Model-based economic evaluation of the effectiveness of "'Hypos' can strike twice", a leaflet-based ambulance clinician referral intervention to prevent recurrent hypoglycaemia

Abstract

"'Hypos' can strike twice" (HS2) is a pragmatic, leaflet-based referral intervention designed for administration by clinicians of the emergency medical services (EMS) to people they have attended and successfully treated for hypoglycaemia. Its main purpose is to encourage the recipient to engage with their general practitioner or diabetic nurse in order that improvements in medical management of their diabetes may be made, thereby reducing their risk of recurrent hypoglycaemia. Herein we build a de novo economic model for purposes of incremental analyses to compare, in 2018-19 prices, HS2 against standard care for recurrent hypoglycaemia in the fortnight following the initial attack from the perspective of the UK National Health Service (NHS). We found that per patient NHS costs incurred by people receiving the HS2 intervention over the fortnight following an initial hypoglycaemia average £49.79, and under standard care costs average £40.50. Target patient benefit assessed over that same period finds the probability of no recurrence of hypoglycaemia averaging 42.4% under HS2 and 39.4% under standard care, a 7.6% reduction in relative risk. We find that implementing HS2 will cost the NHS an additional £309.36 per episode of recurrent hypoglycaemia avoided. Contrary to the favourable support offered in Botan et al., we conclude that in its current form the HS2 intervention is not a cost-effective use of NHS resources when compared to standard NHS care in reducing the risk of hypoglycaemia recurring within a fortnight of an initial attack that was resolved at-scene by EMS ambulance clinicians.

Fig 1. Decision tree structure of the economic model.

(A) stage 1: responses to initial hypoglycaemia. (B) stage 2: responses to recurrent hypoglycaemia. (C) stage 3: secondary care outcomes. Full tree formed by mapping stage 3 onto every non-terminating node of stage 2, then that combination mapped onto every stage 1 branch.

Fig 2. Cost minimisation model.

Model is depicted by patient types in the HS2 intervention arm and in the standard care arm. Once the first fortnight-length cycle concludes exposure to risk of recurrent hypoglycaemia in each subsequent cycle replicates until GP attendance, where the latter defines the absorbing event for the Markov chains.

Fig 3. Tornado diagram for selected sensitivity analyses.

One-way sensitivities of baseline ICER due to low/high variations: (i) HS2 intervention cost (+£6.70 to +£18.43; baseline £9.95), (ii) Hypoglycaemia recurrence rate r_HS2 (0.375 to 0.9; baseline 0.75), (iii) Severe recurrent hypoglycaemia rate (3.5% to 5.5%; baseline 4.5%).

Fig 4. Cost-effectiveness plane (baseline value shown in red).

Cost differential between care under HS2 intervention and standard care plotted against number of recurrent hypoglycaemia cases avoided. 10,000 simulated pairs in which recurrent hypoglycaemia incidence and severity rate vary. Baseline (red) cost difference +£929 for +3.00 avoided cases.

Fig 5. Probability HS2 is least total cost versus standard care by completed cycle, for GP consultation rates 0.3, 0.4 and 0.5.

Estimates of probability that care under HS2 intervention has lesser cost than standard care after n = 1,2,3,4,5 fortnight-length cycles, by absorption probability pGP = 0.3,0.4,0.5. Estimates depicted are proportion of 10,000 simulations in which first-cycle recurrent hypoglycaemia incidence and severity rate vary, then replicate in up to n-1 subsequent cycles until HS2 is least cost.