Journal Article

O19 Modelling inclusion of travel time when determining graft allocation in the UK national liver offering scheme

Webb, Gwilym, Stutchfield Ben

2025

Overview

The UK system allocates donor livers to the blood group and weight compatible individual waitlisted with the highest predicted incremental survival (transplant benefit score, TBS) for each organ under consideration. It does not account for travel time between donating and transplanting centre. Longer travel has implications for cold ischaemic time, energy use and cost. This study simulates including travel time in organ allocation.A probabilistic Monte Carlo simulation was developed incorporating pseudorandom determination of organ offers and recipient characteristics re-sampled from historic data. Pre- and post-transplant survival was estimated using the TBS algorithm. Hospital-specific data on donation and implantation activity 2023–2024 were used for proportional allocation; travel time via Google Maps API. Simulation parameters were set to an initial waitlist of 500 and to transplant approximately 1000/year. Each year-long simulation was repeated 20 times using different pseudorandom number generator seeds with and without allowance for donor-centre travel time. Travel time allocation was to the nearest centre within a variable TBS margin of the national top-ranked recipient.Baseline travel time was estimated at median 2.8 (IQR 1.7–4.5) hours/organ. Travel time was lowest for Birmingham 2.5 (1.8–3.0) hours/organ and highest for Edinburgh at 6.5 (4.7–7.5) hours/organ. In simulation, the median gap in TBS days between first and second ranked compatible recipients was 39.9 (10.0–174.4) days; for travel time 1.6 (0.4–2.9). There was no correlation between these (r2=0.001).Incorporating travel time with a 25-day margin was associated with median travel time reducing to 2.4 (1.3–4.1) hours with no change in predicted mortality. Travel time was reduced in all centres. Further extensions of the TBS margins produced further reductions in travel time (p<0.0001) but not outcomes as assessed by deaths pre-transplant or population life years (p=NS) at 20 simulations [table 1].In simulation, minimising travel time by allocating within a margin from the top-ranked potential recipient by TBS resulted in significant reductions in total travel. At smaller TBS margins and the assessed number of simulations, there was no significant difference in simulated population outcomes. Geography and travel time should be considered for addition to the allocation algorithm.Abstract O19 Table 1 TBS Margin (d) Median travel time (h) Total deaths Deaths pre-transplant Population life years 0/baseline 2.8 (1.7–4.5) 161 (155–168) 118 (114–126) 998.9 (986.2–1012.9) 25 2.4 (1.3–4.1) 161 (155–168) 118 (114–125) 996.9 (984.4–1013.3) 50 2.2 (1.2–3.9) 161 (155–172) 121 (114–124) 999.1 (984.3–1013.3) 75 2.0 (1.1–3.5) 165 (157–170) 121 (116–127) 996.6 (983.8–1013.1) 100 1.8 (1.1–3.3) 164 (158–169) 121 (116–126) 995.7 (985.4–1012.3) 150 1.7 (0.9–3.0) 164 (160–171) 122 (117–127) 996.6 (984.1–1011.2)

• Publisher: BMJ Publishing Group LTD
• Subject: Algorithms / Allografts / Blood groups / Ischemia / Liver transplantation / Monte Carlo simulation / Survival / Time allocation