Why GPs hate Specialists and Vice Versa

GP specialist relations are at an all time low. Why, and how can we make this better?

If one talks to GPs about what annoys them, you will find that it is rarely the patients. Obviously administrators are well up the list, but so also are hospital doctors. If you talk to specialists, their opinion of GPs is not high. Yet these doctors need to work together to help patients, so why is this aspect of the NHS so dysfunctional?

Let us go through the concerns on each side and look for solutions. (Some people will find this biased in favour of GPs)

GPs being asked to do work that would be much simpler and more efficiently done at the hospital.

Some examples:

A patient sees a GP shortly after a visit to outpatients for a broken arm as he needs a sick note for work. The details of his condition are not available to the GP as the letter (yes a printed letter!) will not be sent for a month or two. It would be much more efficient all round if advice about work and the sick note was done at outpatients, but this is often omitted by the clinic, even though officially it is there that this should be done.

A specialist, or his secretary, gets a lab report suggestive of infection. The result is then faxed to the GP, often with no other information about the medical context.  The GP is not informed whether the patient will be notified by the secretary. Sometimes, the GP phones the patient and finds that treatment has already been issued by the specialist. Sometimes attempts to contact the patient fail and the GP has to write a letter, which is expensive in staff time, and slow. The specialist who knows the context could easily contact the patient, decide on treatment, and issue a prescription. This again would result in better care and be much more efficient, and would avoid the chance of the GP dismissing an important result as insignificant.

A patient has a cataract operation and is issued with enough eye drops for a week as per hospital policy, he is then told to get a  prescription for another bottle from his GP who knows little about the intended treatment dose or duration. The hospital is able to negotiate prices for eye drops and pays typically about half the cost that is paid to a pharmacy.  It would be far more efficient for patients and the NHS, as well as ensure that treatment was actually given, if the hospital issued all the medication needed. Similar issues affect for example patients found to have MRSA prior to operations. Rather than the previous nurse having a large stock of antiseptic shower gel to be handed out, a letter goes out to the gp, with an instruction to prescribe. Ditto patients who need anticoagulant injections around the time of surgery. These are actually quite dangerous, and the potential for the GP to make dosing errors on the prescription resulting in the death of the patient is significant.

A patient sees his GP on the advice of the specialist, to discuss a recommendation for treatment or another referral. But routine letters often take 2 months or more and the specialist may well be unaware of this. So the GP has to ask his staff to ring the secretary (not easy), and ask for the letter to be faxed. Often the letter is still in the typing pool and there is a wait, and then another consultation. If the specialist was aware of the problem he would get the letter done faster and faxed, and he would tell the patient not to attend before the letter was likely to arrive.

Similarly we get letters recommending further tests to be done by the GP after a month or so, but the letter arrives several weeks after the proposed action. Hospital letters are rarely checked by the author, and frequently have significant typos and errors.

Of course it is absurd that we still don't have emailed letters. Hospitals do of course actually produce letters on computers, print them and send them,  and then we scan them into our computers, which is slow, labour intensive and uses a lot of computer disc space, but hospitals have no incentive to change to suit us. Similarly hospitals make no attempt whatever to address letters to the referring doctor. Even if specialists try to do this, the patient administration systems change the addressee to the doctor at that practice when the patient first attended the hospital even if that doctor has retired years agos.

Some specialists try to get around the problem by handwriting an "Urgent" prescription request, and either giving it to the patient or faxing it. The patient often expects a prescription to simply be issued. However, this is often difficult, either because of illegibility, or ambiguity about dosage or length of treatment, and it puts the GP in a tricky position, as he will be the one responsible.  GPs may have to ring hospitals to clarify the details, which is time consuming and irritating for the patient who does not expect delay. Commonly, GP prescribing software will warn of interaction with the patients other medication. Hospital doctors very rarely use such software( although their pharmacies have it), and the specialist is unlikely to be aware of less common interactions. Prescribing medication at the clinic, for issuing at the hospital pharmacy, would again be better care and more efficient.

Discharge letters, which in our area are the only hospital letters that are electronic, are a massive problem. They tend to be either late, or erroneous (for example medication given to the patient often does not match the medication list!) or lacking important information about follow up appointments. They are often several pages long with lots of irrelevant detail about DNAR forms and thrombosis risk assessment, with the information we need in a small paragraph somewhere on page 3 of 5. Commonly letters include requests for GPs to arrange further test such as blood tests or scans. Whether patients have been informed of this is usually unclear. Quite often the letter will say “GP to arrange scan”, and yet it has in fact already been requested. Our local trust frequently issues " amended discharge letters" sometimes in several editions, all of which require inputting into the patent record. It is rarely made clear what has been amended, if anything.

Sometimes, letters request that GPs request a repeat scan after an interval, and forward the result to the specialist. This can be tricky, with GPs outside the immediate area often not having access to requesting scans at that hospital, and GPs rarely have a system to arrange tests on certain dates. The potential for the ball being dropped and the patient suffering as a result, is huge. It would not be difficult for all such follow up tests to be arranged and administered by specialists, and it would save huge amounts of GP time.

Then there is the patient who has a scan arranged by specialists, and they are told they will have an appointment to discuss the result. But 2 or 3 weeks later the patient has heard nothing and rings the hospital to be told that their appointment might be a month or two yet, and they should visit their GP. Sometimes GPs can look at scan reports (though not images) on the hospital IT system, but the consultation is difficult as the GP is not in a position to plan further action. If, as is not rare, the report is bad, the GP has to break bad news without the necessary knowledge to inform the patient in detail of their options. The GP then has to ring the specialist to demand an urgent appointment,which can be difficult or impossible.

As well as the extra work pushed onto GP’s there is a huge difference in approach best summarised by the famous William Osler who 100 years ago taught young doctors to Ask not what disease the patient has but rather what person the disease has?

Specialists unfortunately often fail to consider the whole patient. A typical example is a patient with moderate Parkinson's who sees a heart specialist with angina caused by narrowed arteries to his heart. He has a series of tests which last a few months, during which time his ability to walk declines, so he is actually getting much less symptoms from his heart. However, on the basis of the pictures of his heart arteries they decide to offer a heart bypass operation. This may offer a slight survival advantage, but there's a significant risk of damaging the brain during the use of the heart lung machine. The op goes ahead and the patient suffers severe confusion afterwards, which improves very slowly. If the surgeon had looked at the whole patient he would have appreciated that the patients chance of more than a couple of years of quality life was minimal, and that his brain was the important issue, not his heart. The harm easily exceeded the benefit, as was entirely predictable.

Then there is the tendency of many specialists to be economical with the truth about the medication they advise. They very rarely take the trouble to tell patients to stop treatment that is no longer needed. They tend not to explain side effects, and they often prescribe drugs that are not licensed for the patient’s condition without informing the patient. This practice is strongly discouraged by the GMC, but it is regarded as normal practice by many specialists.

 There are many illnesses for which drugs offer very little benefit and often risk significant harm, but specialists rarely feel comfortable not offering any treatment, so it often ends up with the GP explaining the downsides, which is not an easy job. If there was a simple rule that prescribing unlicensed medication was not a job for GPs, it would mean that specialists would have to justify their practise to their hospital peers and argue for the necessary budget.

If GPs and specialists worked together the amount of harmful overprescribing that occurs in the UK would be reduced, and money would be better spent elsewhere.

Another problem occurs when GPs are sent reports typically from so called MDT meetings about patients with cancer. The report may detail spread of a cancer and the need for a particular treatment, but often fails to indicate whether the patient has been informed. Patients can now demand access to their GP records online, and it is not rare that very bad news is discovered accidentally.

Then GPs are often treated in a disrespectful way in terms of referral forms, being expected to simply complete a series of tick boxes, often with insulting boxes of text, for example " Does This patient meet the criteria for referral? ". It has been shown time and again that doctor’s intuition is at least as important a clue to serious illness as the presence of particular symptoms, and yet this is ignored, and GPs are treated as idiots. Paper referral forms also ignore the fact that GP records have all been electronic for at least 20 years. If referrals were simply vetted by the specialist, with dubious ones being discussed on the phone, that would be efficient, educational and preserve good relations.

So what about the faults of GPs? Poor referrals are a major issue. GPs often refer patients to a surgeon for knee or back pain for example, when the symptoms are mild and the patient is high risk. The specialist has to disappoint the patient, which is no fun. GPs often try to bend the rules on which referrals are urgent, querying skin cancer when patients have rashes which are obviously not is a common example. Often GPs feel pressure to comply with patient’s requests even though they are aware that they are ignoring the interests of specialists and of course the patients categorised as non urgent.

Specialists are also painfully aware when GPs miss stuff which should have been obvious. Cases do get easier to diagnose in the so called retrospectoscope.

Patients are regularly admitted to hospital as emergencies for long term problems such as asthma or diabetes, that could easily have been treated by GPs. Some of these cases reflect the fact that the specialists are seeing a selected group and are unaware of the patients who have been successfully treated by GPs. Many of these patients are very poor at managing their illnesses but often the GP care is also at fault.  

So what can be done? Specific changes are obvious from the above but we also need some systemic change.

In many European countries all specialists are required to spend a few months working as a junior GP, which would be a big help. Similarly, it would be reasonable to require some broad education for all specialists in subjects such as communication skills, and keeping up to date on pharmacology , especially on the dangers of multiple medication.

Abolishing the foundation trust v GP purchaser provider split that has been created by the internal market would probably help a lot. Some of the most absurd issues like the lack of swift electronic letters to GPs would quickly get sorted.

The old system whereby a lot of GP education was done at evening meeting with local specialists would be a huge help. This was abolished by Academic GPS several years ago. They wanted GPS to learn from each other, often together with nurses and other staff. This was a laudable aim but the result in many areas has been very poor quality meetings, and that GPS and specialists rarely recognise or speak to each other. Facilitating email and phone liaison would help a lot but at present hospitals discourage it as there is no payment tariff for dealing with a patient without bums on seats.