1 INTRODUCTION

Hypoglycaemia can be life threatening and is a feared complication among people with diabetes using insulin.1-5 Initial symptoms can include anxiety, sweating, tremor, cognitive impairment and behavioural changes.6 Hypoglycaemia can be managed by ingesting carbohydrate-containing foods or liquids (oral carbohydrates). However, factors such as impaired hypoglycaemia awareness, lack of knowledge about or access to treatment and challenging emotions can delay appropriate corrective action,7-11 this delay can lead to a point at which people with diabetes cannot safely swallow carbohydrates. At its worst, severe hypoglycaemia (SH) can cause loss of consciousness, seizures, coma or death.6, 12 Per the International Hypoglycaemia Study Group (IHSG) and the American Diabetes Association (ADA) guidance, during severe events, ‘characterized by altered mental and/or physical status (i.e., cognitive impairment) requiring assistance for recovery’, glucagon may be administered by a third party, such as a caregiver (e.g., spouse/partner, parent), if indicated.12, 13

Unfortunately, SH is common: an estimated 22%–46% of people with type 1 diabetes and 21% of those with type 2 diabetes using insulin experience at least one SH event per year.2, 12, 14 In recognition of the importance of SH management, the ADA recommends that ‘Occurrence and risk for hypoglycaemia should be reviewed at every encounter and investigated as indicated’.13 Despite this, studies have reported that conversations around SH are uncommon in clinical practice. Recently, the multinational Conversations and Reactions Around Severe Hypoglycaemia (CRASH) Study found that most people with diabetes do not bring up recent hypoglycaemic events with their healthcare providers, proactive treatment planning was not happening as often as guidelines recommended and that many people with diabetes did not have glucagon on hand even after experiencing an SH event.10, 15-17

To inform clinical initiatives regarding SH management and help healthcare providers prioritize care appropriately, it is important to understand the experiences of people with diabetes and their caregivers during an actual SH event. However, few studies have evaluated these perspectives in this emergency situation. In one study, 16 people with diabetes in Canada were interviewed about their roles in preventing or self-managing hypoglycaemia.6, 18 Other studies have focused on the fear of hypoglycaemia or the impact of hypoglycaemia and impaired awareness about hypoglycaemia on the day-to-day lives of people with type 1 diabetes and their family members.19-27Generally, however, primary data regarding the experiences of people with diabetes, including those with type 2 diabetes who use insulin, and their caregivers in managing an SH event are lacking. To address this gap in the literature, the present study explored the experiences during SH events among adults in the United States from the perspectives of both people with diabetes who use insulin and caregivers who assist them in order to inform healthcare providers and in hopes of elevating conversations about SH risk and management.

2 METHODS 2.1 Study design and participants

A cross-sectional study was conducted with 60 dyads in the United States (N = 120) comprising people with diabetes who use insulin and who recently experienced SH events as well as their caregivers. Each individual participated in a one-on-one telephone interview. Eligible people with diabetes had diagnosed type 1 diabetes or type 2 diabetes, were using insulin, and had received help from their current caregiver during an SH event occurring after at least 1 year of insulin use and within the prior 3 years. Consistent with IHSG and ADA guidelines, SH events were defined as events in which a person with diabetes’ glycaemic level dropped so low that they needed assistance from someone else to recover.12, 13 A caregiver was defined as a person living with the person with diabetes who was aware of the person's diabetes when they assisted with the SH event. People with diabetes and caregivers were excluded if they were under 18 years old, had professional experience treating patients in a clinical setting as such training could alter their ability and approach to recognizing and managing an SH event, or had a self-reported serious functional limitation that might limit the generalizability of their experiences (e.g., dementia). Finally, the study sample excluded individuals who reported having ever filled a prescription for newer formulations of glucagon (i.e., nasal powder or premixed syringe), as early adopters of treatment technology may not represent the general population of people with diabetes who use insulin.28, 29

The study was exempted from institutional review by the New England Independent Review Board, Needham, MA (now part of WCG IRB, Pullayup, WA).

2.2 Study procedures

Participants were recruited from market research panels maintained by Dynata LLC, comprising individuals who have voluntarily agreed to participate in survey and interview-based studies. In order to enrol in the panels, participants complete eligibility screeners including questions about their medical history which allows targeted recruitment. For this study, participants were contacted using traditional methods (e.g., email, phone) as well as more contemporary modalities such as social media outreach. Interested participants completed a brief online screener to ensure they met the study inclusion and exclusion criteria noted above. Self-reported demographic and clinical characteristics were also collected through the brief online screener.

For eligible participants, telephone interviews were conducted using semi-structured guides developed using findings from previous literature and discussions with experts, including Dr. Stuckey. Each interview lasted 45–60 min and covered topics including circumstances of the most recent SH event, the SH event's impact on perspectives towards hypoglycaemia management and information received from healthcare providers. Interviews with caregivers were conducted separately from people with diabetes and additionally addressed questions around decision making during the SH event. Questions included in the interview guide are presented in Table S1.

Interviews were conducted by a moderator trained by the lead investigator (Dr. Stuckey) and were audio-recorded, transcribed and de-identified for coding. Participants were required to provide informed consent before joining the study.

2.3 Analysis

Participant characteristics were summarized quantitatively for people with diabetes and caregivers using numbers and proportions. Analyses were conducted using SAS Enterprise version 7.15. An inductive thematic analysis was used to identify recurrent concepts from the interview data.30 In the first phase, transcripts were reviewed to identify frequently reported concepts and apply labels (codes) to preliminary groupings of similar phenomena (categories). An initial codebook was developed by HS, then the research team tested the codebook against two transcripts to make further revisions. Two independent researchers coded each transcript using NVivo 12 PRO software, and the text in the transcriptions were not double-coded; in other words, extracts from the participants were only coded once into the code that best described the nature of the text.

Inter-rater reliability was assessed using Cohen's kappa (people with diabetes = 0.870; caregiver = 0.887); discrepancies were adjudicated through discussion with the lead investigator. In the second phase, HS created a coding summary of each code and created a narrative that described the main concepts from the codes. The narrative became the basis for the themes, which was refined further based on discussions with the study team. As the themes and codes were finalized, a few codes were divided further into subcodes (e.g., the code ‘mentioned how to prepare, prevent, treat’ was divided into three subcodes of prepare, prevent, treat). The primary coders recoded the data and another ‘coding summary’ was created to become the basis of the analysis and quotes used in the results section of this paper. For transparency, the frequency with which each code appeared was reported and example quotes for key codes were identified. The complete codebook is included as Table S2.

3 RESULTS 3.1 Participant characteristics

Sixty dyads of people with diabetes and their caregivers (N = 120) were included (Tables 1 and 2). Most had lived together for ≥3 years (n = 52), and in most cases, the caregiver was the person with diabetes’ spouse/partner (n = 48). Most people with diabetes were female (n = 43), while most caregivers were male (n = 38). Mean age was 48 years among people with diabetes and 49 among caregivers. Among people with diabetes, 36 (60%) were type 1 diabetes and 24 (40%) were type 2 diabetes. Most (n = 50) had been using insulin for ≥6 years. One third of people with diabetes reported impaired hypoglycaemia awareness (Gold score 4+31). Over 40% of people with diabetes using insulin (n = 25) reported having experienced 11 or more severe hypoglycaemia events within the prior 3 years, 12 people reported having experienced 7–10 SH events and an additional 17 people reported experiencing 4–6 SH events. For 49 people with diabetes, the most recent SH event occurred within the year prior to the interview. Most recent glycated haemoglobin (HbA1c) was ≤53 mmol/mol (≤7%), 53–86 mmol/mol (7%–10%) and >86 mmol/mol (>10%) in 18, 39 and 3 people with diabetes, respectively. Over half (n = 35) of people with diabetes reported usually receiving diabetes-related care from an endocrinologist or in a diabetes specialty practice while 22 people usually saw a primary care physician and three saw a nurse practitioner (Table 1).

TABLE 1. Characteristics of people with diabetes (N = 60) Male gender, n (%) 17 (28%) Age category, n (%) 18–24 2 (3%) 25–34 9 (15%) 35–44 12 (20%) 45–54 17 (28%) 55–64 13 (22%) 65+ 7 (12%) Age, mean (median) 48 (47) Race/ethnicity, n (%) Black/African American 9 (15%) Latino/Hispanic/Chicano 4 (7%) Non-Hispanic White/Caucasian 45 (75%) Multiple racial/ethnic backgrounds 1 (2%) Other 1 (2%) Employment status, n (%) Working full-time 29 (48%) Working part-time 8 (13%) Not working 23 (38%) Diabetes type n (%) Type 1 36 (60%) Type 2 24 (40%) Initiated insulin 6+ years ago, n (%) 50 (83%) Number of SH events within past 3 years, n (%) 1–3 6 (10%) 4–6 17 (28%) 7–10 12 (20%) 11+ 25 (42%) IAH (Gold score 4+), n (%)a 20 (33%) Type of insulin(s) currently used, n (%) Insulin pump 25 (42%) Basal (long-acting) insulin injections 30 (50%) Meal-time (short-acting) insulin injections 25 (42%) Pre-mixed (long and short-acting) insulin injections 5 (8%) Number of comorbid health conditions, n (%) 0 20 (33%) 1 16 (27%) 2 6 (10%) 3 11 (18%) 4+ 7 (12%) Most recent HbA1c, n (%) ≤53 mmol/mol (≤7%) 18 (30%) 53–86 mmol/mol (>7 to 10%) 39 (65%) >86 mmol/mol (>10%) 3 (5%) Most frequent HCP seen for diabetes, n (%) Doctor in a primary care or family practice 22 (37%) Endocrinologist or doctor in diabetes specialty practice 35 (58%) Nurse practitioner in a primary care or family practice 2 (3%) Nurse practitioner in an endocrinology or diabetes specialty practice 1 (2%) Abbreviations: IAH, impaired awareness of hypoglycaemia; SH event, severe hypoglycaemic event. aAssessed using the 7-point Gold scale for hypoglycaemia awareness (1 = always aware, 7 = never aware). TABLE 2. Caregiver characteristics (N = 60) Male gender, n (%) 38 (63%) Age category, n (%) 18–24 4 (7%) 25–34 9 (15%) 35–44 13 (22%) 45–54 13 (22%) 55–64 10 (17%) 65+ 11 (18%) Age, mean (median) 49 (49) Race/ethnicity, n (%) Black/African American 10 (17%) Latino/Hispanic/Chicano 3 (5%) Native American 1 (2%) Non-Hispanic White/Caucasian 45 (75%) Other 1 (2%) Employment status, n (%) Working full-time 27 (45%) Working part-time 4 (7%) Not working 29 (48%) Most recent year caregiver helped person with diabetes with an SH event, n (%) 2017 1 (2%) 2018 2 (3%) 2019 8 (13%) 2020 49 (82%) Relationship to person with diabetes, n (%) Spouse/partner 48 (80%) Child 5 (8%) Parent/legal guardian 4 (7%) Roommate 2 (3%) Sibling 1 (2%) Length of time person with diabetes and caregiver have lived together, n (%) Less than 1 year 4 (7%) At least a year, but less than 3 years 4 (7%) 3 years or more 52 (87%) Diagnosed with diabetes, n (%)a 12 (20%) Diabetes type, n (%) Type 1 2 (17%) Type 2 10 (83%) Currently using insulin, n (%) 5 (42%) Abbreviations: SH event, severe hypoglycaemic event. aPercentages in the subsequent rows are reported using caregivers diagnosed with diabetes (n = 12) as the denominator. 3.2 Symptoms of hypoglycaemia and SH

Primary initial symptoms reported by people with diabetes or caregivers were sweating (n = 60); confusion (n = 57); shakiness (n = 55), having blood glucose that is, or feels low (n = 46), feeling tired/exhausted (n = 43) and dizziness (n = 29). More than one third of the participants (n = 46) described the people with diabetes as having impaired hypoglycaemia awareness or recognizing initial symptoms (based on subjective assessments; not Gold score). As the hypoglycaemic event progressed, most participants reported that people with diabetes were either unable to follow directions and experienced extreme confusion (n = 80); were unable to stand up or felt general weakness (n = 49); had severe sweating (n = 47); felt they do not look or feel right (n = 32); experienced shakiness (n = 30), or were unable to speak or speak clearly (n = 27). Thirty-four people with diabetes and caregivers mentioned that the person with diabetes lost consciousness during an SH event.

3.3 Management of SH and associated challenges

Four themes were formed from the data: (1) how the caregiver supports the person with diabetes in treating SH; (2) how decision making is impaired during an SH crisis; (3) how people with diabetes and caregivers learn about SH through experience and what worked in the past; and (4) how the healthcare provider could provide more support/education. These themes and supporting data are described below. The frequency of references to key codes is summarized in Table 3.

TABLE 3. Codes used in qualitative analysis related to themes Theme Code Sub-code # of people with diabetes/caregivers % of people with diabetes and caregivers (N = 120) Quotes Theme 1. Scrambling to do the right thing: How the caregiver supports the person with diabetes in treating SH Initial actions of hypoglycaemia Person with diabetes drinks and/or eats something by self 82 68% 192 Person with diabetes asks for help or is given help by caregiver 96 80% 271 Caregiver actions during SH Caregiver gave food and/or juice 113 94% 389 Called 911 or medics 69 58% 142 Glucagon administered by caregiver 16 13% 36 Caregiver/person with diabetes has one or more glucagon kits 39 33% 58 Glucagon Not aware of glucagon 46 38% 81 Additional information needed about glucagon and treatment options 42 35% 61 Theme 2. Facing an emotional challenge: How decision making is impaired during an SH crisis Initial emotions Person with diabetes/caregiver anxiety or worry 82 68% 150 SH emotions Person with diabetes/caregiver rising anxiety, worry or panic 78 65% 187 Theme 3. Figuring it out as we go: How people with diabetes and caregivers learn about SH through experience and what worked in the past Learn about treatment of hypoglycaemia Person with diabetes or caregiver's own experience 57 48% 103 Online or own research 27 23% 35 Person with diabetes’ changes in day-to-day life because of SH Diet, insulin, exercise 88 73% 166 Focus on avoiding hypoglycaemia 38 32% 54 Strategies re: when to correct a low 11 9% 17 Caregiver changes in day-to-day life to support the person with diabetes 46 38% 59 Theme 4. Knowing how to treat SH: How the healthcare provider could provide more support/education Discussion of hypoglycaemia with healthcare provider Healthcare provider doesn't talk about hypoglycaemia 43 36% 71 Healthcare provider mentioned how to prevent, prepare or treat hypoglycaemiab: 79 66% 190 a. Preventing (i.e., check blood glucose) 63 53% 93 b. Preparing (i.e., have things on hand) 31 26% 38 c. Treating (i.e., drink juice) 44 37% 59 Healthcare provider mentioned glucagon as treatment 46 38% 93 Past SH events—did discuss 66 55% 112 Past SH events—did not discuss 31 26% 45 3.3.1 Theme 1. Scrambling to do the right thing: how the caregiver supports the people with diabetes in treating SH For people with diabetes, the first arsenal to combat the onset of hypoglycaemia consisted of oral carbohydrates. Caregivers described being ready and willing to help keep things from getting out of hand, and people with diabetes often received assistance even during milder hypoglycaemia. However, some people with diabetes hesitated to reach out for help, as they did not perceive the event to be very severe and had a desire to handle their condition without feeling like a burden to their caregiver. As hypoglycaemia worsened, people with diabetes looked for what works to shorten the duration. Caregivers typically handled SH by providing more carbohydrates, often scrambling to find more. Some caregivers described needing to urge the person with diabetes to eat or drink more while some had to take the lead on administering the carbohydrates: one described literally dumping Coca-Cola down [the person with diabetes’] throat.

I immediately think of what is it that’s going to act the fastest to get her sugar levels up. That’s really, “What do I have? Do I have orange juice? What do I have?” Bam. “What do I have?” It’s more of what do we have on hand.—Caregiver

He’s depending on me to help him because he can’t get up and walk to the kitchen or somewhere to get something to drink, so I’m in a scuttle trying to hurry up and get to him to make sure that he’s OK. So it’s a little scattered.—Caregiver

If oral carbohydrates did not prevent SH, caregivers sometimes felt helpless, and thought that the only option [was] to call for emergency help. When nothing else worked, caregivers sometimes administered glucagon.

I will ask him, “Do you need help? Do you need me to call 911?” And when he doesn’t respond, saying, “No,” I know it’s time to call.—Caregiver

I knew I wasn’t going to get any juice down her. I squirted the gel in her mouth and a minute goes by, two minutes go by, three minutes go by, and there’s no change, I start thinking, “Oh s***. What’s next?” After maybe seven or eight minutes, probably less than 10, I decided that wasn’t working. That’s when I went for the glucagon.—Caregiver

While some people with diabetes and caregivers had heard of or administered glucagon, nearly a third had not, and some had misconceptions about its use or effects.

3.3.2 Theme 2. Facing an emotional challenge: how decision making is impaired during an SH crisis People with diabetes described SH events as foggy, cloudy periods where things are in slow motion, making it nearly impossible to think quickly at a time when speed was of the essence. People with diabetes often experienced disorientation, making it difficult to collect thoughts and make simple decisions—in one case, after having a cold washcloth applied to her forehead, the person with diabetes asked why there was a wet fish on her head. Many people with diabetes were so disoriented they did not remember the event. One described being in a dream, or virtual reality; others described a separation of body and soul and near-death event.

I’m confused. A lot of times, I’ll wake up and think, “This is easy to deal with.” Then, your mind goes, “What’s next? -- It’s easy to deal with -- What’s next?” It’s good angel, bad angel on your shoulders. “You’re fine -- You’re not fine.”—Person with diabetes

I couldn’t see, I couldn’t talk, I couldn’t move … then when I came back to, I had this huge bruise on my leg and I had three people sitting around me. And I was like, what’s going on?—Person with diabetes

Caregivers also described scenes of confusion, where people with diabetes walked aimlessly, stared at the ceiling, talked like they were in another world, or became uncoordinated.

She’ll get confused. She just doesn’t realize the seriousness of [her blood glucose] getting low, and when she’s in that state of mind, she’s kind of slow in getting [food/drink] into her system.—Caregiver

However, most caregivers recognized the urgency of the situation and were desperate to find a remedy to help the people with diabetes recover from the event. Caregivers described SH events as scary, panicky, alarming and dangerous, and felt fearful that [the person with diabetes] may pass out or worse. Caregivers often reported feeling a rising sense of panic and questioning their actions.

There’s a sense in which she’s always eventually wake up, but there’s a part of me that fears that it only takes one time where that doesn’t happen.—Caregiver

Both people with diabetes and caregivers stated the importance of having a plan in the event the person with diabetes needed to call out for help.

You have something in plan in your mind, because the worst thing you want to do is stand in the middle of your kitchen saying I don’t know how to help myself. I don’