OBJECTIVES : Severe hypoglycemia (SH) is a potentially life-threatening event that requires assistance from others, often resulting in Emergency Medical Services (EMS) and/or Emergency Department (ED) visits. Nasal glucagon (NG) has been demonstrated to be associated with more successful administration of glucagon than Injectable Glucagon (IG). A cost-offset model was developed to compare SH treatment costs for NG and IG, per SH event for which treatment was attempted with glucagon. Probabilistic sensitivity analysis (PSA) was conducted using Monte Carlo Simulation (MCS).

METHODS : The cost-offset model was developed based on literature review, treatment practices, usability studies and claims data, and covered treatment decisions, outcomes, health care resource use and costs. Deterministic sensitivity analyses were performed, where each variable was varied one at time by +10% and −10% changes relative to base case values. One thousand MCS were conducted using random sampling of values from the range of ±10% change relative to base case values. Normal, gamma, triangular and uniform distributions were used for the MCS, based on the characteristics of individual variables.

RESULTS : The base case from a US Medicare perspective showed average cost-savings of $1,008 per SH event with treatment attempted with NG versus IG, when modeled for costs of medications, EMS and ED. In deterministic sensitivity analyses, ED costs for IG followed by probability of successful administration with NG were identified as driving the highest variations in cost-savings assessments. PSA calculated a lower bound of cost savings of $645 for NG relative to IG and upper bound of $1,283, with 25th percentile being $866 and 75th percentile being $1,008.

CONCLUSIONS : Testing 1,000 alternate scenarios using MCS in the cost-offset model supported the robustness of the base case of NG lowering EM costs relative to IG.