I’m not going to answer the question, because “should” implies a right answer. Consider only so far as the patient, the doctor, and the problem, and already the product of the variables is infinite. Even though GPs successfully consult with patients millions of times every week, there is no way to specify a correct amount of time.
But given the pressure on GP capacity it is an important question. It seems to me more useful to examine the evidence on how much time GPs actually take. We have clinical system data from dozens of practices recording the duration of consultations, and patterns emerge.
Telephone consultations last about half as long as face to face
The data for this study is from The Elms in Liverpool, featured as a case study and using the GP Access method since April 2012. Firstly, note the the context
- the service is for a registered list of patients, with full medical records
- the GP phones all patients to begin the episode
- the GP may decide to bring the patient in, and has space today
- the patient may insist on coming in (though rarely does so)
Times are measured from the opening and closing of the patient record, so no extra work was required to produce the data. The mode of 3 minutes may seem short, but as one third of consultations result in the patient coming in, many of these calls will be ended quickly when a face to face is arranged. But the mean is just below six minutes as a few calls are much longer. Note that no norm is applied: the consultation is as long as it needs to be.
Face to face consultations are longer and more variable
Central to this system of operation is that the time allowed for a consultation is determined by the clinician who therefore has greater control over the day and can allocate an appropriate time, based on the earlier brief telephone call.
This means that the time is not constrained to 10 minutes, patients hurried through because of a queue in the waiting room. There is more time to deal with all problems in one go (Norman Lamb is not the first patient to be told, one problem at at time only, you need to rebook. This takes longer overall, for patient and doctor).
The shape of the curve highlights the extent of variation, though the mean is just under 12 minutes and little changed from a traditional face to face system. While the mode is 10 minutes, many are much longer. In the GP contract 2014 the requirement for a 10 minute consultation has sensibly been dropped. Some clamour for 15 minutes – and they are right, but for only a small minority of patients. Many more need under 10 minutes, also right. What is inefficient is allocating the wrong time – too short, and rework results. Too long, throughput falls and waits rise.
How to improve efficiency
True efficiency is hard to measure, because it depends on the quality of the consultation which is hard to measure. There is no point being very fast, but causing rework as problems are left unsolved and frustrated patients come round again. But again the problem is important for capacity, and an inefficient system would result in patients being turned away. This crude measure of efficiency is a product of telephone and face to face consultation times, and the resolve rate – those that needed only a phone call taking much less time overall. In this practice, interestingly despite lengthy experience, consultation times are very stable. We have seen a few others come near to 4 minutes over time.
It is important not to impose targets or expectations: the Doctor First site claims 3 calls can be made in 10 minutes (although analysis of their own evidence shows an average duration over 4.5 minutes). Very few individuals can safely achieve a 3 minute average, and our evidence says it is an unrealistic expectation for most GPs, resulting in disappointment.
Resolve rate, the ratio of telephone to all consultations at the Elms has increased. This is often the case with more experience. We find the range is between 50% and 70%. Below 50% saves no time; above 70% overall may result in rework if patient problems are not solved.
Released from the straitjacket of one size fits all, GPs are able to help patients most efficiently when they can tailor duration to need. Efficiency is however not simply down to individual consulting skills, but a system which allows that flexibility alongside rapid access and continuity.
Our current view is that the simplest most efficient design is for all (barring a few exceptions) requests for a GP to begin with a phone call from the GP. I look forward to see whether this changes over time: meanwhile, there is much to be gained in every practice from what we already know.