Medical practitioners have a responsibility to ensure the timely and direct communication of abnormal medical test results.
The communication of abnormal test results in a clear and timely manner, or at all, is an important element of multidisciplinary care. The inquest into the death of Mr Mettaloka Malinda Halwala and ﬁndings of Coroner Carlin in the Coroners Court of Victoria remind us that erroneous assumptions on the part of medical practitioners about whose responsibility it may be to follow up, and delays in the timely and direct communication of test results, can have serious consequences for patients. The case also illustrates what can happen when the responsibility is shared but no one party is prepared to accept it or to make concessions. As none of the partitioners or health facilities acknowledged responsibility for the communication failure, this inquest proceeded to a hearing and considered the responsibility for a failure to communicate abnormal positron emission tomography (PET) results in a timely and direct way.
Mr Halwala (the patient) was 58 years of age when he died alone in his room at the Tatura Hotel in the Goulburn Valley region of Victoria from complications of chemotherapy for Hodgkin’s lymphoma. The patient had, shortly before his death, received a dose of chemotherapy when abnormal test results indicated that he was experiencing signs of toxicity.
The patient had checked into the hotel on 1 September 2015. He was living alone in the hotel while working as an engineer in Shepparton, his family being in New Zealand. On 7 September, he presented to Goulburn Valley Hospital (GVH) complaining of a sore throat and other symptoms. He underwent tests that were suggestive of Hodgkin’s disease, and had his care transferred to Austin Hospital on 14 September for further investigation.
involving multiple lymph nodes, the spleen and the marrow. The patient was commenced on ABVD chemotherapy on 18 September at the Austin, tolerated this treatment and was discharged on 19 September. He returned home to the Tatura Hotel. After being contacted by a colleague at the Austin, haematologist Dr Robin Filshie agreed to supervise treatment at GVH as he attended there once every 4 weeks. Dr Filshie conducted his practice as a consultant from St Vincent’s Hospital but had been treating the Shepparton-based patient because of his involvement in monthly rural outreach services St Vincent’s provided to GVH.
After his second and third ABVD treatments, the patient saw Dr Filshie for the ﬁrst time on 23 October 2015. Although the patient was undergoing his chemotherapy under the management of Dr Filshie, the Austin had closer involvement with the patient’s care, and Dr Filshie was neither based at the Austin nor at GVH. According to Dr Filshie, the patient “had advanced high-risk disease” carrying a lower chance of cure.
On 27 October 2015 Dr Filshie referred the patient for a further PET scan at the Austin. Dr Filshie completed anAustin Health Oncology Referral for PET Scan form and faxed it to the Austin. He wrote his own contact details at St Vincent’s in the Haematology Department — Level 6, in the “Referring Specialist” box provided on the form, including the phone number to his office and the fax number corresponding to the fax machine closest to his office. He provided instructions to book the PET scan for 10 or 11 November, with the patient’s next treatment to occur on 30 October 2015. Dr Filshie also circled the word “curative” as the management plan intent.
The PET scan ordered by Dr Filshie on 27 October 2015 was booked and performed at Austin on 11 November 2015. The technologist, Mr Farrell, asked the patient a series of questions prior to the scan, as directed by a pro forma questionnaire and the patient did not report any respiratory symptoms, nor did he cough during the procedure. After the scan, a PET scan report was prepared. The report was commenced by a visiting nuclear medicine physician trainee, Dr Leung, and validated by nuclear medicine physician Associate Professor Sze Ting Lee. Associate Professor Lee placed the report in the out tray within the administration area, for post and fax the next day. The address and fax details on the form completed contained a different fax number to that which Dr Filshie had provided, and referred to the “St Vincents Hospital Haematology Clinic 35, Victoria Pde” which was also a different address to that provided by Dr Filshie in his referral form.
The PET scan report revealed “interval development of widespread FDG-avid uptake throughout both lung ﬁelds” and “may be due to bleomycin related pneumonitis. An opportunistic infection is considered less likely if the patient is not clinically septic”. The ﬁndings were “consistent with an excellent complete metabolic response to treatment” and “consistent with bleomycin related toxicity or opportunistic infection if the patient is septic/in the right clinical scenario”. It revealed evidence of lung toxicity.
The results of the PET were not communicated to Dr Filshie or GVH either by Associate Professor Lee or anyone else at the Austin. The patient was administered a further chemotherapy dose 2 days later on 13 November 2015, that being his ﬁfth treatment and the beginning of cycle 3 at GVH. An intern at GVH prescribed pholcodine for a presumed viral chest infection as the patient complained of dry cough of 5 days’ duration and an itchy or sore throat. Routine blood tests showed mild anaemia and neutropenia consistent with the disease and its treatment. The chemotherapy, including bleomycin, was administered as planned.
On 14 November, the patient saw his family in Melbourne and he was described as very weak and severely sick with a heavy chest infection. On 16 November 2015, the patient called Dr Filshie’s rooms and reported feeling unwell but did not mention respiratory symptoms. Upon receipt of the message, Dr Filshie instructed his secretary to call the patient and recommend that he attend hospital for assessment if he felt unwell. That was standard telephone advice to patients. At the time, Dr Filshie did not think to seek the results of the PET scan. At the end of that day he opened his mail and read the PET report for the ﬁrst time. That was several hours after he had instructed his secretary to advise the patient to go to hospital and he assumed the advice would be followed. Dr Filshie planned to contact GVH staff the next day but was called by police advising that the patient did not make it to hospital and was found deceased in his hotel room.
attend hospital for assessment.Adopting the approach he had, Dr Filshie assumed that GVH would call him if the patient did attend, and had no reason to suppose that the patient’s call to his rooms was a complaint of the respiratory symptoms referred to in the PET report as there had been no mention of respiratory distress. Dr Filshie, in evidence, agreed that his approach meant that the hospital would not have had the beneﬁt of the PET report in the event that the patient did attend as assumed. He did however note that he believed that the hospital would call him and that there was no indication that the symptoms experienced by the patient were related to the PET scan.
Communication of the PET report
At the time of the PET scan, Associate Professor Lee noted that the patient showed no sign of respiratory distress. Pulmonary opacities on a PET scan for a patient not clinically unwell would not generally require urgent reporting. Reporting within 24 hours was sufficient and that was what occurred here.
Associate Professor Lee also did not consider the ﬁnding of lung toxicity as clinically signiﬁcant because it is a known complication of ABVD chemotherapy, and the patient was not unwell when he underwent the PET scan. She gave evidence that a PET scan with ﬁndings of lung toxicity were fairly common in patients undergoing ABVD chemotherapy, and as such believed that the ﬁndings were not clinically signiﬁcant and would not generally require urgent reporting.
Nevertheless, Associate Professor Lee noted that she expected the treating doctor would see her report before giving the patient more chemotherapy. If she had known that the patient was to undergo more chemotherapy two days after the PET scan, she conceded that she “might have actually called Dr Filshie”. If the patient had been at the Austin, she would have checked electronically the date of next chemotherapy or medical practice consistent with her usual practice.
Associate Professor Lee’s evidence was that a radiological ﬁnding on a PET scan does not equal a clinical diagnosis. However, regardless of which of the two differential diagnosis was correct, Dr Filshie did agree that both required “fairly urgent” clinical assessment, as in within 24 hours, because the patient could deteriorate rapidly and one could only determine the severity of the situation by clinical assessment. Dr Filshie’s own practice was to call the referring doctor in the case of abnormal test results, and understood that “unexpected results will be notiﬁed in a different way to expected and routine results”. As safety to proceed with treatment is determined by clinical assessment on the day, he was not expecting an adverse reaction and believed he would be informed had a signiﬁcant abnormality been detected. Dr Filshie rejected the notion that he ought to have reviewed the PET scan report before the next dose of ABVD therapy. He believed the nuclear medicine physicians understood from his PET scan referral that ABVD therapy was continuing and was always administered 14 days apart.
The usual practice of the haematologist
Dr Fishlie’s evidence was that a PET scan is usually booked after the fourth treatment as a predetermined investigation. That was the case here, and it was not booked to investigate anticipated abnormal ﬁndings. In most cases it is not anticipated that the PET scan ﬁndings will be available before the next treatment, unless there is an unexpected ﬁnding. He intended to discuss the results with the patient at the next appointment which had been scheduled on 20 November 2015.
Whilst he would have liked to have received the PET scan report prior to the next scheduled treatment on 13 November, and if he had received it would have read it and without doubt withheld treatment; Dr Filshie did not assume he would receive it before 13 November and it was really a prognostic tool. He disagreed with Associate Professor Lee’s evidence about the frequency of lung toxicity and had never seen a report like that of the patient in 19 years’ practice. Mild toxicity, whilst more common, is rare after the ﬁrst two cycles.
Agreeing withAssociate Professor Lee that radiological ﬁndings do not equal a clinical diagnosis, it is important for the treating doctor to be alerted to abnormal or unexpected ﬁndings for fairly urgent clinical assessment within 24 hours of the results.
Concurrent experts cannot agree
Four expert witnesses gave evidence. The consultant radiologist Dr Christopher O’Donnell and haematologist Professor John Seymour were court-appointed expert witnesses. The nuclear medicine physicians, Dr William McKay and Professor Andrew Scott, gave evidence on behalf of Austin Health. They gave evidence concurrently.The radiologists and the haematologists could not agree.
Guidelines for the communication of imaging results
The Royal Australian and New Zealand College of Radiology “Standards of Practice for Diagnostic and Interventional Radiology” expect the provision of nuclear medicine reports to referring practitioners in accordance with the Australian Association of Nuclear Medicine Specialists (AANMS) Standards. The College standard for nuclear medicine dealing speciﬁcally with the timeliness of reporting, expects the provision of nuclear
The timeliness of reporting will vary with the nature and urgency of the clinical problem. In general, the report should be sent to the referring practitioner within 24 hours of completion of the study. If there are urgent or unexpected ﬁndings,the specialist should use reasonable endeavours to communicate directly to the referrer or an appropriate representative who will be providing clinical follow-up [emphasis in original].
The UK has detailed professional guidelines on this issue. At the relevant time, the 2012 Royal College of Radiologists (RCR) “Standards for the Communication of Critical, Urgent and Unexpected Radiological Findings” were in place in the UK. Those were replaced in 2016 by “Standards for the Communication of Radiological Reports and Fail-Safe Alert Notiﬁcation”. The 2012 RCR Standards required special attention for critical ﬁndings, urgent ﬁndings (where medical evaluation is required within 24 hours) and signiﬁcant unexpected ﬁndings. Of interest, the RCR Standards stress that the radiologist, referring doctor and the health care institution all share responsibility for timely communication of radiological ﬁndings.
According to the 2012 RCR Standards, the radiologist should contact the referring clinician if they consider a danger of unexpected relevant information in the report not being acted upon. In addition, the referring clinician’s responsibilities include having a clear policy as to how to access imaging results and an ability to read and act on results as quickly and as efficiently as possible. Similarly, the health care institution is noted to be responsible to provide “appropriate systems”. In 2016, these RCR Standards were expanded, with an expectation that a fail-safe alert be added to the normal communication method.
At the time of the patient’s death, the hospital guidelines at the Austin were limited to the communication of critical test results representing in themselves a clear and immediate threat to the patient’s life or limb, requiring urgent clinical intervention. The policy required such results to be communicated promptly and verbally by the diagnostic service “to a responsible doctor” and to set out procedures for doing so both during and after hours. These guidelines also conﬁrmed that the requesting doctor was responsible for checking and “actioning” test results. Of interest, the Coroner reviewed equivalent guidelines in place at another Melbourne health service and noted that they also were limited to the communication of critical test results, and largely mirrored those in place at the Austin.
So what is direct and timely communication?
The expert conclave accepted that the transmission of the PET report by fax and post was either reasonable or, at least, not unreasonable. This view was largely based on the fact that transmission by fax, followed by post, was a system that had long been successfully used at the Austin. They referred to an audit that noted that 90% of faxes at the Austin go out within 24 hours and 100% if marked as urgent, and that these results are seen by the referring clinician within that time frame. In retrospect however, Dr McKay’s evidence was that all would have made a phone call. The haematologists and the nuclear medicine physicians had different views about what was reasonable.
Dr O’Donnell and Professor Seymour considered the report warranted direct communication by telephone or some method to conﬁrm it had been received and understood and would be acted upon. Fax was not sufficient. From a treating haematologists perspective, the results were unexpected. From a nuclear medicine physician’s perspective, the results were not unexpected.
The ﬁndings The Coroner found a signiﬁcant disconnection between the expectations of the nuclear medicine physician who reported on the scan, and the haematologist who ordered it, in relation to the communication of its results.
Both doctors considered their actions entirely reasonable and relied to a great extent on their expectation as to what the other doctor would do, expectations that proved wrong in each case.
The Coroner noted that while nuclear medicine physicians might never meet the patients, “they are ﬁrst and foremost medical practitioners” with a positive duty and responsibility to communicate results in a manner that is effective and appropriate in the circumstances. The RCR Standards could not be dismissed as irrelevant even if they did not speciﬁcally apply to Associate Professor Lee as an Australian medical practitioner.
Although counsel for Austin Health and Professor Scott objected to references to the RCR Standards that “signiﬁcant unexpected ﬁndings” are a category of ﬁndings requiring special attention, the standards are a useful guide to levels of conduct.
ing means using a method to ensure that the report would be received, understood and acted upon. This could only be done by calling Dr Filshie and speaking to him. There was no evidence whether the faxed copy of the PET report ever reached Dr Filshie, as he received and read the copy sent to him in the post. Austin Health was not able to produce fax transmission reports to conﬁrm transmission in any event.
Dr Filshie’s assumption that Associate Professor Lee would inform him of unexpected results and his failure to follow up for the PET scan results was also a dangerous and ﬂawed approach even if the expectation was reasonable. Professor Seymour’s view was that although scheduled PET scans and investigations at particular intervals were reasonable, the preset chemotherapy timetable should not be subject to a “set and forget approach”, and clinical review and reevaluation could occur after each cycle.
The Coroner was of the view that there was absolute responsibility on both Dr Filshie and Associate Professor Lee. Furthermore, once he read the posted copy of the PET scan report, Dr Filshie should have made further efforts to ascertain if the patient had actually gone to the hospital. His lack of a response fell short of reasonable care, especially as he knew the patient lived alone in an hotel room.
The Coroner invited from the expert conclave ways of improving systems for communicating diagnostic results, and they agreed that electronic distribution with conﬁrmation of receipt should be routine, consistent with both the 2012 and 2016 RCR Standards. The Coroner also invited the expert conclave to suggest a formulation of words to cover types of results requiring direct communication but which the conclave recommended for referral to the relevant specialist colleges for further consideration, and noted and endorsed the introduction of fail-safe mechanisms in the 2016 RCR Standards and considered as a possible fail-safe the routine distribution of diagnostic results to patients and their referring doctors.
As the Coroner observed:
The schism in expectations in this case was no doubt partly attributable to the factAssociate Professor Lee and Dr Filshie worked at different hospitals and did not know each other. However, the fact that an expert conclave of medical professionals from each side of the profession were not able to agree as to what constituted a reasonable means of communication suggests that the problem is more widespread.
In the Australian context, the UK RCR Standards give a useful guide to what might be an appropriate standards framework for cases such as this. The fact that these results are being transmitted by ordinary post and by fax, rather than by some other digital means in circumstances where smartphones and other mobile devices are readily in use, contradicts the expectation of a transmission of results within 24 hours. This remains surprising when the PET report is generated in digital form and converted back into a hard copy for fax transmission. Like other decisions where time-critical information was conveyed by antiquated means, password protected email transmission was recommended.
It is worth noting that the patient, Mr Halwala, died less than 4 months after the ﬁndings of the Coroners Court of South Australia in the inquest of Marjorie Aston, where communications between the specialist who commenced warfarinisation and the general medical practitioner responsible for monitoring it was sent by ordinary post to an adjacent building (separated only by a carpark) only to reach the general practitioner after the patient had died from a subdural haematoma with contributing excessive warfarin anticoagulation. In that case, the recommendations included that:
… the practice of communicating with general practitioners by way of ordinary post should be curtailed and be replaced by a means of communication that would include email and/or facsimile transmission. Any such communication should contain a request that the general practitioner, by return, acknowledge the communication. It may be necessary in some cases for the specialist to communicate with the general practitioner by phone[.]
1. Coroners Court (Vic) Inquest into the Death of: Mettaloka Malinda Halwala (10 May 2018) www.coronerscourt.vic.gov.au/ resources/0a340eaa-9390-4fc4-a3c8-6caf7c3e003f/ mettalokamalindahalwala_585715.pdf.
2. ABVD is an acronym for the drugs adriamycin, bleomycin, vinblastine and dacarbazine.
3. Above n 1, para 19.
4. Above n 1, para 31.
5. Above n 1, paras 32–35.
6. Above n 1, paras 37–38.
7. Above n 1, para 39.
8. Above n 1, para 40.
9. Above n 1, paras 41–43.
10. Above n 1, para 44.
11. Above n 1, para 49.
12. Above n 1, paras 50–1.
13. Above n 1, para 52.
14. Above n 1, para 53.
15. Above n 1, para 58.
16. Above n 1, para 59.
17. Above n 1, para 60.
18. Above n 1, para 61.
19. Above n 1, para 55.
20. Above n 1, para 57.
21. Above n 1, paras 21–24.
22. Above n 1, para 63.
23. Above n 1, para 64.
24. Above n 1 paras 65–66.
25. Above n 1, para 67.
26. Above n 1, para 68.
27. Above n 1, para 69.
28. Above n 1, para 70.
29. Above n 1, para 72.
30. Above n 1, paras 74–75 and 113–14.
31. Above n 1, para 73.
32. Above n 1, para 74.
33. Above n 1, para 113.
34. Above n 1, para 90.
35. Above n 1, para 91.
36. Above n 1, para 97.
37. Above n 1, paras 81–82.
38. Above n 1, paras 83–84.
39. Above n 1, para 119.
40. Above n 1, para 124.
41. Above n 1, para 125.
42. Above n 1, para 114.
43. Coroners Court (SA) Finding of Inquest: Marjorie Irene Aston (17 July 2015) www.courts.sa.gov.au/CoronersFindings/Lists/ Coroners%20Findings/Attachments/620/ASTON%20 Marjorie%20Irene.pdf.
44. Above, para 8.3.