Gardner v Northampton General Hospital NHS Trust [2014] EWHC 4217 (QB)

Case No HQ13X01725[2014] EWHC 4217 (QB)
IN THE HIGH COURT OF JUSTICE
QUEEN'S BENCH DIVISION
Venue Royal Courts of Justice, Strand, London, WC2A 2LLDate 12 December 2014
Before
SIR DAVID EADYSitting as a High Court Judge
Between
COLIN GARDNER (on his own behalf and as widower and Personal Representative of the Estate of Suzanne Gardner Deceased)ClaimantNORTHAMPTON GENERAL HOSPITAL NHS TRUSTDefendant
Gerwyn Samuel (instructed by Access Legal from Shoosmiths) for ClaimantThomas Gibson (instructed by Kennedys) for DefendantHearing Hearing dates: 13, 14, 17 and 18 November 2014
JudgmentSir David Eady :The issues to be resolved
[1]Mrs Gardner was 60 years of age when she attended at 00.34 on 21 December 2010 at the Accident and Emergency department (“A & E”) of the Northampton General Hospital. She was in excruciating pain and, as later emerged, suffering from necrotising fasciitis (“NF”). Her upper limbs were discoloured and swollen and there were weeping blisters (or bullae) on one of her arms. She had fallen and grazed her left elbow on 17 December, at which time she was taking on prescription no less than three immuno-suppressant drugs in connection with rheumatic arthritis (originally diagnosed in March 1994) and these would have rendered her unusually vulnerable to infection. The only treatment she received over the next few hours was in the form of drugs for the relief of pain and the application of gauzes to the suppurating blisters. No one appears to have addressed the possibility of sepsis until 05.00 and she was not given the benefit of the hospital’s sepsis protocol before 08.30.[2]What she needed was an emergency operation for the amputation of both arms and, prior to that by way of preparation, resuscitation by means of oxygen, fluids and antibiotics. By the time her condition was diagnosed (08.30) it was too late and death had become inevitable. The Defendant has admitted some of the breaches of duty pleaded and has apologised to the Claimant and the family for the negligence, but maintains that Mrs Gardner would have died even if there had been earlier diagnosis and she had received reasonably competent treatment within the time available.[3]The Claimant’s surgical expert, Professor Winslet of the Royal Free Hospital, has expressed the opinion that if she had received the necessary surgery at any time up to 07.30 the likelihood is that she would have survived, albeit severely disabled. By contrast, Mr Earnshaw of the Winfield Hospital in Gloucester, called on behalf of the Defendant, is of the view that unless surgery commenced prior to 05.00 it would have been too late to save her. The Defendant has argued that it would be “potentially difficult” to resolve this difference between the experts but that it does not arise, since Mrs Gardner could not have arrived in theatre until 08.30 even if she had received reasonably competent treatment. In any event, timing (“the time line”) is going to be critical in this case. It is necessary, therefore, to scrutinise the history of what took place (and what did not) between 00.34 and 07.30 that morning.[4]Mr Samuel noted an apparent tension between the evidence of Dr Dedi (called by the Defendant as an A & E expert) and that of Mr Earnshaw. Whereas the latter invites the court to conclude that her condition deteriorated up to 05.00, to the point where death became inevitable, the former places reliance on the readings and observations noted at 04.22 in seeking to show that there had been no marked change from her physiological state as at 01.00. That there is such a tension merely illustrates how unfortunate it was that no observations were made at the outset and that her condition was not monitored thereafter. This means that any opinion offered in their absence is bound correspondingly to involve an undesirable element of speculation.[5]There are thus two primary (interlocking) issues to be resolved: How should Mrs Gardner have been treated in A & E from 00.34 onwards by a hypothetical competent medical team? Assuming she had been competently treated, is it probable that emergency surgery could have been carried out in time to save her life? Since the issues of damages have now been resolved, my task is confined to determining these outstanding questions of liability and causation.

What took place following Mrs Gardner’s arrival at the hospital?

[6]It is accepted that the triage which took place in A & E at 01.04 was (just) within the acceptable time limit of half an hour from arrival. She was given Diclofenac at 01.10 and two Co-codamol tablets at 01.15. These analgesics represented the maximum recommended and were in addition to those she had taken before coming to hospital. The triage, it is accepted, was hopelessly inadequate, as no observations were made or recorded save for a Glasgow Coma Scale score of 15. She was then given a triage category of 3 (orange) which required that she be seen by a doctor within one hour of arrival (i.e. at or before 01.34). Mr Samuel submits that she should have been allocated to category 2 (yellow), indicating that a patient should be seen within 10 minutes of arrival. In view of the time lapse prior to triage, that would in effect entail that she should then have been seen by a doctor immediately. In practice, that would probably mean at about 01.15, since it would be unlikely that a doctor would be standing beside her on completion of the triage process.[7]In the light of Mrs Gardner’s unusual symptoms, and specifically the excruciating pain, there is much to be said for that submission. Severe pain, wholly out of proportion to any outward and visible signs, is characteristic of NF. But Mr Mohammed, the emergency medicine consultant called on the Claimant’s behalf, expressly conceded in cross-examination that a reasonable body of professional opinion could have selected category 3. It seems to me that I must accordingly proceed on that basis. It is true that in response to a leading question in re-examination he appeared to qualify the concession, by reference to how much knowledge of the symptoms was to be attributed to the hypothetical reasonable body, but I believe I should take the concession at face value. In practical terms, it may not make very much difference (i.e. the material lapse of time being between 01.15 and 01.34).[8]Unfortunately, Mrs Gardner was not seen by 01.34 or even shortly thereafter. She was not seen by a doctor until 02.57. That is accepted to have fallen short of reasonable standards. The doctor who saw her was an Emergency ST5 (that is to say, the equivalent of an experienced registrar within a year or two of attaining a consultancy) called Dr Obakponovwe. By this time both the swollen arms had weeping blisters. Yet she did not address the possibility of an infection. Her plan was to gain intravenous access for further morphine to be administered and to take a full blood count. In the meantime, Mrs Gardner had been given a dose of Oramorph at 02.45. This would, incidentally, have probably led to a reduction in pulse and respiratory readings (as a consequence of any pain relief). Also, the immuno-suppressants to which I have referred could quite easily have led to a lower temperature than would otherwise have been the case. This was pointed out by Dr Warren, a retired consultant microbiologist, in his report (at para. 37, section 2.2.1). These facts need to be borne in mind when trying to reconstruct what the observations and readings would have revealed, if they had been taken, and whether any should have given cause for alarm.[9]Observations were taken for the first time, and woefully late, at 04.20. It is obvious that, since there had not been any on admission or at triage, there was no observable trend to aid diagnosis. Any deterioration in her physiological state over the past three or four hours would not be available to be taken into account. It is possible that such an aggressive disease as NF might well have caused a deterioration over such a period. It is known, for example, that a steady lowering of blood pressure can be an indicator of NF; yet no such reading was obtained until 09.30. Professor Winslet considers that it would have probably registered as abnormal somewhere between 03.00 and 05.00. Furthermore, as I have said, when using the solitary readings taken at 04.20, little reliance can be placed on them (whether for that specific time or by way of drawing inferences for what might have been the position earlier) unless due allowance is made for the possible impact on any of those readings of the various drugs she was taking.[10]What were the 04.20 observations? Temperature was slightly above 36.5 C. Pulse was recorded at 80 bpm and respiration at 14. Oxygen saturations were at 97%. No early warning score was entered on the form. Had comparable readings been taken between 01.00 and 02.00, they might or might not have been different, but some of the readings could well have been affected, whenever taken, by the drug regime. They would need to be interpreted therefore with some caution.[11]At 05.05 Mrs Gardner was seen by a general physician called Dr Doust. He noted certain matters and did at least address the possibility of infection, but this was linked in his mind to her long standing condition: he needed to exclude the possibility of septic arthritis. He sought a blood culture and also required a CRP test and an x ray of the left elbow. Unfortunately, this was not taken until 06.27 and, when it was, it could not have been properly assessed because a competent physician would have noted areas of gas which pointed strongly to NF.[12]Dr Doust noted that morphine had been taken for pain relief at 02.45 and 04.22. In view of this, he queried whether Mrs Gardner might be suffering from compartment syndrome (caused by bleeding or swelling within an enclosed bundle of muscles). He also recommended intravenous antibiotics (Flucloxacillin 2g), but this instruction was later deleted and the prescribed doses were not administered. Flucloxacillin was not prescribed again until shortly after 08.00. Further morphine was given at 06.20. Nothing else of note happened before 07.00.[13]Meanwhile, the blood sample taken at 03.52 reached the laboratory at 04.50. Since it had clotted (haemolysed), some of the tests could not effectively be carried out. It seems that Dr Doust was aware of the results (probably obtained by telephone from the laboratory), since he noted them at 05.05. At all events, the laboratory asked that another blood sample be taken, although this did not happen until 06.15 and that sample did not reach the laboratory until 11.24, i.e. about six and a half hours afterwards, and eight and a half hours after Dr Obakponovwe made her original request. A further blood sample was taken at 08.50, which was received at 09.44 and reported at 10.27. (When following a hypothetical time line, in order to trace and analyse what would have been the outcomes of reasonably competent treatment, the assumption will be made that any blood sample required would not have clotted.)[14]At 07.00 Mrs Gardner was seen by Dr Wazir, a registrar specialising in trauma and orthopaedics. He lent no support to the septic arthritis theory. It was confirmed an hour later by Mr Campion, an orthopaedics consultant, that this was indeed not the answer. He was the first consultant to apply his mind to her symptoms. At about this time, her tongue began to swell, although that in itself did not point specifically to NF. This seems to have been first considered as a possible diagnosis by another consultant, Dr Kannan, at 08.20. He was a specialist in ITU or intensive care. It was discussed also with a consultant microbiologist called Dr Bentley at about 08.30 and he suggested NF. Meropenem and Clindamycin were prescribed. Since it is now accepted that Mrs Gardner’s life could not have been saved unless emergency surgery had begun by 07.30, at the latest, a consideration of what was happening by this time is largely academic. Yet, for the sake of completeness, I should briefly record what then took place.[15]At 08.30 and 08.35 blood gas analyses were noted (pH 7.305 and lactate 7.5mmol/L). These were consistent with a diagnosis of NF. As noted above, a further blood sample was collected at 08.50, received at 09.44 and reported at 10.27.[16]Observations at 09.00 yielded temperature and pulse readings of 36.1C and 68 respectively. At 09.15 the pulse was at 90. Mrs Gardner’s condition continued to deteriorate. The tongue was still swelling at 09.20 and she had difficulty breathing. In view of her condition, at 09.30 Mr Campion was working on the basis of a NF diagnosis and fluids and antibiotics were at last prescribed with a view to surgical intervention. At that point her pulse was at 100 and blood pressure at 80/40. She was going into shock.[17]At around 09.50 it proved difficult for the anaesthetist to gain access to the right foot and the medication was introduced through the left.[18]At 10.05, a general surgeon, Mr Evans, noted discoloration and swelling in both arms up to the shoulders. It seems finally to have been agreed that NF was the correct diagnosis and that immediate surgery was needed (although far too late). There were two cardiac arrests which Mrs Gardner survived between 10.22 and 10.35.[19]Blood gas analyses revealed readings of pH 6.87 and lactate 14.6, at 10.26, and pH 6.68 and lactate 14.8 at 10.56.[20]Shortly after, at 11.02 Dr Ali, an ITU consultant, also diagnosed NF. Mrs Gardner had difficulty vocalising and her condition was further deteriorating. Yet another cardiac arrest required resuscitation. At 11.05 there was a blood gas analysis of pH 6.690 and lactate 14.8. The next readings were at 11.43: pH 6.870 and lactate 16.0.[21]Plans to operate were abandoned at 12.02 as Mrs Gardner appeared moribund with overwhelming sepsis. Palliative care only was recommended.

The admitted breaches of duty

[22]The following is a summary of the allegations of negligence admitted in the defence:• There was on admission a failure to carry out a complete set of observations, including checking the pulse, the respiratory rate, the blood pressure, urine output, peripheral pulses and capillary refill times for perfusion and to calculate an early warning score so as to track any deterioration (given that she had presented with acutely painful limbs);• If those observations had been carried out, a diagnosis of sepsis would have been made earlier;• There was a failure to test and record blood pressure during the first nine hours (the first reading being at 09.30);• There was a failure to carry out any observations prior to 04.20;• There was a failure to ensure that she was seen by a doctor within 60 minutes (being the appropriate limit for a patient given a triage category of Orange 3, although it was at that point controversial as to whether the 60 minutes began on admission or at the time of triage);• There was a failure by the doctor who saw her (as late as 02.57) to carry out a full examination, including e.g. by testing the pulses, the capillary refill time or the patient’s neurological state (at least one of which would have been abnormal at that time);• There was a failure to report the result of a blood sample taken at 03.50, to establish inter alia the CRP level, before 13.28 (the abnormal reading of 270 would have suggested infection);• There was a failure at each opportunity (during triage at 01.04, and during examinations at 02.57, 05.05 and 07.00) to consider the possibility of underlying infection;• It was negligent for creatinine kinase levels not to be tested at 05.05;• It was negligent for pulses and capillary refill times not to be tested at 05.05 and 07.00;• There was a delay in administering antibiotics from 08.20 to 09.41;• There was a failure to administer clindamycin despite its being prescribed at 08.20;• There was a failure to follow the Hospital’s sepsis protocol.

The outstanding issues on breach of duty

[23]A substantial number of allegations, however, remain in dispute. As I have mentioned, there is the question over the triage category assigned. Also, it is denied that an x ray was called for on first examination (at 02.57), there being no reason to suspect a bony injury. (The Claimant no longer pursues an allegation that an x ray should have been ordered at the first medical review.)[24]So too, it is said that it was reasonable for a doctor to include rheumatoid arthritis in the differential diagnosis while awaiting further investigations.[25]Importantly, it is denied that members of the medical staff were negligent in not diagnosing sepsis on the information available to them, although it is admitted (as noted above) that if appropriate observations had been carried out, such a diagnosis would have been made earlier. This lies at or very near the heart of the case.[26]It is disputed that the low white cell count, combined with raised CRP, should have led clinicians to suspect infection – but that denial is simply on the basis that they had no blood count to review because of the clotting. Obviously, however, if a blood count was appropriate, a fresh sample should have been obtained promptly. Similarly, it is denied that there was any need to check arterial blood gas (which can be done within minutes) or to test, prior to 05.05, creatinine kinase levels.[27]It is not accepted that there was any breach of duty in failing to investigate decreased levels of sensation in the elbows.[28]Although Dr Doust prescribed antibiotics at 05.05, it is denied that it was negligent not to administer any before 08.00.[29]As to the x ray (eventually taken at 06.27), it is not accepted that it was negligent not to review it, despite the fact that it is now known that it did disclose evidence of gas pointing to NF. Another significant question arises as to whether it was appropriate to call in a microbiologist at any time prior to 08.30. The Defendant’s case is that if sepsis had been considered earlier, this would simply have meant that broad spectrum antibiotics should be administered: there would not be any need, in addition, to call for a microbiologist.[30]It is further denied that there was any need for fluids or oxygen within an hour of diagnosing sepsis at 05.05. Sepsis was diagnosed at 09.30 (at which time Mrs Gardner went into shock and it was already at least two hours too late to save her life) and fluid resuscitation proved impossible at that stage because of difficulty in gaining venous access.

What should have happened over the relevant period?

[31]Professor Winslet referred to the need for a “high index of suspicion” with regard to NF, such that a competent middle ranking general or emergency physician ought to have it in mind, despite its rarity, because it is so aggressive and needs to be diagnosed and treated urgently: see also Wong et al., Necrotising Fasciitis: Clinical Presentation, Microbiology and Determinants of Mortality, Journal of Bone and Joint Surgery, 2003, 1454-1460, and Goh et al., Early diagnosis of necrotizing fasciitis, BJS 2014, e119-e125. Both surgical experts agree that it is “a high profile condition”. Nevertheless, Professor Winslet also acknowledged that in 2010 a reasonable body of such physicians might not have thought of NF as a diagnosis when confronted with Mrs Gardner’s particular presentation. Yet, in the light of the expert evidence as a whole, I believe that a reasonably competent emergency physician should have addressed the possibility of infection at an early stage.[32]I was invited to assume in the Defendant’s favour that the A & E department must have been very busy in the early hours of 21 December and, by implication, that there were other patients deserving greater priority than Mrs Gardner. Otherwise, it is said, I would be applying an ideal standard of care rather than a practical or reasonable one. No evidence was adduced from the records or from personal recollection to establish any of this and I do not think it appropriate to speculate.[33]It was important for any assessment of Mrs Gardner’s condition by a clinician, whenever it was made that morning, to take into account her medical history, her drug regime and her recent minor injury. It has been suggested that a cohort of reasonably competent physicians would, even if properly informed of these matters, have been entitled to come to a provisional conclusion or working hypothesis that she was suffering from a “flare up” of her rheumatoid arthritis. There are, however, a number of factors which were or should have been available to any competent clinician making his or her assessment, whether at 01.34 or at 02.57, which(a) pointed away from arthritis as the explanation and(b) required that other possible diagnoses be addressed – specifically infection or sepsis.[34]It was agreed between the experts that, generally, there are eight factors to be considered when addressing the possibility of a sepsis diagnosis. These were identified as follows: a) Systolic blood pressure below 90mmHg b) Pulse over 90 c) Temperature over 38.30C d) Altered mental state e) Raised respiratory rate f) Low or elevated white cell count g) Raised C-reactive Protein (CRP) h) Elevated lactate[35]Obviously, however, this general list is not to be considered in isolation or as being inflexible. The circumstances of an individual patient have to be taken into account. For example, there were four distinctive features of Mrs Gardner’s presentation which, once they were known about, would need to be factored into the equation.[36]I have in mind the following matters: i) The Claimant and Mrs Gardner had experienced the impact on her of earlier arthritic “flare ups”, which were of much shorter duration, and the recent pain that she experienced up to and during her hospital attendance was of a quite different[37]Once arthritis has been discarded as being the likely cause of her unusual cluster of symptoms, a competent clinician should have addressed the possibility of infection (even if not immediately thinking of NF). After all, her GP had the day before (20 December) specifically raised the point in his note of the consultation (“?? infection but odd as bilateral”). Furthermore, the Hospital’s own Guidance for the treatments of patients with Severe Sepsis, first issued in August 2009, warns as follows in the Introduction:
“Sepsis is a leading cause of morbidity and mortality in the UK. Sepsis may occur in any patient of any age group, but the very young, very old and the immuno-compromised are particularly at risk. Also, with medical professionals dealing with ever more complex and immuno-suppressed patients, the presentation of sepsis may be unusual or even insidious.”
[38]Against that background, the hypothetical time line should have been approximately as follows. A middle ranking doctor should have seen Mrs Gardner by 01.34 with the benefit of the observations that should by then have been available. These should have alerted a competent clinician with the relevant experience to the need to investigate the possibility of infection. A blood test should have been arranged and a report could reasonably be expected at some point between (say) 02.45 and 03.15 (a timescale acknowledged as realistic by Dr Warren, the microbiologist whose evidence was relied upon by the Defendant).[39]When received, the report would have shown at least a significantly raised CRP level and confirmed the presence of infection. Accordingly, the sepsis protocol should have been implemented. According to the A&E experts, this process might in practice have taken up to 30 minutes. I will assume, therefore, that the application would have begun by 03.15 or 03.45. This would then have been reviewed after about an hour (i.e. at 04.15 or 04.45). Such a review would have shown no meaningful improvement and a microbiologist should have been consulted by (say) 04.30 or 05.00. The relevant experts are agreed that a competent microbiologist would promptly have diagnosed NF at that point and emphasised the need for urgent surgery.[40]An emergency consultation would then be required with a consultant orthopaedics or general surgery team. The surgical experts were asked if they agreed that it would then take a maximum of one hour from such a provisional diagnosis to prepare the patient for theatre and Professor Winslet agreed. Mr Earnshaw thought, on the basis of experience, that it could take up to two hours. He highlighted the obtaining of consent as a particular delaying factor. I concluded that if a proper sense of urgency were shown, an hour would be manageable. As for consent, there was very little choice and Mr Gardner said consent was actually given very promptly (albeit about five hours too late).[41]There would then be a need for the patient to go to the anaesthetics room to be prepared for surgery and, so far as necessary, resuscitated to improve her physiological state so as to give the optimum chance of coping with that surgery. If the sepsis protocol had been effectively applied, as I am supposing, that should at least have achieved some of the objectives of resuscitation through the means of fluids and antibiotics. It seems reasonable to conclude, even if one still allows an hour for the resuscitation, that the patient could and should have been ready for surgery by 05.40 or 06.10. Either way, this would allow sufficient time for successful completion on Professor Winslet’s assessment, although not according to that of Mr Earnshaw. In these circumstances, contrary to the Defendant’s case, it becomes necessary for me to consider the conflict between them and attempt to resolve it.

What was the latest point at which a successful operation could have been carried out?

[42]Professor Winslet, in expressing his view that an operation could have been successfully carried out up to 07.30, placed considerable reliance on the absence of any evidence that Mrs Gardner’s physiological condition had materially worsened after 05.00. I found his reasoning more convincing in this respect than that of Mr Earnshaw, since the latter’s attempts to show such a deterioration seemed rather weak. The only two pointers he mentioned were the sodium reading and the fact that she appeared rather “sweaty” at around 05.00.[43]To take “sweatiness” first, this is far too vague and ambiguous to justify in itself the conclusion that it had become too late for a potentially life-saving operation. Secondly, Mr Earnshaw at one point in his evidence stated that the recorded sodium level of 129 could be characterised as an early sign that she was in shock. Yet it went up to 131 by 08.00 and, moreover, it is reasonably clear that she did not go into shock until 09.00 or 09.30.[44]Furthermore, the necessary hypothesis requires me to assume that Mrs Gardner had been competently treated and the earlier application of the sepsis protocol would probably have improved her physiological state prior to 05.00. Moreover, Professor Winslet drew attention to the fact that, even without that support, Mrs Gardner had sufficient resources to resist and survive two cardiac arrests between 10.22 and 10.35 and indeed a third as late as approximately 11.02.[45]I note that Professor Winslet considers that if Mrs Gardner had been assigned to triage Category 2 she could have reached the theatre by 04.00, in which case she would have needed bilateral above elbow amputation, rather than the more drastic four quarter amputation. Nevertheless, if a four quarter amputation was required, he still believes that “despite a stormy post-operative period” Mrs Gardner would have still survived, provided surgery commenced no later than 07.30 (paragraph 10.1.22). He also expressed the view that on a balance of probability she would not have required a four quarter amputation, but only the “relatively straightforward procedure” of bilateral upper arm amputation (paragraph 12.1.6).[46]I therefore find, consistently with Professor Winslet’s evidence, that an operation at any time up to 07.30 would on the balance of probabilities have enabled Mrs Gardner to survive.

The appropriate test for causation

[47]I was addressed on the correct test to apply on causation. Mr Samuel puts his case in the alternative. Either the facts justify the conclusion that Mrs Gardner could probably have been successfully operated upon but for the negligent treatment meted out to her in the hospital, or that the Defendant’s negligence made a material contribution to the fact that it became too late to operate with any reasonable prospect of her survival: see e.g. Bailey v Ministry of Defence [2009] 1 WLR 1052. In this context, the law recognises a material contribution as one which is more than de minimis or negligible: ibid. at [46]. (See also the speech of Lord Reid in Bonnington Castings Ltd v Wardlaw [1956] AC 613, 620-621.)[48]I conclude that the Claimant succeeds on either test. Had Mrs Gardner been treated non-negligently during and following her arrival at the hospital, she could have been ready for an amputation well before 07.30 and would then probably have survived (i.e. she would have lived but for the delay caused by the breaches of duty identified above). It would also have been true to say, however, that the aggressive disease would have been making steady (if not necessarily linear) progress in the period between 00.34 and 07.30, but the Defendant’s negligence made a material contribution to her death because it led to such a delay that the advance of the infection could not be averted by surgery.

Conclusion

[49]I am now in a position to state my conclusion on the outstanding questions identified above. As to the first, I find that a competent medical team would have assessed and treated Mrs Gardner in accordance with the time line I have given above at paragraphs [31] to [41]. I do not consider this to be unreasonable or to reflect the application of an inappropriate or “ideal” test. There is some room for “slack”, which allows for the kind of delays or hitches which occur in practice. It could certainly be argued that a competent clinician would, for example, have pressed for blood results in less than one and a half hours, or implemented the sepsis protocol more promptly, or called in a microbiologist at an earlier stage, but I believe that I have incorporated a sufficient margin to allow for the fact that some reasonable practitioners would have taken a more cautious approach. Nevertheless, a greater sense of urgency was clearly required than that displayed on the night in question.[50]As to the second question, I find on a balance of probability that Mrs Gardner would have survived an emergency operation for bilateral amputation and debridement for which she could and should have been ready by no later than 06.10 (i.e. just over five and a half hours after she arrived).

order

she could not bear to be touched. ii) The blisters on her arm, weeping or not, bore no relation to arthritis. iii) The discoloration and swelling, away from the joints, could not reasonably be interpreted either as symptomatic of arthritis. iv) The fact that she had a few days earlier grazed her elbow, while on a regime of three immuno-suppressant drugs, meant that she was more vulnerable than usual to infection: warning bells should have rung.