Letters

In reply: We thank Mallows for his constructive comments, and would like to reply as follows. 1. Intention-to-treat analysis is normally associated with randomised controlled trials (RCTs). Our study was not an RCT, as patients who refused vertebroplasty formed the comparison group.1 We used intention-to-treat analysis to indicate that all patients who were assessed as eligible for our study at the outset were enrolled and were all included in the results presented, irrespective of subsequent events (ie, whether they were completely compliant, had died or were lost to follow-up). We thought it was important to make clear that the study was prospective, that all patients entered into the study were represented in the outcomes, and that patients had not been selected on the basis of later events or results. 2. A description of the analgesic requirements of a subgroup of the patients in our study has been published.2 More than 90% of the vertebroplasty-treated cohort were either able to cease or reduce their analgesia dose by at least 50% within 24 hours after the procedure. Before the procedure, an intravenous injection of pethidine (50–100 mg) was the only parenteral analgesia given routinely and would not have affected the pain scores at 24 hours. 3. Pain scores at 6 weeks were statistically lower in the vertebroplasty-treated group compared with the control group. The clinical significance was not apparent, as measured by the changes in the Barthel indices. This indicator is only a crude assessment of patients’ wellbeing, and a more detailed questionnaire of activities of daily living may have shown a difference. Pain scores recorded at 24, 48, 72 and 96 hours after the procedure would have been more sensitive, so that the additional pain, without vertebroplasty, could have been calculated from the area under the pain curve; the more data points, the more accurate the curve. 4. We elected to report the results as means ± 1 SD instead of 95% CIs. We used SDs to be consistent with our previous publications and other reports in this field. We agree that we could have used 95% CIs. 5. Patients were referred to Sydney Imaging Group from the inpatients and outpatients departments of St George public and private hospitals. All patients were treated without any out-of-pocket patient expenses. Private hospital inpatients were billed according to standard Medicare fees and outpatients were bulk-billed, which was revenue neutral for the radiology practice.

Terrence H Diamond · Carl Bryant · Lois Browne · William A Clark

To the Editor: We now know the outcome of the parliamentary vote on mifepristone (RU-486), which restored responsibility for its use to the Therapeutic Goods Association (TGA). Politicians from both houses used their conscience votes to support the scientific scrutiny of medical abortion. van Gend will now be worried about Australian women “demanding” abortions.1 However, abortion on demand in Australia does not exist. I refer van Gend to state laws which specify under what circumstances termination of pregnancy can take place. In no state can women “demand” an abortion whenever, wherever or however they wish. Regulations exist in all states and territories and, as a family doctor, van Gend must be aware of the multiple requirements. While there remains a lack of clarity about various state laws,2 the current position in Australia is that termination of pregnancy is available under certain conditions and in particular cases. The attempt by the Minister for Health Tony Abbott to influence women’s decisions about abortion by providing Medicare-funded, church-affiliated counselling for pregnant women3 has only further entrenched the view that the Minister is not able to speak for the majority of Australians. The previous situation whereby any Minister for Health, rather than the TGA, had the power to decide on the safety and efficacy of new medications before their entry into the pharmaceutical market place was ludicrous. Despite the endorsement of science over theology in health, and potential access to medical abortion, we still have the freedom of our own conscience. No one can force medical practitioners to prescribe mifepristone and no one can force women to accept medical (or surgical) abortions. Morals in Australia are a private matter and these decisions should be left to individuals and their families. Who would decide the authenticity of the medical grounds for abortion, mentioned by van Gend — doctors or priests, or academic ethicists, or feminists? van Gend is clearly not in favour of women deciding. I agree that more attention should be paid to the reasons for women stating they do not want to continue with a pregnancy, and, yes, we could do more to assist them. But I do not agree that excluding non-medical reasons is the answer — which, as van Gend points out, are financial hardship, relationship problems, single motherhood, and a completed family. To many, these appear convincing reasons to choose abortion. While this may not sit comfortably with van Gend’s medical paradigm, the “non-medical” reasons include the mental health of the woman (see the Menhennitt ruling which stipulates that an abortion is lawful if a doctor believes that the abortion is necessary to preserve her physical or mental health).4

Suzanne Belton

In reply: Belton is correct that “safety and efficacy of medications” is a matter for the Therapeutic Goods Association. The dispute was whether such limited criteria can meaningfully assess a drug designed to take life. The government needed to consider higher criteria for RU-486 — its ethical and medical justifiability. Doctors needed to advise the government on justifiable indications for RU-486, in contrast to the corrupt practice of abortion for non-medical reasons. That advice was withheld. The Australian Medical Association advised only on the ethically neutral question of “. . . who is best qualified to scientifically assess the safety and efficacy of a drug”.1 Such marginalisation of ethical concerns is consistent with the AMA’s earlier response in the context of late-term abortion: “There is no place for third parties — governments, over-zealous politicians and lawyers, hospital committees, or even the spectre of legal action”.2 This assertion of unchallengeable medical power over an unborn life is wrong. Belton’s notion that the morality of abortion is “a private matter” is wrong; neither parents nor doctors are above the moral and legal prohibition on intentional killing. “The law in this state has not abdicated its responsibility as guardian of the silent innocence of the unborn”,3 even if medical leaders have.

David van Gend

To the Editor: The report by Brooks and Ellis on the Health Workforce Innovation Conference held in November 2005 was, in my opinion, very misleading.1 I was one of about five doctors who attended this conference; the other 200 attendees were non-medical health care workers. It was fortuitous I attended — no invitation was extended to the Australian Society of Anaesthetists. Not only was the audience nearly exclusively composed of people dedicated to the introduction of non-doctors to replace doctors, but the presentations themselves included nobody expressing a contrary view. Such an unbalanced 2-day meeting therefore failed to truly examine medical task substitution. It failed to explore whether there truly is an inadequate number of doctors in Australia and whether the introduction of non-doctors to do medical work would actually save any money at all. In fact, the presentation by Sibbald indicated that nurse practitioners in the United Kingdom are no more cost effective: although they cost half as much, they take twice as long, so the overall cost is the same. The issue of the quality of Australian health care and how that quality would be affected by the introduction of non-doctors to do the doctoring was not addressed at all. There was also no examination of the consequences on the workforce of the rather illogical proposal to greatly expand the nurses’ scope of practice into medical work when there are already too few nurses in Australia. Such a proposal would surely only worsen the nursing workforce problem. Overall, the meeting was very disappointing, as it failed to approach the topic of medical task substitution in a balanced fashion, failed to justify why the proposal should be contemplated to begin with, failed to address whether there would be any improvement in health care delivery, failed to address whether there would be any reduction in health care costs, and absolutely failed to address how the proposed medical task substitution would not lead to an inevitable reduction in the quality of health care in Australia.

Gregory J Deacon

In reply: Deacon has unfortunately missed the whole point of the Health Workforce Innovation Conference. It was about innovation — new ways of doing things. It was also about health, not medicine — a concept that some doctors might find difficult to accept, but is actually the reality. Deacon seems to suggest that we still need to debate whether there is an “inadequate number of doctors in Australia” — surely most of us have moved on from there. One of the consequences of the “rather illogical proposal to greatly expand the nurses’ scope of practice” might be to retain nurses in the workforce. This, as most of us know, is a major issue. The Health Workforce Innovation Conference was not necessarily about saving costs; it was about producing a more effective health system and trying to provide for those who cannot access services because they are not available or there are long waiting lists for procedures that could well be done by other health professionals. A number of papers presented at the conference demonstrated that care could be provided by groups such as nurse practitioners or physician assistants without any reduction in quality, and it behoves Deacon (whom I assume is a believer in evidence-based practice) to present data to the contrary if he wishes to make those assertions. Deacon commented from the floor on a number of occasions, making the assertions that we would expect from an organisation which is dedicated to maintaining the status quo. I really think the time has come to move on.

Peter M Brooks

To the Editor: I would like to clarify several issues relating to the competing interests statement for my editorial in the 16 January issue of the Journal.1 In my statement, I declared that I have used several meal replacement products in clinical settings and that I am not employed by, and do not receive benefit from, any companies producing these products. By way of further information, I have used Optifast (Novartis) and Dr MacLeod’s (Orfam) products in clinical work in the past. These were always sold at cost to patients, or patients were referred to a chemist, so there was no direct financial benefit to me. I currently use KicStart (Pharmacy Health Solutions) as part of the kit for “Professor Trim’s weight loss program for men”. I purchase these in bulk at wholesale prices from the manufacturer and include them as part of the total program (which is much more than just meal replacement).

Garry J Egger

To the Editor: Reading the excellent Christmas edition of the Journal, I was struck by the symmetry of the computed tomography scans of rectus sheath haematomas in a husband and wife, reported by James and colleagues1 (Box). I was wondering if the couple had not been involved in a minor car accident or incident of heavy braking several days earlier, where the lap–sash or buckle of a seatbelt might explain the mirror injuries?

Mark R Nelson

In reply: The rare and spontaneous nature of rectus sheath haematoma leaves its origins open to such interesting questions. Neither of our patients was able to recall a history of motor vehicle trauma, but minor braking trauma certainly remains a possibility, as does respiratory infection leading to protracted cough. Computed tomography scans of matching rectus sheath haematomas in a 62-year-old woman (A) and her husband (B).

David James

To the Editor: I found the recent letter to the Editor from Riley1 fascinating, and the comment from Teo2 depressing in the extreme. Of the many factors that drove me from anaesthetic practice and into the houseboat business 4 years ago, the selfish attitude of many surgeons towards our communal working environment was high on the list. If all parties in an operating theatre wish to be “entertained” with music while operating on their fellow man, then I suppose it might be permissible (but one wonders what many patients would say if they knew). However, it seems quite beyond the autocratic mindset of many surgeons to understand that auditory input is important to anaesthetists for monitoring the patient and for communication between the anaesthetist, surgeon and nursing staff. To impose music as background noise is unacceptable to many of us, especially as auditory discrimination decreases with age. Riley’s description of video as well as audio to distract from the primary function of surgery beggars belief. I note that Teo agrees with me here. When are surgeons going to realise that surgery is only one part of the professional work that goes on in an operating theatre?

Douglas N Gow

To the Editor: The Ear, Nose and Throat Department at the Royal Darwin Hospital services an area of about 550 000 km in the Northern Territory. As part of our remit involves dealing with conditions affecting the organs of speech and hearing, a proportion of our patients have communication handicaps. We have frequently encountered the use of “Hand Talk” among Indigenous patients with communication handicaps, and we feel this merits highlighting to your readership. Hand Talk is an established sign language within and between Aboriginal groups in the NT. Although various groups have different signing systems, there are enough similarities between them to enable inter-group communication. Its existence is thought to date as far back as other spoken Aboriginal languages that have now been lost. Theories about its conception include a means of overcoming language barriers between different language groups, a silent form of communication during hunting expeditions, a means of conversation for women during long periods of mourning when speech is prohibited, and a means of communication for deaf or aphonic individuals.1 While completely separate from established Western sign languages, Hand Talk is a sophisticated and intricate means of communication that allows complex interactions to take place and proficient individuals to integrate into their society.2 In the course of our service delivery in Darwin and to outlying districts, we have been struck by both the frequency with which Hand Talk is used in an impromptu manner and the proficiency users exhibit in communicating with it. The significance of this requires an appreciation of the difficulties of applying Western-type communication rehabilitation to Indigenous Australians in remote regions. For example, providing hearing aids, a basic and standard form of communication rehabilitation in urban areas of Australia, can face numerous obstacles in a remote setting. These include lack of a service provider and technical and maintenance support, poor compliance, and logistical problems (something as simple as running out of hearing aid batteries can be a major problem for those in remote areas). While no data are available on the ubiquity of Hand Talk, it clearly represents a valuable part of Aboriginal culture with an important practical function in a situation where Western models of communication rehabilitation are difficult to apply. While attempts to improve conventional communication rehabilitation should continue, these should be combined with efforts to foster Hand Talk through education and facilitating its dissemination by existing users so it does not suffer the fate of other lost Indigenous languages.3

Patrick Jassar · Garrett F Hunter

To the Editor: On 1 November 2005, the Australian states and territories introduced quadrivalent, pentavalent or hexavalent vaccines for childhood immunisations. This simplifies vaccination for young children, but may impair the ability of health services to provide primary immunisation for refugees over the age of 8 years. The Australian refugee and humanitarian program targets refugees from many countries that have poor primary health infrastructure. In the 2004–05 financial year intake, at least 75% of the 12 096 entrants under the offshore resettlement program came from countries that had immunisation coverage rates below 50% in the 1990s.1,2 Adolescents and adults from these countries generally have patchy vaccination histories and no records. According to Australian guidelines, they warrant full catch-up vaccination, often involving a primary vaccination course.3 Primary vaccination against tetanus, diphtheria and pertussis requires three doses of vaccine. The dose of diphtheria toxoid in vaccines for children or adults over 8 years of age is significantly lower than in early childhood preparations because of potential adverse effects. In 2004, the conjugated pertussis–adult diphtheria–tetanus vaccine for adolescents (Boostrix, GlaxoSmithKline) was introduced to the immunisation schedule to provide boosters against pertussis, diphtheria and tetanus. However, Boostrix has no proven efficacy for primary vaccination against pertussis and is not recommended for adolescents and adults who have no primary cover against pertussis.3,4 As monovalent pertussis vaccine is not available, refugees over the age of 8 years cannot be provided with a primary vaccination course against pertussis. Adult diphtheria–tetanus vaccination (ADT) is the most-used primary vaccine for refugees over the age of 8 years. After the introduction of Boostrix, many state and territory health departments reduced their supply of ADT to immunisation providers. Some refugee health services have attempted to meet demand for ADT by collating individual doctors’ stocks provided under the Emergency Drug (Doctors Bag) supplies section of the federally-funded Pharmaceutical Benefits Scheme, which provides for up to 15 doses of ADT per month. But this is a cumbersome and unsustainable strategy. Some jurisdictions, such as the Australian Capital Territory, supply ADT directly to refugee health service providers. All the new polyvalent childhood vaccines include inactivated polio vaccine. Unless states and territories procure monovalent polio vaccine, primary vaccination against polio for people over 8 years will remain inadequate. People from refugee backgrounds warrant the same level of protection against vaccine-preventable diseases as other Australians. The level of protection may be reduced by failure to provide suitable vaccines. We encourage state and territory health departments to stock sufficient vaccines for adult and adolescent refugees, including ADT and monovalent inactivated polio vaccine. We also recommend that the Australian Technical Advisory Group on Immunisation provide detailed advice on the needs of refugees when crafting immunisation guidelines.

Christine B Phillips MPH, MA, FRACGP · Mahomed Patel FRACCP, FFAPHM

To the Editor: Leprosy rates in Australia are low (less than one case per million population)1 and predominantly occur in Indigenous Australians and immigrants from leprosy-endemic areas.2 A 21-year-old pregnant Burundian woman had migrated to Australia in 2005 from a refugee camp in Tanzania. In the year before her arrival, she had received intermittent courses of steroids for an undefined illness characterised by fever, nightsweats and painful symmetrical peripheral polyarthritis. Three months after arriving in Australia, the patient presented to a rural hospital with a recurrence of the previous symptoms. The symptoms improved on recommencement of prednisolone treatment. The patient was transferred to the Royal North Shore Hospital, where examination revealed bilateral peripheral sensory neuropathy (confirmed by nerve conduction studies); bilateral, enlarged, tender ulnar nerves; and tender hyperpigmented 2–3 cm nodules on the upper arms, but no other skin lesions or infiltrations. Skin biopsy revealed features consistent with erythema nodosum leprosum (ENL), but no acid-fast bacilli (AFB) were detected. Slit-skin smears were also negative for AFB. Leprosy was confirmed by histopathological examination of a sural nerve biopsy, which showed AFB and granulomatous changes of leprosy. The patient commenced multidrug therapy for multibacillary leprosy, with prednisolone for ENL. Leprosy is a chronic granulomatous infection of skin and peripheral nerves with Mycobacterium leprae. Host immune responses determine the spectrum of clinical presentations. Leprosy is classified into either multibacillary disease (≥ 6 skin lesions and/or skin smears positive for AFB) or paucibacillary disease (< 6 skin lesions, with no bacilli on skin smears).3 Type 1 (reversal) reactions are delayed-type hypersensitivity reactions and manifest as neuritis and increased inflammation of pre-existing skin lesions. Type 2 reactions (ENL) are a systemic response to immune complex deposition and manifest with multiple tender nodules, fevers, neuritis, arthritis and iritis.4,5 ENL occurs exclusively in multibacillary disease in 10%–20% of patients, and negative slit-skin smears (as in our patient) are unusual. Possible explanations include undisclosed diagnosis and treatment of leprosy in Tanzania or the combination of steroid therapy and immune changes that occur during pregnancy.6 Multidrug therapy is well established and regarded as safe for pregnant women. Diagnosis of infections that are uncommon in Western countries, especially leprosy, is often delayed.7 For refugees living in remote areas, access to expertise and support may be limited. Therefore, doctors, especially those involved in refugee health, should be aware of “exotic” infections and their varied presentations. Furthermore, effective referral networks should be encouraged, as this resulted in a swift positive outcome in our case.

Sebastiaan J van Hal MB ChB · Bernard J Hudson FRACP, FRCPA

To the Editor: Australia and the other member nations of the World Health Organization’s Western Pacific Region were declared free of circulating endemic poliovirus in 2000, although the last case of endemic polio in Australia occurred in the 1970s.1 Nevertheless, the low risk of vaccine-associated paralytic poliomyelitis (VAPP) persisted through the continued use of the Sabin live attenuated oral polio vaccine (OPV) until it was replaced by the Salk inactivated polio vaccine in the National Immunisation Program from 1 November 2005.2 Despite the eradication of indigenous wild poliovirus and the removal of the risk of VAPP, Australia cannot afford to be complacent with surveillance for cases of poliomyelitis. Polio is a highly infectious disease and is quickly spread through international travel. All countries risk importation of wild poliovirus from the four remaining endemic countries (Afghanistan, India, Nigeria and Pakistan) — as occurred in Indonesia and 11 other countries during 2005.3 Until the latest outbreak, involving over 300 cases, Indonesia had not reported a single case of poliomyelitis since 1995. Genetic sequencing of the wild polioviruses from Indonesia determined that they originated in Nigeria and were related to strains isolated in Sudan, Saudi Arabia and Yemen. Australia is also at risk from imported vaccine-related strains of poliovirus, as indicated by two reports from the United States in 2005. The first was a case of imported VAPP in an unimmunised adult, who had been in close contact with an infant recently immunised with OPV, while in Costa Rica.4 The second report described isolation of OPV poliovirus type 1 with a significant number of mutations (referred to as vaccine-derived poliovirus [VDPV]) from unvaccinated members of a religious community.5 Given that OPV has not been used in the USA since 2000, the source of the virus is unknown. VDPVs have been associated with paralytic polio worldwide. It is imperative that the Australian community maintains the current high rate of polio vaccination coverage, especially for travellers, which remains the best defence against all forms of imported polio. A surveillance scheme for investigation of children with acute flaccid paralysis, the major clinical presentation of poliomyelitis, was established in Australia in 1995. It is coordinated by the National Poliovirus Reference Laboratory and the Australian Paediatric Surveillance Unit (Box). While the scheme focuses on children, specimens from patients of all ages are tested. Notification of all cases with a clinical suspicion of poliomyelitis is essential for the detection of imported polio. Surveillance for acute flaccid paralysis (AFP) within Australia Paediatricians notify cases of AFP via a monthly report card to the Australian Paediatric Surveillance Unit and submit a clinical questionnaire to the National Poliovirus Reference Laboratory (NPRL). Stool specimens from AFP cases are tested at the NPRL for isolation of poliovirus. The Australian Polio Expert Committee reviews the clinical and laboratory data to determine whether the case is compatible with poliomyelitis. The Committee reports to the Australian Government Department of Health and Ageing and the World Health Organization. Protocol for investigation of suspected polio cases Clinicians should phone the NPRL to notify the case and arrange for two stool specimens to be collected 24 hours apart (due to intermittent virus shedding) and within 14 days of onset of symptoms, for testing at the NPRL. Polio antibody testing requires acute and convalescent serum, and is only performed when there is a clinical suspicion of poliomyelitis. Contacts National Poliovirus Reference Laboratory, Victorian Infectious Diseases Reference Laboratory Phone: (03) 9342 2607; fax: (03) 9342 2665; email: polioATmh.org.au; website: http://www.vidrl.org.au/labsandunits/polio/polio_activity.htm Australian Paediatric Surveillance Unit, Children’s Hospital at Westmead Phone: (02) 9845 3005/ 9845 2200; fax: (02) 9845 3082; email: apsuATchw.edu.au; website: http://www.apsu.org.au

Bruce Thorley PhD · Kerri Anne Brussen BAppSc · Elizabeth J Elliott MD, FRACP, FRCPCH · Heath A Kelly MB BS, MPH, FAFPHM

To the Editor: Rickettsia typhi is an endemic cause of atypical pyrexial illness worldwide.1 Its non-specific presentation can lead to misdiagnoses, with overseas reports of unwarranted laparotomies in affected patients.2,3 We describe a patient with R. typhi infection presenting as cholecystitis, in whom a cholecystectomy was avoided by vigilance for R. typhi. A 51-year-old businessman presented to a general practitioner with a 5-day history of fever, sore throat, headaches and myalgia. The illness had begun a week after his return to Sydney from a business trip to major cities in Asia, his last stop being Hong Kong. Investigations revealed mild lymphopenia and thrombocytopenia, negative results on screening for malaria, and normal results on chest x-ray. A non-specific viral illness was provisionally diagnosed. A week later, the patient presented again to a GP with fever, abdominal pain, cough, dehydration and confusion. Investigations revealed lymphopenia (0.7 × 109/L; reference range [RR], 1.5–4 × 109/L], thrombocytopenia (93 × 109/L; RR, 150–400 × 109/L), and raised serum levels of bilirubin (25 μmol/L; RR, 0–17 μmol/L) and hepatic enzymes (alanine aminotransferase, 307 U/L [RR, 5–40 U/L]; alkaline phosphatase, 381 U/L [RR, 30–115 U/L]; aspartate aminotransferase, 389 U/L [RR, 5–40 U/L]; and γ-glutamyl transferase, 364 U/L [RR, < 66 U/L]. Serological tests were negative for hepatitis viruses A, B and C, and dengue and Epstein–Barr viruses. The patient was referred to an emergency department. On presentation to the hospital, the patient was febrile, with severe right upper quadrant abdominal tenderness, and a slight truncal macular rash. Abdominal computed tomography and ultrasound examination indicated cholecystitis (Box). Acute cholecystitis was diagnosed, and treatment begun with intravenous fluids, ampicillin, gentamicin and metronidazole. Following clinical improvement, the patient was discharged on Day 8 with plans for an elective cholecystectomy. In view of the atypical symptom complex, serological testing for leptospirosis, syphilis, and rickettsial, amoebic and HIV infection had been requested during his admission. Results received after discharge indicated a R. typhi antibody titre of 1:1024 (RR, < 1:128). Results of the remaining serological tests were negative. The patient was contacted, and the cholecystectomy cancelled. He has remained well. Murine typhus is a zoonosis caused by R. typhi, and is acquired from rodent flea faeces, either by bite inoculation or inhalation. Hepatobiliary involvement occurs in up to 34% of cases. Histopathology specimens show neutrophilic sinusoidal infiltrates and cloudy swelling of hepatocytes,2 but hepatocyte injury and cholestasis are transient, resolving over 1–3 weeks. Diagnosis is by serological testing: a single indirect immunofluorescent antibody (IFA) titre against R. typhi of at least 1:400; or a fourfold rise in IFA titre from the acute to the convalescent phase (2 weeks apart). The treatment of choice is doxycycline. Although the clinical course is usually benign, the mortality rate can reach 4%.1 Rickettsial diseases remain an under-reported cause of febrile illness.4 As R. typhi has now been described throughout Australasia,1,5 it is important that murine typhus is excluded in patients with atypical pyrexial illnesses and abnormal liver function results. Computed tomography and ultrasound examination in a patient with murine typhus Computed tomography on admission showed pericholecystic inflammation, suggesting acute cholecystitis, and a possible gallstone at the lower pole (arrow), which was later noted to be a fibrous septum on ultrasound examination. Ultrasound examination also showed a thickened gall bladder wall and pericholecystic fluid.

Vidyut P Suttor MB BS, BSc(Med) · Robert B Feller MB BS, PhD, FRACP

To the Editor: Routine school vision screening has been discontinued in many regions.1 Since 2000, children around Newcastle and Lake Macquarie in New South Wales only have their vision checked at school if their parents request it. I am particularly concerned that this may disadvantage young children with diabetes, who may also have undetected amblyopia. These children are already at risk of diabetes-related vision impairment, and simple screening could prevent further disability related to amblyopia. Amblyopia, commonly known as “lazy eye”, is an asymptomatic, potentially treatable condition of poor vision in a “normal” eye. It is caused by the brain suppressing an unclear image from the affected eye. Amblyopia occurs in 2.5%–3.2% of the population.2 The condition needs to be detected early, and treatment needs to be instituted before the end of practical vision development at about 7–8 years of age, otherwise even intensive treatment is unlikely to restore normal vision.3 This is especially important because people with untreated amblyopia have an increased lifetime risk of loss or impairment of vision in their good eye,4 as well as the poor vision in their amblyopic eye. The current Australian screening guidelines for children with diabetes recommend screening for retinopathy after 5 years of diabetes in those who are prepubertal, and annually in adolescents after 2 years of diabetes.5 The International Society for Pediatric and Adolescent Diabetes recommends retinopathy screening in children with diabetes of prepubertal onset at 5 years after the onset of diabetes, or 11 years of age, or at puberty, whichever is earlier.6 Neither document specifies other visual screening (although the National Health and Medical Research Council guidelines do recommend a clinical examination of the eyes for cataract soon after diagnosis). Thus, a 6-year-old child with diabetes would not have his or her vision screened until 11 years of age. An eye with significant amblyopia detected at this age will not achieve normal vision, and the child would be reliant on only one eye for his or her lifetime. Even a 3-year-old child with diabetes would not have visual screening until 8 years of age, the end of practical vision development. The case has been made recently for biennial retinopathy screening for children with diabetes.7 I propose that diabetic children under 9 years of age have their vision fully assessed soon after the diagnosis of diabetes. Should amblyopia be detected then, treatment could commence before the end of active vision development.

Catherine Dunlop

To the Editor: I agree with Jones1 that the body surface area (BSA) formula printed in the position statement on reporting of estimated glomerular filtration rate (eGFR)2 is wrong, even though the authors say that he is mistaken.3 As stated by Jones, the correct formula4 for BSA in m2, for a body weight W kg and height H cm is: (i) BSA = W0.425 × H0.725 × 0.007184. However, the position statement2 gave the following formulas: (ii) BSA = W0.425 × H0.725 × 0.007184/1.73; and (iii) Uncorrected eGFR = GFR estimate (mL/min/1.73m2) × BSA. It appears that the denominator “1.73” has migrated from formula (iii) to formula (ii), so in fact both of these formulas are incorrect. This is potentially misleading for doctors and others who may want to calculate the eGFR for someone who is unusually big or small. Formula (iii) should in fact be: (iii) Uncorrected eGFR = GFR estimate (mL/min/1.73m2) × BSA/1.73. It is interesting that the same two errors in BSA calculations are present on the US National Kidney Disease Education Program website,5 which was presumably the source of the formulas used by the Australian Creatinine Consensus Working Group.

Alan McNeil PhD, FRACP, FRCPA

In reply: McNeil draws attention to the detail in the correction factor we published in an attempt to assist users to “uncorrect” the eGFR derived from the MDRD (Modification of Diet in Renal Disease) equation used in calculating GFR from a serum creatinine concentration. Recalculating the eGFR to remove the adjustment for body surface area (BSA) in an individual is unnecessary except at extremes of body size.1 Readers can be reassured that the formulas published in the position statement,2 if used as directed, will not lead to any error. However, it would have been clearer if we had labelled the “BSA” equation as “correction factor” instead of “BSA”. In the position statement2 it can be misinterpreted that the BSA formula has a denominator, whereas, when used primarily to calculate BSA, it of course does not. Both versions of the formulas (ours in the position statement and McNeil’s) therefore lead to identical answers. The reader can choose which one to use.

Timothy H Mathew MB BS, MRACP, FRACP · Graham Jones MB BS, DPhil, FRCPA · David Johnson MB BS, FRACP, PhD

To the Editor: This question was recently posed by Massie and colleagues in an editorial in the Journal.1 We have experience in providing cystic fibrosis (CF) carrier testing in pregnancy and in those planning pregnancy. In 1998, we ran a 12-month pilot program, the Double Testing Program, offering ΔF508 carrier testing to couples attending the John Hunter Hospital, Newcastle, NSW, for nuchal translucency screening. Of 491 participants, 84% chose CF carrier testing; 23 carrier–non-carrier couples were identified, and no carrier–carrier couples. A postnatal questionnaire compared knowledge of CF, anxiety levels and perceptions of the service between non-carrier couples and couples in which one partner was a carrier.2 Both groups were very satisfied with the service, with no increased levels of anxiety. The second initiative has been to offer CF carrier testing to clients attending the “drop-in” clinic at Hunter Genetics, Newcastle. This clinic has been available for 10 years and provides genetic counselling for clients referred by their general practitioners for pre-pregnancy or pregnancy counselling. Over the past 3 years, there have been 560 occasions of service with CF carrier testing in 499 individuals. Indications for testing include pregnancy screening or a family history of CF. With couples, both are tested for ΔF508, and, if one partner is identified as a carrier, the other partner is tested for 28 other mutations. The area health service, Hunter New England Health, meets the cost of the service and testing — $100 per couple for the ΔF508 test, and $250 for the full mutation screen. Of the 499 individuals, 65 (13%) were found to be CF carriers; after excluding those with a family history of CF, the carrier rate was 5.8%. Twenty carrier–non-carrier couples and one carrier–carrier couple with a family history of CF were identified. Benefits included enabling the carrier–carrier couple to know their 1 in 4 risk of having a child with CF. Most women who have a child with CF want to avoid having further affected children, and most who have a subsequent pregnancy choose prenatal diagnosis.3 Cascade testing can be offered to carrier families. Non-carrier couples can be reassured that they have a low risk. The uptake of CF couple screening is low, given that there are over 3500 births annually at the John Hunter Hospital. The main obstacles are lack of awareness and costs. CF tests are not covered by public health funding or by private health insurance, and this issue needs urgent attention. We believe that GPs are best placed to offer CF couple testing. Testing could be incorporated into routine first trimester pregnancy care. We can provide a distance CF learning package, pamphlets and informed consent material to those interested.

Louise M Christie · Elvira M Zilliacus · Angela J Ingrey · Gillian Turner

To the Editor: We were interested to note that the evolution of morning handover at Launceston General Hospital, as described by Fassett and Bollipo,1 closely parallels our own experience at the Canberra Hospital, and we endorse their points about running a successful meeting. Our hospital has a large Geriatric Unit and all subspecialties are covered, but we do not have a general medicine unit. In 2002, we began a formal morning handover meeting from 08:00 to 08:30 for junior medical officers (JMOs) in the Department of Medicine, with the initial intention of providing an opportunity for Royal Australasian College of Physicians (RACP) basic trainees to present cases they had seen overnight. Scrutiny of the individual’s clinical approach by consultants, in preparation for the RACP examination, was the main emphasis, and “interesting” cases were chosen. The meeting was also used for case presentations by specialty units. Attendance was variable, and many junior staff reported feeling somewhat threatened by having their patient management approach examined in a public forum. Handover of most newly admitted patients did not occur during this meeting. The format was incrementally modified over the following 3 years so that, by 2005, the meeting had become a formal handover of all patients admitted during the previous evening and overnight. Attendance is now compulsory (except for staff attending medical emergencies), and breakfast of brewed coffee and tea with fruit and muffins is provided (funded by the Canberra Hospital). We have over 40 daily attendees (comprising registrars, residents, interns, medical students and 5–10 consultants). We have minimised the number of specialty presentations: these now usually take the form of “red flag” sessions, in which a specialist unit highlights areas of common and/or life-threatening importance (eg, a patient with unstable angina needs admission, regardless of their troponin level; recurrent rigors in a middle-aged person usually signal a bacterial infection). A survey of 57 of the attendees this year revealed that over 90% thought the format and duration of meetings and attendance by consultants was appropriate; 54% and 39%, respectively, said they learned new information every day or every week. Over the past 4 years, the handover has become embedded in the clinical culture of the hospital. The long-term commitment of a small group of consultants has demonstrated that this is a safe and encouraging environment for clinical teaching, and the level of discomfort of the JMOs appears to have receded. The morning handover has been an important means of ensuring that young doctors are exposed to a broad perspective on patient care and that their after-hours patient care can be supervised.

Francis J Bowden MB BS, FRACP, MD · Christian Lueck FRCP, FRACP, PhD · Mark Hurwitz MB BCh, FCCO, FRACP · Karina Kennedy MB BS, FRACP

To the Editor: Healthy ageing is listed as a National Research Priority by the Australian Government. The higher prevalence of sensory, cognitive and mobility impairments in older people presents a major challenge in achieving this goal. The effects of single impairments are recognised,1,2 but the cumulative effects of multiple impairments have not been reported from population-based samples. We aimed to assess the impact of multiple impairments (vision, hearing, cognitive, mobility) on health-related quality of life (HRQOL), hospitalisation, and aged-care service use in an older Australian population. In the second cross-sectional Blue Mountains Eye Study,3 HRQOL was measured by means of the self-administered Short Form 36-item Health Survey (SF-36)4 (n = 3509; mean age, 66.7 years; 57% women). Visual impairment was defined as best-corrected visual acuity (after refraction) of less than 6/12 (better eye). Hearing impairment was defined as average hearing threshold (pure-tone air conduction, frequencies 500–4000 Hz) over 25 decibels (better ear). Possible cognitive impairment was defined as Mini Mental State Examination scores less than 24/30. Mobility impairment was recorded. General linear regression was used to calculate age-adjusted SF-36 mean scores,5 and logistic regression was used to estimate likelihood ratios for use of health and aged-care services. Models were age-adjusted to eliminate confounding. For 2873 participants who had completed the SF-36 (90.9%), the mean physical component score (PCS) was 44.9 (95% CI, 44.5–45.3) and the mean mental component score (MCS) was 51.9 (95% CI, 51.5–52.2). Age was significantly associated with the prevalence of these impairments (P < 0.001). After adjusting for age, people with any of the impairments had poorer mean PCS and MCS than those without the impairment (Box 1). Hospitalisation within the last year was reported by 743 participants (23.5%; 58.3% women), and 97 (3.1%; 65.0% women) reported regular use of community support services. Use of community support services was reported more frequently by people with any impairment, except possible cognitive impairment (Box 1). The presence of two or more impairments was associated with a cumulative, linear decline in HRQOL (Box 2). The successive addition of each impairment was associated with a decrease of 4.0 in mean PCS and 2.1 in mean MCS, and with greatly increased reporting of regular community support service use. The likelihood of participating in or completing the SF-36 decreased with increasing number of impairments. Hence, the prevalence of impairments and the extent of detrimental impacts on HRQOL may be underestimated. Nevertheless, our data highlight a linear increasing pattern of cumulative effects from multiple impairments on HRQOL, hospitalisation, and use of aged-care services. Preventing and reducing these impairments is crucial in maximising healthy ageing. 1 Prevalence, mean SF-36 physical and mental component scores, and use of services by impairment Impairments Prevalence (%) Age-adjusted mean SF-36 scores (95% CI) Use of services: %, age-adjusted and sex-adjusted odds ratio (95% CI) Physical component score Mental component score Hospitalisation in past 12 months Regular use of community services Visual impairment 2.7 42.8 (39.9–45.7) 47.6 (44.8–50.3)* 34.9%, 1.3 (0.8–2.2) 24.2%, 2.9 (1.4–6.0) Hearing impairment 33.4 43.8 (43.0–44.7)* 51.1 (50.3–51.9)* 27.5%, 1.1 (0.9–1.3) 7.4%, 2.7 (1.4–5.0) Cognitive impairment 2.2 42.2 (39.5–44.8)* 46.0 (43.4–48.5)* 28.2%, 1.0 (0.6–1.6) 14.1%, 1.7 (0.8–3.7) Mobility impairment 7.6 32.3 (30.8–33.7)* 48.1 (46.7–49.5)* 41.0%, 2.0 (1.5–2.7) 21.3%, 6.8 (4.2–11.0) All mean values adjusted to 66.7 years, the overall sample mean age. SF-36 = Short Form 36-item Health Survey.4 * Significantly lower than without disability. 2 Mean physical and mental component scores and use of services by increasing number of impairments No. of impairments* Age-adjusted mean SF-36 scores (95% CI) Use of services: %, age-adjusted and sex-adjusted odds ratio (95% CI) Physical component score Mental component score Hospitalisation in past 12 months Regular use of community services 0 (n = 1031) 46.6 (45.9–47.2) 52.8 (52.6–53.8) 22.0%, 1.0 0.4%, 1.0 1 (n = 616) 42.6 (41.8–43.3) 51.0 (50.3–51.7) 25.3%, 1.1 (0.8–1.3) 4.2%, 7.4 (2.7–19.8) 2 (n = 121) 38.6 (37.1–40.0) 48.8 (47.4–50.2) 35.5%, 1.5 (1.0–2.3) 19.0%, 24.9 (8.5–73.2) ≥ 3 (n = 31) 34.5 (32.2–36.8) 46.6 (44.5–48.8) 45.2%, 2.0 (0.9–4.1) 41.9%, 47.4 (13.1–171.4) SF-36 = Short Form 36-item Health Survey.4 * Includes vision, hearing, cognitive and mobility impairments.

Ee-Munn Chia · Jie Jin Wang · Elena Rochtchina · Paul Mitchell

To the Editor: Despite a new survey with some interesting results, the MJA’s Medicine and the Media special failed overall to advance the topic. Van der Weyden and Armstrong enthusiastically cite Schwartz and Woloshin: “don’t report preliminary findings”.1 (In fact, they wrote, “In general, don’t report preliminary findings”,2 but let’s allow this journalistic context tweak.) In their article, Schwartz and Woloshin contend that this is because “what is new may turn out to be wrong”. Many things reported in newspapers are subject to subsequent change. So, to be on the safe side, let’s exclude them all. No more Cabinet leaks: let’s wait till everything is resolved and perfect-bound by the government stationery office. An end to covering murder trials: what if the accused is found not guilty? Forget about half-time match scores, and definitely no celebrity weddings because they’ll be divorced within the year. A moratorium on discussing preliminary findings would put off-limits highly respected annual meetings such as those of the American Society of Clinical Oncology and the American Society for Reproductive Medicine, which set the treatment agenda annually for clinicians and patients in the fastest-paced medical specialties. It is nonsense to suggest the media should censor early results, a point that has been made previously.3 The world has moved on. News is no longer a series of monolithic reports, each entirely true and complete. Like life, news is a work in progress — a rolling tide of updates, each modifying the last. Everyone loves a winner, and it is gratifying to note the Sydney Morning Herald is currently top of the Media Doctor league table.4 This website is a useful focal point, and Smith and colleagues’ suggestion that researchers take some responsibility for how their results are presented to the public is helpful.5 But of the three solicited commentaries,6-8 not one was from a newsroom health reporter, or — better still — a daily news editor or producer who decides, amid the controlled chaos of breaking and evolving stories, which reports should run and how prominently. Was this because they were not approached? Imagine a five-article package on immunisation practice without a view from a general practitioner, or one on appendicectomy without the insight of a surgeon. It is a serious omission that undermines the credibility of the MJA’s package and calls its motivation into question. Genuine rapprochement might threaten the sport of media sniping that has become a lively sideline for some medical journals.

Julie Robotham BA (Oxon)

In reply: Would that we could convince newspaper editors not to publish inaccurate stories, cabinet leaks, sensational and one-sided details of trials in progress, hope-raising interim football scores or the details of celebrity weddings. However, our media package was not that ambitious! In the context of medical reporting, we stand by our advice against publishing interim results. The reason that clinical trials have protocols, power calculations and stopping rules is that (just as football is a game of two halves) the results are really not relevant or reliable until all the data have been collected and analysed. Our media package included contributions from two highly respected medical writers1,2 and a representative of Australian journalism’s peak body.3 Robotham is correct in surmising that we did not approach a newsroom health reporter or a news editor, but the immunisation analogy seems spurious. If we published a research paper that identified, for instance, deficiencies in general practitioners’ delivery of vaccinations, we would not necessarily accompany it with a commentary from a GP. Whatever their discipline, we would assign the task to someone who could point the way towards best practice. Nevertheless, we are grateful for Robotham’s interest and acknowledge the responsibilities shared by both journalists and medical journal editors, albeit with differing emphasis — the dissemination of accurate and timely information.

Ruth M Armstrong BMed · Martin B Van Der Weyden MD, FRACP, PhD

To the Editor: During a national multicentre study of snake bites — the Australian Snakebite Project (ASP), involving over 40 hospitals — we have recently noted a high rate of early allergic reactions following the administration of tiger snake antivenom in Australia. People with suspected or definite snake envenoming are recruited to ASP, and laboratory and clinical data and serial blood samples are collected to measure venom and antivenom concentrations. From 1 November 2005 to 31 January 2006, 14 patients who had been given tiger snake antivenom (CSL Limited, Parkville, VIC) were recruited. These patients are briefly described in the Box, and include bites from several different groups of snakes. Of the 14 patients, 11 exhibited immediate systemic hypersensitivity reactions to antivenom infusion. Reactions were mild in five patients, moderate in three, and severe in three, according to the grading system by Brown.1 The six patients in the moderate and severe groups fulfilled the criteria for anaphylaxis according to a recent consensus definition.2 All patients required specific treatment in addition to ceasing antivenom therapy, and nine were treated with adrenalin. Antivenom was recommenced in all patients at a slower rate, although an adrenalin infusion was required in four of these and repeat doses of intramuscular or subcutaneous adrenalin in another four while the antivenom infusion continued. Over the same period, there were seven administrations of brown snake antivenom (over 30 vials of antivenom) reported to ASP without any hypersensitivity reactions. There has been a previous report of allergic reactions to tiger snake antivenom in a single hospital,3 and we are concerned that there may be a particular problem with tiger snake antivenom. The reaction rate in this series is similar to reported rates in parts of the world where high reaction rates have been attributed to relatively impure antivenom preparations.4,5 The reactions here were traced to at least four different batches of tiger snake antivenom. We have informed CSL of this high rate of reactions and the antivenom batch numbers, and we have encouraged the treating doctors to make formal reports of each adverse reaction to CSL and the Adverse Drug Reactions Advisory Committee. Health care professionals treating patients with tiger snake antivenom need to be aware of the possible higher risk of anaphylaxis with tiger snake antivenom and be prepared to treat with adrenalin. Recommendations for the diagnosis and treatment of anaphylaxis have recently been reviewed.6 However, this current problem with CSL tiger snake antivenom should not cause health professionals to reduce or cease its use. In all patients described here, control of the adverse reaction and continuation of antivenom was possible. The rapid identification of this problem over a short period was only possible because of our large multicentre collaborative study, and supports such studies for recognising uncommon envenoming syndromes. Fourteen patients administered tiger snake antivenom for snake envenoming Age/sex Previous antivenom Snake Clinical features Grading Treatment of reaction 47 M No; SH BHS No reaction Nil Nil 13 M No TSG No reaction Nil Nil 17 M No TSG Nil Nil Nil 20 M No TSG Generalised erythema, urticaria, tachycardia Mild IM adrenalin (0.5 mg), then IV adrenalin infusion 12 M No RBBS Generalised erythema Mild IM adrenalin (0.2 mg × 3) 12 M No RBBS Generalised erythema, urticaria Mild IM adrenalin (0.25 mg 3 2) 28 F No TSG Pruritus, erythema and moist cough (no wheeze) Mild Promethazine (10 mg) 53 M No TSG Generalised pruritus Mild Promethazine (25 mg) 32 M No TSG Dizziness, chest tightness, tachycardia, vomiting Mod. SC adrenalin (0.3 mg) 21 M Yes; SH SBS Generalised rash and pruritus, vomiting Mod. IV adrenalin infusion for 1 hour 9 M No SBS Urticarial rash, chest tightness Mod. IV adrenalin infusion 55 M No; SH PHS Generalised pruritus, diaphoresis, confusion, hypotension Severe SC adrenalin (0.5 + 0.5 + 1 mg), IV fluid, IV hydrocortisone (100 mg) M No RBBS Rash, wheeze and hypotension Severe Adrenalin, IV fluid, antihistamines, steroids 45 F No TSG Hypotension, sweaty and unwell appearance Severe IV adrenalin infusion BHS = Broad-headed snake (Hoplocephalus bungaroides); IM = intramuscular; IV = intravenous; Mod. = moderate; PHS = Pale-headed snake (H. bitorquatus); RBBS = Red-bellied black snake (Pseudechis porphyriacus); SBS = Stephens’ banded snake (H. stephensii); SC = subcutaneous; SH = snake handler; TSG = Tiger snake group (any snake from the Notechis, Hoplocephalus, Tropidechis, and Austrelaps genera).

Geoffrey K Isbister · Alan Tankel · Julian White · Mark Little · Simon G Brown · David J Spain · Chris F Gavaghan · Bart J Currie

Comment: Thank you for the opportunity to comment on the letter by Isbister et al regarding hypersensitivity reactions to tiger snake antivenom. CSL has been notified about the cases of hypersensitivity reactions to tiger snake antivenom in general, but has only received two individual case reports. We have been in contact with the study investigators and have requested more detailed information on the other patients so that we can further our investigations. A thorough check of product manufacturing records revealed no deviations from approved specifications for tiger snake antivenom. It is important to note Isbister et al have advised health professionals not to reduce or cease the use of tiger snake antivenom. We would like to draw physicians’ attention to the approved Product Information before use of the product. The tiger snake (and other antivenom) Product Information lists the possibility of both anaphylactic and anaphylactoid reactions. Hypersensitivity and skin reactions (including urticaria, rash, hypotension, bronchospasm, anaphylaxis and delayed serum sickness) are listed as common, and are more likely to occur in people who have had previous exposure to equine-based products. In addition, the Product Information describes an anaphylactoid reaction which can occur because the antivenom has the ability to bind complement. The risk of this reaction can be minimised by adequate dilution of the antivenom (1 : 10 for adults and 1 : 5 in small children) before infusion. Further, the Product Information states that a syringe already loaded with 1 : 1000 adrenalin must be available during antivenom therapy. CSL is continuing to monitor this situation closely and is awaiting further details on the patients from the reporting physicians. In the meantime, we encourage users of all antivenom products to report any untoward reaction to CSL so that these can be fully evaluated.

Jane Leong

To the Editor: Community strains of methicillin-resistant Staphylococcus aureus (MRSA) are increasingly seen in Australia, particularly in certain population subgroups, such as Pacific Islander1 and Aboriginal2 people. We report the case of an African with community MRSA to highlight its existence in yet another subgroup. Given the increase in people arriving in Australia from Africa under the Humanitarian Program (with around 8500 arrivals from Africa in 2004–2005)3 and their wide dispersal around the country, it is possible that African community MRSA will be seen increasingly in Australia. A 53-year-old Burundi refugee presented with an infected wound overlying the left lateral malleolus after laceration 6 weeks previously in a Tanzanian refugee camp. An unknown antibiotic was given for 2 weeks before travel to Australia. On the patient’s arrival in this country, the wound appeared purulent, erythrocyte sedimentation rate was 82 mm/h (reference range [RR] < 10 mm/h), and C-reactive protein level was 9 IU/L (RR < 5 IU/L). Plain x-rays and bone scans suggested osteomyelitis. A wound swab grew S. aureus, Streptococcus pyogenes and Pseudomonas species. The patient was initially given intravenous cefazolin, and then definitive therapy (for MRSA and S. pyogenes, ignoring the colonising pseudomonad) with oral clindamycin (450 mg three times daily). Clinical resolution was complete, and levels of acute-phase reactants returned to normal. The antibiotic sensitivity pattern of the S. aureus isolate raised suspicion that it might be an unusual strain: it was resistant to methicillin, tetracycline and trimethoprim–sulfamethoxazole, but sensitive to erythromycin, clindamycin, ciprofloxacin, gentamicin, vancomycin, linezolid, mupirocin, rifampicin, fusidic acid and chloramphenicol. The mecA gene was detected by polymerase chain reaction testing, confirming methicillin resistance. The organism possessed staphylococcal cassette chromosome mec (SCCmec) element type IV. The Panton–Valentine leukocidin gene, staphylococcal enterotoxins A to E and toxic shock syndrome toxin-1 were not detected. As DNA fingerprinting with standard pulsed-field gel electrophoresis showed a novel banding pattern, the “gold standard” of multilocus sequence typing was used for identification. This confirmed an ST140 allelic profile, which has not been seen previously in Australia.4 On the balance of probabilities, the isolate represents an African community MRSA strain, not previously detected in Australia. Non-multiresistant community MRSA is not widely recognised in African countries. Hospital MRSA rates vary widely in Africa (eg, between 21% and 30% of all S. aureus isolates in Nigeria, Kenya and Cameroon, and fewer than 10% in Tunisia and Algeria5), but most are multiresistant. Medical practitioners in Australia who treat African refugees need to be aware that pyogenic soft tissue infections could be caused by community MRSA, and these MRSA strains may have a different antibiotic sensitivity profile to Australian community MRSA strains. It is essential to take appropriate specimens for microbiological analysis (wound swabs and possibly blood cultures and/or tissue samples), as antibiotic susceptibility profiles are increasingly unpredictable.

Annabelle Donaldson MB ChB · Iain B Gosbell MD, FRACP, FRCPA