Objective: The objective of the present study was to evaluate the adherence, safety, and effectiveness of a paper versus an electronic insulin infusion protocol.
Design: This quasi-experimental implementation study compared experimental and control groups using a nonrandomized prospective cohort design.
Setting: The study was performed at 2 surgical units within a federal tertiary care teaching hospital.
Sample: Fifty-eight registered nurses volunteered.
Methods: We compared time intervals using electronic time stamps from glucometers and insulin infusion devices to measure protocol adherence. We assessed perceived adherence using a nurse survey, and, to evaluate safety, we reviewed each paper protocol infusion calculation for correctness.
Findings: Median times from blood glucose acquisition to infusion rate adjustment did not differ significantly between groups (P = .215). The majority of infusions (96.6%) had glucose acquisition times within the acceptable range. Median values of time to next "glucose due" did not differ significantly (P = .88), and relative variation in median glucose reporting times did not differ significantly between groups (P = .16). Evaluation of 877 paper protocol entries demonstrated a 10.7% (n = 94) calculation error rate. Registered nurses within the electronic group reported greater ease in balancing workload when compared with paper protocol use (P = .03). Attitudes did not differ significantly between groups in areas of determining infusion adjustment, bolus insulin dose, next glucose due, ease of access, understanding protocol, or overall satisfaction.
Conclusions: This study demonstrates that one can adhere to an insulin infusion protocol, regardless of the format (paper or electronic), in the medical-surgical setting. Our results suggest there are safety and nurse workload benefits when an electronic protocol was used.
Implications: Adherence, safety, and effectiveness can be achieved when using insulin infusion in the medical-surgical setting.
Author Affiliations: Medical-Surgical Clinical Nurse Specialist, Portland Veterans Affairs (VA) Medical Center, Oregon, and Adjunct Faculty at Oregon Health & Science University, Portland (Dr Locke); Registered Nurse and Certified Diabetes Educator (Ms McEuen); Computer Specialist, Portland VA Medical Center, Portland (Dr Felder); Hospitalist and Informatician, Portland VA Medical Center, Oregon, and Assistant Professor of Medicine and Medical Informatics, Oregon Health & Science University, Portland (Dr Lesselroth); Hospitalist, Portland VA Medical Center, Oregon, and Assistant Professor of Medicine, Oregon Health & Science University, Portland (Dr Barrett); Staff Nurse and Informaticist, Portland VA Medical Center, Portland (Mr Stevens); Nurse Scientist and Epidemiologist, Portland VA Medical Center, Oregon (Dr Pope).
This study was funded through the Quality Enhancement Research Initiative (QUERI), Office of Research and Development, Veterans Health Administration ($27 000).
The authors report no conflicts of interest.
Correspondence: Christy Locke, DNP, CNS, CNOR, 22117 NE 267th Ct, Battle Ground, WA 98604 (Christine.firstname.lastname@example.org).