A183outcomes. One-way, two-way, and probabilistic sensitivity analyses were used to examine uncertainty in the estimates. Results: Clozapine dominated the three therapies it was compared to and had negative ICERs compared to each of these therapies: -£8,773/QALY, -£20,659/QALY, and -£2,226/QALY, compared to olanzapine, quetiapine, and risperidone respectively. The results from the one-way and two-way sensitivity analyses suggested the findings were insensitive to changes in single model parameters of the top five key drivers and double model parameters of the top two key drivers for all treatment scenarios. The mean ICER values from the probabilistic sensitivity analysis were -£8,737.40/QALY for clozapine versus olanzapine, and -£20,562.56/QALY and -£2,020.44/QALY for quetiapine and risperidone respectively. The probability of clozapine being cost effective compared to olanzapine at a threshold value of £20,000/QALY was 99.98% and 98.44% compared to risperidone. However the probability of clozapine being cost effective compared to quetiapine was 100% at any threshold value. ConClusions: Clozapine was the dominant treatment for treatment-resistant schizophrenia as it was associated with the lowest costs and highest QALYs. Based on these findings the treatment strategy for treatment-resistant schizophrenia ought to be reviewed and clozapine considered after one unsuccessful trial with any other antipsychotic drug.

S121 studied. MethodS: A retrospective analysis was conducted of claims data, acquired from third party funders that contract the ICON Network, for the year 2016. The data (n 1 = 9444) was governed by a requirement to adhere to treatment protocols. The control group (n 2 = 2008) was under no protocol restriction and seen by a different group of oncologists. The hospital admission rate and total stay duration were measured per patient where the treating doctor for the hospital event was an oncologist. Case mix was adjusted by diagnosis (using admission ICD10 code). The smaller cohort dataset (n2) was extrapolated proportionally by diagnosis (random sampling with replacement). The same methodology was applied to 2015 data to validate the methodology. ReSultS: The hospitalisation rate for patients under protocolised treatment, 16.07%, was significantly less (99% CI, p< 0.0022, Chi-squared test) than the control group, 17.64%. The average hospital stay duration, 1.62 days, for patients under protocolised treatment was also significantly less (99% CI, p< 0.0022, student's t test) than the control group, 1.95 days. The results were validated against data from 2015 (n 1 = 8559, n 2 = 1870): that showed significantly lower hospitalisation rate and hospital stay duration in comparison to the control group. ConCluSionS: Using hospitalisation metrics as a proxy outcome of patient well-being and cost shifting, the ICON solution demonstrates significant value without compromising care.

No abstract

No abstract

No abstract

No abstract

No abstract

No abstract