Inequitable use of the Cancer Drug Fund (CDF) according to region of England and Primary Care Trust (PCT)


Year:

Session type:

James Clark1, Peter Jones2
1University of Cambridge School of Clinical Medicine, Cambridge, UK, 2Health Services Unit, Keele University, Keele, UK

Background

The CDF was created in 2010 to provide additional funding for cancer drugs in England and was administered until April 2013 by 10 separate regional CDFs. This audit examines application rates between and within these regions.

Method

Freedom of Information requests were sent to all CDFs requesting application rates by PCT for the 3 most used drugs (bevacizumab, cetuximab in colorectal, abiraterone in prostate cancers) and for all drugs between October 2010 and February 2013. Data were received from 4 CDFs: South East Coast(SEC), South Central(SC), North West(NW) and Yorkshire & Humber(YH). Applications rates/million and the latest average score for Index of Multiple Deprivation (IMD) were recorded by PCT.

Results

The number of cohort indications for routine CDF funding in 2013 differed: 33 in SEC, 46 in SC, 57 in NW, 35 in YH. Mean rates of application/million for all drugs differed between regions with substantial intra-regional variation: 738 [SD 176] in SEC, 562 [213] in SC, 381 [114] in NW and 134 [63] in YH. The YH application rate was significantly lower than the other 3; rates for SEC and SC were significantly higher than NW and YH.Greater variation was noted within CDFs for the 3 individual drugs.

The correlations (r) of total application rates per PCT and IMD were -0.41 for YH (p=.13),-0.61 for NW (p=.002), -0.85 for SC (p=.02), -0.87 for SEC (p=.005). For NW and YH combinedr=-0.27 ( p=.07); and for SC and SEC combined, r= -0.71 (p=.003); and for all data r= -.59 (p<.001).

Conclusion

There is substantial variation in CDF application rates both within and between regions. The relation between PCT IMD and CDF use appears stronger in southern than northern CDFs. Many factors contribute to these variations but socio-economic inequality appears to be an important predictor. These findings deserve further investigation.