Abstract P061

Variation in the delivery of radiotherapy as part of frontline curative treatment for Hodgkin lymphoma across healthcare providers in England

Background: The addition of radiotherapy (RT) to chemotherapy confers superior disease free-survival in limited-stage Hodgkin lymphoma (HL). However, the consequences in terms of late effects are currently unclear. Given this uncertainty, we seek to understand the extent to which receipt of frontline RT varies as a result of the provider at which a given patient receives treatment in England.

Methods: Cancer registry data was obtained for all classical HL patients diagnosed 1st Jan 2014 to 31st Dec 2020 in England. Multivariate logistic regression was used to assess associations between patient characteristics (age, sex and Index of Multiple Deprivation (IMD) quintile) and odds of receiving frontline RT. Greater than expected variation across provider (NHS Trust) in the case-mix adjusted rate of delivery was assessed via funnel plots. A hierarchal logistic regression with random intercepts for treating NHS Trust was specified and a likelihood ratio test performed to assess improvement of fit. Variation across NHS Trusts was quantified through the variance partition coefficient (VPC) and median odds ratio (MOR).

Results: 2019 of 9743 HL patients treated at 128 different NHS Trusts received frontline RT. The percentage receiving RT stayed consistent at 20% across the 7 years, ranging from 23% (2015) to 19% (2018). The case-mix adjusted rate of RT delivery was outside 2δ (95%) control limits for 33% of NHS Trusts (10 above, 32 below). Hierarchal specification led to a statistically significant increase in goodness-of-fit. Both suggestive of hospital-level effects. Being of male sex had a positive effect on the odds of receiving RT (OR = 0.122, p = 0.095). Similarly, patients in the least deprived IMD quintile had an increased odds of receiving RT (OR = 0.223, p = 0.010) compared to the most deprived. Older age at diagnosis had a non-statistically significant negative effect on the odds of receiving frontline RT (OR = -0.002, p = 0.092). The resulting VPC estimate suggests 10% of variation in the odds a patient receives RT is attributable to the NHS Trust-level. The increase in the MOR of receiving RT were the same patient to move from a lower-RT delivery rate NHS Trust to a higher-RT delivery rate NHS Trust was 1.405.

Conclusions: Healthcare providers had a statistically significant influence on the odds of receiving frontline RT. This effect size was greater than that of patient sex. Improved knowledge to allow optimal patient selection for RT is required.