Medical Resource Use, Disturbance of Daily Life and Burden of Hypoglycemia in Insulin-treated Patients With Diabetes

Results From a European Online Survey

William D Willis; Jesús I Diago-Cabezudo; Anne Madec-Hily; Aftab Aslam


Expert Rev Pharmacoeconomics Outcomes Res. 2013;13(1):123-130. 

In This Article


This large-scale, online survey, conducted in the UK, France and Germany, clearly shows a considerable incidence of self-reported hypoglycemic events among individuals with insulin-treated diabetes. Indeed, 18% of patients had at least one major hypoglycemic event and 5% had three to five such events, during the previous 12 months. The mean overall number of major hypoglycemic events was 2.4 per patient. Minor hypoglycemic events also occurred frequently; overall, 90% of all patients across the three countries reported minor hypoglycemic episodes, and significantly more patients with Type 1 than Type 2 diabetes reported minor hypoglycemic episodes during the previous 12 months. Importantly, many patients appear to self-manage their hypoglycemic events, or are helped by people in the community, general practitioners or ambulance staff, since the overall frequency of emergency room visits was low (4%), as was the overall frequency of hospitalizations due to hypoglycemia (3%). Slightly more than half the survey population reported discussing hypoglycemia with a healthcare practitioner, and 6.5% of patients discussed hypoglycemic events with a pharmacist during the previous 12 months. Added to this, 10% of patients indicated that they had taken days off work due to hypoglycemia, and the mean rate of absenteeism among workers who took at least 1 sick day in the previous 12 months was 6 days per year. Importantly, all these findings suggest that the total costs associated with hypoglycemia in patients with diabetes may be markedly greater than at first considered.

Generally, throughout this survey, no major differences in patient responses were identified between the three countries assessed, or between individuals with insulin-treated Type 1 or Type 2 diabetes. However, major hypoglycemic events were significantly more common in patients with Type 1 rather than Type 2 diabetes (30 vs 16% of patients; p < 0.05), as also were minor hypoglycemic events during the previous 12 months (95 vs 85% of patients; p < 0.05). These findings are to be expected, since patients with Type 1 rather than Type 2 diabetes often have fewer warning signs of hypoglycemia.[26]

Furthermore, the current survey revealed that patients in the UK, compared with patients in France and Germany, were less likely to talk with a healthcare practitioner about hypoglycemia (2.4 vs 3.4 consultations per patient), and patients in France (especially those with Type 2 diabetes) were more likely than patients in the UK and Germany to consult with a pharmacist (2.2 vs 1.1 consultations per patient). The latter statistic can perhaps be attributed to cultural reasons and to issues such as greater patient contact with a pharmacist in France each time a BGM is issued.

Although the mean number of sick days off work each year was relatively low (4–7 days), this statistic highlights a potential loss of productivity, and it should not be forgotten that some patients with diabetes self-manage hypoglycemia, and some may have a staunch, 'solider-on' attitude and may return to work before full recovery from hypoglycemia; working when 'below par', may also contribute to an overall loss of productivity. Moreover, although the number of emergency room visits and hospitalizations due to hypoglycemic events was relatively low, estimated total costs for healthcare systems would remain substantial, as absolute numbers based on our survey equate (approximately) to: emergency room (UK: 25,000 patients, assuming 5% of patients [n = 16,000] with Type 1 and 2% of patients [n = 9000] with Type 2 diabetes; France: 39,000 patients [6% with Type 1, 4% with Type 2 diabetes]; Germany: 55,500 patients [5% with Type 1, 2% with Type 2 diabetes]); hospital admissions (UK: 22,000 visits; France: 83,000 visits; Germany: 46,500 visits).

Interestingly, patients with Type 1 rather than Type 2 diabetes were significantly more likely to test their blood glucose level every time before driving (31 vs 20% of patients), and this again may reflect a fear of hypoglycemia and an awareness among patients with Type 1 diabetes that sometimes they have problems recognizing the warning signs of hypoglycemia by symptoms alone.[26]

As already emphasized, the over-riding implication of the current survey is that fear of hypoglycemia causes patients with diabetes to maintain their blood glucose at levels higher than recommended by their healthcare team, to try to reduce the risk of hypoglycemia. Clearly, this strategy is likely to carry an attendant risk of raised HbA1c levels,[21] increased diabetic complications[22,23] and increased healthcare costs.[25] Another publication of additional data from the current survey documented that 40% of diabetics overall "…were very worried about hypoglycemia", and 37% stated that "I tend to keep my blood glucose levels a little higher than I should to help avoid hypoglycemia".[20] Thus, several important issues arise in light of the substantial rate of deliberately maintained hyperglycemia:

  • How many patients are actually aware that their 'hyperglycemic strategy' or 'running high' (i.e., keeping blood glucose levels a little higher than they should be) may have a detrimental effect on long-term health? Indeed, many patients with diabetes may be unaware of the correlation between blood glucose concentration and HbA1c level,[21] and may be unaware of their last HbA1c test result and when the test was conducted. Alternatively, is it a form of 'time discounting',[27,28] where a desired result in the future is perceived as less valuable than one in the present (i.e., the short-term avoidance of hypoglycemia)?

  • Which patients are at greatest risk from running high? For instance, some patients may have a major fear of hypoglycemia and may incorporate a large 'buffer' above 70 mg/dl (3.9 mmol/l) at which to maintain their blood glucose values. Other patients may not be so fearful of hypoglycemia and may allow their blood glucose levels to be maintained at a level closer to the hypoglycemic threshold of 70 mg/dl (3.9 mmol/l);

  • It will be particularly pertinent in future studies and surveys to clarify exactly what patients mean when they say "their blood glucose values are running high", and to have some consistency in the definition of hyperglycemia. For example, are patients recording random or fasting blood glucose values, values in the preprandial or postprandial state or values just before going to bed? Also, what percentage, above-normal increase in blood glucose levels (i.e., a range of values or a single fixed value) do patients consider to define running high? A goal of future studies and surveys should be to try to accurately define and quantify the average extent to which fear of hypoglycemia raises particular blood glucose levels, increases HbA1c levels, as can be extrapolated from the ADAG study,[21] and increases diabetic complications, as can be extrapolated from UKPDS data.[23] Indeed, the UKPDS showed that for each 1% increase in HbA1c level, relative risk increases were 37% for microvascular complications and 21% for diabetes-related end points and death.[23] Interestingly, a strong correlation (r = 0.92) has been defined between mean blood glucose concentration and HbA1c levels: thus, even a small increase in mean blood glucose concentration (28–29 mg/dl) has been linked with an increase of 1% in HbA1c value, and therefore with a marked increase in diabetic complications;[29]

  • In light of the current survey findings that 80% of patients with diabetes would value a meter that tells them when their blood glucose level is getting high/low at a particular time of the day, and that 74% of patients with diabetes would value a meter giving an advanced warning of potential hypoglycemia, it would be especially useful if future studies were specifically designed to quantify the efficacy of SMBG with novel BGMs in terms of reducing fear of hypoglycemia and subsequent complications of diabetes. Projected short and long-term cost savings resulting from reduced complications of diabetes could then be calculated.

A limitation of the current study is obviously the subjective nature of the online questionnaire, even though large numbers of patients with diabetes were surveyed in France, Germany and the UK. Other limitations creating the potential for bias, or potentially limiting the generalizability of results, include the slightly larger sample size surveyed in Germany than the other two countries; cultural and language differences between countries, possibly making it difficult to ensure questionnaire consistency; and the fact that only individuals with insulin-treated diabetes, and not patients treated with antidiabetes agents other than insulin, were included. Furthermore, survey data are based on self-reported patient recall of hypoglycemic events and, as reported elsewhere by other groups, inaccurate patient recall has the potential to cause some under-reporting or bias.[30,31] Nonetheless, the survey provides important information from a real-world setting about the true extent of hypoglycemia and fear of hypoglycemia in patients with diabetes, and about the potential for such fear to increase complications of diabetes and dramatically increase the emotional and economic burden of diabetes.