Further supplementary evidence submitted
by the Department of Health (WP 01D)
MEDICAL WORKFORCE PLANNING 2005-30
1. This paper brings together analysis of
the future demand and supply of the medical workforce including
analysis of:
— The impact of recent medical school
expansion on medical supply—there will be more doctors in
training to find jobs for;
— The implementation of Modernising
Medical Careers (MMC)—a higher proportion of medical care
will be delivered fully-trained doctors;
— The European Working Time Directive
(EWTD)—doctors in training working hours could be lower;
— The impact of reduced flows of International
Medical Graduates (IMGs) on medical supply—there will be
more job opportunities for the increasing number of UK graduates;
— Future reduction in workforce participation—a
lower contribution from each doctor, creating need more individuals;
— Costs and affordability of growth
in the medical workforce;
— Changes in skill mix—potential
affecting the demand for doctors but lacking hard evidence and
comparative costs;
— Estimates of demand for healthcare—demographic
trends suggest a higher demand.
BACKGROUND
2. Recent expansion of medical school numbers
will mean the medical workforce will continue to expand for the
next 30-40 years until a steady-state is reached. The larger medical
workforce will mean the service will have capacity to meet growing
demand for healthcare whilst gradually reducing reliance on IMGs.
As UK graduates progress through specialty training, the workforce
will be more richly skilled than at present, and average labour
costs will therefore increase commensurately.
3. Over the coming decades factors like
the EWTD, trends to more part-time working, the growing "feminisation"
of the workforce, and demands for a more satisfactory work-life
balance could have a negative impact on the supply of doctors.
4. Other factors, such as implementation
of MMC or future changes in skill mix add further uncertainty
to medical workforce planning.
MEDICAL WORKFORCE
SUPPLY
5. The shape and composition of the medical
workforce is likely to change in coming years.
— Recent increases in the UK medical
school intake will enable the service to become more self sufficient
in doctors, reducing reliance on IMGs, whilst still increasing
the overall size of the workforce.
— As the expanded UK workforce progress
through specialty or GP training, we will have a richer medical
skill mix than present. This implies however, that service currently
delivered at junior level, will in future be delivered by fewer,
more efficient specialist doctors.
6. The Workforce Review Team (WRT) have
modelled a number of medical supply scenarios. The "central"
scenario is set out below.
— Average annual growth in headcount
(HC) trained doctors (CCST, CCT and GPs) 2005-30 is projected
to be about 3.8% per annum. The number of trained doctors will
continue to grow until about 2040.
— WRT project no growth in total numbers
of juniors/non specialist career grade doctors. A small increase
in trainees is offset by a decline in numbers of career doctors.
7. The changing composition of the workforce
implies a radical shift in the way service is delivered. Service
currently delivered by doctors in training will have to be performed
by other healthcare professionals—specialist doctors and
non-medical staff. The chart below sets out the changing medical
skill mix over time. A key question is whether the service is
prepared for and wants the changing size and skill mix of the
medical workforce.
KEY ASSUMPTIONS:
— 2,500 training posts for GPs each
year.
— Total entrants into the system =
UKGs + headroom.
— Specialty training places for all
UKGs + headroom.
PARTICIPATION
8. Changes in the working culture may affect
future participation of the medical workforce.
— Some of the current stock of doctors
trained by working 60+ hours per week, whereas doctors in training
are currently limited to an average maximum working week of 56
hours, which will reduce from 2009 to a maximum of 48 hours. It
is possible that when future doctors in training complete specialty
training they will expect to work less than current consultants
do.
— It is asserted that in the past consultants
have worked longer than their contracted hours "for free",
in future, they may only work their contracted hours, which will
lower participation.
— A growing feminisation of the workforce
may reduce overall participation as doctors take career breaks
or work part time to have children.
— There is a growing societal trend
(for men and women) towards more part time working and taking
career breaks.
9. The chart below shows projected participation
rates for GPs, consultants and career doctors and trainee doctors
used in the analysis. The projections take into account past trends
in participation and likely reduction in hours worked. They project
a relatively quick decline in participation.
MMC
10. The MMC reforms will change the training
and career pathway for doctors. Though the framework for the new
career structure has been agreed, there are several key factors
that remain to be determined. These will affect projected medical
supply:
— The average length of specialist
training in a competency-based system is hard to predict (six
to eight years).
— The proportion of doctors with a
Certificate of Completion of Training (CCT) will increase and
will deliver a higher proportion of medical care. This may knock
on to future demand.
— The pay/cost/contribution of MMC
doctors.
— Whether there will be enough specialist
and GP places for current SHOs and the future contribution of
IMGs.
— How many doctors will take up Fixed
Term Specialist Training Appointments (FTSTA).
— How many doctors will become Non
Consultant Career Grades (NCCGs) and not progress to CCT.
11. To start examining some of these issues
the WRT have produced 28 supply scenarios by varying some of these
and other assumptions (though many more scenarios are possible).
12. The assumptions used as our current
central supply estimate are as follows and is shown in chart 3
above:
— 20% of current SHO posts (about 4,000)
are cut. These could include some of the 8,500 IMG SHOs.
— Average length of specialty training
of seven years.
— No flows into FTSTA or career grades.
13. The precise role of MMC trainee and
specialist doctors is likely to change. This means that their
productivity is likely to be different. Analysts from DAT and
WRT canvassed service experts to estimate the relative productivity
of MMC and current doctors; this is set out in the table in the
appendix. The relative productivity reflects (1) the relative
amount of time each grade of doctor spends providing service rather
than doing other activities (such as providing training, being
trained or administration and management) and (2) the relative
expertise of each grade of doctor. The assumptions are however
debatable and represent our current best estimate for modelling
purposes.
— Run Through (RT) doctors are assumed
in the scenario to be much less productive than SpRs. This reflects
the belief that specialty training for MMC doctors might focus
on learning rather than providing service.
— CCT doctors are assumed to be slightly
less productive than CCST consultants to reflect different levels
of training and expertise.
— Foundation programme year 2 doctors
are assumed to deliver no service, unlike current SHO level 1.
Experience is now however showing that is assumption undervalues
the contribution of F2 doctors.
14. The relative productivity of doctors
enables us to translate demand modelled under the current medical
workforce structure, to demand expressed as numbers of MMC doctors.
WORKING TIME
DIRECTIVE
15. We model the impact of the EWTD by crudely
assuming it will reduce the amount of time that juniors are available
to work from a maximum of 56 hours to 48 hours (about a 14% drop
in the available supply of doctors in training). In practice,
Trusts will adopt various strategies such as service redesign
to meet growing demand in the context of the EWTD.
16. The analysis assumes that the EWTD is
an additional demand pressure—ie additional doctors will
be required to deliver the projected demand for healthcare following
the introduction of the EWTD. New Deal contract monitoring shows
that a third of doctors in training are already working a 48 hour
week. By 2009, when the EWTD is fully implemented, it adds about
5,000 registrars and SHOs to demand, though Trusts will not simply
employ more junior staff to manage the EWTD.
SKILL MIX
SUBSTITUTION BETWEEN
DOCTORS AND
NURSES
17. The current evidence base for using
non-medical staff to substitute for medical staff is equivocal.
There is evidence that skill mix changes either improve patient
outcomes and patient experience, but there is little evidence
that the changes are cost effective in hospital settings. However,
in primary care there is evidence that practice nurses can cost-effectively
substitute for GPs.
18. A recent review by DAT identified many
new or extended roles. As these roles are new, there is little
evidence on their cost effectiveness or impact on demand for medical
staff. New roles included:
— Advanced practitioners such as Medical
Care, Surgical Care, Anaesthesia Practitioners who work alongside
qualified medical staff;
— Emergency Care Practitioners;
— Consultant Allied Health Professionals;
— More assistant practitioners to work
with AHPs;
— Medical assistants and Administrative
assistants to work with doctors;
19. Apart from assistant practitioners and
Consultant AHPs the new roles cannot be identified on the census[1]
so it is difficult to assess their current impact on the delivery
of healthcare.
20. The modelling assumes that in the long
term growth in practice nurse consultations will be faster than
for GPs, as practice nurses perform an increasing proportion of
primary care activity. It is assumed that by 2030 practice nurses
could do about 44% of all primary care activity, up from about
38% now.
WORKFORCE DEMAND
21. Growth in demand for health services
is driven by many factors:
— An increasing population. Total population
growth is projected to be about ½% per year.
— An aging population. The population
growth of the over 65s is projected to be about 1.8% per year.
— Changing culture and lifestyle that
affect morbidity. For example, incidence of obesity are rising,
smoking rates have fallen but may have levelled off.
— Changing public expectations. Rising
living standards mean growing patient expectations.
— Greater medical advances increase
the range of treatments available.
22. We have modelled various scenarios around
healthcare demand (set out in more detail in table 1).
— The lowest assume demand for healthcare
increases only as a result of demographic growth.
— Past trendsin activity can be taken
as an indication of future demand for healthcare. Trends in growth
of activity have been significantly above demographic growth.
— The Wanless review modelled activity
up to 2020 around different scenarios. These all projected activity
growth some way above recent trends.
23. Our central assumption on healthcare
demand is activity to follow Wanless projections up to 2020 then
revert to past trends.
Table 1
KEY ASSUMPTIONS—DEMAND SCENARIOS
Scenario 1:
Demographic growth only
| This scenario assumes that growth in demand for healthcare will only come from demographic growth. Current activity rates per person will continue, ie the impact of changing morbidity and patient expectations will be zero. This is a low growth scenario as historically activity rates per person have been rising faster than population growth.
— Low secondary care growth (0.8%)
— Low primary care growth (0.7%)
|
| Scenario 2: past activity trends continue
| This scenario extrapolates forward past (1990-2003) trends in activity.
— high growth for secondary care (2.5% per year),
— moderate growth of primary care activity (1.2% per year).
|
Scenario 3:
Wanless activity growth assumptions—Fully engaged
| This scenario extrapolates forward long-term (2015-20) activity growth assumptions used in the Wanless review of 2002.
— high life expectancy;
— high improvement in health status;
— high growth for secondary care activity (3% per year);
— 2.0% per year growth in GP primary care activity + practice nurse activity set at 3.3% per year growth.
|
| |
MEDICAL DEMAND
AND SUPPLY
24. We model demand and supply separately for trained
doctors (specialists and GPs) and trainee doctors/NCCGs. Demand
is expressed in MMC workforce grades. Any gap between workforce
demand and supply is presented as numbers of CCT doctors. Charts
5 and 6 show supply and demand below.
25. The modelling projects an over supply of trained
specialist and GP doctors of about 20,000 wtes by 2030 as the
supply of trained doctors increases faster than demand.
26. This is offset by a forecast under supply of trainee
doctors equivalent to about 26,000 CCT trained wtes by 2030. It
is debatable how far (fewer more experienced and productive) CCT
doctors can substitute for the service currently provided by trainee
doctors as this implies much fewer doctors per patient covering
rotas than at present. If the gap cannot be filled by fewer CCT
doctors, the total gap could by much larger[2].
Specialist doctors and GPs
Note: demand and supply in charts 5 and 6 is assessed as
Full Time Equivalent (fte) figures unlike chart 1.
Note: the gap between demand and supply is expressed as numbers
of wte CCT doctors.
Trainee and NCCGs
Note: the gap between demand and supply is expressed as numbers
of wte CCT doctors. If the gap were expressed as numbers of trainee
doctors it would be much larger.
AFFORDABILITY
27. The changing medical supply will have implications
on the paybill. One way of estimating whether the projected supply
of doctors is affordable is to compare the projected pay bill
with growth in NHS expenditure. If pay bill growth is higher than
NHS expenditure, this implies medical pay will take up an increasing
proportion of NHS expenditure and is unaffordable. If pay bill
growth is lower than NHS expenditure, this implies growth in the
workforce is affordable.
28. The table below sets out the projected paybill costs
given the base medical supply scenario and non-medical workforce
growth based on Wanless[3].
It also sets out an estimate of the long run (1971-2007) growth
in annual real NHS expenditure. A trend in expenditure covering
the period since 1999 would show higher average growth. Further
analysis of the trends in NHS expenditure is contained in the
appendix.
Table 2
PROJECTED GROWTH IN PAY BILL
| | With accredited specialist grade paid 75% of a consultant
| All CCT holders become consultants
|
Workforce category | 2004-05
| 2030-31
projected
paybill (2004
prices)
| Average
real p/a
growth | 2030-31
projected
paybill
(2004 prices)
| Average
real p/a
growth |
Medical pay bill | 9,525m
| 21,688m | 3.2% | 24,684m
| 3.7% |
| Non medical pay bill | 24,006m
| 68,416m | 4.1% | 68,416m
| 4.1% |
| Trend in NHS expenditure (1971-1999)—Low
| | | |
| |
| 3.2% | | |
| | |
| Trend in NHS expenditure (1971-2007)—High
| | | |
| |
| 4.1% | | |
| | |
| | |
| | |
Non medical paybill growth assumes staff growth is 2.2% per
year (equal to Wanless demand minus ½%)
Hospital staff pay drift assumed to be 1% per year
GP pay drift assumed to be 0%
Annual pay settlement assumed to be inflation + ½%
Accredited doctor grade assumes newly qualified CCT doctors
earn 75% of current consultants
29. Growth in the medical pay bill of 3.2% and the non-medical
paybill of 4.1% shows that Wanless demand could be affordable
in the long run, only if the long run trend (1971-2007) in real
growth in NHS expenditure continues.
30. The growth in the medical pay bill will depend on
the future pay and grade structure of specialist doctors. Two
scenarios are modelled, one where all new CCT holders are appointed
to doctor grade whose pay is on average 75% of a consultant and
another where all CCT holders are appointed to consultant positions.
CONCLUSION
31. Many variables could affect the supply and demand
of staff over the coming decades, which make projections difficult
to make. However, we can draw the following conclusions from the
analysis.
— Recent medical school expansion means that the
trained medical workforce will continue to expand past 2030 until
reaching a steady state 2040-50.
— The trainee workforce is not projected to grow.
MMC trainees are likely to have lower productivity than current
trainees.
— The service therefore needs to be prepared to
deliver growing activity with a medical workforce with proportionately
fewer junior grades of doctors.
— The expansion in the medical workforce is sufficient
to meet Wanless demand only if fewer, more productive,
CCT doctors can substitute for work currently performed by doctors
in training. Otherwise, there is likely to be excess supply of
CCT doctors and continued imports of junior IMGs.
— Increases in medical workforce would not offset
the impact of WTD by 2009 based on current design of services.
Higher medical supply could offset reductions in supply from the
WTD and falling participation only with service redesign. The
impact of changing participation rates is not certain and should
be monitored and managed closely.
— There are some uncertainties in MMC which might
affect the supply and cost of the medical workforce. These include:
pay, length of training, the popularity of FTSTAs and career grades.
— The evidence that using different skill mix solutions
to substitute non-medical staff for doctors in the secondary care
sector is equivocal. The impact of skill mix particularly from
new roles should be monitored.
— Future growth in NHS Expenditure is unknown; however,
if growth reverts to its 1971-2007 trend then Wanless growth could
be affordable, but lower rates of growth could limit affordability.
Department of Health
11 December 2006
1
Modern Matrons can be identified on the census from 2005. Back
2
For example, if the gap were filled by (less productive) career
doctors it would be about 35,000. If it were filled by (much less
productive) Run Through grades, it would be over 100,000 ftes. Back
3
Wanless growth minus ½% per year. Back
| 
