If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Address for correspondence: Shawn Bugden, BSc(Pharm), MSc, PharmD, College of Pharmacy, Rady Faculty of Health Sciences, University of Manitoba, 750 McDermot Avenue, Winnipeg, Manitoba R3E 0T5, Canada.
Considerable attention has been paid to the rising costs of the use of blood glucose test strips (BGTS). Insulin users have generally been treated as a single homogeneous group, resulting in policies that cap usage (8.2 strips/day) in provincial drug insurance programs. The objective of this study was to conduct a utilization review of BGTS by insulin users and to evaluate use patterns against current insulin use patterns and BGTS policy.
Methods
BGTS usage was examined in a cohort of insulin users with type 1 and type 2 diabetes over a 12-year period (2001 to 2013) using the population-based administrative data in Manitoba, Canada.
Results
Total BGTS strip use increased by 121%, from $4.3 to $9.5 million. However, the number of insulin users also increased by 115%. Use has been stable at 1.5 strips per day per person since 2004 by insulin users with type 2 diabetes but has risen from 1.9 to 3.0 strips per day per person in those with type 1 diabetes. Mean daily test strip use was below the number of daily tests recommended for patients using insulin as per the current Canadian guidelines, with 11% and 15% of insulin users with type 1 and type 2 diabetes not claiming any BGTS use and a further 15% (type 1) and 28% (type 2) using fewer than 1 strip per day.
Conclusions
BGTS use per insulin user has been stable for most of the past decade, and the vast majority of use falls well below provincial insurance caps. The amount of low-level testing (0 to <1 strip/day) suggests that greater attention should be directed to ensuring a safe level of testing by all insulin users.
Résumé
Objectifs
Les coûts croissants de l'utilisation des bandelettes de mesure de la glycémie ont fortement retenu l'attention. Les patients sous insuline ont en général été considérés comme un seul groupe homogène, d'où des politiques de plafonnement de l'utilisation (8,2 bandelettes par jour) par les programmes provinciaux d'assurance médicaments. L'objectif de cette étude était d'effectuer un examen de l'utilisation des bandelettes de mesure de la glycémie par les patients sous insuline et d'évaluer les profils d'utilisation par rapport aux schémas d'insulinothérapie courants et aux politiques actuelles relatives à ces bandelettes.
Méthodes
L'utilisation des bandelettes de mesure de la glycémie a été évaluée pendant une période de 12 ans (2001 à 2013) au sein d'une cohorte de personnes sous insuline atteintes de diabète de type 1 ou de type 2, à partir des données administratives populationnelles du Manitoba (Canada).
Résultats
Le coût total de l'utilisation des bandelettes de mesure de la glycémie s'est accru de 121 %, passant de 4,3 à 9,5 millions de dollars. Toutefois, le nombre d'utilisateurs d'insuline a aussi augmenté de 115 %. L'utilisation est demeurée stable à 1,5 bandelette par jour par personne depuis 2004 chez les patients sous insuline atteints de diabète de type 2, mais est passée de 1,9 à 3,0 bandelettes par jour par personne chez les patients atteints de diabète de type 1. Une utilisation en deçà des recommandations énoncées dans les lignes directrices a été mise en évidence, puisque 11 % et 15 % des patients sous insuline atteints de diabète de type 1 et de diabète de type 2, respectivement, n'ont utilisé aucune bandelette et que 15 % (type 1) et 28 % (type 2) des autres ont utilisé moins de 1 bandelette par jour.
Conclusions
L'utilisation de bandelettes de mesure de la glycémie par patient sous insuline est demeurée généralement stable au cours des dix dernières années, et était en majeure partie bien inférieure aux plafonds imposés par les régimes provinciaux d'assurance médicaments. La proportion de patients qui vérifient peu leur glycémie (0 à<1 bandelette par jour) donne à penser que plus d'efforts devraient être consentis pour garantir que tous les patients sous insuline vérifient leur glycémie à une fréquence qui garantisse leur sécurité.
). In publicly funded drug programs, a great deal of attention has been paid to the cost of blood glucose test strips (BGTS). In the province of Ontario, BGTS were found to be the third largest “drug” expenditure (
). The value of this expenditure has been questioned. The efficacy of SMBG in people with type 2 diabetes not using insulin is limited, producing only modest reductions in glycated hemoglobin (A1C) levels (0.3%, 95% CI 0.1 to 0.4) at 6 months and no statistically significant change at 12 months (
Self-monitoring of blood glucose (SMBG) for type 2 diabetes patients treated with oral anti-diabetes drugs and with a recent history of monitoring: Cost-effectiveness in the US.
). In response, a number of Canadian provinces (British Columbia, Saskatchewan and Ontario) and Health Canada's Non-Insured Health Benefits Program have implemented caps on the number of BGTS they will cover: 3000 strips per year for insulin users, 400 strips per year for users of oral medications that may cause hypoglycemia, 200 strips per year for all others (
). Ontario is expected to save more than $100 million over the next 5 years by implementing these caps, with the vast majority of savings coming from people with diabetes who are not using insulin (
Intensified blood glucose monitoring improves glycemic control in stable, insulin-treated veterans with type 2 diabetes: The Diabetes Outcomes in Veterans Study (DOVES).
Pre- and post-prandial capillary glucose self-monitoring achieves better glycaemic control than pre-prandial only monitoring: A study in insulin-treated diabetic patients.
Evidence of a strong association between frequency of self-monitoring of blood glucose and hemoglobin A1c levels in T1D exchange clinic registry participants.
). The Canadian Diabetes Association 2013 Clinical Practice Guidelines for the Prevention and Management of Diabetes in Canada endorse individualized SMBG based on patient-specific characteristics, with a recommended frequency of at least 3 times per day (including pre- and postprandial measurements) for individuals using insulin more than once a day and at least once a day for those with type 2 diabetes on once-daily insulin (
). Clearly, a great deal of variety exists in the intensity of management of insulin users; however, a single current cap of 3000 strips per year (approximately 8 strips/day) applies to all insulin users. The majority of use and the majority of growth in BGTS use has been with those who use insulin, while the expenditures on BGTS in people who do not use insulin has stabilized since 2010 (
). Given the level of use, the increase of BGTS expenditure for insulin users and the heterogeneity of SMBG needs of these people, greater evaluation of the utilization of BGTS in this group is needed. The objective of this study was to identify and evaluate trends in test-strip utilization in relation to treatment intensity in people with diabetes using insulin in Manitoba.
Methods
A retrospective, longitudinal study using a population-based cohort of people with diabetes using insulin was conducted to examine BGTS use between January 1, 2001, and December 31, 2013, in the province of Manitoba, Canada. Data were obtained through the Manitoba Centre for Health Policy's Population Research Data Repository, which contains administrative healthcare databases. The repository contains linked, deidentified records of virtually every contact between Manitoba residents and the province's universal healthcare system (
). The sources used included the hospital discharge abstracts database, which contains summary records of all hospital stays; and data from the provincial Drug Program Information Network (DPIN), an online, point-of-service prescription database system that connects all community pharmacies to the provincial Pharmacare program and other third-party payers. The Pharmacare program provides full prescription coverage after an annual, income-based deductible has been fulfilled. This program covers BGTS for all users up to a maximum of 4000 strips per year.
Approval for this study was granted by the University of Manitoba Health Research Ethics board and the provincial Health Information Privacy Committee.
BGTS use was evaluated within a cohort of insulin users with type 1 and 2 diabetes. Prescription records for hypoglycemia drugs or insulin, and medical claims or hospital discharge abstracts with International Classification of Disease-9 (ICD-9) or ICD-10 codes for diabetes mellitus were used to assess cohort entry. Persons were categorized as having diabetes, provided they met at least 1 of 3 criteria: 1) 1 or more hospitalizations with a diabetes code, 2) 2 or more medical claims with diabetes codes within a 3-year window, or 3) 1 diabetes medical claim plus a prescription for a diabetes drug. The definition of diabetes has been validated previously and has shown 88.4% sensitivity and 98.8% specificity (
). This cohort was further restricted to those who had received 2 or more prescriptions for insulin within the study period. Individuals were classified as having type 2 diabetes if they had had 1 or more prescriptions for a noninsulin hypoglycemic drug.
Individuals were further classified by the type of insulin regimen, defined by the combination of insulin types they were prescribed within each study year. This was done by first creating 5 categories of insulin preparations based on their pharmacokinetic profiles and identified by their Anatomical Therapeutic Chemical (ATC) classification system code. These groups were 1) rapid-acting insulins: insulin lispro, insulin aspart, insulin glulisine (ATC codes A10AB04, 05, 06); 2) regular (Toronto) insulin (A10AB01); 3) intermediate-acting NPH and Lente insulins (A10AC); 4) premixed insulin combinations containing regular or rapid-acting insulin with a longer-acting insulin (A10AD) and 5) long-acting insulins: ultralente insulin, insulin glargine and insulin detemir (A10AE).
Regimens were then defined using various combinations of these classes and were rank ordered in terms of level of intensity (increasing number of variable dose injections per day) of that regimen (Table 1). This grouping of regimens is adapted from the clinical practice guidelines approaches to treatment and testing (
The quantity of BGTS dispensed for each person within each study year in which they were treated with insulin was then tabulated, and the number of strips per day was calculated by dividing by the days in the year. In cases in which an individual was hospitalized within a particular year, the total days in hospital were subtracted from the days of the year when determining the rate of BGTS use. Person-years with more than 180 hospital days were excluded from the analysis. Persons were also removed from the cohort upon their being admitted to a personal-care or nursing home. In addition for persons dying or leaving the province, the final partial year was removed from the analysis.
Statistical analysis
The total number of test strips dispensed and mean daily test strip use were calculated within each year across all insulin users and were stratified by diabetes type and insulin regimen type. In addition, the number of people within each of these strata was determined and analyzed over time. A Poisson regression model, using the SAS GENMOD procedure (SAS Institute, Cary, North Carolina, United States), was used to evaluate the association between several independent parameters and annual test strip use in insulin users with type 1 or type 2 diabetes (separately). We modelled annual BGTS use, per person, as a count variable using a log-linear model with an AR (
) correlation structure, taking into account the effect of repeated measures in our analysis. Variables included in our final model were age and calendar time, both as continuous variables; and sex and insulin regimen, treated as categorical variables.
Results
Overall test strip use in insulin users
Total BGTS usage increased by 148%, from 4.8 to 11.9 million strips per year over the study period (Figure 1). Total cost increased by 121% from $4.3 million in 2001 to $9.5 million in 2013. However, the number of insulin users also increased by 115% over that same period. As a result, the BGTS use per person has been relatively stable at approximately 1.9 strips per person per day since 2004 (Figure 2).
Figure 1Number of insulin users with type 1 and type 2 diabetes mellitus, by year, in Manitoba, Canada. The red line indicates the total number of blood glucose test strips dispensed to insulin users annually. BGTS, blood glucose test strips.
Test strip use in insulin users with type 2 diabetes
The increase in total use was driven by an increase in insulin users with type 2 diabetes, which increased by 231% from 4295 to 14,225 (Figure 1), resulting in an increase in test strip cost from $1.8 to $6.3 million/year. However, BGTS use per person in this group has been stable at approximately 1.5 strips per person per day since 2004 (Figure 2).
Test strip use in insulin users with type 1 diabetes
The number of insulin users with type 1 diabetes has been relatively stable (Figure 1; 4053 in 2001, 3745 in 2013). The total cost of test strips in this group rose from 2.5 to 3.2 million per year over the study period. In the context of a relatively stable number of insulin users, this appears to be driven by an increase in use of BGTS per person, which has risen from 1.9 to 3.0 strip per day between 2001 and 2013 (Figure 2).
Test strip use and insulin regimen
The average test strip usage varied among insulin regimen groups from approximately 1 to 3 test strips per day, with frequency of use increasing with the intensity of the insulin regimen used (i.e. from basal only to premix to basal/regular-acting bolus to basal/rapid-acting bolus) (Table 2). Persons with type 2 diabetes had lower average test strip use for most insulin regimens, resulting in an overall average of 1 test strip per day fewer compared to people with type 1 diabetes. The basal/rapid-acting insulin regimen represented the largest number of person-years in type 1 diabetes and was the second most common regimen in type 2 diabetes. Here too, people with type 2 diabetes averaged 1.1 test strip fewer per day when compared to people with type 1 diabetes using this regimen.
Table 2Overall mean strip use and daily insulin dose by diabetes type and insulin regimen
The average test strip use per person increased by 1.1 tests strips per day over the study period in people with type 1 diabetes (Figure 2). Test strip usage for each insulin regimen was plotted over time (Figure 3). This figure also superimposes the average test strip use per user by insulin regimen. This allows for a more detailed examination of the patterns of use responsible for the average increase in test strip use. The increasing average level of test strip use was driven primarily by the greater use of the basal/rapid-acting insulin regimen over time. In addition, this regimen was associated with an increase of 0.4 strips per day over the study period. There was also an increase of 0.6 strips per day for the basal/regular-acting insulin regimen but, because of the lower level and declining use of this regime, this increase would have minimal impact on the overall average test strip use (Figure 3).
Figure 3Bubble plot of total blood glucose test strips dispensed (lines), and mean frequency of daily test strip use (bubbles) by insulin regimen. Frequency of use is proportional to the diameter of each bubble.
In insulin users with type 2 diabetes, total annual strip use increased almost 4-fold, but there was no substantive change in the average test strip use per person over the study period (Figure 2). The average test strip use per day was also relatively stable in each of the insulin regimen groups. However, total strip use increased 10-fold in the basal/rapid-acting insulin regimen group over the study period (Figure 3).
Test strip usage and policy limits
This higher level of test strip usage per person in people with type 1 diabetes is also reflected in Figure 4, which illustrates the range of test strip use per day stratified by insulin regimen during the final year of the study period (2013). During the study period, 11% of people with type 1 diabetes and 15% of people with type 2 diabetes using insulin did not have any dispensation claims for test strips. In addition, 15% of people with type 1 diabetes and 28% of people with type 2 diabetes using insulin used fewer than 1 strip per day. The median BGTS usage was 2.7 strips per day in people with type 1 diabetes and 1.1 strips per day in people with type 2 diabetes. The 90th percentile was 6.6 strips per day for people with type 1 diabetes and 3.6 strips per day for people with type 2 diabetes. The current test strip cap used in a number of Canadian provinces of 3000 strips per year corresponds to a use of 8.2 strips per day. Of people with type 1 diabetes, 95%, and of people with type 2 diabetes, 99.6% have usage below this level.
Figure 4Daily test strip use in 2013. Percent of cohort, by diabetes type, by mean daily test strip use, broken down by insulin treatment regimen in the last year of the study period.
A Poisson regression model, adjusting for age and sex, was used to examine the relationship of both insulin treatment regimen and the passage of time on the use of BGTS dispensed by persons with type 1 or 2 diabetes using insulin (Supplementary Table 1). In both models, a higher number of BGTS dispensed in a year was most strongly associated with more intensive insulin regimens. In insulin users with type 1 diabetes, use of basal/rapid-acting insulin regimen was associated with 85% more test strips per year compared to patients with type 1 diabetes using basal insulin only. In persons with type 2 diabetes, basal/rapid insulin regimens were associated with 54% greater use of BGTS than the type 2 diabetes reference group (basal insulin only). After controlling for other variables, we found that calendar time had a significant effect on testing frequency in type 1 diabetes, with BGTS use increasing by 2.5% per year, or 28% over the entire study period. There was no effect of calendar time in type 2 diabetes.
Discussion
Test strip usage by insulin users doubled during the 12-year study period, but the majority of this change can be attributed to the increase in the number of people with type 2 diabetes who use insulin rather than to an increase in use per person. In fact, mean daily test strip use per person for people with type 2 diabetes using insulin has been stable at 1.9 strips per day since 2004. Increases in mean strip use per day from 1.9 to 3.0 strips per day in people with type 1 diabetes is explained largely by changes to more intensive insulin, specifically basal/rapid-acting insulin regimens. This shift to basal/rapid-acting insulin regimens is also seen in those with type 2 diabetes but is less pronounced, with the majority maintained on NPH insulin. Even when basal/rapid insulin regimens were used in type 2 diabetes they were associated with use of fewer test strips (2.05 strips/day) than in type 1 diabetes (3.16 strips/day). Despite the emphasis on high levels of test strip use in the literature (
), this detailed examination of insulin users revealed considerable potential underutilization. The study showed approximately 1 in 7 insulin users did not test at all, and an additional 1 in 4 tested less often than once per day. The vast majority of insulin users had a level of test strip use far below the 3000 annual test strip limit used in some provinces and virtually no one exceeded Manitoba's limit of 4000 strips per year.
Most previous analyses of this issue did not use population-level data and did not differentiate between type 1 and type 2 diabetes. Yeaw et al made use of a sample of private insurance and limited public coverage data and found an overall test strip usage of 3.0 strips per day and higher use with basal/bolus regimens (3.6 strips/day) for a study period from 2007 to 2009 (
). Using 2006 data from private insurance plans, the Canadian Optimal Medication Prescribing and Utilization Service (COMPUS) found an average usage of 3.18 strips per day in insulin-only users (all ages, all regimens, type 1 and type 2 diabetes combined) and 1.90 strips per day in people who used insulin along with oral antidiabetes drugs. More recently, Tavares et al used data (2006 to 2012) from private drug plans combined with Ontario Public Drug Plans (>65 years of age and/or on social assistance/disability) to conduct a more detailed analysis of test strip utilization in type 2 diabetes (
). This study found a modestly higher level of utilization, but the same pattern of increased testing with increased intensity of insulin regimen (basal 1.51 strips/day, premixed insulin 1.84 strips/day, basal/bolus 2.60 strips/day).
The current Canadian Diabetes Association 2013 Clinical Practice Guidelines for the Prevention and Management of Diabetes in Canada recommend that individuals using once-daily insulin test at least once a day. It is recommended that more frequent insulin administration associated with more intensive regimens requires more frequent testing, with a testing frequency of at least 3 times daily (
). Based on these guidelines, there appears to be considerable underutilization, with 14% and 16% of the type 1 and type 2 insulin-using population not testing at all and 21% of type 1 and 31% of the type 2 insulin-using population testing less often than once per day. This population of low test strip users contained people using all types of insulin regimens, including the basal/rapid-acting insulin regimen.
Particularly for people with type 2 diabetes, there is some controversy regarding the maximal level of testing and a general agreement that testing should be optimized on an individual basis. The Canadian Optimal Medication Prescribing and Utilization Service suggested that 2 strips per day would be the maximum necessary for most adults with type 2 diabetes using insulin (
). Current coverage maximums of 3000 strips per year (Saskatchewan, British Columbia, Ontario; 8.2 strips/day) or Manitoba's 4000 strips per year (11.0 strips/day) exceed the current use requirements for virtually all people with type 2 diabetes using insulin and most people with type 1 diabetes, even for the most intensive regimens (basal/rapid) and an average use of 3.16 strips per day for people with type 1 diabetes and 2.05 strips per day by people with type 2 diabetes using insulin.
Current coverage limits (3000 to 4000 strips/year) treat all insulin use as equal and vastly exceed the usage requirements of the average insulin user, suggesting that current coverage policy seems to have been designed to include the most extreme user without regard to guideline suggestions, diabetes type or insulin regimen used. A more effective policy may strongly link to guideline recommendations and be informed by reasonable usage patterns. A 50% reduction of the cap for insulin users with type 2 diabetes to 1500 strips per year (
) would allow for testing 4 times per day and include more than 95% of the population, with special authorization required to justify usage above this level.
Results
This study has strengths and limitations worthy of discussion. The most important strength is the use of full population data, with all ages included. Past studies have been limited by selected populations (>65 years of age), insurance coverage (private insurance) or inferred populations from a selected sample (
The relationship between type of drug therapy and blood glucose self-monitoring test strips claimed by beneficiaries of the Seniors' Pharmacare Program in Nova Scotia, Canada.
). This study's focus on insulin users over an extended period of time (13 years) is also a strength. However, like other administrative data studies, this study is limited by reliance on pharmacy-dispensing claims. Although test strips dispensed might imply test strips used, there is no assurance of actual utilization. Insulin and test strips are available without prescriptions, so there is the potential for missing data. In Manitoba, these missing data are expected to be minimal because access to universal drug coverage requires that all use be entered into the drug information system. The classification of type 2 diabetes was based on the use of noninsulin hypoglycemic drugs. Early in the study period, a minimum of 1-year look-back was used to assess prior noninsulin hypoglycemic drugs. A small amount of misclassification of diabetes type may have occurred by using this pragmatic approach. The study assigned an insulin regimen based on the products used in a year. This may result in some misclassification because people may change insulin regimens during the year. Given the periodic purchase of test strips and the uncertainty of the actual insulin regimen based on pharmacy dispensing claims, a more granular approach to assigning time was felt to be equally problematic. Last, it is not possible to determine how many doses of insulin are given per day, which may be a stronger indicator of intensity and requirement for blood glucose testing.
Conclusions
Test strip use doubled during the study, but the majority of the increase was related to the increasing number of insulin users who had type 2 diabetes. The average test strip use per person has remained relatively stable since 2004. There has been an increased use per person in type 1 diabetes, but this is explained by an increase in use of basal/rapid-acting regimens that require more frequent insulin administration. The vast majority of insulin users have test strip use requirements far below current coverage limits. Hidden within the population level numbers that indicate increased use is a sizable portion of insulin users who are not testing or are testing below levels recommended by practice guidelines. Future efforts should perhaps be directed at the more than one third of the people using insulin who appear to be undertesting, so as to determine the risks and outcomes associated with this practice.
Acknowledgments
This study was funded by a grant from the University of Manitoba Project #45526. The authors acknowledge the Manitoba Centre for Health Policy for use of data contained in the Population Health Research Data Repository under project HIPC#2015/2016-29. The results and conclusions are those of the authors, and no official endorsement by the Manitoba Centre for Health Policy, Manitoba Health, Seniors and Active Living or other data providers is intended or should be inferred.
Appendix
Supplementary Table. Multiple variable regression models of daily test strip use. Shown are the best-fit models for insulin users with type 1 and 2 diabetes (separately).
Type 1 Diabetes.
Tabled
1
Model: Strips/day = Intercept + Year + Age + Sex + Basal-rapid + Basal-regular + Other + Premix
Model: Strips/day = Intercept + Year + Age + Sex + Basal-rapid + Basal-regular + Other + Premix
Source
Degrees of freedom
SS
MS
F-value
p value
Model
7
26 986
3855
1894
<0.0001
Error
120 721
245 671
2.04
Total
120 728
272 657
Parameter
Estimate
Std err
t-value
p value
Intercept
0.5
2.26
0.21
0.83
Year
0.0003
0.001
–0.30
0.7700
Insulin regimen
Basal-rapid
1.09
0.011
96.4
<0.0001
Basal-regular
0.45
0.016
29.1
<0.0001
Other
0.98
0.031
31.9
<0.0001
Premix
0.24
0.011
22.1
<0.0001
Basal
(ref)
age
0.02
0.0003
67.9
<0.0001
Sex (female)
0.06
0.008
7.4
<0.0001
Sex (male)
(ref)
MS, mean square; SS, sum of squares.
Note: Models using additional parameters, including insulin dose and duration of insulin use, were also explored but discarded due to lack of meaningful predictive value.
Self-monitoring of blood glucose (SMBG) for type 2 diabetes patients treated with oral anti-diabetes drugs and with a recent history of monitoring: Cost-effectiveness in the US.
Intensified blood glucose monitoring improves glycemic control in stable, insulin-treated veterans with type 2 diabetes: The Diabetes Outcomes in Veterans Study (DOVES).
Pre- and post-prandial capillary glucose self-monitoring achieves better glycaemic control than pre-prandial only monitoring: A study in insulin-treated diabetic patients.
Evidence of a strong association between frequency of self-monitoring of blood glucose and hemoglobin A1c levels in T1D exchange clinic registry participants.
The relationship between type of drug therapy and blood glucose self-monitoring test strips claimed by beneficiaries of the Seniors' Pharmacare Program in Nova Scotia, Canada.
30560-3&id=jcjd819-fig-0001.jpg){kind=link}
30560-3&id=jcjd819-fig-0002.jpg){kind=link}
30560-3&id=jcjd819-fig-0003.jpg){kind=link}
30560-3&id=jcjd819-fig-0004.jpg){kind=link}