Letters

To the Editor: We read with interest the recent publication by Rowe and colleagues.1 The authors reported low influenza vaccine coverage (39%) among pregnant women in Victoria from 2015 to 2017. Individual‐level factors associated with this finding included greater maternal age, primigravidity, early antenatal care and GP‐led antenatal care.1 As the authors accurately concluded, integrating vaccine delivery into antenatal care pathways is important to improve pregnant women's vaccination coverage.1 Our team reported on this previously, with coverage approximating 90% achieved by introducing standing orders for midwives.2 In collaboration with key stakeholders from six Victorian maternity services, a Monash University‐led project funded by Better Care Victoria is currently underway to implement integrated vaccination strategies and measure the cost and magnitude of improvement in maternal immunisation coverage in Victoria,3 the results of which will be available by the end of 2019. One of the key findings in the article by Rowe and colleagues1 is higher odds of influenza vaccination in women who gave birth after 37 weeks' gestation compared with women who gave birth before 28 weeks (adjusted odds ratio [aOR], 4.74; 95% CI, 3.54–6.35). A similar finding was reported for women who gave birth between 28 and 36 weeks gestation (aOR, 4.13; 95% CI, 3.07–5.56).1 This finding has two important implications. Firstly, it may indicate a potential beneficial effect of influenza vaccine received by pregnant women in reducing pre‐term birth (< 37 weeks' gestation). This is consistent with a recent systematic review and meta‐analysis that reported inactivated influenza vaccine to have a protective effect against pre‐term birth and low birth weight.4 Secondly, this finding may serve as an opportunity to emphasise the safety and benefits of influenza vaccines on perinatal outcomes. As the authors alluded to in their discussion, pregnant women tend to view influenza as primarily a health risk for themselves rather than for their infants.1 Given the importance of health care providers' recommendations in encouraging influenza vaccination among pregnant women, timely dissemination of the potential benefit in lowering the chance of pre‐term birth could further empower health care providers to recommend influenza vaccines to pregnant women.5

Khai Lin Kong · Michelle L Giles · Euan M Wallace

To the Editor: Last month, the Medical Journal of Australia called for manuscript submissions on the topic of “Women in medicine and medical leadership in Australia — is there gender equity?” We answer with a resounding no. Indeed, we believe the question itself perpetuates gender disparity by suggesting that the answer is up for debate. There is overwhelming evidence to demonstrate that gender equity in medicine and medical leadership in Australia has not been achieved. Women have had gender parity in Australian medical schools for decades; however, they represent only 28% of medical deans and 12.5% of hospital chief executive officers.1 In February 2019, The Lancet dedicated an entire issue on advancing women in science, medicine and global health.2 The MJA has also reported on capacity, capability and credibility barriers for women in health leadership.3 These disparities are even greater for Aboriginal and Torres Strait Islander women, women of colour and women with disabilities. There is an urgent need to shift our focus from asking whether gender inequity exists to implementing and evaluating sustainable strategies to change the status quo. This year, the Australian Medical Association of Victoria changed its constitution to include a 40% gender quota for its board.4 The Royal Australasian College of Surgeons has established a business plan with tangible indicators to promote leadership and flexible training for its female surgeons.5 Both the Women in Tropical Health Catalyse Program6 in Australia and Wāhine Connect (www.wahineconnect.nz) in New Zealand offer mentoring for women in medicine and medical leadership. We need to bolster current strategies aimed at improving the number of women in medical leadership. Moreover, we need to keep our workplaces, colleges, committees, professional associations and academic journals accountable for the role they play in the persistent gender inequities in medicine and medical leadership in Australia. We invite the MJA to follow The Lancet's example and dedicate an entire issue to strategies that advance women in medicine and medical leadership. We implore it not to ask “is there gender equity?” when the answer to this question is patently clear. The answer is no.

Allison Hempenstall · Jillian Tomlinson · Marie M Bismark

To the Editor: I thank Whitfield and colleagues for their article about hyperemesis gravidarum and abnormal liver function in pregnancy.1 In a 15‐month prospective study in South West Wales, abnormalities of liver function were present in 3% of pregnancies.2 In managing pregnant women with hepatic dysfunction, it is important to consider uncommon causes of liver disease that may be associated with serious maternal and fetal morbidity and mortality if untreated. Addison disease is a rare but potentially life‐threatening condition that may imitate hyperemesis gravidarum in presenting with vomiting, weight loss, postural hypotension and hyponatraemia.3,4 Addison disease has also been associated with elevated hepatic transaminases in 16 published cases, reversing with glucocorticoid replacement.5 In excluding Addison disease, the physiological rise in cortisol during pregnancy must be considered using trimester‐specific reference ranges for short synacthen testing.6 In the pregnant woman with unexplained liver disease and fever, acyclovir should be administered empirically, given the absence of cutaneous vesicles in up to 80% of affected patients and the extreme maternal and fetal mortality associated with untreated herpes simplex virus hepatitis.7 Budd–Chiari syndrome should be considered with abnormal liver function in pregnancy with abdominal pain, hepatomegaly and ascites. Additionally, the use of herbal and over‐the‐counter medications should be sought in pregnant women with abnormal liver function, given the high rates of complementary and alternative medicine use in pregnancy and their potential to cause liver injury.8 Investigations need to be interpreted with regard to gestational physiological changes, as copper, ceruloplasmin, α‐1 antitrypsin and alkaline phosphatase levels rise significantly in pregnancy. Serum lipase levels are commonly elevated in hyperemesis gravidarum — levels up to ten times normal have been reported in the absence of pancreatitis.9 Antithrombin III levels may be useful to distinguish acute fatty liver of pregnancy from pre‐eclampsia with haemolysis, elevated liver enzymes and low platelets.10 Bile acid levels are not specific for intrahepatic cholestasis of pregnancy, being elevated in many hepatic disorders including non‐alcoholic fatty liver disease. Twenty per cent of women with pruritus typical of intrahepatic cholestasis of pregnancy have normal bile acids and liver function at presentation, and symptoms may precede abnormal biochemistry by up to 6 weeks.11 In addition to ondansetron and glucocorticoids, mirtazapine has been effective in the management of hyperemesis gravidarum in case reports.12

Adam Morton

To the Editor: In their article, De Costa and colleagues1 clearly demonstrate the importance of making mifepristone freely available for prescription to all registered Australian medical practitioners, and they emphasise that the current need for special registration discourages general practitioners to become involved in medical abortion provision. We conducted a cross‐sectional study of 39 GPs and 30 primary health care nurses from regional or rural Victoria and identified additional uptake barriers.2 Most study participants showed important gaps in medical abortion knowledge, despite their overall positive stance on abortion and extensive experience with women with unplanned pregnancies, and only a few indicated to be current medical abortion providers. Although nearly all health practitioners indicated they would support a colleague in providing abortions, fewer GPs than nurses were interested in medical abortion training. The main reported uptake barriers to medical abortion provision included a lack of training opportunities as well as the absence of local support services required in Australia for the recommended pre‐abortion ultrasound and for surgical back‐up in the case of complications. Participants additionally worried about the legality of providing abortions, and some indicated that their practices would not allow the provision of this service. Abortion access for Australian women in regional and rural regions is still very restricted. By moving early medical abortion provision into the primary health care setting of underserved regions, and particularly in general practice, this situation can be considerably improved. However, the uptake among GPs remains low.3 In addition to addressing uptake barriers, alternative solutions to improve abortion access in underserved areas should be further explored as well, such as the use of telemedicine (until recently provided by the Tabbot Foundation) and the inclusion of primary health care nurses in the abortion provision process — an evidence‐based practice that is already extensively implemented in a range of high income countries.4,5 A nurse‐led model approach not only addresses the shortage of physicians but also the time‐intensive aspect of the medical abortion process, and it provides women with choice and flexibility, which is indispensable to their reproductive autonomy and, thus, to their overall welfare.

Caroline Moel‐Mandel · Melissa Graham

To the Editor: Recreational misuse of nitrous oxide remains a significant public health problem,1 sustained in part by the ready availability online of gas‐containing canisters intended for use in the catering industry. Known as “nangs” or “whippits” and usually purchased in bulk, each canister contains 8 g of nitrous oxide. When inhaled, this gives a seconds‐long “high”, which is typically prolonged by using several “nangs” in a single session. Some individuals can consume hundreds each day. Prolonged exposure to nitrous oxide leads to the oxidisation of vitamin B12, rendering it unusable in key enzymatic reactions necessary for normal myelin synthesis.2 Over time, this leads to a potentially devastating neuropsychiatric syndrome that commonly presents with ataxia.3 Notably, the culprit shortage of vitamin B12 is a qualitative one and can be purely so, meaning that marked clinical deficits emerge in the presence of serum B12 levels that appear normal on standard laboratory assays. Furthermore, with continued exposure to nitrous oxide, these deficits will respond poorly to vitamin B12 supplementation. In a year‐long clinical audit at Royal Prince Alfred Hospital (2017–2018), seven nitrous oxide users, all aged between 20 and 30 years, presented with ataxia that ranged from mild to severe (Box 1). Most patients also had psychiatric symptoms. Nearly every patient estimated using 100 or more canisters of nitrous oxide per day in the months before being seen. Four patients also reported engaging in B12 supplementation (both oral and parenteral), aiming to circumvent the harmful sequelae of prolonged nitrous oxide misuse. Laboratory studies showed that all seven patients had accumulated homocysteine, as is usually seen when vitamin B12 is in short supply in the body.2 Individuals who reported taking supplements had serum B12 levels that were either normal or in excess of normal, implicating a qualitative deficiency of metabolically useful B12. Evidence of demyelination was seen on spinal cord imaging in six patients, including all those who used supplements, with the “inverted V” sign4 visible on T2‐weighted magnetic resonance imaging sequences (Box 2). Despite treatment according to best practice guidelines, all patients left hospital with persistent symptoms, and most were unable to walk or to attend to their bodily needs without the assistance of family members (modified Rankin score, 4). Sadly, one of the least affected individuals re‐presented to hospital with worsened symptoms because of continued nitrous oxide misuse. At every opportunity nitrous oxide users should be reminded of the futility of B12 supplementation, as one of many reasons why they should choose to avoid this profoundly destructive drug. Box 1 – Patients presenting with symptoms due to nitrous oxide misuse Age (years) Sex Canister use Duration of use B12 supplementation Ataxia severity* Psychiatric symptoms† Homocysteine level Serum B12 (active) MRI: “inverted V” sign‡ mRS: Day 1 mRS: discharge 20 Female 250/day 1 year No Severe Yes High Low (low) Yes 4 4 30 Male 60/day 1 year No Moderate Yes High Low (low) Yes 1 1 30 Male 100/day 6 months No Mild No High Low (normal) No 1 1 21 Male 200/day 1 year Yes Severe Yes High Normal (normal) Yes 4 4 23 Female 300/day 2 months Yes Severe Yes High Normal (high) Yes 4 4 23 Female 200/day 2 months Yes Severe Yes High High (high) Yes 4 4 28 Male 300/day 1 year Yes Mild No High Normal (normal) Yes 1 1 MRI = magnetic resonance imaging; mRS = modified Rankin score of neurological disability. * Ataxia: mild = visible gait disturbance; moderate = frequent falls; severe = inability to walk without assistance. † Psychiatric symptoms included mood disturbance, memory impairment and psychosis. ‡ MRI findings: “inverted V” sign on T2‐weighted MRI spinal cord imaging (Box 2). mRS: 0 = no symptoms; 1 = no significant disability despite symptoms; 2 = slight disability; 3 = moderate disability; 4 = moderately severe disability, unable to walk or attend to bodily needs without assistance; 5 = severe disability, bedridden; 6 = dead. Box 2 – T2‐weighted magnetic resonance imaging sequence showing “inverted V” sign, indicating the presence of dorsal column demyelination

Christopher Blair · Chris Tremonti · Leon Edwards · Paul S Haber · G Michael Halmagyi

To the Editor: Recreational use of nitrous oxide is increasing among regular drug users in Australia1 and internationally.2 In a New South Wales survey of participants who had used ecstasy or other stimulants in the past 6 months, 75% of respondents reported recent use of nitrous oxide in 2018, up from 20% in 2013.3 Nitrous oxide use while bingeing on stimulants rose from 4% of respondents in 2013 to 16% in 2017.4,5 Access to nitrous oxide has been facilitated by businesses offering 24/7 delivery of large quantities of canisters, ostensibly for whipping cream.6 We examined presentations to 60 emergency departments (EDs) across NSW from January 2012 to December 2018 (covering about 82% of all NSW ED presentations) in which the patient reported inhaling nitrous oxide outside a therapeutic setting. We extracted records from the Rapid Emergency Department Data for Surveillance (REDDS) dataset in which the person was aged 16 years or over and “nitrous oxide” or related terms were mentioned in the presenting problem, nursing assessment or diagnosis fields. We manually reviewed records for inclusion; this method may underestimate true presentation counts. ED presentations fitting the criteria were infrequent (n = 118) but increased over time, particularly from 2016 to 2018 (Box). Most presentations were in persons aged 16–30 years (n = 98, 83%) and just over half were men (n = 66, 56%). Almost half indicated polydrug use (n = 54, 46%), and one quarter indicated chronic or heavy use of nitrous oxide (n = 28, 24%). Consistent with the case literature,7 presenting problems and diagnoses included injury (n = 15, 13%), neurological symptoms (n = 14, 12%), loss of consciousness or syncope (n = 13, 11%), respiratory arrest (n = 2, 2%), self‐harm or suicidal ideation (n = 16, 14%), or other mental health conditions (n = 28, 24%). This increase in presentations may reflect changes in the underlying population or in data collection, rather than changes in drug use. However, the trend is consistent with drug use survey findings,1,3 suggesting that recreational nitrous oxide use may be an emerging health problem in Australia. Clinicians should include nitrous oxide use as part of a drug history, consider potential use among young patients presenting with neurological symptoms resembling B12 deficiency, and educate users on the health impacts, harm minimisation strategies and available support services. NSW Health is educating at‐risk groups through multilingual fact sheets8 and targeted social media and other messaging. This report is the result of an investigation carried out by the NSW Ministry of Health under the provisions of the NSW Health Administration Act 1982; therefore, specific ethics approval was not required. Data were sourced from the REDDS, which is maintained by the Ministry of Health, and were analysed by Ministry of Health staff for the purpose of the investigation. Box – Emergency department presentations in New South Wales in which the patient reported inhaling nitrous oxide in a non‐therapeutic setting (from January 2012 to December 2018) Data source: Rapid Emergency Department Data for Surveillance (REDDS), held by NSW Ministry of Health.

Anna Bethmont · Claire E Harper · Betty SH Chan · Andrew H Dawson · Jeremy McAnulty

To the Editor: The authors of a recent Guttmacher–Lancet Commission article1 point out that Australia is a signatory to the United Nations Sustainable Development Goals, which nominate sexual and reproductive health as rights. The key focus of the article on the Guttmacher–Lancet Commission is around human immunodeficiency virus and sexually transmitted infections, unintended pregnancy, contraception, abortion, and sexual violence.1 These are all important reproductive health rights to address. While the Guttmacher–Lancet Commission also includes maternal and newborn health, there is no mention of reproductive carrier screening. Reproductive carrier screening involves testing prospective parents — before pregnancy, ideally, or in the early stage of pregnancy — for carrier status for autosomal recessive and X‐linked recessive disorders, and giving reproductive choices to people at increased risk of having an affected child. These choices include pre‐implantation genetic diagnosis, prenatal diagnosis by chorionic villus sampling or amniocentesis, donor gametes or embryos, adoption, having no children, or ignoring the risks. Most couples with or at risk of having an affected child have no family history, which is typical for recessively inherited diseases. Reproductive carrier screening is available in Australia, although only through a fee‐for‐service mechanism, but most couples are unaware of its availability. Currently, screening for cystic fibrosis, fragile X syndrome, and spinal muscular atrophy is available,2 and in the future we may be able to screen for a vastly expanded number of diseases. The Royal Australian and New Zealand College of Obstetricians and Gynaecologists has recently released a position statement to recommend that all women, either before pregnancy or in the first trimester, should be offered carrier screening for inherited conditions.3 Reproductive carrier screening should be a routine part of pregnancy care and should be considered a health care right.

R John Massie · Martin B Delatycki

To the Editor: We welcome the research letter by Roche and Kostadinov,1 who have reported an increasing trend in risky alcohol consumption among adults aged over 50 years. The Report of the Chief Health Officer Queensland in 20162 presented the trends for lifetime risky drinking3 and found a similar trend of risky drinking for males aged 65 years and older. While agreeing with the authors' overall message, there are concerns on the method used to assess trend, in particular, the highly significant P values for the relatively small change in prevalence. Population‐weighted counts, as presented in the research, are used to ensure that prevalence estimates are representative of the national population, but are inappropriate to use when determining the precision of the estimate. Using the weighted counts to calculate the χ2 statistic for trend effectively assumes that the population is the sample size (ie, 8 015 139 in 2016), when in fact the sample size was 23 772 in 2016.4 The effect of this is an overprecise estimate that results in a highly significant P value. Using the authors' table, a simple linear regression with prevalence as the outcome and year as the predictor can be used as a quick face validity check. This results in non‐significant trends for both risky and high risk groups; however, when these groups are combined a significant result is observed (significance level P < 0.05). My research into lifetime risky drinking also encountered this issue of incorporating population weights into trend analysis and this was resolved by using various regression techniques.5 Another advantage of using regression is the ability to incorporate interactions in models to test differences in trend between subgroups. Using this technique identified a statistical difference in lifetime risky drinking trends between people aged 18–29 years and people aged 65 years and older (P < 0.001) (Box). I believe that the authors' findings are important, but we need to take care that results are not discredited by choice of methodology. Interactive data visualisations of alcohol consumption and other risk factors trends for Queensland can be found at the Queensland health website.6 Box – Trends in prevalence of lifetime risky drinking, by age group, Queensland, 2010–2018* * Reproduced from Queensland Health's website.6

Tim Roselli

To the Editor: A 40‐year‐old man with no past medical history presented to the emergency department complaining of a painful red rash across his back, which he noticed when he undressed at home after work. On examination, the patient was systemically well with normal vital signs. He had a non‐blanching, band‐shaped erythematous rash across his upper back, with no associated vesicles, consistent with a first degree burn. The patient worked as a field environmental engineer and was required to wear high visibility shirts throughout his working day. It was noticed that the band‐shaped rash coincided with the upper high visibility band on his work shirt (Box). The patient also reported that the high visibility tape on his shirt often becomes extremely hot when he works out in the sun, and he occasionally has to change position so the shirt does not touch his skin in that area. The rash was managed as a first degree burn, with emollient aloe vera cream and simple analgesia. Retroreflective tape is used on work clothing to increase the wearer's visibility to others, especially in the dark. It usually consists of minute glass beads or prismatic elements encapsulated in a transparent film, which reflect light back towards its source. Safety warnings mention cases of increased heat build‐up around shoulders, neck and ears, but no cases have been published in the medical literature. Skin abrasions have also been described when aged tape cracked and frayed was brushed against the forehead.1 The manufacturers also warn about the tape risk to smoulder or melt when subjected to heat. To the authors' knowledge, this is the first reported case of skin burns occurring secondary to overheating of retroreflective tape. Even though not life‐threatening, it caused discomfort to the patient for a few days. Workplaces mandating clothes with retroreflective tape should ensure that garments with the tape in areas touching the skin are not worn in very hot and sunny conditions and consider using removable vests instead. Manufacturers should consider designing shirts that decrease direct contact between retroreflective tape and skin, potentially by increasing the number or thickness of cloth layers under it. Box – Band‐shaped erythematous rash across the patient's upper back coinciding with the upper retroreflective band on the work shirt

Ioana Vlad

To the Editor: The efficacy of evidence‐based doses of statins is well established. The poor compliance with high intensity lipid‐lowering pharmacotherapy reported by Brieger and colleagues1 in the CONCORDANCE study has been noted in many studies.2 Compliance is related to several factors, including patients’ perspectives and concerns about quality of life and possible adverse effects3 compared with potential benefits. Adverse effects of statins are extensively documented, are dose‐related, and contribute to suboptimal compliance.2 Outcomes with high intensity lipid‐lowering doses are predominantly extrapolated from trials and epidemiological endpoints. Treatment to specific target cholesterol levels is not supported by any direct trial evidence, acknowledged in the United States lipid guidelines since 2013.4 Further, a Cochrane meta‐analysis has failed to show any reduction in hard clinical endpoints such as myocardial infarction, stroke or mortality when receiving statin treatment in the first 3–6 months after acute coronary syndrome.5 The maximum reduction in total mortality reported on statins, around 15% by 3 years, is seen with about 40 mg of simvastatin, equivalent to about 5 mg of atorvastatin, associated with over a 25% reduction in myocardial infarction6 — impressive for a single coronary preventive intervention. The safety and efficacy of only a 10 mg dose of simvastatin led to its approval for over‐the‐counter sale in the United Kingdom in 2005.7 Being competitive enzyme inhibitors, as approved statin doses are increased, plateauing efficacy is overtaken by increases in a variety of adverse effects and potential harms. For example, high dose compared with conventional dose statin (2.5–10 mg of atorvastatin) has no impact on survival (Box) but increases myopathy by up to 29‐fold and liver dysfunction by up to ninefold.6 Higher intensity statin may achieve a small reduction in coronary events (only statistically significant in the Treating to New Targets [TNT] trial10) but at a price with respect to safety, tolerability, overall survival and compliance, particularly in older patients with multiple comorbidities. Instead of increasing statin dose, a greater reduction in cardiovascular risk may be achieved by combining smoking cessation, antithrombotic therapies, control of blood pressure and diabetes, weight loss and other lifestyle measures, each of which can reduce coronary events by 10–20%. Box – Trials comparing conventional versus high intensity statin dose Trials (years of follow‐up) Number of patients Statin Doses* (mg) Major CHD events† (% of cohort) Mortality† (% of cohort) Total CHD SEARCH8 12 064 Simvastatin 2.5 20% 16% 7% 7 years (mean) Simvastatin 10 20% 16% 7% A to Z9 4497 Simvastatin 2.5 12% 7% 5% 2 years (median) Simvastatin 10 12% 6% 4% TNT10 10 001 Atorvastatin 10 8% 6% 2% 5 years (median) Atorvastatin 80 7% 6% 2% IDEAL11 8888 Simvastatin 2 10% 8% 4% 5 years (median) Atorvastatin 80 9% 8% 4% PROVE‐IT TIMI12 4162 Pravastatin 2 9% 3% 1% 2 years (mean) Atorvastatin 80 8% 2% 1% A to Z = Aggrastat to Zocor; CHD = coronary heart disease; IDEAL = Incremental Decrease in Endpoints through Aggressive Lipid Lowering; PROVE‐IT TIMI = Pravastatin or Atorvastatin in Evaluation and Infection Therapy–Thrombolysis in Myocardial Infarction; SD = standard deviation; SEARCH = Study of the Effectiveness of Additional Reductions in Cholesterol and Homocysteine; TNT = Treating to New Targets. * Statin doses expressed as equivalent atorvastatin dose, based on mean low‐density lipoprotein‐lowering in a meta‐analysis (Law et al13). †None of the differences in major CHD events and mortality (total and CHD) were significantly different, except major CHD events in TNT (P = 0.002).

Simon B Dimmitt · Jennifer H Martin

To the Editor: We read with interest the recent review by Adie and colleagues1 discussing the optimisation of patients planned for total knee arthroplasty (TKA) and the prevention of adverse post‐operative outcomes. However, anaemia, which is a common, major and modifiable risk factor for this patient cohort,2 was not addressed by the authors. A recent observational study found pre‐operative anaemia — defined as haemoglobin concentration < 130 g/L — in 32% of patients undergoing elective TKA, and, based on the results for the orthopaedic surgical cohort from this study, it is likely that most of these patients were iron depleted.3 A 2017 single‐centre retrospective study similarly reported anaemia in 24% of a similar cohort of patients, and found it was associated with an increased risk of hospital stay in excess of 6 days (unadjusted odds ratio [OR], 1.97; 95% confidence interval [CI], 1.53–2.53; P < 0.001), which escalated further and proportionally with decreasing haemoglobin cconcentration.4 In another observational study, anaemia was associated with an increased risk of post‐operative complications in aseptic revision joint arthroplasty (OR, 1.45; 95% CI, 1.24–1.70; P < 0.001), mortality (OR, 2.18; 95% CI, 1.09–4.36; P = 0.028), and increased hospital length of stay (adjusted coefficient, 1.02 days; 95% CI, 0.73–1.31; P < 0.001).5 Perhaps the most compelling argument for the recognition of anaemia in the pre‐operative TKA patient is that it is a risk factor that can be addressed even as little as 2–5 days pre‐operatively. A retrospective review of ultrashort term parenteral iron supplementation in major orthopaedic surgery suggested intravenous iron reduced allogeneic blood transfusion rates (8.9% v 30.1%; P = 0.001), which are a likely contributor to post‐operative outcome. Intravenous iron was also associated with reduced length of hospital stay (8.4 days v 10.7 days; P = 0.001).6 Newer parenteral iron preparations are widely used as pre‐operative rehabilitative interventions, are characterised by their ease of administration and favourable side‐effect profile, and are recommended as part of best practice patient blood management guidelines.7 At present, large scale prospective and randomised data assessing the safety and efficacy of iron supplementation for correction of anaemia before elective TKA remain lacking. The importance of this area and the potential capacity for clinical and economic effect warrant both further prospective research and consideration by clinicians practising perioperative medicine.

Lachlan F Miles · Kate L Burbury · David A Story

To the Editor: The period from January to March marks the peak season for travellers returning to Australia, and typhoid is a key illness of concern. Since 2016, an extensively drug‐resistant (XDR) typhoid clade has emerged in Pakistan, showing resistance to all first‐line agents.1,2 Over the past 2 years, seven cases have been reported in returned travellers — mostly children — from Pakistan to England, Germany and the United States.1,3,4 We report here the first case of XDR typhoid identified in Australia. A 20‐month‐old Australian‐born girl presented to the Children's Hospital at Westmead with features of enteric fever 14 days after her return from a 3‐month trip to Karachi. Diarrhoea began in Pakistan 6 weeks before her return, and continued despite the use of oral antibiotics prescribed locally in Pakistan. Ten days of high fevers, irritability, vomiting and reduced oral intake prompted admission to our hospital. She was a previously well child and had received routine vaccines. No additional pre‐travel vaccinations had been recommended. Blood and stool cultures grew extended spectrum β‐lactamase‐producing Salmonella enterica serovar Typhi. The isolate showed microbiological features typical for the XDR clade, with resistance to chloramphenicol, ampicillin, trimethoprim–sulfamethoxazole, fluoroquinolones and third‐generation cephalosporins (Box). The isolate was susceptible to meropenem and had an azithromycin minimum inhibitory concentration of 12 mg/L. The child responded to intravenous meropenem and oral azithromycin and was discharged after an uncomplicated 8‐day admission to complete a further week of azithromycin. Notification to the local public health unit facilitated contact tracing. This case was one of 12 travel‐associated, culture‐positive enteric fever cases managed at our hospital in the first 3 months of 2019. This continues a trend toward a higher incidence of enteric fever, as previously reported at our hospital from 2003 to 2015.5 This case highlights the emerging threat of XDR typhoid and the broader global issue of escalating antimicrobial resistance, to which Australia is not immune, especially given increasing travel connectivity. Typhoid must be considered as a diagnosis for febrile returned travellers from endemic regions, including South and South‐East Asia. Typhoid vaccination is recommended from 2 years of age if travel is planned to these regions. The important role of general practitioners in providing travel‐related vaccine advice and care to returning travellers must not be underestimated. Box – European Committee on Antimicrobial Susceptibility Testing (EUCAST) disc diffusion demonstrating zones of inhibition to cefotaxime (COX), amoxicillin–clavulanate (AMC), cefepime (FEP), meropenem (MEM), ampicillin (AMP), imipenem (IMP), ciprofloxacin (CIP), amikacin (AKN), piperacillin–tazobactam (PTZ), fosfomycin (FOS); trimethoprim–sulfamethoxazole (SXT) and gentamicin (GMN)

Annaleise Howard‐Jones · Alison M Kesson · Alexander C Outhred · Philip N Britton

Over two decades of Dutch experience can inform deliberations about the nature of a regulatory framework in Australian jurisdictions

Vicky Sheppeard · Robin Gilmour · Sean Tobin

To the Editor: We read with interest the article regarding Australia's approach to managing the challenges of pharmaceutical opioid utilisation.1 The various regulatory, service delivery and educational activities described would appear to provide a comprehensive response to this problem. However, we were surprised to note the absence of any reference to the potential of pharmacogenomics in identifying patients at increased risk of opioid toxicity. Various factors dictate how much opioid reaches the brain, how long it stays there, and how sensitive a patient may be to central nervous system (CNS) depression. Many risk factors for opioid toxicity are familiar to doctors. These include high daily opioid dose, extremes of age (neonates and older patients), comorbid conditions (eg, liver, kidney and respiratory disease), concomitant CNS depressants (sedatives and alcohol), and strong inhibitors or inducers of metabolism. In contrast, pharmacogenomics is unfamiliar to many Australian doctors. A recent position statement from major medical colleges describes a coordinated effort by doctors to develop the role of pharmacogenomics in Australian clinical practice.2 Codeine is converted to the active metabolite morphine by the cytochrome P450 enzyme encoded by the gene CYP2D6. Due to variants in CYP2D6, some patients are poor metabolisers (5–10%), have low CNS exposure to morphine, and are unlikely to benefit from or be harmed by codeine. Conversely, other patients are ultrarapid metabolisers of codeine (up to 10%), resulting in high and prolonged CNS exposure to morphine and an increased risk of CNS depression.3 There is an ongoing debate about the place of pharmacogenomics in managing the prescribing of opioids. There is strong support for testing children and nursing mothers taking codeine due to fatal cases of respiratory failure in ultrarapid metabolisers.4 International clinical guidelines recommend that codeine be avoided in poor metabolisers and ultrarapid metabolisers because of lack of efficacy and risk of CNS depression, respectively.3 Considerable research is underway to determine whether pre‐emptive pharmacogenomic testing can help with the pharmaceutical opioid problem.5 To reduce harm from opioids, we recommend testing for CYP2D6 variants as part of a benefit–risk assessment when prescribing codeine, especially for patients with other risk factors for opioid toxicity.

Thomas Polasek · Melody Caramins · Graeme Suthers

To the Editor: Sodium–glucose cotransporter type 2 (SGLT2) inhibitors — dapagliflozin, empagliflozin and now ertugliflozin — have become established second line options for type 2 diabetes, with favourable potential for weight loss and cardiovascular protection.1 However, it soon became clear post‐marketing that they had potential for several pronounced side effects, including euglycaemic ketoacidosis — an unusual form of diabetic ketoacidosis where blood sugar levels remained relatively normal.2 The Therapeutic Goods Administration (TGA) first sent an alert about euglycaemic ketoacidosis in relation to SGLT2 inhibitors in 2015; subsequent alerts in 2018 from the TGA and the Australian Diabetes Society warned specifically about periprocedural risks.3,4 Austin Health has a well developed culture of adverse drug reaction reporting. A multidisciplinary committee includes representation from pharmacy, clinical pharmacology, dermatology and infectious diseases. During 2018, our adverse drug reaction committee forwarded 302 reports to the TGA, estimated to be around 15% of all reports received from Australian hospitals. Since 2016, our adverse drug reaction committee has received 12 reports of patients with diabetic ketoacidosis related to SGLT2 inhibitors, including eight in 2018. The growth in incidence locally in such a short period is alarming. Most patients (75%) had a blood sugar level of 11 mmol/L or lower at presentation. Our committee reviewed the cases in the Box to evaluate severity and causality. SGLT2 inhibitors were considered a probable cause in ten cases; the reaction was considered severe in nine cases, with one death during admission. We report our cases with the aim of increasing awareness around contributing factors, particularly concurrent illness resulting in poor oral intake. Only two of the 12 cases related to a perioperative setting, and in neither situation was the SGLT2 inhibitor withheld prior to surgery. We remind clinicians that the precipitants for diabetic ketoacidosis extend beyond the perioperative period. We advise caution when patients are experiencing other contributing factors illustrated by our case series, including acute illness, reducing insulin doses, poor oral intake, severe dehydration and low carbohydrate diet. Patients should be counselled about the signs of ketoacidosis and advised to seek medical help if they occur. SGLT2 inhibitors should be withheld if a patient is acutely unwell or undergoing surgery, and should only be restarted when the patient is eating and drinking normally.5 Box – Cases of ketoacidosis related to sodium–glucose cotransporter type 2 inhibitors Case Year Medication Dose Severity Causality Potential contributing factors 1 2016 Empagliflozin 10 mg daily Moderate Probable Low dietary intake in perioperative setting 2 2016 Dapagliflozin 5 mg twice a day Severe Probable Perioperative setting 3 2017 Dapagliflozin 10 mg daily Severe Probable Unwell for 3 days prior to presentation — patient had type 1 diabetes 4 2017 Empagliflozin 10 mg daily Severe Possible Concurrent influenza 5 2018 Empagliflozin 10 mg daily Severe Probable Narcosis leading to poor oral intake 6 2018 Empagliflozin 12.5 mg twice a day Severe Probable Weight loss since commencing — worse in the month prior to admission 7 2018 Empagliflozin 25 mg daily Moderate Probable Concurrent pneumonia 8 2018 Dapagliflozin 10 mg daily Moderate Possible Low carbohydrate diet 9 2018 Empagliflozin 10 mg daily Severe Probable Patient unwell with some vomiting for several days before admission 10 2018 Dapagliflozin 5 mg twice a day Severe Probable 5–7 days of loss of appetite 11 2018 Dapagliflozin 10 mg daily Severe (died during admission) Probable Illness for 10 days before admission Pancreatitis 12 2018 Empagliflozin 25 mg daily Severe Probable 5 days of gastroenteritis before admission Weaning insulin doses

Gina McLachlan · Claire Keith · Albert Frauman