2Assessing the health workforce

Density of health workers per 10 000 population

The density of health workers shows the number (physical persons or full-time equivalent) of doctors, nurses, midwives, dentists, pharmacists and physiotherapists per population (usually expressed in per 1000 or 10 000). This is one of the key indicators to assess the supply of the health workforce and to monitor as per the WHO Resolution (WHA69.19). The WHO Global Health Workforce database and the Eurostat and OECD databases provide internationally comparable data on the six main categories of health professionals listed above.

Europe has the highest density of health workers compared to other WHO regions, with 37 doctors, 80 nurses, 4.1 midwives, 6.7 dentists and 6.9 pharmacists per 10 000 people (WHO Regional Office for Europe, 2022). There is nevertheless wide variation between countries, with the density of doctors, nurses and midwives ranging from 54 per 10 000 population in Türkiye to over 200 in Sweden, Belgium and Finland (Fig. 2.1). This reflects differences in investment in education, training, recruitment and retention, and capacity for planning, as well as variations in health system organization and service delivery.

Limitations and challenges of interpreting this indicator

Determining what counts as a sufficient number of health workers in a given country is challenging. It depends on a complex analysis of demand-side factors such as population characteristics and patterns of health service utilization, as well as supply-side factors including health workforce characteristics (for example, numbers working full-time versus part-time).

Sufficient density levels will depend on health system objectives and how healthcare is organized and delivered. A country that is focused on delivering team-based care or on prevention may need a wider range of health workers with different skills and competencies than other countries. Expanding the skill-mix of the workforce can reduce the number of doctors and nurses needed to deliver effective care, making it important for countries to monitor the composition (skill-mix) of the workforce and substitution of tasks between occupations (see Section 2.2). Countries that make greater use of digital technologies that improve the efficiency of care delivery may also need a lower density of health workers compared to others (see Section 3). It should also be noted that some countries with lower health worker densities have relatively good health outcomes compared to countries with higher densities, for example as a result of having a younger population or having a more effective welfare state that contributes to better population health.

Figure 2.1

Total density of medical doctors, nurses, and midwives, 2020 or latest available. Source: WHO 2023, National health workforce accounts database

Figure 2.2

Physician density, metropolitan and remote areas, 2021 or latest available. Note: Remote areas are defined as regions far from metropolitan areas and regions near small urban areas with a population of less than 250 000 people. Source: OECD, 2023, Regional (more...)

Comparisons of density levels across countries are challenging owing to lack of data harmonization. Very often countries report the numbers of health workers without providing information on whether they work full-time or part-time. There are further definitional issues. For instance, the classification of a general medical practitioner or medical specialist may vary between countries. The definition of nursing professionals is not always clear-cut (for example, whether nurses with lower levels of qualifications are counted as nurses or as nursing assistants). Furthermore, even within Europe, not all countries are able to report the number of practising health professionals, and instead report only the number of professionally active or licensed to practice health workers, thus inflating the number of healthcare staff working on the ground. Also, where dual public and private practice is permitted, few countries are able to report the number and activity level of health professionals working in the public sector.

Limitations and challenges of interpreting this indicator

This indicator does not provide data on why geographical disparities are present. Many factors influence decisions to work in different geographic areas and these need to be assessed to design effective policies. Due to a concentration of specialist services in larger hospitals in urban areas, some variations in density levels may be warranted and should be taken into account. However, there are well documented disparities in the availability of GPs and nurses that result in lack of availability of certain types of care in underserved areas. This makes it useful to also consider types of skill-mix, especially for delivery of primary care at the subnational level (see Section 2.2). Disparities will also likely be related to challenges around recruitment and retention. Indicators on recruitment and retention (see Section 2.3) – collected through surveys of staff, graduates, and students – are therefore needed to understand these factors in order to develop appropriate policies to address them.

Measuring health worker density in isolation may mask other challenges in accessing services in rural areas. For instance, density levels may be deemed sufficient, but poor transport infrastructure or long distances may impede access to care and also need to be considered. At the same time, the use of telemedicine and other digital health technologies can support the delivery of services in rural areas even if health workers are not physically present. Monitoring the accessibility and use of digital health tools to support care delivery in these areas is thus important (see Section 3).

Some countries may not collect data at the subnational level, or may only do so for doctors and nurses. Even when data are collected, they may often be unreliable. For example, a recent review of the nursing workforce in Lithuania showed that data on numbers of nurses reported by the municipalities differ from those reported by the Ministry of Health five-fold in some areas (OECD/European Observatory on health systems and policies, 2023).

Limitations and challenges of interpreting this indicator

There are no standard estimates on what the graduate replacement rate should be, as this will be context specific. It will depend firstly on the current stock of health workers (see previous indicator) and retirement patterns. Policy-makers also need to take account of how demand-side factors (for example population health needs, population age structure, etc.) will change in the future. Forecasting will also be needed to assess how care delivery may progress in the future, such as in relation to the advancement of digital health technologies or an increased focus on prevention, and what this may mean for skill-mix requirements (see Section 2.2). The indicator does not capture potential supply from foreign-trained workers (see next indicator), which may be particularly important for those countries that attract a lot of foreign workers, and other sources of recruitment (for example, inactive workers), and should therefore be analysed in combination with data on other potential sources of supply.

Indicators on the number of graduates do not show how many of these graduates will end up working in the health system or for how long, or how many will work in currently underserved areas. Many health workers may not work in the health system after graduation for many reasons, including: not completing their training, not finding jobs sufficiently attractive (in terms of pay or other working conditions), choosing to work in other sectors or other countries; lack of effective recruitment strategies; and insufficient effective demand and job offers (McPake et al., 2023). Countries therefore need to monitor and take into account education and training completion rates and attrition rates, especially in the first few years after graduation, to understand supply needs and how to reduce waste in human capital. Finally, this indicator also does not reveal anything about the quality or level of education received; this is important for countries to monitor, such as by assessing whether educational institutions are accredited and provide competency-based learning.

Figure 2.3

Number of annual nursing graduates per 100 000 population, 2020 or latest available. Source: WHO, 2023, National health workforce accounts database

Limitations and challenges of interpreting this indicator

Many countries do not collect data on the age of health workers, or only do so for doctors and nurses and not other health professionals (WHO Regional Office for Europe, 2022). In addition, these data do not capture exactly when workers will retire or why. Further research could help establish more precisely the effective average retirement age in countries, and identify drivers of early or late retirement age. The increased use of digital health technologies may (positively or negatively) affect retirement rates, as may the introduction of age-friendly workplace policies, and their impact on older health workers should be assessed. Retirement rates may also differ according to medical specialty or by geographical area. The indicator does not take into account other reasons for attrition, and is therefore only one input into calculating current and future exit rates from the health workforce.

Migration of health workforce

A number of countries in Europe lose a substantial proportion of health workers to migration each year, while others are “net gainers”. In the former group of countries, this undermines access to care and the sustainability of health systems, represents a waste of investment in the education and training of these workers, and needs to be taken into account when analysing sufficiency of current and future supply. However, few countries are able to monitor outward migration of health workers. One method to capture this information is through surveys of leavers or data on emigration collected in professional registries (for example, as conducted in the United Kingdom). Another method to support countries in understanding rates of outward migration is through collecting immigration data by country of origin of foreign-trained workers who are working in various destination countries. The OECD/Eurostat/WHO-Europe Joint Questionnaire in Europe asks countries to provide this information, although data are only reported for doctors and nurses and not all countries report data on foreign-trained doctors and nurses by country of origin. Table 2.1 provides an example of the results from the Joint Questionnaire on the emigration of nurses trained in Romania to other EU and OECD countries measured in terms of annual outflow.

Figure 2.4

Percentage of medical doctors aged 55 and over, 2020 or latest year. Source: WHO, 2023, National health workforce accounts database

Because there are often no direct data available on the emigration of doctors and nurses from source countries, the approach of collecting data on immigration from destination countries is in many cases the best way to monitor emigration from source countries on a routine basis. Monitoring the country of training of health workers is also useful for destination countries. It can help show if countries are relying too much on foreign-trained workers to overcome shortages, potentially due to issues over retention (see Section 2.3) or insufficient investment in education. This is an unsustainable strategy to address workforce issues in the longer term as it is highly susceptible to changes in visa regulations, and is dependent on recruitment and retention policies in other countries that are competing for foreign-trained workers and the motivation of health workers to migrate (Williams et al., 2020). In addition, it deprives other – often poor – countries of health professionals, undermining the sustainability of their health systems and representing a huge loss of investment in training and education. Nevertheless, it should be noted that migration is not a wholly negative phenomenon and can have benefits for health workers as well as source and destination country health systems (OECD, 2015; WHO, 2016).

Table 2.1

Emigration of nurses trained in Romania to other OECD countries, annual outflow, 2010–2022.

Limitations and challenges of interpreting this indicator

Many countries do not accurately report data on location of training for health workers, or only do so for doctors (Williams et al., 2020). This means that relying on information from other countries to assess outward migration trends will often be incomplete. The indicator also does not provide information on why health workers are migrating, making it important to capture this information from surveys of leavers as it will often be tied to poor retention strategies (see Section 2.3). In addition, it may be that domestic students study abroad and return to their country of origin once their education is completed. This cannot be considered a “brain drain”. In some cases, countries (or medical schools) deliberately target foreign students as a means to get funding for education institutions and train more graduates than needed for their own replacement rates. Alternatively, other countries (for example, the Philippines) overtrain domestic students in recognition that many will go abroad and send remittances.

Ratio of doctors to nurses

All countries have invested in expanding the nursing workforce (including midwives) to support the delivery of cost-effective, patient-centred care and to reduce the over-medicalization of treatment. Doctors and nurses have distinct but complementary skills, and ensuring an appropriate doctor to nurse ratio can support the efficient allocation of resources, ensure a more balanced and effective distribution of tasks and responsibilities, and promote team-based working, which can support the delivery of high-quality care.

The ratio of doctors to nurses varies considerably across Europe. Figures 2.5 and 2.6 show two examples of how the ratio can be visualized. A ratio of high numbers of nurses to low numbers of physicians may represent countries with a greater focus on team-based working and task-sharing with physicians. However, it may also reflect a lack of investment in recruiting and retaining doctors (see Section 2.3). Conversely, a ratio of high numbers of doctors to low numbers of nurses may instead represent countries where physician-led care dominates, or highlight challenges in recruiting and retaining nurses.

Figure 2.5

Ratio of nurses to doctors in OECD countries, 2021 or latest available. Note: In Greece and Portugal, data refer to all doctors licensed to practice, resulting in an overestimation of the number of practising doctors. In Greece, the number of nurses is (more...)

Figure 2.6

Doctors and nurses per 1000 population in the EU, 2021 or latest available. Notes: 1. For countries that have not provided data on practising nurses and/or practising doctors, numbers relate to the “professionally active” concept for both (more...)

Share of GPs in the physician workforce

Efficient, equitable and people-centred health systems are founded on strong primary health care and public health principles (WHO, 2016; OECD, 2021). Yet despite many countries implementing policy actions to strengthen the primary health care workforce, the ratio of GPs to specialists has declined in many countries in Europe (Fig. 2.7), and shortages of GPs are now a critical issue in many countries, especially in rural and remote areas (OECD, 2020). While a decline in the share of GPs may reflect decisions to invest in other types of health workers in primary care or specialist care instead, it is very often related to challenges in recruitment and retention. Primary care is often viewed as less attractive to work in compared to specialist care, owing to the work environment, working conditions, remuneration, and work-life balance (Kroezen, Rajan & Richardson, 2023).

Insufficient numbers of GPs can affect timely access to care and early intervention for health issues. GPs also play a gatekeeping role in many health systems, but adequate numbers are required to ensure appropriate utilization and access to specialized care, preventing unnecessary referrals and thus reducing healthcare costs. Monitoring this indicator can shed light on (lack of) investment in primary care, and highlight the need to increase training places in general medicine, implement additional incentives for graduates to work in primary care, and/or increase recruitment in this sector.

Figure 2.7

General practitioners as a share of physicians (%). Source: OECD Health Statistics, 2023

How do these indicators help to monitor and transform skill mix?

Addressing skills gaps can help support the delivery of innovative models of team-based care by strengthening primary health care and prevention. Skill-mix reforms that facilitate task-sharing with physicians are generally a cost-effective way to meet population health needs and can also help address physician shortages as other types of health workers can be trained in a shorter amount of time.

The indicators suggested in this section only capture a very high-level overview of skill-mix in the health workforce owing to current data limitations in many countries beyond the EU or OECD. This includes a lack of data on health workers beyond doctors, nurses and midwives, and insufficient disaggregation by geographic location and type of care. Capturing more complete and accurate data on all types of health workers, which should be standardized across countries, is a prerequisite for monitoring skill-mix. Further data disaggregation according to scope of practice (for example, advanced practice nurses versus registered nurses with a Bachelor’s degree) could provide more insights for optimizing skill-mix, although various countries are at different stages and taking different approaches in implementing new advanced roles for nurses and related titles (for example, nurse practitioners, community and family nurses, etc.). Improving data on available skill-mix can be used to inform the development of education and training programmes and the use of joint workforce planning and forecasting, which are essential to effectively overcome shortages and skills gaps.

Intention to leave in the next 12 months

Monitoring intention to leave in the next 12 months can help shed light on whether policy-makers at the regional or national level need to take action to address various issues connected to retention, or may need to step up recruitment initiatives to address shortages. However, few countries routinely collect such data. Some exceptions include Belgium, Germany, Ireland, Netherlands (Kingdom of the), Poland and England in the United Kingdom.

In England, one of the largest staff surveys in the world is conducted annually (since 2003) to understand staff experiences. Interactive results for the national, regional and local levels are published online and are publicly available (NHS, 2023). While the survey contains core questions which must be included, local areas may opt to include other questions of local relevance. In 2022, 46% of the NHS workforce of 1.3 million workers took part in the survey (King’s Fund, 2023).

The NHS staff survey is open to all staff and covers a wide range of questions related to retention, including intention to leave (see Fig. 2.8). These results can be broken down by NHS provider, care area (for example, ambulance trusts, community settings, etc.), occupation group, workforce characteristics (for example, gender, ethnicity, childcare commitments, etc). Breaking down data to this level allows policy-makers and managers at different levels to understand where specific retention efforts may need to be targeted. For example, Fig. 2.9 shows that intention to leave for younger age groups is highest among ambulance workers and lowest for acute specialists, with the reverse patterns seen for acute specialists.

Limitations and challenges of interpreting this indicator

This indicator does not show how many health workers will actually leave in the next 12 months. Monitoring exit rates to establish how many people have left the health workforce in the last 12 months is also needed to inform retention efforts and for accurate health workforce planning and forecasting. Data on exit rates can most reliably be collected from professional registers (for example, as in the UK for doctors and nurses), although these have not yet been established for all professions in many countries. Identifying other data sources that provide data on health workers who have actually left their job (going beyond the “intention to leave” indicator) would be useful to inform understandings of exit rates.

Figure 2.8

Results of the NHS staff survey. Source: NHS (2023)

The indicator on its own also does not enable policy-makers to understand why health workers are intending to leave. Establishing regular staff surveys with questions on reasons for intention to leave are therefore needed as this can inform the development of appropriate policy actions (for example, efforts to improve mental health support, work-life balance, remuneration, etc.). Harmonizing a small set of variables in national staff surveys across countries would be useful to be able to obtain cross-country comparable data. As in England, it is important this information is broken down by different categories, including local area and occupation time, to ensure policy actions can be targeted. Of course, just because this information exists, it does not mean that it will be acted upon by managers or policy-makers. Some level of monitoring of actions to address issues raised by staff surveys could be useful in ensuring accountability for relevant actors. Staff surveys are not likely to be mandatory so efforts may be needed to encourage participation among certain groups who may be less likely to respond.

Job satisfaction rates by occupational group

If health workers are not satisfied with their jobs, they may be more likely to leave to work in other health facilities, outside the health sector, or in another country. Lack of job satisfaction may also lead to loss of morale, and contribute to higher rates of absenteeism and reduced productivity. Understanding whether staff are satisfied with their jobs is therefore important to inform retention efforts at the local and national levels. These data can be captured through regular staff surveys; as noted, however, these are not undertaken regularly in many countries. In the NHS England staff survey, an overall question on job satisfaction is not asked. Instead, satisfaction with different elements – including recognition, feeling valued, opportunities for flexible working hours and level of pay (see Fig. 2.9) – is captured, providing more information on reasons for health workers being satisfied or unsatisfied with their jobs. Fig. 2.10 shows that satisfaction with pay among respondents to the NHS staff survey is lowest for ambulance workers, but has been declining among occupations in all care areas since 2020, likely partly as a result of the cost-of-living crisis and growth of wages below the inflation rate.

Figure 2.9

Share of respondents to the NHS staff survey 2022 feeling satisfied or very satisfied with their pay. Source: NHS (2023)

Share of health workers experiencing burnout by occupation

Health workers have always experienced high rates of stress owing to the nature of their jobs. However, the COVID-19 pandemic has exacerbated rates of burnout and mental health conditions resulting from intensive workloads over a sustained period and the distress of treating COVID-19 patients (Greenberg et al., 2021; Rimmer, 2021; Santabárbara et al., 2021). Fig. 2.10, using data from the NHS staff survey, shows that almost one third of non-clerical and non-administrative staff working in the NHS reported experiencing burnout often or always in 2021 and 2022, with these rates highest among ambulance staff and registered nurses and midwives. This emphasizes the need for employers at local and national levels to understand and take actions to address the causes of burnout for different occupational groups and to provide mental health and well-being support.

Ensuring that health workers are provided with appropriate mental health and well-being support has become a key area of focus for policy-makers across the region and a central pillar of the WHO Regional Office for Europe’s Framework for Action on the Health and Care Workforce. Providing mental health and well-being support is not just a moral imperative to protect health workers, but can also help reduce absenteeism, turnover and attrition, and improve labour productivity.

Figure 2.10

Share of respondents to the NHS staff survey reporting often or always experiencing burnout because of work, 2021–2022, by selected occupational group. Source: WHO (2023)

How do these indicators help to monitor and transform health workforce recruitment and retention policies?

Together, these indicators can help provide some indication on whether retention strategies in a country are sufficient. Taking action to retain existing health workers to ensure they do not leave the sector is one of the best health workforce investments that can be made. If health workers are not supported, are burnt-out, overworked and feel undervalued, they will not be satisfied with their job or be able to perform optimally, and may drop out of the workforce entirely. This not only exacerbates shortages, but is a huge loss of skills and knowledge and a waste of investment in education and training. The indicators also help provide some guidance on what level of training and recruitment of new workers may be needed to offset potential outflows from attrition.

These indicators are, however, not sufficient to gain a full picture of satisfaction and morale among the workforce, to develop policy options to improve retention, or to understand rates of current and expected attrition. Many other indicators are also needed to address these evidence gaps, which have been described throughout this section. These include levels of remuneration, monitoring exit rates from the workforce together with reasons for leaving, as well as reasons for low job satisfaction, low morale, burnout and intention to leave among practising health workers. The impact of different policies to improve retention for different occupations and demographic groups should also be monitored and assessed to learn what works. Data on these indicators can be obtained from surveys of leavers and staff surveys of current health workers, as well as from professional registries. However, few countries currently undertake regular staff surveys, while some do not have professional registries or may only have them for doctors and nurses. This undermines retention and recruitment strategies and makes planning and forecasting to overcome shortages and skills gaps challenging.