Advertisement

The Impact of Non-Severe Hypoglycemic Events on Work Productivity and Diabetes Management

Open ArchivePublished:June 06, 2011DOI:https://doi.org/10.1016/j.jval.2011.02.001

      Abstract

      Objectives

      Hypoglycemia is a common complication of treatment with certain diabetes drugs. Non-severe hypoglycemic events (NSHEs) occur more frequently than severe events and account for the majority of total events. The objective of this multi-country study was to identify how NSHEs in a working population affect productivity, costs, and self-management behaviors.

      Methods

      A 20-minute survey assessing the impact of NSHEs was administered via the Internet to individuals (≥ 18 years of age) with self-reported diabetes in the United States, United Kingdom, Germany, and France. The analysis sample consisted of all respondents who reported an NSHE in the past month. Topics included: reasons for, duration of, and impact of NSHE(s) on productivity and diabetes self-management.

      Results

      A total of 1404 respondents were included in this analysis. Lost productivity was estimated to range from $15.26 to $93.47 (USD) per NSHE, representing 8.3 to 15.9 hours of lost work time per month. Among individuals reporting an NSHE at work (n = 972), 18.3% missed work for an average of 9.9 hours (SD 8.4). Among respondents experiencing an NSHE outside working hours (including nocturnal), 22.7% arrived late for work or missed a full day. Productivity loss was highest for NSHEs occurring during sleep, with an average of 14.7 (SD 11.6) working hours lost. In the week following the NSHE, respondents required an average of 5.6 extra blood glucose test strips. Among respondents using insulin, 25% decreased their insulin dose following the NSHE.

      Conclusions

      NSHEs are associated with substantial economic consequences for employers and patients. Greater attention to treatments that reduce NSHEs could have a major, positive impact on lost work productivity and overall diabetes management.

      Keywords

      Introduction

      Hypoglycemia is a common complication of treatment with antidiabetic medications. Symptoms of hypoglycemia often include, but are not limited to: pounding heart (8%–62%), trembling (32%–78%), hunger (39%–49%), sweating (47%–84%), difficulty concentrating (31%–75%), and/or frank confusion (13%–53%) [
      • Hepburn D.A.
      Symptoms of hypoglycemia.
      ]. According to the American Diabetes Association, hypoglycemic episodes are categorized as either severe (events that require the assistance of another person to administer a remedy) or non-severe (events that do not require assistance from another individual). These non-severe hypoglycemic events (NSHEs) may be symptomatic and confirmed by a low blood glucose reading. However, NSHEs may also be asymptomatic (patient does not exhibit typical symptoms, but hypoglycemia is confirmed by a low blood glucose reading), or probable (patient has typical symptoms, but low blood glucose is not confirmed by a blood glucose reading) [
      American Diabetes Association Workgroup on Hypoglycemia
      Defining and reporting hypoglycemia in diabetes: a report from the American Diabetes Association Workgroup on Hypoglycemia.
      ].
      Data from multiple studies indicate that NSHEs occur in 24% to 60% of patients with diabetes. They account for 88% of total hypoglycemic events [
      • Miller C.D.
      • Phillips L.S.
      • Ziemer D.C.
      • et al.
      Hypoglycemia in patients with type 2 diabetes mellitus.
      ,
      • Vexiau P.
      • Mavros P.
      • Krishnarajah G.
      • et al.
      Hypoglycaemia in patients with type 2 diabetes treated with a combination of metformin and sulphonylurea therapy in France.
      ,
      • Davis R.E.
      • Morrissey M.
      • Peters J.R.
      • et al.
      Impact of hypoglycaemia on quality of life and productivity in type 1 and type 2 diabetes.
      ,
      • Pramming S.
      • Thorsteinsson B.
      • Bendtson I.
      • Binder C.
      Symptomatic hypoglycaemia in 411 type 1 diabetic patients.
      ]. According to a retrospective study conducted in Canada to examine the costs of hypoglycemia and hyperglycemia, NSHE management accounts for up to 13% of all out-of-pocket costs related to diabetes [
      • Harris S.B.
      • Leiter L.A.
      • Yale J.F.
      • et al.
      Out of pocket costs of managing hypoglycemia and hypoglycemia in patients with type 1 diabetes and insulin-treated type 2 diabetes.
      ]. In addition, approximately one-third of patients with diabetes report that NSHEs interrupt and affect their ability to carry out day-to-day tasks including housework, social activities, sporting activities, and sleep [
      • Jermendy G.
      • Erdesz D.
      • Nagy L.
      • et al.
      Outcomes of adding second hypoglycemic drug after metformin monotherapy failure among type 2 diabetes in Hungary.
      ,
      Diabetes UK
      Survey reveals hidden incidence of 'hypos' among people with type 2 diabetes [Internet].
      ]. NSHEs can also affect health-related quality of life. Patients report that fear of hypoglycemia may lead to increased panic and anxiety, or deliberate overeating in order to boost blood glucose levels [
      • Levy A.R.
      • Christensen T.L.
      • Johnson J.A.
      Utility values for symptomatic non-severe hypoglycaemia elicited from persons with and without diabetes in Canada and the United Kingdom.
      ,
      • Currie C.J.
      • Morgan C.L.
      • Poole C.D.
      • et al.
      Multivariate models of health-related utility and the fear of hypoglycaemia in people with diabetes.
      ]. In addition, hypoglycemic events, and concerns regarding future events, can lead patients with diabetes to maintain higher glucose levels overall and/or fail to use adequate doses of insulin or oral antidiabetic medications [
      • Frier B.M.
      How hypoglycaemia can affect the life of a person with diabetes.
      ,
      • Hauber A.B.
      • Mohamed A.F.
      • Johnson F.R.
      • Falvey H.
      Treatment preferences and medication adherence of people with type 2 diabetes using oral glucose-lowering agents.
      ,
      • Wild D.
      • von Maltzahn R.
      • Brohan E.
      • et al.
      A critical review of the literature on fear of hypoglycemia in diabetes: Implications for diabetes management and patient education.
      ,
      • Shiu A.T.
      • Wong R.Y.
      Fear of hypoglycaemia among insulin-treated Hong Kong Chinese patients: implications for diabetes patient education.
      ]. Patient concerns about potential hypoglycemia are also associated with an increased frequency of self-monitoring of blood glucose [
      • Hansen M.V.
      • Pedersen-Bjergaard U.
      • Heller S.R.
      • et al.
      Frequency and motives of blood glucose self-monitoring in type 1 diabetes.
      ] and an attendant increase in blood glucose strip use [
      • Farmer A.
      • Balman E.
      • Gadsby R.
      • et al.
      Frequency of self-monitoring of blood glucose in patients with type 2 diabetes: association with hypoglycaemic events.
      ].
      Existing data indicate that NSHEs occur more frequently in insulin-treated than non-insulin treated patients, are more likely to occur on working compared with non-working days, and are associated with missed work [
      • Pramming S.
      • Thorsteinsson B.
      • Bendtson I.
      • Binder C.
      Symptomatic hypoglycaemia in 411 type 1 diabetic patients.
      ,
      Diabetes UK
      Survey reveals hidden incidence of 'hypos' among people with type 2 diabetes [Internet].
      ]. Additionally, the impact on work productivity for the majority of NSHEs which occur outside of the work place, such as during sleep, is unknown. Data documenting the effect of NSHEs on work productivity are limited; one 12-month, prospective study found that, among insulin-treated patients with diabetes, up to 30% of minor hypoglycemic events occur at work [
      • Leckie A.M.
      • Graham M.K.
      • Grant J.B.
      • et al.
      Frequency, severity, and morbidity of hypoglycemia occurring in the workplace in people with insulin-treated diabetes.
      ]. Likewise, NSHEs have been shown to decrease work productivity in individuals with type 2 diabetes. Results from a survey-based, retrospective study (N = 200) found that 6.7% to 10.3% reported leaving work/school early due to the hypoglycemic event, and 9.3% stayed home the day after the hypoglycemic event (type 2 diabetes only) [
      • Leiter L.A.
      • Yale J.F.
      • Chiasson J.L.
      • et al.
      Assessment of the impact of fear of hypoglycemic episodes on glycemic and hypoglycemia management.
      ].
      The objective of this multi-country study was to examine the impact of NSHEs occurring during work, outside working hours, and nocturnally, on productivity (both absenteeism and reduced performance while at work) and diabetes self-management in patients with type 1 and type 2 diabetes. In order to address the scarcity of currently available published data regarding the impact of NSHEs, this study addresses productivity and identifies the impact on work productivity for the majority of NSHEs which occur outside of the work place, such as during sleep.

      Methods

      To facilitate survey development, extensive background qualitative research was conducted using focus groups and individual telephone interviews among a total of 68 patients in the United States, United Kingdom, and France who had recently experienced an NSHE. As Germany was not originally included in the countries to be surveyed, they did not participate in the interview process. A total of eight focus groups were conducted, consisting of individuals aged >18 years with type 1 or type 2 diabetes, who reported experiencing ≥1 NSHE in the past 3 months. Topics discussed, from a patient perspective, included how NSHEs impact daily life, functioning, and well-being, as well as diabetes treatment and self-management.
      Feedback received from these interviews was combined with a review of the current literature and expert input, and a survey was developed to assess the impact of NSHEs on work productivity and diabetes self-management. The survey (in English) was then cognitive-debriefed and pilot-tested in six additional patients. These steps were undertaken to ensure content validity (relevant questions) and face-validity with respondents (i.e., no unfamiliar/strange words or concepts). After six interviews there were no significant changes to the survey required and debriefing was concluded. The final questionnaire was translated into all relevant languages using a linguistic validation method [
      • Wild D.
      • Eremenco S.
      • Mear I.
      • et al.
      Multinational trials—recommendations on the translations required approaches to using the same language in different countries, and the approaches to support pooling the data: the ISPOR patient-reported outcomes translation and linguistic validation good research practices task force report.
      ].
      The survey was administered in the United States, United Kingdom, Germany, and France via a secure Internet server. Respondents were selected by applying a sampling frame to a pre-existing panel of individuals with self-reported type 1 or 2 diabetes. In order to ensure the generalizability of the results from the panel, the panel structure and recruitment used the following strategies. The panel used for the survey was multisource: panellists were mainly recruited online via a wide range of permission, e-mail recruitment, affiliate networks and web site advertising, avoiding potential bias associated with single source recruitment methodology. Patients were recruited from literally hundreds of web sites as well as from face to face and telephone surveys where appropriate to include members who were not frequent online users. Additionally, the panel was used for research only. Panelists were not exposed to third party advertising or direct marketing campaigns, nor were their personal data sold to third parties. The panel was also frequently refreshed to ensure that the panel was dynamic in nature and reflected any changes in the online population that may have been occurring. Lastly, the incentive was very low to help ensure that there was not undue incentive to participate in the panel.
      All respondents were required to first respond to a healthcare profiler (screening) to ensure that their diabetes had been diagnosed by a physician, and that diabetes treatment had been initiated. In addition, eligible respondents were required to be ≥18 years of age, able to read in the predominant language of their country of residence, and to have experienced ≥1 NSHE in the past month (to reduce recall bias, a maximum of 1-month recall was applied). During the survey, respondents answered questions regarding what they perceived to have caused the hypoglycemic event, the duration of the event, and the impact of the NSHE on productivity. The survey took approximately 20 minutes to complete and respondents were remunerated the equivalent of $10.00 US dollars (USD) for completing the survey.
      A stratified sampling procedure (to account for disproportionate response rates between stratification categories) was applied to ensure adequate sample size in relevant subcategories. This included extending a survey invitation to individuals who expressed interest in participating and met the criteria for an unfilled stratification category. Stratification variables applied were: age (18-29 years, 30-49 years, 50-64 years, and ≥ 65 years) and diabetes type (type 1 diabetes and type 2 diabetes). Based on current guidelines [
      American Diabetes Association Workgroup on Hypoglycemia
      Defining and reporting hypoglycemia in diabetes: a report from the American Diabetes Association Workgroup on Hypoglycemia.
      ], NSHEs were defined as a hypoglycemic situation in which the patient had low blood glucose, but did not require help from anyone else to manage the episode. Respondents were also permitted to classify events as NSHEs if they had hypoglycemic symptoms such as sweating and/or confusion, with or without confirmatory blood glucose measurements, or if they experienced no symptoms, but noted a hypoglycemic episode when recording blood glucose levels.
      The time/place of the NSHE was classified as follows: a) at work, b) during sleep, or c) or during the day, but not at work. In addition, the survey incorporated several real-time validation steps (e.g., it affirmed any responses, such as patient age, that fell outside the plausible minimum-maximum input values) and used skip-patterns based on respondent replies.
      Results by country are presented via frequencies or descriptives (means and standard deviations) with differences explored using analysis of variance (ANOVA) for continuous variables and chi square for proportions. Responses for amount of work time lost contained outliers, or observations that appeared to be inconsistent with other observations in the data set. To account for these departures from normality, a 5% trim was employed [
      • Wilcox R.R.
      The goals and strategies of robust methods.
      ].
      Prior to conducting the focus groups, ethics approval was received from New England IRB (approval # 08-288) for the patient interview and focus group portion of the study groups and from Independent Review Consulting (approval #09033-01) for the survey panel portion of the study. Questions on ethnicity were removed from the German survey.

      Results

       Respondent characteristics

      A total of 6756 respondents with self-reported diabetes were screened. Of these, 2669 reported an NSHE during the last month; 54% of these individuals (n = 1431) reported working for pay. Finally, 27 subjects (24 subjects listed “other” and 3 reported, “I don't remember”) were removed from the analysis because their work status was unable to be classified; this left 1404 respondents in the analysis.
      For most demographic variables (Table 1), no substantial differences were observed across the total sample for key variables (e.g., percentage of respondents who used oral vs. insulin treatment or had type 1 vs. type 2 diabetes). Significant cross-country differences, however, were found for age (ANOVA, P < 0.001), gender (chi-square, P < 0.001), and diabetes duration (ANOVA, P < 0.001). The United States had the highest mean respondent age (49.4 ± 13.2) and France had the youngest mean age (37.6 ± 12.3). The greatest percentage of females was found in the United States (64.1%), whereas Germany had the smallest proportion of females (42.8%). The mean duration of diabetes was longest in the US (15.8 years, SD 12.7) and shortest in Germany (8.5 years, SD 7.7).
      Table 1Demographic characteristics.
      TotalUSAUKGermanyFranceP value
      Sample size with NSHE within last month (N)1404409385236374
      Age, mean (SD)42.4 (13.9)49.4 (13.2)40.4 (13.8)41.2 (13.2)37.6 (12.3)<0.001
      Analysis of variance (ANOVA).
      Gender: Female, N (%)755 (53.8%)262 (64.1%)180 (46.8%)101 (42.8%)212 (56.7%)<0.001
      Chi square.
      Ethnicity, N (%)
       White/Caucasian1046 (89.6%)360 (88.0%)345 (89.6%)NA341 (91.2%)0.142
      Chi square.
       Latino/Hispanic/Mexican American50 (4.3%)17 (4.2%)15 (3.9%)NA18 (4.8%)
       Black/African American29 (2.5%)17 (4.2%)8 (2.1%)NA4 (1.1%)
       Other (Asian/American Indian/Mixed Race)43 (3.7%)15 (3.7%)17 (4.4%)NA11 (2.9%)
       Missing236 (16.8%)0 (0.0%)0 (0.0%)236 (100%)0 (0.0%)
      Type of diabetes
       Type 1, N (%)713 (50.8%)200 (48.9%)193 (50.1%)120 (50.8%)200 (53.5%)0.630
      Chi square.
       Type 2, N (%)691 (49.2%)209 (51.1%)192 (49.9%)116 (49.2%)174 (46.5%)
      Diabetes duration (years), mean (SD)12.1 (12.4)15.8 (12.7)11.9 (11.2)8.5 (7.7)10.5 (14.4)<0.001
      Analysis of variance (ANOVA).
      Diabetes treatment, N (%)
       Insulin1024 (72.9%)290 (70.9%)279 (72.5%)175 (74.2%)280 (74.9%)0.619
      Chi square.
       Orals380 (27.1%)119 (29.1%)106 (27.5%)61 (25.8%)94 (25.1%)
      Hours worked per week, mean (SD)34.6 (11.8)34.2 (13.4)33.2 (12.2)35.8 (12.1)35.6 (8.7)0.013
      Analysis of variance (ANOVA).
      Status when last NSHE happened, N (%):
       Awake, and at work484 (34.5%)123 (30.1%)111 (28.8%)73 (30.9%)177 (47.3%)<0.001
      Chi square.
       Awake, but not at work687 (48.9%)204 (49.9%)207 (53.8%)129 (48.9%)147 (39.3%)
       During sleep at night233 (16.6%)82 (20.0%)67 (17.4%)233 (16.6%)50 (13.4%)
      Last NSHE was, N (%):
       Symptomatic, confirmed by blood glucose test875 (62.3%)270 (66.0%)238 (61.8%)146 (62.3%)221 (59.1%)0.102
      Chi square.
       Symptomatic, not confirmed by blood glucose test457 (32.5%)112 (27.4%)128 (33.2%)83 (35.2%)134 (35.8%)
       No symptoms, but confirmed by blood glucose test72 (5.1%)27 (6.6%)19 (4.9%)7 (3.0%)19 (5.1%)
      Frequency of NSHE, N (%)
       Type 1
        “Daily” to “about 1/week”365 (51.2%)129 (64.5%)91 (47.2%)52 (43.3%)93 (46.5%)<0.001
      Chi square.
        “1/month” to “several times/month”193 (27.1%)47 (23.5%)54 (28.0%)32 (26.7%)60 (30.0%)
        “Only a few times/year”/“very rarely”155 (21.7%)24 (12.0%)48 (24.9%)36 (21.7%)47 (23.5%)
       Type 2
        “Daily” to “about 1/week”196 (28.4%)52 (24.9%)52 (27.1%)33 (28.4%)59 (33.9%)0.607
      Chi square.
        “1/month” to “several times/month”234 (33.9%)73 (34.9%)69 (35.9%)40 (34.5%)52 (29.9%)
        “Only a few times/year”/“very rarely”261 (37.8%)84 (40.2%)71 (37.0%)43 (37.1%)63 (36.2%)
      NSHE, non-severe hypoglycemic event; SD, standard deviation.
      low asterisk Analysis of variance (ANOVA).
      Chi square.
      Fifty-one percent of individuals with type 1 diabetes (range: 43.3% in Germany, 64.5% in the United States, P < 0.001) and 28.4% of individuals with type 2 diabetes (range: 24.9% in the United States, 30.1% in France, P = 0.607) reported having an NSHE at least weekly. Eighty percent of respondents (n = 1,123) reported having an NSHE within the last 2 weeks. A total of 69.2% (n = 972) respondents reported having an NSHE at work in the past month. A total of 75.0% (n = 1053) reported having an NSHE during the day, but not at work, and 44.2% (n = 620) reported having an NSHE during sleep. These percentage estimates are not mutually exclusive, as some respondents experienced multiple NSHEs. When looking only at the most recent NSHE, 34.0% of respondents reported that the event occurred at work, 48.9% indicated during the day, but not at work, and 16.6% reported during sleep at night.
      Most documented NSHEs (∼≥60%) were symptomatic and confirmed by a blood glucose test. The majority of respondents (94.8%) identified their NSHE through hypoglycemic symptoms (confirmed or unconfirmed by blood glucose test). A small proportion (range 12.5%–17.1%) had no symptoms, but identified their NSHE solely using blood glucose testing. The percent of asymptomatic events per country were: United States – 16.9%, United Kingdom – 12.5%, Germany – 12.7%, and France – 17.1%. The primary self-reported causes for NSHEs were irregular or insufficient food intake or diet (48.2%), physical exercise or overexertion (34.1%), and/or stress (32.0%). Among respondents who reported intentionally trying to lower their blood glucose levels as causative, a miscalculated insulin dose (16.7%) or general poor control (11.4%) were the most commonly reported events.

       Impact of NSHEs on diabetes management

      Respondents reported that their NSHEs lasted an average of 33.0 minutes (SD 60.6). As shown in Table 2, across all respondents the predominant strategy for managing the most recent NSHE was the use of sugar drinks and packs. Approximately one-third of the total sample (range 25% Germany to 43% United States) also ate a meal to deal with their event. In the week following the event, compared with usual blood glucose monitoring practices, respondents conducted an average 5.6 (SD 8.5) extra self-monitoring tests. Additionally, 24.9% of the sample contacted a health-care professional (either primary care physician, hospital, diabetes clinic, or other health-care worker) as a result of the event. Last, among respondents using insulin, 25% reported decreasing their insulin dose following their NSHE. The average decrease was 6.5 (SD 13.1) insulin units for an average of 3.6 (SD 5.1) days.
      Table 2Diabetes management after last NSHE.
      TotalUSAUKGermanyFranceP value
      Used to recover from last NSHE, N (%)
       Glucose tablets or gel295 (21.0%)95 (23.2%)94 (24.4%)44 (18.6%)62 (16.6%)0.028
      Chi square.
       Sugar packs, candy or cake373 (26.6%)84 (20.5%)66 (17.1%)65 (27.5%)158 (42.2%)<0.001
      Chi square.
       Soda, juice, sweet tea or milk421 (30.0%)145 (35.5%)113 (29.4%)66 (28.0%)97 (25.9%)0.026
      Chi square.
       Light meal (e.g. sandwich)324 (23.1%)118 (28.9%)98 (25.5%)45 (19.1%)63 (16.8%)<0.001
      Chi square.
       Full meal (e.g. lunch or dinner)150 (10.7%)57 (13.9%)31 (8.1%)14 (5.9%)48 (12.8%)0.002
      Chi square.
       Other66 (4.7%)28 (6.8%)23 (6.0%)4 (1.7%)11 (2.9%)0.005
      Chi square.
      Number of extra blood sugar tests after last NSHE, mean (SD)
       First day (day of NSHE)1.9 (2.9)1.7 (3.2)1.8 (1.9)1.8 (1.6)2.1 (3.9)0.173
       Second day1.1 (2.5)0.8 (3.7)1.1 (1.7)0.9 (1.4)1.4 (2.1)0.017
       Third day0.8 (2.8)0.7 (4.4)0.8 (1.5)0.7 (1.2)1.1 (2.2)0.251
       Fourth to seventh day2.2 (5.2)1.4 (3.7)2.5 (6.3)1.8 (4.2)3.1 (5.7)<0.001
       Total (7 days after NSHE)5.6 (8.5)3.9 (6.8)6.2 (9.7)5.1 (7.5)7.3 (9.1)<0.001
      Contacted a health-care professional
      Primary care doctor, hospital, diabetes clinic, other.
      after last NSHE
       Yes, N (%)349 (24.9%)56 (13.7%)99 (25.7%)56 (23.7%)138 (36.9%)<0.001
      Chi square.
      Did your LAST NSHE cause you to decrease your normal insulin dose?
       Yes, N (%)351 (25.0%)109 (26.7%)101 (26.2%)33 (14.0%)108 (28.9%)<0.001
      Chi square.
       Total units decreased, mean (SD)6.5 (13.1)6.3 (8.0)6.2 (9.9)6.5 (9.3)6.8 (19.4)0.988
      Analysis of variance (ANOVA).
       Days decreased, mean (SD)3.6 (5.1)4.7 (6.5)3.1 (4.2)3.1 (4.8)3.2 (4.0)0.062
      Analysis of variance (ANOVA).
      NSHE, non-severe hypoglycemic event; SD, standard deviation.
      low asterisk Chi square.
      Primary care doctor, hospital, diabetes clinic, other.
      Analysis of variance (ANOVA).

       Impact on productivity of NSHEs occurring during working hours

      Of the total sample reporting an NSHE at work during the past month (n = 972), 9 patients (4.9%) were removed from this analysis as outlying data (5% upper-limit trim). As shown in Table 3, among the 963 evaluated respondents, 18.3% (n = 176) either left work early or missed a full day. For NSHEs occurring during work hours, the average time lost from work was 9.9 hours (SD 8.4). In addition, 23.8% (n = 231) of respondents reported missing a meeting or work appointment, or not finishing a work task on time. The number of people who missed work or a meeting/appointment due to their NSHE was significantly greater in the UK (23.3%) and France (27.2%) than in Germany (8.8%) or the US (10.8%).
      Table 3Productivity impact of an NSHE.
      TotalUSAUKGermanyFranceP value
      NSHE during working hours, N (with 5% trim)
      The top 5% was trimmed for these results. This dropped 9 patients (1 from the USA, 4 from the UK, 1 from Germany, and 3 from France).
      N = 963N = 278N = 232N = 170N = 283
       Missing work time, N (%)176 (18.3%)30 (10.8%)54 (23.3%)15 (8.8%)77 (27.2%)<0.001
      Chi square.
       If missing work time, what was the amount of work time lost, mean (SD)9.9 (8.4)10.2 (9.5)11.4 (9.2)8.3 (7.1)8.9 (7.6)0.467
      Analysis of variance (ANOVA).
       Missing a meeting/appointment or not finishing a project in due time because of the NSHE, (N) %231 (23.8%)48 (17.2%)68 (28.8%)29 (17.0%)86 (30.1%)<0.001
      Chi square.
      NSHE outside working hours, N (with 5% trim)
      The top 5% was trimmed for these results. This dropped 7 patients (2 from the UK, 3 from Germany, and 2 from France).
      N = 1046N = 307N = 287N = 173N = 279
       Missing work time, N (%)150 (14.3%)26 (8.5%)41 (14.3%)14 (8.1%)69 (24.7%)<0.001
      Chi square.
       If missing work time, what was the amount of work time lost, mean (SD)12.6 (11.0)11.1 (10.5)15.1 (13.6)9.2 (7.4)12.4 (9.9)0.268
      Analysis of variance (ANOVA).
       Missing a meeting/appointment or not finishing a project in due time because of the NSHE, N (%)173 (16.4%)39 (12.7%)49 (17.0%)17 (9.7%)68 (24.2%)<0.001
      Chi square.
      Nocturnal NSHE, N (with 5% trim)
      The top 5% was trimmed for these results. This dropped 8 patients (1 from the USA, 1 from the UK, 2 from Germany, and 4 from France).
      N = 612N = 205N = 153N = 88N = 166
       Missing work time, N (%)139 (22.7%)29 (14.1%)43 (28.1%)14 (15.9%)53 (31.9%)<0.001
      Chi square.
       If missing work time, what was the amount of work time lost, mean (SD)14.7 (11.6)14.3 (11.8)14.2 (10.2)12.5 (12.7)15.9 (12.5)0.787
      Analysis of variance (ANOVA).
       Missing a meeting/appointment or not finishing a project in due time because of the NSHE, N (%)197 (31.8%)48 (23.3%)62 (40.3%)20 (22.2%)67 (39.4%)<0.001
      Chi square.
      NSHE, non-severe hypoglycemic events.
      low asterisk The top 5% was trimmed for these results. This dropped 9 patients (1 from the USA, 4 from the UK, 1 from Germany, and 3 from France).
      The top 5% was trimmed for these results. This dropped 7 patients (2 from the UK, 3 from Germany, and 2 from France).
      The top 5% was trimmed for these results. This dropped 8 patients (1 from the USA, 1 from the UK, 2 from Germany, and 4 from France).
      § Chi square.
      Analysis of variance (ANOVA).

       Impact on productivity of NSHEs occurring outside working hours

      NSHEs that occurred outside working hours also had an impact on respondents' work productivity, and resulted in both absenteeism and reduced productivity while at work. Among respondents experiencing a nocturnal NSHE, 22.7% (139 of 612) arrived late for work or missed a full day of work (Table 3). For those reporting missed work, an average of 14.7 working hours (SD 11.6) were lost (after applying a 5% trim of upper outliers). In addition, 31.8% (197 of 612) of respondents reported that they missed a meeting or work appointment or did not finish a work task on time due to the nocturnal NSHE.
      Among respondents having an NSHE outside working hours, but not during sleep, 14.3% (150 of 1046) reported absenteeism from work. For respondents who reported missed work, an average of 12.6 (SD 11.0) working hours were lost (using a 5% trim). Moreover, 16.4% (173 of 1046) reported that they missed a meeting or work appointment or did not finish a work task on time. No significant differences related to NSHE timing and work time lost were observed between any countries.

       Economic burden of NSHEs

      The human capital method, which uses average wages to estimate productivity, was applied to draw cost estimates for NSHEs [
      • Van den Hout W.B.
      The value of productivity: human-capital versus friction-cost method.
      ]. Based on respondents' self-reported working hours, an average of 35 working hours per week over an estimated 47 working weeks per year (a total of 1,645 working hours per year) was used for the calculations. The 2009 gross domestic product per capita [
      Central Intelligence Agency (CIA)
      The World Factbook: United Kingdom [Internet].
      ] was used to estimate annual income. An average income of $28.2 USD/hour in the United States, $21.5 USD/hour in the United Kingdom, $19.9 USD/hour in France, and $20.8 USD/hour in Germany [
      Central Intelligence Agency (CIA)
      The World Factbook: United Kingdom [Internet].
      ] were used to estimate the value of lost productivity. The conversion of Euros and British Pounds to US dollars was based on the April 9, 2010 conversion rate (1 € to 0.747 USD, 1£ to 1.54 USD).
      As shown in Table 4, the estimated productivity loss per NSHE due to absenteeism ranged from $26.43 to $55.16 (USD) in the United States, $46.30 to $83.59 (USD) in the United Kingdom, $15.26 to $35.58 (USD) in Germany, and $48.33 to $93.47 (USD) in France. These estimates were calculated based on the proportion of respondents reporting missed work, multiplied by hourly income and hours missed. For example, for an NSHE occurring during working hours in the US, 10.8% of the sample reported missing an average of 10.2 hours of work at a cost of $28.2/hour (equivalent to approximately $31.12 in productivity loss per NSHE). An estimation of the yearly costs for these events, based on the number of reported events per month, is $2,293.81 (range $1,939.06–$2,986.28). Overall, NSHEs were most costly in France, and in each country, observed costs associated with lost productivity were the highest for NSHEs that took place during sleep (i.e., nocturnal hypoglycemia).
      Table 4Productivity loss
      Costs provided in US dollars (USD).
      of an NSHE.
      USAUKGermanyFranceP value
      NSHE outside working hours $, (SD), N
      • 26.43 (121.26)
      • N = 307
      • 46.30 (157.60)
      • N = 287
      • 15.50 (67.24)
      • N = 173
      • 61.12 (144.41)
      • N = 279
      <0.001
      NSHE at work $, (SD), N
      • 31.12 (124.91)
      • N = 278
      • 57.21 (140.51)
      • N = 232
      • 15.26 (65.16)
      • N = 170
      • 48.33 (111.58)
      • N = 283
      <0.001
      NSHE at sleep at night $, (SD), N
      • 55.16 (184.17)
      • N = 205
      • 83.59 (177.30)
      • N = 153
      • 35.58 (130.27)
      • N = 88
      • 93.47 (197.62)
      • N = 166
      0.002
      NSHE, non-severe hypoglycemic event.
      low asterisk Costs provided in US dollars (USD).

       Out-of-pocket costs

      Respondents also reported their out-of-pocket costs required to manage or prepare for NSHEs (Table 5). The total average monthly out-of-pocket cost was $25.29/month, with US patients spending the most ($35.36) and the UK patients spending the least ($16.94). The largest expenses were extra/special groceries, extra test strips and lancets, and transportation services.
      Table 5Monthly out-of-pocket costs.
      TotalUSAUKGermanyFranceP value
      Total out of pocket costs, $,
      Costs provided in US dollars (USD).
      (SD), N
      • $ 25.29 (30.74)
      • N = 1044
      • $ 35.56 (37.04)
      • N = 298
      • $ 16.94 (23.36)
      • N = 299
      • $ 22.71 (25.50)
      • N = 178
      • $ 24.91 (30.37)
      • N = 269
      <0.001
       Extra/special groceries
      • $ 21.48 (21.84)
      • N = 375
      • $ 32.94 (28.25)
      • N = 104
      • $ 12.38 (12.45)
      • N = 121
      • $ 19.59 (16.41)
      • N = 51
      • $ 21.55 (20.30)
      • N = 99
      <0.001
       Food or drink from vending machines or café/restaurants
      • $ 7.29 (7.75)
      • N = 609
      • $ 7.83 (8.44)
      • N = 183
      • $ 5.33 (6.47)
      • N = 195
      • $ 7.70 (7.52)
      • N = 93
      • $ 9.09 (8.12)
      • N = 138
      <0.001
       Glucose products such as tablets, gum, gel tubes, etc.
      • $ 8.24 (8.44)
      • N = 513
      • $ 8.86 (7.71)
      • N = 159
      • $ 5.79 (8.33)
      • N = 165
      • $ 8.46 (7.53)
      • N = 85
      • $ 11.01 (9.38)
      • N = 104
      <0.001
       Extra test strips, lancets
      • $ 17.23 (19.51)
      • N = 385
      • $ 23.58 (22.40)
      • N = 159
      • $ 8.01 (12.50)
      • N = 85
      • $ 17.81 (17.79)
      • N = 59
      • $ 14.07 (16.09)
      • N = 82
      <0.001
       Using a taxi, bus or other transportation
      • $ 19.43 (19.10)
      • N = 143
      • $ 35.83 (23.32)
      • N = 12
      • $ 11.97 (13.13)
      • N = 67
      • $ 20.95 (18.26)
      • N = 26
      • $ 26.39 (21.59)
      • N = 38
      <0.001
       Other (specify)
      • $ 8.33 (11.95)
      • N = 32
      • $ 12.50 (15.16)
      • N = 12
      • $ 7.53 (11.38)
      • N = 12
      • $ 2.69 (2.72)
      • N = 5
      • $ 4.23 (3.02)
      • N = 3
      0.410
      low asterisk Costs provided in US dollars (USD).

      Discussion

      This study confirms prior research indicating that mild to moderate hypoglycemic episodes are common in patients with both type 1 and type 2 diabetes [
      • Miller C.D.
      • Phillips L.S.
      • Ziemer D.C.
      • et al.
      Hypoglycemia in patients with type 2 diabetes mellitus.
      ,
      • Vexiau P.
      • Mavros P.
      • Krishnarajah G.
      • et al.
      Hypoglycaemia in patients with type 2 diabetes treated with a combination of metformin and sulphonylurea therapy in France.
      ,
      • Murata G.H.
      • Duckworth W.C.
      • Shah J.H.
      • et al.
      Hypoglycemia in stable, insulin-treated veterans with type 2 diabetes: a prospective study of 1662 episodes.
      ,
      • Donnelly L.A.
      • Morris A.D.
      • Frier B.M.
      • et al.
      DARTS/MEMO collaboration
      Frequency and predictors of hypoglycaemia in type 1 and insulin-treated type 2 diabetes: a population-based study.
      ], and are associated with substantial additional costs that are borne by patients and/or payers [
      • Harris S.B.
      • Leiter L.A.
      • Yale J.F.
      • et al.
      Out of pocket costs of managing hypoglycemia and hypoglycemia in patients with type 1 diabetes and insulin-treated type 2 diabetes.
      ,
      • Jonsson L.
      • Bolinder B.
      • Lundkvist J.
      Cost of hypoglycemia in patients with type 2 diabetes in Sweden.
      ]. The current results also validate existing data indicating that patients who have experienced NSHEs fear future hypoglycemic events [
      • Wild D.
      • von Maltzahn R.
      • Brohan E.
      • et al.
      A critical review of the literature on fear of hypoglycemia in diabetes: Implications for diabetes management and patient education.
      ,
      • Leiter L.A.
      • Yale J.F.
      • Chiasson J.L.
      • et al.
      Assessment of the impact of fear of hypoglycemic episodes on glycemic and hypoglycemia management.
      ], and that NSHEs are associated with patient-initiated changes to diabetes self-management. These include an increased frequency of SMBG (self-monitoring of blood glucose) testing and greater use of blood glucose strips [
      • Hansen M.V.
      • Pedersen-Bjergaard U.
      • Heller S.R.
      • et al.
      Frequency and motives of blood glucose self-monitoring in type 1 diabetes.
      ,
      • Farmer A.
      • Balman E.
      • Gadsby R.
      • et al.
      Frequency of self-monitoring of blood glucose in patients with type 2 diabetes: association with hypoglycaemic events.
      ], as well as adjustments to insulin dose and/or increased food consumption [
      • Leiter L.A.
      • Yale J.F.
      • Chiasson J.L.
      • et al.
      Assessment of the impact of fear of hypoglycemic episodes on glycemic and hypoglycemia management.
      ]. In this study, across all four countries, costs for extra blood glucose strips, as well as patient out-of-pocket expenses for food items to control NSHEs, ranged from ∼$17 to $35 USD. The number of extra blood glucose tests per NSHE was estimated to range from 3.9 to 7.3.
      The current research also confirms limited, existing data indicating that NSHEs have a substantial effect on productivity [
      • Davis R.E.
      • Morrissey M.
      • Peters J.R.
      • et al.
      Impact of hypoglycaemia on quality of life and productivity in type 1 and type 2 diabetes.
      ,
      • Leiter L.A.
      • Yale J.F.
      • Chiasson J.L.
      • et al.
      Assessment of the impact of fear of hypoglycemic episodes on glycemic and hypoglycemia management.
      ]. Across the four countries evaluated, a high proportion of respondents reported missed work time due to NSHEs (mean hours lost per incident ranged from 8.3 to 15.9). Using a human capital approach, the lost productivity per NSHE was estimated to range from $15.26 to $93.47 USD. Of interest, nocturnal hypoglycemia, the least-frequently reported NSHE type (occurring in 44.2% of respondents vs. 70.0%–75.0% of respondents reporting daytime events), was associated with the highest per-incident work time lost and costs for lost productivity.
      Compared with existing research, the current findings appear to contradict only those obtained by Leckie and colleagues [
      • Leckie A.M.
      • Graham M.K.
      • Grant J.B.
      • et al.
      Frequency, severity, and morbidity of hypoglycemia occurring in the workplace in people with insulin-treated diabetes.
      ] in a 2005, 12-month, UK based prospective evaluation of 243 employed individuals with insulin-treated diabetes. In this study, among patients who experienced mild (defined as self-treated) hypoglycemia at work (563 events), only 19 individuals reported requiring time off from work (mean time off: 12.8 minutes; maximum time off: 30 minutes). The definitions applied in this study to distinguish mild from severe hypoglycemia, however, were not as stringent as those used in the current research, and moderate hypoglycemia was not defined at all. Furthermore, this study only evaluated the productivity impact of hypoglycemic events that took place at work.
      As with all research, this study was not able to address or answer the wide range of important issues regarding the impact of NSHEs. We suggest that future research in this area is needed. As reported here, there are country differences in important demographic factors that may impact outcomes. For example, do patients with diabetes for a longer duration of time adjust their management differently than those newly diagnosed? Are patients using insulin pumps more or less impacted by NSHEs? Additional research is also needed to better understand the impact of sociocultural factors and country differences which were identified in this study. It is of note that the productivity impact on employers of NSHEs during working hours varies by country, but the impact of NSHEs outside working hours does not. This suggests that the influence of country-specific work place ethics and standards warrants greater examination. Can the finding that the impact of work place NSHEs is greater in the United Kingdom and France than in the United States and Germany be due to a greater tolerance for work place absence in the United Kingdom and France? Further, what is the impact of nocturnal NSHEs, as lost productivity costs were the highest for events occurring while asleep in every country? The burden of illness for nocturnal NSHEs warrants considerable focus in future research as does the clinical implications and impact of reduced insulin use following an NSHE on the course of disease and treatment compliance. The clinical implication of reduced insulin use following an NSHE by a quarter of patients suggests that patient's trade-off adequate glycemic control to avoid future hypoglycemic events. Inclusion of this information in both patient diabetes education and discussion between patients and their physicians may be beneficial.
      Several limitations of this study should be mentioned. First, since recall bias can influence findings, accuracy of reporting is a consideration with any survey research project. A 1991 study comparing prospective and retrospective recordings of hypoglycemic episodes, however, found no statistically significant difference between the accuracy of prospectively and retrospectively recorded recordings, and concluded that recall of up to 1 week could be considered relatively accurate [
      • Pramming S.
      • Thorsteinsson B.
      • Bendtson I.
      • Binder C.
      Symptomatic hypoglycaemia in 411 type 1 diabetic patients.
      ]. In the case of this study, based on findings obtained from focus groups conducted prior to survey implementation, recall was posited to be accurate for up to 1 month.
      It is also feasible that data collection via an Internet-based survey could introduce selection bias, since only literate respondents with computer access were able to participate. The literacy rates and proportion of Internet users in all four sampled countries, however, are high (e.g., the United Kingdom has a 99% literacy rate and 48.7 million computer users out of 61.1 million inhabitants) [
      Central Intelligence Agency (CIA)
      The World Factbook: United Kingdom [Internet].
      ]. Accuracy could also be affected by the incentive administered to respondents for completion of the survey. However, the amount of the incentive was minimal (approximately $10 USD) and should not have affected participant response. In addition, because all countries that participated in the study were in Western Europe or North America, it is not certain whether a similar study, conducted in countries with different cultures or diabetes management systems, would yield the same results.
      Another consideration is that this survey was not designed to adequately capture the extent of respondent contact with health-care professionals following an NSHE. For example, the survey did not distinguish between actual visits and telephone consultations, and respondents could report both scheduled and unscheduled contacts (although only unscheduled visits would incur extra costs). If available, such data might add to the cost equation identified by this study, and future research should be considered to examine the increased costs associated with contacting health professionals following an NSHE. Also, given the panel nature of the survey it was not possible to have a physician confirmed diagnosis. However, it was not known to the patients who completed the screener beforehand that only those with diabetes would be administered the survey. In the screener, the subjects were provided with several medical conditions and asked to check which they had been diagnosed with by a physician. Only those who checked diabetes, among the multiple possibilities, were invited to complete the full survey. Additionally, only a small incentive was given to complete the survey. Given these safeguards, we believe that the incidence of misrepresentation of diagnosis was unlikely and that this group was not large enough to influence findings.
      Lastly, the current study may underestimate the full extent of the burden of NSHEs on work productivity from the patient perspective because NSHEs not only affects absenteeism, but the ability to be productive while at work (presenteeism), and the effects of presenteeism were not completely captured with the current study design. The extrapolated yearly costs for these events were in the range of $1,939.06 to $2,986.28 USD per patient. The implications for the impact of these costs should be evaluated within the context of the costs of goods and services and purchasing power within each country.
      From the employer perspective, these costs may be overestimated as the human capital cost approach does not take into consideration the ability of the employer to have another employee accomplish the work [
      • Koopmanschap M.A.
      • Rutten F.F.
      • van Ineveld B.M.
      • van Roijen L.
      The friction cost method for measuring indirect costs of disease.
      ]. Calculations using the friction cost method in future studies would be of value to address this issue.
      The goal of new drugs and devices for diabetes treatment is to improve glycemic control and reduce the frequency of adverse events such as hypoglycemia [
      American Diabetes Association Workgroup on Hypoglycemia
      Defining and reporting hypoglycemia in diabetes: a report from the American Diabetes Association Workgroup on Hypoglycemia.
      ]. Clearly, addressing NSHEs should be a factor considered in this goal. Discussing NSHEs at regular diabetes health care visits, and reviewing with patients how to recognize NSHEs and appropriately treat and deal with their sequelae would be a beneficial addition to routine care and diabetes management. In addition, antidiabetic therapies that are not associated with additive hypoglycemic risk could have a substantial impact on the individual burden imposed by such events.
      As our survey indicates, NSHEs are associated with measurable productivity and financial consequences for patients and employers. Diabetes management strategies and/or treatments that reduce NSHEs may have a major impact on reducing costs of care and lost work productivity while increasing psychological health, and daily functioning in individuals with diabetes.
      Source of financial support: This study was funded by Novo Nordisk.

      Supplementary material

      References

        • Hepburn D.A.
        Symptoms of hypoglycemia.
        in: Frier B.M. Hypoglycemia and Diabetes: Clinical and Physiological Aspects. Edward Arnold, London1993
        • American Diabetes Association Workgroup on Hypoglycemia
        Defining and reporting hypoglycemia in diabetes: a report from the American Diabetes Association Workgroup on Hypoglycemia.
        Diabetes Care. 2005; 28: 1245-1249
        • Miller C.D.
        • Phillips L.S.
        • Ziemer D.C.
        • et al.
        Hypoglycemia in patients with type 2 diabetes mellitus.
        Arch Intern Med. 2001; 161: 1653-1659
        • Vexiau P.
        • Mavros P.
        • Krishnarajah G.
        • et al.
        Hypoglycaemia in patients with type 2 diabetes treated with a combination of metformin and sulphonylurea therapy in France.
        Diabetes Obes Metab. 2008; 10: 16-24
        • Davis R.E.
        • Morrissey M.
        • Peters J.R.
        • et al.
        Impact of hypoglycaemia on quality of life and productivity in type 1 and type 2 diabetes.
        Curr Med Res Opin. 2005; 21: 1477-1483
        • Pramming S.
        • Thorsteinsson B.
        • Bendtson I.
        • Binder C.
        Symptomatic hypoglycaemia in 411 type 1 diabetic patients.
        Diabet Med. 1991; 8: 217-222
        • Harris S.B.
        • Leiter L.A.
        • Yale J.F.
        • et al.
        Out of pocket costs of managing hypoglycemia and hypoglycemia in patients with type 1 diabetes and insulin-treated type 2 diabetes.
        Can J Diabetes. 2007; 31: 25-33
        • Jermendy G.
        • Erdesz D.
        • Nagy L.
        • et al.
        Outcomes of adding second hypoglycemic drug after metformin monotherapy failure among type 2 diabetes in Hungary.
        Health Qual Life Outcomes. 2008; 6: 88
        • Diabetes UK
        Survey reveals hidden incidence of 'hypos' among people with type 2 diabetes [Internet].
        ([Accessed: January 13, 2010])
        • Levy A.R.
        • Christensen T.L.
        • Johnson J.A.
        Utility values for symptomatic non-severe hypoglycaemia elicited from persons with and without diabetes in Canada and the United Kingdom.
        Health Qual Life Outcomes. 2008; 6: 73
        • Currie C.J.
        • Morgan C.L.
        • Poole C.D.
        • et al.
        Multivariate models of health-related utility and the fear of hypoglycaemia in people with diabetes.
        Curr Med Res Opin. 2006; 22: 1523-1534
        • Frier B.M.
        How hypoglycaemia can affect the life of a person with diabetes.
        Diabetes Metab Res Rev. 2008; 24: 87-92
        • Hauber A.B.
        • Mohamed A.F.
        • Johnson F.R.
        • Falvey H.
        Treatment preferences and medication adherence of people with type 2 diabetes using oral glucose-lowering agents.
        Diabet Med. 2009; 26: 416-424
        • Wild D.
        • von Maltzahn R.
        • Brohan E.
        • et al.
        A critical review of the literature on fear of hypoglycemia in diabetes: Implications for diabetes management and patient education.
        Patient Educ Couns. 2007; 68: 10-15
        • Shiu A.T.
        • Wong R.Y.
        Fear of hypoglycaemia among insulin-treated Hong Kong Chinese patients: implications for diabetes patient education.
        Patient Educ Couns. 2000; 41: 251-261
        • Hansen M.V.
        • Pedersen-Bjergaard U.
        • Heller S.R.
        • et al.
        Frequency and motives of blood glucose self-monitoring in type 1 diabetes.
        Diabetes Res Clin Pract. 2009; 85: 183-188
        • Farmer A.
        • Balman E.
        • Gadsby R.
        • et al.
        Frequency of self-monitoring of blood glucose in patients with type 2 diabetes: association with hypoglycaemic events.
        Curr Med Res Opin. 2008; 24: 3097-3104
        • Leckie A.M.
        • Graham M.K.
        • Grant J.B.
        • et al.
        Frequency, severity, and morbidity of hypoglycemia occurring in the workplace in people with insulin-treated diabetes.
        Diabetes Care. 2005; 28: 1333-1338
        • Leiter L.A.
        • Yale J.F.
        • Chiasson J.L.
        • et al.
        Assessment of the impact of fear of hypoglycemic episodes on glycemic and hypoglycemia management.
        Can J Diabetes. 2005; 29: 186-192
        • Wild D.
        • Eremenco S.
        • Mear I.
        • et al.
        Multinational trials—recommendations on the translations required approaches to using the same language in different countries, and the approaches to support pooling the data: the ISPOR patient-reported outcomes translation and linguistic validation good research practices task force report.
        Value Health. 2009; 12: 430-440
        • Wilcox R.R.
        The goals and strategies of robust methods.
        Brit J Math Stat Psychol. 1998; 51: 1-39
        • Van den Hout W.B.
        The value of productivity: human-capital versus friction-cost method.
        Ann Rheum Dis. 2010; 69: i89-i91
        • Central Intelligence Agency (CIA)
        The World Factbook: United Kingdom [Internet].
        CIA, Washington, DC2009 ([Accessed: March 22, 2010])
        • Murata G.H.
        • Duckworth W.C.
        • Shah J.H.
        • et al.
        Hypoglycemia in stable, insulin-treated veterans with type 2 diabetes: a prospective study of 1662 episodes.
        J Diabetes Complications. 2005; 19: 10-17
        • Donnelly L.A.
        • Morris A.D.
        • Frier B.M.
        • et al.
        • DARTS/MEMO collaboration
        Frequency and predictors of hypoglycaemia in type 1 and insulin-treated type 2 diabetes: a population-based study.
        Diabet Med. 2005; 22: 749-755
        • Jonsson L.
        • Bolinder B.
        • Lundkvist J.
        Cost of hypoglycemia in patients with type 2 diabetes in Sweden.
        Value Health. 2006; 9: 193-198
        • Koopmanschap M.A.
        • Rutten F.F.
        • van Ineveld B.M.
        • van Roijen L.
        The friction cost method for measuring indirect costs of disease.
        J Health Econ. 1995; 14: 171-189