Patient and Prehospital Predictors of Hospital Admission for Patients With and Without Histories of Diabetes Treated by Paramedics for Hypoglycemia: A Health Record Review Study

Abstract

Objectives

The objectives of this study were to describe the characteristics, management, and outcomes of patients treated by paramedics for hypoglycemia, and to determine the predictors of hospital admission for these patients within 72 hours of the initial hypoglycemia event.

Methods

We performed a health record review of paramedic call reports and emergency department records over a 12-month period. We queried prehospital databases to identify cases, which included all patients ⩾18 years with prehospital glucose readings of <72 mg/dl (<4.0 mmol/L) and excluded terminally ill and cardiac arrest patients. We developed and piloted a standardized data collection tool and obtained consensus on all data definitions before initiation of data extraction by trained investigators. Data analyses included descriptive statistics univariate and logistic regression presented as adjusted odds ratios (aOR) with 95% confidence intervals (95%CI).

Results

There were 791 patients with the following characteristics: mean age 56.2, male 52.3%, type 1 diabetes 11.6%, on insulin 43.3%, median initial glucose 54.0 mg/dl (3.0 mmol/L), from home 56.4%. They were treated by advanced care paramedics 80.1%, received intravenous D50 37.8%, intramuscular glucagon 17.8%, oral complex carbs/protein 25.7%, and accepted transport to hospital 70.2%. Among those transported, 134 (24.3%) were initially admitted and four more were admitted within 72 hours. One patient was admitted, discharged, and admitted again within 72 hours. Patients without documented histories of diabetes (aOR 2.35, CI 1.13–4.86), with cardiovascular disease (aOR 1.81, CI 1.10–3.00), on corticosteroids (aOR 4.63, CI 2.15–9.96), on oral hypoglycemic agent(s) (aOR 1.92, CI 1.02–3.62), or those given glucagon (aOR 1.77, CI 1.07–2.93) on scene were more likely to be admitted to hospital, whereas patients on insulin (aOR 0.49, CI 0.27–0.91), able to tolerate complex oral carbs/protein (aOR 0.22, CI 0.10–0.48), with final GCS scores of 15 (aOR 0.53, CI 0.34–0.83), or from public locations (aOR 0.40, CI 0.21–0.75) were less likely to be admitted.

Conclusions

There are several patient and prehospital management characteristics which, in combination, could be incorporated into a safe clinical decision tool for patients who present with hypoglycemia.

Acknowledgments

Thank you to the University of Ottawa, Department of Emergency Medicine Academic Grant for funding this study. We thank Jason Rouleau for providing administrative support, data abstraction and entry and for reviewing the format of the manuscript. We also thank Jeff Yantha for his assistance with the statistical analysis, Ben de Mendonca for providing feedback, and Dianne Cummins for her help editing the manuscript.

Disclosure Statement

Julie E. Sinclair served as an advisor on the Eli Lilly National Nasal Glucagon Advisory Board. The other authors report there are no potential competing interests to declare.

Prior Presentations

Parts of this work were presented in abstract and oral presentation at the 2016 National Association for EMS Physicians Annual Meeting, San Diego, CA.

Authors’ Contributions

JES was responsible for the conception of the project, methods, data collection, data analysis, draft manuscript, and integrity of the paper as a whole. MAA was involved in data analysis and drafting the manuscript. SL involved in data collected, data analysis, and drafting the manuscript. RD was involved in project development, obtaining funding, data analysis, and drafting the manuscript. MF was involved in project development and drafting the manuscript. JM was involved with data collection and entry, and formatting and reviewing the manuscript, and CV was involved in project development, data analysis, and drafting the manuscript.

Additional information

Funding

This study was funded by a peer-reviewed University of Ottawa, Department of Emergency Medicine Academic Grant.