Changes in the consumption of blood sugar test strips among patients in Germany who use FGM systems

In Germany, around 7.5 million people live with diabetes (IDF2017) [1]. A total of 95% of them have type 2 diabetes (T2D) and 5% have diabetes type 1 (T1D) or another type according to the diabetes classification [2]. In many cases, oral antidiabetic drugs for T2D or insulin for T1D and T2D can be used very successfully [3-4]. The goal of diabetes management is always to achieve and maintain good blood sugar control. As recommended by the American Diabetes Association, a glucose level of 70-180 mg/dl (3.8-10 mmol/l) should be achieved. This value can of course be adjusted individually as required. The second goal is to prevent hypoglycemia (<70 mg/dl or
<3.8 mmol/l) and hyperglycemia (>180 mg/dl or
10 mmol/l) should be avoided as far as possible. The long-term goal of diabetes treatment is the prevention of diabetes complications such as retinopathy, nephropathy and neuropathy [5-10]. The biggest burden in the daily life of a patient with diabetes is monitoring blood sugar. Many fingertip blood samples are necessary every day in order to be able to adjust the individual insulin regimen and achieve good blood sugar control. In order to achieve optimal blood sugar control with intensive therapy, frequent blood sugar measurements are necessary (self-monitoring of blood sugar levels (SMBG)). Unfortunately, capillary blood sampling is not able to adequately prevent hypoglycemia and hyperglycemia. Hypoglycemia and hyperglycemia affect quality of life, and hypoglycemia in particular is a major burden in living with diabetes [11]. Blood glucose testing is very time-consuming and painful and many factors must be taken into account. In addition, capillary blood glucose measurements only represent a snapshot of the current glucose level. Flash Glucose Monitoring (FGM) has been available in Germany since 2014 and is reimbursed by health insurance companies for adults and children aged 4 and over with T1D and T2D.

>> Flash Glucose Monitoring (FGM) is a very easy-to-use, painless and user-friendly method for bloodless monitoring of blood sugar levels. A small sensor is inserted under the skin on one arm and remains there for 14 days. The patient can insert the sensor themselves and replace it with a new sensor when the current sensor has expired. The current interstitial glucose value is determined by the patient passing the FGM reader over the sensor. This can be done as often and in any situation and is very discreet and quick. The current value is displayed on the reader screen. FGM systems provide a greater number of glucose readings so that a glucose curve can be created over a 24-hour period, which, along with the rate and direction of blood glucose changes, helps the patient reduce glucose variability and increase time in target range [12 ].
The FGM system works without calibration, meaning no regular capillary blood testing is required to keep the system functional. This can be an advantage or disadvantage depending on the accuracy of the sensor. The CGM systems available on the German market could facilitate diabetes management, support patients in their everyday lives by reducing their burden, and help them achieve and maintain good blood sugar control [13].
However, self-management of blood sugar levels is not yet completely replaced by FGM or other CGM (Continuous Glucose Monitoring) systems.
To date, no real-world data has been published on the use and number of blood glucose test strips (BZTS) after FGM initiation. Therefore, the aim of this study was to compare the daily consumption of BZTS before and after the start of FGM use in T1D and T2D patients.
material and methods
Database
The present study is based on data from the Disease Analyzer database (IQVIA). This collects information on medication prescriptions and diagnoses as well as basic medical and demographic data, which is provided directly and in anonymized form by the computer systems in general and specialist practices [14]. The database includes approximately 3% of all outpatient practices in Germany. Diagnoses (according to the International Classification of Diseases, 10th Revision [ICD-10]), prescriptions (according to the Anatomical Therapeutic Chemical [ATC] classification system), and the quality of reported data are monitored by IQVIA. The sampling methods used to select medical practices are appropriate to create a representative database of general medical and specialist medical practices in Germany [14].

study population
This retrospective cohort study included adult (≥18 years) T1D and T2D patients who received an initial prescription for a flash glucose monitoring sensor between January 2015 and July 2018 (index date) in 323 primary care (HA) and diabetology practices in Germany. This sensor is required to use the Flash Glucose Monitoring System. Additional inclusion criteria were as follows: (1) an observation period of at least 12 months before the index date and a follow-up period of at least 12 months after the index date and (2) an FGM sensor prescription within 7 to 12 months after the index date to ensure that the patients used the FGM sensor for a period of at least 6 months.

Study outcomes and covariates
The main outcome measure of the study was daily consumption of BZTS within 0 to 183 days after the index date compared to the period from 1 to 183 days before the index date. Calculated daily BZTS consumption was assessed using prescription information (pack size multiplied by the number of packs divided by the number of days between two consecutive prescriptions). The mean values per patient were then calculated over a period of six months before the index date and six months after the index date.
This analysis was performed for all FGM patients together and separately for men and women, T1D and T2D patients, the four age groups (≤40, 41-50, 51-60 and >60) and patients treated by HAs and diabetologists. In addition, to describe the study populations, the baseline prevalence of the following diagnoses was estimated: renal complications (ICD-10: E10.2, E11.2, N18, N19), neurological complications (ICD-10: E10.4, E11.4), peripheral vascular disease (ICD-10: E10.5, E11.5, I73.9), hypertension (ICD-10: I10), dyslipidemia (ICD-10: E78), stroke or TIA (ICD-10: G45, I63, I64) and myocardial infarction (ICD-10: I21-I23, I25.2). Since obesity and ophthalmological complications are not regularly documented by primary care physicians, these diagnoses were not used in the study. The proportions of patients treated with insulin alone, oral antidiabetic agents (OADs) alone, or insulin plus OAD were also reported.

Statistical analyses
Descriptive comparison of average daily BZTS consumption after and before the index date was performed using the Wilcoxon test. To examine the association between predefined variables and the probability of stopping using BZTS after the index date, a multivariate regression model was run with “no BZTS after the index date” (1= no BZTS, 0≥1 BZTS per day) as the dependent Variable equipped. Age, gender, diabetes type, practice specialty, oral antidiabetic drugs (OAD), HbA1c level, codiagnoses, and average daily use of BZTS before the index date served as predictors. P values of <0.05 were considered statistically significant. Analyzes were performed using SAS version 9.4.
Results
Baseline characteristics of the study patients
The present study included 1,343 patients prescribed flash glucose monitoring sensors. A total of 48.5% of them were T1D patients, the remaining 51.8% were T2D patients. The mean age [SD] was 51.4 [16.0] years; 39.1% were female. The mean baseline HbA1c value was 7.9% (SD: 1.4).
Of the patients, 79.2% received insulin therapy, 20.8% received insulin + OAD and 5.4% received OAD only. In total, 34.9% were diagnosed with neurological complications, 24.1% with renal complications and 15.6% with peripheral vascular disease (Table 1).

Average daily consumption of blood glucose test strips within six months before and six months after the start of FGM
Daily BZTS consumption is a non-normally distributed variable. Table 2 shows both mean and median values for daily consumption before and after the index date. Overall, average consumption fell from 2.7 (SD: 3.0) before the index date to 1.4 (SD: 2.4) after the index date. BZTS consumption decreased to a similar extent in T1D and T2D patients, women and men in the different age groups and in diabetology and HA practices (Table 2). However, 40% continued to use BZTS, and 30% used it at least once daily.

Association between predefined variables and the probability of no longer using BZTS after the index date
Table 3 shows the results of the multivariate regression model. Three variables (male gender (OR: 1.63, p<0.001), HbA1c <6.5% (OR: 1.67, p=0.045) and use of less than one BZTS per day before the index date) were significant associated with non-use of BZTS after the start of FGM.
discussion
Key findings
The present study, which was carried out between January 2015 and July 2018 in general practitioner and diabetology practices in Germany, led to two important findings. First, the average consumption of blood glucose test strips per day decreased sharply once patients began using FGM systems. This reduction in BZTS use was independent of the type of diabetes, gender, age group and treatment in diabetological or home care settings.
medical practices. Second, the proportion of patients who continued to use a large number of BCTs despite using an FGM system was relatively high. In addition, the average HbA1c value decreased in the population of FGM users.

Reduction of BZTS use
In the present study, the average daily consumption decreased from 2.7 to 1.4 test strips per day. In the study by Tyndal et al. A reduction from 3.8 to 0.6 test strips was reported with T1D patients in Scotland [15].
Bolinder et al. found in their study that the use of FGM systems in adults with well-controlled T1D reduced the time patients spent in a hypoglycemic state compared to patients measuring with capillary test strips [16].
Diabetes patients who do not use the FGM system must monitor their blood sugar regularly using test strips. FGM systems can replace some of these SMBG measurements, either helping to improve metabolic control and reducing glucose variability, or relying on sensor glucose readings without the need for double checking against blood glucose measurements. Another study in Germany and Austria showed that the moderate decline in the average frequency of self-monitoring of blood glucose from 2016 to 2017, particularly in the youngest age groups with the highest CGM rates, may be due to the increase in non-adjunctive CGM use, which requires less routine SMBG measurements [17].

Further use of BZTS
Szadkowska et al. have shown that FGM in children provides accurate measurements, but the accuracy depends on the glucose trend. When glucose levels fell, glucose levels were not as reliable, so the authors recommended blood glucose monitoring under these conditions [18]. Another study agreed with this finding and showed that the flash glucose monitoring system did not measure as accurately during physical activity and in the hypoglycemic state.
In such situations, it is inconvenient and difficult but very necessary for people with T1D to perform capillary blood glucose measurements and patients must be encouraged to do so [19]. There may be many patients who still choose capillary blood glucose monitoring because they doubt the accuracy of FGM measurements, especially in the first days of using this system.
Physiologically, there is a difference between blood glucose values and sensor glucose values when blood glucose rises or falls (e.g. after a meal or after an insulin injection). This is the interval in which glucose must diffuse from the bloodstream into the interstitium. In these circumstances, the patient will need to have a capillary blood test to make therapeutic adjustments.
Variables associated with discontinued use of BZTS
Male gender was associated with a higher probability of not using BZTS in addition to the FGM system. Other studies did not report gender differences in the number of BZTS during CGM device use [17].
HbA1c <6.5 was associated with non-use of BZTS after the index date. This result is consistent with another study. Aleppo et al. found that in adults with type 1 diabetes with good glycemic control, using a CGM device without regular monitoring measurements (SMBG measurements) is as safe and effective as using a CGM device with monitoring measurements [20].
Patients who used less than one BZTS per day before the index date were more likely to stop using BZTS after the index date. This result seems very logical. Regular SMBG measurements are necessary in any diabetes management with glucose-lowering agents to achieve good glycemic control. The load here is very high [21]. Patients who struggle with this burden before starting FGM are likely to replace any SMBG measurement with FGM scans and not take blood glucose measurements.

Study limitations
In general, retrospective primary care database analyzes are limited by the validity and completeness of the data on which they are based. The current study had several limitations that should be mentioned here.
First, hypoglycemia is rarely documented in the database, and the association between hypoglycemia and BZTS use could not be investigated.
Second, for most patients, only baseline HbA1c levels were available. Only in a small number of patients were the values during the period of use of FGM devices available, allowing the analysis of the associations between the use of FGM devices and HbA1c reduction in the entire study population.
Third, data on socioeconomic status (such as education and income) and lifestyle-related risk factors (such as smoking, alcohol consumption, and physical activity) were also missing. Fourth, data on diabetes duration were not available. Finally, information from hospitals was not available. The strengths of this work are the number of patients available for analysis and the use of real data, which allowed an unbiased exposure assessment (no recall bias).
conclusion

The daily consumption of BZTS was greatly reduced after the start of FGM in diabetic patients. However, a relatively high proportion of patients continued to use BZTS in addition to FGM measurements. Further longitudinal studies are needed to better understand the association between FGM use and BZTS. Future studies also need to examine the real-world outcomes of FGM device use, including improved HbA1c levels, reduction in hypoglycemic events, and reduced incidence of diabetes-related complications. <<