Date of Award

Fall 1-8-2021

Degree Type


Degree Name

Doctor of Philosophy (PhD)


Public Health

First Advisor

Ike. S. Okosun

Second Advisor

Ruiyan Luo

Third Advisor

David Ziemer


Introduction: Hypoglycemia is the most common acute adverse complication of glucose lowering therapy among patients with diabetes and is associated with poor outcomes, including death. A recent study reported that the rate of hypoglycemia-related hospitalization in the United States increased by 11.7% (from 94 to 105 admissions per 100,000 person years) between 1991 and 2011 and that total annual costs increased to $1.6 billion. Despite these concerns, diabetes care quality established a decade ago has primarily focused on prevention and reducing rates of complications such as hypoglycemia. Although these efforts have been successful and the proportion of patients achieving HbA1c levels below a 7% threshold improved significantly, consequences of these changes may include increased hospital admission rates of hypoglycemia, which now exceed rates of hyperglycemia. Overall, the three studies presented provide an understanding of the risks of hypoglycemia and multifaceted interventions that reduce the rate of hypoglycemia in the hospital setting and U.S. population.

Study One: Study one is a literature review of multifaceted interventions that have used an educational component to reduce the rate of hypoglycemia in inpatient settings. The literature search used a combination of search terms, which include ‘inpatient hypoglycemia’, ‘inpatient low glucose’ ‘educational prevention’, and ‘educational intervention’. Fifteen articles were selected for the review. Three articles focused on the safety and efficacy of insulin order sets, three articles implemented structured insulin order sets and protocols, seven studies focused on computerized insulin order sets and alert systems and two studies focused on root cause analysis and alert systems. The review suggested a need for randomized control trials and interventions that target at-risk populations, interventions that focus on the root causes of hypoglycemia and interventions that use technological enabled education tools (i.e. mobile health). The review also found that education in combination of multifaceted intervention that have used technology have been successful in reducing the rates of hypoglycemia.

Study Two: Study two is a pilot randomized control study that examined the risk factors of hypoglycemia and provided a tailored feedback intervention to address those risk factors in the hospital setting. In this RCT study, a convenient sample of diabetes patients (N=85) were recruited in Atlanta. Baseline surveys were conducted to administer survey instrument. Nine months after recruitment, electronic medical reviews were conducted to determine whether the intervention reduced the incidence of hypoglycemia. Univariate and multivariate logistic regression analysis was conducted to determine the predisposing risk factors of hypoglycemia. Logistic regression analysis was also conducted to determine whether the intervention reduced the incidence of hypoglycemia. Results indicated that insulin use and older age (60+) were significant predictors of hypoglycemia occurrence. Results also indicated that the intervention reduced the incidence nine months post recruitment (P=0.046). Ongoing clinician education regarding insulin use and dosing and educational tools and resources targeted for an older population could continue to lower the rate of hypoglycemia occurrence.

Study Three: Study three is a cross-sectional analysis of data on 1,642 older adults (≥ 65 years) with diabetes from the National Health and Nutrition Examination Survey (NHANES) from 2009 to 2018 who had an HbA1c measurement. The study examined glycemic control levels among older adults with diabetes mellitus by health status and to estimate the potential of overtreatment with hypoglycemia causing medications (insulin or sulfonylureas). A weighted proportion of survey participants with glycemia that was poorly, moderately or tightly controlled across health status categories was conducted. Weighted proportions of participants whose glycemia was tightly controlled and were treated with insulin or sulfonylureas was also conducted. Logistic regression analysis was conducted to assess linear trends in participants with tightly controlled glycemia, their health status and patterns of treatment during the five NHANES surveys. Results indicated that during the 10 years there were no significant changes in proportion with an HbA1c level less than 7% who had complex/intermediate or very complex/poor health (P=0.444). There were significant changes in the proportion with an HbA1c <7% and were treated with insulin or sulfonylureas who had complex/intermediate or very complex/poor health (P=0.005). Therefore, findings indicated overtreatment in participants with complex/intermediate or very complex/poor health. Quality and overtreatment measures can reduce the risk of complications such as hypoglycemia for glycemic control.

Conclusion: Collectively, the three studies in this dissertation describe risk factors for hypoglycemia and multifaceted interventions that have used an educational component to reduce the rate of hypoglycemia in the hospital setting and the U.S. population. Implications suggest that there is a shift toward determining the root causes/risk factors of hypoglycemia and that there is a need to educate, individualize and tailor treatment to the patient to successfully reduce the rates of hypoglycemia. However, there are gaps in clinician education on insulin use and other hypoglycemia causing medications, which can lead to overtreatment. Therefore, there is a need in the assistance of a diverse and specialized diabetes management team that can include education and effective communication in the physician workflow. Furthermore, overtreatment measures, technological advancements (i.e. mobile health, smartphones) and improvements in electronic medical records to flag hypoglycemic events can further reduce hypoglycemic events and overall healthcare costs.


File Upload Confirmation