Appendix A: Independent clinical advice to Commissioner

The following independent advice was obtained from Dr Penelope Jeffrey:

‘The office of the Commissioner is seeking my opinion on the care provided by [the] District Health board … to [Mrs A] during her two visits to the  Emergency Department (ED) during the period [Days 1–3].

I have read and agree to follow the Commissioner’s Guidelines for Independent Advisors, and have prepared my report in accordance with these guidelines.

My Qualification is MBChB (1993, Otago).

I am vocationally registered in Emergency Medicine, as a fellow of the Australasian College for Emergency Medicine since 2008.

I have practised as an Emergency Physician in a tertiary hospital in New Zealand from 2009–present.

I advise that commenting on the initial Emergency Department (ED) assessment, management and discharge of [Mrs A] on [Day 1], and the subsequent assessment in the ED on [Day 2] falls within my area of expertise.

I have commented on the admission to the Intensive care unit on [Day 3] under the care of a Surgeon and an Anaesthetist however this is my opinion as an Emergency Physician.

I have no personal or professional conflict of interest with regard to this case.

I have been provided and have read copies of the following documents:

1.  Complaint by [Mrs B] dated 12/07/18

2. Response from [the DHB], consisting of a response from [Dr C] a Rural hospital medicine and General practice Registrar and a response from [Dr E] PGY (post graduate year) 3. There was no response provided by the [DHB] from other doctors involved in the admission on [Day 3] — [radiologist], Surgeon [Dr F], Admitting ICU [anaesthetist].

3.  Copy of clinical records and Radiology images …

4.  Ambulance care summaries from [Day 1] and [Day 2]

Advice requested

I was asked to review the above documentation and advise

1. Whether I consider the care provided to [Mrs A] during her admission to ED on Day 1, was appropriate and consistent with the accepted standards of practice at the time;

2. Whether I consider that [Mrs A’s] presenting symptoms and clinical history were appropriately assessed and treated on Day 1;

3. Whether I consider the care provided to [Mrs A] following her readmission to ED on Day 2, was appropriate and consistent with the accepted standards of practice at the time;

4. Any other matters in this care that I consider warrant comment

For each question, I was asked to advise:

a. What is the standard of care/accepted practice?

b. If there has been a departure from the standard of care or accepted practice, how significant a departure (mild, moderate, or severe) do I consider this to be?

c. How would it be viewed by my peers? (Note — as this opinion is my own, and unless stated I have not consulted with peers, so I have not commented specifically on the view of my peers).

d. Recommendations for improvement that may help to prevent a similar occurrence in future.

Summary of Events:

[Mrs A] was [a woman in her seventies] with a history of high blood pressure and dyslipidaemia (risk factors for heart and blood vessel disease) and depression. She had previously been assessed for chest pain in 2013, including an Exercise test, and the advice to future doctors in the clinical record was to ‘have a low threshold for angiography if further chest pain’.

[Mrs A] took heart and blood pressure medications (Metoprolol, Quinapril, Hydrochlorthiazide and Aspirin) as well as a cholesterol lowering medication and an antidepressant.

At 09.30 on Day 1 [Mrs A] developed chest pain while at rest at home, and after 10 minutes an ambulance was called. This was dispatched immediately and paramedics attended at 10.21.

[Mrs A] was found lying on the floor. [Mrs B] states that [Mrs A] fainted in her chair, paramedic notes state “family lay [her] on the floor”, but did not document loss of consciousness as such.

A nurse in attendance at the scene had inserted an IV and was administering IV fluids — 100ml was given.

The paramedics report that [Mrs A] described pain lasting 3–5 minutes — onset while sitting at a table ([Mrs B] reports onset while bending over), felt in the centre and left chest, crushing in nature, and associated with nausea. At the time of this initial paramedic assessment, however, she had no chest pain (her pain score was 0/10), no shortness of breath, no nausea or abdominal pain, no bowel or bladder symptoms, and she was not dizzy on standing.

She was pale and her skin was cool to touch but a strong radial pulse was present.

[Mrs A’s] heart rate was low (not unusual in a patient taking Metoprolol), and her blood pressure was very low at 80/60.

Her breathing rate and oxygen levels were normal.

En route to hospital [Mrs A’s] blood pressure normalised and her pallor resolved. At some time after 11am, she voiced that she had developed pain in the chest on taking a deep breath.

An ECG was done by the paramedics at 10.37 — this was abnormal. T wave inversion was present in leads 2, 3 and AVF, and likely but less obviously T wave flattening in leads V4–6.

The ECG machine interpreted the T wave inversions correctly as “Abnormal ECG. Inferior/lateral T wave abnormality may be due to myocardial ischaemia”. Other less concerning abnormalities of “Sinus bradycardia, poor R wave progression and long QT — probable normal variant” were also automatically interpreted on the ECG.

It should also be noted here that the less concerning findings were more prominently displayed on the ECG, but also that automatic reports should always be backed up by human interpretation. The ECG was not mentioned in the paramedic documentation.

The Ambulance officers recorded the “Primary clinical impression” as “Atypical chest pain” and “code 3: unlikely threat to life” was assigned.

At 11.20 the ambulance arrived at hospital and Triage assessment occurred at 11.27. [Mrs A’s] pain was 2/10 in severity, worse on breathing in, and was noted to be radiating into the neck. The nurse noted there was no shortness of breath or cough.

The triage nurse noted the low blood pressure in the ambulance but that this had now normalised after the 100ml of saline.

The nurse assigned a triage code of 2 (= should be seen within 10 mins).

The triage nurse did not note seeing the ambulance ECG and did not document any concern reported by paramedics regarding the ECG.

An Emergency department ECG was done at 11.27.

This was also abnormal — showing now clear T wave inversion in leads V5–V6. The T wave inversion in leads 2, 3 and AVF seen in the ambulance had resolved.

[Dr C] viewed the ECG and wrote a correct interpretation noting a low heart rate and the presence of T wave inversion in leads V5 and V6.

Blood tests were taken.

At 11.45 [Mrs A] was assessed by [Dr C]. The documentation of this assessment is in the form of the final discharge summary completed at 16.57.

[Dr C] documented that [Mrs A] had experienced “acute onset” of chest pain radiating to the neck and associated with diaphoresis (sweatiness), feeling cold and clammy, with lightheadedness and nausea. The pain was noted to be worse on inspiration, and improved on sitting forward. [Dr C] noted that there was “no history of SOBOE” (shortness of breath on exertion). It is not clear if this is relating to the current symptoms or days leading up to the presentation, as [Mrs A] had not mobilised since the onset of pain it would suggest in the lead up period — and no shortness of breath is documented by paramedics or at triage, however in [Dr C’s] response letter to the commissioner he states that [Mrs A] was experiencing shortness of breath at the time of his assessment, as is also described by [Mrs B].

[Dr C] noted that vital signs were normal.

Cardiovascular Respiratory and Abdominal Examination was noted as normal other than some fine crepitations (audible sounds) in the base of the right lung.

Vital signs were documented between 12.15 and 14.50 and were relatively normal. No further vital signs were done after this (in [Mrs B’s] complaint she states that the blood pressure cuff and oxygen monitor were not reattached to [Mrs A] after a trip to the bathroom).

No further pain score was documented during the admission apart from the 2/10 on arrival.

A Troponin (a highly sensitive blood test looking for heart muscle damage) was normal. Other blood results were also unremarkable.

At 12.10 Intravenous fluid of 1 litre was given as a “bolus” (given all at once) with intravenous Diclofenac (anti-inflammatory and pain relief) plus oral Aspirin (used if a doctor is suspecting impending or current heart attack).

At 12.50 Paracetamol pain relief was given — suggesting ongoing pain at this time.

At 13.50 a “Pink lady” was given — this is a term sometimes still used for liquid antacid mixed with local anaesthetic and is used to relieve pain from acid reflux.

At 13.50 Nursing notes state that [Mrs A] had been for a Chest Xray and had “slight nausea”.

The Xray showed abnormal changes at both lung bases. This was reported by [the radiologist] as possibly due to pulmonary oedema (fluid developing in the lungs due to heart problems) or evolving infection. It is not clear whether the report was available to [Dr C] — likely not as the report is documented as being dictated at 07.09 the next day. The Aorta does not appear abnormal on this Xray however this is common in Aortic Dissection.

[Dr C’s] documentation does not mention his interpretation of the Chest Xray.

At 14.30 Nursing notes state that [Mrs A] was having cups of tea and eating sandwiches.

At 15.16 a further ECG was done — the T wave inversion in V5 and V6 were the same as that on the first arrival ECG. New T wave inversion was also now present in lead AVL and probably lead 1 also, although this is subtle.

A further Troponin blood test taken at 15.30 — the result of this (normal) was reported at 16.34.

[Mrs A] was discharged at 1700.

The discharge summary states [Mrs A’s] chest pain had been relieved by the “Pink lady”, that [Dr C’s] impression was “chest pain of oesophageal cause” and she was discharged with a prescription for anti-reflux medication. The discharge summary indicated written and verbal advice was given but not what that advice was. [Dr C] in his response advises that he told [Mrs A] to return if her pain or shortness of breath recurred.

At 19.42 on [Day 2] an ambulance was again called to [Mrs A’s] home, was dispatched immediately, and reached her at 20.21. She had told family she was experiencing nausea, and she had then been observed to collapse and become unconscious with seizure activity and urinary incontinence.

The same nurse as the day before was in attendance at the scene and reported to paramedics that on her arrival [Mrs A] was sweaty and grey, not fully conscious, and not answering questions coherently, with laboured breathing, but a good pulse and saying “no” when asked about pain. The collapse of the previous day was noted, and that [Mrs A] had been discharged with likely indigestion as she had “responded well to a Pink Lady” and Troponins had been negative.

[Mrs A] was observed by the paramedics now to be conscious but pale, with grey and mottled skin colour, and she was sweaty and short of breath. She denied pain but reported she had been experiencing chest pain through the night after discharge and that day before the collapse. Her blood pressure and oxygen levels were low. Her heart rate was normal.

Oxygen was applied, Aspirin was given and an IV inserted, with 200ml of fluid given.

An ECG was done — this was reasonably unremarkable — the T wave inversion seen the previous day had resolved.

En route to hospital [Mrs A] improved although she still remained pale. Her blood pressure which had remained very low during transit had improved slightly but was still low for someone with hypertension at 100/80.

She still denied pain and the pain score was 0/10 throughout transport. The clinical impression noted was “Collapse and Hypotension” (low blood pressure) and a code of “2 — possible threat to life” was assigned.

[Mrs A] arrived at hospital at 21.45 and was seen by the triage nurse immediately.

The nurse noted that [Mrs A] had experienced chest pain throughout the day with nausea, and a collapse with seizure activity and incontinence. [Mrs A] looked pale with mottled legs and was agitated.

At 21.50 another IV was inserted and blood tests sent.

At 22.35 1000ml of fluid was given over an hour.

At 2200 [Mrs A] was assessed by [Dr E]. Her documentation is thorough and found on the ED Medical assessment form and a later Medical admission form.

[Dr E] began by noting the presentation to ED the day before with chest pain, “improved with a Pink Lady”, with stable ECG, no interval changes, and unremarkable blood test results.

She noted [Mrs A’s] past history of Hypertension and Depression and also of Chest pain in 2005 and 2013 with the ECG and Exercise test results from 2013 and the recommendation for low threshold for angiography if ongoing chest pain. She also noted a syncopal (collapse) episode in 2004.

[Dr E] then took the history of the current presentation — she noted that [Mrs A’s] chest pain had been gradually worsening since the day before and reported it as “? like reflux” (but that [Mrs A] had not suffered from reflux before), a “gripping” pain, radiating into the neck. She noted [Mrs A] also reported “some” but “not significant” amount of abdominal pain, and that she felt short of breath, lightheaded, dizzy and unwell.

[Dr E] noted the history of collapse with warning of feeling lightheaded, the family’s observation of a seizure lasting 5 minutes, and then decreased responsiveness and looking like she was going to die, and the hypotension at the scene.

[Dr E] documented that [Mrs A] had vomited once since arriving in ED, with blood speckling in the vomit, and had now passed 2 large bowel movements with approx. 150 ml fresh red blood mixed into the stool.

She also documented that [Mrs A] complained of neck pain which was a chronic (longstanding) problem for her.

[Dr E] examined [Mrs A] and noted she looked unwell, was pale and cool to touch, with prolonged central and peripheral capillary refill (signifying poor blood supply to the skin common in many causes of severe unwellness), an irregular pulse, normal heart sounds, otherwise unremarkable chest examination (but only examined at the front and sides of the chest — as is often the case when a patient is very unwell and Xrays are planned) and a soft non-distended abdomen with upper abdominal tenderness to palpation, and normal bowel sounds.

Vital signs are charted every 20–30 minutes — [Mrs A’s] heart rate was steady at 70–80 and blood pressure initially low, improving a little but remaining low after fluid administration. Oxygen levels were slightly below normal to low normal — requiring a small amount of oxygen to achieve this initially but no oxygen required after a few hours. Pain score was not charted.

An ECG was relatively normal with only occasional extra beats seen.

Blood results showed initially normal haemoglobin (which can be a measure of blood loss but can be normal early in even significant blood loss), with moderately raised white cells and moderately raised CRP (the former can be a stress response or sign of infection or inflammation, the latter a more reliable sign of infection).

Kidney function blood tests had deteriorated slightly compared with the day prior. Troponin was raised, compared with normal the day before. (Troponin elevation can be a non-specific sign: causes include a heart attack, but also heart strain due to severe unwellness of any cause, or in the context of abnormal kidney function can be elevated in the absence of heart problems.)

Liver tests were mildly abnormal. Glucose was mildly raised which can be a non-specific illness response.

A Chest Xray was taken. This showed worsening abnormal lung changes plus fluid accumulation in the lung bases (these changes are reasonably non specific and can be due to problems in the lung such as infection, or problems with the heart, or can occur in other general severe unwellness including with cholecystitis).

The heart and the mediastinum (which includes the first part of the aorta) appeared unchanged from the day before given the differing projection of the two films.

The report of this Xray was dictated at 02.24 — the time at which the formal report was available to [Dr E] is not clear and she did not note her own interpretation of the Chest Xray findings in her documentation. Given that by this time [Dr E] had discussed the case with [the radiologist] who had agreed to perform an abdominal CT, I think this is an oversight and [Dr E] and [the radiologist] had both likely seen the CXR abnormalities (which were also visible on CT and reported in [Dr E’s] documentation of the CT), and the lack of documentation is an oversight.

At 00.05 another litre of fluid was given to [Mrs A] over an hour, antibiotics were started and paracetamol given.

At 23.37 [Dr E] discussed [Mrs A’s] presentation with the surgeon on call who advised a CT abdomen. This was reported by [the radiologist] as showing thickening of the low end of the bowel indicating likely colitis (inflamed bowel, which can be due to many causes — including low blood flow to the bowel) and which was thought by [the radiologist] to be the likely cause of blood in bowel motions, plus free fluid in the abdominal cavity.

The CT scan also showed thickening of the gallbladder wall suspicious for Acute Cholecystitis (recent onset inflammation of the gallbladder). In addition changes in the liver and spleen were present that [the radiologist] reported as suggesting poor cardiac function and decreased venous return to the heart. [The radiologist] noted that the abdominal Aorta and its branches appeared normal.

The whole chest had not been scanned, but the bottom of the lungs and heart were visible on this abdominal scan. [The radiologist] noted pleural effusions (fluid around the lung) on both sides, and a large pericardial effusion (fluid around the heart).

The findings were discussed by [the radiologist] with [Dr E] at 01.40, and at 01.52 [Dr E] called [the surgeon] with the results of the CT scan.

A plan was made for [Mrs A] to be admitted to Intensive care unit (ICU) under the care of [the surgeon]. [Dr E] also discussed the admission with the ICU admitting anaesthetist who also agreed with the plan and advised that the ICU Physician on duty would review [Mrs A] the next day with respect to “work-up” (diagnosis of the cause) of the pericardial effusion.

[Dr E] documented her admitting impression as “GI (Gastrointestinal) bleed”, “? Acute Cholecystitis” and “Chest pain ? significance”, and that the collapse and seizure at home had likely been due to low blood pressure.

In line with the plan agreed with [the surgeon], [Dr E] prescribed IV fluids and antibiotics, pain relief and thromboembolic prevention stockings, gave instructions for urine and bowel motion monitoring. ECG was to be repeated if chest pain recurred, and repeat Haemoglobin and Troponin tested at 0200. A plan was made for blood transfusion and Tranexamic acid (a medication that aids clotting) if a significant haemoglobin or blood pressure drop occurred or if further significant bleeding in the bowel motions was seen.

The repeat blood tests at 0200 showed a small drop in haemoglobin to 128 (insignificant given the diluting fluid that had been given).

At 02.20 [Mrs A] passed further bowel motions with blood.

At 03.30 ICU nursing staff note the admission to ICU. I was not provided with the ICU vital signs chart. [Dr E] in her response notes that pain scores of 1/10 and 4/10 were recorded (the latter for neck pain but the neck pain was noted to be a long term pain for [Mrs A]).

Chest and Abdominal pain was noted on the ICU admission visual body diagram.

In the nursing notes [Mrs A] is noted to remain stable during the night, other than needing a small amount of oxygen to maintain low normal oxygen levels and being very tired. Her heart rate was stable in the 70s and blood pressure was noted to have now improved to 140/- .

A further 1000ml of fluid was given over 4 hours.

At 0600 [Mrs A’s] blood pressure was noted to be very high — 180/122. No note is made of whether this was notified to medical staff.

At approximately 06.55 [Mrs A’s] heart rate dropped and she collapsed. CPR was commenced and multiple doses of Adrenaline were given.

[Dr E] in her response documents that needle aspiration of the fluid around the heart was attempted on 3 occasions during resuscitation but I cannot see this documented in the notes. I believe there is a page of resuscitation notes that documents the resuscitation between 7.09 and 07.27 that is missing, however as it is clear that [Mrs A] would not have survived the resuscitation, I have not asked for these notes to be found.

The [on call physician] was present during the initial resuscitation, as was the medical house officer [Dr C], and [Dr E].

[The anaesthetist] was documented at the scene at 07.28, performing cardiac ultrasound. Resuscitation was continued for 31 minutes, however was not successful and [Mrs A] passed away at 07.30 on [Day 3].

My Opinion on the questions posed by the Commissioner

As a preface to my opinion:

Chest pain assessment in the Emergency department is a complex task due to the large number of potential causes of chest pain, and because any one particular cause for chest pain may present in many different ways.

Aortic Dissection can be an exceedingly difficult diagnosis to make. It can present in a vast number of ways and symptoms may overlap with the symptoms of many other diseases¹.

Even Emergency specialists, who are the clinicians most experienced in diagnosis of Aortic Dissection, often struggle to identify the condition. As correctly stated by [Dr C] in his response, rates of misdiagnosis in the Emergency department are reported to be as high as 38%².

As Aortic dissection is also a rare disorder³, many clinicians will never have personally diagnosed a case. This is especially true of doctors in the first few years of clinical practice. In my experience and that of my peers, a junior doctor would commonly not be able to identify the diagnosis of Aortic dissection without senior help.

The definitive test for Aortic dissection is CT Aortography (CTA). This cannot be used in every patient presenting to the Emergency department with chest pain however, as this would cause harm through contrast and radiation exposure, and also as discovery of other incidental findings can expose a patient to the harm of further diagnostic tests or treatments for a condition that may never have troubled them. Over-use of finite health resource also precludes this approach.

A high degree of suspicion for Aortic Dissection is needed, combined with a systematic clinical rule-out approach, using CTA when clinically indicated^1,2 .

Emergency Medicine Specialists are doctors who specialise in achieving a diagnosis from undifferentiated symptoms.

Bedside or point of care Echocardiography can greatly aid diagnosis of Aortic dissection⁴.

Echocardiography skills are usually acquired at a mid–late stage of training in some specialties, for example in Emergency, Intensive care, Rural Care or Cardiology specialties.

Were [Mrs A’s] presenting symptoms and clinical history appropriately assessed and treated on Day 1?

DIAGNOSTIC PROCESS

This was difficult to assess due to the brevity of contemporaneous documentation by [Dr C].

The standard of care in cases of chest pain, is for life-threatening causes to be considered and systematically ruled out. Some of these conditions are mentioned by [Dr C] in his response — Aortic Dissection, Myocardial infarction, and Pulmonary Embolism.

Aortic dissection

It is not apparent from the very brief clinical documentation on [Day 1] that [Dr C] did consider Aortic Dissection as he states in his response. For example there is no documentation of enquiry into the character of the pain, any migration of pain, associated neurologic, abdominal or limb symptoms, the past medical history of hypertension, family history, examination features such as bilateral pulses or blood pressures, or the appearance of the mediastinum (including the Aorta) on the Chest X-ray.

In addition, an episode of collapse and hypotension at the scene was not mentioned by [Dr C].

[Dr C] states in his response that he did consider Aortic Dissection. He states he discounted the diagnosis due to the lack of severe ripping or tearing pain radiating to the back, lack of neurologic symptoms, lack of widened mediastinum on Chest X-ray, documented low pain score of 2/10, the normal vital signs, lack of risk factors such as smoking or connective tissue disorders, and the rarity of Aortic dissection.

[Dr C] states due to “rapid relief of pain with ‘Pink Lady’, an impression of reflux seemed most likely”. Relying on a reduction of chest pain following oral antacid/local anaesthetic (ie using this as an indicator of reflux as the cause for the pain) in my opinion and that of my peers is a common pitfall in the assessment of chest pain.

I do note the documented low pain scores in the ambulance and at triage and other factors that would be less common in Aortic dissection such as pain on inspiration, improved on sitting forward.

From [Dr C’s] response, he is now aware of the use of the ADD-RS+ D dimer clinical decision tool. Use of this tool is common in 2022 but was not standard of care in 2018.

Given [Mrs A’s] concerning symptoms of sudden onset pain radiating to the throat and associated collapse, her age and past history of hypertension, and while I recognise the known difficulties with diagnosis of Aortic Dissection, in my opinion it was a severe departure from accepted practice to discount the diagnosis at this stage.

As noted in my preface above however, a training doctor would usually need senior supervision to recognise this — especially in a stable patient as [Mrs A] appeared to be at that time.

Adequacy of assessment for life threatening conditions other than Aortic Dissection

Myocardial Infarction

[Dr C] did take steps to rule out Myocardial infarction by reviewing ECGs and ordering two troponin blood tests.

[Dr C’s] clinical documentation at the time and in his response however, does not show he saw the abnormalities present on the ambulance ECG. In his response he describes the automated report of the less concerning abnormalities of “Sinus bradycardia, poor R wave progression and long QT — probable normal variant”, but not the “Inferior/lateral T wave abnormality may be due to myocardial ischaemia”.

Regarding the abnormal lateral T wave inversion [Dr C] saw on the ECG in the Emergency department, he did not document whether he took steps to check whether those changes were present on old ECGs. It is likely that the old ECGs would only have been available if the hard copy notes were called down to the Emergency department. A description of an ECG in the previous record of prior presentations can be helpful in this situation — the past notes are documented as describing lateral ST depression but not T inversion and it is unclear whether those changes were present during the exercise test or at rest.

[Dr C] also stated that there were no changes on the second ED ECG — however there is new T wave inversion in AVL and 1 (subtle).

It is not apparent whether [Dr C] used the EDACS tool which would be standard of care at the time. Even if the ECGs had been normal and with the negative Troponins, [Mrs A] would not be low risk using this tool. Given the previous advice to have a low threshold for cardiology review in the event of further chest pain, in my opinion a Cardiology opinion should have been sought.

Not reviewing the ambulance ECG, or not recognizing the abnormality on the ambulance ECG, or not taking steps to ensure that the abnormal findings of the Emergency department ECG had been present on old ECGs or not using the EDACS tool would all be a severe departure from accepted practice.

Not seeking Cardiology advice before discharge is in my opinion a severe departure from accepted practice.

Pulmonary embolism

[Dr C] in his response mentions discounting Pulmonary embolism due to the lack of risk factors, but not that he used a tool such as Well’s Criteria +/- D dimer to rule out the diagnosis. Nor is this mentioned in the documentation on the day of assessment. This is a relevant illness to exclude due to [Mrs A’s] pain on inspiration, shortness of breath (not documented at the time but described by [Dr C] in his response and [Mrs B’s] complaint) and collapse — all of which would not be usual with pain related to reflux.

Although not the cause of [Mrs A’s] illness, if these methods were not used, this would be a severe departure from accepted practice.

Chest X-Ray review

This was ordered by [Dr C] as [Mrs A] was presenting with acute chest pain, worse on inspiration, shortness of breath (as above — not documented at the time but described by [Dr C] in his response and by [Mrs B]) and had an abnormal chest examination. It is also relevant as assessing the Aortic contour on a Chest X-ray is part of the assessment for Aortic dissection (although not a sensitive test — and the Aortic contour was not abnormal that day.)

Neither [Dr C’s] interpretation of the Chest X-ray nor any knowledge of a formal Radiology report of the Chest X-ray are mentioned in his documentation that day.

The Chest X-ray did not show signs of Aortic dissection, but was not normal with abnormalities not explained by a diagnosis of reflux.

The Chest X-ray findings were either not appreciated by [Dr C], or were seen but not further pursued.

As above I am not sure if the formal report of the abnormalities was available to [Dr C].

Discharging [Mrs A] with the abnormal Chest X-Ray findings not explained is a severe departure from accepted practice.

DISCHARGE ADVICE

It is not clear what advice was given on discharge and in what form.

The discharge summary states that written advice was given to [Mrs A] but this was not supplied to me. The standard of care would be a discharge summary given to [Mrs A] at the time of discharge containing clear instructions regarding symptoms of concern that should prompt re-presentation to the Emergency department or GP follow-up. [Dr C] in his response states that [Mrs A] was told to return if her pain recurred or concerns arose, and that it is his routine practice to advise patients to return if symptoms worsen.

If such advice was not at least verbally provided, it is a severe departure from accepted practice. If provided verbally but not in written form this would be a moderate departure from accepted practice.

DOCUMENTATION

[Dr C’s] documentation was extremely brief and appears to be found only on the discharge summary — presumably completed at the time of discharge. There is a tick on the paper clinical record in a box saying “see electronic attendance summary” suggesting this was common practice in the department however this is accepted standard only in a quick assessment of a minor case. This is because documentation of the medical assessment must be available to nursing and other staff during the patient’s stay to enable information sharing and minimise error.

It must be noted workload pressures especially in a busy or understaffed Emergency department may result in brief and non-contemporaneous documentation. For example a doctor may be called urgently from a patient they have just seen to attend to a new urgent case requiring resuscitation, before they have had a chance to write notes on the first patient.

Whatever the reason on that day, I regard the brevity of documentation and the lack of contemporaneous documentation in a patient presenting with collapse and chest pain and in the ED for over 5 hours as a severe departure from accepted practice.

PROVISION OF SPECIALIST MEDICAL EXPERTISE IN THE EMERGENCY DEPARTMENT

[Dr C] in his response stated that he was a Medical officer in [the] ED and a Rural hospital medicine and General practice Registrar (training doctor). From my question to the DHB, [Dr C] was working in [the] ED as a training placement, to contribute to his qualification.

I am not clear whether [Dr C] had access to direct senior supervision in his workplace on the day in question. To my questioning on this point, the DHB stated that [Dr C] was supervised in his remote and rural qualification but not whether that supervisor was available to him on site on the day. My assumption is that they were a remote supervisor and not available in the workplace. The DHB also stated that [Dr C] “worked alongside” two senior medical officers that day. It is not clear to me whether [Dr C] was employed to work independently to those senior doctors or directly under their supervision. This is an important distinction, as especially in a busy Emergency department, unless a doctor knows that there is a direct line of supervision provided, concern of interrupting a busy colleague that one is “working alongside” is a significant barrier to seeking out an opinion from a more experienced diagnostician. In addition, due to the recognised stages of learning⁵ trainee doctors other than those in a very late stage of training should not be left to decide which of their cases they need help with, but instead should be expected to discuss all except the most minor of cases with an on-site supervisor.

It is difficult to successfully diagnose Aortic Dissection in the ED without the help of an Emergency Specialist.

Requirements for Emergency Departments in smaller centres in NZ to provide training doctors with senior supervision by an Emergency Specialist is out of my scope of practice so I will not comment as to whether any lack of supervision for [Dr C] that day is a departure from accepted practice.

Was the care provided to [Mrs A] during her admission to ED on [Day 1] appropriate and consistent with the accepted standards of practice at the time?

Triage and Medical/Nursing assessment were completed in a timely manner.

MONITORING

Vital signs were not measured from 14.50 until discharge at 1700. [Mrs B] reports that monitoring was removed after [Mrs A] went to the toilet. Given that the second Troponin (ruling out myocardial infarction) was still awaited until 16.34, in my opinion this a moderate departure from accepted practice.

PAIN RELIEF

After the initial IV Diclofenac at 12.10, additional paracetamol given at 12.50 and the pink lady at 13.50 — which all suggest [Mrs A] was experiencing ongoing pain. The pain score is not filled out during the admission however. I note the low (2/10) pain score at triage and [Dr C’s] documentation that pain had resolved at discharge (after the pink lady).

Although not ideal, pain scores are commonly omitted from documentation as other factors may be used to determine administration of pain relief eg a question such as “would you like some more pain relief?”. For this reason the lack of documentation of pain score is only a mild departure from accepted practice.

[Mrs B] states that despite the documentation above, [Mrs A] was in severe pain during this admission and was refused stronger pain relief due to concerns it might drop blood pressure. Actually there is almost never a scenario in which severe pain cannot be immediately treated even when blood pressure is very low.

If severe pain was present and not treated, this would be a severe departure from accepted practice.

Recommendations for improvement that may help to prevent a similar occurrence in future:

1. [The] Emergency Department be staffed to ensure Emergency specialist supervision of training doctors. This would provide clinical diagnostic expertise including bedside ultrasound skills. A bedside Echo of the Aortic root and heart may have helped make the diagnosis.

2. [The] Emergency Department review with the [medical school] whether they are providing suitable supervision to Rural Hospital medicine trainees on their clinical placements.

3. Routine daily use of a structured risk assessment approach to Aortic Dissection be introduced to [the] Emergency Department. An example of such an approach is use of a tool such as ADD-RS + d-Dimer.

In my department our Aortic Dissection pathway involves pre-test probability assessment (a modified ADD-RS tool), combined with requirement for ED SMO review of the patient if low risk (ie not going straight to CTA), CXR, Bedside Echocardiography, and use of D dimer and CTA as needed. This has been described by one of my colleagues as a “game changer”.

It is my personal experience that our Emergency Specialists routinely need to remind less experienced staff (who have usually never seen a case of Aortic Dissection) to use the pathway.

If training doctors are required to practise in [the] ED without Emergency specialist supervision, the use of whatever risk assessment tool is chosen would need to be continually promoted by another means. It could be included in orientation material for doctors (along with other scoring systems such as EDACS for acute coronary syndrome and Wells/PERC/YEARS for Pulmonary embolism), but in the absence of senior supervision, routine consideration of the diagnosis/use of the clinical risk assessment tool would need to be promoted by other means. An example would be a check box on documentation.

I cannot stress highly enough that the diagnosis of Aortic Dissection needs to be considered in every patient with chest pain, and this needs frequent if not daily promotion amongst all medical and nursing staff, not just formal intermittent education sessions (although these help).

In addition orientation to the Emergency department for doctors could include a specific session on Emergency department chest pain assessment with particular focus on strategies for ruling out life threats.

4. Routine request of past hard copy medical records in all patients with chest pain should be made to enable comparison of ECGs (in my institution this is done for all patients given a triage code 2).

5. [The] Emergency and Radiology departments review reporting of radiology from the Emergency department.

Ideally reporting Emergency images should be prioritised, and reports issued as soon as possible with consideration to the feasibility of providing clinicians with a report while the patient is still in the department. The reports should go straight onto the digital information system, allowing Emergency doctors to access them. The timely “sign off” of reports by Emergency clinicians should also be prioritised and documented by the ED.

The reason for my advice here is that if the Chest Xray formal report had been available to [Dr C], even if he had not recognised the abnormalities himself, having seen the report he may not have discharged [Mrs A] but instead consulted with a Medical physician or a Cardiologist, providing another opportunity for correct diagnosis. Alternatively if the Xray was not reported till the next day but when issued was immediately seen for sign-off by an Emergency clinician, [Mrs A] may have been called back to hospital before her condition deteriorated.

6. Digital clinical information system be used in the ED (not just for discharge documentation). These commonly used systems can be accessed anytime by treating staff from anywhere in the department, allowing contemporaneous documentation and immediate information transfer between staff. It saves clinician time (eg cut and pasting instead of writing notes manually and then typing the same information onto the discharge summary). It can also enable documentation of vital signs and a supervising doctor’s input into the case.

7. The discharge summary template be pre-populated with mandatory fields such as “Specific advice given to patient” and/or “Reasons to seek medical attention after discharge”. If digital information systems communicate, discharge summaries can be auto-populated with Pathology and Radiology reports.

8. Staff education around use of pain scoring on the vital signs chart, pain relief options in case of low blood pressure concerns, and removal of monitoring and replacement after patient movements.

9. [Dr C] be provided with my opinion. Depending on his area of current practice, [Dr C] could consider

— Reviewing his approach to diagnosis of chest pain — particularly with regard to systematic rule-out of life threats before entertaining a reflux diagnosis.

— Ensuring routine use of validated decision tools (ADD-RS, EDACS, Wells/PERC/Years)

— Review of ECG and CXR interpretation skills

— Acquisition of Bedside Ultrasound/Echo skills

— Strategies to enable adequate contemporaneous documentation especially when under time constraint.

I note that [Dr C] in the intervening 4 years may have done all of the above.

It is clear from his response that he had changed his medical practice already at that time.

Do I consider the care provided to [Mrs A] following her readmission to ED on [Day 2], was appropriate and consistent with the accepted standards of practice at the time?

In my opinion both the initial assessment and documentation of this by [Dr E] in the Emergency department on [Day 2]–[Day 3] was thorough and of appropriate standard for her level of experience.

[Mrs A] was by now clearly very unwell on this second presentation. The clinical picture had progressed. There was by now an extremely complex and unusual constellation of symptoms, signs and abnormal investigation results — confounding the assessment and making the correct diagnosis more obscure rather than more obvious — especially given the normal appearance of the abdominal Aorta on CT scan.

With this clinical picture, I am in absolutely no doubt that a PGY 3 doctor would not be able to achieve the correct underlying diagnosis without guidance from a more experienced doctor.

I have assumed that [Dr E] at that time did not have an Emergency Specialist supervising her in the ED. She instead discussed the case with the surgeon on call, which given the now very prominent symptom of lower GI bleeding/vomiting and some degree of abdominal tenderness on examination was appropriate.

The plan for ICU admission was made, with subsequent consultation with [the anaesthetist] — described as the anaesthetist covering ICU admissions.

It appears [Mrs A’s] symptoms at this time, including her collapse, were thought by this senior admitting team of [the surgeon] and [the anaesthetist], to be due to gastrointestinal blood loss due to colitis, +/- infection from the inflamed gallbladder. The chest pain was apparently still not explained (although it should be noted that an inflamed gallbladder can at times cause chest pain), nor the pericardial and pleural effusions.

There is no contemporaneous documentation from either [Dr F] or [the anaesthetist] to enable insight into their decision making process — for example whether these doctors were aware of [Mrs A’s] presentation the day before and the extent of the chest pain she had experienced over the previous 36 hr.

No response to the commissioner was provided by the DHB from these two senior admitting doctors nor from [the radiologist].

In addition it is not clear to me whether [the surgeon] and [the anaesthetist] +/- [the radiologist] spoke together in person regarding [Mrs A’s] case, or all just spoke with [Dr E], which seems more likely from [Dr E’s] documentation.

It is evident from the documentation supplied, the planned overnight management pathway, and the verbal information relayed to family as reported by [Mrs B], that the unifying diagnosis of Aortic dissection was not considered at any time. As above I do not think this is unusual given the now extremely confusing clinical picture, especially with a normal looking descending aorta on CT. I note that a general surgeon and an anaesthetist would be unlikely to have experience diagnosing Aortic dissection.

The following is the perspective of an Emergency Physician:

I am critical that given the findings on CT of both pericardial effusion and signs of decreased venous return to the heart, along with the history of chest pain and collapse and [Mrs A’s] signs of severe illness, that none of the three senior doctors (Surgeon/Anaesthetist/Radiologist) advised an urgent bedside Echocardiogram to rule out the immediate life threat of Pericardial tamponade (which can be due to Aortic dissection — or due to another cause).

I am presuming Echo was within the scope of practice of [the anaesthetist], as it is documented he was performing it during the attempted resuscitation of [Mrs A]. It may be that the appearance of the pericardial effusion on CT was reassuring with regard to the possibility of tamponade, but it is out of my scope of practice to comment on this.

I regard this as a severe departure from expected standard.

Recommendations for improvement that may help to prevent a similar occurrence in future.

1.  As above, Emergency Specialist supervision be provided for junior doctors in [the] ED — providing diagnostic expertise and ultrasound skills.

2.  Review in person by a more senior doctor than a PGY 3 would seem reasonable for patients sick enough to require ICU admission, even if they appear stable at the time.

In addition SMO to SMO discussion between admitting specialty SMO (in this case surgical) and admitting ICU SMO may help.

Other matters in this care that I consider warrant comment

I am critical that the DHB did not provide a response to the commissioner from the 3 senior doctors involved with [Mrs A’s] care.

A Radiology opinion on the imaging would be relevant — particularly with regard to whether any signs of dissection on the CT were missed and whether it was reasonable that the appearance of pericardial effusion and impaired venous return to the heart did not raise suspicion of cardiac tamponade.

References

1.  The International Registry of Acute Aortic Dissection (IRAD): new insights into an old disease JAMA. 2000 Feb 16;283(7):897–903.

2.  Diagnosing acute aortic syndrome: a Canadian clinical practice guideline

CMAJ July 20, 2020 192 (29) E832–E843; DOI: https://doi.org/10.1503/cmaj.200021

3.  Ethnic differences in incidence and outcomes of acute aortic syndromes in the Midland region of New Zealand. J Vasc Surg. 2022 Feb;75(2):455–463.e2. doi: 10.1016/j.jvs.2021.08.066. Epub 2021 Sep 24.

4.  Emergency physician focused cardiac ultrasound improves diagnosis of ascending aortic dissection. Am J Emerg Med. 2016 Mar;34(3):486–92.

doi: 10.1016/j.ajem.2015.12.005. Epub 2015 Dec 12  

5.  https://clinical-supervision-toolkit.qut.edu.au/content/Clinical.html’