NICE guidelines for cancer treatment set out recommended treatment for patients. However, how consistently these guideline recommendations are followed in care settings is not known.
These two projects aimed to establish levels of concordance to as many NICE treatment guideline recommendations as possible for both colon cancer and (NSCLC) non-small cell lung cancer. This work also investigated variation in concordance by patient characteristic and Cancer Alliance.
This analysis was done in partnership with NHS England.
Concordance was assessed for two colon cancer treatment recommendations, with concordance ranging from 53.4% to 98.9%.
There was no significant association between gender and concordance, but concordance to each recommendation varied across patient age groups. The pattern of this association was different for each recommendation.
Deprivation was slightly associated with concordance to one recommendation on adjuvant chemotherapy with those living in the most deprived areas the least likely to be guideline concordance.
Those of Black, Asian, or 'Other' ethnicity less likely to have concordant treatment on one recommendation on adjuvant chemotherapy than those of White ethnicity. Those with increasing comorbidity scores were also less likely to receive concordant treatment around adjuvant chemotherapy.
Concordance to recommendations was generally more likely in later years. There was also significant variation between Cancer Alliances.
Read the report on lung cancer in fullConcordance was assessed for four NSCLC treatment recommendations, with concordance ranging from 18.1% to 99.5%.
There was no significant association between gender and concordance, but concordance to each recommendation varied across patient age groups. Patients aged 45-54 and 55-64 were significantly more likely to have concordant treatment compared to those aged 65-74, while those aged 75+ were less likely.
There were mixed results for the relationship between deprivation quintile and recommendation concordance, with some recommendations having no significant association. Where there was an association, it was found that those living in the most deprived quintile were less likely to receive concordant treatment than those in the least
deprived quintile. There were no significant relationships between ethnicity and receiving concordant treatment, where ethnicity was known. Those with increasing comorbidity scores were also less likely to receive concordant treatment around adjuvant chemotherapy.
For two of the four recommendations, concordance was more likely in later years. There was also significant variation between Cancer Alliances.
Read the report on colon cancer in fullThe variation in treatment recommendation concordance across patient characteristics, suggest that there are inequalities in the access to, and provision of, cancer treatment. The significant variation by Cancer Alliance suggests there may be geographical variation in the use of optimal treatments and could potentially highlight areas for improvement.
Our data offers no clear explanation for the link between deprivation and concordance, but many factors could be driving this. These could include the high frequency of hospital attendances required for chemoradiotherapy (so travel cost and time required may be having a greater impact for those most deprived), the effect of morbidity not being accounted for in this analysis, or different responses to treatment benefits and risks by clinicians and/or patients.
Overall, investigating the levels of concordance to the treatment guideline recommendations could help to highlight gaps between recommended and actual practice and suggest potential areas for improving the delivery of evidence-based treatment.
Lorna Wills, Jess Fraser, Scarlett Cartwright-Hughes, Rosie Hinchliffe, Shane Johnson, Sophia Lowes, Fahmina Fardus-Reid, Lucy Ironmonger, Jon Shelton, Lyndsy Ambler, Sam Harrison, Sean McPhail, Lizz Paley, Yoryos Lyratzopoulos.
If you have any comments, questions or feedback, please reach out to our team.
