Samir Dawlatly: Can you measure what is good about general practice?

The Health Foundation, at the behest of the government, is gathering thoughts from professionals and the public on the use of data to inform the quality of primary care until 24 August 2015. Although they say that this will not be used to “rate” practices, data about so-called quality has been used to “band” or should that be “brand” practices in the past, with the media immediately interpreting them as ratings. Putting that aside I would like to make three broad points:

1. Much of the variation seen in health outcomes is not due to the quality of healthcare.

2. Many of the qualities of care cannot be quantified (that is why it is a quality and not a quantity).

3. Efforts to quantify healthcare do not reliably correlate with outcomes.

What causes variation in health?

Variation in health outcomes is the argument that is often used to justify top down, widespread policy. The rationale seems to be that areas that have better health outcomes are served by better GP surgeries (or hospitals for that matter) and therefore all other health institutions must be doing something wrong. Finding outcomes to measure (sometimes surrogate endpoints, though life expectancy does vary between areas) will show that areas are different. Taking life expectancy, one of the most consistent factors determining life expectancy is social class and wealth. The rich live longer than the poor. They smoke less, exercise more, and take up offers of health interventions. Does this mean that a GP surgery in a deprived inner city area of Glasgow is no better than one in Guildford, Surrey? Of course not; but by trying to use the same data set to describe these two hypothetical practices one is making the implied assumption that social determinants of health are not important. But they are.

To provide “quality” data sets on GP surgeries is inviting comparison, but it is not a comparison of like with like. This begs the question, “Why bother?” and if it is in an effort to raise standards, then I would suggest that taking seriously the recruitment and retention crisis in GP would be a big step. For general practice to function properly there needs to be enough GPs.

What is good care?

Taking a step backwards from general practice towards medical school, I would pose the following question:

“How does a medical school know that they have produced good doctors?”

The measurements that are available to quantify whether a doctor is “good” are extremely limited. Academic prowess in terms of publications doesn’t necessarily reflect clinical ability. In fact perhaps the only true measure is simply not being struck off. Brusque surgeons can operate brilliantly. Kindly GPs can prescribe inappropriately. The point being that if it is difficult to measure what a good doctor is, to quantify the quality of care that they provide, then it appears logical that it is almost as difficult to measure the care of organisations run by those doctors.

Do measured outcomes correlate with good care?

One would think that the Quality Outcomes Framework would give the opportunity to show whether there is a correlation between definite end points such as life expectancy and QOF achievements. Recent research has shown that no such correlation exists. Those that achieve high on quality markers do not necessarily improve the health of their patients. Whether this is due to the confounding social factors briefly discussed above or because these measurements are not a true measure of quality is difficult to say.

GP surgeries have to carry out the “Friends and Family” test with their patients. There is no evidence that a good achievement in this test correlates with good quality clinical care. Sometimes the most clinically appropriate thing to do is to deny a patient inappropriate antibiotics, which risks upsetting them.


I believe that the idea of creating a national dataset for “clinical quality” is inherently flawed due to the difficulties in quantifying a quality. Previous efforts have not correlated well with clinical outcomes. To press ahead with this will inevitably lead to invalid and unfair comparisons between practices, due to variation beyond the scope and remit of primary care. It will simply provide yet another stick for politicians and certain sections of the press to beat general practice with. My suspicion is that was always the plan.

Samir Dawlatly is a GP partner at Jiggins Lane Surgery in Birmingham. He combines clinical practice with being a part time house husband and an interest in social media, as well as publishing poems, essays, and blogs. He can be found on Twitter as @sdawlatly.

I have read and understood BMJ policy on declaration of interests and declare the following interests: I am a member of the RCGP online working group on overdiagnosis.