Intended for healthcare professionals

Education And Debate

Mind the gap: the extent of the NHS nursing shortage

BMJ 2002; 325 doi: (Published 07 September 2002) Cite this as: BMJ 2002;325:538
  1. Belinda Finlayson, research officer (Bfinlayson{at},
  2. Jennifer Dixon, director,
  3. Sandra Meadows, visiting fellow,
  4. George Blair, adviser on workforce issues
  1. Health Care Policy Programme, King's Fund, London W1G 0AN
  1. Correspondence to: B Finlayson
  • Accepted 1 May 2002

The NHS is struggling to recruit and retain nursing and midwifery staff in a time of high turnover rates and low morale. The problems are most acute in inner cities and teaching trusts. The government is tackling the crisis, but the reasons behind the staffing shortages are complex

The government has a mission to “modernise” Britain's NHS. Success will depend on NHS staff—in particular, whether their numbers can be boosted, whether staff can change how they work, and whether they can be motivated to “go the extra mile” for the NHS. Yet the service is struggling to attract and retain staff in crucial areas, particularly in nursing and midwifery. Here we assess the extent of recruitment and retention problems in nursing in England, comparing acute NHS trusts in London with those in other cities. In another article in this same issue we examine the government's initiatives for tackling these problems.1

Summary points

  • The NHS has serious problems in recruiting and retaining nursing and midwifery staff

  • The underlying causes of these problems include pay, the changing nature of jobs, how valued the staff feel, and other employment opportunities

  • The crisis is most acute in inner cities and teaching trusts, particularly in London, where some turnover rates range from 11% to 38%

  • High turnover results in higher costs and lower morale and may affect patient care

  • The government is tackling the crisis but change is slow and the problems are complex

  • Why 34% of new graduate nurses are not registering to practise needs further study

Numbers and trends in nursing and midwifery

The nursing and midwifery workforce comprises two broad groups of staff working in the NHS. The first group comprises registered nurses and registered midwives, who have a diploma or degree and who have registered with the Nursing and Midwifery Council (before 1 April 2002, the UK Central Council for Nursing, Midwifery and Health Visiting). The second group comprises nurse auxiliaries, nursing assistants, and healthcare assistants; these staff may have received training up to the level of national vocational qualifications but are not registered with any regulatory body. We focus on registered nurses working in the acute sector.

A total of 634 529 nurses and midwives were registered with the UK Central Council for Nursing, Midwifery and Health Visiting in 1999-2000, of whom 90% were women2 and about 11% were from ethnic minorities.3 The NHS in Great Britain employs 302 400 whole time equivalent registered nurses and midwives.4 In 2000, 23 140 whole time equivalent healthcare assistants were employed in the NHS in England.5

Results from the 2000 annual NHS vacancies survey6 suggest there are about 10 000 vacancies for registered nurses and midwives. But the Royal College of Nursing suggests that the figure is nearer 22 000 (whole time equivalents).4 The difference between the two figures is explained by the way vacancies are calculated. The NHS vacancies survey counts only posts that have been vacant for three months and that NHS trusts are actively trying to fill. The Royal College of Nursing counts a post as vacant on the day it becomes vacant and includes posts that have been frozen. On an average day, about 20 000 nurses provide vacancy cover for hospitals in England and Wales, costing the NHS almost £810m ($1235m; £;1273m) a year.7

Registered nurses and midwives

Nurses and midwives must renew their registration every three years. Figure 1 shows changes in the number of nurses registering since 1990. The number of registrants peaked in 1997 and then declined. However, the figure masks three trends. Firstly, although the overall number of registrations increased by about 30 000 between 1990 and 2000, the number of nurses and midwives (on the register) who had trained in the United Kingdom declined by about a third (6000) between 1990-91 and 1998-99. The decline was steepest in the early 1990s, and there has been a modest recovery in the past five years.2 This trend may be explained partly by a reduction in the overall number of preregistration training places in the early 1990s and then an increase since 1995.

Fig 1
Fig 1

Number of nurses registered with UK Central Council for Nursing, Midwifery and Health Visiting, 1990-2001.2 Data from UK Central Council for Nursing, Midwifery and Health Visiting and from a council press release8

Secondly, the number of nurses and midwives joining the register who qualified overseas has been steadily rising (5300 joined in 1998, 7705 in 2001). This partly results from NHS recruitment campaigns. These numbers are expected to jump substantially in 2002 as many of the 29 000 overseas qualifiers who applied to join the register during 2001 complete adaptation courses.8

Thirdly, since 1997 the number of leavers has outstripped the number of entrants. In 1997-8, for example, 16 392 nurses and midwives joined the register and 27 173 left.2 This may be due to three factors: an increase in the number of nurses and midwives retiring; changes in post-registration education and practice (PREP) requirements in 1997 (nurses and midwives who had not maintained their practice would be removed from the register); and overseas registrants allowing their membership to lapse.4 Partly as a result, the average age of the nursing and midwifery population (in the NHS and on the register) is rising. Already nearly half of NHS nurses and midwives are aged over 40, and the number of retirements are projected to rise from 5500 in the late 1990s to over 10 000 a year by 2005.9


The table shows that the total number of undergraduate nursing and midwifery students has almost doubled between 1995-6 and 1999-2000, from 62 010 to 117 680. But there are several caveats to this. Training places were significantly cut during the late 1980s and early ‘90s, so the increase in numbers towards the end of the 1990s may do little more than compensate for the earlier cuts. Moreover, not all those who begin training become registered nurses or midwives—on average, a fifth are estimated to leave during a three year course.10 The numbers do not show the proportion of training places filled by overseas students (who may be less likely to work for the NHS), currently estimated at 4%.1

Total number of students enrolled in nursing courses at UK educational institutions, in 1995-6, 1997-8, and 1999-2000. Data from Higher Education Statistics Agency

View this table:

We estimate that the number of newly qualified nurses and midwives eligible to register with the UK Central Council for Nursing, Midwifery and Health Visiting in, for example, 1997-8, was 24 686. (This calculation is based on the assumption that each course lasts three years, that a third of students are in their last year of education, that a fifth do not complete their course, and that 4% return to their home country.) In fact, only 16 392 nurses registered in 1997-8, suggesting that about a third of new graduates do not register to practise. Furthermore, in the first year after qualifying and registering to practise, 10% of nurses do not work for the NHS.11

No nationally available comparable data exist for higher education institutions on attrition rates from NHS funded training places,10 and it is difficult to link data on students enrolled in training courses with those qualifying and those who go on to join the register. This means, at best, that calculations on attrition rates from courses and on the difference between the number of nurses and midwives qualifying and not registering are crude. To improve workforce planning this problem must be tackled urgently.

Reasons for not joining or for leaving the profession

The data presented above suggest that recruitment in nursing and midwifery is less problematic than retention. Reasons for retention problems in the NHS can be grouped under four broad headings: pay and the cost of living,9 the changing nature of the job, 11 12 perceptions of being “valued,”9 and other employment opportunities.11

Almost half of all nurses who leave the NHS remain in nursing, typically in non-NHS nursing posts or general practice nursing.11 Other reasons for leaving the NHS include career breaks, retirement, maternity leave, nurse education, non-nursing work, and travel.

Effect in acute hospitals in London and other cities

The reasons cited above for not joining or for leaving nursing and midwifery probably have profound effects in cities, particularly London. In this section we examine turnover rates in acute trusts in several large cities in England. We focus on registered nurses working with adults in acute NHS trusts because data on this group is more readily available.

Two indicators normally used to reflect retention are the number of vacancies and staff turnover. Vacancies are measured as the number of unfilled established nursing posts at a particular point in time—they would typically be all or partly filled by agency staff or by staff doing overtime. Staff turnover represents the number of nursing staff who have left a post and moved on to another NHS organisation or who are known to have left the NHS altogether, over one year.

There are two drawbacks to using vacancies: the data represent only one particular day (usually 31 March), which may not be representative of the whole year; and the definition of “unfilled post” is loose. Accordingly, we focus our analysis on data relating to staff turnover.

Figure 2 shows turnover for registered nurses working with adults in London's acute NHS trusts in 1999-2000. The turnover rates for such nurses in 1999-2000 were high, with one in three leaving their post in almost a third of London's acute trusts. Of the 33 acute trusts in London, 25 had turnover rates above 20%, 18 trusts had rates exceeding 25%, and nine had rates exceeding 30%. Trusts with turnover rates higher than 25% are more likely to be in inner London and to be teaching trusts.

Fig 2
Fig 2

Turnover rates for registered nurses in adult care in acute NHS trusts in London, 1999-2000. Trust No 4 employs less than 100 whole time equivalent registered nurses in adult care. Data from the Education and Training Consortia in London

Turnover rates for registered nurses working with children and for registered midwives tend to be lower than those for registered nurses working with adults. However, the numbers of registered nurses working with children and of registered midwives are much smaller than those caring for adults, which may make the turnover rates for these subgroups more difficult to interpret. Of the seven NHS trusts employing over 100 whole time equivalent registered nursing posts (care of children), five had turnover rates of 30% or above. Of the 12 trusts employing over 100 whole time equivalent registered midwifery posts, five exceeded 20% and none exceeded 30%. For both indicators, the rates were higher in inner city and teaching hospitals.

Comparison of London with other cities

To help assess whether London is a special case, we analysed turnover rates of registered nurses in acute trusts in two other large conurbations in England. Figure 3 shows the rates (for nurses in adult care) for NHS trusts in Birmingham and Solihull and in Manchester.

Fig 3
Fig 3

Turnover rates for registered nurses in adult care in acute NHS trusts in Birmingham and Solihull and in Manchester, 1999-2000. Trust No 1 in Birmingham and Solihull employs less than 100 whole time equivalent registered nurses in adult care. Data from the Education and Training Consortia in Birmingham and Solihull and in Manchester

In Birmingham and Solihull, where most trusts are in the outer city, turnover rates are higher among teaching trusts. Overall, however, when comparing trusts employing over 100 whole time equivalent registered nursing posts in adult care, turnover rates in Birmingham and Solihull are lower than those experienced by some acute trusts in London. Like London, however, turnover rates are lower among registered midwives and registered nurses working with children than among registered nurses in adult care.

In Manchester, as in London and in Birmingham and Solihull, the turnover rate for registered nurses in adult care is higher in the inner city and in the teaching trust. Overall, however, turnover rates in acute trusts in Manchester are much lower than in London and in most trusts in Birmingham and Solihull.

Why are the turnover rates in these three conurbations higher in inner city areas and in teaching trusts and more acute in larger cities, particularly London? Evidence is limited, but there are several possible reasons. Firstly, teaching hospitals are a training ground for nursing students and new graduates. They are generally able to recruit newly qualified nurses in grades D and E (the lowest grades for registered nurses) but struggle to retain them beyond two years. Secondly, accommodation and living costs are expensive in inner cities. As nurses start wanting better accommodation, they tend to move to the outer suburbs or beyond, where accommodation is cheaper. Difficulties travelling to the inner city may then make working close to home more attractive. These problems may be more acute in London because there is a concentration of teaching hospitals; accommodation and living costs are more expensive than in other cities in England; and transport can be more arduous.

Consequently, inner city and teaching trusts experience high turnover among lowest grade registered nurses.13 Outer city trusts have difficulty recruiting nurses in the lowest grades but experience lower turnover in higher grades, such as F and G.

If turnover rates are high (particularly in London), should we be worried? Clearly high turnover has a knock-on effect on staff morale, 9 14 organisational finances, 15 16 and perhaps patient care. 17 18 Little consensus exists over whether there is an appropriate level of turnover that can be applied to all trusts since different, and at times uncontrollable, factors operate in different places and to varying degrees. Consensus is firm, however, among acute trusts in London that turnover rates are too high.


The serious problems facing acute trusts in England in retaining and recruiting nurses result in high financial costs and low morale and may affect patient care. These staffing problems are most stark in inner city and teaching trusts—particularly in London, where they will need to be tackled if the government's modernisation agenda for the NHS is to be realised.


We thank the former education and training consortiums for use of their data and insights into the story behind the data; Isabella Kpobie and Diane Gray for help with formatting the data; and Steve Dewar and Pippa Gough for helpful comments on earlier drafts of this paper.

Contributors: BF helped to design the paper and collate and analyse the data; she also wrote and edited the paper. JD helped to design and edit the paper and analyse the data. SM helped to design the paper. GB helped to find the relevant data and collate and analyse them. BF and JD are guarantors for the study.


  • Funding None.

  • Competing interests None declared.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
View Abstract