Systematic Diabetes Screening Using Point-of-Care HbA1c Testing Facilitates Identification of Prediabetes

Heather P. Whitley; Courtney Hanson; Jason M. Parton

doi:10.1370/afm.2035

Published eLetters

If you would like to comment on this article, click on Submit a Response to This article, below. We welcome your input.

Submit a Response to This Article

Compose eLetter

Title *

Author Information

Contributors
First Name and Middle Initial * First or given name, e.g. 'Peter'. Last Name * Your last, or family, name, e.g. 'MacMoody'. Email Address * Your email address, e.g. higgs-boson@gmail.com Occupation * Your role and/or occupation, e.g. 'Orthopedic Surgeon'. Affiliation * Your organization or institution (if applicable), e.g. 'Royal Free Hospital'.

Statement of Competing Interests

Competing interests? *

Yes

Please describe the competing interests

CAPTCHA

This question is for testing whether or not you are a human visitor and to prevent automated spam submissions.

What code is in the image? *

Enter the characters shown in the image.

Vertical Tabs

Jump to comment:

Author Response to the Letter: Apples to Apples
Heather P. Whitley

Published on: 05 May 2017
Evaluating Point-of-Care HbA1c Testing in Primary Care: Comparing Apples to Apples
Jason L. Salemi

Published on: 01 May 2017

Published on: (5 May 2017)
Page navigation anchor for Author Response to the Letter: Apples to Apples
Author Response to the Letter: Apples to Apples
Heather P. Whitley, Associate Clinical Professor
Other Contributors:
Kind thanks to Drs Salemi and Zoorob for their thoughtful comment in response to our article "Systematic Diabetes Screening Using Point-of-Care HbA1c Testing Facilitates Identification of Prediabetes."[1] We appreciate the ability to further expound on the study outside of the brief published report.
Firstly, the authors inquire about the application of exclusion criteria between arms. In fact eligibility criter...
Show More

Kind thanks to Drs Salemi and Zoorob for their thoughtful comment in response to our article "Systematic Diabetes Screening Using Point-of-Care HbA1c Testing Facilitates Identification of Prediabetes."[1] We appreciate the ability to further expound on the study outside of the brief published report.
Firstly, the authors inquire about the application of exclusion criteria between arms. In fact eligibility criteria were applied to both the 'standard practice' and the 'active screening' arm. In the 'standard practice arm' patient charts were retrospectively reviewed to first verify criteria (age e 45 years, lack of pregnancy, lack of oral or injectable steroid use in the past 3 months, past medical history of diabetes, or A1C test in the past year). As the consort diagram shows, this resulted in exclusion of 385 individuals. The 'active screening arm' also underwent evaluation of these same criteria for eligibility, albeit prospectively, but uniquely were verbally confirmed. This addition of the verbal evaluation of eligibility criteria potentially reduced the false positive rate of elevated A1C outcomes in the 'active screening' arm of the study. This aspect of the study design allowed for 25 individuals (n=23 of whom were verbally excluded for recent steroid use) to be removed from the 'active screening arm', whose A1C outcome would have otherwise been higher, thus falsely inflating the outcome of interest (increasing the likelihood of a positive result). While the application of verbal confirmation of eligibly criteria was not possible for the 'standard practice arm' doing so would have further widened the difference in likelihood of positive outcomes between the two study arms.
The authors of the comment critique the comparative rates of screening between the arms (active 100% versus standard practice 22%) due to study design. The prospective EMR review identified 324 of the possibly 689 patients who met criteria to be offered an A1C screening test in the 'active screening arm'. However, 117 were unable to be offered due to missed appointments or conflicts with the investigator's schedule. After further subtracting the 25 patients that were excluded by verbal confirmation and the 17 who declined participation a resulting 51% (n=185) were screened in the 'active arm'. Reviewing this calculation confirms that those in the proactive screening arm were more than twice as likely to be screened for diabetes than did the standard screening arm (p=0.005). Still, irrespective of comparative screening rates to the 'active arm', the low screening rate of 22% under standard practice certainly highlights the low frequency of diabetes screening occurring when otherwise indicated by national guidelines.
Secondly, the authors draw attention to the discrepancy in body mass index (BMI) between the two study arms, noting that those in the active arm tended to have a higher BMI (1.7 kg/m2 difference). Therefore, test outcomes could have been influenced, to some degree, by these baseline demographics. They further mention that the POC A1C tests were provided free-of-charge, while the standard practice arm, which used venipuncture blood glucose 96% of the time, would have been commonly accompanied by some cost to the patient. In fact, 93-95% of patients in both arms had health insurance, and therefore were not likely to assume additional costs outside of their copayment for the visit, which was assumed by all individuals regardless of participation in the study.
Lastly, the authors point out performance variability of point-of-care A1C instruments. As discussed in your reference [2], all A1C devices, POC instruments as well as floor models, are governed by the same standards set forth by the National Glycohemoglobin Standardization Program.[3] These require all A1C instruments to meet acceptable performance (+/- 6% of the target value) in terms of accuracy and precision. Similarly, the International Organization for Standardization (ISO) regulates the performance measures for blood glucose meters.[4] The most recently published standards (ISO 2013:15197) specify performance measures, which must report outcomes within 15% of the reference measurement 95% of the time. Thus, one should not forget that any instrument of measurement comes with inherent outcome variability in assessing the true reference. Therefore, this rationale cannot simply be pointed toward A1C devices, but rather should be considered for all such devices for that matter. We agree that inclusion of, or at least reference to, performance measures of study instruments would strengthen the evaluation of study designs.
Again, we thank the authors for their comment and questions for clarification, and for allowing us to further explain our findings. We appreciate their shared enthusiasm for POC A1C devices in the clinical setting to efficiently and effectively screen patients for diabetes and prediabetes.
REFERENCES
1. Whitley HP, Hanson C, Parton JM. Systematic Diabetes Screening Using Point-of-Care HbA1c Testing Facilitates Identification of Prediabetes. Annals of family medicine. 2017;15(2):162-164.
2. Whitley HP, Yong EV, Rasinen C. Selecting an A1C Point-of-Care Instrument. Diabetes spectrum : a publication of the American Diabetes Association. 2015;28(3):201-208.
3. National Glycohemoglobin Standardization Program homepage. Available from http://www.ngsp.org. Accessed 3 May 2017
4. International Organization for Standardization. In vitro diagnostic test systems -- Requirements for blood-glucose monitoring systems for self-testing in managing diabetes mellitus. Available from https://www.iso.org/obp/ui/#iso:std:iso:15197:ed-2:v1:en. Accessed 3 May 2017
Competing interests: None declared
Show Less

Competing Interests: None declared.
Published on: (1 May 2017)
Page navigation anchor for Evaluating Point-of-Care HbA1c Testing in Primary Care: Comparing Apples to Apples
Evaluating Point-of-Care HbA1c Testing in Primary Care: Comparing Apples to Apples
Jason L. Salemi, Assistant Professor of Epidemiology
Other Contributors:
LETTER TO THE EDITOR
There's little not to appreciate about the potential advantages of point-of-care (POC) hemoglobin A1c (HbA1c) testing for diabetes screening and diabetes management in primary care settings. Patients and family physicians benefit from the immediacy of both the testing itself and the receipt of test results, the rapidity of feedback, and timely decision- making. POC HbA1c testing has even pro...
Show More

LETTER TO THE EDITOR
There's little not to appreciate about the potential advantages of point-of-care (POC) hemoglobin A1c (HbA1c) testing for diabetes screening and diabetes management in primary care settings. Patients and family physicians benefit from the immediacy of both the testing itself and the receipt of test results, the rapidity of feedback, and timely decision- making. POC HbA1c testing has even proven effective in emergency department settings as a fast and reliable tool to detect unknown and poorly controlled diabetes and to appropriately select antidiabetic therapies that may improve clinical outcomes and reduce complications. To add to the growing literature in this arena, in the March/April 2017 issue of Annals of Family Medicine, Whitley et al. assess the extent to which POC testing, as compared to usual care, increases screening rates and facilitates the identification of prediabetes in a single-physician family medicine clinic.[1] Although we agree with the authors' conclusions regarding POC HbA1c testing and the proposed direction of future research, we feel there are a few concerns to which the authors' response would be beneficial in assessing the findings from their important study.
First, we are concerned with the validity of the comparison of screening rates between the 'active' (POC HbA1c) and 'standard' (usual care) screening arms due to seemingly differential determination of the samples included in the study. The authors report a 23% screening rate among the 324 included participants in the 'standard' arm - those identified from scheduled Wednesday medical appointments between April 2013 and March 2014 and whose electronic medical record (EMR) were not indicative of any exclusion criteria (age <45 years, pregnant, past medical history of diabetes, HbA1c test in the past 12 months, steroid use in the past 3 months). The authors then compare this to the 100% screening rate observed in the 'active' arm. However, after application of exclusion criteria to these patients (with Tuesday appointments) that were identical to the 'standard practice' arm, there were also 324 eligible participants. However, the consort diagram in the supplemental appendix then also depicts exclusion of another 117 patients who did not show to their medical appointment or were otherwise not offered participation due to time constraints, 17 patients who outright declined participation, and 1 patient whose POC HbA1c measurement was deemed 'inaccurate'. The result was an 'active' arm with 164 participants, only half of the 324 eligible patients. We feel the comparison of screening rates in these two differentially constructed arms is misleading. It would seem obvious that POC HbA1c screening rates would be 100% if a sample were restricted only to patients who agreed to be screened.
Second, we are concerned about the comparison of these two arms to estimate the ability of POC HbA1c testing to identify prediabetes and even previously undiagnosed diabetes. The authors state that the "22% difference in hyperglycemic detectability between screening arms is best explained by test utility".[1] However, in addition to the differential determination of the final samples in each arm, statistical tests applied to the Table 1 mean body mass index (BMI) (a Wilcoxon-Mann-Whitney test) and categorized BMI distributions (a chi-square test) would reveal statistically significant differences between the arms. The 'active' arm, which had higher rates of diabetes and prediabetes had fewer health weight individuals (17%) compared to the 'standard' arm (28%), and even the mean BMI among the morbidly obese in the 'active' arm was higher than the same group in the 'active' arm. This underscores the likelihood that baseline differences, and not test utility, may have explained at least some, although likely not all, of the differences in screening outcomes. Although analyses that adjusted for these differences may be beyond the scope of the study, acknowledgment of these differences in explaining the primary study outcomes seems warranted.
Third, the precision, accuracy, and bias of POC HbA1c instruments vary depending on the device/system used.[2-5] Therefore, it would be valuable to readers and as a point of comparison to future studies that the authors provide details on the specific POC instrument(s) used as part of their study. Last, as the authors discuss the improvement in screening rates conferred by POC testing, they should reiterate that their findings were based on comparing free POC HbA1c testing to usual care ('standard' arm) that often carries with it at least some cost to patients (e.g., copayment, coinsurance).
These above-mentioned clarifications notwithstanding, we commend the authors on their study, and agree that future studies should assess the impact on clinical outcomes, health-related quality of life, and economic costs associated with the POC transformation in diabetic screening and management.
REFERENCES
1. Whitley HP, Hanson C, Parton JM. Systematic Diabetes Screening Using Point-of-Care HbA1c Testing Facilitates Identification of Prediabetes. Annals of family medicine. 2017;15(2):162-164.
2. Gomez-Peralta F, Abreu C, Andreu-Urioste L, et al. Point-of-care capillary HbA1c measurement in the emergency department: a useful tool to detect unrecognized and uncontrolled diabetes. International journal of emergency medicine. 2016;9(1):7.
3. Paknikar S, Sarmah R, Sivaganeshan L, et al. Long-Term Performance of Point-of-Care Hemoglobin A1c Assays. Journal of diabetes science and technology. 2016;10(6):1308-1315.
4. Torregrosa ME, Molina J, Argente CR, Ena J. Accuracy of three hemoglobin A1c point-of-care systems for glucose monitoring in patients with diabetes mellitus. Endocrinologia y nutricion : organo de la Sociedad Espanola de Endocrinologia y Nutricion. 2015;62(10):478-484.
5. Whitley HP, Yong EV, Rasinen C. Selecting an A1C Point-of-Care Instrument. Diabetes spectrum : a publication of the American Diabetes Association. 2015;28(3):201-208.
Competing interests: None declared
Show Less

Competing Interests: None declared.