Journal Pre-proof

Consensus Recommendations for Sick Day Medication Guidance for People With
Diabetes, Kidney, or Cardiovascular Disease: A Modified Delphi Process

Kaitlyn E. Watson, B. Pharm (Hons), PhD, Kirnvir Dhaliwal, RN, MN, PhD, Sandra
Robertshaw, Nancy Verdin, Eleanor Benterud, RN, MN, Nicole Lamont, MBT, BHSc,
Kelsea M. Drall, MSc, Kerry McBrien, MD, MPH, CCFP, Maoliosa Donald, PhD,
BScPT, Ross T. Tsuyuki, BSc(Pharm), PharmD, MSc, FCSHP, FACC, FCAHS,
David J.T. Campbell, MD, MSc, PhD, FRCPC, Neesh Pannu, MD, SM, Matthew
T. James, MD, PhD, FRCPC, On behalf of the PAUSE (Preventing Medication
Complications During Acute Illness Through Symptom Evaluation And Sick Day
Guidance) Medication Safety Advisory Panel
PII: S0272-6386(22)01054-X
DOI: https://doi.org/10.1053/j.ajkd.2022.10.012
Reference: YAJKD 57826

To appear in: American Journal of Kidney Diseases

Received Date: 1 June 2022

Accepted Date: 12 October 2022

Please cite this article as: Watson KE, Dhaliwal K, Robertshaw S, Verdin N, Benterud E, Lamont N, Drall
KM, McBrien K, Donald M, Tsuyuki RT, Campbell DJT, Pannu N, James MT, On behalf of the PAUSE
(Preventing Medication Complications During Acute Illness Through Symptom Evaluation And Sick Day
Guidance) Medication Safety Advisory Panel, Consensus Recommendations for Sick Day Medication
Guidance for People With Diabetes, Kidney, or Cardiovascular Disease: A Modified Delphi Process,
American Journal of Kidney Diseases (2023), doi: https://doi.org/10.1053/j.ajkd.2022.10.012.

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
© 2022 Published by Elsevier Inc. on behalf of the National Kidney Foundation, Inc.
Consensus Recommendations for Sick Day Medication Guidance for
People With Diabetes, Kidney, or Cardiovascular Disease: A
Modified Delphi Process
Kaitlyn E. Watson, B. Pharm (Hons), PhD, GradCertAppPharmPrac, FHEA
EPICORE Centre, Department of Medicine, University of Alberta, AB T6G 2V2, Canada

Kirnvir Dhaliwal RN, MN, PhD

Department of Medicine, Cumming School of Medicine, University of Calgary

Sandra Robertshaw

of
Patient partner, Department of Medicine, Cumming School of Medicine, University of Calgary
Nancy Verdin

ro
Patient partner, Department of Medicine, Cumming School of Medicine, University of Calgary
Eleanor Benterud RN, MN,

-p
Department of Medicine, Cumming School of Medicine, University of Calgary
Nicole Lamont MBT, BHSc
re
Department of Medicine, Cumming School of Medicine, University of Calgary
Kelsea M. Drall, MSc,
lP

Division of Nephrology, Department of Medicine, University of Alberta, Edmonton AB, Canada

Kerry McBrien, MD, MPH, CCFP
Departments of Family Medicine and Community Health Sciences, Cumming School of
na

Medicine, University of Calgary, Alberta Canada

Maoliosa Donald, PhD, BScPT
ur

Departments of Medicine and Community Health Sciences, Cumming School of Medicine,

University of Calgary, Alberta, Canada
Jo

Ross T. Tsuyuki, BSc(Pharm), PharmD, MSc, FCSHP, FACC, FCAHS, ISHF

EPICORE Centre, Department of Medicine, University of Alberta, AB T6G 2S2, Canada

David J.T. Campbell, MD, MSc, PhD, FRCPC

Departments of Medicine, Community Health Sciences and Cardiac Sciences, Cumming School
of Medicine, University of Calgary, Alberta, Canada

Neesh Pannu MD, SM

Department of Medicine, Faculty of Medicine and Dentistry, University of Alberta
Matthew T. James, MD, PhD, FRCPC
Departments of Medicine and Community Health Sciences, Cumming School of Medicine,
University of Calgary, Alberta Canada

On behalf of the PAUSE (Preventing Medication Complications During Acute Illness Through
Symptom Evaluation And Sick Day Guidance) Medication Safety Advisory Panel

1
Complete author and article information (including a list of the members of the PAUSE Medication

Safety Advisory Panel) provided before references.

Corresponding Author:

Matthew T. James, MD, PhD, FRCPC

Departments of Medicine and Community Health Sciences
Cumming School of Medicine, University of Calgary
3280 Hospital Drive NW, Calgary, AB
Canada, T2N 4Z6.
Email: mjames@ucalgary.ca

of
ro
-p
re
lP
na
ur
Jo

2
Abstract

Rationale and Objective: Sick day medication guidance (SDMG) involves withholding or

adjusting specific medications in the setting of acute illnesses that could contribute to

complications such as hypotension, acute kidney injury (AKI), or hypoglycemia. We sought to

achieve consensus among clinical experts on recommendations for SDMG that could be studied

in future intervention studies.

Study Design: A modified Delphi process following the Conducting and Reporting Delphi

of
Studies reporting guidelines.

ro
Setting & Participants: An international group of clinicians with expertise relevant to SDMG

-p
was recruited through purposive and snowball sampling. A scoping review of the literature was
re
presented, followed by three sequential rounds of development, refinement, and voting on
lP

recommendations. Meetings were held virtually and structured to allow participants to provide
na

their input and rapidly prioritize and refine ideas.

Outcomes: Opinions of participants were measured as the percentage who agreed with each
ur

recommendation, whereas consensus was defined as >75% agreement.

Jo

Analytical Approach: Quantitative data were summarized using counts and percentages. A

qualitative content analysis was performed to capture the context of the discussion around

recommendations and any additional considerations brought forward by participants.

Results: The final panel included 26 clinician participants from four countries and 10 clinical

disciplines. Participants reached a consensus on 42 specific recommendations: five regarding the

signs and symptoms accompanying volume depletion that should trigger SDMG; six regarding

signs that should prompt urgent contact with a health care provider including a reduced level of

consciousness, severe vomiting, low blood pressure, presence of ketones, tachycardia, and fever;

3
and 14 related to scenarios and strategies for patient self-management, including frequent

glucose monitoring, checking ketones, fluid intake, and consumption of food to prevent low

blood sugars. There was consensus that renin-angiotensin system inhibitors, diuretics, non-

steroidal anti-inflammatory drugs, sodium-glucose cotransporter-2 inhibitors, and metformin

should be temporarily stopped. Participants recommended that insulin, sulfonylureas, and

meglitinides be held only if blood glucose was low and that basal and bolus insulin be increased

by 10-20% if blood glucose was elevated. There was consensus on six recommendations related

of
to the resumption of medications within 24-48 hours of the resolution of symptoms and the

ro
presence of normal patterns of eating and drinking.

-p
Limitations: Participants were from high-income countries, predominantly Canada. Findings
re
may not be generalizable to implementation in other settings.
lP

Conclusion: A multidisciplinary panel of clinicians reached a consensus on recommendations

na

for SDMG in the presence of signs and symptoms of volume depletion, as well as self-

management strategies and medication instructions in this setting. These recommendations may
ur

inform the design of future trials of SDMG strategies.

Jo

Index words: medications, safety, diabetes, kidney disease, cardiovascular disease, modified
Delphi process

4
Plain Language Summary

Sick day medication guidance (SDMG) is intended to prevent adverse events during acute

illness; however, varying recommendations exist. This study included 26 clinical experts in a

modified Delphi process to develop consensus SDMG recommendations for patients with

diabetes, kidney, or cardiovascular disease. Participants reached a consensus on 42

recommendations for SDMG, including recommendations on the signs and symptoms that

should trigger SDMG, the signs that should prompt urgent contact with a health care provider,

of
and scenarios and strategies for patient self-management. Eleven medication classes were

ro
recommended to be temporarily stopped or adjusted and guidelines were provided for the

-p
resumption of medications. These consensus recommendations may inform the design of studies
re
that examine the effectiveness of different strategies for implementing SDMG.
lP
na
ur
Jo

5
Introduction

Sick day medication guidance (SDMG) has been recommended by several organizations to

prevent potential complications that can arise when people taking medications for chronic

conditions, including diabetes mellitus, kidney, and cardiovascular diseases, experience an acute

illness.1-8 SDMG typically involves recommendations for withholding or adjusting specific

medications in the setting of acute dehydrating illness that could contribute to complications

such as hypotension, acute kidney injury (AKI), diabetic ketoacidosis, or hypoglycemia.9-11 This

of
guidance is intended to mitigate serious adverse medication complications in the setting of

ro
intercurrent illness that could contribute to death or hospitalization.10,12-17

-p
A previous scoping review identified 74 documents pertaining to SDMG, however, the majority
re
were guidelines or educational resources, and only 19 were primary research studies.18 The
lP

review highlighted that there was little empirical evidence available to assess the effectiveness of
na

approaches for implementing SDMG into practice, suggesting that further research to design and

evaluate SDMG is required. However, there was also notable variation in the specific
ur

recommendations included in SDMG resources from different organizations. Before intervention

Jo

studies can be designed to test the clinical effectiveness of SDMG, additional efforts are needed

to establish consensus on the SDMG recommendations for inclusion in future intervention

studies.

The objective of this study was to engage expert clinicians in a modified Delphi process to

generate consensus recommendations for SDMG that could be used by clinicians and researchers

designing future intervention studies.

Methods

Study Design

6
We conducted a modified Delphi process that followed the Conducting and REporting DElphi

Studies reporting guidelines.19 The items presented in the modified Delphi process were

informed by our scoping review of SDMG, a qualitative needs assessment that included primary

care clinicians (i.e., family physicians and pharmacists) and people with a chronic condition of

interest, specifically diabetes mellitus type 2 (DM-2), chronic kidney disease (CKD), or

cardiovascular disease. All session questions were developed and pilot-tested by team members

and patient partners to ensure they were appropriate, clear, and comprehensive. Each round of

of
the Delphi process was conducted virtually using the Zoom videoconferencing platform and

ro
lasted 90 minutes in duration. Ethics approval for this study was granted by the University of

-p
Alberta and University of Calgary Health Research Ethics Boards (ethics approval numbers:
re
Pro00114350 and pSite-21-0024) and all participants provided informed consent.
lP

Recruitment of Participants
na

International stakeholders were recruited through purposive and snowball sampling and invited

to participate in the modified Delphi process if they had clinical expertise in one or more content
ur

areas relevant to SDMG including: primary care, pharmacy, nursing, and medical subspecialties
Jo

(including general internal medicine, endocrinology/diabetology, cardiology/heart failure, and

nephrology). Invitations were sent to authors of published primary research studies, guideline

statements, reviews, commentaries, and patient or care provider educational resources that

addressed the topic of SDMG, regardless of the findings, interpretation, or perspective provided

in the publication. Additionally, snowball sampling was used, where invitees could also suggest

other individuals to include that had expertise relevant to SDMG. Clinician participants received

no financial compensation.

Patient Engagement

7
Two patient partners (SR and NV) participated in the study as active (non-voting) participants in

all three rounds of the modified Delphi process. SR and NV assisted in structuring the research

question and designing the Delphi rounds. They presented their stories of lived experience

managing medications in the setting of an acute illness in the first session to provide context and

framing of the importance of the topic from a patient perspective. In subsequent sessions both

patient partners participated in the small group sessions to help ground the discussions of SDMG

in a patient-centred context. In all sessions, patient partners contributed to the group discussion,

of
were involved in the interpretation of this study’s findings, and in the development of this

ro
manuscript. See supplement for the GRIPP2-Short Form Checklist for the Reporting of Patient

Engagement in Research.20
-p
re
Structure of the Rounds
lP

This modified Delphi process involved three rounds with discussion and voting (Figure 1).
na

Round One began with stories about personal experiences with SDMG provided by our two

patient partners, followed by a presentation of the findings from existing literature identified in
ur

the recent scoping review. Subsequently, a full group discussion of current knowledge about
Jo

SDMG was held, followed by voting on an initial set of recommendations compiled from all

resources identified by the scoping review.18 The Round One statements were categorised into

three domains: (1) symptoms or signs of acute illness that should trigger SDMG (n=15 items),

(2) actions and self-management advice that should be included in SDMG (n=18 items), and (3)

patient groups that would qualify for SDMG and/or specific modifications (n=14 items) (Figure

1). Participants rated (based on importance of individual items) their level of agreement on a 6-

point Likert scale (from 0 = strongly disagree to 5 = strongly agree). Responses were measured

and reported to participants in real time using Mentimeter Interactive Software

8
(www.mentimeter.com) during the session. A summary document of the results from the first

round were emailed to participants after the first session for further review prior to the second

round.

Round Two involved small group discussions based on clinical expertise to further refine the

Round One statements in four clinical groups: 1) patients with DM-2 using medications with the

potential to cause hypoglycemia (sulfonylureas, meglitinides, insulin), 2) patients with DM-2

using medications that may contribute to volume depletion or hypotension (sodium glucose

of
cotransporter-2 [SGLT-2] inhibitors, glucagon-like peptide-1 [GLP-1] receptor agonists,

ro
diuretics, renin-angiotensin-aldosterone system inhibitors), 3) patients with CKD, AKI, or at risk

-p
of AKI, and 4) patients with heart failure (HF), with or without CKD. Each group produced
re
revised statements that were subsequently collated and refined by the facilitators from each
lP

group into a final list of recommendations. Members of each group were provided with an email
na

summary of their group’s revised statement and invited to provide any additional feedback to

their group facilitator before finalizing the recommendations for final review in Round Three.
ur

In the final round (Round 3), we presented the revised statements and accompanying contextual
Jo

statements generated from the discussion to frame each group of recommendations. Participants

then voted on their agreement with each recommendation on a binary scale (Disagree or Agree)

using Mentimeter Interactive Software. The final recommendations were categorised into three

domains (Figure 1): 1) what symptoms or signs of acute illness should trigger SDMG? (n=15

items), 2) what clinical actions should be included in SDMG? (n=15 items) and 3) what

medication instructions should be included in SDMG? (n=19 items) A summary document of the

results from the final round were emailed to all participants after the session accompanied by a

9
survey to provide anonymous feedback on their satisfaction with the process and their perception

whether they felt their opinions were heard during the process.

Data Analysis

Recommendations from Rounds 1 and 3 were voted on and agreement was measured as the

percentage of voting participants who agreed with each individual statement. Participants were

able to abstain from voting on items they deemed were outside of their area of clinical expertise.

The threshold for consensus was pre-specified at 75% agreement for each statement. Qualitative

of
content analysis was also performed to capture the context around the recommendations

ro
developed and the accompanying discussion by participants. Data was obtained from reviewing

-p
the session transcripts, comments typed in the Zoom chat box, and field notes collected by three
re
research coordinators attending the sessions. The data was coded using descriptive and pattern
lP

coding into categories, and high-level themes.

na

Results

Participants
ur

From 60 clinicians who were sent an invitation to participate, a total of 26 participated in the
Jo

modified Delphi process, representing 10 clinical areas of expertise and four countries, including

Canada, the United States, Australia, and the UK (Table 1; Supplement 1). Participants worked

in various practice settings including outpatient specialty clinics, hospitals, primary care clinics,

and community pharmacies. There were 13 male and 13 female participants, with many (n=11)

having more than two decades of clinical experience. Nine participants stated they provide

SDMG frequently or always to their patients, 10 stated they only sometimes provide SDMG, and

seven said they rarely or never provide SDMG.

Round One

10
Forty-seven recommendations in total were initially put forward to the participants in the first

round, and 30 (63.8%) statements reached consensus with 75% or more of the participants voting

in agreement (slightly agree to strongly agree) (Supplement 2). Participants agreed (slightly or

strongly) with 12 of the 15 statements related to “what symptoms or signs of acute illness should

trigger SDMG?”, six of the 18 statements related to “what actions should be included in

SDMG?”, and 12 of the 14 statements related to “which patients should receive SDMG

intervention?”. There were three main themes identified from the discussion in Round 1

of
including 1) the lack of evidence supporting SDMG, 2) the effectiveness of current SDMG

ro
strategies, and 3) challenges for patients identifying sick days and appropriate responses. These

-p
themes and representative quotes are highlighted in Table 2. Participants emphasized that
re
recommendations for SDMG needed to be placed in the context of individual patient needs and
lP

abilities, and that recommendations should be used to design interventions for future research,
na

rather than used as guidelines for current clinical practice. In the absence of clinical evidence that

interventions for SDMG can prevent harm, participants generally expressed a preference for
ur

more conservative general recommendations where they perceived greater potential for benefit
Jo

over harm.

Round Two

Three overarching themes emerged from the discussions held within the four small group

sessions; 1) distinguishing appropriate situations for self-management versus those needing

health care provider support, 2) triaging and clarifying symptoms to guide SDMG, and 3) need

for refinement of parameters for SDMG recommendations.

1. Self-Management versus Health Care Provider Support

11
Groups identified that SDMG should be centered on patient self-management but provided in

tandem with support from their health care provider (HCP). Often there can be limited support

immediately available when patients are sick (e.g., overnight, weekends, etc.) and that SDMG

could be designed for patients and their care partners to self manage their sick days. It was also

acknowledged this may not work for all patients and mechanisms for collaboration and oversight

from their HCPs is still essential in providing SDMG. The need for individualising SDMG to the

patient and their health literacy was also stressed as important, as illustrated in this quote from

of
one participant:

ro
“Self management is always appropriate as there is limited support and care available for

-p
patients when sick, but it should be provided in tandem with the patient trying to engage with
their HCP (pharmacists might be the easiest to get in contact with during an acute illness). Also,
re
need to consider patient’s capabilities, cognitive function, support network, and health literacy
to individualize and implement self management.” [Participant 12]
lP

2. Triaging and clarifying symptoms

na

All groups highlighted that focusing SDMG on symptoms and signs of volume depletion was
ur

appropriate. However, it was identified that not all SDMG monitoring recommendations listed
Jo

would apply to all patients and it should be tailored to their chronic condition and personal

monitoring access and abilities (e.g., weight, blood pressure, ketones, blood glucose, etc.).

One group identified that signs and symptoms could be presented as a traffic light or triage

approach. For mild symptoms, patients could self-manage with SDMG and for severe symptoms

be informed when to contact emergency care (e.g., syncope), as highlighted in the following

comment:

“So, green light would suggest that you can continue and that you are doing well. Yellow might
be alerting a healthcare provider (HCP), but not necessarily, it could be a pharmacist or
primary care provider, and red-light symptoms would prompt an emergency department visit or
calling 911.” [Participant 7]

12
Discussions accentuated the need to emphasize “new or worsening” signs and symptoms, as

many patients can experience some of these symptoms as side-effects of their medications or part

of their chronic condition:

“So, that really needs to be clear in the guidance that this was a change or worsening and that
this applies to all the symptoms, vomiting and diarrhea as well, because those were sometimes
common symptoms that these patients would experience at baseline.”
[Participant 5]

3. Refining Parameters of SDMG Recommendations

of
Each group discussed areas for refinement of specific SDMG recommendations (e.g., signs and

ro
symptoms, medications, and appropriate timeframes). Participants recognised the need for

-p
SDMG interventions to be further studied for effectiveness, as well as implementation strategies
re
and education to be tailored for patients and their care partners. Key questions such as, “how to
lP

ensure patients correctly identify which medications they need to temporarily stop or resume”,
na

while important, were deemed beyond the scope of this modified Delphi process.

Round Three
ur

The revised contextual statements, and list of recommendations created following the synthesis
Jo

of Round Two discussion are shown in Tables 3 and 4, respectively, categorized under three

domains: 1) what signs and symptoms should trigger SDMG? 2) what clinical actions should be

included in SDMG?, and 3) what medications should be included in SDMG. Forty-nine

recommendations in total were put forward to participants in the final round, and 42 (86%) of

them reached consensus with >75% of voting participants agreeing with the recommendation.

Domain 1: Signs and Symptoms to Trigger SDMG

Recommendations that were agreed upon addressed triaging of responses based on severity of

signs and symptoms and a patient’s ability to replace their fluids (Table 4; Supplement 3). For

13
example, participants recommended that vomiting resulting in significant fluid loss should

trigger a SDMG intervention but that greater than four episodes of vomiting in 12 hours or a

patient’s inability to keep fluids down should prompt contact with a patient’s HCP.

Domain 2: Clinical Actions That Should be Included in SDMG

Participants agreed that the SDMG was appropriate for patient (or caregiver) self-management

when there is an absence of severe symptoms, patients are competent and patients (or caregivers)

feel capable of coping, and patients can keep up their fluid intake. Alternatively, participants

of
agreed that patients not coping with self-management, with symptoms that have not resolved

ro
after 72 hours, or who are unable to keep fluids down should seek assistance and support from

-p
their HCP. Participants agreed that SDMG should only be used for temporary self-management
re
until symptoms resolve or for a maximum of 72 hours, whichever comes first. This was in
lP

recognition that even mild symptoms that last longer than 72 hours should involve management
na

and support from a patient’s HCP. Additionally, participants agreed that patients with DM-2 that

have major changes in their blood glucose levels should contact their HCP for advice.
ur

Domain 3: Specific Medication Instructions for SDMG

Jo

Out of the 13 recommendations put forward to the participants, 11 medication instruction

recommendations achieved consensus for inclusion as part of SDMG. Participants agreed with

recommendations for SDMG related to withholding SGLT2i and metformin, adjusting insulin

depending on blood glucose and ketones, and withholding sulfonylurea/meglitinide only if blood

glucose is low and until it recovers. Participants agreed with including recommendations to

withhold angiotensin-converting enzyme inhibitors (ACE-I)/ angiotensin II receptor blockers

(ARBs), angiotensin receptor - neprilysin inhibitor (ARNI), diuretics (loop, thiazides, and

14
potassium sparing), direct renin inhibitors, and non-steroidal anti-inflammatory drugs (NSAIDs)

as part of SDMG.

Delphi Process Evaluation

From the 26 participants, 19 responded to the evaluation survey following completion of the

Delphi process with all stating they were satisfied or very satisfied with the process and 17

stating they felt the process identified valuable SDMG recommendations to be evaluated in

future studies.

of
Discussion

ro
This modified Delphi process involved an international panel of clinicians from four countries

-p
and 10 clinical disciplines. Participants reached consensus on 42 recommendations that can be
re
incorporated into interventions for testing in future clinical trials of SDMG. These included five
lP

recommendations for signs and symptoms of volume depletion that should trigger SDMG, six
na

severe signs (reduced level of consciousness, severe vomiting, low blood pressure, presence of

ketones, tachycardia, and fever) that should prompt contact with a HCP, and 14
ur

recommendations related to appropriate scenarios and strategies for patient self-management,

Jo

including frequent glucose monitoring, checking ketones, fluid intake, and consumption of food

to prevent low blood sugars. Participants also reached consensus on recommendations related to

withholding renin-angiotensin system inhibitors, diuretics, non-steroidal anti-inflammatories,

SGLT-2 inhibitors, and metformin, and that insulin, sulfonylureas, meglitinides should be held

only if blood glucose was low, while a 10-20% increase in basal and bolus insulin should be

made if blood glucose was high. There were six recommendations to guide resumption of

medications within 24-48 hours of resolution of symptoms and when eating and drinking

normally. Achieving consensus on these clinical recommendations is a fundamental first step to

15
inform the design of consistent and acceptable SDMG interventions for patients with diabetes,

kidney disease or cardiovascular disease experiencing acute dehydrating illnesses. However,

further research is required to design the best implementation strategies to support uptake of

these recommendations within the setting of clinical care and patient self-management.

This modified Delphi process builds upon our previous scoping review, where we identified

several areas of inconsistences in SDMG between organizations and published resources.18 In

particular, existing resources provide variable guidance on use of antihyperglycemic medications

of
and insulin in the setting of intercurrent illness, with some recommending patients continue these

ro
medicines, others recommending to stop them, and some suggesting to continue or stop

-p
according to blood glucose levels.1,3,4,8,21-32 Our panel was able to come to consensus in this area
re
and agreed with the recommendations that “if blood glucose levels are low, hold
lP

insulin/sulfonylurea/meglitinide until blood glucose levels recover” and that “if blood glucose is
na

more elevated than usual, an empiric 10-20% increase in basal and bolus insulin doses [is

recommended]”. Furthermore, although some recent resources for SDMG identified GLP-1
ur

receptor agonists3,4,26-29,31-34 and sedative medications as medicines to withhold on sick days, our
Jo

panel did not reach consensus to include these medications in recommendations due to most

GLP-1 agonists having long half-lives, risks of adverse events with rapid withdrawal of sedative

agents, and since they are not expected to worsen volume depletion during an acute dehydrating

illness.

A strength of this modified Delphi process includes the participation of a diverse group of

international clinicians with relevant multidisciplinary expertise and experience with

development of educational resources for SDMG. However, it is possible that the use of

snowball recruitment could have resulted in selection of a more likeminded group of participants.

16
Reassuringly, participants included a mix of clinicians who provided SDMG rarely as well as

frequently, suggesting inclusion of individuals with differing practice behaviours. Due to the

travel restrictions associated with the COVID-19 pandemic, this modified Delphi process was

undertaken virtually using the Zoom videoconferencing platform and real-time feedback through

the Mentimeter Interactive Software. To counter the challenges a virtual environment could

potentially pose, we scheduled time within each of the three rounds for questions, discussion,

and small group sessions to ensure all voices had an opportunity to be heard and incorporated

of
into the recommendations. This modified Delphi process focused specifically on SDMG for

ro
adults, and thus guidance or inference to paediatric settings was beyond the scope of the study.

-p
Participants also recognised that the management of type 1 diabetes mellitus is associated with a
re
higher risk of DKA and requires individualized approaches to its management, and that advice
lP

should be given early and directly from a patient’s HCP. Therefore, development of
na

recommendations specific to this population were not included in this process. Additionally, this

modified Delphi process was designed to focus on clinical content for inclusion in SDMG and
ur

recommendations for implementation strategies, and thus modes of delivery to patients and
Jo

HCPs, were considered outside the scope of the study. Further steps are required to develop

resources and strategies to effectively communicate these recommendations to patients. Finally,

participants were from high-income countries and most were from Canada, therefore the findings

may not necessarily be generalizable to other settings, particularly low- and middle-income

countries.

Our modified Delphi process has resolved some of the uncertainty and inconsistencies identified

from various published studies and resources for SDMG. The recommendations that we

developed and achieved consensus on can be used to design interventions to test the

17
effectiveness of SDMG. However, further research will be required to design and test effective

strategies to implement these recommendations into patient care. Previous research has reported

that traditional approaches to delivering SDMG are prone to patient error identifying the

symptoms that should trigger SDMG and recognizing the appropriate medications to adjust.35

Future research should test educational strategies, support mechanism, and self-management

tools to ensure interventions for SDMG can implement these recommendations as intended. For

practising clinicians, we recognize that the findings of this study might help guide support to

of
people with SDMG where they deem it appropriate. However, these recommendations are not

ro
intended to form general treatment recommendations or guidelines for current clinical practice

-p
but have instead been proposed to promote a consistent and acceptable set of interventions for
re
further implementation and evaluation to close the evidence gap around the effectiveness of
lP

SDMG in community settings.

na

In conclusion, we brought together a multidisciplinary international panel of experts and used a

systematic process to establish consensus on specific recommendations for signs and symptoms
ur

that should trigger SDMG, scenarios and strategies for self-management versus HCP responses
Jo

to sick days, and guidance on withdrawal, adjustment, and resumption of medications during and

after sick days. These recommendations can be used to identify information for inclusion in

clinician and patient facing resources and inform future studies to investigate the effectiveness of

SDMG within clinical care and patient self-management strategies.

Supplementary Material

18
Item S1: Relationships between Disciplines, Professions, and Countries of Participants in the

Delphi process (Numbers represent the number of participants in each category and the size of

ribbons is proportional to the number participants with each characteristic)

Item S2: Modified Delphi Process Round One Results

Item S3: Modified Delphi Process Round Three Results

Article Information

of
Members of the PAUSE Medication Safety Advisory Panel: Bibiana Cujec, University of

ro
Alberta, Canada; David Campbell, University of Calgary, Canada; David Dyjur, Alberta Health

-p
Services, Canada; Edward Siew, Vanderbilt University Medical Centre, United States; Eddy
re
Lang, Alberta Health Services, Canada; Jane de Lemos, Vancouver Coastal Health Research
lP

Institute, University of British Columbia, Canada; Jay L. Koyner, University of Chicago, United
na

States; Julie McKeen, University of Calgary, Canada; Justin Ezekowitz, University of Alberta,

Canada; Kaitlyn Watson, University of Alberta, Canada; Kerry Porter, Queensland Health,
ur

Australia; Maeve O’Beirne, University of Calgary, Canada; Maoliosa Donald, University of

Jo

Calgary, Canada; Matthew James, University of Calgary, Canada; Meghan J. Ho, University of

British Columbia, Canada; Neesh Pannu, University of Alberta, Canada; Nicholas Selby,

University of Nottingham, United Kingdom; Rhonda Roedler, Alberta Health Services, Canada;

Roseanne O. Yeung, University of Alberta, Canada; Ross Tsuyuki, University of Alberta,

Canada; Samuel Silver, Queens University, Kingston Ontario, Canada; Samira Bell,

University of Dundee, United Kingdom; Simon Sawhney, University of Aberdeen, United

Kingdom; Susie Jin, Clinical Community Pharmacist, Ontario, Canada; Tom Blakeman,

University of Manchester, United Kingdom; Vicky Parkins, University of Calgary, Canada.

19
Additional Information: ORCiDs are 0000-0001-6617-9398 (KEW), 0000-0003-3480-1415

(KD), 0000-0001-5653-6558 (EB), 0000-0002-6024-7828 (KM), 0000-0002-8484-8361 (MD),

0000-0002-3724-598X (RTT), 0000-0002-5570-3630 (DJTC), 0000-0002-1876-3917 (MTJ).

Authors’ Contributions: Conceived and designed the study: KEW, EB, RTT, DJTC, NP, MTJ;

contributed to study design: SR, NV, NL, KMD, KM, MD; co-facilitated the modified Delphi

sessions: KEW, KD, SR, NV, KMD, DJTC, NP, MTJ; contributed to data collection and

synthesis: KEW, NL, EB, KD, KMD, MD, RTT, DJTC, NP, MTJ. Each author contributed

of
important intellectual content during manuscript drafting or revision and agrees to be personally

ro
accountable for the individual’s own contributions and to ensure that questions pertaining to the

-p
accuracy or integrity of any portion of the work, even one in which the author was not directly
re
involved, are appropriately investigated and resolved, including with documentation in the
lP

literature if appropriate.
na

Support: This work was supported by a Canadian Institutes for Health Research Strategic

Patient Oriented Multi-Year Research grant number 433773. The funder had no role in study
ur

design, data collection, analysis, reporting, or the decision to submit for publication.
Jo

Financial Disclosure: MTJ was the principal investigator of an investigator-initiated research

grant from Amgen Canada, outside the submitted work. RTT receives consulting fees from

Shoppers Drug Mart and Emergent BioSolutions and is a paid Editor-in-Chief of the Canadian

Pharmacists Journal. He has received investigator-initiated grants from Merck and Sanofi. The

remaining authors declare that they have no relevant financial interests.

Acknowledgements: We would like to thank the clinical experts who participated in the

modified Delphi panel and provided their knowledge and expertise. We also thank Sarah Gil and

Terry Smith for technical assistance.

20
Peer Review: Received June 1, 2022. Evaluated by 2 external peer reviewers, with direct

editorial input from an Associate Editor and the Editor-in-Chief. Accepted in revised form

October 12, 2022.

References

1. Diabetes Canada Clinical Practice Guidelines Expert Committee. Diabetes Canada 2018 Clinical
Practice Guidelines for the Prevention and Management of Diabetes in Canada. Can J Diabetes.
2018;42:S1-S325.
2. American Diabetes Association. Be Prepared: Sick Day Management. Diabetes Spectr.
2002;15(1):54-54.10.2337/diaspect.15.1.54.

of
3. GP Notebook. Sick day rules: how to manage Type 2 diabetes if become unwell with coronavirus
(COVID) and what to do with your medication [Internet].

ro
https://gpnotebook.com/simplepage.cfm?ID=x20200923862158319. Published 2021. Accessed
September 5, 2021.
4.
-p
GP Notebook. Sick day rules - type 2 diabetes [Internet].
https://gpnotebook.com/simplepage.cfm?ID=x2017060421413191130. Published 2021.
re
Accessed September 5, 2021.
5. Healthcare Improvement Scotland. Medicines and Dehydration: Updated Briefing for
lP

Professionals on the Medicine Sick Day Rules card [Internet]. NHS.

https://ihub.scot/media/5798/medicine-sick-day-rules-professionals-leaflet-dispensing-
practices-web.pdf. Accessed August 5, 2021.
na

6. HealthLink BC. Sick-Day Guidelines for People With Diabetes [Internet].

https://www.healthlinkbc.ca/health-topics/uq2659spec. Published 2021. Accessed September
ur

5, 2021.
7. International Diabetes Federation Europe. How to manage diabetes during an illness? “SICK DAY
Jo

RULES”
[Internet].https://www.idf.org/component/attachments/?task=download&id=2155:IDFE-Sick-
day-management. Accessed August 5, 2021.
8. Vancouver Coastal Health. Sick Day Medication Management (for physicians and pharmacists).
http://bcpslscentral.ca/wp-
content/uploads/2021/03/sick_day_medication_management_for_physicians_and_pharmacists
.pdf. Published 2017. Accessed September 5, 2021.
9. Morris RL, Ashcroft D, Phipps D, et al. Preventing Acute Kidney Injury: a qualitative study
exploring 'sick day rules' implementation in primary care. BMC Fam Pract.
2016;17:91.10.1186/s12875-016-0480-5.
10. Whiting P, Morden A, Tomlinson LA, et al. What are the risks and benefits of temporarily
discontinuing medications to prevent acute kidney injury? A systematic review and meta-
analysis. BMJ Open. 2017;7(4):e012674.10.1136/bmjopen-2016-012674.
11. Scott J, Jones T, Redaniel MT, May MT, Ben-Shlomo Y, Caskey F. Estimating the risk of acute
kidney injury associated with use of diuretics and renin angiotensin aldosterone system
inhibitors: A population based cohort study using the clinical practice research datalink. BMC
Nephrol. 2019;20(1):481.10.1186/s12882-019-1633-2.

21
12. Faber SJ, Scherpbier ND, Peters HJG, Uijen AA. Preventing acute kidney injury in high-risk
patients by temporarily discontinuing medication – an observational study in general practice.
BMC Nephrology. 2019;20(1):449.10.1186/s12882-019-1636-z.
13. Ronksley PE, Hemmelgarn BR, Manns BJ, et al. Potentially Preventable Hospitalization among
Patients with CKD and High Inpatient Use. Clin j Am Soc Nephrol. 2016;11(11):2022-
2031.10.2215/cjn.04690416.
14. 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(12):362-365.10.15585/mmwr.mm6712a3.
15. McMurray J, Matthews DM. Consequences of fluid loss in patients treated with ACE inhibitors.
Postgraduate Medical Journal. 1987;63(739):385.10.1136/pgmj.63.739.385.
16. Stirling C, Houston J, Robertson S, et al. Diarrhoea, vomiting and ACE inhibitors: – an important
cause of acute renal failure. J Hum Hypertens. 2003;17(6):419-423.10.1038/sj.jhh.1001571.
17. Ronksley PE, Tonelli M, Manns BJ, et al. Emergency Department Use among Patients with CKD: A

of
Population-Based Analysis. Clin j Am Soc Nephrol. 2017;12(2):304-314.10.2215/cjn.06280616.
18. Watson KE, Dhaliwal K, McMurtry E, et al. Sick Day Medication Guidance for People With

ro
Diabetes, Kidney, or Cardiovascular Disease: A Systematic Scoping Review. Kidney
Medicine.10.1016/j.xkme.2022.100491.

-p
19. Jünger S, Payne SA, Brine J, Radbruch L, Brearley SG. Guidance on Conducting and REporting
DElphi Studies (CREDES) in palliative care: Recommendations based on a methodological
re
systematic review. Palliat Med. 2017;31(8):684-706.10.1177/0269216317690685.
20. Staniszewska S, Brett J, Simera I, et al. GRIPP2 reporting checklists: tools to improve reporting of
lP

patient and public involvement in research. BMJ. 2017;358:j3453.10.1136/bmj.j3453.

21. Lea-Henry TN, Baird-Gunning J, Petzel E, Roberts DM. Medication management on sick days.
Aust Prescr. 2017;40(5):168-173.http://dx.doi.org/10.18773/austprescr.2017.057.
na

22. Diabetes Canada. Stay Safe When You Have Diabetes and Are Sick or at Risk of Dehydration
[Internet]. Diabetes Canada. http://guidelines.diabetes.ca/docs/patient-resources/stay-safe-
when-you-have-diabetes-and-sick-or-at-risk-of-dehydration.pdf. Accessed May 4, 2020.
ur

23. RxFiles. What To Do With Heart Failure Medications If I’m Sick [Internet]. University of
Saskatchewan. https://www.rxfiles.ca/rxfiles/uploads/documents/Heart-Failure-Sick-Days.pdf.
Jo

Accessed August 5, 2021.

24. RxFiles. Type 2 Diabetes And Sick Days Medications To Pause [Internet]. University of
Saskatchewan,. https://www.rxfiles.ca/rxfiles/uploads/documents/SADMANS-Rx.pdf. Accessed
August 5, 2021.
25. Christopoulos S, St-Jean J. What should you do when you are ill and have type 2 diabetes?
[Internet] Diabetes Quebec,. https://www.diabete.qc.ca/en/understand-
diabetes/resources/useful-documents/gerer-le-diabete-de-type-2-lors-des-jours-de-maladie/.
Published 2019. Accessed August 5, 2021.
26. Kaur G, Coane S. Primary Care Sick Day Guidance for the management of adult patients with
diabetes mellitus. In: NHS Diabetes Medicines Management Advisory Group, ed: NHS
Birmingham, Solihull, Sandwell and Environs Area Prescribing Committee (APC); 2020.
27. GP Notebook. Sick day rules - how to manage Type 1 diabetes if become unwell with
coronavirus (COVID) [Internet].
https://gpnotebook.com/simplepage.cfm?ID=x2020092381741158319. Published 2021.
Accessed September 5, 2021.
28. GP Notebook. Sick day rules - type 1 diabetes [Internet].
https://gpnotebook.com/simplepage.cfm?ID=x20170604205420191130. Published 2021.
Accessed September 5, 2021.

22
29. NHS London. Sick day rules: how to manage Type 2 diabetes if you become unwell with
coronavirus and what to do with your medication [Internet]. NHS.
https://www.england.nhs.uk/london/wp-content/uploads/sites/8/2020/04/3.-Covid-19-Type-2-
Sick-Day-Rules-Crib-Sheet-06042020.pdf . Published 2020. Accessed August 5, 2021.
30. Northumbria Healthcare Diabetes Service. Sick Day Rules for People with Diabetes [Internet].
NHS. https://www.northumbria.nhs.uk/sites/default/files/images/PIN451%5B1%5D.pdf.
Published 2017. Accessed August 5, 2021.
31. Leicester Diabetes Centre. What To Do When You Have Type 2 Diabetes And Are Ill: Information
Booklet [Internet]. NHS. https://www.pennine-gp-
training.co.uk/res/Type%202%20SickDay_LDC_JAN2015.pdf. Published 2015. Accessed August
5, 2021.
32. Trend UK. Type 2 Diabetes: What To Do When You Are Ill. Published 2020. Accessed August 5,
2021.
33. My diabetes My Way, NHS Scotland. Sick Day Guidance for Type 2 Diabetes [Internet]. NHS.

of
https://mydiabetesmyway.scot.nhs.uk/resources/internal/sick-day-guidance-for-type-2-
diabetes/. Published 2020. Accessed August 5, 2021.

ro
34. East and North Herts Institute of Diabetes and Endocrinology. Diabetes Medication Sick Day
Rules [Internet]. ENHIDE. https://www.enherts-tr.nhs.uk/content/uploads/2020/03/Sick-Day-

-p
Rule-Card.pdf. Accessed August 5, 2021.
35. Doerfler RM, Diamantidis CJ, Wagner L-A, et al. Usability Testing of a Sick-Day Protocol in CKD.
re
Clin j Am Soc Nephrol. 2019;14(4):583.10.2215/CJN.13221118.
lP
na
ur
Jo

23
Table 1: Participant Characteristics

Number of participants Percentage (%)

(N = 26)
Clinical Discipline
• Nephrologist 8 30.8%
• Endocrinologist 4 15.4%
• Pharmacist 4 15.4%
• Cardiologist 2 7.7%
• Diabetes educator 2 7.7%
• Primary Care Physician 2 7.7%
• General Internist 1 3.8%
• Emergency Physician 1 3.8%
• Nurse Practitioner 1 3.8%
• Clinician Researcher 1 3.8%

of
Years of Clinical Practice
• <5 years 2 7.7%

ro
• 5-10 years 3 11.5%
• 10-15 years 6 23.1%
• 15-20 years

-p
4 15.4%
• >20 years 11 42.3%
Country
re
• Canada 19 73.1%
• United Kingdom 4 15.4%
lP

• United States 2 7.7%

• Australia 1 3.8%
Type of Practice
na

(total does not sum to 25 as multiple answers were possible)

• Outpatient Clinic 19 73.1%
• Hospital 15 57.7%
ur

• Primary Care 3 11.5%

• Community Pharmacy 1 3.8%
Jo

Age
• 31-45 7 26.9%
• 46-55 11 42.3%
• 56-65 8 30.8%
Gender
• Woman 13 50.0%
• Man 13 50.0%
Self-reported Ethnicity
• Caucasian/White 14 53.8%
• Visible Minority 8 30.8%
• Other 2 7.7%
• Prefer not to answer 2 7.7%
Frequency of Sick Day Medication Guidance
• Never 1 3.8%
• Rarely 6 23.1%
• Sometimes 10 38.4%
• Frequently 7 26.9%
• Always 2 7.7%
Population Served
• Between 30,000 and 99,999  1 3.8%

24
• Between 100,000 and 499,999 4 15.4%
• Between 500,000 and 999,999 4 15.4%
• 1,000,000 and greater 17 65.3%

of
ro
-p
re
lP
na
ur
Jo

25
Table 2: Round One Themes
Theme Quote
Level of evidence to “So, I just think we’ve got to be super careful that this is not exactly a robust area
support SDMG of evidence.” [P25]
“…not to say you can't stop medicines in the context of individual assessment, but
real caution about systematic rollout of sick day guidance without a robust
evidence base.” [P22]

Effectiveness of current “I think that there’s more than just about the medications and the context and
SDMG strategies I’ve been thinking about what I say to individual patients and how much I
struggle to get the nuance right for an individual and then I’ll say something
completely different to the next person that comes into clinic.” [P2]

“…generating some parameters or guidance that we might design studies to

of
evaluate whether these strategies are effective and safe rather than.., try to

ro
synthesize this limited evidence right now to make any kind of clinical
recommendations would not be the direction we are intending to go in.” [PI]
“But I know for us it was just, how do you tell people who this pertains to. You

-p
Challenges with defining
sick days know, it’s not just the common cold, it’s not just, oh I’ve got a runny nose, so in
our tools we really tried to identify when you are at risk at dehydration and it
re
wasn’t even that, it was more like when you were at risk of dehydration, or when
you are dehydrated and cannot replace your fluids because technically if you are,
lP

you know, at risk of dehydration, but you are able to replenish then you can
continue your medications - you aren’t dehydrated then, right. So, anyway even
the term sick day was actually something that we got stuck on to tell people.”
na

[P19]
ur

“And, it’s partly about when to stop something, what to stop, but also what to
continue and how to make sure that somebody doesn’t just think about the
Jo

medications, but then thinks about well, how do I decide whenever I’m sick
enough or it’s too complex that I need to ask for advice.” [P2]

26
Table 1: Delphi Round Three Contextual Statements

Domain 1: What symptoms or signs of acute illness should trigger SDMG?

Context:  Symptoms and signs of acute illness that trigger SDMG should be
readily understandable by patients (or caregivers), and help patients identify situations when
they are vulnerable or may be developing volume depletion or dehydration in the community.
Patients with chronic disease may already experience some degree of these symptoms due to
underlying chronic conditions, and some of these symptoms may occur after taking their
medications (e.g., nausea and satiety after taking a GLP-1 analogue). Guidance should thus
emphasize new or worsening of symptoms or signs, particularly when intake or fluids may not
be keeping up with losses. We acknowledge that not all recommendations will apply to all
patients.  For example, changes in weight, blood pressure, blood glucose, and ketones would
only be applicable to those who monitor these at home. 
Domain 2: What clinical actions should be included in SDMG?

of
Context: A graded approach can be used to guide the intensity of support provided for sick
day guidance, which may include self-management as well as assistance provided to a patient

ro
at home from a health care provider. Self management should be provided in tandem with
education and the ability for the patient to engage with their health care provider. The ability

-p
to self-manage should be guided by a patient’s capabilities, cognitive function, support
network, and health literacy. Patients can self-manage if they feel capable, have support, and
re
feel able to cope with monitoring and keeping up with fluid intake (green light) or adjusting
insulin in response to blood glucose. Patients that are not coping or who develop severe signs
lP

or symptoms of hypovolemia, or those related to heart failure/volume overload or

hyperglycemia while holding medications, (red light), should seek medical assistance.
na

Domain 3: What medication instructions should be included in SDMG?

Context: SDMG includes instructions for patients to temporarily stop medications for a short period
of time. This guidance requires appropriate education and tools to allow patients or their care givers to
ur

identify the appropriate medications to be stopped during acute illness. These approaches should be
co-designed and developed with patients and are beyond the scope of this modified Delphi
Jo

process. However, it should be made clear that sick day medication is intended only to temporarily
stop medications during acute illness and that it is important to resume medications for these chronic
conditions when illness has resolved.
**SDMG = sick day medication guidance; GLP-1 = Glucagon-like peptide-1 receptor agonist

27
Table 4: Delphi Round Three Recommendations and Voting Results
Domain 1: What symptoms or signs of acute illness should trigger SDMG?
One or more of the following symptoms or signs of volume depletion, when new or
more frequent or severe than usual, can be considered triggers to initiate SDMG:
Response, (%)
(No. agreed /
No. of
respondents)
Vomiting or diarrhea, resulting in significant fluid losses  25 / 25 100
Anorexia or nausea, resulting is significant decrease in
22/25 88
fluid intake        
New lightheaded, dizziness, or fainting, particularly with
22//25 88
sitting or standing up  
Decreased weight (3kg in 2 days) 

of
20/24 83
Decreased urine output                          18/24 75

ro
New weakness, lethargy, or fatigue  12/24 50
Increased thirst 7/25 28

-p
New dry mouth, lips, or eyes  2/24 8
The following symptoms and signs should be considered severe enough to prompt contact
re
with a health care provider:
Reduced level of consciousness or new confusion  25/25 100
lP

Vomiting > 4 times in 12 hours or cannot keep fluids

24/25 96
down 
Low blood pressure (systolic blood pressure < 80 mmHg;
na

23/25 92
drop of 20 mmHg in systolic; or 10 mmHg in diastolic) 
Moderate or high ketones (for patients taking SGLT2i
21/23 91
ur

or insulin) 
Increased heart rate (increase by 30 bpm)  19/24 79
Jo

Fever (temperature > 38 degrees C (101°F)

18/24 75
on two measurements) 
Extreme thirst  7/24 29
Domain 2: What clinical actions should be included in SDMG?
Self-management is appropriate when:
There is an absence of severe symptoms  24/25 96
Patients feel they are able to cope  23/25 92
Patients can keep up with their fluid intake. 21/24 88
Assistance/support from a health care provider should be sought when:
Patients who feel they are not coping  25/25 100
Signs and symptoms have not resolved within 72 hours 25/25 100
Patients cannot keep up with intake of foods or fluids  24/24 100
Patients are experiencing recurrent low blood glucose
24/25 96
readings 
Patients experience significant increase in blood glucose
24/25 96
not coming down with self-adjustment after 24 h 
911, emergency or urgent care should be sought for:

28
Difficulty or rapid breathing  24/24 100
Reduced level of consciousness or new confusion  23/24 96
Fainting or falls  17/24 71
Sick day guidance should include the following instructions to reverse volume depletion
or dehydration and avoid hypoglycemia or ketoacidosis:
Patients receiving insulin should receive
instructions for more frequent self-monitoring of blood
24/24 100
glucose q4-6 hours while awake and for the duration of
symptoms 
Patients receiving SGLT2i, insulin, or on ketogenic
19/20 95
diets should check ketones 
Increase fluid intake with limited caffeine, and
22/24 92
consider electrolyte replacement solutions 
Patients who took their daily dose of sulfonylurea should

of
be instructed to try to eat foods to prevent low blood
18/23 78
sugars until the effect of the medication has worn off (~12-

ro
24 hours) 

-p
Domain 3: What medication instructions should be included in SDMG?
SDMG should include instructions to temporarily stop these medications: 
re
SGLT2i (e.g., empagliflozin)  22/23 96
If blood glucose low, hold insulin/sulfonylurea/meglitinide
22/23 96
lP

until blood glucose recovers

NSAIDS  21/22 95
Potassium sparing diuretics (e.g., amiloride,
na

18/19 95
spironolactone) 
Loop diuretics (e.g., furosemide)  18/19 95
ur

ACE-I/ ARBs (e.g., perindopril, candesartan)   18/20 90

Thiazides/thiazide-like diuretics (e.g., HCTZ,
18/20 90
Jo

indapamide) 
ARNI (sacubitril/valsartan) 15/17 88
If blood glucose more elevated than usual, empiric 10-
20% increase in basal and bolus insulin doses (if 20/23 87
unsuccessful at lowering blood glucose, contact HCP)
Metformin  19/22 86
Direct Renin Inhibitors (aliskiren)   14/17 82
GLP-1 analogues (e.g., liraglutide)  12/21 57
Sedative medications (e.g. benzodiazepines, Z drugs)  8/17 47
For medication that can be temporarily stopped, stop for: 
Up to three days  24 100
Until signs and symptoms have resolved  21 88
Resuming medications:
For those that can cause hypoglycemia, they should
be resumed at usual doses as soon as symptoms improve 23/23 100
and normal eating and drinking resume

29
 Seek assistance from health care provider about their
23/23 100
medications when symptoms last > 72h
For those that are volume depleting, they should be
resumed at usual doses with 24-48 h or eating and drinking 21/22 95
normally 
Others than those immediately above, they should be
resumed at usually doses within 24-48h of eating and 19/20 95
drinking normally 
*Dominator varies, as participants were allowed to abstain from voting on items they deemed
outside their expertise. Consensus was pre-specified as ≥75% agreement. Items that achieved
consensus are written in black, while those that did not are written in grey.
**No. = number; SDMG = sick day medication guidance; HCP = health care provider; GLP-1 =
Glucagon-like peptide-1 receptor agonist; NSAIDs = Non-steroidal anti-inflammatory drugs;

of
ARNI = Angiotensin receptor - neprilysin inhibitor;  HCTZ = hydrochlorothiazide; SGLT2i =
Sodium/glucose cotransporter-2 inhibitors

ro
-p
re
lP
na
ur
Jo

30
List of Figures
Figure 1 – Modified Delphi Process Flow Diagram
Abbreviations: SDMG=Sick Day Medication Guidance, T2DM= Type-2 Diabetes Mellitus,
BP=Blood Pressure, AKI=Acute Kidney Injury, CKD=Chronic Kidney Disease

of
ro
-p
re
lP
na
ur
Jo

31
 Research Project

Needs Assessment Scoping Review

Interviews/focus group with physicians, pharmacists and patients Primary research studies, guidelines, educational resources

Modified Delphi Consensus Process Domains

3 Rounds of Stakeholder meetings

Round 1: Overview and Survey

of
ro
What symptoms or signs of
-p
What actions should be Which patients should receive
re
acute illness should trigger included in SDMG? SDMG intervention?
SDMG? (n=15 items) (n=18 items) (n=14 items)
lP

Ranking of general statements (6-point Likert scale and qualitative analysis of comments/feedback)
na

Round 2: Focused session statements

ur

Small groups
Jo

T2DM-Hypoglycemics T2DM-Volume/BP CKD-Post AKI Heart failure +/- CKD

Refine general statements based on clinical expertise for specific conditions in small groups &
Qualitative analysis of comments/feedback provided

Round 3: Obtain consensus on SDMG recommendations

What symptoms or signs of What should be included in What medication instructions

acute illness should trigger SDMG? (n=15 items) should be included in SDMG
SDMG? (n=15 items) (n=19 items)

Ranking of revised statements (Binary scale) for inclusion as SDMG recommendations