How Many Doctors Do We Really Need?

In a previous post we blogged about the changing nature of medical practice: the influences of regulation, guidelines, sub-specialisation, and patient expectations. We mentioned skills substitution, whereby less experienced staff take on tasks previously carried out by doctors. We also mentioned the role of Information Technology, but shied away from discussing the implications for medical manpower. However, it seems important to ask whether Information Technology could reduce the need for medical input by increasing the scope for skill substitution. Some patients have complex needs or vague symptoms, and such patients we assume will need to be seen by someone with deep medical knowledge to underpin professional judgements, and to provide patients with such an informed account of the probable causes of their illness and the risks and benefits of viable options. But much of medicine is rather algorithmic. A patient presents with back pain – follow the guidelines and refer the patient if any ‘red flags’ appear, for example. Many of the criteria for referral and treatment are specified in guidelines. Meanwhile, computers increasingly find abnormal patterns in a patient’s data that the doctor has overlooked. Work in CLAHRC WM shows that many patients do not receive indicated medicines.[1] Health promotion can be delivered by nurse and routine follow-up cases triaged by Physician Assistants. A technician can be trained to perform many surgical operations, such as hernia repair and varicose vein removals, and Physician Assistants already administer anaesthetics safely in many parts of the world.[2] Surely we should re-define medicine to cover the cognitively demanding aspect of care and those where judgements must be made under considerable uncertainty.

In the USA they talk about “people working up to their license”. What they mean is that it is inefficient for people to work for extended periods at cognitive or skill levels well below those they have attained by virtue of their intellect and education. Working way below the level is not only inefficient, but deeply frustrating for the clinician involved, predisposing them to burn out. Use doctors to doctor, not to fill in forms and perform routine surgical operations.

We conclude by suggesting that there is a case for re-engineering medical care or at least articulating a forward vision. The next step is some careful modelling, informed by experts, to map patterns of practice, assign tasks to cognitive categories, and calculate manpower configurations that are both safe and economical. Such a process would likely identify a more specific, cognitively elite role for expensive personnel who have trained for 15 years to obtain their license. In turn, this may suggest that less people of this type will be needed in the future.

While high-income countries should address the question “how much should we reduce the medical workforce, if at all?”, low-income countries face the reciprocal question, “by how much should we increase the medical work-force?” Countries such as Kenya have only two doctors per 10,000 population, compared to 28 in the UK, and 25 in the United States.[3] Much of the shortfall is covered by other cadres, especially medical officers (who work independently), and nurses. Health personnel are strongly buttressed by community health workers, a type of health worker that we have discussed in previous posts.[4] [5] Information Technology is unsurprisingly very under-developed in low-income countries, although telemedicine is increasingly used. It is particularly difficult to attract doctors to work in rural areas, and there is the perennial issue of the medical brain drain. The time is thus propitious to consider carefully the human resource needs not just of high-, but also of low- and middle-income countries, and consider how these may be affected by improving Information Technology infrastructure.

— Richard Lilford, CLAHRC WM Director


  1. Wu J, Yao GL, Zhu S, Mohammed MA. Marshall T. Patient factors influencing the prescribing of lipid lowering drugs for primary prevention of cardiovascular disease in UK general practice: a national retrospective cohort study. PLoS One. 2013; 8(7): e67611.
  2. Mullan F & Frehywot S. Non-Physician Clinicians in 47 Sub-Saharan African Countries. Lancet. 2007; 370: 2158-63.
  3. World Health Organization. Health Workforce: Density of Physicians (total number per 1000 population): Latest available year. 2015.
  4. Lilford RJ. Lay Community Health Workers. NIHR CLAHRC West Midlands News Blog. 10 April 2015.
  5. Lilford RJ. An Intervention So Big You Can see it From Space. NIHR CLAHRC West Midlands News Blog. 4 December 2015.

3 thoughts on “How Many Doctors Do We Really Need?”

  1. Richard,

    Interesting article, and I tend to agree with the premise that we probably do not need more doctors, and there is a large amount of care (procedures, notably) that could probably be provided by appropriately trained practitioners, but I do disagree with some of our points.

    In particular, back pain in primary care is a hard problem: it’s very common, very likely benign, and has a few serious causes. The only investigation of significant worth is MRI (which numerous studies have shown has a high FP rate and is often abnormal in asymptomatic patients), a high cost investigation.

    ‘Red flag’ features unfortunately don’t work ( Algorithmic management of problems like pain doesn’t work particularly well, because it’s just too common and the history and examination findings too broad. The same holds true for many common problems – the sensitivity and specificity of the protocols are not good enough.

    The other key problem is multi-morbidity. There’s a neat paper in the BMJ i’m sure you’ve seen ( that shows that guidelines for common, often co-existing conditions often interact. Guideline or algorithmic management for these patients, particularly in primary care is difficult. Should I stop the beta blocker as there is some evidence of airways obstructive disease? (probably not). Those are the decisions that are hard, and require complex thinking and high level doctor skills. I particularly like this example:(

    Primary care has both the most amount of low risk, simple patients that are ideal to be managed without complex management, but also has a large number of complex, multi-morbid, vague patients – and you don’t know which one you’re going to see!

    I do agree that a huge number of technical hospital work could be replaced, especially in elective surgery, but I certainly think that primary care (and probably internal medicine/hospitalism) is the place I want the high level care and skills.


  2. Thank you CLAHRC WM team for another entertaining blog. Really enjoyed reading the article on “HOW MANY DOCTORS DO WE REALLY NEED?”. I think that medicine can take inspiration (again) from aviation, where the pilots only need to take over from the auto-landing system during landing in extreme weather conditions. Routine procedures can be conducted by an on-board computer.

    Thank you to Tom Marshall for providing insight around national screening programmes. I think even a lay person would recognise the national diabetes prevention programme as public screening. Why can’t policy makers? If you are going to expense of funding and effort of setting up a public health intervention, it needs every chance of being successful in its primary objective.

    I look forward to the next blog.

  3. I should like to bring the ideas from two different areas where I hve some experience.
    1. It is not true that pilots use autolanding routinely. Even on the modern aircraft they do it mistly ‘by hands’, and employ the autolanding only in special circumstances (very low visibility). I believe that if to learn from aviation, doctors may be relaxed: while many specific task are shown may be done by ‘technicians’ successfully, the whole system of care (including the patient as a subsystem) is so complicated that docteor is needed. And will be needed for a long.
    2. There are nnot only countries who want more doctors and who want less doctors. There are some, like Russia, where there are plenty of Drs in the big cities, and Drs are almost not available in the desert areas. It is impossible to arrange the medical care provided by Drs, if you have the population density like 2 people per square kilometer. USSR did try to arrange the care for such areas by nurses, trained for independent practice. But even this model failed (my opinion) because of the same problem – long distances, rare population. Now, in 21 century Russia is develping the model of care provision for sparcely populated areas by lay care providers.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s