radiotherapy capacity required in scotland in 2015 sara c erridge carrie featherstone
TRANSCRIPT
Radiotherapy capacity required in Scotland in 2015
Sara C Erridge
Carrie Featherstone
Scottish RT models
RT capacity required depends on
Number of new cancer patients Types of cancer Proportion needing RT Fractionation schedule used Complexity of treatments e.g. IGRT Spare capacity to avoid waiting times /R&D etc
Scottish Models
In 2005 Scottish Activity Report suggested that by 2015 potential 38% to 81% increase in machine capacity required
270,000 and 354,000 fractions Given as total number of fractions as can be
delivered on variable number of machines depending on hours worked, through-put etc
Scottish Models –weaknesses
Some sites no Scottish data could be identified (e.g. prostate) – we had to use NYCRIS or the Australian data
Predictions of future can be affected by other factors e.g. informal PSA screening
Models are just models……… some will always work better than others
Calculating demand
Proportion of patients who fit into different cancer scenarios
Population-based (>90% cases) Factors
Cancer type Stage Performance status Co-morbidity Patient/physician preference
Difficu
lty
Revised models
Step 1 – predicting the number of cancer patients
Data and methods
Historical incidence data 1978-2007 GRO population estimates 1978-2007 GRO population projections 2008-2022 Age-Period-Cohort models using software
developed by the Norwegian Cancer Registry (NORDPRED)
Conducted by Sam Oduro, David Brewster, Roger Black @ISD
Lung cancer risk varies by birth cohort
Source: Swerdlow et al, 1998
Age-Period-Cohort models
Model works by estimating the simultaneous influence of age, period and birth cohort on individual risk
Cohort effects highly predictable Future period effects not so, e.g. new
screening programme, diagnostic test
Figure 6 Actual and projected total population compared with previous projections, 1983-2033
0
1,000
2,000
3,000
4,000
5,000
6,000
1983 1988 1993 1998 2003 2008 2013 2018 2023 2028 2033
Year
Per
son
s ('
000s
)
Actual 2008-based 2006-based 2004-based 2003-based
Source: GRO(S)
We have consistently underestimated population growth in Scotland …
Source: GRO(S)
… and most of the growth is in people aged 75+
Actual Projected2003-2007 2013-2017 % change
All Cancer(excluding C44) 27352 32256 18Lung 4692 4691 0Kidney 717 912 27Stomach 800 741 -7Melanoma of skin 927 1350 46Leukaemia 651 811 25Colorectal 3513 4202 20Colon 2332 2748 18Rectum 1180 1453 23Brain & CNS 382 398 4Head & neck 1040 1288 24Bladder 1610 1824 13Oesophagus 832 1030 24Hodgkins 141 156 11Non-Hodgkins 916 1212 32Panceas 634 711 12Other 3152 3690 17Breast 4004 4820 20Corpus Uteri 518 610 18Cervix 285 242 -15Ovarian 623 694 11Prostate 2577 3805 48Testis 197 221 12
Results(average number of new cases per annum)
ValidationActual Projected
2003-2007 2003-2007 % differenceAll Cancer(excluding C44) 27352 28090 3Lung 4692 4481 -4Kidney 717 714 0Stomach 800 864 8Melanoma of skin 927 865 -7Leukaemia 651 767 18Colorectal 3513 3765 7Colon 2332 2458 5Rectum 1180 1307 11Brain & CNS 382 404 6Head & neck 1040 1153 11Bladder 1610 1636 2Oesophagus 832 878 5Hodgkins 141 123 -13Non-Hodgkins 916 991 8Panceas 634 623 -2Other 3152 3295 5Breast 4004 4073 2Corpus Uteri 518 505 -3Cervix 285 312 10Ovarian 623 676 9Prostate 2577 2496 -3Testis 197 234 19
Comment on results
Results reasonably reliable for most of the common cancers and the total
Concern about prostate cancer (period effect) We usually adopted the most conservative
results from the range of model predictions Model choice is always to some extent
subjective
Conclusion
New cancer cases expected to increase by approximately 18% every 10 years
This is mainly due to our aging population Results for Scotland compatible with those
for England (33% increase from 2001 to 2020).