Barriers to Patient Use of Control Solution for Glucose Meters: Surveys of Patients,Pharmacists,and Providers in a Metropolitan Area

Abstract

Background:

An important factor in controlling diabetes is self-monitoring of blood glucose. Manufacturers of glucose meters recommend routine use of control solution to ensure accuracy. Previous studies have demonstrated that glucose meters vary in accuracy and that patients are not using control solution as recommended. The purpose of this study is to identify potential barriers to control solution use from multiple perspectives including patient, pharmacist, and provider.

Methods:

This study was a prospective, observational survey design. First, 25 randomly selected chain and independent pharmacies in the Tulsa metropolitan area were audited for control solution accessibility. These pharmacies were then used to survey pharmacists, via telephone, regarding control solution inventory and perception of importance of use. Next, providers were electronically surveyed on their routine practice recommendations, while 60 patients with diabetes were randomly selected for telephone survey on use and perceptions of control solution.

Results:

Twenty-five pharmacies were audited and 23 pharmacists, 60 patients, and 29 providers were surveyed. Only 39% of pharmacies stated they supplied control solution, however, only 1 pharmacy visibly stocked it. The only patient factor that appeared to have an impact on control solution usage was having type 1 versus type 2 diabetes (38% vs 15%). Providers are aware of what control solution is (62%), but only half felt it should be routine practice with 44% of those never recommending it.

Conclusion:

This study raises awareness for the need to educate patients, providers, and pharmacists about use of control solution to ensure glucose meter accuracy.

The importance of the management of diabetes cannot be over emphasized considering the potential devastating impact of the complications of uncontrolled disease. The American Diabetes Association publishes annual Standards of Care in practice to provide the best evidence to reduce the risks of complications of uncontrolled diabetes. ¹ The landmark Diabetes Control and Complications Trial and UK Prospective Diabetes Study emphasized the importance of intensive A1c reduction for the prevention of microvascular events in patients with diabetes.^2-3 The best method to accomplish this is maintaining blood glucose levels in desired therapeutic ranges. ¹ This can be achieved by balancing diet with physical activity and medications. ¹ Patients should use a glucose meter with diabetes self-management to monitor blood glucose, adjust therapy, and modify diet or exercise to find that balance that will achieve glycemic targets, which, in turn, should lead to controlled A1c values. Patients also rely on self-monitoring of blood glucose (SMBG) values to direct acute treatment for hypo- or hyperglycemia. ¹ An inaccurate reading could lead to either over- or undercorrection with medications, potentially putting the patient at risk for severe hypo- or hyperglycemic complications or fatal consequences.

Glucose meter manufacturers recommend using control solution (CS) (which is brand specific) to confirm proper function when a new glucose meter or new vial of test strips is used, or when there is a suspected error or malfunction of the glucose meter or test strips. The test is performed by applying a drop of CS to a test strip that is inserted into a glucose meter. If the reading on the glucose meter is out of the control range (printed on the side of a test strip vial), this indicates a problem with the strips or the glucose meter. The patient can then contact the manufacturer (toll-free number listed on the back of the glucose meter) to gain assistance resolving the problem. CS is not always included with the purchase of a glucose meter and has an expiration date which puts a time limitation on its use.

In 2013, a study by Chaudhry and Klonoff determined that while 82% of their study patients with type 1 diabetes had knowledge of CS, only 26% were using it as intended. ⁴ A major concern identified was the lack of availability of CS, as found through a survey of pharmacies indicating that they did not stock or had limited supply of CS. Pharmacies reported that this was due to a lack of patient demand.

The purpose of this study was (1) to determine the availability of CS in a sampling of community pharmacies and (2) to identify patient, pharmacist, and provider perceptions of CS use that may be associated with barriers to use.

Methods

Design

This study used a prospective observational design conducted in 4 parts: (1) community pharmacy environment audit for visible CS, (2) community pharmacist telephone surveys, (3) patient telephone surveys, and (4) provider electronic surveys. Pharmacies, pharmacists, patients, and providers located in the metropolitan area of Tulsa, Oklahoma, were each audited and/or surveyed from January 2014 through May 2014 using standardized scripts. This study was approved by the University of Oklahoma Health Sciences Center Institutional Review Board.

Setting

The Tulsa metropolitan area was divided into 5 geographic sections (north, south, east, west, and central) to represent unique socioeconomic demographics. Patients and providers were from an academic outpatient group practice.

Pharmacy Audit

A map of the city was used to identify all pharmacies within each geographic section. Each pharmacy was then numbered and a random number generator was used to select 4 chain community pharmacies and 1 independent pharmacy for each section for a total of 25 pharmacies.

An investigator visited each pharmacy for an on-site audit to determine if the pharmacy had CS visibly stocked or space allocated on a shelf and the time it took to locate the CS. The location of the CS in the store was documented along with which brands were stocked.

Pharmacist Survey

Once the pharmacy audits were completed, a pharmacist from each of the pharmacies was contacted by an investigator for a telephone survey. Survey questions included pharmacist’s knowledge of CS, perception of importance of usage, usage recommendations, availability of CS, brand availability, location in pharmacy where CS is stocked if at all, whether it is ordered by the pharmacy or the “front end” of the store, the ordering process, and knowledge of insurance coverage.

Patient Survey

The next phase surveyed patients. The institution’s electronic medical record (EMR) was used to generate a report identifying all patients diagnosed with type 1 or type 2 diabetes within the following clinics: internal medicine, family medicine, pediatric, and pediatric diabetes. The patients from each clinic were then numbered and a random number generator was used to select 15 patients from each clinic for a total of 60 patients. Patients were included if they were assigned a primary care provider and had at least 1 visit in the past year, had the diagnosis of diabetes for at least 1 year, utilized SMBG at least once a week, had a working telephone number, and able to communicate in English or through a translator. If a selected random patient did not meet the inclusion criteria or if a patient did not consent to participate, the random number generator was used to select another patient until 15 eligible patients were identified from each clinic. Survey questions included CS awareness, CS usage or barriers to usage, perception of the importance of CS, perception of their diabetes control and importance of obtaining regular A1c testing, frequency of SMBG monitoring and the perception of the importance of SMBG, the glucose meter brand used, which pharmacy was used, whether formal diabetes education had been received, if they had received care from a certified diabetes educator or diabetes specialist (defined by the patient perception of a specialist), and medication usage. Other information collected included type of diabetes, value of the most recent A1c, clinic the patient received care from, years since diagnosis, age, and insurance type. For pediatric subjects (age <18), patient caregivers were surveyed for their responses unless the patient was responsible for his/her own diabetes management.

Provider Survey

While patients were being surveyed, all faculty providers and medical residents employed by each of the 4 clinics were simultaneously emailed an electronic link to complete a web-based survey. The survey asked questions regarding provider knowledge and perception of CS importance, practice habits and frequency of CS recommendation, clinical practice location, and level of provider (faculty, resident, or endocrinologist). After 2 weeks, each provider was sent a reminder email to complete the survey and given 2 additional days.

Data Analysis

Data collected from each phase of research were entered into Microsoft Excel 2013 and then analyzed using SAS v9.3 (SAS Institute Inc, Cary, NC). Patient demographics were explored using frequency distributions and measures of central tendency, as appropriate. Descriptive statistics were used to analyze all survey responses as well as the environmental audit. Descriptive information from open-ended questions was grouped by theme.

Results

Pharmacy Audit

Only 1 of the 25 pharmacies displayed CS where it was easily visible. It should be noted that this pharmacy was independently owned and specialized in diabetes care. The CS was displayed behind the counter, but in clear sight to patients. None of the other pharmacies stocked CS where a patient could easily find it, whether visibly behind the counter or on a shelf outside of the pharmacy area.

Pharmacist Survey

Of the 25 pharmacies audited, 23 pharmacists consented and were surveyed. Two did not want to participate without supervisor permission. All pharmacists surveyed (100%) reported being familiar with CS. However, only 61% felt that its use should be part of routine practice. Of those, only 14% “always recommend” its use, 43% “recommend only in certain situations,” and 43% “never recommend” its use. Reasons for never recommending include it “not being a priority,” “patients don’t ask for it,” “time constraints,” and it is “not stocked.” Only 39% indicated that they stocked CS. Of those, 56% said it is stocked behind the counter, 33% report it is stocked with the over-the-counter diabetes supplies, and 11% didn’t know. In addition, 87% of the pharmacists stated that they can order CS if the patient asks for it.

Patient Survey

Fifteen patients participated from each of the 4 clinics (n = 60). Of participants, 43% had the diagnosis of type 1 diabetes, while 57% had type 2 diabetes. The mean age was 36 years (range 6 to 94 years). Adults made up 70% of the population while 30% were pediatric. Of all patients surveyed, 62% carried State Medicaid or Medicare coverage, 33% had private insurance, and 5% had no insurance. Looking at time with the diagnosis of diabetes, 26% (9) of those with the diagnosis for 1-9 years used CS, 31% (4) with it for 10-20 years used CS, and 17% (1) with it for ≥21 years used CS (P = .9). 72% of patients (43) reported that they had received formal diabetes education at some point. Only 17% of all patients had a controlled A1c of <7% (20% of them used CS), 53% had an A1c ≥ 7% (25% of them used CS), and 30% had no A1c documented (22% of them used CS).

Although 98% of patients perceived SMBG as an important aspect of achieving glycemic control, only 23% (14) indicated that they used CS, despite 52% (31) claiming awareness of it. Of the remaining 77% who did not use CS, 67% (31) were unaware of it (P = .05), 28% (13) lacked understanding of it, and 2% (1) indicated their pharmacy did not stock it. Patients with type 1 diabetes were significantly more likely to use CS than those with type 2 diabetes (38% vs 15%; P = .03). Also, 100% of patients who use CS claim to see a diabetes specialist (P = .01). Use was not influenced by time since diagnosis (P = .9), receipt of formal diabetes education (P = .7), or the level of diabetes control (P = .1).

Provider Survey

Electronic surveys were sent to all providers (217) employed by each of the 4 clinics. Responses were received from 32 providers; 13 from family medicine, 8 from internal medicine, 10 from pediatrics, and 1 from pediatric diabetes. Of the 32 providers, 13 were residents, 18 were faculty physicians, and 1 was an endocrinologist. Three responses were excluded because the respondents “never work with diabetes patients.” Of the 29 responses included, 62% of the providers (18) were familiar with CS. Only half (50%) of these 18 providers felt that usage should be routine practice, and of those, 56% “recommend” or “recommend in certain situations” while 44% “never recommend” it. Reasons for not recommending include “don’t think about it,” “use a clinic glucose meter to compare against for accuracy,” and “not sure it is needed with new technology.” Type of clinic did not influence whether a provider recommended using CS (P = .2) or whether the patient actually used CS (P = .2). Whether the provider was a resident or faculty provider did not influence recommendation of CS use (P = .5).

Discussion

Almost all of the patients surveyed believe that monitoring blood glucose levels is important. Yet, just over half of them are generally aware of CS, and just under half of these actually use CS; just 23% of the study population, which is similar to the 26% found in the study by Chaudhry and Klonoff. ⁴ Patient factors did not influence the use of CS. The length of time a person had diabetes was not associated with use of CS, implying that those with any certain length of time with the diagnosis are no more motivated to use CS than others. Receiving formal diabetes education in the past was also not associated with use of CS. However, there seems to be something about having type 1 diabetes that associates those patients with CS use more than patients with type 2 diabetes. Involvement of a diabetes specialist was also associated with CS use. Perhaps patients with type 1 diabetes are more likely to get care in a setting that is more likely to employ diabetes specialists, as was the case in this study.

The pharmacy audit revealed that an unexpectedly low number, only 1, of the pharmacies stocked CS in a location visible to patients. This independent pharmacy may have only stocked the product due to its specialization in diabetes services. Patients of the other 96% of the pharmacies audited would not be able to find CS without asking for assistance. Expectedly, 87% of pharmacists, a number consistent with the finding of 85% in the Chaudhry and Klonoff study, will special order CS for patients that ask for it. ⁴ However, this presents an extra step, and therefore, a barrier that could decrease the likelihood of a patient obtaining and using CS.

Of the 43% of pharmacists that reported they “never recommend” CS use, some of their reasons imply a lack of demand from patients. Pharmacies will not want to stock a product that won’t sell for obvious business reasons. However, pharmacists may be able to create patient demand by recommending regular CS use in appropriate scenarios. Some pharmacists may not be fully aware of whether they stock it or not; 11% said they were not aware and 33% said it was with the OTC items, which was not observed by the investigator store audits. It is also possible in some chain stores that this responsibility falls to the front-end manager of the store and not the pharmacy staff.

Despite all pharmacists (100%) and most providers (62%) being familiar with CS, many still do not recommend its use; in fact 43% of pharmacists that felt it should be routine practice say they never recommend it. It seems pharmacists and providers are either not familiar with manufacturer recommendations, they don’t agree with them, or just don’t see this as a priority relative to other daily responsibilities. Now that most glucose meters don’t require a code chip, some providers believe that our technology has advanced to the point that we no longer need to “calibrate,” or test for accuracy. The FDA requires manufacturers and device-user facilities to report when their device may have caused or contributed to, or may be likely to cause or contribute to a death or serious injury due to malfunction. Reports are sent to the FDA Manufacturer and User Facility Device Experience (MAUDE) system. A search of the MAUDE database using the term ‘glucometer’ revealed 438 issues reported to the FDA for errors associated with these products from various manufacturers. ⁵ Two common examples include incorrect display of SMBG results and incorrect glucose readings. Melker ⁵ points out that many manufacturers fail to investigate reports of incorrect glucose readings (ie, glucose meter reads a hyperglycemic episode while laboratory testing reveals hypoglycemia) claiming the patient or provider’s failure to use CS as recommended prior to the inaccurate reading is at fault; in a sense, making this a user error. ⁶

Accurate glucose meter readings play an important role in the attainment of diabetes control; however, as with any measure with variations in sensitivity, an acceptable degree of variance from the true glucose value must be defined. At a 20%-error level, self-monitored hypoglycemia may be missed by 1 in 10; in contrast to <1 in 100 at a 10%-error level. ⁷ Glucose meter accuracy standards are regulated by the US Food and Drug Administration (FDA) using guidelines from the International Organization for Standardization (ISO). The 2003 ISO guidelines require 95% of SMBG results from a glucose meter to read within ±20 mg/dL or ±20% of actual blood glucose levels ≥ 75 mg/dL. ⁸ According to data presented at the Diabetes Technology Society meeting in 2013, The American Association of Diabetes Educators (AADE) reports that many glucose meters fall short of the ISO 2003 accuracy standard. ⁹ Stricter 2013 ISO recommendations now require a) 95% of meter results to fall within ±15 mg/dL or ±15% of actual blood glucose levels ≥ 100 mg/dL; and b) 99% of all values within zones A and B of Parkes Error Grid ⁸ . Glucose meter manufacturers will have to further improve the accuracy of their devices to meet these more stringent standards. A 2014 study by Freckmann et al ¹⁰ assessed the accuracy of 12 SMBG systems for adherence to the ISO 2003 standard as well as the ISO 2013 standard. They reported that 83% of the systems fulfilled the 2003 ISO standard while only 67% met the 2013 ISO standard. This demonstrates that while most glucose meters meet accuracy standards, there are still many that may not. Since patients may not be aware of which glucose meters meet the accuracy standards, use of CS to self-assess their meter accuracy could help to identify potential inaccuracies. In addition, a survey conducted by the AADE found that most patients using insulin (84%) perceive glucose meter use to be important, but they do not perceive the reality that many glucose meters do not hold up to current accuracy standards. The survey further identified that 1/3 of patients fail to receive information about the importance of selecting an accurate glucose meter from a health care provider. ¹¹ These findings illustrate the increased need for patient education about the accuracy of their glucose meters, as well as guidance from health care providers regarding selection of accurate devices.

This study has been useful to inform investigators of CS use and availability in the Tulsa metropolitan area, as well as to some barriers to use. However, this study is not without limitations. The small sample size from each clinic may have limited the ability to determine a significant difference between certain parameters. Also, independent samples limit making direct relationships. The patients surveyed did not necessarily see the providers surveyed or use the pharmacies audited and surveyed. Despite this, the study has been valuable by informing investigators of potential barriers to CS use that can be further investigated and educational programs that could be implemented to better inform patients, pharmacists, and providers of the value of using CS.

Conclusions

Ideally, providers should inform patients of the importance of CS use and this should then be reinforced by pharmacists. Hearing this from “both sides” would likely increase patient awareness and, therefore, appropriate use of CS. Looking more closely at the practice habits of services that focus on type 1 diabetes could unveil practices that could be used to increase CS use in type 2 diabetes populations. Further research should be conducted to determine if this difference is due to the clinic environment or the education received in the differing clinics.

The results of this study identify a need to inform patients, providers, and pharmacists about the importance of CS usage. Findings have informed investigators of local barriers to CS use and a need to find effective methods of informing all concerned parties to potentially improve use to recommended standards which could improve patient safety and clinical outcomes.