The psychosocial and financial impact of non-severe hypoglycemic events on people with diabetes: two international surveys

Abstract

Objectives:

To understand the impact of nocturnal and daytime non-severe hypoglycemic events on healthcare systems, work productivity and quality of life in people with type 1 or type 2 diabetes.

Methods:

People with diabetes who experienced a non-severe hypoglycemic event in the 4 weeks prior to the survey were eligible to participate in a nocturnal and/or daytime hypoglycemia survey. Surveys were conducted in Argentina, Australia, Brazil, Israel, Mexico and South Africa.

Results:

In total, 300 respondents were included in nocturnal/daytime hypoglycemia surveys (50/participating country/survey). All respondents with type 1 diabetes and 68%/62% (nocturnal/daytime) with type 2 diabetes were on insulin treatment. After an event, 25%/30% (nocturnal/daytime) of respondents decreased their insulin dose and 39%/36% (nocturnal/daytime) contacted a healthcare professional. In the week after an event, respondents performed an average of 5.6/6.4 (nocturnal/daytime) additional blood glucose tests. Almost half of the respondents (44%) reported that the event had a high impact on the quality of their sleep. Among nocturnal survey respondents working for pay, 29% went to work late, 16% left work early and 12% reported missing one or more full work days due to the surveyed event. In addition, 50%/39% (nocturnal/daytime) indicated that the event had a high impact on their fear of future hypoglycemia.

Conclusions:

The findings suggest that nocturnal and daytime non-severe hypoglycemic events have a large financial and psychosocial impact. Diabetes management that minimizes hypoglycemia while maintaining good glycemic control may positively impact upon the psychological wellbeing of people with diabetes, as well as reducing healthcare costs and increasing work productivity.

Keywords::

• Non-severe hypoglycemia
• Nocturnal
• Daytime
• Diabetes
• Survey
• Healthcare
• Economics
• Quality of life

Introduction

One of the main clinical goals of diabetes treatment is to achieve good glycemic control, to prevent or delay the development or progression of long-term microvascular complications, and reduce the risk of macrovascular complications and all-cause mortality1,2. However, one of the most serious adverse effects associated with glucose-lowering treatment is hypoglycemia, a major barrier to initiating and optimizing therapy as well as long-term adherence3,4.

In many people with diabetes, the negative consequences and unpleasant symptoms associated with hypoglycemia may lead to significant worry or fear of hypoglycemia4. In the short-term, this impact relates to the symptoms associated with the actual hypoglycemic event, such as trembling, altered cognitive functions and, in some severe cases, seizures, coma or even death. The long-term consequences may include impaired social functioning, employment, compromised travel and leisure activities and cognitive impairment over time5–8. Severe and non-severe nocturnal episodes also disrupt sleep, and impair functioning at work and other daily routines9–12. Concern about hypoglycemic events, even non-severe ones, can lead people with diabetes to intentionally omit medication or take sub-optimal doses of glucose-lowering medications, which could impact upon glycemic control and thereby negatively affect long-term outcome3,4,7.

Hypoglycemia is not only associated with considerable burden to the individual in terms of quality of life, but also carries cost implications as a result of healthcare resources used to treat hypoglycemic events (direct costs) and loss of productivity due to absence from work (indirect costs)13. Furthermore, in some countries, an important direct cost related to hypoglycemic events is the increased frequency of blood glucose testing after an event14,15.

The American Diabetes Association (ADA) defines hypoglycemic events as an episode of abnormally low plasma glucose concentration, which can be categorized as either severe (requiring assistance of another person) or non-severe (not requiring the assistance of another person). On average, an individual with type 1 diabetes may experience two episodes of symptomatic non-severe hypoglycemia per week and one episode of severe hypoglycemia per year16. The incidence of hypoglycemia in people with type 2 diabetes is about one-third of that seen in type 1 diabetes17,18; however, due to the higher prevalence of type 2 compared with type 1 diabetes, the absolute number of hypoglycemic episodes occurring in people with type 2 diabetes is greater19,20. Although most glucose-lowering medications increase the risk of hypoglycemia to some extent, the majority of episodes occur in people treated with glucose-lowering medications that raise insulin levels independently of blood glucose, such as sulfonylureas, glinides and insulins21,22. Although non-severe events may not have the same noticeable effect as a severe event on productivity and healthcare costs, they still have a significant impact, especially due to the relative frequency with which they occur3,9. Recent qualitative and quantitative research on non-severe hypoglycemia suggests that such events are, indeed, important barriers to optimal glycemic control and contribute to healthcare costs while reducing quality of life3,10,23–26. A future increase in the burden of hypoglycemia can be expected, considering the rising prevalence of diabetes, especially type 2 diabetes, and increasing insulin use19.

Hypoglycemia has been widely assessed in controlled research settings and, although there is increasing knowledge of the impact of non-severe hypoglycemia in North America and Europe, there is little information on the impact of hypoglycemia in a real-life setting in other parts of the world. The aim of this study was to understand the impact of nocturnal and daytime non-severe hypoglycemic events on healthcare systems, work productivity and quality of life in people with type 1 or type 2 diabetes, the majority being insulin users. Data from North America and Europe have been published previously9,23 and here we report additional data from six countries across five continents.

Patients and methods

Two separate surveys, one designed to assess the impact of non-severe nocturnal hypoglycemic events and the other assessing the impact of non-severe daytime hypoglycemic events, were conducted in Argentina, Australia, Brazil, Israel, Mexico and South Africa.

People with diabetes type 1 or type 2 (self-reported) who had experienced a non-severe nocturnal or daytime hypoglycemic event in the past 4 weeks were eligible for each of the surveys. Participants were recruited from pre-existing panels and additionally through healthcare professionals (HCPs) who were asked to refer their patients. People who had experienced a nocturnal episode were eligible for the nocturnal survey and those who had experienced a daytime episode were eligible for the daytime survey. An individual experiencing both a nocturnal and daytime episode in the past 4 weeks could participate in both surveys if willing. Respondents were screened to confirm the non-severe nature of the hypoglycemic event and determine eligibility. A non-severe hypoglycemic event in this survey was defined as any self-reported event of low blood glucose during the day or night for which respondents were able to treat themselves. Nocturnal non-severe events were judged based on the respondents waking at night (11 pm to 6 am) or noticing symptoms the following morning. For the nocturnal event survey, those who had experienced an event the night before taking part in the survey were excluded, as these people may not have been sufficiently alert to give considered answers and may not have enough information about their experiences in the following day(s).

The surveys were conducted online (using a secure link), or face-to-face to capture parts of the population where internet coverage and internet use may be low. The surveys were based on previous surveys on non-severe hypoglycemia carried out in the US, UK, Germany and France9,23,24; all questions were validated and some modified to be appropriate for local conditions (e.g., patient education). To ensure the validity of the content and avoid use of unfamiliar words and concepts, the survey was developed based on the literature, and qualitative analysis of expert input and interviews with 78 people with diabetes in focus groups from the US, UK, Germany and France10.

The surveys were estimated to take ∼20 (daytime) and 25 min (nocturnal) to complete, and included sections for respondent background, detailed information about the non-severe hypoglycemic event before, during and after the episode, and impact on the healthcare system (requirement of medications and blood glucose tests and involvement of HCPs), economics (loss in working time and productivity at work) and quality of life (negative impacts on daily routines, social environment, emotions and mental state).

Respondents were asked to rate difficulty of going back to sleep after their last non-severe nocturnal hypoglycemic event and the impact of the event on quality of sleep on a scale of 0–10, where 0–3 = easy going back to sleep/low impact on quality of sleep and 7–10 = difficult to go back to sleep/high impact on quality of sleep. In the two surveys, respondents were asked to rate the impact of the non-severe nocturnal hypoglycemic event, on a scale of 0–10, on daily routines, social life and their emotional state, where 0–3 = low impact and 7–10 = high impact. When respondents were asked to estimate how much they spent on treating the surveyed hypoglycemic event, the answer was given in local currency, which was then converted to euros (€) for comparison. The respondents received a small remuneration for completing the survey; however, the incentive was minimal and should not have affected participant response. The respondent data were used only for the purposes of this research report.

Results from the surveys are available for the overall cohort, as well as by diabetes type and country, and are presented as frequencies or descriptive statistics (means and standard deviations). With a sample size of 50 respondents per country and survey, a normal distribution can be assumed that allows for interpretation of results on a country level. However, as it was not an objective of the study to make between-country comparisons, no statistical calculations were performed and only observations are reported.

Results

Respondent characteristics

A total of 300 people were included in each survey (50 from each participating country), and their characteristics are summarized in Table 1. The majority of respondents (76% in the nocturnal and 66% in the daytime survey) had experienced a non-severe hypoglycemic episode less than 2 weeks prior to the survey. All respondents with type 1 diabetes were using insulin. Sixty-eight per cent of the nocturnal survey respondents and 62% of the daytime survey respondents with type 2 diabetes were insulin users (Table 1).

Table 1. Respondent demographics.

	Total	Type 1 diabetes	Type 2 diabetes	Argentina	Australia	Brazil	Israel	Mexico	South Africa
Nocturnal hypoglycemia
n	300	64	236	50	50	50	50	50	50
Male, %	42	50	39	42	44	46	50	38	30
Female, %	58	50	61	58	56	54	50	62	70
Age, mean	48	39	51	50	53	41	50	52	45
Living alone, %	8	2	10	18	4	4	14	2	8
Living with others, %	92	98	90	82	96	96	86	98	92
Working for pay, %	59	77	55	54	56	88	82	46	30
Working hours/week, mean	36.5	37.4	36.4	33.7	30.2	38.7	37.0	39.5	40.9
Type 1 diabetes, %	21	NA	NA	30	20	20	20	18	20
Type 2 diabetes, %	79	NA	NA	70	80	80	80	82	80
Diabetes duration, years	11.3	13.1	10.8	12.7	15.6	8.3	9.9	10.2	11.0
Diabetes treatment, %
Insulin	68^a	100	68^a	54	100	70	60	32	93
Oral medication	62^a	NA	62^a	74	78	60	55	95	13
Non-insulin injectable	4^a	NA	4^a	3	0	8	13	0	0
General hypoglycemia awareness, %
Usually notice early	55	56	55	64	40	60	54	70	42
Usually notice, when BG is already low	25	28	24	22	14	36	28	16	36
Sometimes no, but low BG when checked	18	14	19	10	46	4	16	10	22
No symptoms, event not noticed	2	2	2	4	0	0	2	4	0
Frequency of nocturnal hypoglycemia, %
>1/week	15	14	15	14	0	12	24	18	22
∼1/week	15	20	14	20	0	26	16	16	12
Several times/month–1/month	38	39	38	38	46	32	40	42	32
Few times/year–very rarely	32	27	33	28	54	30	20	24	34
Time since last nocturnal hypoglycemia, %
<2 weeks ago	76	69	65	58	68	82	68	70	50
2–4 weeks ago	34	31	35	32	32	18	32	30	50
Daytime hypoglycemia
n	300	67	233	50	50	50	50	50	50
Male, %	42	51	40	46	44	38	60	42	24
Female, %	58	49	60	54	56	62	40	58	76
Age, mean	47	37	50	50	50	40	49	53	43
Living alone, %	10	3	12	16	10	4	20	4	4
Living with others, %	90	97	88	84	90	96	80	96	96
Working for pay, %	59	79	53	56	68	86	76	42	24
Working hours/week, mean	35.8	34.4	36.4	37.5	29.5	39.3	33.5	37.7	40.1
Type 1 diabetes, %	22	NA	NA	30	20	24	20	20	20
Type 2 diabetes, %	78	NA	NA	70	80	76	80	80	80
Diabetes duration, years	10.7	12.8	10.1	12.8	15.0	6.9	10.0	10.4	9.4
Diabetes treatment (type 2 only), %
Insulin	62^a	100	62	37	100	53	50	38	93
Oral medication	65^a	NA	65	77	73	61	73	95	15
Non-insulin injectable	3^a	NA	3	0	3	0	13	0	0
General hypoglycemia awareness, %
Usually notice early	62	58	63	60	66	68	60	64	52
Usually notice, when BG is already low	25	34	22	24	12	28	28	28	30
Sometimes no, but low BG when checked	10	8	11	6	20	4	8	4	18
No symptoms, event not noticed	3	0	4	10	2	0	4	4	0
Frequency of daytime hypoglycemia, %
>1/week	19	19	19	18	2	20	20	30	22
∼1/week	14	18	12	16	2	22	10	14	18
Several times/month–1/month	26	30	25	22	38	20	34	24	20
Few times/year–very rarely	41	33	44	44	58	38	36	32	40
Time since last daytime hypoglycemia, %
<2 weeks ago	66	70	65	70	62	76	68	76	44
2–3 weeks ago	34	30	35	30	38	34	32	24	56

^aData for people with type 2 diabetes only. 100% of people with type 1 diabetes were treated with insulin.

BG, blood glucose.

Overall, 30% (nocturnal) and 33% (daytime) of participants experienced a non-severe hypoglycemic episode at least once a week, with similar percentages reported across most of the surveyed countries. The exception was Australia, where no respondents reported experiencing a non-severe nocturnal episode ≥1 a week and only 4% experienced a non-severe daytime episode ≥1 a week. A numerically higher percentage of respondents with type 1 diabetes reported having ≥1 episode a week compared with those with type 2 diabetes (34% vs 30% in the nocturnal survey and 38% vs 31% in the daytime survey).

Characteristics of non-severe hypoglycemic events

Almost half of the respondents (49%) reported that they identified their most recent non-severe nocturnal hypoglycemic episode by experiencing symptoms and confirmed the event with a blood glucose test (Table 2). Only 14% reported that their non-severe nocturnal hypoglycemic event was asymptomatic, but that they recognized the event based on a low blood glucose test or by experiencing symptoms the following morning. The frequency of respondents who did not experience any symptoms of the non-severe nocturnal event was similar in those with type 1 and type 2 diabetes (14%) (Table 2). Where causes of nocturnal hypoglycemia were reported, food intake (59%) and exercise (15%) were among the most common. Respondents also reported general poor glycemic control (15%), miscalculation of insulin doses (12%) and too aggressive glucose lowering (3%) as probable causes for their most recent hypoglycemic event. The mean (SD) self-reported cost to treat a single nocturnal event was €2.20 (€3.9), ranging from €0.20 (€0.70) in South Africa to €4.40 (€5.50) in Israel.

Table 2. Details of hypoglycemic event, impact on diabetes management and healthcare utility.

	Total	Type 1 diabetes	Type 2 diabetes	Argentina	Australia	Brazil	Israel	Mexico	South Africa
Nocturnal hypoglycemia
n	300	64	236	50	50	50	50	50	50
Identification of last hypoglycemia, %
Symptomatic, confirmed by BG test	49	55	47	48	40	64	54	28	58
Symptomatic, not confirmed by BG test	37	31	39	38	34	30	32	64	26
Non-symptomatic, confirmed by BG test	14	14	14	14	26	6	14	8	16
Used to recover from hypoglycemia, %
Nutritional or sweet drink	45	50	43	89	48	25	31	62	16
Candy, sweets, biscuit/cookie	26	31	24	25	13	37	36	28	14
Sandwich, light meal, or snack	23	24	22	11	48	35	13	6	26
Glucose tablets or gel	20	17	20	0	21	40	24	18	52
Sugar packs, honey, sweet syrup	14	12	14	14	5	10	22	13	19
Full meal	5	3	5	2	5	0	2	2	16
Other	7	4	8	2	0	12	4	19	0
Contacted HCP^a after last event, %	39	30	41	26	64	38	42	50	14
Daytime hypoglycemia
n	300	67	233	50	50	50	50	50	50
Identification of last hypoglycemia, %
Symptomatic, confirmed by BG test	52	60	50	58	34	62	54	30	72
Symptomatic, not confirmed by BG test	42	34	44	38	62	30	36	62	24
Non-symptomatic, confirmed by BG test	6	6	6	4	4	8	10	8	4
Used to recover from hypoglycemia, %
Nutritional or sweet drink	33	40	31	60	32	22	26	48	10
Candy, sweets, biscuit/cookie	34	30	35	26	36	42	40	30	28
Sandwich, light meal, or snack	19	22	19	14	26	32	22	8	14
Glucose tablets or gel	24	28	22	6	38	26	22	12	40
Sugar packs, honey, sweet syrup	11	10	11	18	0	8	12	6	22
Full meal	13	9	14	6	4	4	6	16	40
Other	4	2	5	4	0	4	2	10	4
Contacted HCP^a after last event, %	36	39	36	24	54	50	20	58	12

^aPrimary care physician, hospital or diabetes clinic, other.

BG, blood glucose; HCP, healthcare professional.

In the daytime hypoglycemia survey, 52% of respondents reported that they identified their most recent event by experiencing symptoms and subsequently confirmed this event with a blood glucose test. Only 6% of the overall respondents reported having no symptoms but confirmed the event with a blood glucose test. The frequency of asymptomatic non-severe daytime events was similar in respondents with type 1 or type 2 diabetes (6%) (Table 2). At the time of their most recent daytime hypoglycemia, the majority of respondents reported being involved in normal daily activities (79%) and 21% were relaxing. The mean (SD) self-reported cost to treat a single daytime event was €2.50 (€3.60), ranging from €0.40 (€0.80) in South Africa to €3.50 (€4.40) in Brazil.

Impact of hypoglycemia on diabetes management and healthcare utilization

Across all of the respondents, the most common treatments for non-severe hypoglycemic events (both nocturnal and daytime) were a nutritional or sweet drink, eating sweets or biscuits, a sandwich, light meal or a snack (Table 2). There were some observed differences between countries in terms of methods of recovery, with regard to the size of the meal/amount of food used to counter the low blood glucose (Table 2).

In the week following their most recent non-severe nocturnal hypoglycemic event, respondents reported adding 5.6 blood glucose tests on average to their mean weekly total of 13.4 tests. Diabetes type and country-specific figures are illustrated in Figure 1. With regard to utilization of healthcare resources, 36% of all respondents reported contacting an HCP (primary care physician, hospital or diabetes clinic) following their non-severe nocturnal hypoglycemic event (Table 2). In terms of insulin dose adjustments, 38% (n = 24) of the respondents with type 1 diabetes and 24% (n = 39) with insulin-treated type 2 diabetes reported decreasing their insulin dose due to a nocturnal hypoglycemic event. Differences between surveyed countries on post-event changes to insulin dose are shown in Figure 2.

Figure 1. Number of blood glucose tests in a normal week and following a nocturnal (A) or daytime (B) non-severe hypoglycemic event.

Figure 2. Percentage of respondents who decreased their insulin dose following a nocturnal or daytime non-severe hypoglycemic event.

In the daytime survey, in the week following their most recent non-severe hypoglycemic event, respondents reported using a mean of 6.4 blood glucose tests in addition to a weekly average of 12.6 glucose tests. Country-specific changes are shown in Figure 1. Among the 212 respondents in the overall cohort using insulin, 30% (n = 64) decreased their insulin dose following a non-severe daytime event. For diabetes type and country-specific changes please refer to Figure 2.

Impact of hypoglycemia on work absenteeism and work productivity

Details of the impact of the non-severe hypoglycemic events on work productivity, including both absenteeism due to the hypoglycemic event and impact on productivity while at work, are summarized in Table 3. Among the 178 respondents in the nocturnal survey who reported working for pay, 29% (n = 52) went to work late, 16% (n = 29) left work early, and 12% (n = 21) reported missing ≥1 full work days due to their most recent non-severe nocturnal hypoglycemic event. Among those who spent time at work following the non-severe hypoglycemic event, 49% (n = 87) reported having difficulties focusing/concentrating, 31% (n = 55) had to reschedule their work day, 29% (n = 52) did not complete tasks in time and 23% (n = 41) postponed appointments. The proportion of respondents with type 2 diabetes, compared with type 1 diabetes, who reported having to reschedule their work day (36% vs 16%), not completing their tasks on time (33% vs 16%) or postponing appointments (26% vs 12%) was about double.

Table 3. Impact of hypoglycemic event on work productivity among respondents working for pay.

	Total	Type 1 diabetes	Type 2 diabetes	Argentina	Australia	Brazil	Israel	Mexico	South Africa
Nocturnal hypoglycemia
Working for pay, n	178	49	129	27	28	44	41	23	15
Went to work late, %	29	27	30	11	25	46	34	17	13
Work time lost, mean (SD)	2 h 30 min (1 h 55 min)	1 h 46 min (1 h 7 min)	2 h 45 min (2 h 4 min)	2 h (1 h 30 min)	3 h 51 min (2 h 22 min)	2 h 26 min (1 h 57 min)	1 h 54 min (1 h 1 min)	3 h 40 min (2 h 57 min)	1 h 0
Left work early, %	16	8	19	4	14	30	10	13	20
Work time lost, mean (SD)	2 h 27 min (1 h 51 min)	2 h 15 min (1 h 16 min)	2 h 29 min (1 h 57 min)	2 h –	3 h 23 min (2 h 26 min)	1 h 42 min (1 h 12 min)	2 h (1 h 47 min)	4 h 40 min (2 h 31 min)	2 h 55 min (2 h)
Missed ≥1 full days, %	12	12	12	7	11	23	7	13	0
Work time lost, mean (SD)	4 d 2 h (4 d 7 h)	4 d 7 h (4 d 5 h)	4 d (4 d 12 h)	2 d 12 h (17 h)	2 d (1 d)	5 d 10 h (5 d 7 h)	2 d 17 h (2 d 2 h)	4 d 7 h (5 d 19 h)	0 0
At work next day
Difficult to focus/  concentrate, %	49	41	52	37	36	66	54	30	60
Rescheduled work day, %	31	16	36	11	29	48	42	22	7
Tasks done on time, %	29	16	33	22	32	43	17	30	20
Postpone appointments, %	23	12	26	7	25	43	20	13	7
Daytime hypoglycemia
Working for pay, n	176	53	123	28	34	43	38	21	12
Went to work late, %	17	19	15	4	0	35	16	24	17
Work time lost, mean (SD)	2 h 10 min (1 h 55 min)	1 h 36 min (1 h 19 min)	2 h 29 min (2 h 8 min)	2 h 0	0 0	1 h 45 min (1 h 49 min)	2 h 41 min (2 h 18 min)	3 h 6 min (2 h 18 min)	1 h 35 min (35 min)
Left work early, %	21	19	21	4	15	30	13	43	25
Work time lost, mean (SD)	2 h 21 min (1 h 42 min)	1 h 51 min (1 h 38 min)	2 h 32 min (1 h 44 min)	2 h 0	3 h 54 min (54 min)	1 h 44 min (1 h 29 min)	1 h 54 min (1 h 30 min)	2 h 25 min (1 h 56 min)	3 h (2 h 39 min)
Missed ≥1 full days, %	8	6	9	4	5	9	5	24	0
Work time lost, mean (SD)	3 d 10 h (4 d 10 h)	4 d 7 h (4 d 5 h)	3 d 2 h (4 d 17 h)	3 d 0	1 d 0	8 d 7 h (6 d 5 h)	2 d (1 d 10 h)	1 d 0	0 0
Work appointments missed because of event, %	26	30	28	11	24	44	29	33	17

d, day; h, hour; min, minute.

Among the 176 respondents in the daytime hypoglycemia survey working for pay, 17% (n = 30) went to work late, 21% (n = 40) left work early and 8% (n = 14) reported missing ≥1 full working day due to their most recent non-severe daytime hypoglycemic event. Additionally, 26% (n = 46) reported missing a work appointment following their most recent event.

Impact of non-severe nocturnal hypoglycemic event on sleep quality

The most commonly reported time respondents experienced nocturnal hypoglycemia was between midnight and 2 am (47%), followed by the time between 2 am and 4 am (30%). About 5% of all respondents with type 1 and type 2 diabetes reported waking up regularly during the night to check their blood glucose.

In response to how easy or difficult it was to get back to sleep, compared with a typical night, 40% of the overall population reported that it was difficult (rating 7–10). In the overall population, almost half of the respondents (44%) reported that the event had a high impact (rating 7–10) on the quality of their sleep compared to an average night. The disrupted sleep also affected the respondents the following day, with about two-thirds of respondents reporting that they wanted to nap and/or rest and about half of the respondents wanting to go to bed earlier the following day. In the nocturnal survey, respondents also answered questions regarding how their hypoglycemic event may have affected other members of their household. Among the respondents who had a bed partner present during the nocturnal hypoglycemic event, 59% made an effort not to wake them, 51% reported that they unintentionally woke up their bed partners, and 20% woke them up on purpose for support or assistance.

Impact of non-severe hypoglycemic event on quality of life

In the overall survey, 28% of respondents reported that the event had a high impact (rating 7–10) on their daily routine, 28% avoided driving the next day and about half of all respondents limited doing household chores and running errands (46% and 52%, respectively). The non-severe nocturnal hypoglycemic event had a moderate impact on wellbeing, with an overall mean rating for emotional impact of 4.9. The most frequent negative feelings reported following their most recent non-severe nocturnal episode were tiredness and/or fatigue (77%), feeling emotionally down or low (56%) and feeling less alert (52%).

In the daytime survey, 28% responded that the event had a high impact (rating 7–10) on their usual daily activities (outside of work hours).

Fear of future nocturnal hypoglycemic episodes was reported by 50% of participants (range = 30% [Argentina]–80% [Australia]), while 39% (range = 18% [South Africa]–74% [Argentina]) reported having a high level of fear of having future daytime hypoglycemic episodes. When respondents in the nocturnal survey were asked whether they feared nocturnal or daytime episodes more, 42% reported that their nocturnal episodes were more frightening than the daytime events, and 42% reported an equal fear of any event, irrespective of the time of occurrence.

Discussion

These two surveys, conducted across five continents with middle- and upper middle-income countries with differing social and healthcare systems, confirm previously reported findings from Europe and North America that non-severe daytime and nocturnal hypoglycemic events have an observable impact on diabetes management and healthcare utilization, productivity at work and daily activities and emotional wellbeing of people with diabetes3,9,23,25,26.

About one-third of respondents reported experiencing a non-severe event ≥1 times per week, and this was similar in both type 1 and type 2 diabetes. The relatively common occurrence of hypoglycemic events in this survey is particularly alarming considering that, for each non-severe hypoglycemic event, there is an associated reduction in societal and healthcare utilities in people with type 1 or type 2 diabetes6. Furthermore, it has been shown that these utility reductions and clinical importance of the hypoglycemic episode increased with the frequency of episodes (from quarterly to monthly to weekly episodes)6.

There are a number of studies that have evaluated the cost of severe hypoglycemia27–30, the associated costs of which are fairly well defined, including hospitalization, emergency services and treatment. Although the cost of a severe episode is higher than that of a non-severe episode, the cost of non-severe hypoglycemic episodes should not be overlooked, especially as non-severe episodes are relatively frequent13,20. Although this study was not designed to calculate direct costs for non-severe hypoglycemic events, it does confirm the increased use of blood glucose test strips, an important cost component related to hypoglycemia. An average of two extra tests in the 24 h following a non-severe event3 and 5.6 extra tests in the week following a non-severe event have been previously reported9. The number of extra glucose tests reported in the current nocturnal survey ranged from 2.0 (Mexico) to 11.3 (Israel). The average number of blood glucose tests used in a week with no hypoglycemic events also differed between the surveyed countries. These differences between the countries may reflect the way people are advised to respond to a hypoglycemic event, the financial ability of the respondents and/or the cost and level of reimbursement of blood glucose tests. The results presented here support the findings from prior publications that non-severe hypoglycemic events can impact upon work productivity, both in terms of reduced productivity while at work and missed work time, ranging from hours to full work days3,9,12. Data from the US and Europe indicate that a high proportion of survey respondents missed work due to non-severe hypoglycemic events9,24 and that nocturnal non-severe hypoglycemic events, compared with daytime events, were associated with the highest loss of work time9,24. A number of respondents in the current surveys reported going to work late and leaving work early, and even missing full days of work, due to their most recent non-severe event; this percentage was higher in the nocturnal survey, but the mean work time lost for those who stayed home was similar in the two surveys. An unexplained finding was that the work productivity of the respondents with type 2 diabetes was affected more compared with those with type 1. It is possible that people with type 1 diabetes have lived with hypoglycemia for longer and are, therefore, adapted to recover more readily than those with type 2 diabetes. Hypoglycemic events led to reduced productivity across all surveyed countries, although the extent of absence from work differed between countries; this may depend on different workplace cultures and ethics, but, in this study, these results may also have been affected by the relatively low frequency of respondents working for pay in some countries (especially South Africa).

In two previous surveys conducted in the US and Europe9,24 and in the present survey, non-severe nocturnal hypoglycemia affected the sleep quality of respondents with either type 1 or type 2 diabetes. Nocturnal events are disruptive and produce symptoms that awaken and frighten people who experience them10. Good sleep is necessary for optimal cognitive functioning31 and, in the current nocturnal survey, almost half of the respondents reported that the non-severe event had a high impact on their quality of sleep compared with an average night. Furthermore, repeated periods of disrupted sleep contribute to a number of molecular, immune and neural changes that play a role in disease development and are associated with greater mortality32. This may be concerning considering the relative frequency with which non-severe events occur.

As seen in the current nocturnal survey, a non-severe hypoglycemic event does not only affect the person with diabetes, but the emotional and physical impact also extends to their supporting family members. In this survey, many respondents reported that a bed partner or other household members are awake with the person experiencing the event, disrupting their sleep as well. In addition, in the second Diabetes Attitudes, Wishes and Needs (DAWN2) study, more than half of the family members of people with diabetes reported that they worried about hypoglycemia and that diabetes had an impact on aspects of their lives33.

Fear of hypoglycemia is common in people with diabetes, and a history of hypoglycemia, even non-severe events, is associated with anxiety and an increased fear of such events3,26,34. It is therefore unsurprising that, in this study, 50% and 40% of the respondents reported a high fear of nocturnal or daytime hypoglycemia, respectively. Likewise, in a survey conducted in Germany, France and the UK, one-third of respondents were very worried about hypoglycemia, and a similar proportion reported maintaining hyperglycemia26. In a survey conducted in the US and Europe, 25% of respondents decreased their insulin dose following a non-severe hypoglycemic episode9, which is very similar to the findings of our two surveys. More people with type 1 diabetes decreased their insulin dose after the event compared with people with type 2 (38% vs 24%). In contrast, more people with type 2 diabetes contacted an HCP after their nocturnal hypoglycemic episode than those with type 1 diabetes (41% vs 30%). It is possible that people with type 1 diabetes are taught to self-adjust their insulin dose depending on their glucose level and food consumption and, hence, are less likely to seek guidance from their HCP. In order to avoid hypoglycemia, people with diabetes may reduce or even skip an oral anti-diabetic drug or insulin dose, which may lead to sub-optimal glucose control and long-term complications3,4,9. Another detrimental self-corrective measure carried out in the hope of avoiding hypoglycemia reported by people with diabetes is deliberate over-eating6. Although our surveys did not collect information on weight gain associated with snacking and ove-eating as a preventive measure, 5% and 13% of the respondents ate a full meal to remedy their nocturnal and daytime hypoglycemia, respectively. Of note, in Mexico, 40% of the respondents reported eating a full meal to remedy their daytime hypoglycemia. It has been shown that educational programs and increased awareness about hypoglycemia among patients and their relatives, compared with standard care, is associated with improved glycemic control and lower rates of hypoglycemia, even with intensive glucose-lowering treatment35. Evidence from a structured group education program for patients and physicians implemented in 10 Latin American countries showed long-term improvement in psychological outcomes and socioeconomic costs36. While this survey did not capture whether the respondents had gone through any structured diabetes education and the effect this may have had on the impact of hypoglycemia, most respondents had diabetes for a relatively long period of time, seeking healthcare which included advice and counseling for years and, therefore, education in the surveyed population would have a lower impact compared to people with recently diagnosed diabetes.

There are potential limitations to these surveys. This survey had a relatively small respondent group of people with self-reported diabetes, with a larger proportion of people with type 1 diabetes compared with the general population. However, even though the total number of respondents is low, the consistency of the frequency across countries strengthens the validity of the outcome. Recall bias can influence findings in any study based on self-reported data, although one study demonstrated that the recall of a non-severe hypoglycemic event of up to 1 week could be considered relatively accurate37. Previous similar surveys and findings from focus groups conducted prior to survey development have indicated that recall was accurate for up to 1 month9,10. There may also have been some selection bias resulting from lack of universal access to the internet and from the choices of patient referral made by HCPs. Also, patients were recruited based on reporting a recent hypoglycemic episode and, although the respondents reported a wide range in the frequency of events experienced (from once a year to daily), the number of people who participated in this survey suffering from frequent hypoglycemia may be disproportionally higher than that seen in the general population An inherent limitation of any survey is the self-reported nature of the data and possible reporting errors, such as primary or secondary diabetes, and the possibility of experiencing symptoms that are attributed to, but unconfirmed as, biochemical hypoglycemia. However, the multi-country approach to the survey provides interesting exploratory information about different cultures, economies and healthcare systems. Furthermore, the availability of the survey both online and as face-to-face interviews allows for the inclusion of a wider and more diverse population, regardless of literacy level. The current surveys did not collect information on the level of glycemic control, which means that some people might have been experiencing frequent hypoglycemia while trying to achieve strict control with an intensive regimen, while others may have been in poor control overall with large glucose fluctuations. The mean age of the sub-population with type 2 diabetes in these surveys was relatively low compared to the general type 2 diabetes population; however, this may lend strength to the observations on the impact of hypoglycemia on work productivity by capturing the younger working-age population. In certain regional cohorts, a large proportion of respondents were elderly and/or not employed, so the impact of non-severe hypoglycemia on work productivity is most likely understated in these regions. These surveys were also not designed to capture healthcare utility to full extent, as the questions on HCP contact did not distinguish between scheduled and unscheduled contact, nor whether the contact was face-to-face or via telephone. Nonetheless, these surveys provide important information from a real-world setting, about the extent and burden of hypoglycemia in patients’ lives.

In many middle-income countries, including some represented in the current nocturnal and daytime survey, there has been a rapid increase in the prevalence of diabetes. The extent of the financial burden this imposes in terms of costs in human, social and economic terms is not well known38–41. Therefore, identification of the underlying factors and where impact is greatest may be of interest for future decision-making in terms of clinical practice, patient education and health policies in individual countries. Structured education on hypoglycemia and treatment strategies to decrease the frequency of these events could reduce care costs and offer substantial health benefits. In conclusion, this survey indicates that non-severe hypoglycemic events cause worry and fear in affected individuals with diabetes. The fear of future hypoglycemic episodes influences how people with diabetes manage their disease and their glucose-lowering medication. Non-severe hypoglycemic events may also impair their ability to perform everyday activities at home and in the workplace. Furthermore, there are direct and indirect economic consequences of a non-severe event. We have found that one episode increases the utilization of healthcare resources, with regard to HCP contact and extra use of blood glucose test strips, and both nocturnal and daytime hypoglycemic events lead to a decrease in productivity at work and increased absenteeism. As shown in these surveys, non-severe hypoglycemic events are a great burden for people with diabetes, as well as their families and society, and reducing the risk of hypoglycemia should be a factor when choosing and optimizing glucose-lowering treatments. Working towards education and efficacious treatment of people with diabetes, while avoiding hypoglycemia, will have a positive impact on the psychological wellbeing and daily life of people with diabetes and their families, as well as on society, by reducing healthcare costs and increasing work productivity.

Transparency

Declaration of funding

The surveys on which this manuscript are based were also funded by Novo Nordisk. The authors take full responsibility for the content of the manuscript, but are grateful to Watermeadow Medical for writing and editorial assistance, funded by Novo Nordisk.

Declaration of financial/other relationships

GF has received fees from Novo Nordisk for participating in an investigator-initiated study, advisory board participation and sponsorship to attend meetings. JS has been a consultant/advisor to Novo Nordisk, AstraZeneca, and Eli Lilly. JS has received a travel grant to present research from participated in research funded by Novo Nordisk. MB has been a consultant to Novo Nordisk. JME peer reviewers on this manuscript have no relevant financial or other relationships to disclose. This manuscript has not been previously published and is not under consideration in the same or substantially similar form in any other peer-reviewed media. Partial data from the individual study countries have been presented as posters at ISPOR-EU 2013 and IDF 2013.