Select Committee on Health Minutes of Evidence

Supplementary evidence by the Department of Health: Follow up information


  The 1998 White Paper "Smoking Kills" set out an overall strategy on tobacco control including expenditure of some £50m on health education and up to £60m on cessation services.

  Since "Smoking Kills" the Government has made first bupropion (Zyban) and more recently all Nicotine Replacement Therapy products available on NHS prescription and appropriate increases have been made in the unified budgets to meet the expected costs of these developments (£32.5m in 2001/02). [2]This expenditure together with £53m on smoking cessation services and £42.9m on the education campaign means that we will significantly exceed the sums committed in "Smoking Kills". This reflects the need to keep our strategy under constant review


  The Secretary of State offered to provide a note setting out figures on the number of GPs coming through and how many in the same timescale are retiring and whether there is a net gain or loss.

  The first point to make is that all targets set out in the NHS plan represent net increases in the numbers. They take full account of the numbers expected to leave the profession due to retirement. The numbers appointed will be sufficient both to replace those who have left and to secure significant increases.

  The figures in Table 1 below show the trends in GP numbers in recent years. The figures show that we have had continued net growth in the overall numbers of GPs, despite a considerable reduction in the number of GP registrar training places in the early 1990s and despite suggestions that numbers of retirement were set to rise.

  We have now reversed the decline in GP registrar training places and the increases that we have already seen (an increase of over 300 since October 1997), are beginning to feed through. Under the NHS plan, the number of GP registrar training places is set to increase further, by a further 450, to over 2,100.

  Under the NHS plan, the number of GPs (excluding GP registrars and retainers) will increase by 2000 above the October 1999 baseline. Increases in training places will play a part on this, but there will also be a number of other measures, including international recruitment and speeding up appointment of those who have completed their GP registrar training. The figures take full account of those leaving the profession, which (including retirements) account for 3 to 4% of the workforce each year.

Table 1

19901995 19961997 199819992000
General medical practitioners (1) 27,523 28,86929,11629,389 29,69729,98730,252
Unrestricted principals and equivalents (2) 25,62226,70226,855 27,09927,39227,591 27,704
GP registrars 1,5621,404 1,3051,3431,446 1,5201,659
GP Retainersn/an/a n/an/an/a 9721,117
Other practitioners (3)339 763956947 859876889

  Source: Department of Health publication Statistics for General Medical Practitioners in England: 1990—2000.


  (1)  Excludes GP Retainers, data on which were first collected in 1999.

  (2)  Unrestricted Principals and Equivalents (UPEs) includes GMC unrestricted Principals, PMS contracted GPs and PMS salaried GPs.

  (3)  Other Practitioners include Assistants, Restricted Principals, Salaried Doctors (para 52) and PMS Other GPs, but exclude GP retainers.

  We do not have details of the specific reasons why GPs left the workforce and cannot therefore identify which GPs retired (as opposed to, for example, taking a short career break). However Table 2 below shows the balance of joiners and leavers (for whatever reason) for the UPE workforce over the last 10 years.

Table 2

England1990-91 (1) 1991-921992-93 1993-941994-95 1995-961996-97 1997-981998-99 1999-2000
UPEs (2) 25,622 25,686 25,96826,28926,567 26,70226,85527,099 27,39227,59127,704
Joiners1,4961,453 1,3511,3261,384 1,3371,3021,333 1,2291,203
Composed of: new
joiners (not seen before
in a census)
1,3131,2521,134 1,1361,1871,059 1,0291,069967 958
re-joiners (appeared in a previous census) 183201217 190197278 273264262 245
Leavers1,4321,171 1,0301,0481,249 1,1841,0581,040 1,0301,090
Change in UPEs64282 321278135 153244293 199113

Source: Department of Health publication Statistics for General Medical Practitioners in England: 1990-2000.
   (1)   Earliest available data.
   (2)   Numbers of UPEs at 1 October each year from 1990 to 1999 and 30 September from 2000. Other rows relate to changes between successive October/September Censuses.


  The table below contains details of 20 or so common procedures (which, taken together amount to about half of all surgical activity in the NHS) and the number of each procedure which were carried out in 1996/97 and 1999/00 (the latest year available).

  The table shows that the fastest growing procedures are a mixture of relatively straightforward (for example cataract removal), and longer more resource intensive operations such as knee replacements and PTCAs. What characterises all of these fast growing procedures is that they are clinically effective; producing real health gains for patients.

  At the other extreme, those procedures which have become less common since 1996/97 are those which are considered to be less clinically effective in some cases—for example removal of teeth in hospital and drainage of the middle ear.

Increase between 1996-97 and 1999-2000
NoPer cent
Colonoscopy250,090344,373 94,28338
Cataracts160,013213,303 53,29033
Miscellaneous operations including radiograms chemotherapy 728,576950,093221,517 30
PTCA17,04521,968 4,92329
Total prosthetic replacement of knee joint 24,47429,8935,419 22
Endoscopic operations on stomach and intestines 479,246536,89957,653 12
Investigations and minor operations on joints 91,368101,2879,919 11
Total hip replacement64,505 69,6005,0958
Endoscopic operations on bladder280,890 296,20915,3195
Minor skin operations including minor skin cancers 182,464190,5488,084 4
Mastectomy52,73753,823 1,0862
Inguinal hernia81,398 81,7583600
Operation on cervis uteri80,775 76,148-4,627-6
Varicose vein54,470 49,080-5,390-10
Removal of teeth95,069 84,689-10,380-11
Fallopian tube operations139,860 119,566-20,294-15
Hysterectomy72,23260,955 -11,277-16
Drainage of middle ear74,259 61,000-13,259-18
Tonsillectomy75,407 61,449-13,958-19
Other evacuation of contents of uterus 144,577117,611-29,966 -19
Endoscopic operations on outlet of male bladder 55,72339,339-16,384 -29
Curretage of uterus87,408 43,212-44,196-51


  The figures for 1989-90 to 1997-98 are grossed for both coverage and unknown/invalid clinical data.

  But the figures for 1998-99 to 2000-01 have not yet been adjusted for shortfalls in data.

  The above operations are mentioned in any of the four fields in the HES data set.

  The total for all operations is not equal to the sum of the rows because some operations (episodes) have procedures on more than one row.

  Several endoscopic procedures have dedicated "single" codes in the OPCS4 classification of operative procedures, but many others require the prescence of a secondary code to qualify the reported operative procedure as performed by an endoscopic approach.

  Source: Hospital Episode Statistics (HES), Department.


Modelling the Reduction of Waiting Times for Hospital Treatment


  This following note covers:

    —  The waiting times models

    —  the assumptions behind the analysis; and

    —  a measure of the sensitivity around those assumptions.

  Our analysis has been done within a worst case framework. It assumes that new objectives will be met without changing the way we work and manage waiting lists, which is definitely not the case. Moreover, the implementation of booking across the NHS represents a fundamental change to the dynamics of the system.


  The goal is to deliver a three month maximum wait for first attendance at an outpatient clinic and a six month maximum wait for elective inpatient (ordinary and day case) treatment by 2005.

Waiting Times Models

  Two basic models are used: an outpatient and an inpatient waiting times model. They follow a similar methodology and can be used in conjunction. Specialty level versions and a day case and ordinary admission version of the inpatient model are also used. The models use the way waiting lists and times have behaved in the past to estimate future results within a given set of parameters, e.g.

    —  growth in GP referrals;

    —  removals from waiting lists without being seen;

    —  conversion rate from outpatient waiting list to inpatient waiting list;

    —  how patients are seen/admitted from waiting lists; and

    —  additional outpatient/inpatient activity.

  From the information above, the models then estimate the size and distribution of the waiting lists at given points in time. The models attempt to replicate the differing clinical priorities of patients by assuming that a given proportion of new additions to waiting lists (different for outpatient and inpatient) will need to be seen very quickly. Another group of patients will need to be seen fairly quickly and some will be able to wait. The proportions that will need to be seen within given intervals are estimated from historical data.

  A very crude simplification of the model is shown below

  The amount of activity required to keep the outpatient waiting list steady is first modelled, after taking account of removals from the outpatient list. The amount of additional activity required to reduce long waits is then calculated.

  The number of additions to the inpatient waiting list is calculated by applying the converstion rate from outpatients. This has stayed fairly constant over the last few years.

  We have two different versions of the inpatient model. These vary in the assumptions they make about removals:

    —  the "Constant" model assumes that the risk of removal from the list is constant however long a patient has been waiting; and

    —  the "Sutherland" model assumes that the risk of removal increases with time spent on the list.

  The constant model provides a more realistic view of removals when comparing predicted levels of removal with past trends. Throughout the modelling of the central scenario, the constant model has been used.


  At each stage of the process, we need to make assumptions about the behaviour of the system. This is covered below:


  We have assumed underlying demand of 3% for outpatient treatment following GP written referral. This level of demand is the average level of growth seen since the mid-1990s but around 1% higher than the trend over the last couple of years. Any supply induced demand could be offset by the impact of referral protocols and new incentives and levers to be put in place around PCGs. This has not happened during the waiting list initiative during the last two years (when waiting times have also been falling), where referral growth has remained fairly stable. [Research from CHE at York University suggests that the elasticity between average waiting time and demand is around—0.2.]

  Conversion rates from outpatient attendance to the inpatient waiting list have remained very stable in the last few years. [We have assumed that this level will continue in the future.]

  However, it is important to consider the case-mix changes that will occur. As long-wait specialties are tackled, the number of Decision To Admits (DTAs) for ordinary admission may increase but the overall conversion rate will remain unchanged.


  Long waits in inpatients have been tackled by treating urgent cases first and then targeting additional activity.

  With the reduction in waiting times there should be a corresponding fall in those removed from the list. Reasons for removal include that the patient cannot be traced, no longer requires treatment or has moved away. For each of these reasons a shorter waiting time will decrease the chances of the event happening. This has proved a realistic view of removal patterns when comparing predicted levels with past trends.

Reducing cancellations—pre-condition for booking

  A necessary pre-condition to successfully introducing booking systems for all elective hospital admissions is to reduce the risk of cancelling elective admissions to almost zero. This can be achieved through increasing NHS capacity to meet the surges in need for emergency inpatient treatment. This extra spare capacity (not surplus) would reduce the risk of cancellations but also increases the average cost of treatment. There are two capacity requirements to avoid cancellations: the availability of staffed beds and operating theatre time.

Simulations of emergency admissions at an acute hospital found that a mean bed occupancy of 80% kept the risk of cancellations to almost zero. Occupancy rates for non-urgent elective patients can be run at a higher level (90%) because their admission can be planned.

  Similar considerations apply in theatres.


  The Committee requested to see details of the standards in care homes for the elderly. For information, enclosed are two copies of the report from March 2001 "Care Homes for the Elderly—National Minimum Standards" which outlines these standards.


  The performance ratings website ( contains summary data for each acute trust, as stated by SoS in the HSC. The website shows, for each Trust, the level achieved against each indicator as a tick, dash or cross. A table is provided which shows the range of values covered by these three categories for each indicator. The website also shows, for each Trust, a further breakdown of the performance against each of the core targets and focus areas.

  Supporting methodology and threshold levels to determine the ratings are stored under `Methodology' and/or `Indicators'. The individual report cards containing further information can be found by selecting the following options on the website:


    Trust Clusters

    (eg) Small acute

    (eg) Princess Alexandra Hospital

    Performance Briefing

  The underlying data was not provided, although some of it (such as waiting times) is already in the public domain. We are making arrangements for the data to be forwarded to the committee

2   £32.5m was included in unified allocations to health authorities for NRT and Zyban. Between April to August 2001, the net ingredient cost of NRT/Zyban products prescribed in GP practices which were dispensed was £11.2m. This figure excludes the costs of NRT products delivered via other routes other than on prescription. The decision on how best to spend unified allocations is made locally, depending on local priorities. Back

previous page contents

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2002
Prepared 8 January 2002