Abstract
This prospective longitudinal study compares diabetes screenings between standard practices vs systematically offered point-of-care (POC) hemoglobin A1c (HbA1c) tests in patients aged 45 years or older. Systematically screened participants (n = 164) identified 63% (n = 104) with unknown hyperglycemia and 53% (n = 88) in prediabetes. The standard practice (n = 324) screened 22% (n = 73), most commonly by blood glucose (96%); 8% (n = 6) and 33% (n = 24) were found to have diabetes and prediabetes, respectively. The association between screening outcome and screening method was statistically significant (P = 0.005) in favor of HbA1C. HbA1c may be the most effective method to identify patients unknowingly living in hyperglycemia. Point-of-care tests further facilitate screening evaluation in a timely and feasible fashion.
INTRODUCTION
Millions of Americans are unknowingly living with chronic hyperglycemia. Early identification through a reliable and convenient screening test may lead to prompt identification, intervention, and improved health outcomes.1 Fasting blood glucose measured through a chemistry panel is inconvenient, highly variable, and delays care, unlike HbA1c point-of-care (POC) tests. This investigates the population of adults 45 years of age and older for frequency of diabetes screening by standard practices vs a systematically offered HbA1c POC test and determines differences in identifying unknown chronic hyperglycemia in a single-physician family medicine clinic.
METHODS
Patients with scheduled medical appointments on Tuesdays (active screening arm) and Wednesdays (standard practice arm) between April 2013 and March 2014 were evaluated by the clinical pharmacist. Patients’ electronic medical records (EMR) were assessed for eligibility criteria: those aged 45 years or older were included; pregnancy, past medical history of diabetes (type 1 or type 2), HbA1c test in the past 12 months, or steroid use (injectable or oral) in the past 3 months were exclusion criteria. Those meeting EMR criteria in the active screening arm were offered a free POC HbA1c test by the clinical pharmacist during their scheduled medical appointment after verbally confirming eligibility.2 The EMR of eligible “standard practice arm” patients was assessed for diabetes screenings under usual care. Study characteristics were examined with descriptive statistics; proportional differences were evaluated using a Pearson’s χ2 analysis of independence.
RESULTS
The standard practice arm evaluated 709 patients; however, 324 were included in final analysis per exclusion criteria. The active screening arm evaluated 689 patients. Exclusion criteria removed 390, and 117 were unreachable due to logistics; 182 were offered participation, 17 declined, and 164 were screened. (Supplemental Appendix 1, http://www.annfammed.org/content/15/2/162/suppl/DC1.) Demographics were similar between the 2 study arms (Table 1). The majority was Caucasian (87%) and was comprised of more females (55%) than males. Age ranged from 45 to 91 (mean 63) years. They were predominately obese with an average height, weight, and BMI of 1.69 meters, 86.5 kg, and 31.0 kg/m2, respectively.
In the active screening arm only 37% (n = 60) of patients had an HbA1c ≤5.6%. Whereas 53% (n = 88) had an HbA1c in the prediabetes range (5.7%–6.4%) and 10% (n = 16) of patients had an HbA1c ≥6.5%. The average HbA1c values for patients with euglycemia, prediabetes, and diabetes were 5.34%, 5.93%, and 6.68%, respectively.
The standard practice arm tested 73 (22%) patients, most commonly via blood glucose (n = 70, 96%). Only 4 individuals were screened by HbA1c test; 1 of whom received both HbA1c and blood glucose. Through this process, 33% (n = 24) were in the prediabetes glycemic range, and 8% (n = 6) tested in the diabetes glycemic range. The majority was euglycemic (n = 43, 59%).
Glycemic outcomes were compared between the active and standard practices arms. The χ2 analysis (Table 2) showed that the glycemic outcomes vs methods (active vs standard) were statistically different from one another (P = 0.005).
DISCUSSION
This study demonstrates that systematically screening adults (aged ≥45 years) for diabetes using a POC HbA1c test, vs standard practices, greatly increases the chances for a screen to occur (P = 0.005). Outcomes show that 63% (n = 104) of patients systematically screened were unknowing living in chronic hyperglycemia compared to 41% (n = 30) under standard practices. Interestingly, the most profound impact was in ability to identify prediabetes (53% vs 33% respectively). Under standard practice, blood glucose was the most common screening method used (96%, n = 70), which was often captured as part of a larger venipuncture chemistry panel. The 22% difference in hyperglycemic detectability between screening arms is best explained by test utility.
HbA1c is durable and more accurately reflects sustained hyperglycemia over a 3-month period as compared to fasting or random glucose. Diagnosis by glucose alone may initially be missed until the glycemic curve fully shifts upward, crosses the diagnostic threshold, and bothersome symptoms develop, impacting daily life. Therefore, HbA1c may be a superior screening method,3 due to effectively identifying individuals early on in the course of the disease, which accounts for this study’s difference in identified hyperglycemia. Furthermore, POC HbA1c devices improve patient care by allowing for immediate assessment, patient education, and early management initiated by the practitioner.4–6 Identifying and treating chronic hyperglycemia early can result in clinically meaningful patient outcomes, which is most feasible by HbA1c screenings and further facilitated by POC devices.
The American Diabetes Association recommends screening adults aged ≥45 years or those younger with at least 1 risk factor.1 Conversely, United States Preventative Services Task Force (USPSTF) recommendations focus on overweight or obese patients aged 40 to 70 years.7 Post-hoc analysis applying USPSTF recommendations reduced screenings from 164 to 104 patients, with similar results: average HbA1c 5.78%, 10% HbA1c ≥6.5%, and 50% HbA1c 5.7%–6.4%, but resulted in a higher average HbA1c for outcomes ≥6.5% vs ADA criteria. (Table 1) Tighter USPSTF screening parameters, however, missed identifying 36 patients with HbA1c 5.7%–6.4% and 6 with HbA1c ≥6.5%. Regardless of guidelines used, the post-hoc analysis shows that systematically screening patients is more effective than standard screening practices.
Limitations
While this investigation effectively articulated the frequency of diabetes screenings conducted under standard practice compared to a systematic approach and the outcomes of those screenings, it did not investigate what prompted or prevented the clinicians from screening patients. Future research should focus on (1) how to improve screening programs for chronic hyperglycemia, (2) methods to facilitate patients’ medical and personal behaviors to safely, effectively, and efficiently lower their glycemic curve, and (3) benefits of these screening and health initiatives on outcomes, including reductions in microvascular complications, economic cost impact, and change in quality of life.
Acknowledgements
Special thanks to the Diabetes Hands Foundation for funding the study and to Meg Sheppard, PhD, University of Alabama Institute for Rural Health Research for her support in developing the data collection form, and Larry Skelton, MD for welcoming this project into his rural family medicine clinic in Moundville, Alabama. Thanks to the Auburn University Institutional Review Board whom reviewed this protocol and approved it as expedited.
Footnotes
Conflicts of interest: authors report none.
Funding support: This project was funded through the Diabetes Hands Foundation through the Big Blue Test.
Prior presentations: Whitley HP, Hanson C. Improving the diagnostic rate of diabetes and prediabetes in an underserved rural clinic using a structured screening method and point-of-care A1C device. Presented at the American Diabetes Association 29th Annual Clinical Conference on Diabetes; May 22–25, 2014; Marco Island, Florida.
Supplementary materials: Available at http://www.AnnFamMed.org/content/15/2/162/suppl/DC1/
- Received for publication April 26, 2016.
- Revision received October 4, 2016.
- Accepted for publication October 16, 2016.
- © 2017 Annals of Family Medicine, Inc.