
• Adolescent age insulin doses factors could also inform the eligibility of patients who
• Female sex SGLT • Diets with reduced could use SGLT inhibitors and help to weigh the benefits
• Ethnicity or migration inhibitor- carbohydrate intake
• Previous DKA related • Dehydration against the risks.18 Accordingly, SGLT inhibitor use
• Elevated HbA1c DKA risk • Low BMI (<25 kg/m²) should be evaluated with particular care in the adolescent
• Non-adherence • Insulin pump use
• Alcohol or drug abuse
age group and in those with high HbA1c, a previous
• Quality of care history of DKA, problems with adherence, insulin pump
• Poor mental health use, and poor mental health. Identification of DKA risk
• Somatic comorbidities
factors can thus also contribute to reducing the DKA risk
associated with SGLT inhibitors use (figure 2).
Preventive strategy 2: reduction of
Addressing Addressing
risk factors
• Ketone monitoring
Reduction of DKA risk in patients who use SGLT
• Diabetes self-management education inhibitors
• Education of health-care professionals It should be emphasised that SGLT inhibitor use in the
• Risk communication
• Withholding SGLT inhibitors in RCTs summarised by Taylor and colleagues26 was
high-risk situations (eg, surgery) accompanied by an elevated DKA risk compared with the
control group despite intensive education and risk
Risk enhancing Risk reducing
mitigation protocols. Such protocols are likely to be less
Figure 2: Conceptual model of preventive strategies for SGLT inhibitor- rigorously used in the real world. Thus, further preventive
related DKA risk strategies are needed. In addition to the use of risk factors
Preventive strategy I: the general risk factors and the SGLT inhibitor-specific risk
factors for DKA are applied to identify the eligibility of patients who could
to inform eligibility for SGLT inhibitor use, a second
potentially use SGLT inhibitors. These factors can be used to identify people with preventive strategy is to address these risk factors. We
an increased DKA risk during SGLT inhibitor therapy (green arrows). Preventive provide a conceptual model for these preventive strategies
strategy II: preventive actions should address general factors and SGLT in figure 2. The most important measure is to raise
inhibitor-specific risk factors to reduce their negative effect on DKA risk.
DKA=diabetic ketoacidosis.
awareness that with SGLT inhibitors, DKA can also occur
at normal or moderately elevated glucose concentrations
and that non-specific bodily symptoms (eg, lethargy,
Search strategy and selection criteria nausea, vomiting, abdominal pain, thirst) should receive
We searched PubMed using the following terms: “type 1 more attention.25,28 Patients who use SGLT inhibitors
diabetes”, “ketosis”, “ketoacidosis”, “risk factor”, and “risk should also be equipped with ketone strips and require
factors” in conjunction with “SGLT” and “sodium glucose knowledge and skills regarding how to respond to a
cotransporter” (both SGLT2 and SGLT1 combined, and SGLT2 positive ketone test. Diabetes self-management education
inhibitors were included). To review the role of diabetes can be a promising strategy to provide these skills and
education, we used the terms “self-management education” knowledge and to address modifiable risk factors for
and “structured education” in combination with “type 1 DKA. Consequently, diabetes education is heavily
diabetes”, “ketosis”, and “ketoacidosis”. No criteria on advertised and endorsed in current risk mitigation
publication data were set, and all articles in English or German protocols.25,28 Education should also be provided to health-
were included if published before Jan 27, 2020. We also care professionals, and general risk communication
checked reference lists in relevant articles and Google Scholar strategies should be used.
for additional references. We excluded studies that As the quality and experience of care centres was
investigated risk factors for DKA at the onset of diabetes or associated with general DKA risk, it seems sensible to
DKA as a risk factor for other complications. suggest that only centres that could provide the necessary
education and are experienced in treating people with
type 1 diabetes should prescribe SGLT inhibitors.116
consensus statement by Danne and colleagues,28 which Withholding SGLT inhibitors in high-risk situations such
provides guidance on the risk manage­ment of DKA in as infection, surgery, trauma, dehydration, excessive
patients who use SGLT inhibitors. alcohol intake, and low carbohydrate diets is also an

442 www.thelancet.com/diabetes-endocrinology Vol 8 May 2020


important prevention strategy for DKA in patients who
use SGLT inhibitors.25,28 However, these strategies are
based solely on consensus statements,25,28 whereas
empirical evidence from methodologically sound studies
about the preventive effects of these strategies,
particularly of structured diabetes education in patients
who use SGLT inhibitors, is scarce.
Conclusion
DKA remains a pronounced problem for people with
type 1 diabetes and has become an even more relevant
concern with new advancements in non-insulin adjunct
therapy options such as SGLT inhibitor use. Therefore,
effective prevention strategies, such as identifying people
at risk and the use of efficacious intervention to reduce
the risk of DKA, are needed. Further research from
prospective register studies might improve under­
standing of the relative importance of risk factors and
facilitate a quantification of DKA risk. A second preventive
strategy is structured diabetes self-manage­ment education,
coaching, and the treatment of comorbidities. These
interventions should be sensitive to age, sex, and cultural
and socioeconomic background. Evidence from RCTs
regarding the effectiveness of existing and new DKA-
prevention strategies is needed. These prevention
strategies also appear to be relevant for people with type 1
diabetes on SGLT inhibitor therapy to mitigate DKA risk.
Contributors
DE did the literature search, evaluated and summarised the studies, and
wrote the manuscript. BK was involved with summarising the studies
and revised the manuscript for content. TR evaluated and summarised
the studies and revised the manuscript for content. TH revised the
manuscript for content. MA-K revised the manuscript for content.
NH was involved with summarising the studies and wrote the
manuscript. No medical writer was involved in this review.
Declaration of interests
DE reports grants from Berlin-Chemie, Dexcom, Roche Diabetes Care,
Abbott Diabetes Care, AstraZeneca, and Ypsomed; personal fees from
Berlin-Chemie, Dexcom, Roche Diabetes Care, Abbott Diabetes Care,
and Medtronic; and non-financial support from Berlin-Chemie.
BK reports grants from AstraZeneca, Berlin-Chemie, Roche Diabetes
Care, Abbott Diabetes Care, AstraZeneca, Dexcom, and Ypsomed; and
personal fees from Berlin-Chemie, Roche Diabetes Care, Novo Nordisk,
Medtronic, Ascensia Diabetes Care, Abbott Diabetes Care, and Eli Lilly.
TR reports grants, personal fees, and non-financial support from Berlin-
Chemie. TH reports grants from AstraZeneca, Abbott Diabetes Care,
and Boehringer Ingelheim; and personal fees from MSD, AstraZeneca,
Roche Diabetes Care, Abbott Diabetes Care, Novo Nordisk, Eli Lilly, and
Berlin-Chemie. NH reports grants from AstraZeneca, Berlin-Chemie,
Roche Diabetes Care, Abbott Diabetes Care, AstraZeneca, Ypsomed, and
Dexcom; personal fees from Berlin-Chemie, Roche Diabetes Care, Novo
Nordisk, Abbott Diabetes Care, Eli Lilly, Ypsomed, and Dexcom; and
non-financial support from Berlin-Chemie. NH and BK are supported by
an educational grant from AstraZeneca for an investigator-initiated
project about diabetic ketoacidosis; however, AstraZeneca had no
involvement in the present Review. MA-K declares no competing
interests.
References
1 Umpierrez G, Korytkowski M. Diabetic emergencies—ketoacidosis,
hyperglycaemic hyperosmolar state and hypoglycaemia.
Nat Rev Endocrinol 2016; 12: 222–32.
2 Dhatariya KK. Defining and characterising diabetic ketoacidosis in
adults. Diabetes Res Clin Pract 2019; 155: 107797.
www.thelancet.com/diabetes-endocrinology Vol 8 May 2020
Review
3 Dhatariya KK, Umpierrez GE. Guidelines for management of
diabetic ketoacidosis: time to revise? Lancet Diabetes Endocrinol 2017;
5: 321–23.
4 Wolfsdorf JI, Glaser N, Agus M, et al. ISPAD clinical practice
consensus guidelines 2018: diabetic ketoacidosis and the
hyperglycemic hyperosmolar state. Pediatr Diabetes 2018;
19 (suppl 27): 155–77.
5 Neu A, Hofer SE, Karges B, Oeverink R, Rosenbauer J, Holl RW.
Ketoacidosis at diabetes onset is still frequent in children and
adolescents: a multicenter analysis of 14,664 patients from 106
institutions. Diabetes Care 2009; 32: 1647–48.
6 Dhatariya KK, Nunney I, Higgins K, Sampson MJ, Iceton G.
National survey of the management of diabetic ketoacidosis (DKA)
in the UK in 2014. Diabet Med 2016; 33: 252–60.
7 Schober E, Rami B, Waldhoer T. Diabetic ketoacidosis at diagnosis
in Austrian children in 1989–2008: a population-based analysis.
Diabetologia 2010; 53: 1057–61.
8 Dabelea D, Rewers A, Stafford JM, et al. Trends in the prevalence of
ketoacidosis at diabetes diagnosis: the SEARCH for diabetes in
youth study. Pediatrics 2014; 133: e938–45.
9 Jefferies C, Cutfield SW, Derraik JGB, et al. 15-year incidence of
diabetic ketoacidosis at onset of type 1 diabetes in children from
a regional setting (Auckland, New Zealand). Sci Rep 2015; 5: 10358.
10 Limenis E, Shulman R, Daneman D. Is the frequency of
ketoacidosis at onset of type 1 diabetes a child health indicator that
is related to income inequality? Diabetes Care 2012; 35: e5.
11 Farsani SF, Brodovicz K, Soleymanlou N, Marquard J, Wissinger E,
Maiese BA. Incidence and prevalence of diabetic ketoacidosis (DKA)
among adults with type 1 diabetes mellitus (T1D): a systematic
literature review. BMJ Open 2017; 7: e016587.
12 Große J, Hornstein H, Manuwald U, Kugler J, Glauche I, Rothe U.
Incidence of diabetic ketoacidosis of new-onset type 1 diabetes in
children and adolescents in different countries correlates with human
development index (HDI): an updated systematic review, meta-
analysis, and meta-regression. Horm Metab Res 2018; 50: 209–22.
13 Benoit SR, Zhang Y, Geiss LS, Gregg EW, Albright A. Trends in
diabetic ketoacidosis hospitalizations and in-hospital mortality­–
United States, 2000–2014. MMWR Morb Mortal Wkly Rep 2018;
67: 362–65.
14 Karges B, Rosenbauer J, Holterhus PM, et al. Hospital admission
for diabetic ketoacidosis or severe hypoglycemia in 31,330 young
patients with type 1 diabetes. Eur J Endocrinol 2015; 173: 341–50.
15 Icks A, Strassburger K, Baechle C, et al. Frequency and cost of
diabetic ketoacidosis in Germany--study in 12,001 paediatric
patients. Exp Clin Endocrinol Diabetes 2013; 121: 58–59.
16 Dhatariya KK, Skedgel C, Fordham R. The cost of treating diabetic
ketoacidosis in the UK: a national survey of hospital resource use.
Diabet Med 2017; 34: 1361–66.
17 Desai D, Mehta D, Mathias P, Menon G, Schubart UK. Health
care utilization and burden of diabetic ketoacidosis in the U.S.
over the past decade: a nationwide analysis. Diabetes Care 2018;
41: 1631–38.
18 Fernando SM, Bagshaw SM, Rochwerg B, et al. Comparison of
outcomes and costs between adult diabetic ketoacidosis patients
admitted to the ICU and step-down unit. J Crit Care 2019;
50: 257–61.
19 Snorgaard O, Eskildsen P, Vadstrup S, Nerup J. Diabetic
ketoacidosis in Denmark: epidemiology, incidence rates,
precipitating factors and mortality rates. J Intern Med 1989;
226: 223–28.
20 Livingstone SJ, Levin D, Looker HC, et al. Estimated life expectancy
in a Scottish cohort with type 1 diabetes, 2008–2010. JAMA 2015;
313: 37–44.
21 Gibb FW, Teoh WL, Graham J, Lockman KA. Risk of death
following admission to a UK hospital with diabetic ketoacidosis.
Diabetologia 2016; 59: 2082–87.
22 Macfarlane J, Dhatariya K. Incidence of euglycemic diabetic
ketoacidosis in adults with type 1 diabetes in the United Kingdom
before the widespread use of sodium glucose cotransporter 2
inhibitors. Mayo Clin Proc 2019; 94: 1909–10.
23 Mathieu C, Dandona P, Gillard P, et al. Efficacy and safety of
dapagliflozin in patients with inadequately controlled type 1
diabetes (the DEPICT-2 study): 24-week results from a randomized
controlled trial. Diabetes Care 2018; 41: 1938–46.
443
 Review
24 Garg SK, Henry RR, Banks P, et al. Effects of sotagliflozin added to
insulin in patients with type 1 diabetes. N Engl J Med 2017;
377: 2337–48.
25 Goldenberg RM, Gilbert JD, Hramiak IM, Woo VC, Zinman B.
Sodium-glucose co-transporter inhibitors, their role in type 1
diabetes treatment and a risk mitigation strategy for preventing
diabetic ketoacidosis: the STOP DKA protocol. Diabetes Obes Metab
2019; 21: 2192–202.
26 Taylor SI, Blau JE, Rother KI, Beitelshees AL. SGLT2 inhibitors as
adjunctive therapy for type 1 diabetes: balancing benefits and risks.
Lancet Diabetes Endocrinol 2019; 7: 949–58.
27 Handelsman Y, Henry RR, Bloomgarden ZT, et al. American
Association of Clinical Endocrinologists and American College of
Endocrinology position statement on the association of SGLT-2
inhibitors and diabetic ketoacidosis. Endocr Pract 2016; 22: 753–62.
28 Danne T, Garg S, Peters AL, et al. International consensus on risk
management of diabetic ketoacidosis in patients with type 1
diabetes treated with sodium–glucose cotransporter (SGLT)
inhibitors. Diabetes Care 2019; 42: 1147–54.
29 Garg SK, Peters AL, Buse JB, Danne T. Strategy for mitigating DKA
risk in patients with type 1 diabetes on adjunctive treatment with
SGLT inhibitors: a STICH protocol. Diabetes Technol Ther 2018;
20: 571–75.
30 Scavini M, Bertuzzi F, Girelli A, et al. Off-label prescription of
SGLT-2 inhibitors to patients with type 1 diabetes: an online survey
of diabetes specialists in Italy. Acta Diabetol 2019; 56: 1235–38.
31 Fadini GP, Bonora BM, Avogaro A. SGLT2 inhibitors and diabetic
ketoacidosis: data from the FDA adverse event reporting system.
Diabetologia 2017; 60: 1385–89.
32 Garrett CJ, Choudhary P, Amiel SA, Fonagy P, Ismail K. Recurrent
diabetic ketoacidosis and a brief history of brittle diabetes research:
contemporary and past evidence in diabetic ketoacidosis research
including mortality, mental health and prevention. Diabet Med 2019;
36: 1329–35.
33 Li J, Yang D, Yan J, et al. Secondary diabetic ketoacidosis and severe
hypoglycaemia in patients with established type 1 diabetes mellitus
in China: a multicentre registration study. Diabetes Metab Res Rev
2014; 30: 497–504.
34 Yan JW, Hamelin AL, Gushulak KM, Van Aarsen K, Columbus MP,
Stiell IG. Hyperglycemia in young adults with types 1 and 2
diabetes seen in the emergency department: a health records
review. Can J Diabetes 2018; 42: 296–301.e5.
35 Everett E, Mathioudakis NN. Association of socioeconomic status
and DKA readmission in adults with type 1 diabetes: analysis of the
US National Readmission Database. BMJ Open Diabetes Res Care
2019; 7: e000621.
36 Govan L, Maietti E, Torsney B, et al. The effect of deprivation and
HbA1c on admission to hospital for diabetic ketoacidosis in type 1
diabetes. Diabetologia 2012; 55: 2356–60.
37 Kalscheuer H, Seufert J, Lanzinger S, et al. Event rates and risk
factors for the development of diabetic ketoacidosis in adult patients
with type 1 diabetes: analysis from the DPV registry based on
46,966 patients. Diabetes Care 2019; 42: e34–36.
38 Weinstock RS, Xing D, Maahs DM, et al. Severe hypoglycemia and
diabetic ketoacidosis in adults with type 1 diabetes: results from the
T1D exchange clinic registry. J Clin Endocrinol Metab 2013;
98: 3411–19.
39 Foster NC, Beck RW, Miller KM, et al. State of type 1 diabetes
management and outcomes from the T1D exchange in 2016–2018.
Diabetes Technol Ther 2019; 21: 66–72.
40 Fritsch M, Rosenbauer J, Schober E, et al. Predictors of diabetic
ketoacidosis in children and adolescents with type 1 diabetes.
Experience from a large multicentre database. Pediatr Diabetes 2011;
12: 307–12.
41 Ampt A, van Gemert T, Craig ME, Donaghue KC, Lain SB,
Nassar N. Using population data to understand the epidemiology
and risk factors for diabetic ketoacidosis in Australian children with
type 1 diabetes. Pediatr Diabetes 2019; 20: 901–08.
42 Everett E, Mathioudakis N. Association of area deprivation and
diabetic ketoacidosis readmissions: comparative risk analysis of
adults vs children with type 1 diabetes. J Clin Endocrinol Metab 2019;
104: 3473–80.
43 Skinner TC. Recurrent diabetic ketoacidosis: causes, prevention
and management. Horm Res Paediatr 2002; 57 (suppl 1): 78–80.
444
44 Lindner LME, Rathmann W, Rosenbauer J. Inequalities in glycaemic
control, hypoglycaemia and diabetic ketoacidosis according to
socio-economic status and area-level deprivation in type 1 diabetes
mellitus: a systematic review. Diabet Med 2018; 35: 12–32.
45 Lindner LME, Gontscharuk V, Bachle C, et al. Severe hypoglycemia
and diabetic ketoacidosis in young persons with preschool onset of
type 1 diabetes mellitus: an analysis of three nationwide population-
based surveys. Pediatr Diabetes 2018; 19: 713–20.
46 Randall L, Begovic J, Hudson M, et al. Recurrent diabetic ketoacidosis
in inner-city minority patients: behavioral, socioeconomic, and
psychosocial factors. Diabetes Care 2011; 34: 1891–96.
47 Sayed MH, Hegazi MA, Abdulwahed K, et al. Risk factors and
predictors of uncontrolled hyperglycemia and diabetic ketoacidosis
in children and adolescents with type 1 diabetes mellitus in Jeddah,
western Saudi Arabia. J Diabetes 2017; 9: 190–99.
48 Keenan HT, Foster CM, Bratton SL. Social factors associated with
prolonged hospitalization among diabetic children. Pediatrics 2002;
109: 40–44.
49 Rewers A, Chase HP, Mackenzie T, et al. Predictors of acute
complications in children with type 1 diabetes. JAMA 2002;
287: 2511–18.
50 Semenkovich K, Berlin KS, Ankney RL, et al. Predictors of diabetic
ketoacidosis hospitalizations and hemoglobin A1c among youth with
type 1 diabetes. Health Psychol 2019; 38: 577–85.
51 Lievre M, Marre M, Robert JJ, et al. Cross-sectional study of care,
socio-economic status and complications in young French patients
with type 1 diabetes mellitus. Diabetes Metab 2005; 31: 41–46.
52 Otieno CF, Kayima JK, Omonge EO, Oyoo GO. Diabetic ketoacidosis:
risk factors, mechanisms and management strategies in sub-Saharan
Africa: a review. East Afr Med J 2005; 82 (suppl 12): 197–203.
53 Shulman R, Luo J, Shah BR. Mental health visits and low
socio-economic status in adolescence are associated with
complications of type 1 diabetes in early adulthood: a population-
based cohort study. Diabet Med 2018; 35: 920–28.
54 Maahs DM, Hermann JM, Holman N, et al. Rates of diabetic
ketoacidosis: international comparison with 49,859 pediatric
patients with type 1 diabetes from England, Wales, the U.S.,
Austria, and Germany. Diabetes Care 2015; 38: 1876–82.
55 Wright J, Ruck K, Rabbitts R, et al. Diabetic ketoacidosis (DKA)
in Birmingham, UK, 2000–2009: an evaluation of risk factors for
recurrence and mortality. Br J Diabetes Vasc Dis 2009; 9: 278–82.
56 Snoek FJ, Bremmer MA, Hermanns N. Constructs of depression
and distress in diabetes: time for an appraisal.
Lancet Diabetes Endocrinol 2015; 3: 450–60.
57 Young-Hyman D, de Groot M, Hill-Briggs F, Gonzalez JS, Hood K,
Peyrot M. Psychosocial care for people with diabetes: a position
statement of the American Diabetes Association. Diabetes Care
2016; 39: 2126–40.
58 Barski L, Eshkoli T, Brandstaetter E, Jotkowitz A. Euglycemic
diabetic ketoacidosis. Eur J Intern Med 2019; 63: 9–14.
59 Rosenbloom AL. The management of diabetic ketoacidosis in
children. Diabetes Ther 2010; 1: 103–20.
60 Thomas M, Harjutsalo V, Feodoroff M, Forsblom C, Gordin D,
Groop PH. The long-term incidence of hospitalization for
ketoacidosis in adults with established T1D-a prospective cohort
study. J Clin Endocrinol Metab 2020; 105: 231–241.
61 Hurtado CR, Lemor A, Vallejo F, et al. Causes and predictors for
30-day re-admissions in adult patients with diabetic ketoacidosis in
the United States: a nationwide analysis, 2010–2014. Endocr Pract
2019; 25: 242–53.
62 Elliott J, Jacques RM, Kruger J, et al. Substantial reductions in the
number of diabetic ketoacidosis and severe hypoglycaemia episodes
requiring emergency treatment lead to reduced costs after structured
education in adults with type 1 diabetes. Diabet Med 2014; 31: 847–53.
63 Cooper H, Tekiteki A, Khanolkar M, Braatvedt G. Risk factors for
recurrent admissions with diabetic ketoacidosis: a case-control
observational study. Diabet Med 2016; 33: 523–28.
64 Gupta A, Taylor F, O’Sullivan T, Simmons D. Characteristics of
young adults with multiple episodes of diabetic ketoacidosis.
Intern Med J 2019; 49: 911–14.
65 Al-Hayek AA, Robert AA, Braham RB, Turki AS, Al-Sabaan FS.
Frequency and associated risk factors of recurrent diabetic
ketoacidosis among Saudi adolescents with type 1 diabetes mellitus.
Saudi Med J 2015; 36: 216–20.
www.thelancet.com/diabetes-endocrinology Vol 8 May 2020

66 Butalia S, Johnson JA, Ghali WA, Rabi DM. Clinical and
socio-demographic factors associated with diabetic ketoacidosis
hospitalization in adults with type 1 diabetes. Diabet Med 2013;
30: 567–73.
67 Del Degan S, Dube F, Gagnon C, Boulet G. Risk factors for
recurrent diabetic ketoacidosis in adults with type 1 diabetes.
Can J Diabetes 2019; 43: 472–76.e1.
68 Nyenwe EA, Loganathan RS, Blum S, et al. Active use of cocaine:
an independent risk factor for recurrent diabetic ketoacidosis in a
city hospital. Endocr Pract 2007; 13: 22–29.
69 Wolfsdorf J, Glaser N, Sperling MA. Diabetic ketoacidosis in
infants, children, and adolescents: a consensus statement from the
American Diabetes Association. Diabetes Care 2006; 29: 1150–59.
70 Realsen J, Goettle H, Chase HP. Morbidity and mortality of diabetic
ketoacidosis with and without insulin pump care.
Diabetes Technol Ther 2012; 14: 1149–54.
71 Jefferies CA, Nakhla M, Derraik JGB, Gunn AJ, Daneman D,
Cutfield WS. Preventing diabetic ketoacidosis.
Pediatr Clin North Am 2015; 62: 857–71.
72 Racicka E, Brynska A. Eating disorders in children and adolescents
with type 1 and type 2 diabetes: prevalence, risk factors, warning
signs. Psychiatr Pol 2015; 49: 1017–24.
73 Trief PM, Xing D, Foster NC, et al. Depression in adults in the T1D
exchange clinic registry. Diabetes Care 2014; 37: 1563–72.
74 Goueslard K, Petit JM, Cottenet J, Chauvet-Gelinier JC, Jollant F,
Quantin C. Increased risk of rehospitalization for acute diabetes
complications and suicide attempts in patients with type 1 diabetes
and comorbid schizophrenia. Diabetes Care 2018; 41: 2316–21.
75 Galler A, Bollow E, Meusers M, et al. Comparison of glycemic and
metabolic control in youth with type 1 diabetes with and without
antipsychotic medication: analysis from the nationwide German/
Austrian Diabetes Survey (DPV). Diabetes Care 2015; 38: 1051–57.
76 Holt RIG. Association between antipsychotic medication use and
diabetes. Curr Diab Rep 2019; 19: 96.
77 Gonzalez JS, Peyrot M, McCarl LA, et al. Depression and diabetes
treatment nonadherence: a meta-analysis. Diabetes Care 2008;
31: 2398–403.
78 Fisher L, Glasgow RE, Strycker LA. The relationship between
diabetes distress and clinical depression with glycemic control
among patients with type 2 diabetes. Diabetes Care 2010;
33: 1034–36.
79 Lustman PJ, Anderson RJ, Freedland KE, De Groot M, Carney RM,
Clouse RE. Depression and poor glycemic control: a meta-analytic
review of the literature. Diabetes Care 2000; 23: 934–42.
80 Kitabchi AE, Umpierrez GE, Miles JM, Fisher JN. Hyperglycemic
crises in adult patients with diabetes. Diabetes Care 2009;
32: 1335–43.
81 Schober E, Otto KP, Dost A, Jorch N, Holl R. Association of epilepsy
and type 1 diabetes mellitus in children and adolescents: is there an
increased risk for diabetic ketoacidosis? J Pediatr 2012; 160: 662–66.e1.
82 Wright JJ, Salem JE, Johnson DB, et al. Increased reporting of
immune checkpoint inhibitor-associated diabetes. Diabetes Care
2018; 41: e150–51.
83 Beulens JW, Kruidhof JS, Grobbee DE, Chaturvedi N, Fuller JH,
Soedamah-Muthu SS. Alcohol consumption and risk of
microvascular complications in type 1 diabetes patients:
the EURODIAB Prospective Complications Study. Diabetologia
2008; 51: 1631–38.
84 Hermann JM, Meusers M, Bachran R, et al. Self-reported regular
alcohol consumption in adolescents and emerging adults with
type 1 diabetes: a neglected risk factor for diabetic ketoacidosis?
Multicenter analysis of 29 630 patients from the DPV registry.
Pediatr Diabetes 2017; 18: 817–23.
85 Kinney GL, Akturk HK, Taylor DD, Foster NC, Shah VN. Cannabis
use is associated with increased risk for diabetic ketoacidosis in
adults with type 1 diabetes: findings from the T1D exchange clinic
registry. Diabetes Care 2019; 43: 247–49.
86 Hoshina S, Andersen GS, Jorgensen ME, Ridderstrale M,
Vistisen D, Andersen HU. Treatment modality-dependent risk of
diabetic ketoacidosis in patients with type 1 diabetes: Danish adult
diabetes database study. Diabetes Technol Ther 2018; 20: 229–34.
87 Crossen SS, Wilson DM, Saynina O, Sanders LM. Outpatient care
preceding hospitalization for diabetic ketoacidosis. Pediatrics
2016; 137(6): e20153497.
www.thelancet.com/diabetes-endocrinology Vol 8 May 2020
Review
88 Mays JA, Jackson KL, Derby TA, et al. An evaluation of recurrent
diabetic ketoacidosis, fragmentation of care, and mortality across
Chicago, Illinois. Diabetes Care 2016; 39: 1671–76.
89 Egger M, Davey Smith G, Stettler C, Diem P. Risk of adverse effects
of intensified treatment in insulin-dependent diabetes mellitus:
a meta-analysis. Diabet Med 1997; 14: 919–28.
90 Ehrmann D, Kulzer B, Schipfer M, Lippmann-Grob B, Haak T,
Hermanns N. Efficacy of an education program for people with
diabetes and insulin pump treatment (INPUT): results from a
randomized controlled trial. Diabetes Care 2018; 41: 2453–62.
91 Chatterjee S, Davies MJ, Heller S, Speight J, Snoek FJ, Khunti K.
Diabetes structured self-management education programmes:
a narrative review and current innovations. Lancet Diabetes
Endocrinol 2018; 6: 130–42.
92 Pillay J, Armstrong MJ, Butalia S, et al. Behavioral programs for
type 1 diabetes mellitus: a systematic review and meta-analysis.
Ann Intern Med 2015; 163: 836–47.
93 Muehlhauser I, Bruckner I, Berger M, et al. Evaluation of an
intensified insulin treatment and teaching programme as routine
management of type 1 (insulin-dependent) diabetes. The Bucharest-
Dusseldorf Study. Diabetologia 1987; 30: 681–90.
94 DAFNE Study Group. Training in flexible, intensive insulin
management to enable dietary freedom in people with type 1
diabetes: dose adjustment for normal eating (DAFNE) randomised
controlled trial. BMJ 2002; 325: 746.
95 Hermanns N, Kulzer B, Ehrmann D, Bergis-Jurgan N, Haak T.
The effect of a diabetes education programme (PRIMAS) for people
with type 1 diabetes: results of a randomized trial.
Diabetes Res Clin Pract 2013; 102: 149–57.
96 Bergis D, Ehrmann D, Albrecht C, Haak T, Kulzer B, Hermanns N.
Comparison of the efficacy of an education program for people with
diabetes and insulin pump treatment (INPUT) in a randomized
controlled trial setting and the effectiveness in a routine care
setting: results of a comparative effectiveness study.
Patient Educ Couns 2019; 102: 1868–74.
97 Hermanns N, Ehrmann D, Schipfer M, Kroger J, Haak T, Kulzer B.
The impact of a structured education and treatment programme
(FLASH) for people with diabetes using a flash sensor-based
glucose monitoring system: results of a randomized controlled trial.
Diabetes Res Clin Pract 2019; 150: 111–21.
98 Tankova T, Dakovska G, Koev D. Education of diabetic patients-
-a one year experience. Patient Educ Couns 2001; 43: 139–45.
99 Lemozy-Cadroy S, Crognier S, Gourdy P, et al. Intensified
treatment of type 1 diabetes: prospective evaluation at one year
of a therapeutic patient education programme. Diabetes Metab
2002; 28: 287–94.
100 Samann A, Muehlhauser I, Bender R, Hunger-Dathe W, Kloos C,
Muller UA. Flexible intensive insulin therapy in adults with type 1
diabetes and high risk for severe hypoglycemia and diabetic
ketoacidosis. Diabetes Care 2006; 29: 2196–99.
101 Speight J, Holmes-Truscott E, Harvey DM, et al. Structured type 1
diabetes education delivered in routine care in Australia reduces
diabetes-related emergencies and severe diabetes-related distress:
the OzDAFNE program. Diabetes Res Clin Pract 2016; 112: 65–72.
102 Ehrmann D, Bergis-Jurgan N, Haak T, Kulzer B, Hermanns N.
Comparison of the efficacy of a diabetes education programme for
type 1 diabetes (PRIMAS) in a randomised controlled trial setting
and the effectiveness in a routine care setting: results of a
comparative effectiveness study. PLoS One 2016; 11: e0147581.
103 Humayun MA, Jenkins E, Knott J, et al. Intensive structured
education for type 1 diabetes management using BERTIE: long-
term follow-up to assess impact on glycaemic control and quality of
life indices. Diabetes Res Clin Pract 2018; 143: 275–81.
104 Kulzer B, Ehrmann D, Albrecht C, Haak T, Hermanns N. Incidence
and risk as well as protective factors regarding diabetic ketosis in
people with type 1 diabetes: a secondary analysis of the PRIMAS
and INPUT studies. Diabetologia 2019; 62 (suppl 1): S293.
105 Baumeister H, Hutter N, Bengel J. Psychological and
pharmacological interventions for depression in patients with
diabetes mellitus and depression. Cochrane Database Syst Rev 2012;
12: CD008381.
106 Wong-Rieger D, Rieger FP. Health coaching in diabetes:
empowering patients to self-manage. Can J Diabetes 2013;
37: 41–44.
445
 Review
107 Ellis D, Naar-King S, Templin T, et al. Multisystemic therapy for
adolescents with poorly controlled type 1 diabetes: reduced diabetic
ketoacidosis admissions and related costs over 24 months.
Diabetes Care 2008; 31: 1746–47.
108 Farrell K, Fernandez R, Salamonson Y, Griffiths R,
Holmes-Walker DJ. Health outcomes for youth with type 1 diabetes
at 18 months and 30 months post transition from pediatric to adult
care. Diabetes Res Clin Pract 2018; 139: 163–69.
109 Tauschmann M, Hermann JM, Freiberg C, et al. Reduction in
diabetic ketoacidosis and severe hypoglycemia in pediatric type 1
diabetes during the first year of continuous glucose monitoring:
a multicenter analysis of 3,553 subjects from the DPV registry.
Diabetes Care 2020; published online Jan 22. DOI:10.2337/dc19-1358.
110 Prahalad P, Tanenbaum M, Hood K, Maahs D. Diabetes technology:
improving care, improving patient-reported outcomes and
preventing complications in young people with type 1 diabetes.
Diabet Med 2018; 35: 419–29.
111 Zinman B, Wanner C, Lachin JM, et al. Empagliflozin,
cardiovascular outcomes, and mortality in type 2 diabetes.
N Engl J Med 2015; 373: 2117–28.
112 Wiviott SD, Raz I, Bonaca MP, et al. Dapagliflozin and
cardiovascular outcomes in type 2 diabetes. N Engl J Med 2019;
380: 347–57.
113 Inzucchi SE, Zinman B, Wanner C, et al. SGLT-2 inhibitors and
cardiovascular risk: proposed pathways and review of ongoing
outcome trials. Diab Vasc Dis Res 2015; 12: 90–100.
446
114 Shyangdan DS, Uthman OA, Waugh N. SGLT-2 receptor inhibitors
for treating patients with type 2 diabetes mellitus: a systematic
review and network meta-analysis. BMJ Open 2016; 6: e009417.
115 Nathan DM. The diabetes control and complications trial/
epidemiology of diabetes interventions and complications study at
30 years: overview. Diabetes Care 2014; 37: 9–16.
116 Wolfsdorf JI, Ratner RE. SGLT inhibitors for type 1 diabetes:
proceed with extreme caution. Diabetes Care 2019; 42: 991–93.
117 Boeder S, Edelman SV. Sodium-glucose co-transporter inhibitors as
adjunctive treatment to insulin in type 1 diabetes: a review of
randomized controlled trials. Diabetes Obes Metab 2019;
21 (suppl 2): 62–77.
118 Musso G, Gambino R, Cassader M, Paschetta E. Efficacy and safety
of dual SGLT 1/2 inhibitor sotagliflozin in type 1 diabetes: meta-
analysis of randomised controlled trials. BMJ 2019; 365: l1328.
119 Karges B, Schwandt A, Heidtmann B, et al. Association of insulin
pump therapy vs insulin injection therapy with severe
hypoglycemia, ketoacidosis, and glycemic control among children,
adolescents, and young adults with type 1 diabetes. JAMA 2017;
318: 1358–66.
120 Yang Y, Pan H, Wang B, Chen S, Zhu H. Efficacy and safety of
SGLT2 inhibitors in patients with type 1 diabetes: a meta-analysis of
randomized controlled trials. Chin Med Sci J 2017; 32: 22–27.
© 2020 Elsevier Ltd. All rights reserved.
www.thelancet.com/diabetes-endocrinology Vol 8 May 2020