The experience of blood glucose monitoring in people with type 2 diabetes mellitus (T2DM)

Abstract Background Finger prick blood glucose (BG) monitoring remains a mainstay of management in people with type 2 diabetes (T2DM) who take sulphonylurea (SU) drugs or insulin. We recently examined patient experience of BG monitoring in people with type 1 diabetes (T1DM). There has not been any recent comprehensive assessment of the performance of BG monitoring strips or the patient experience of BG strips in people with T2DM in the UK. Methods An online self‐reported questionnaire containing 44 questions, prepared following consultation with clinicians and patients, was circulated to people with T2DM. 186 responders provided completed responses (25.5% return rate). Fixed responses were coded numerically (eg not confident = 0 fairly confident = 1). Results Of responders, 84% were treated with insulin in addition to other agents. 75% reported having had an HbA1c check in the previous 6 months. For those with reported HbA1c ≥ 65 mmol/mol, a majority of people (70%) were concerned or really concerned about the shorter term consequences of running a high HbA1c This contrasted with those who did not know their recent HbA1c, of whom only 33% were concerned/really concerned and those with HbA1c <65 mmol/mol of whom 35% were concerned. Regarding BG monitoring/insulin adjustment, only 25% of responders reported having sufficient information with 13% believing that the accuracy and precision of their BG metre was being independently checked. Only 9% recalled discussing BG metre accuracy when their latest metre was provided and only 7% were aware of the International Standardisation Organisation (ISO) standards for BG metres. 77% did not recall discussing BG metre performance with a healthcare professional. Conclusion The group surveyed comprised engaged people with T2DM but even within this group there was significant variation in (a) awareness of shorter term risks, (b) confidence in their ability to implement appropriate insulin dosage (c) awareness of the limitations of BG monitoring technology. There is clearly an area where changes in education/support would benefit many.


| INTRODUC TI ON
In the last 40 years, self-monitoring blood glucose (SMBG) has revolutionized the treatment of diabetes mellitus (DM). BG monitoring remains a mainstay of management in people with type 2 diabetes (T2DM) who take sulphonylurea (SU) drugs or insulin. However it has also been found to improve outcomes in people who are not taking these agents. 1,2 Although self-monitoring of blood glucose is now widely accepted as part of the management of people with type 2 diabetes, 3 Polonsky et al 4 showed that lack of understanding and limited skills to apply self-monitoring data to aid insulin dose adjustment, avoidance of thinking about BG values and diabetes, and a sense of pointlessness of self-monitoring were associated with infrequent self-monitoring and limited use of self-monitoring data for insulin dose adjustments.
At a general practice level using National Diabetes Audit data, we have found that there is a significant variation in BG metre strip precision, which is linked to a variation in measured BG and in glycated haemoglobin (HbA1c). 5 We recommended that account be taken of International Organisation for Standardisation (ISO) BG strip performance standards when BG metres are provided. 6 Furthermore, we previously showed a link between analytical BG metre precision and an established qualitative error grid, highlighting the potential impact of accuracy on clinical decision and outcomes. 7 Specifically, those metres with a variability of readings between 10% and 20% versus the standard laboratory method fall into the category of potentially affecting clinical outcomes for patients. This difference in BG monitor strip performance could mean that people make potentially harmful decisions about their food intake and insulin dose, based on imprecise BG readings.
We recently examined patient experience of BG monitoring in type 1 diabetes (T1DM) and drew some important conclusions. 8 There has not been any recent comprehensive assessment of the performance of BG monitoring strips or the patient experience of BG strips in T2DM in the UK. Our study aimed to start to address this by asking people with T2DM about their experience of day-to-day BG monitoring and how this influenced their decisions about insulin dosing. Furthermore, confidence in the BG monitoring equipment is essential, so how patients felt and behaved in this critical area was also examined.

| ME THODS
A digital questionnaire containing 44 questions (See Appendix 1) was prepared in consultation with clinicians and patients and was sent by email to patients on the Research for the Future (RftF) consent for approach database. 9 This is a National Institute for Health Responses from the survey were allocated specific numerical values (eg not confident = 0, fairly confident = 1, and so on) on a Likert Scale and the responses to certain questions were related to self-reported HbA1c.

Ethical approval was obtained from the West Midlands Research
Ethics Committee: REC Reference: 19/WM/0075.

| Statistics
Categorical responses were shown as simple percentages by levels and qualitative responses aggregated under headings. These are shown in the Figures.

| Demographics & description of study population
In relation to the online survey, 186/730 (25.5%) of those individuals approached to complete the online questionnaire responded. 84% were treated with insulin in addition to other agents. 23% of respondents were 60 years old or younger and 72% had been diagnosed with T2DM for more than 10 years. Of all respondents, 37% were women. The non-responder rate was 62%.
(b) confidence in their ability to implement appropriate insulin dosage (c) awareness of the limitations of BG monitoring technology. There is clearly an area where changes in education/support would benefit many.

K E Y W O R D S
blood glucose, diabetes education, HbA1c, monitoring, patient experience, type 2 diabetes The characteristics of non-responders were not materially different with 39% being women and 26% 60 years old or younger.
Of the respondents on insulin, 48% were injecting rapid acting insulin and 39% of insulin treated respondents said that they adjusted the dose of insulin that they administered. 24% reported giving insulin 4 or more times a day. 30% reported a daily insulin dose of up to 30 units per day, 35% between 31 and 60 units per day and 32% gave more than 60 units per day, with 3% unsure.
When asked about glycaemic control (4% of those questioned did not reply to this question), 51% of patients self-reported their last HbA1c result as ≤64 mmol/mol (≤8.0%) and 7% reported their last HbA1c to be >86 mmol/mol (>10.0%). 75% reported having had an HbA1c check in the previous 6 months.

| BG metre use
In relation to frequency of monitoring, 13% were testing once a day, 32% twice a day, 27% three times a day, 12% four times a day and 9% more than four times per day with 7% not at all. 63% stated that they were confident in the accuracy of their metres.
Regarding duration of BG metre use, 67% said that they had used a BG metre for more than 10 years. 51% had used the same BG metre for 3 years or more. 44% reported having been trained to use their BG metre. Concerning the matter of keeping a BG diary, 42% reported keeping a diary of BG readings consistently, 25% sometimes and 33% not at all.
In regards to difference between 2 consecutive readings when assessing metre accuracy, 18% of people reported a difference of more than 1.0 mmol/L, while 24% reported having to adjust their dose of insulin after a double check of the blood glucose reading.
The distribution of reported BG target ( Figure 1A) and actual readings ( Figure 1B) are described. 50% were setting a target for pre-meal readings ≥7 mmol/L with 45% setting a post-meal target of ≥9 mmol/L. This relates to the concerns that respondents reported in relation to the consideration of potential hypoglycaemia as reported below. The actual recorded %BG readings pre-meal ≥7 mmol/L was 70% with the actual recorded post-meal ≥9 mmol/L being 67% of BG readings. For bedtime readings 15% were set at ≥9 mmol/L and for actual bedtime readings 50% were ≥9 mmol/L.

| Respondent concerns and confidence
We next asked questions around how people with T2DM felt about BG levels and insulin dosing, 30% stated that they keep BG level high at times, to avoid hypoglycaemic episodes; this was reflected in the targets they set for pre-meal and post-meal BG levels. Furthermore, 52% were concerned that they might be over-or under-dosing their insulin (Breakdown by HbA1c is shown in the Figure S1, which illustrates that this proportion increased to 66% in those whose last reported HbA1c was ≥65 mmol/mol).
In relation to those with a last reported HbA1c of ≥65 mmol/ mol, a majority of people (70%) were concerned or really concerned about the consequences of running a high HbA1c (Figure 2A). This contrasted with those who did not know their recent HbA1c, of whom only 33% were concerned or really concerned about the consequences of running a high HbA1c and those with an HbA1c ≤64 mmol/mol of whom only 35% were concerned about the shorter term consequences of high BG levels.
For longer term consequences in relation to HbA1c, for both for HbA1c ≤64 mmol/mol and ≥65 mmol/mol, there was concern F I G U R E 1 (A) Target blood glucose (BG) Level (mmol/L). (B) Actual blood glucose (BG) Level (mmol/L) in 85% and 80%, respectively, about their BG readings ( Figure 2B).
This was also true to a lesser extent in those who did not know their HbA1c (concern in 58% of those who did not know their recent HbA1c).

| Patient views on and use of BG metres
We next explored respondents' knowledge of BG metre performance in terms of accuracy and how this was reflected in their day-to-day practice. We found that 77% of respondents said that they had not ever discussed BG metre performance with a healthcare professional and 33% said that they had not been HbA1c being 67% ( Figure 3).
When questioned about adequacy of information about BG monitoring, 25% of individuals responded that they had sufficient information with 13% believing that the accuracy of their BG metre was being independently checked. Only 9% remembered discussing BG metre accuracy when their latest metre was provided.

Participants in our survey gave a wide variety of responses when
asked what might reduce their concern. These provide us with insight into their day-to-day experience of BG monitoring and are listed in Appendix 2.
Examples included: I do not expect these devices to be perfect but they should be reasonably correct.
I would like to have regular discussion about diabetes level and metre checking.
I don't know. The device was given to me by the hospital so have always presumed they were confident about the accuracy.
I have poor experience of professional advice because of stark changeover at my surgery.
Very little help is given regarding diet. As I have aged, now 75, I experience more Hypos.
As I am quite old (81 yrs), I don't feel the need to be anxious about how much life there is left in me! I have several 'chronic' conditions other than diabetes.
I need more education.

| DISCUSS ION
This study has shown that in a group of people with T2DM, concerns regarding long-term complications were highly prevalent. This was seen in groups with both high and low levels of HbA1c, and if HbA1c was not known. This was in contrast to views of current BG levels where there was a noticeable lack of concern when HbA1c was not known. This is similar to our previous findings in people with T1DM. 7,8 We found significant concern about over-or under-dosing of insulin. Only one-fifth of patients responded that they had sufficient information. Only 9% remembered discussion of BG metre accuracy when their latest metre was provided. This indicates a large gap in patient education in this area, including about ISO standards. 6 This is reflected in the fact that 30% of those questioned reported keeping their BG levels high to avoid hypoglycaemic episodes. These findings are of direct relevance to all involved in management of diabetes in a primary care setting, where increasingly T2DM individuals are exclusively managed.
Our findings provide important insights into the way that people view and react to their BG readings. This study would suggest that concern / anxiety about the longer term consequences of high BG levels is an effective motivator to encourage tighter glycaemic management. This was also described in an interview-based qualitative study 10 in people with diabetes undertaken to develop the Health and In the same study, 10 some of those interviewed reported concerns over whether they were managing their BG levels correctly, achieving stable and consistent BG levels over a period of time, administering medication correctly, whether their diabetes was 'stable' and/or whether there had been any implications of having diabetes on areas of their health (such as neuropathy). Individuals noted that there was a degree of stress with the 'review appointments', and spoke of concerns of being 'told off' at these reviews.
In our study, we found that 52% of respondents were concerned that they might be over-or under-dosing their insulin. In an important outpatient questionnaire study from 2015, Ward et al 18 described in a survey of people with T2DM that respondents' most frequent personal pattern was to test "occasionally, as needed", which did not differ by insulin use status, gender or age.
Interestingly, in those people on insulin, significantly better control was found in those never experiencing being "too busy" than in those who were "too busy" several times per week. Similarly, never "feeling discouraged" as a barrier to self BG monitoring was associated with better metabolic control than experiencing the barrier a few times per week. Certainly this was borne out by some of the comments that patients provided to us such as: Important especially if you are reacting to high or low reading or planning your meal, salad with low carbs. I take less insulin if reading is under 6 or more if having high carbs and reading is higher than 8.
It helps to understand how your diabetes is controlled. Plus when driving it is important for safety.
It needs to be correct.
Long as it's not higher than 11 or lower than 5. Fine by me.
The group who was contacted for our survey comprised long-term We have previously shown that decisions taken in GP practices, including the decision to move to insulin treatment and provision of a BG metre have a profound influence on glycaemic outcomes in T2DM. 11 Our analysis has shown both in modelling and in real world data that the effect of longer term multiple use of less precise BG monitoring strips resulted in an increase in longer term variability of actual BG levels in both models and as measured by HbA1c. 5 We found that the increase in BG variability was over twice the change in variability in BG strips.
It should be pointed out that none of the people who replied to our survey were using a Flash blood glucose monitor. 20,21 This technology is increasingly being utilized by people with T1DM and in some with T2DM. 22 Nevertheless, traditional BG monitoring will continue to be the way that most people who take insulin to treat diabetes monitor their diabetes for some time to come.
The study population was everyone who has signed up to 'Research for the Future (RfTF)'.
We had no control of who responded to the survey which was presented online. However, tt is likely that many people with T2DM and on insulin were likely to reply to the survey, hence the high proportion of people with T2DM on insulin who responded to the survey.

| Strengths and limitations
The survey used in this study was comprehensive and covered many of the aspects of day-to-day diabetes management and living with diabetes. Respondents were from across the spectrum of age and duration of T2DM.
Those contacted had already expressed an interest in participating in diabetes research and so there will be some degree of responder bias. Nevertheless the characteristics of responders were similar to those of non-responders. The non-responder rate was 62%. However this is not unusual for an online survey of this kind.
There will be recall bias for the self-reported HbA1c results reported.
We do not have specific information on demographic characteristics as this was an online survey. Ethnicity was not captured in our study as this is not generally self-reported accurately.
The survey was online, so people who do not have access to a computer, tablet or smart phone were not able to respond.
Furthermore, we have relied on self-reported HbA1c and BG values.
However Gonder-Frederick et al 23 reported that misrepresentation of test values by respondents was extremely rare. This is more likely to be true in well-motivated groups such as those in our study.

| CON CLUS ION
In conclusion, there have been a large number of major devel- in education/support would benefit many people.

CO N FLI C T O F I NTE R E S T
No author has any conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.

Accuracy
• I do not expect these devices to be perfect but they should be reasonably correct.
• I feel that any reading is at that moment. When I have a reading of 6 and minutes later 4, that worries me.
• I think this refers to how close the reading is to the actual value.
• I would expect the meter to be accurate.
• Only that insulin take is dependent on the reading. Wrong reading = wrong dosage.
• Had never thought of checking meter with control solution more fre-

Why it is important to control BG levels
• If you don't control your blood sugars you can get more health issues.
• Managing your health probably.
• So you can keep your levels low because of risk of stroke etc.
• That accuracy is important to avoid hypo or hyper.
• To keep below 7 mmol if possible. • Carry out more comparisons and/or consult my GP.
• Change meter/control my diet better with more exercise.
• Check BG more often.
• Check bloods at clinic.
• Check glucose levels more than twice a day.
• Check its accuracy.
• Check with a control solution.
• Constant vigilance regarding what I eat and the amount of exercise I do.
• Does the meter really recall correct sugar reading?
• Double testing is important.
• More training for patients on how to use the meter regarding how to understand results.
• Find out of I should have been given control solution.
• Get my meter callibrated against a control solution.
• Have better professionally advice than I currently have.
• HAVE MY METER CHECKED ON MY ANNUAL CHECK UP.
• I don't know. The device was given to me by the hospital so have always presumed they were confident about the accuracy.
• I would ask my nurse specialist at my next appointment.
• Keep on eye on my diet.
• More professional monitoring and information. •