Haywood v University Hospitals Of North Midlands NHS Trust

The Defendant is the NHS Trust responsible for the provision of medical services at the University Hospital of North Staffordshire (“the hospital”), and is liable for the acts and omissions of the medical and nursing staff there. On 17^th September 2010, in the Maternity Unit of the hospital, Sarah Haywood (“the Claimant”) gave birth to her first child. The baby was delivered by emergency caesarean section. Whilst in the hospital, the Claimant developed a persistent tachycardia: her pulse rate was measured at 130 beats per minute (“bpm”) both on the night of 18^th September and early in the morning of 19^th September. Testing of a sample of her blood taken at around 0825 on 19^th September showed an elevated white cell count. Later that morning her pulse rate dropped, and shortly before midday it was measured at 80 bpm, which is within the normal range. No further measurement of her pulse rate and no further blood test was carried out before she was discharged from the hospital, on the afternoon of 19^th September 2010, to her home and to the care of the community midwives.

It is now agreed between the parties that in the course of the caesarean section, bacteria entered the Claimant’s surgical wound at a deep level and developed into infection. It is further now agreed between the parties that a reactive thrombocytosis subsequently developed, secondary to the wound infection: this led to the Claimant’s platelet count becoming grossly elevated and a thrombosis developing in her left carotid artery. As a result, the Claimant suffered a stroke on the 3^rd October 2010. Unfortunately, she continues to suffer a right-sided hemiparesis, speech impairment (evident as she gave her oral testimony) and cognitive deficit.

It was not necessary for me to consider the expert evidence which had been gathered in the fields of haematology and stroke medicine, because it is agreed between the parties that the Claimant’s stroke was caused by her post-operative infection. The fact that she suffered that infection does not in itself give rise to any allegation of negligence. It is however said on her behalf that the Defendant was negligent in discharging her from the hospital on the afternoon of 19^th September 2010. In particular, it is said that there was a negligent failure to investigate the cause of the persistent tachycardia so as to exclude the possibility of post-operative infection, and that it was negligent to discharge her on the basis of the single pulse rate measurement of 80 bpm. The Claimant’s case is that her tachycardia, and her raised white cell count, were caused by, and were signs or symptoms of, her developing infection. She should not have been discharged when she was. The Defendant should have considered the possibility of infection and carried out further tests. Had that been done, the results of the tests would have been abnormal, consistent with infection, and the Claimant – before being discharged - would have been given broad-spectrum antibiotics. Her infection would have been successfully treated by those antibiotics, and she would not have suffered her subsequent stroke.

The Defendant admits that treatment with broad-spectrum antibiotics on 19^th September 2010 would have resolved the infection, so that the Claimant would not have suffered her subsequent stroke. The Defendant however denies negligence. The Defendant’s case is that although bacteria were inoculated into the Claimant’s body on 17^th September 2010, her wound infection did not develop until 21^st September 2010. Her tachycardia on 18^th/19^th September, and her raised white blood cell count on 19^th September, were caused not by the infection, but by other factors related to her labour and delivery. There were no indications that further investigation was necessary, it would not have been appropriate to treat with broad spectrum antibiotics at that stage, and it was reasonable to discharge the Claimant from the hospital on the afternoon of the 19^th September.

In this trial, I am concerned only with the issue of whether negligence on the part of the Defendant caused or aggravated the Claimant’s stroke. If the Defendant is held liable to compensate the Claimant, the quantum of damages will be determined at a later date.

As will be apparent from the above brief introductory remarks, the parties have in the course of these proceedings reached sensible and helpful agreement as to a number of important aspects of the case. The matters which remain in issue between them, and which I must resolve, are these:

If those issues be resolved in the Claimant’s favour, the Defendant accepts its liability to compensate her for the harm which resulted from her untreated post-operative infection, even though the stroke which she suffered was a rare consequence of that infection. It is of course for the Claimant to prove her case, and to establish the relevant facts on the balance of probabilities.

The relevant law is not in dispute: I must apply the familiar test stated by McNair J in Bolam v Friern Hospital Management Committee [1957] 1 WLR 582 at p587, when he directed the jury that a medical practitioner –

“… is not guilty of negligence if he has acted in accordance with a practice accepted as proper by a responsible body of medical men skilled in that particular art. … Putting it the other way round, a man is not negligent, if he is acting in accordance with such a practice, merely because there is a body of opinion who would take a contrary view.”

I begin by summarising the evidence I heard about the period when the Claimant was in hospital.

For convenience, I attach as an appendix a table summarising the evidence as to measurements which were made in hospital of the Claimant’s heart rate, temperature, CRP, haemoglobin level, white cell count, and neutrophils level.

The table also shows the scores which were recorded in hospital on the Claimant’s Modified Early Warning Signs (“MEWS”) chart. The purpose of such a chart is to monitor postnatal signs in seven respects, including by the measuring of pulse rate and temperature. If any of the measurements shows a value outside the normal range, a score must be assigned to that measurement in accordance with the instructions shown on the chart. So far as pulse rate is concerned, the chart requires a score of 3 where the heart rate is 130 bpm or more; a score of 2 where it is between 111 and 129; and a score of 1 where it is between 101 and 110. The instructions which accompany the chart show that all 7 parameters must be recorded, the scores added together and the escalation pathway followed. The escalation pathway requires that where the MEWS score is 2-3, or the patient is causing clinical concern, the following action must be taken:

“inform nurse/midwife; contact obstetric registrar to review patient within 30 minutes; initiate treatment as required and document all MEWS scores and any actions taken; increase frequency of observations until MEWS score below 3; if no improvement in MEWS score, seek senior registrar review.”

In the Claimant’s case, the postnatal management plan, noted in the medical records, included completing the MEWS chart five times daily. The evidence, as summarised in the appendix, showed that this was not in fact done.

The Claimant’s evidence in chief was that her pregnancy had progressed well. On the morning of 16^th September 2010, two weeks after her expected date of delivery, she was admitted to the hospital for a planned induction. Her labour did not commence, and she returned home for a short time. She was re-admitted at 2040 that evening. In the early hours of 17^th September 2010, decelerations recorded by the foetal heart monitor gave rise to concern. The midwives sounded the emergency buzzer at 0530, and a registrar attended. At 0540 the decision was taken to carry out a caesarean section. The Claimant was admitted to theatre at 0546, at which time her heart rate was measured at 80 bpm. After an unsuccessful trial of forceps, the baby was delivered by caesarean section at 0610 on 17^th September. The Claimant was transferred to the postnatal ward, where she was first assessed at 1110 that morning. Her pulse rate at that time was measured at 83 bpm.

The Claimant’s evidence was that she felt very tired and drained after the birth. She recollected that in the course of that evening, and the following day, her temperature and pulse were checked a number of times, and she believed there was some concern about her heart rate, which she was told would be monitored. She told the nursing staff that she felt unwell and extremely tired. She did not sleep on the night of 18^th September. She recalled an ECG being carried out on the 19^th September.

In cross-examination the Claimant said more than once that she was unwell, and did not feel normal, but that it was the first time she had given birth and she did not know how she should feel. She denied the suggestion that at 2220 on 18^th September she had told a midwife (Miss Deacon: see below) that she felt well: she said she did not feel well, she had to be helped to the lavatory when she visited it for the first time after being catheterised, and she asked for painkillers. On 19^th September she felt shaky and could not sleep, and she did not remember saying that she was “otherwise well” (as Dr Suraweera recorded: see below). Indeed, the hospital notes show that she had difficulties feeding her daughter, and did not manage much sleep during the preceding night. She denied that she had been able to sleep during that morning.

She was discharged home at about 1610 on the 19^th September. Her evidence was that she explained to the nursing staff at that time that she still felt tired, weak and shaky. She stated that before she left hospital she was told by a nurse, or possibly by a doctor, that she had a high pulse rate but it would settle. She could not remember who it was who had told her that. She did not accept the suggestion that she had been told that she had had a high pulse rate. It was pointed out to her that the discharge letter said that her tachycardia had resolved: that, she said, was not what she was told.

The evidence in chief of Mr Mirko Budimir, the Claimant’s partner, was that she was cold, shaky and in a lot of pain when she returned to the ward after the birth. On 19^th September there was talk of the Claimant being discharged later that day: his evidence was that he and the Claimant were concerned about that, and both of them informed the staff that she felt tired, weak and shaky. He too says they were told that she had a high pulse rate but that it would settle.

Midwife Emma Deacon gave evidence that the ward where the Claimant was cared for contained 28 beds. Unsurprisingly, it was a busy place, where it could be difficult for new mothers to sleep because of the activity around them. For that reason, she said, it is common for new mothers to be tired, and so a complaint of tiredness or fatigue would not necessarily cause her to think the woman was unwell. Similarly, it is common for a new mother to feel pain after a caesarean section, and so a complaint of pain would not necessarily indicate that the woman was unwell. She said that all mothers on the ward were encouraged to tell staff if they had any concerns.

She explained that MEWS charts were at that time a recent introduction. Maternal observations would be carried out approximately every 4 hours during the first 24 hours after a caesarean section, but would usually be discontinued at night (unless there were any particular concerns) because it was considered beneficial to allow a new mother to sleep.

Miss Deacon was on duty on the night of 17^th/18^th September 2010. At 2230 she noted that the Claimant’s temperature was above the normal range, at 37.6 degrees, to which she assigned a MEWS score of 1. She regarded that as only a mild pyrexia and not did not consider calling for medical assistance. She also noted the pulse rate as 91 bpm. She confirmed that the records showed that on 4 occasions whilst on the ward, the Claimant had been given diclofenac, which Miss Deacon knew could suppress temperature. She was not aware (though it is common ground between the parties) that the risk of post-operative infection is higher when a caesarean section is carried out as an emergency, as opposed to an elective, procedure.

The next observations were conducted by one of Miss Deacon’s colleagues, a midwife who did not give evidence. At 0650 on the morning of 18^th September, an entry was made on the MEWS chart which showed a pulse rate in the range 51-100 bpm (MEWS score 0). Further observations at 0940 showed that the pulse rate had risen to 105 bpm (MEWS score 1).

Miss Deacon next carried out observations at 2220 on 18^th September, when she noted in the medical records that the Claimant’s pulse rate was 130 bpm (a figure which she underlined). On the MEWS chart, she recorded the pulse rate as 129 bpm, with a score of 2: she could not explain why she had recorded two different values, but thought the note on the MEWS chart was the more likely to be accurate. All other observations were within normal limits. She stated that the record she made at the time - to the effect that the Claimant was feeling well (something which the Claimant does not accept) and had just mobilised to the toilet – causes her now to think that she considered the tachycardia to be the result of mobilisation and pain, which she would consider normal on day one following a caesarean section. She also noted, however, that she intended to re-check the pulse rate shortly. She did not in fact do so, but could not now recall why that was. If she had done so, and had found that the Claimant remained tachycardic, she would have contacted the on-call obstetric registrar.

As it was, Miss Deacon’s next entry in the records was made more than 8 hours later, at 0700 on 19^th September, when the pulse rate was 130 (which she again underlined), to which she assigned a MEWS score of 3. In accordance with the MEWS escalation pathway, she contacted the on-call registrar, Dr Suraweera, who instructed her to repeat the observations in one hour and to arrange for a full blood count to be performed that day. Dr Suraweera in his evidence confirmed that he had given that advice. Miss Deacon then handed over care to the day staff.

She agreed in cross-examination that the MEWS system is designed to pick up any underlying issues. She also agreed that a management plan had been completed by a midwife to the effect that observations of the Claimant should be carried out five times daily, which had not been done. She said that more commonly the plan would require the MEWS chart to be completed four times a day, but she accepted that even that had not been done: on 18^th September entries had been made on the chart on only three occasions. She agreed that when she measured the Claimant’s pulse rate at 2220 on 18^th September it was outside the normal range. 129 bpm scored 2 on the MEWS sheet, 130 bpm scored 3; but whichever it was, she accepted that she had not followed the acceleration pathway. She also accepted that during the night of the 18^th there should have been further observations of the Claimant, but that had not happened either. At 0700 on the following morning the pulse rate was again measured at 130 bpm, but Miss Deacon said it did not follow that it had been at that level throughout the night, because it might have varied.

At 0800 on 19^th September a midwife, who did not give evidence before me, recorded the Claimant’ pulse rate as 115 bpm. She completed an “assessment of maternal well-being” form. In answer to the question “are there any concerns about the following”, she ticked the “yes” box in two sections:

“Temperature, pulse, respirations and blood pressure

Infection, fever, chills, headache, visual disturbances”

and

“Fatigue

Unable to sleep, restless sleep, extreme tiredness”.

Dr Suraweera, now a locum consultant in obstetrics and gynaecology, was then a junior registrar. He carried out his review at 0825 on 19^th September. He noted slight pallor. He stated that in response to his questions, the Claimant did not report breathlessness or chest pain: the only issue which she raised was that she had not slept well the previous night and was feeling tired. The note which he made at the time recorded

“c/o tiredness, hasn’t slept through night, otherwise well”.

In carrying out his examination, he had in mind the possibility of a deep venous thrombosis in the calf with pulmonary emboli. He found no clinical evidence of such a condition and did not consider it to be the cause of the tachycardia. He recorded his impression as follows:

“? anaemia”

The blood sample had been taken only shortly before Dr Suraweera’s examination, and so the results were not available to him. He arranged for an ECG, the results of which showed a normal regular rhythm but faster than normal. The pulse rate recorded at the time of that test was 110 bpm. His evidence was that he did not suspect infection: the Claimant’s temperature was normal, and there was no overt focus of infection; her chest was clear when he listened to it; and there was no abdominal tenderness, which he would expect to find if there was developing infection. He said that overall, therefore, he had no reason to suspect infection, and did not suspect it. His provisional diagnosis was anaemia (because of the pallor), which he felt would account for the Claimant’s tachycardia, compounded by a lack of rest. He did not request a CRP test because he would expect some rise in the CRP level after a caesarean section, and therefore a raised level would not necessarily indicate infection. As he had no suspicion of infection, he did not perceive any significant benefit in requesting a CRP. Nor was there any reason to prescribe antibiotics, which he said should not be used where there is no clinical need to use them and no benefit will be derived from them. Unnecessary use can be to the detriment of the patient, because antibiotics might lessen the body’s natural resistance to certain bacteria. His plan was to await the result of the full blood count and to allow the Claimant to rest.

The results of the blood tests showed that the white cell count and neutrophil count had risen well above the normal range. However, Dr Suraweera had finished his duty at 0900 that morning, and therefore did not see the results.

In cross-examination Dr Suraweera agreed that any caesarean section carries a risk of infection entering the surgical wound, and that an emergency caesarean section carries a rather higher risk. Tachycardia is one of the signs of infection, though by itself it is not enough to demonstrate the presence of infection. He agreed that when there is tachycardia persisting over 24 hours it is necessary to consider the possibility of infection and to be able to conclude that infection is not the cause of the raised pulse rate. In this case, he said, his initial impression was of anaemia. He asked for an ECG in order to exclude any possibility of a pulmonary embolism. After the ECG he felt able to conclude that embolism was unlikely. He agreed that it was necessary to look at the haemoglobin levels to test for anaemia, but his shift ended before the results of the blood tests were received. Looking at the results now, he felt that the drop in the haemoglobin level coupled with tiredness – both of which can be expected after a caesarean section - provides an explanation for the Claimant’s tachycardia.

As to the white cell count shown by the blood test results, he agreed that a raised white cell count can be a marker for infection, and that the value of 21.5 was almost double what would be expected in a non-pregnant woman. After labour and delivery, however, the level can go up to 25, so he felt it was still normal. He accepted that tachycardia may be a sign of an infection which has not shown itself at the surface, but his view was that he had reasonably excluded infection as a cause of the Claimant’s tachycardia. He stated that the pulse rate had come down before he saw the Claimant, and continued a downward trend, whereas with infection he would expect tachycardia to persist. On his examination of the Claimant there were no signs to suggest infection, so that possibility was not in his mind. He accepted however that the blood test results did not exclude infection, and that infection should have been “kept in the frame”.

Dr Hamad Kayani, then a Senior House Officer, saw the Claimant at 1143 that morning in the course of his routine ward round. He stated that he reviewed her medical records before assessing her: he was therefore aware of her tachycardia, and he knew of the risk of post-operative infection after a caesarean section. The results of the full blood count and ECG were available to him. He said he reviewed the situation afresh. Although he had no recollection of this assessment, he stated that the Claimant’s responses to his questions had led him to record that there was nothing of particular concern: he noted in the records

“Now comfortable. Has not had good sleep last night. Now feeling better had some sleep”.

He measured the Claimant’s pulse rate and found it to be 80 bpm, within the normal range: there was no challenge to the accuracy of that measurement. He listened to her breathing and heard no sounds indicative of fluid retention or a chest infection. His evidence was that she appeared to be stable, with her pulse and heart rate within the normal range and with no evidence of any underlying infection. He took the view that the blood test results were consistent with recent major surgery and did not indicate anything more serious.

Dr Kayani therefore did not accept the suggestion in cross-examination that the Claimant’s tachycardia was unexplained. He said there was a clear sequence of her pulse rate going down since the early morning, and the tests and history pointed to anaemia as the cause of her tachycardia. He commented that in cases of developing infection, the temperature “keeps going up, whatever you do”. Accordingly, whilst infection crossed his mind, there were no signs other than the tachycardia pointing to it, and he felt the tachycardia could be explained by anaemia. He therefore did not accept that he should have advised that the Claimant’s pulse rate should be checked around 4pm and that she should not be discharged until after that was done.

The plan which Dr Kayani had noted in the medical records was “Home when safe”. His evidence was that at that time he habitually used that form of words, because he would carry out a ward round in the morning but it would be usual for a new mother to be discharged later in the day, and her condition could potentially change between his assessment and the time of discharge. He stated that if this did happen, the midwives would seek a further medical assessment rather than discharge the patient home. But his note did not mean that anyone else had to do anything before the woman could be discharged: he would have had no problem with it if the Claimant had left the hospital immediately after he had seen her.

Although Miss Deacon had by this time gone off duty, she gave evidence that from a midwife’s point of view she understood “Home when safe” to mean that a woman could be discharged home when all checks and documentation were complete if both mother and baby were well. Miss Deacon did not agree that further observations should have been made in this case: the Claimant had been fully examined earlier that day, and had then been assessed as fit for discharge. Miss Deacon pointed out that the Claimant might have gone home straight away, in which case there would of course have been no further observations. In her view, therefore, the need to make entries on the MEWS chart ceased once the doctor had assessed the Claimant as fit for discharge. She had also said in cross-examination that if a new mother had been medically discharged, she would discharge her even if the woman said (as the Claimant says she did) that she felt tired, weak and shaky.

It appears that the Claimant left the hospital at some time between about 1620 and about 1700. The hospital’s “discharge summary” recorded no infections, but noted “tachycardia – ecg nad”. A midwife, who did not give evidence, signed the section of the sheet headed “assessed fit for discharge”. That, I think, can only be a reference to Dr Kayani’s assessment, for there is no evidence that any assessment had been made of the Claimant in the hours since Dr Kayani had seen her. As I have noted above, a discharge letter addressed to the Claimant’s GP recorded “tachycardia postnatally - now resolved”, but the Claimant says that is not what she was told.

Mr Gourab Misra, a consultant obstetrician and gynaecologist, was the named consultant during the Claimant’s period as an inpatient, and was therefore responsible for her care, though he had no personal contact with her. He had been contacted in connection with the decision to carry out an emergency caesarean section on the 17^th September. He had not originally been expected to give oral evidence before me, but was called because in cross-examination of the junior doctors Mr Weitzman had suggested that a CRP test should have been carried out, not because a high value would have identified the presence of infection but because a low reading would have excluded infection.

Mr Misra’s evidence was that if Dr Suraweera had consulted him on the 19^th September, he would have endorsed the investigations which Dr Suraweera had carried out, and would not have asked for any further investigations to be made. In the absence of clinical signs of infection, he said, the tachycardia could be explained by slight anaemia coupled with a stress response to the surgery. He said he would expect the clinical team to investigate infection if, for example, there was a temperature of 38 degrees or clinical evidence of a wound infection or a chest infection. An isolated finding of a pulse rate in excess of 100 bpm would not have caused him to request blood cultures or the use of a broad spectrum antibiotic. If he had been called in to review the Claimant, he would not have requested a CRP measurement, because he would in any event have expected it to be raised after a caesarean section and therefore the value would not have affected the management of the patient. If he had been asked to review the Claimant between the 17^th and 19^th September, he would have been reassured that her ECG was normal and her pulse rate had fallen to within normal limits. He stated that he would have regarded the earlier elevated pulse rate as consistent with her post-operative state with a mild degree of anaemia. He would not have asked her to remain in hospital for further monitoring of her pulse rate, and he would not have felt there were clinical grounds to give her antibiotics. Rather, he would have discharged her from hospital on the 19^th September, as in fact happened. Miss Harrison understandably relied on this evidence as showing that Dr Kayani’s actions were entirely reasonable.

Mr Misra stated that the Defendant did not have a specific sepsis guideline at that time. He produced a book “Surgical Guidelines 2010/11” published by the Bedside Clinical Guidelines Partnership (of which the Defendant was a member). In the section dealing with wound infection, this book gives the following advisory guidance:

I turn next to summarise the evidence relating to the days after the Claimant left hospital. She was visited at her home on 20^th September 2010 by a community midwife, Mrs Heather Stone. The Claimant’s evidence was that she told the midwife that she had not had much sleep and was still feeling wobbly and weak. She assumed this was normal after a caesarean section and the midwife confirmed that was so. The Claimant stated that she also reported to the midwife that her wound was sore, which she again assumed was normal. The midwife asked to see the wound, which the Claimant stated was dry but a little pink at that time: the Claimant’s evidence was that the midwife only glanced at it, without looking closely, and said it looked fine. Mr Budimir, who was present, gave evidence to a similar effect.

Mrs Stone told me that when she visits a mother who has been delivered by caesarean section, she expects to see a scar 3-4” in length. Any reddening or oozing is likely to be a sign of infection. If she saw that, she would telephone the mother’s GP from the house. On this visit, she said, the only specific issue which she recalls the Claimant mentioning was some nipple soreness, for which she provided a cream and recommended contacting a support group, “mum2mum”. The note she made at the time was that the wound was clean and dry. She stated that if there had been any evidence of infection – in particular, if the wound had been red or draining fluid - she would have documented it and referred the Claimant to her GP; but she saw nothing untoward. She did not accept the evidence of the Claimant and Mr Budimir that the wound was then pink.

On 21^st September the Claimant was then visited by a member of the “mum2mum” scheme. Asked why she did not say anything to this lady about feeling unwell, the Claimant replied that there was no reason why she should: the visit was in connection with supporting her in breast feeding, and was “all about the baby”, so she did not feel it appropriate to talk about her own welfare.

The Claimant stated that the condition of the wound deteriorated over the next couple of days, and she became increasingly concerned about it. She noticed pus oozing from the wound and so on 22^nd September went to see her GP, who told her that the wound had become infected and prescribed antibiotics.

The GP did not give evidence. The note made at the time reads:

“E: postop. wound infection, unspec

O: distinct flare”

The Claimant’s temperature and pulse either were not taken, or were not recorded; and there is no indication that the GP took a wound swab.

Later on 22^nd September, Mrs Stone made a further visit to the Claimant’s home. She stated that she examined the wound and noted that it was slightly red and that the GP had prescribed antibiotics. She said she did not see any oozing from the wound: if she had, she would have recorded it, and she would have arranged a further visit the next day if she had had any concerns about wound infection. In cross-examination the point was made that the Claimant’s evidence (supported by a note of what she had said at the hospital a few days later, to which I refer below) was that the wound had been slightly oozing when she went to the GP on 22^nd September: Mrs Stone accepted that may have been so, but repeated that she did not see any oozing.

The Claimant’s evidence was that the surgical wound appeared to be getting worse, and in the early hours of 25^th September she went to the hospital. She stated that the wound was oozing and that she was in a great deal of pain and very distressed. She was examined: she understood that both her temperature and her pulse rate were slightly elevated. She said she was shocked to be told that there was infection “four fingers deep”. The wound was drained, though there was still a little leaking after that had been done. A drain was put in place at her side, and she was discharged home.

The nurse who first saw the Claimant on this attendance noted that the Claimant told her that she had been to the GP on 22^nd September with a “slightly oozing wound”. Antibiotics had been prescribed but no wound swab taken. The nurse noted that the Claimant “appears uncomfortable”. She bleeped the registrar to review the Claimant. The registrar’s note says in part:

“Wound redness and oozing from wound.

Generally well. No pyrexia.

About 50ml of fluid (offensive) drained from 1-2cm gape on left side. No deep tissue infection.

NB no signs of systemic infection”

A wound swab, blood sample and CRP test were sent for analysis. Arrangements were made for the Claimant to return for review on 28^th September, and she was told to continue her antibiotics.

The Claimant was asked in cross-examination about the notes that she “appeared uncomfortable” and was “generally well”. She did not regard either as an accurate description of her condition at that time. She pointed out that she was given gas and air whilst the wound was drained: that, she said, was because she was in pain, and had said so at the time.

Laboratory test results from the samples sent for analysis showed that the Claimant’s white cell count and platelet count were elevated at 8.3 g/dL and 556 x 10⁹/L respectively, and her CRP level greatly elevated at 314 mg/L.

On 28^th September, before the Claimant went to hospital, Mrs Stone visited her and noted that the wound was still slightly red, with oozing down the right side.

The Claimant initially attended the hospital at 1400, but could not be seen at that time and was advised to return later. She did so, at 2200, and was reviewed by Dr Fiona Ross, then a Foundation Year 2 doctor. The Claimant’s evidence was that her wound was tender and red, with a gap at each end from which there was a yellowish discharge, and that she reported that she was not feeling at all well. She states that she was nonetheless told that all was fine and was sent home.

Dr Ross’ note of this attendance recorded “1cm gap in the wound at the right hand side; redness much reduced; slight discharge from both ends, yellowish; mildly tender”. She felt that the Claimant seemed much better than she had been when seen by a different doctor on 25^th September. She advised the Claimant to continue oral antibiotics for 5 days and then attend her GP for wound review. Although Dr Ross had no independent recollection of this review, she was also able to refer to a statement which she had been asked to make on 21^st December 2010: this was to the same effect as her note, but also included a record that the size of the gap in the wound had decreased since antibiotics had started, and that the wound was responding to the antibiotics. She had also said in that statement that the Claimant was happy with the plan. Her evidence was based on, and consistent with, these records.

The Claimant denied a suggestion by Miss Harrison that she was feeling much better by 28^th September. She felt awful, she said. She had not waited at the hospital when she could not be seen on her first attendance because there was no one to look after her daughter. She therefore went back home, but returned at 2200 because she was not getting any better and felt she needed to be seen. This part of the Claimant’s evidence was criticised, because Dr Ross’ witness statement had been agreed. With respect to Miss Harrison, I do not attach any significance to the point: the Claimant’s perception of how she was feeling might understandably have differed from the opinion of a doctor comparing present appearance with what had been recorded in a previous doctor’s note.

In the early hours of 3^rd October 2010, after feeding her baby, the Claimant collapsed. An ambulance was called. Investigations at hospital showed that she had suffered a stroke.

It was put to the Claimant in cross-examination that the overall picture was that she did not begin to feel unwell until a few days after the birth. The point was made that the pre-action protocol letter sent by her solicitors had referred to her discharge from hospital on 19^th September 2010 and had then continued “After a few days the Claimant started to feel unwell and she noticed that her wound appeared to be infected.” She denied the suggestion, and repeated that she did not feel right, but she did not know how she was supposed to feel.

Mr Budimir was cross-examined about the terms of a letter which he sent to the hospital on 4^th November 2010. He agreed that it did not refer to the Claimant being ill when she was discharged from hospital (though it did say that she was discharged “despite her having a rapid heart rate”), but explained that was because he wanted to keep the letter short and to emphasise the main point of the seriously infected wound. He had ended the letter by asking:

“Can you please help me understand how my partner, a healthy and fit woman of 36 years of age, could have had such a serious stroke after routine child birth? Could you please send me her medical records so I can try to understand how this has happened?”

He repeated in cross-examination that the Claimant had been shaking and unwell whilst she was in hospital, and whilst at home she had been in pain, suffering constant headaches and unable to sleep.

He was also cross-examined about the fact that large parts of his witness statement are in identical terms to that of his wife, despite which he insisted that he had drafted his statement. I think he must be wrong about that; but it does not cause me to doubt his credibility on other matters, because it may in part be a matter of semantics and in part an example of the complications which can arise when statements are drafted by lawyers on the basis of accounts given to them by witnesses.

My conclusions and findings in relation to the factual evidence are as follows.

I accept the Claimant and her partner as truthful witnesses, doing their best to remember how the Claimant felt at a difficult time more than 6 years ago, and I found their descriptions of the Claimant’s condition convincing. Although Miss Harrison skilfully made forensic points about the terms in which the complaint to the hospital was first expressed, and about the suggested changes from the accounts given by the Claimant and Mr Budimir in their witness statements, I do not regard those points as reasons not to accept their oral testimony. I specifically reject the submission that they have altered their account because of a conscious or unconscious wish to strengthen a claim for compensation. So far as Mr Budimir’s letter of 4^th November 2010 is concerned, his focus was understandably on the question which he posed in that letter, and he would probably have expressed himself in different terms if he had understood that a more precise analysis was necessary of when his wife first felt unwell. As to the Claimant, my assessment is that she is not by nature a forceful character. During the time she was in hospital her baby was surely the focus of her attention, and I readily accept her evidence that although she felt unwell, she did not know whether that was how she was supposed to feel shortly after being delivered of a child. It is not surprising that a new mother in such circumstances, experiencing difficulties and pain in trying to breast feed her baby, should not say much about her own general condition. I accept her evidence that she was told that she had (not had had) a high pulse rate but it would settle: in my view, a statement which would have conveyed to her that there was nothing for her to worry or complain about. It may be that in response to questions asked of her from time to time, she expressed herself in terms which did not forcefully indicate that she was feeling unwell; but nothing in the notes suggests she ever stated in terms that she was feeling entirely well. In those circumstances the terms in which entries have been made in the various medical records do not cause me to doubt the Claimant’s account of her condition.

The evidence of the Claimant and Mr Budimir was that she felt unwell when she left the hospital, that her wound was pink when it was seen by the community midwife on 20^th September, and that it began to ooze on 21^st September. The Defendant disputes that evidence, contending that it is inconsistent with other evidence and unreliable. However, I did not find the community midwife Mrs Stone an impressive witness. Her notes, on which she relied, did not record any pinkness of the wound on 20^th September; but nor did they record any oozing on 22^nd September, when the Claimant visited her GP. Strikingly, her notes say nothing about any gaping of the wound on 28^th September, the date on which a 1cm gap at the right hand side of the wound was recorded when the Claimant attended hospital. In my view, Mrs Stone’s notes are at best incomplete, and do not cast any doubt on the evidence of the Claimant.

The Defendant also relies on the fact that the GP’s record of the Claimant’s attendance on 22^nd September refers to a “distinct flare” but says nothing of any oozing. However, the GP’s note does not record any temperature or pulse measurement, and he does not appear to have taken a swab from the wound. Those are surprising omissions, given that the GP referred to wound infection and prescribed broad spectrum antibiotics. They suggest that it was a brief examination, recorded in very brief terms. This record does not cause me to doubt the reliability of the Claimant’s evidence.

In those circumstances I accept what the Claimant and Mr Budimir told me about the condition of the wound in the days after her discharge from hospital. I find that the wound was oozing on 21^st September, which is why she went to the GP on the following day. I further find that after her visit to the GP on 22^nd September the Claimant continued to feel unwell, and for that reason went to the hospital on 25^th and 28^th September. On the second of those occasions, I readily accept her evidence that the reason she did not remain at the hospital was because she wanted to care for her baby at home and not (as was suggested in cross-examination) because she did not feel unwell. The fact that she returned to the hospital that night clearly shows, to my mind, that she felt seriously unwell.

I therefore accept and find that the Claimant did in fact feel unwell whilst in hospital and in the days after being discharged. It follows that I reject that part of the Defendant’s case which is based on the proposition that the Claimant did not begin to feel unwell until a few days after the birth.

I turn to the expert evidence. Oral evidence was given by Professor KAV Cartwright (retired consultant medical microbiologist, called by the Claimant), Professor Robert Masterton (retired consultant microbiologist, called by the Defendant), Professor Ian Greer (consultant obstetrician, called by the Claimant) and Mr DJ Tufnell (consultant obstetrician, called by the Defendant). In addition to their individual reports, the microbiologists and the obstetricians had prepared joint reports identifying the areas of agreement and disagreement in their respective disciplines. All are well qualified and highly experienced in their respective disciplines.

At the court’s request, the expert witnesses prepared a helpful note of definitions which they agreed for the purposes of this case. They define infection as the invasion of an organism's body tissues by disease-causing agents, their multiplication, and the reaction of host tissues to these organisms and the toxins they produce. Infection can be clinically evident, in that it demonstrates recognisable signs (such as fever or tachycardia) or symptoms (such as general malaise or pain); or it may be subclinical, in that there are no recognisable clinical findings.

The expert witnesses define sepsis as a syndrome shaped by pathogen factors and host factors (e.g. sex, race and other genetic determinants, age, comorbidities, environment) with characteristics that evolve over time. They say that what differentiates sepsis from infection is an aberrant or dysregulated host response and the presence of organ dysfunction. They point out that sepsis-induced organ dysfunction may be occult, and it is therefore necessary to consider its presence in any patient who presents with infection. A first step in this is to consider whether the systemic inflammatory response syndrome (SIRS) is present.

In relation to the microbiological issues in this case, there was a substantial measure of agreement between Professor Cartwright and Professor Masterton. They agreed that there is a well-recognised risk, occurring in about 5-10% of cases, of postoperative wound infection following caesarean section. Infection may be introduced into the wound despite the taking of all reasonable care. In this case, they agreed that the Claimant developed a deep post-surgical infection when tissues were exposed at the time of the surgery. They both regarded the Claimant’s tachycardia on and after 18^th September as consistent with infection at the time of surgery, though the tachycardia could have had other causes, eg haemorrhage. They agreed that the slight rise in the Claimant’s temperature on the evening of 17^th September was consistent with infection, but could have had other causes. It was “at most a marginal pyrexia that taken alone would not have caused concern in the immediate post-partum period”. They agreed that the diclofenac which was prescribed for the Claimant could have reduced her temperature and so could have masked any pyrexia caused by infection, but Professor Masterton felt it unlikely that the effect of diclofenac would have completely masked any significant pyrexia caused by infection.

Professor Cartwright and Professor Masterton agreed that the white cell count and neutrophil count on 19^th September were unusually high (even taking into account the fact that both values can rise during labour and after surgery) and were both consistent with, and probably caused by, post-operative wound infection. They agreed that on 19^th September there was no evidence of chest or urinary tract infection.

The expert microbiologists differed in their opinions as to when clinical signs or symptoms of the post-operative infection first appeared. Professor Cartwright’s opinion was that the Claimant’s tachycardia on 18^th/19^th September 2010, and her raised white cell and neutrophil counts on 19^th September, probably were caused by the infection at the time of surgery. Professor Masterton took a different view: in his opinion the white cell count on 19^th September (which was not associated with either tachycardia or a raised temperature) was probably caused by post-operative wound infection which had not yet declared itself; but he felt that the Claimant’s tachycardia during the period 17^th – 19^th September probably was not caused by the infection, because there was no significant pyrexia during that period, and the tachycardia showed a reduction to a normal level on the day when the high white cell count was found.

Both Professor Cartwright and Professor Masterton agreed that infection was “much the most likely cause” of the raised platelet count on 25^th September. As to how long it would have taken for the platelet count to reach that level, they agreed that there was no literature which enabled an accurate assessment to be made. Professor Masterton’s opinion in the first joint statement was that it would have taken “a few days”, whilst Professor Cartwright’s opinion was that it would have taken “around 5 – 7 days. It takes a minimum of around 5 days for the platelet count to rise consequent on infection”.

They agreed that the CRP level on 25^th September was grossly elevated: it was indicative of “serious, well-established bacterial infection” and was the result of that infection. In view of that elevated level, Professor Cartwright’s opinion was that the infection was likely to have been present for “several days or more”. Professor Masterton, however, thought it was consistent with “the start of the severe clinical infection around 21/22 September”.

They agreed that the raised platelet count on 3^rd October, when the Claimant suffered a stroke, was caused by the infection.

They were asked in their first joint report to consider what would have been found if further tests had been carried out on 19^th September. They agreed that if there had been a test of the Claimant’s CRP level it would probably have been moderately raised, perhaps in the range 50-100 mg/l. In Professor Cartwright’s view that would have added considerable further weight to the other pointers to the possibility of wound infection. In Professor Masterton’s view, such a level would have been consistent with bacterial infection but not indicative of it: it could, for example, be related to the surgery on 17^th September. He agreed however that if further tests had shown a raised CRP level in conjunction with a white cell count of 21.5 and neutrophils of 18.7, that would have been indicative of a bacterial infection and the patient should have been managed accordingly.

They agreed that, if a broad spectrum antibiotic had been prescribed for the Claimant on 19^th September, the probability was that her infection would have been successfully treated, the raised platelet count on the 3^rd October would have been avoided, and she would not have suffered a stroke on that date.

In their second joint report, Professors Cartwright and Masterton commented further on the Claimant’s tachycardia. Professor Cartwright accepted that there were other potential causes of tachycardia, such as anxiety and anaemia, but in his opinion the most likely cause of the Claimant’s tachycardia on 18^th and 19^th September was deep infection. He regarded infection as the only credible cause of the tachycardia on 25^th September, which occurred in the absence of fever. He was unable to offer any explanation for the pulse rate of 80 bpm at 1143 on 19^th September, “other than observing that the pulse rate can vary quite widely over relatively short periods of time (eg a few hours) when infection is present”. Professor Masterton remained of the view that, whilst the tachycardia on 19^th September was consistent with the presence of infection, it probably was not caused by infection, having regard to the absence of any associated significant features of infection and the presence of other possible explanations. He too was unable to explain the differences in the Claimant’s pulse rates during 19^th September, and agreed that there is a well-recognised variability of pulse rates. He felt that the pulse rate of 80 bpm was consistent with the records showing that the Claimant was “generally well” and had no pyrexia.

As to what was likely to have happened over the next 48 hours after 19^th September, Professor Cartwright’s opinion was that the pulse rate was likely to have risen again from 80 bpm because deep infection was present and continuing untreated. It would probably have returned within about 4 – 5 hours to around 100 bpm, possibly more. Having risen to that level, his opinion was that it would probably have remained raised for most of the period 19^th – 25^th September. He therefore felt that if the Claimant’s pulse had been taken prior to her discharge between 4pm and 5pm on 19^th September, it would probably have been 100 bpm or more. Professor Masterton remained of the view that the pulse rates recorded on 17^th – 19^th September probably were not due to infection, and he concluded that over the following days the Claimant would have demonstrated a normal pulse rate as “there was no evidence that she was unwell or experiencing any event that would give rise to a significant tachycardia”. However, if the court were to find that the Claimant was suffering from an active infection at the time of the discharge from hospital, then he agreed with Professor Cartwright’s evidence as to the likely pulse rates over the following days.

The two experts had considered papers by Hartmann and others in 2000 (Footnote: ¹), and by Sivasankar and others in 2014 (Footnote: ²). They summarised those papers as showing that white blood cell counts in pregnancy often exceed moderately the values in the same woman when not pregnant. White cell counts commonly rise much more during labour, and during and shortly after a caesarean section, especially a caesarean section carried out after a period of labour. The white cell counts return to normal quite quickly after labour and after caesarean section. The Hartmann paper tried to define more helpfully than previous research the contribution of infection to a rise in the white cell count. It showed that the greater the increase in the white cell count on the first day after caesarean section (compared with antepartum counts), the greater the likelihood that infection is present. The conclusion of that paper was that labour influenced the white cell count in women who had undergone caesarean sections, but did not obscure changes associated with infection: information gained from changes in white cell counts could therefore be used to assess the risk of infection. The Sivasankar paper reported a study aimed at establishing the normal range of white cell counts in pregnancy. It concluded that, whereas previous research had shown that a moderately raised white cell count may be normal during pregnancy, even a white cell count raised as high as 25 x 10⁹/L or more could not be regarded as diagnostic of infection. The authors commented that if a woman appears well, there could be little justification for giving prophylactic antibiotics on the basis of her white blood cell count alone.

The witnesses had also considered a paper by James, Steer and others in 2011. (Footnote: ³) This showed that in a study of women sampled 1 to 3 hours after delivery, white cell counts were in the range 9.4 – 25. However, women who had been delivered by caesarean section were excluded from the study: Professor Cartwright expected such women to show higher white cell counts.

The only antepartum white blood cell count recorded in the Claimant’s case was a value of 12.7 on 16^th June 2010. On 19^th September, the second day after delivery, it was 21.5. In their second joint report, the experts agreed it was therefore likely that if her white cell count had been measured on the first day after delivery, it would probably have been around 100% higher than the antepartum level, which would have given a likelihood ratio of 5.8 for a clinical diagnosis of infection. A white cell count greater than 25 is very unusual even in labour. They concluded that the Claimant’s white cell count on 19^th September did not predict the presence of infection but raised the likelihood of infection by about fivefold.

In his oral evidence about the raised white cell count value of 21.5 on 19^th September, Professor Cartwright said that a value of 21-22 can occur in an uninfected patient but is very unusual on day 2. Therefore the white cell count of 21.5 in this case, whilst not diagnostic of infection, raised very considerably the possibility that infection was present. He drew support from the Hartmann paper, which he said showed that a high white cell count was not diagnostic but would push the clinician to think more closely about the possibility of infection.

Professor Cartwright said that, after further consideration of the Hartmann paper, his opinion was that the white cell count on the first day after the birth would have been at a level which had a 36% predictive value: ie, about a third of patients with such a white cell count will have infection. He explained that from that point onwards, two factors would be at work: the white cell count could be expected to fall as time passed after labour and delivery; but in the presence of infection, in its early stages, the white cell count would rise. He felt therefore that the white cell count as measured on the 19^th September would probably have stayed around the same level until the antibiotics prescribed on 22^nd September brought it down towards the level of 13.2 measured on the 25^th September.

In cross-examination he accepted that, even if the white cell count had indicated a 36% likelihood of infection, that would mean there was a 64% chance that there was no infection. He said however that if he had seen such a white cell count in a woman in the Claimant’s position, with some clinical changes which were pointers to infection (although they may have had other causes), he would have advised further observations and investigations. He would therefore not have discharged the Claimant from hospital as was done. His opinion was that if a further blood test had been done on the 20^th it would have shown that the white cell count had not fallen, which would have been a further indication of infection.

As to the CRP level, the evidence was that in the presence of infection it starts to rise after 12-18 hours. It is then expected to double every 8 hours until it reaches a level of around 300-400, after which it rises no further. Professor Cartwright stated that the Claimant’s CRP level of 314 measured on 25^th September had been suppressed by the antibiotics she had been given on 22^nd September. He referred to further papers which he had seen since preparing his initial report. Taking those into account, he now felt that if the Claimant’s CRP had been measured on 19^th September it would probably have been in the range 100 – 150 mg/l, and the level would have increased subsequently. He therefore accepted in cross-examination that, whilst a low value would have been an indication against infection, the likelihood is that a CRP test of the Claimant on the 2^nd day after her baby was born would in any event have shown a value of 100 or more.

Professor Cartwright explained that when bacteria which have been introduced into the body during surgery establish themselves, they will probably start to multiply in 12 or 18 hours. They do so slowly at first, but there is then an exponential rise, with the number of bacteria doubling in an hour or two. He agreed that it is not commonplace to give prophylactic antibiotics after caesarean sections, and that clinicians are cautious about prescribing antibiotics because of the risk that bacteria will develop which are resistant to antibiotics. He accepted that this Claimant suffered a wound infection which did not develop into sepsis. She did so because bacteria were introduced during the surgery. She did not suffer a pyrexia.

Professor Cartwright explained that the Claimant’s infection did not respond to antibiotics on the 22^nd September because by then there was an accumulation of fluid in the infected wound: he said that antibiotics do not work well with a collection of fluid. If however antibiotics had been given on the 20^th, when there was no, or less, fluid in the wound, then he would have expected them to be effective.

As to the Claimant’s tachycardia, Professor Cartwright said in cross-examination that he thought it very unlikely that a pulse rate of 130 bpm had been caused by anxiety and/or anaemia. Nor could such a value on day 2 be explained by the labour and delivery. It was much more likely to have been caused by infection: he pointed out that the pulse rate rose to 101 even after antibiotic treatment. Although there had been a downward trend, his opinion was that if further measurements had been taken over the 12 hours after the last observation, they would have been 100 bpm or more, because the infection was developing. He therefore regarded the pulse rate of 80 taken by Dr Kayani on the morning of 19^th September as aberrant and inexplicable. He thought that in the context of tachycardia due to infection, there was “no great difference” between the values of 130 and 110: it is not rare for pulse rates to vary by 20-30 bpm. He regarded the value of 80 however as bizarre. It did not however disprove the presence of infection, even if it was made accurately, because there was a slowly progressing infection deep within the wound.

The absence of pyrexia or signs of local infection when the Claimant’s pulse rate was 130 did not strike him as atypical when the tachycardia was due to infection. The bacteria were very deep within the wound, and therefore changes at the surface would not be expected to be apparent at the outset. Further, in his opinion the organisms were anaerobic, and would therefore cause infection which was slow to develop. He said it was not unusual to have an absence of pyrexia, and a wound which was not tender for a day or two, but nonetheless to have another sign of infection, which in this case was the tachycardia.

In re-examination, Professor Cartwright said that if (as the Defendant contends) the infection only started to develop late, despite having been inoculated into the Claimant’s body at surgery, he found that inconsistent with the drainage of 50 ml fluid on 25^th September 2010 despite the administration of antibiotics on 22^nd September, and inconsistent with the rise in the platelets. He thought it very unlikely that the organisms lay dormant deep within the wound for several days, and in his view the platelet count made it “a virtual certainty” that infection was developing during that period, because it takes many days for the platelet count to rise as a result of infection: an absolute minimum of 5 days to rise to the level of 500 or above.

Professor Masterton began his oral evidence by summarising his opinion that if the Claimant’s tachycardia of 130 bpm on 18^th/19^th September had been a clinical expression of the post-operative infection, he would have expected also to observe a raised temperature, the Claimant feeling generally unwell and her condition deteriorating over 24-48 hours rather than her remaining clinically well. He accepted that organisms had probably been introduced into her body at or around the time of surgery, and therefore would be multiplying in her body during the time she was in hospital, but his view is that they were not producing clinical signs at that stage. He commented that a raised pulse normally goes with a raised temperature; an increase of about 10 bpm for each half-degree rise in temperature. But here, there were no significant temperature differences.

Professor Masterton did not regard the white cell count of 21.5 on 19^th September as being related to bacteria introduced during surgery. He agreed that such a value should have prompted a consideration of the possibility of infection, but the evidence of the clinicians was that they had looked for signs of infection and found none except the tachycardia which had returned to a normal level. He concluded that the elevated pulse and white cell count did not reflect infection at that time. He thought that the infection became symptomatic on the 21^st September. He indicated however that if the court found infection was present on the 19^th, then he would accept that the pulse rate and white cell count were due to infection.

Professor Masterton thought it highly probable that the raised platelet level on 3^rd October was caused by infection. He and Professor Cartwright agreed that there was no literature which enabled them to say how long it would have taken for the platelets to reach that level, and their personal experience differed: Professor Cartwright believed it would have taken 10-14 days, with the result that the infection was present on 17^th September; but Professor Masterton thought the platelets would rise after 3-5 days because of reactive thrombocytosis, and though it would take longer than that to reach this level, he did not agree it would take as long as 10-14 days.

As to the collection of fluid, he agreed a significant quantity had been drained on 25^th September, but thought it could have collected within 3 days, not the longer period suggested by Professor Cartwright. He said they both agreed that the infection started on the 17^th September, but they differed as to when it became clinically evident. In his view, the collection of fluid did not assist in determining the timeline of the infection.

The white cell count on the 19^th September could be explained by infection, or it could just be at the top of the range which is normal in the Claimant’s circumstances. It was a high value in the expected range and should have prompted consideration of infection because such a value carries a significant association with infection. Applying the research in the Hartmann paper, there was a 1 in 3 chance that it was caused by infection. His opinion was that it was unrelated to infection and was merely high in the range. If however the court found that it was elevated on the 19^th partly because of infection, then he would accept that the tachycardia was also likely to be due to infection.

In his opinion, if further readings had been taken on 19^th September, the temperature, pulse rate and blood pressure would all have been normal.

In cross-examination he acknowledged that after the reading at 1110 on 17^th September, the measurements of the Claimant’s pulse were all tachycardic until the reading at 0800 on the 19^th. He also agreed that a raised pulse is not always accompanied by a raised temperature, and that although tachycardia usually goes along with changes at the surface, it can precede them, though it would be unusual to precede them by several days. If the wound had been pink on the 20^th, and oozing on the 21^st, he would expect the white cell count and CRP to have begun to be abnormal on the 19^th. If antibiotics had been given on the 19^th or the 20^th, the subsequent development of the infection, and the stroke, would have been avoided.

He agreed that the platelet level on 25^th September, 556, was significantly elevated. It must have been reacting to infection for some time. Professor Cartwright thought it would have taken 5-7 days to reach that level; Professor Masterton thought it would have taken 5 days.

His overall view was that he accepted that all the readings and findings were consistent with the tachycardia being caused by developing infection, but he did not agree that it was probably so. He emphasised the importance of considering all the information rather than viewing the pulse rate in isolation. The falling pulse rate indicated normality, and whilst it can go down during infection it is not likely to do so. The white cell count was within expected ranges, and the Claimant was clinically stable.

Both Professor Cartwright and Professor Masterton were willing to make appropriate concessions, and both were impressive witnesses. After careful consideration, I am persuaded that Professor Cartwright’s evidence is to be preferred. With all due respect to Professor Masterton’s expertise, there are in my judgment a number of features of his evidence which collectively lead me to the conclusion that his opinion is less likely to be correct than that of Professor Cartwright:

In short, the defence case, and Professor Masterton’s evidence, depend on the coincidence that for three days the Claimant showed no sign of the developing post-operative infection from which she was in fact suffering, but instead showed signs – consistent with infection – of some other problems such as anaemia, tiredness or anxiety. Professor Cartwright’s evidence persuades me that it is very improbable that an infection as severe as that which the Claimant suffered would have failed to declare itself at all until 21^st/22^nd September.

For those reasons I accept the evidence of Professor Cartwright that on the balance of probabilities, the Claimant’s post-operative infection was the cause of her persistent tachycardia on 18^th/19^th September 2010 and of her elevated white cell count on 19^th September. I accept Mr Weitzman’s submission that there was a gradual progression of symptom as the multiplying bacteria caused the infection to develop, and that the tachycardia and raised white cell count were signs of that developing infection. I find that the subsequent findings of pinkness of the scar, oozing from the scar on 21^st and 22^nd September, and the collection of about 50 mls of fluid by 25^th September, were further signs of the developing infection and not the first declarations of infection.

Turning to the expert obstetricians, Professor Greer and Mr Tufnell, they too agreed that the risk of infection after an emergency caesarean section is around 5-10%. They agreed that the pulse rate readings on 18^th and 19^th September showed a persisting tachycardia. Mr Tufnell’s opinion was that it had resolved by the time she was discharged from hospital. Professor Greer however took the view that the single pulse rate of 80 bpm measured on 19^th September was inadequate to confirm that the problem causing the tachycardia – which had not yet been identified – had resolved. He noted that the pulse rate fluctuates in the presence of infection, so that more than one check is needed. They agreed that postnatal tachycardia may have a number of causes, including anxiety and tiredness, blood loss/anaemia, pain, dehydration, a cardiac problem, pulmonary embolism or infection. They also agreed that diclofenac would suppress pyrexia but would not completely mask it in the presence of significant infection. They agreed that the ECG carried out on the morning of 19^th September neither excluded infection nor established that the Claimant was no longer suffering from persistent tachycardia.

Professor Greer was the chair of the advisory committee, and Mr Tufnell was a regional advisor, involved in the production of the British Journal of Obstetrics and Gynaecology 2011 supplement, “Saving Mothers’ Lives” (Footnote: ⁵). It was produced because infection had become the most common cause of postnatal mortality. A section entitled “Back to basics: sepsis” identified a number of “red flag” signs and symptoms which should prompt urgent referral for hospital assessment: these included “sustained tachycardia >100 bpm”. The section also warned that –

“A normal temperature does not exclude sepsis. Paracetamol and other analgesics may mask pyrexia, and this should be taken into account when assessing women who are unwell.”

A later chapter on sepsis, which is directed specifically to genital tract sepsis but is nonetheless relevant to the present case, includes a box which lists the signs and symptoms of sepsis. It includes the following:

“Persistent tachycardia > 100 bpm is an important sign which may indicate serious underlying disease and should be fully investigated.”

In the same chapter, warning is given that sepsis is often insidious in onset, and carers need to be alert to any changes that may indicate developing infection.

Professor Greer referred to that publication as showing the need to consider infection and sepsis where there is persistent tachycardia. He also referred to other reports which emphasise the value of the MEWS charts in alerting clinicians to problems. In those charts, a pulse rate of more than 120 bpm on a single occasion is a red flag sign for medical assessment.

As to the circumstances in which the Claimant was discharged from hospital, Professor Greer’s opinion was that her tachycardia had been of a level seen as a red flag for possible infection. Although no other physical signs were observed, infection may be subclinical. The white cell count was substantially elevated, which was consistent with (though not diagnostic of) infection. The haemoglobin reading on the morning of 19^th September excluded anaemia as a cause of the tachycardia. In his opinion, which he supported by reference to guidance published by the RCOG, the absence of other physical signs did not exclude infection, and assessment for sepsis (though it was not in fact present here) was necessary. The white cell count was higher than usually expected after a caesarean section, and combined with the tachycardia it emphasised the risk that infection was present. In his view, the cause of the raised white cell count was infection together with the change in response to labour and section. No explanation had been found for the tachycardia, the possibility of infection had not been investigated or excluded, and so in his view further tests – CRP, pulse rate/vital signs and full blood count - should have been carried out before discharge. The appropriate treatment would have depended on the results, but antibiotics would have been given if the CRP had been substantially elevated.

In his opinion, the possibility of infection had not been adequately investigated, as subsequent events confirmed. The single pulse rate reading of 80 bpm was not sufficient to ensure that the condition had resolved, and the observations carried out earlier in the day should have been repeated. He regarded the Defendant as being in breach of duty because the Claimant should not have been discharged without ascertaining the cause of her tachycardia. He said that after Dr Kayani’s examination on the 19^th September, there was still no explanation for the overnight tachycardia, and it was inappropriate to discharge the Claimant on the basis of one measurement of her pulse rate. He said that there should have been at least one further assessment before she was discharged: at the very least, measurements of her temperature, pulse and blood pressure. If that had been done, and if it had shown that the pulse rate had risen, then the position would have been: a 5-10% risk of infection following an emergency caesarean section; an unexplained tachycardia overnight; a full blood count which showed only mild anaemia, and a high white cell count, which raised the possibility of infection; therefore overnight observation was necessary.

Professor Greer agreed in cross-examination that there was no pyrexia in the Claimant’s case, even when allowance is made for the effect of the diclofenac which she was given. He described that as a well-recognised but unusual feature in the presence of infection. He acknowledged that he had not been involved in regular work on a labour ward since about 2006, but said that this case concerns the management of tachycardia in a postnatal woman, which has not changed much in recent years. He agreed that the pulse measurements of 115/110/80 showed a downward trajectory. He repeated however that the tachycardia remained unexplained, because the mild anaemia was not sufficient to have caused it. He said that the white cell count of 21.5 was “pretty high” and could not be dismissed: he relied on the Hartmann paper, pointing out that women who had laboured and then undergone a caesarean section was the very group which Hartmann had sampled. If the white cell count was due to delivery, it would be expected to fall; but in the presence of infection it would stay the same or rise. A further CRP test would in his opinion have been elevated, and if it were very elevated, that would suggest infection.

Professor Greer accepted that the clinicians who cared for the Claimant thought the tachycardia was explained. But, he said, none of the findings had ruled out infection, which is life-threatening; so further assessment should have been carried out to see whether infection could be excluded and it was safe to discharge. The fact that she would be discharged to the care of the community midwife did not make it reasonable to discharge her, because the hospital staff could not rule out infection and therefore did not know whether she would develop sepsis. It was not appropriate to rely on a single pulse measurement, because pulse rates go up and down: he said that you could not, on the basis of one reading, “dismiss 24 hours of tachycardia including several red flag readings”. It was therefore essential to carry out at least one further set of observations and make a further assessment of the Claimant’s condition. It was inappropriate to discharge her without doing so.

Miss Harrison suggested that the approach which Professor Greer was suggesting would result in very many postnatal women being kept in hospital. He did not accept that: he said that if you kept in everyone with the Claimant’s symptoms, you would not be keeping many. He did not suggest that antibiotics should have been given on the basis of the tachycardia alone; but if further tests had shown that the pulse rate had gone up, the CRP level was around 150 and he white blood cell count was around 21.5 (instead of having gone down as would be expected in the absence of infection), then it would have been unreasonable not to start antibiotics, because of the risk of severe infection. In view of the tachycardia and the white cell count, it was in his opinion inappropriate to discharge her on 19^th September without further tests after about 4 hours.

Professor Greer said he would not expect obstetricians to be aware of the Hartmann paper, but he would expect them to know that changes in white cell counts can be relevant.

Mr Tufnell in contrast was of the opinion that the blood test had shown mild anaemia, there was no pyrexia, the ECG was normal and the tachycardia had settled. The fact that the tachycardia had settled made it unlikely that infection was present. The white cell count in his view was in the normal range for a woman two days after a caesarean section following labour, and was explained by the labour and birth. In those circumstances, whilst infection was in the differential diagnosis, there were no clear signs of it, and no further checks were necessary. In his view, any further checks of the Claimant’s pulse and temperature – had they been made – would probably have been normal. Infection was unlikely, and if it developed it could be treated. It was therefore reasonable, and entirely in line with the expected standards of care, for the Claimant to be allowed home and transferred to community services.

Mr Tufnell’s observation about the published guidance relied on by Professor Greer was that it related to persistent tachycardia in the context of sepsis or severe sepsis, which was not present at any stage in this case. He accepted that the infection which developed several days after the birth had its origins in the Claimant’s surgery, but said she did not have a clinical infection when discharged from hospital. He said that clinicians are encouraged to be careful about the use of antibiotics, and he would not prescribe them on the basis of raised inflammatory markers without evidence of a clinical focus of infection or signs of sepsis rather than non-specific signs of infection.

In his oral evidence, Mr Tufnell said he did not think that the Claimant’s tachycardia was caused by infection: on the basis of his experience in practice, he thought it was very unusual for wound infection to result, within 48 hours, in signs such as were seen in this case; and he would have expected the tachycardia to be accompanied by pyrexia or visible infection. He said that if the tachycardia had been caused by infection, he would have expected the Claimant’s temperature to rise, her tachycardia to worsen and she would have felt rotten. He agreed with Mr Misra’s evidence as to the appropriate management, and said that you would want to see a local sign of infection before deciding to keep a mother in hospital or carry out more tests or prescribe antibiotics. Sufficient tests had been done, and in his opinion it was reasonable to discharge her that afternoon to the care of the community midwife. He said that he commonly saw white cell counts in the high teens or early twenties on day 2, and so the Claimant’s white cell count of 21.5 on the 19^th September did not help him say there was infection present: it might be, so one would look for a focal clinical sign of infection. In the absence of that, the white cell count could be accepted as a normal post partum result. In his opinion, the clinicians acted reasonably in discharging when they did. In view of the downwards trajectory of the pulse rates, he thought that any further checks would have shown a heart rate in the normal range. He would have expected the white cell count to have gone down later on the 19^th September and it would probably have declined significantly on the following day. A responsible body of doctors could have formed the view that the tachycardia was explained by tiredness, mild anaemia and dehydration. In general, it is not a good idea to keep women in hospital after birth, and it would have been reasonable to discharge her even if a further pulse measurement had been taken and found to be around 100 bpm. The purpose of the red flag designation is to remind clinicians of the need to perform a reasonable clinical assessment: in his opinion, they did so in this case.

Mr Tufnell pointed out that there is no such thing as clinical certainty and that infection cannot be excluded with uncertainty. It is therefore necessary to consider what care is reasonable in particular circumstances. He also commented that most post-operative wound infections are identified in the community rather than in hospital, because they take time to develop.

In cross-examination of Mr Tufnell, Mr Weitzman placed considerable weight on passages in Mr Tufnell’s initial report which he suggested showed a failure by Mr Tufnell to take significant observations into account. Mr Tufnell had referred in his report to the Claimant’s “initial pulse in the morning” of 19^th September, when the rate was 115 bpm, and said that “in the afternoon her pulse had settled to 80 bpm”. He did not mention the readings of 130 bpm on the night of 18^th September and early in the morning of 19^th September. He had then referred to the pleaded case that it was inappropriate to discharge the Claimant from hospital in the absence of any explanation for her tachycardia on the basis of a single reading of 80 bpm and that her tachycardia should have been recognised as a risk factor for subsequent complications. He responded to that case as follows:

“My view is that this is a completely unreasonable approach to a woman in the immediate postnatal period. It is not uncommon to have a slightly raised pulse which settles. Tests were performed to exclude a cardiac problem. Her mild anaemia would be an explanation for a slightly raised pulse. There were no signs of infection. It was entirely reasonable for her to be discharged home on the 19^th September 2010.”

In his answers to cross-examination, Mr Tufnell acknowledged that the Claimant’s pulse rate of 130 bpm was significantly raised and could not be described as “slightly raised”. He also acknowledged that a significantly raised pulse could not be explained by the Claimant’s mild anaemia. He said that his omission to mention certain observations and measurements in his report did not mean that he had failed to take them into account. He said he had taken all the information into account, and accepted the possibility that the Claimant might have had a pulse rate around 130 bpm throughout the night. He accepted that in the presence of significant tachycardia one should think of infection, but her pulse had settled on the morning of the 19^th September and the approach of the clinicians was therefore reasonable.

Mr Tufnell said that women are commonly dehydrated and tired in the 36 hours after birth (though he accepted there was no evidence of dehydration in this case). She had a mild anaemia, and pain could also have been a factor after she had mobilised for the first time after the birth. He thought these, rather than the infection which was present, explained the tachycardia. He said however that if the court found that the tachycardia was caused by wound infection, then the other explanations he had put forward would only be minor factors.

I found Professor Greer an impressive witness, willing to make concessions where appropriate and clear in his criticisms of the Defendant’s care of the Claimant. I reject the submission that his evidence is undermined because he has less experience than Mr Tufnell of current everyday clinical practice on an obstetric ward. Although now retired from such practice, Professor Greer has comparatively recent experience of it; and in any event, no one suggests that there is any less need to monitor a newly-delivered mother than there was a few years ago. I agree with Mr Weitzman that Professor Greer’s evidence as to what should have happened, and why the Defendant was in breach of duty to the Claimant, did not involve any counsel of perfection or the attaining of an unrealistic standard of care. I accept his evidence.

I am afraid that Mr Tufnell’s evidence was in my view seriously undermined by his failure to refer in his report to the high pulse rates. It seems to me that that omission devalues the report in which he set the course for his subsequent oral evidence: it was a wrong course, because it was based on an incomplete recital of the details of the Claimant’s persistent tachycardia. In giving his opinion that the Defendant had treated the Claimant with reasonable care, he failed expressly to acknowledge the significant findings as to her persistently-elevated pulse rate which were on any view consistent with the presence of infection. I do not accept the submission that Mr Tufnell was lying, and I am prepared to accept his evidence that he had those findings in mind when he wrote his report, even though he did not mention them. But if he did, then in my view he failed to give sufficient weight to them, and the conclusions in his report cannot be regarded as reliable, because they did not address the most important factors militating against his conclusions. Although he did address those factors in his oral evidence, his conclusions did not in my view survive that scrutiny.

In addition, Mr Tufnell’s opinion – like that of Professor Masterton – involves the coincidence that signs and symptoms were observed which were consistent with the infection which was in fact developing in the Claimant’s body, but were not caused by that infection. With all respect to his experience and expertise, he was unable to persuade me that in all the circumstances of this case, it was reasonable to regard the Claimant’s persistent tachycardia as satisfactorily explained by mild anaemia, fatigue and/or anxiety.

My conclusions are as follows.

Were the Claimant’s tachycardia and raised white cell count whilst she in the hospital caused by her developing post-operative infection? In my judgment they very probably were. I have already said that I accept Claimant’s evidence as to her condition in hospital and after discharge. As to the cause of those signs and symptoms, I accept the expert evidence of Professor Cartwright ascribing them to her developing post-operative infection. I reject the contrary opinion of Professor Masterton.

Would further tests on 19^th September have revealed the presence of post-operative infection and so led the Defendant to administer broad spectrum antibiotic? In my judgment, it is very probable that further tests would have shown that post-operative infection was present or likely to be present and that after further review the Claimant would have been prescribed antibiotics. In particular, I accept Professor Cartwright’s evidence that if the Claimant’s pulse rate had been taken on the afternoon of 19^th September it would have been elevated to 100 bpm or more. It would therefore have shown that it was not safe to view her pulse rate as having returned to the normal range.

I therefore find on the balance of probabilities that by the morning of 19^th September the Claimant was afflicted by a developing post-operative infection which could and would have been detected by further tests and could and would then have been successfully treated.

Was it negligent to discharge the Claimant from hospital at around 4.30pm on 19^th September 2010 without first carrying out further tests to exclude the risk of post-operative infection? In my judgment, it was, for the following reasons:

I conclude that in all the circumstances of this case, no reasonable body of doctors would have discharged the Claimant on 19^th September 2010 without having carried out further tests, which I find would have shown the presence or likely presence of infection and would have led to the Claimant being given broad spectrum antibiotics which (as is common ground) would have saved her from her subsequent ill health and stroke. I therefore find that there was breach of duty on the part of the Defendant, and that the Claimant’s stroke, and its continuing consequences, were caused by that breach of duty.

I accordingly find in the Claimant’s favour, and will receive written submissions as to any consequential orders.