Tucker v Griffiths

On 13 January 2011, Mr Stephen Tucker was admitted to the Royal Hampshire County Hospital having suffered an acute ischaemic stroke. The cause of the stroke was infective endocarditis of the mechanical aortic valve with which he had been fitted in 2006.

In November 2010, Mr Tucker had consulted a general practitioner, Dr Rosemary Griffiths, at his GP practice, the Andover Health Centre. The question that arises in this case is whether she should have appreciated that he was, or might have been, suffering from infective endocarditis and urgently admitted him to hospital. Had she done so, it is agreed that his condition would have been diagnosed and treated, and he would not have suffered a stroke. The value of the claim, if breach of duty is established, is also agreed.

Infective endocarditis is an infection of the endocardial surface of the heart. It is agreed that it has significant morbidity and mortality risk; 20-25% of those admitted with a diagnosis of infective endocarditis die; 80% have major complication during admission, such as heart failure. There is 95% mortality with untreated infective endocarditis.

Infective endocarditis is an uncommon presentation but a well-recognised one. The condition can present acutely with heart failure or sub-acutely where the condition worsens over a period of days or weeks. The consequences of delayed treatment can be very serious. As Dr Colin Tidy puts it in an internet article entitled “Infective Endocarditis”:

“It is a disease that is easily overlooked or misdiagnosed and clinicians should be vigilant and well versed in the manifestations of IE to avoid missing the diagnosis… The clinical presentation is very variable. A high index of suspicion and low threshold for investigation to exclude IE are therefore essential in at-risk groups.”

Of particular relevance amongst the at-risk groups, for present purposes, are patients with a prosthetic valve. In an article in the British Medical Journal by Beynon and others, it is said that the incidents of prosthetic valve endocarditis range from 0.1% to 2.3% a year and account for 10-15% of cases.

The NHS Choices website says this:

“Endocarditis is a rare condition in England, even in those with a higher risk. It is estimated to affect around one in every three thousand people every year… Although it may sound strange, rates of endocarditis are increasing because of advances in medical care. This is due to an increasing number of people being treated with valve replacement surgery or surgery to repair congenital heart disease.”

The clinical presentation of the condition can be variable and non-specific. It is described by Ineson in his book “General Practice” as being “notoriously difficult to diagnose”. Dr Nicholas Kearsley, the expert GP witness called by the Claimant, says that “sub-acute infective endocarditis is commonly characterised by (a) fever, (b) fatigue, (c) anorexia, (d) back pain and (e) weight loss”. Dr Stephanie Chadwick, the GP expert called by the Defendant, says that the condition:

“may present as an acute, rapidly progressive infection but also as a sub-acute or chronic disease with non specific symptoms – e.g. fatigue, low-grade fever, flu-like illness, polymyalgia-like symptoms, loss of appetite, back pain, pleuritic pain, abdominal symptoms… and weight loss. The majority of patients present with fever, often associated with systemic symptoms of chills, poor appetite and weight loss. Heart murmurs are found in up to 85% and new murmurs have been recently reported in 48% of patients.”

The NHS Choices website lists as the most common symptoms of endocarditis:

“high temperature (fever) of 38c or above, chills, night sweats, headaches, shortness of breath, especially during physical activity, cough, heart murmurs… tiredness… muscle and joint pain… the symptoms of endocarditis are similar of those of other conditions, so it is important that other possible causes are ruled out. Sometimes you may be referred for further tests.”

The Claimant was a serving police officer in the Serious Organised Crime Agency (“SOCA”). In 2006 he underwent aortic valve replacement surgery. That surgery was successful and he remained in his previous work in SOCA and remained fit and healthy.

In August 2010, Mr Tucker developed some chest pain whilst at work. He says that he “had a couple of days off work but I was still just not right”. Mr Tucker attended his GP practice on two occasions during August 2010, seeing first Dr Rosemary Griffiths, the Defendant, and subsequently Dr Katrina Webster. An x-ray of his chest was ordered. He was then seen by another GP at the practice, Dr Samuel White on 2 September 2010. He diagnosed a chest infection. Mr Tucker was treated with antibiotics and says in his statement that “there was a bit of an improvement”.

It appears that that improvement subsisted during September and October 2010 because Mr Tucker did not seek further treatment until 14 November. On the morning of that day he attended an out-of-hours GP service. His presenting condition was recorded as follows: “hasn’t been to the toilet for two weeks, shaking, fever, pain in stomach, nauseous, aortic valve replacement five years ago – concerned about heart rhythm”. Under the heading “history” it was noted that “bowels normal until two weeks ago… wife describes temp and rigors last night (she is a nurse)”. The Claimant was diagnosed with new onset constipation. The impression given by the symptoms was said to be “? diverticulitis, ?? growth, needs to have sigmoidoscopy or colonoscopy. See GP tomorrow for review and prescriptions as infection pro tem”. The prescriptions noted were metronidazole and cefradine, both of which were to be taken for seven days.

As advised, the Claimant telephoned the surgery the following day and spoke to a Ms Vicki Lloyd, a nurse experienced in telephone triage. She recorded the following: “patient used OOH service yesterday with history of constipation for two weeks. High fevers. Doctor advised to see GP today for prescription and referral for sigmoidoscopy. Plan – refer GP same day”. The Claimant attended and saw the Defendant at the Andover Health Centre. She recorded his presenting problem as constipation. She noted that he had had his bowels open twice and was feeling better. On examination she noted that he was apyrexial. He was advised to finish the course of antibiotics and Dr Griffiths recommended that he should be reviewed in two weeks if his symptoms had not settled. Because there had been blood in his urine the previous day at the OOH she arranged a urine test.

At around 19.43 on 21 November 2010, the Claimant attended the Andover War Memorial Hospital Minor Injuries Unit (the “MIU”). He presented with a fast heart beat (a tachycardia) and a slightly raised respiratory rate. It was noted that he used to get tachycardia prior to the aortic valve replacement operation in 2006. That tachycardia was confirmed on ECG.

The examination findings were recorded at some length at the MIU. They included the fact that the Claimant was alert and otherwise well, there was no chest pain, no abdomen pain, no back pain and no neck pain. It was said that there was no acute abdominal cramping and no pain on opening the bowel or passing urine. There was tachycardia but no atrial fibrillation and no acute cardiac arrhythmia evident. The impression recorded was of “dehydration? causing tachy”. It was said that there was no evidence of acute inflammation and no acute coryzal symptoms. It was said that there was “nil to suggest acute cardiac or respiratory event. Nil to suggest acute abdomen… May need bloods. No indication for emergency admission”. The follow up treatment was said to be review by the GP.

The notes at the MIU were made by a nurse practitioner by the name of Henning. However she confirmed her conclusions in a telephone conversion with the medical registrar. The records from the MIU were sent to the GP practice the same day and were recorded as being seen on 23 November 2010.

The Claimant returned to the GP surgery on 26 November 2010 and was again seen by the Defendant. She recorded the presenting problem, history and examination as follows:

“Problem: malaise – symptoms.

History – cough six months, just feels out of sorts. Had bad reaction to antibiotics and had to go to (the MIU) with tachycardia. No other symptoms.

Examination - chest clear.”

The Defendant arranged blood tests, including tests for inflammatory markers for infection, and a chest x-ray. The results came back to the surgery on 29 November 2010 and were reviewed by her. The chest x-ray was clear. The blood tests indicated inflammatory markers, including erythrocyte sedimentation rate at 34 and C-reactive protein at 35. The neutrophil reading was said to be 8.3. The normal range for ESR is 1-10, CRP is 0-7.5 and for neutrophil is up to 7.5. These results were considered by the Defendant that same day. She marked the Claimant’s file as “MAKE ROUTINE APPOINTMENT”.

The Claimant made an appointment and was seen again at the surgery by Dr White on 8 December 2010. Dr White recorded the complaint as:

“malaise…non-specific symptoms. Generally unwell for 3 months, multiple areas of muscle soreness? Secondary to statin which has been discontinued. Aortic valve repair 5 years ago, Echo 5 months ago ok. No shortness of breath/chest pain. No bowel/urinary problems.

On examination - no lymph nodes. Has prosthetic aortic click, chest clear. Abdomen soft, non-tender, no masses. For further tests – raised CRP and ESR noted. ? Early inflammatory arthritis.”

He directed a further review when test results were to hand.

On 13 January 2011, the Claimant was admitted to Winchester Hospital having suffered a right sided cerebral vascular accident at home. The hospital notes described him as having been unwell in the previous 1-2 months with “a non-specific malaise and lethargy”. He was transferred to the stroke unit at Southampton General Hospital. As a result of the stroke the Claimant has sustained a brain injury, which is said to have left him with a right sided paresis, effecting his right leg and right hand as well as problems with speech and language, memory and concentration. His ability to mobilise is impaired and he is now unable to continue working for the police.

When issued, the Claimant’s claim was directed against three defendants. First, Dr Griffiths, who is now the sole defendant. Second, Dr White, the other GP at the Andover Health Centre to whom reference is made above. The third defendant was Hampshire Hospital’s NHS Trust, which administered the Andover Memorial Hospital Minor Injuries Unit where the Claimant attended on 21 November 2010. The claim against Dr White has not been pursued. The claim against the Hampshire Hospitals Trust has been settled.

Only two particulars of negligence are pleaded against the First Defendant. They are as follows:

The Defendant maintained that, although she had no recollection of the consultation, infective endocarditis would have been on her “long list” of possible diagnoses. That was not subject to substantive challenge until final submissions. I deal with that issue at paragraph 70 below.

The focus of the case, and accordingly of this judgment, is on the second particular of negligence, namely whether any reasonably competent general practitioner would have done other than refer the Claimant to hospital for further investigations on an urgent basis having seen the blood test results on 29 November 2010.

The legal principles to be applied in clinical negligence actions are well established and are not in dispute here. They can be summarised as follows:

Both the Claimant and his wife attended the hearing but neither gave evidence because their witness statements were agreed. I have read both statements with some care. The Claimant describes his visit to the out-of-hours GP on 14 November 2010 as follows:

“One night in mid-November Sharon told me that I was suffering rigours. I attended the out of hours GP the next day, 14 November 2010, and told them of my symptoms and my past medical history of a valve replacement. I was diagnosed with constipation and prescribed antibiotics and told to visit my GP the next day. On 15 November I had an appointment with Dr Griffiths who told me continue taking the antibiotics.”

Mr Tucker describes the events of the 21 November 2010 as follows:

“I was suffering from a very fast heart rate which would not settle down and I was still feeling awful so I attended the minor injuries unit again… An ECG was then performed but the nurse told us that my heart was fine. She diagnosed me with dehydration and then sent us home. On 26 November 2010 I attended a further appointment at my GP surgery with Dr Griffiths because I continued to feel unwell… It is difficult for me to remember the details of the appointment but I remember that bloods were taken and an x-ray arranged.”

Mrs Tucker’s statement is to similar effect. She describes explaining to the nurse at the MIU on 21 November 2010 the symptoms her husband had been experiencing, particularly that he had been suffering from rigours. She says:

“at the time of the attendance Steve’s heart rate was fast again… An ECG was performed by the nurse following which she told us that it had shown sinus tachycardia which was okay and nothing to be worried about. The nurse discussed Steven’s symptoms with a doctor at Winchester Accident and Emergency department on the telephone and then came back to us to advise that she thought Steven may be dehydrated…”

Dr Griffiths was the first witness to give oral evidence. She made it clear from the beginning of her account that she had no independent recollection of the events in November 2010. She was dependent on her notes and her knowledge of what would have been her usual practice. Nonetheless, I found Dr Griffiths a singularly impressive witness. She repeatedly acknowledged the limitations of her recollection. She was considered and careful in the answers she gave. She struck me as a careful and conscientious clinician.

Dr Griffiths confirmed the accuracy of her witness statement. In that statement she described the consultation on 15 November 2010, so far as she was able. She said the Claimant’s symptoms were consistent with a working diagnosis of diverticulitis. Accordingly, she advised him to finish his antibiotics and prescribed what she regarded as a “gentle laxative”. She said that she discussed with him the possibility of referral to hospital, but that given that there were no “red flag symptoms” she decided instead to arrange a review in two weeks’ time.

She described the appointment on 26 November 2010. She said that prior to that consultation she would have reviewed his notes and read the note of the consultation at the MIU on 21 November. She would have recorded what he described as his symptoms. She says that due to Mr Tucker’s persisting symptoms she thought further investigation was required and ordered a number of standard general malaise blood tests. In view of his persistent cough she also ordered an x-ray. She says that she did not give him a follow up appointment because she wanted to await the test results before deciding how to proceed.

She says that she then reviewed the tests on 29 November 2010. She said the chest x-ray was normal and she was pleased to note a normal white cell count which showed there was no significant infection. She goes on:

“his neutrophil count was 8.3. Given that the upper limit is 7.5, I did not consider this to be significantly raised. Similarly, his ESR was only moderately raised at 34. Mr Tucker’s CRP was also slightly raised at 35 but none of these results warranted in my view an urgent referral. Moreover, there was still no substantive diagnosis for Mr Tucker’s problems so would have been difficult to decide an appropriate referral anyway. In light of the fact that Mr Tucker’s results were only moderately raised I took the view that I needed to see him again in order to reach a decision as to how best to proceed in my investigation of his condition.”

Dr Griffiths continues in her statement:

“My note on 29 November 2010 states ‘Make routine appt’. Mr Tucker would have been advised to ring for his results at the previous consultation. I note that Mr Tucker made the appointment with Dr White on 6 December 2010 and it would be reasonable to assume that this was when he rang for his blood results.”

That assertion was not challenged.

In her oral evidence Dr Griffiths said that she would have regarded it as reasonable to wait for a few days to review the patient after 29 November 2010. She acknowledged that she had not recorded Mr Tucker’s pulse and temperature on 26 November, but said that the condition he was then suffering from was not acute; he was essentially complaining of a cough of some six months duration. She had examined the chest and would have been able to tell if he was hot or if there was a racing heart. She said that infective endocarditis would have been one possible diagnosis on a “reasonably long list” of possibilities.

She said that in her view there were no “cardinal” signs of infective endocarditis, in particular there was no persistent fever, no night sweats, no weight loss, no tachycardia, no raised pulse and no heart murmurs. Having considered the blood test results on 29 November, she was aware of inflammatory markers with a variety of possible causes. The patient’s condition needed further clarification and further assessment and it was for that reason that she booked a further appointment. She said that the raised CRP and ESR were not in themselves significant given that Mr Tucker had recently suffered from diverticulitis and a possible urine infection. He had also had a sore throat, which suggested a recent viral illness.

Dr Griffiths said that there was no evidence that Mr Tucker had had a high temperature in the previous two weeks. She would have noted that his wife described a high temperature and rigors on 14 November, but that what had been suspected thereafter was an infection of the bowel. By the time of the consultation on 26 November, he was presenting as feeling much better and was apyrexial. She appreciated that the Claimant had sought primary care assistance on four occasions in 12 days. Her long list of possible diagnoses would have included diverticulitis, urine infection, malignancy, respiratory problems causing the cough, general malaise, including aches and pains blamed on statins. But she emphasised that he was not acutely unwell when she saw him.

Dr Griffiths said that she had examined the Claimant’s chest on 26 November 2010; she had listened to the front and back of the chest using a stethoscope. This was the examination of his lungs rather than his heart but if he had a loud heart murmur she would have heard it and recorded it. She accepted that she might not have recorded a soft murmur but she was confident that he had no tachycardia at the time. Dr Griffiths pointed to the fact that Dr White did listen to the Claimant’s heart on 8 December 2010; he noted “prosthetic aortic click” but no other heart murmur. Dr Griffiths said she too would have heard the click of the valve because it is very loud.

Dr Griffiths said that any GP with a patient with a prosthetic heart valve would have infective endocarditis in mind as a potential diagnosis but she would only admit such a patient to a hospital if that diagnosis was a probability. She would reach that judgment by considering the position of the patient in front of her. She would have noted the variable presentations, the fact that a number of his symptoms were commonplace. She was aware that infective endocarditis had serious mortality and morbidity risk and that overlooking it could have serious consequences. However, she made the point that she had many patients with artificial heart valves on her list and it was her duty to exercise a clinical judgement as to whether referral to hospital was necessary. Not all patients with artificial heart valves who developed infection could properly be referred. She was aware that Mr Tucker had been treated with antibiotics and knew that the blood test results might be modified as a result of that treatment.

When she reconsidered the Claimant’s case on 29 November 2010 with the blood test results in front of her, infective endocarditis was still on her list of possible diagnoses. She would have noted that he had had one episode of pyrexia on 14 November, but that was consistent with the presumed diagnosis of diverticulitis. She did not feel it necessary to order an urgent follow up of Mr Tucker; she thought it sufficient to make another routine appointment. She did not regard it as significant whether that subsequent appointment was with her or with one of her colleagues. She said she had found the record of the MIU somewhat reassuring given the lengthy list of negative symptoms recorded there.

I also heard from Dr Kearsley and Dr Chadwick, the two GP experts.

In Dr Kearsley’s opinion, there was here a reasonable cause to suspect infective endocarditis. He said there was persistent illness resulting in multiple presentations. He said there were reports of fever, rigors and tachycardia. He said that there were four visits in 12 days with an underlying malaise and muscular pain. Those symptoms certainly justified taking a blood test and those blood tests revealed inflammatory markers. In his view, no obvious diagnosis fitted the bill but infective endocarditis was a real possibility. In those circumstances it was mandatory to refer the patient to hospital.

Dr Kearsley said that it was acceptable for a GP to delay referral after 26 November 2010 in order to obtain the results of the blood test. But in his view any further delay was unacceptable. He said Dr Griffiths should have effected an immediate transfer to hospital. When pressed he was willing to countenance a delay of perhaps 24 hours but he said no responsible body of general practice opinion would delay beyond that.

Dr Chadwick described the “cardinal signs” of infective endocarditis as persistent fever, persistent fatigue, anorexia and sweats, usually nocturnal. Whilst she maintained that there were cardinal signs of the condition, she agreed presentation could be variable. She said that a critical diagnostic criteria was the presence of persistent fever and that a one off fever would not be sufficient. In her view, a competent GP would review the blood test results with the patient. Blood tests like these might prompt a GP to ask further questions about the patient’s history or condition.

In Dr Chadwick’s opinion it was safe and reasonable of Dr Griffiths to ask the Claimant to come back to the surgery after she had reviewed the blood test results. In her view there were a range of other possible diagnoses, including diverticulitis, urinary tract infection, pulmonary fibrosis, a general viral illness or rheumatic arthritis. It was entirely reasonable of the Defendant to arrange to see the Claimant again. She noted that it took only two days from the Claimant telephoning the surgery to his being seen by Dr White.

I also had the benefit of a joint report prepared by two expert cardiologists, Professor Kevin Channer, on behalf of Hampshire Hospitals NHS Foundation Trust, and Dr Stephen Becker, prepared for the Claimant. These two cardiologists had each prepared their own report but also prepared responses to questions from the parties. The following answers provided by the cardiologists are of particular significance. The cardiologists agreed first that:

“On the basis of the history and the medical records the only time he had rigours suggesting bacteremia was in early November 2010, leading to his presentation on 14.11.10. In our opinion this is when the endocarditis probably started. He had a weak antibiotic which would have modified the natural history but not eradicated infection on the valve. As time progressed, after stopping the antibiotics (after 21.11.10) the infection would have recurred causing damage to the valve ring support structures.”

Second, the cardiologists said that the paroxysmal tachycardia settled before he got home after attending the out of hours service on 21 November 2010. Third, they agreed that it is likely that the Claimant’s heart rate would have been normal after 21 November 2010 (unless there had been further fever or rigors).

There was a good deal of common ground between the parties. It was common ground that infective endocarditis is a rare condition but is one of which every doctor is aware; that the presence of a prosthetic heart valve significantly increases the incidence of the disease; that infective endocarditis is a disease with significant morbidity and mortality. As a result the illness is said by some, including Dr Kearsley, to be a “can’t miss disease”, by which is meant not that it is easy to recognise, but that the possibility of the condition should never be overlooked.

It was common ground that the diagnosis and treatment of infective endocarditis was a matter for hospital clinicians not GPs, and that a GP who “suspected” it, should refer the patient to hospital; but not every patient with a prosthetic aortic valve who develops a fever could or should be sent to hospital. It is a matter for careful clinical judgment whether referral is necessary.

It was also common ground that this was a difficult clinical judgment; in sub-acute infective endocarditis, the presenting symptoms can be vague and various. The common presenting symptoms are not exclusive to infective endocarditis. There is not even a universally recognised list of symptoms to be looked for; different authors suggest different lists. The symptoms which appear most commonly in the textbooks and guidance to which I have been referred appear to be the following: fever, fatigue, anorexia, back pain (all taken from Dr Kearsley’s report); night sweats, heart murmurs (taken from “NHS Choices”); and flu-like illness, polymyalgia, abdominal pain (Dr Chadwick and “Patient Plus”). According to Dr Kearsley, “99% of cases of infective endocarditis have a detectable heart murmur” (a rather higher figure than that given by Dr Chadwick).

There are few, if any, “red flags”, symptoms that point to infective endocarditis and which in themselves demand immediate hospital admission. In its sub-acute form this disease is both indolent and insidious; the condition can persist for weeks or months (as happened here) without reaching a crisis point. It is difficult to recognise and easy to overlook. The nearest thing to a red flag is found in the Oxford Handbook of General Practice, 3^rd edition published in 2010. Above the description of the condition appears the following “! New murmur + fever = endocarditis until proven otherwise”.

It was common ground that the trigger for mandatory referral is “suspicion” of infective endocarditis, and there was no real challenge to the suggestion that this disease merited “a high index of suspicion and a low threshold for investigation”. In the Oxford Handbook of General Practice, the advice is “Admit as an emergency if suspected”. The “Patient Plus” website advises that “You will be admitted to hospital if infective endocarditis is suspected”.

It was also agreed by all concerned that there was a continuum of suspicion ranging from near certainty at one end to faint possibility at the other. The critical difference between the parties concerned the degree of strength of the suspicion of infective endocarditis required to mandate referral. It was agreed that not all competent GPs would require the same strength of suspicion to make a referral; some clinicians are more cautious than others. There are those who would readily refer if infective endocarditis was a possibility; there are others who would want a greater degree of confidence in the possible diagnosis before admitting a patient.

According to Dr Chadwick, it was where infective endocarditis was “probable” that a GP was required to refer. According to Dr Kearsley, referral was mandated if there was “reasonable suspicion of infective endocarditis”. I prefer Dr Kearsley on this point. In my judgment, given the mortality and morbidity risks of infective endocarditis, a GP who failed to refer to hospital a patient with a prosthetic valve, with whom there were reasonable grounds to suspect infective endocarditis, would be falling below the standards of a competent GP.

Dr Griffiths, like Dr Chadwick, said she would admit if she thought that infective endocarditis was the probable diagnosis. For the reasons set out above, in my judgment that is to pitch the test too high. It follows that if, on a proper analysis, it can be said that that made a difference, and that applying the correct test the Defendant should immediately have admitted the Claimant, this claim would succeed.

I emphasis the word “reasonable” in Dr Kearsley’s test; it is not any suspicion that will require referral. I reject Ms Hallissey’s suggestion in her closing submissions that, because the risks associated with infective endocarditis are so marked, any suspicion of the condition would be enough to make immediate referral mandatory; there is no support for that proposition.

In applying this test it is essential that hindsight is not employed. The test must be applied to the Defendant as at 29 November 2010, as she sat at her desk reviewing the blood tests results. It is tempting, but impermissible, to have regard to what it is now known was going to happen to Mr Tucker. Dr Griffiths could not know that and must not be judged by that standard.

What Dr Griffiths knew was that the Claimant had come in to see her 14 days earlier with a history of constipation. He had been to the out-of-hours service the previous day and been diagnosed with constipation. It was thought possible that he had diverticulitis, an inflammatory condition. He had reported a temperature and rigors, and had been treated with antibiotics. Blood was detected in his urine.

By the following day he had been able to open his bowels and was feeling better. The triage nurse at the surgery had recorded from him a history of high fever. However, on 15 November, he had no fever. It is also of note that there was no complaint of weight loss.

Dr Griffiths was also aware on 29 November that Mr Tucker had gone to the MIU on 21 November with tachycardia, apparently caused by dehydration. There had been no back pain, nothing to suggest cardiac arrhythmia or other acute cardiac event, no coryzal symptoms and nothing to indicate that an emergency admission was required.

He had returned to see the Defendant on 26 November with a cough and general feelings of malaise. She had not heard a heart murmur whilst listening to his lungs.

As at 26 November, there was, in my judgment, very little to suggest this was infective endocarditis. There had been reports of fever a fortnight before, involving “spiking temperature” and a competent GP would have considered the possible continuing effect of the antibiotics which might be “muting” (to adopt Ms Hallissey’s expression) signs of infection. But there was nothing to suggest that fever had persisted or returned. The evidence of the Claimant and his wife does not suggest it returned and it was certainly not evident on 26 November. There was no obvious heart murmur. There was malaise but no anorexia, no back pain, no night sweats, no flu-like illness (other than the malaise) and no abdominal pain. In my judgment, considering this history and presentation, this pattern of symptoms tested against the expected symptoms of infective endocarditis noted above, Dr Griffiths could not possibly be criticised for not referring on 26 November.

On 29 November, Dr Griffiths had the blood test results, which showed non-specific inflammatory markers. In my judgment that plainly required further investigation. There were a number of possible causes for the inflammatory markers, including diverticulitis, inflammatory arthritis, urinary tract infection and underlying malignancy, as well as infective endocarditis. Had Dr Griffiths done nothing more she would have been subject to legitimate criticism. But she directed a further routine appointment and that was arranged.

Dr Kearsley suggested in evidence that Dr Griffiths should have admitted the Claimant immediately on reading the test results. But on questioning he changed that stance and suggested it was perhaps acceptable to have referred the Claimant the following day. He said it might also have been acceptable to call the Claimant in for an urgent appointment to discuss the results, but that a routine appointment was unacceptable. I found this evidence troubling; there was no rationale behind the selection of a point in time beyond which it was not acceptable to delay admission. Dr Kearsley accepted that what in fact happened on 19 January 2011 could have happened at any point after early November 2010. In my judgment, if the Claimant’s condition mandated admission on reading the test results, there could be no justification for delaying a moment more.

In contrast to that oral evidence, in his report Dr Kearsley did not suggest immediate referral was required. He said: “Another appointment was needed with the blood test results unless (the Claimant) was completely better and the blood tests were normal” and “the blood tests did not mandate admission by themselves but they did mandate Dr Griffiths revisiting the overall situation”. I accept that evidence. But a reconsideration of the situation was precisely what Dr Griffiths proposed on 29 November and what she then arranged. It was that that led to the consultation with Dr White on 8 December.

Dr Kearsley suggests that Dr Griffiths should have taken steps to ensure that she, and she alone, saw the Claimant on his return. I see no basis for that. It is not a pleaded allegation, but in any event I see no reason why the Claimant’s condition should not be reviewed by another GP at the practice. That GP would have all the notes, the out-of-hours and hospital records and all the test results available, just as Dr Griffiths did, and would be in as good a position as she was to decide what should happen next. Arguably he or she would be in a better position because he or she would be bringing a new set of eyes to the problem and would be able to discuss the matter through with the patient in the light of the test results. That GP would have been as able to admit him to hospital if necessary as was Dr Griffiths.

In my judgment, the history, presentation and test results on 29 November 2010 did not ground a reasonable suspicion of infective endocarditis. Faced with non-specific inflammatory markers in the blood tests, and the history that is revealed in the notes, to invite Mr Tucker to make another appointment to discuss the results was sensible and reasonable. In my view, to decide to direct another appointment rather than immediately to admit was a course which could be expected of a reasonable body of competent GPs, applying what I have held is the proper test.

It is, in that regard, of some interest that 10 days later another apparently competent GP, Dr White, faced with the same patient with the same history and the same test results, also decided not to refer the Claimant to hospital. It is apparent from his notes that he did listen to Mr Tucker’s heart. It appears he heard no murmur other than the click of the prosthetic valve. It appears he took the view that the proper course was to run further blood tests and re-review the Claimant thereafter.

The Claimant discontinued his claim against Dr White in respect of the conduct of his consultation on 8 December. It is, of course, a matter for any individual litigant against whom he chooses to proceed but it is of note that he would have been in as good a position to allege negligence against Dr White as he was against Dr Griffiths. He chose not to do so.

A fair point was made by Ms Hallissey about the heart murmur. Dr Griffiths agreed that when she listened to the Claimant’s chest she was directing her attention to the lungs not the heart. Whilst she was confident she would have detected a loud heart murmur she said she might not have heard a soft one. Mr Hyam responded making the legitimate point that we know that in fact no murmur would have been heard because Dr White did listen to the heart on 8 December and he records no murmur. So far as it goes, that is a sound point; the presence or absence of a heart murmur would be significant in diagnosing infective endocarditis or in deciding whether to refer a patient who might have the condition.

But the forensic significance Ms Hallissey attaches to the point is rather more subtle; she does not suggest, given Dr White’s notes, that a murmur would have been heard but says that the fact that Dr Griffiths did not listen expressly for it suggests that she was not as concerned about the possibility of infective endocarditis as she says she was. That was not a point put to the Defendant. Having carefully considered the possibility that Dr Griffiths was, accidentally or deliberately, misleading me when she said she would have had infective endocarditis in mind as a possibility, I reject it.

As Dr Griffiths said, the fact that the Claimant had a prosthetic heart valve was “all over the notes”. That patients with prosthetic valves are at greater risk of infective endocarditis was very well recognised by GPs in general and Dr Griffiths in particular. I am confident that it was indeed on the long list of possible diagnoses in Dr Griffiths’ mind. Even if she did not listen to the chest specifically with that in mind, she was sufficiently aware of its potential significance to have been attuned to the possibility of hearing a murmur it if was obvious. More particularly, I accept that she reviewed the signs and symptoms on 26 and 29 November with infective endocarditis on her list of possible conditions when she was deciding on the appropriate next step.

I conclude that on 29 November 2010 Dr Griffiths did have infective endocarditis on her mind as a possible, if unlikely, diagnosis, and that it was the response of a competent GP to direct another routine appointment rather than immediate admission to hospital.

In those circumstances, this claim must fail.