Letters

To the Editor:We thank Rakhra and colleagues,1 whose report highlighted a group of potential donors yet to be fully exploited for organ transplantation. While this untapped potential appears to largely affect kidney and lung donor pools, we must not exclude the effect that relaxing donor suitability criteria may have on the heart. While not largely emphasised, the data collected for the heart are compelling, particularly for an organ so susceptible to ischaemic injury. Similar to the kidney and lungs, the data show that when including ideal and expanded criteria, potential donors more than doubled the number of donor hearts available for heart transplantation.1 This is an important finding, particularly if the authors have conservatively underestimated the actual potential for organ donation. In Australia, donation after circulatory death (DCD) kidney and lung donors comprise 25–30% of the total donor pool for these organs.2,3 Since 2014, the number of national and international hospitals now investigating DCD heart donation has increased, and we anticipate that DCD heart transplantation will follow the same trend as DCD lung and kidney transplant over time. The Critical Care Research Group at the Prince Charles Hospital has developed a clinically relevant 24-hour ovine model of brain stem death,4 and these donor hearts are being used to transplant into recipient animals. Transplant clinicians and researchers from the Prince Charles Hospital, the Alfred Hospital and St Vincent’s Hospital will collaboratively employ an ex vivo hypothermic perfusion system, capable of safely storing animal hearts up to 24 hours,5 together with novel cardioprotective therapeutics to increase the allowable ischaemic times for donor hearts without impairing graft function. This technology will hopefully extend to DCD donors in the future, which are currently only preserved using a warm blood perfusion system. By relaxing donor heart suitability criteria, and using novel machine perfusion technology to increase the storage time of the donor heart,5 thus eliminating time and travel distance constraints in Australia, the donor heart pool could be greatly increased. While the numbers may ultimately be modest, the positive effect on those recipients would be substantial.

Louise E See Hoe · David McGiffin · John F Fraser

To the Editor:In October 2016, we achieved the eradication and control of hepatitis C virus (HCV) in a New South Wales correctional centre, which we believe to be a first of its kind in NSW. HCV prevalence in NSW prisons is 30–40 times higher than in the community, where prevalence is about 1%.1,2 Elevated risk of HCV infection is associated with the high proportion of prisoners who have injected drugs, the rate of injecting in prison, and restricted or limited access to bleach and needle and syringe programs.3 The Justice Health and Forensic Mental Health Network (the Network) is responsible for health care in the NSW forensic mental health and criminal justice systems. The availability of direct-acting antivirals on the Pharmaceutical Benefits Scheme in March 20164 created an opportunity for the Network to cure all patients with HCV infection in one prison. The Compulsory Drug Treatment Program (CDTP) is run at the Compulsory Drug Treatment Correctional Centre — a stand-alone prison with a stable sentenced inmate population, where patients with repeat drug-related charges participate in comprehensive drug treatment and rehabilitation. Patients at this correctional centre have longer sentences than those in other centres, which allowed for the full course of treatment. The Network, Corrective Services NSW and Hepatitis NSW formed a partnership to ensure that patients were able to access health centres, have their medication scripted and administered, could undertake monitoring and were supported through the Network’s established nurse-led model of care.5 All 58 patients in the CDTP were offered screening, and 54 patients with risk factors were screened; of these, 18 patients had chronic HCV infection. After further work-up, including liver elastography to measure fibrosis, all patients were concurrently commenced on treatment. Of the remaining four patients who were not screened, all had recent negative HCV pathology results. Three months after the treatment, 15 patients achieved sustained virological response equating to cure of their chronic HCV infection, and three patients were released before final assessment.6 Concurrent treatment commencement with the direct-acting antivirals was recognised as an innovative measure in reducing re-infection, in conjunction with the more common practices of harm minimisation education and use of the hospital-grade disinfectant for general cleaning purposes offered by Corrective Services NSW to all incarcerated people. Throughout the course of the project, two new patients were admitted, screened, and returned negative HCV pathology results. A proactive screening approach with patient consent was adopted to ensure that new cases were able to be identified and treated to maintain elimination. Maintaining a prison HCV-free may mean that patients have to take some responsibility with regard to sharing needles with new inmates. A peer education approach is being developed to increase patients’ ownership of a prison’s HCV-free status. The CDTP treatment model, combined with ongoing screening of new admissions, is an innovative approach for eliminating HCV, and is considered suitable for adoption in similar-sized prisons across Australia. The Network is currently rolling out this approach within NSW.

James Blogg · James Wood · Colette McGrath · Camilla Lobo

To the Editor:Your publication of a review on global approaches to antimicrobial resistance is timely.1 We especially note that antibiotic-resistant pathogens are not limited by borders, have greater impact on disadvantaged communities, and will require coordinated, high level government commitment to minimise their threat.1 In Australia, Indigenous communities bear a disproportionate burden of infectious diseases. This burden arises on a background of overcrowding, poorly built and maintained water and sanitation infrastructure, and colonisation of companion animals by human pathogens. The delivery of biomedically oriented health services leads to frequent use of broad spectrum antibiotics, promoting the development of multiresistant pathogens.2 The prominent multiresistant pathogen methicillin-resistant Staphylococcus aureus first emerged in hospitals, but, in Australia, it was soon identified in remote Indigenous communities.2 Health services have been unable to control its development and spread. As a consequence, community-acquired methicillin-resistant S. aureus is now the dominant strain of this bacterium in Central Australia, where Indigenous people are one-quarter of the population, but bear three-quarters of the S. aureus disease burden in Alice Springs Hospital.3 Primary health care is founded on full community participation and an intersectoral approach, incorporating education, housing and other sectors to complement health services.4 Housing for Indigenous communities remains inadequate, and government responses deficient, particularly in remote regions.5 As a result, even high quality health services have limited impact on Indigenous people’s health and wellbeing. Safe, secure, functioning housing that is appropriate for its occupants is a building block to manage other areas of Indigenous disadvantage.5 The deficit in appropriate housing contributes to bacterial colonisation, infection and development of antimicrobial resistance among Indigenous Australians.2 “Illness is a weapon” was intended as a metaphor for the resistance of Indigenous people to their ongoing colonisation.6 However, the threat of antibiotic resistance evolving through the neglected conditions in which some communities find themselves could make this metaphor more real than was likely intended. The spread of antibiotic-resistant pathogens in Indigenous communities and elsewhere is a global threat, which highlights the need to transform services for Indigenous people using approaches driven by communities and focused on their strengths.

Rosalie Schultz

To the Editor: Cullen and colleagues are to be congratulated on their most recent contribution to the assessment of emergency department patients presenting with possible acute coronary syndrome.1 The ability to safely reduce length of stay for a larger proportion of patients compared with the ADAPT study, pending external validation, is promising.1 The timing of the article also presents an invaluable opportunity to remind all clinicians that despite the progressive improvements in accelerated diagnostic pathways for chest pain, not all chest pain is cardiac. Accelerated diagnostic pathways, when used incorrectly, run the risk of introducing cognitive dispositions to respond such as availability bias (where a clinician may choose a diagnosis that is more familiar, such as acute coronary syndrome), omission bias (doing something easy and omitting something hard, such as ordering serial troponins instead of a computed tomography aortogram) and Sutton’s slip (settling on the most obvious interpretation of the problem).2 The potential for error is further compounded by error-producing conditions inherent in busy emergency departments, such as overcrowding, narrow time windows for assessment, surge phenomena and multiple transitions of care (eg, movement to short stay).3 It is pleasing to see that additional features were utilised in stratifying patients compared with previous accelerated diagnostic pathways.1 High-risk features such as prolonged duration and recurrence of chest pain as well as syncope may also be present in aortic dissection.4 It is possible that a patient presenting with an aortic dissection could be deemed low risk with criteria such as those in the ADAPT accelerated diagnostic pathway.5 Troponin levels do not provide a safeguard, as evidenced in a previous case where a 75-year-old woman was discharged home after two normal troponin test results only to die later the same day from progression of a type A aortic dissection.4 Aortic dissection is a lethal cardiovascular emergency that boasts significant morbidity and mortality.4 It has been astutely described as the “subarachnoid haemorrhage of chest pain”.4 Although it is rare and sometimes difficult to diagnose compared with acute coronary syndrome, clinicians are urged to include aortic dissection in their initial assessment of chest pain before implementing an accelerated diagnostic pathway.

Joe-Anthony Rotella

To the Editor: We read with interest Mendelson and colleagues’ article regarding the underutilisation of computed tomography colonography (CTC) for colorectal cancer detection in Australia.1 The authors state that “CTC is less accurate in the diagnosis of small or diminutive polyps … However, in the context of symptomatic patients, this is not relevant”.1 This overlooks the importance of detecting small adenomas as well as flat, right-sided colonic lesions such as sessile serrated polyps. It is well established that the early detection and treatment of these lesions reduces interval colorectal cancer.2 To state that these are “not relevant” in symptomatic patients is inaccurate. High definition white light with the aid of chromoendoscopy tools such as narrow band imaging in optical colonoscopy (OC) has significantly improved the ability of endoscopists to detect diminutive and subtle lesions. CTC has a markedly limited ability to detect small and flat lesions when compared with OC.3 The authors reference phase 2 data from the National Bowel Cancer Screening Program, stating that “the great majority of OCs (about nine in ten) were normal”. However, polyps were detected in 34.9% of participants.4 The authors appear to define normality as the absence of a cancer or advanced adenoma (one in ten colonoscopies). We argue that the detection of any adenoma is important and is not “normal”, as it highlights a cohort of patients at risk of colorectal cancer and requiring ongoing polyp surveillance. We also question the authors’ suggestion that a negative CTC obviates the need for OC in lower risk patients with positive faecal immunochemical test results. OC provides the opportunity to detect and treat a range of polypoid and non-polypoid colonic pathology and although a negative CTC may help exclude a cancer, it does not address the range of other potentially morbid or precancerous causes of occult faecal blood. Given these limitations, a cautious approach should be adopted if CTC is to be used as an alternative to OC in all lower risk patients.

Simon Hew · Zaid SM Ardalan

In reply

Richard M Mendelson · Tom R Sutherland · Andrew F Little

To the Editor: We commend Cohen and colleagues1 on their recently published study, which is the largest randomised controlled trial (RCT) of acupuncture in the emergency department (ED). We recently completed a systematic review and meta-analysis on the role of acupuncture for analgesia in the emergency setting.2 Our meta-analysis incorporated 19 RCTs and included non-English language publications. The trial by Cohen and colleagues1 was not published at the time of our review; however, it strengthens our main conclusion that acupuncture was non-inferior to standard analgesia in the emergency setting. We also found similar evidence of improved patient satisfaction. It was interesting that the authors reported an adverse effects rate of 51% for acupuncture, whereas our study found an overall rate of 5%, with significant adverse effects being 1%. Our figures are consistent with other reviews3 and almost certainly highlight the difficulties in developing agreed definitions on adverse effects in acupuncture. Our review found that acupuncture in two out of four RCTs decreased pain medication requirements, whereas Cohen and colleagues’ study had the potential to inform this outcome, but did not report such data. The study by Cohen and colleagues1 illustrates many of the challenges in acupuncture trials, including having no sham comparator group. Some acutely painful conditions might resolve simply because of time or careful patient attention. Sham acupuncture is difficult to deliver as a control4 and needs to be plausible, realistic and, if possible, blinded. Our meta-analysis showed acupuncture to be superior, with clinically significant reductions in acute pain scores compared with sham. This latest significant RCT gives further impetus to carefully designed research on acupuncture in the emergency setting, which will require acupuncture techniques applicable to the time-constrained ED environment (eg, ear acupuncture), provision of a suitable sham acupuncture technique, and minimisation of assessment bias. We suggest that the specific outcomes to be assessed should include the impact of acupuncture as an adjunct to standard analgesia, side effects recorded using standard definitions, and reductions in medication use. Most importantly, the analgesic effect of acupuncture is unlikely to be equal for all pain presentations in the emergency setting and, therefore, the conditions for which its role is most beneficial need to be delineated.

Andrew L Jan · Ian Rogers · Eric J Visser

To the Editor: The conclusions made by Cohen and colleagues1 can be challenged on the basis of study design and results. A significant flaw in the design is that of all patients not excluded due to additional medical problems, 38% of potential study participants declined to be included, possibly because they were not prepared to be assigned to the acupuncture-only arm. This flaw introduces a critical bias when analysing the 62% who did participate, as it is reasonable to assume that they are statistically more inclined to experience a placebo benefit from acupuncture and more likely to be positive at 48 hours about repeating this therapy. Even with this bias, however, the results of the study do not support a broad interpretation of non-inferiority, given that 39% of acupuncture-only patients required rescue therapy at or after one hour, compared with 22.5% of pharmacotherapy-only patients. In the context of an emergency department (ED) and patients with mean verbal numerical rating scale score of 8.5 at presentation, this difference in pain reduction and satisfaction with the initial therapy during the first hour of treatment should render the acupuncture-only option unacceptable. The authors have selectively and speculatively interpreted their data to support their enthusiasm for acupuncture. Results for patients with migraine are given minimum discussion or reference in the conclusions. The inferior performance of acupuncture-only patients’ willingness to repeat the therapy at one hour is dismissed as being influenced by patients’ concerns about length of stay and transport arrangements, whereas they indicate that the same parameter at 48 hours is more meaningful. I suggest that the results at one hour more accurately reflect a patient’s experience of their pain and desire for abatement, whereas the improved satisfaction at 48 hours is the result of shifted perspective following self-resolution — in a selected patient population with an existing positive bias for acupuncture. The authors’ suggestion that acupuncture use in EDs would reduce opioid addiction is an opportunistic grab at exploiting popular sentiment among people who do not understand the causes and parameters of this serious problem. This article does not support diversion of resources towards acupuncture use in EDs or a change in evidence-based treatment protocols in emergency medicine.

Anitra J Wenden

In reply

Marc M Cohen · Michael Ben-Meir · Nick Andrianopoulos

The following study was conducted independently of the MJA. The MJA has not verified the accuracy of the data

Yassar Alamri · Mohamud Osman

To the Editor: After reading the recent article by van Driel and colleagues,1 we want to report on hepatitis C treatment outcomes in a Sydney general practice. New direct-acting antiviral (DAA) therapy for the treatment of chronic hepatitis C became widely available in Australia on 1 March 2016 via the Pharmaceutical Benefits Scheme, and general practitioners are able to prescribe it in consultation with a specialist. We describe here the treatment outcomes from the first 60 days of DAA prescribing in a single general practice clinic. We searched the clinic database to extract demographic and clinical data for all patients prescribed DAA from 1 March to 30 April 2016. We found that 47 patients had been prescribed DAA agents by a GP, five had received DAA therapy via an early access program, 41 had genotype 1 hepatitis C virus (HCV), six patients had genotype 3 HCV, and 11 patients had co-infection with HIV. All treated patients had an assessment of liver fibrosis performed with a FibroScan. Most patients had early liver disease, with three having cirrhosis. On 1 May 2017, we assessed the outcome data: 33 patients were treated with ledipasvir and sofosbuvir (all had genotype 1 HCV), nine were treated with daclatasvir and sofosbuvir (three had genotype 1 HCV, and six had genotype 3 HCV), and five patients were treated with paritaprevir + ritonavir + ombitasvir + dasabuvir (all had genotype 1 HCV). Forty-six patients had started treatment, with sustained virological response (SVR16; ie, undetectable virus 16 weeks after the end of treatment) results available for 45 patients. One patient had treatment failure, one had not started treatment and one patient was waiting on SVR results. On-treatment SVR (cure) rate was 96%. A steadily increasing percentage of patients in Australia are receiving hepatitis C treatment prescribed by their GPs, with 19% of prescriptions provided in this setting in September 2016.2 GPs are well placed to provide care for patients living with chronic hepatitis C, with reassuringly high cure rates.

David Baker · Marilyn McMurchie · Vanessa Farr

To the Editor: I read with some concern the study by Ibrahim and colleagues1 published in the Journal. In their epidemiological study of deaths of nursing home residents reported to the coroner, the authors reported a more than quadrupling of the rate of deaths due to falls over the decade from 2002 to 2012, and deemed these deaths to be preventable and premature. In their discussion, Ibrahim and colleagues1 insinuate that this increased rate of deaths due to falls relates to some extent to a decline in the standard of care provided to nursing home residents. As this study lacks a control group, fundamentally, there is a problem in making any assumptions about preventable deaths and care provided to nursing home residents. This problem is particularly pertinent in light of data published by the Australian Institute of Health and Welfare which show an almost 40% increase in the rate of injury hospitalisations for older Australians, predominantly due to falls, over a similar 10-year period.2 A subsequent publication by the Australian Institute of Health and Welfare reports that 50% of all falls in older people resulting in hospitalisation occur in or around the home and that only 22% occur in an aged care facility.3 Therefore, one could ask whether there has been a similar increase in the rate of premature and preventable deaths due to falls of older Australians who live in their own homes. Clearly, further research on falls in general is required before any conclusions about possible causes can be made.

Henry Zeimer

To the Editor: Ibrahim and colleagues1 are to be congratulated for finding a source of information that throws some light on what is happening in Australian aged care facilities, because there is so much positive rhetoric and so little reliable data about aged care coming from the sector itself. The study by Ibrahim and colleagues1 reveals an increase in deaths from external causes, including falls and choking, but we need context. In his 1993–94 report, Gregory2 indicated that in the proposed market-driven aged care system “neither the current standards monitoring system, nor any alternatives considered, would be able to prevent the diversion of funding from nursing and personal care to profit”. However, the government policy is driving consolidation and corporatisation using a competitive profit-driven model. International data indicate that an increased focus on profitability in aged care is associated with poorer staffing and increased failures in care.3 Studies in Australia and the United States over the past 35 years indicate that, on average, 4 hours or more per person per day of nursing care are required for safe care. Available data indicate that, in Australia, only 2.8 hours are provided. As acuity has increased, the proportion of trained staff has fallen. Our residents receive less than half the nursing time from registered and trained nurses compared with the US and an hour less of total nursing care each day.3 Safe care cannot be provided with these staffing levels; yet, over 95% of facilities are accredited by the Australian Aged Care Quality Agency. Braithwaite and colleagues,4 who have studied the regulation of aged care, conclude that aged care regulation has been captured by the market and is ineffective and warn that “the community should be concerned”. While politicians and the industry talk up our system, it is becoming increasingly clear that the many failures (often in recently accredited facilities) reported in the press are red flags to systemic problems that a succession of captured inquiries into a system that is anything but world class have avoided addressing. With some exceptions, our profession seems to have disengaged from the sector and one wonders if they have been captured too.

J Michael Wynne

To the Editor: The study published by Ibrahim and colleagues1 raises questions about the quality of care delivered in residential aged care facilities. This analysis of routinely recorded coronial data identified that 15% of premature and preventable residents’ deaths resulted from external causes, with falls being the most frequent culprit. Falls in residential aged care facilities are well recognised, with about half of all residents falling within a given year.2 Residents tend to be older and frailer, have higher rates of dementia and higher rates of psychotropic medication use compared with community-dwelling older people, which contributes to higher rates of falls. A falls risk assessment, using validated tools and appropriately qualified personnel, has the potential to reduce the rate of falls and related injuries by addressing individual and environmental risk factors.3 While the residential aged care sector is subject to variation in staffing numbers and skill mix, mandatory education related to falls prevention is also needed. The Australian Commission on Safety and Quality in Health Care’s Preventing falls and harm from falls in older people — best practice guidelines for residential aged care facilities4 provide a helpful framework. In a separate study, Ibrahim, the lead author of the MJA article, and Davis5 state that residents in aged care facilities are entitled to the “dignity of risk” principles that allow them autonomy to accept risks that may be associated with short term increases in their quality of life. However, the reality is that a person’s quality of life is often determined by risk management rather than autonomy,6 which reflects a form of age discrimination whereby the risk of injury of falls and its consequences outweighs the older person’s quality of life, by making assumptions about their ability to make choices about their everyday activities. Promoting dignity and autonomy for older residents in aged care may see greater falls rates but it will allow residents to enact their choices. The work by Ibrahim and colleagues1 needs to be understood within this context.

Judy A Lowthian · Claudia Meyer · Dianne Goeman · Colette Browning

In reply

Joseph E Ibrahim · David L Ranson · Lyndal Bugeja

To the Editor:The recent tragic suicides of young doctors have highlighted concerns regarding the welfare of trainees in our profession. The Trainee Committee of the College of Intensive Care Medicine of Australian and New Zealand (CICM) met in Melbourne in March 2017, and the recent results of the survey1 into bullying and harassment, in addition to the deaths of several junior medical officers to suicide, provided for a solemn meeting. The committee wants to highlight the factors that adversely affect intensive care medicine trainees: bullying, discrimination and sexual harassment;1 staffing and intensity as, while patients’ needs are more complex, staffing has not increased with this intensity of practice; rapid response teams, which are often under-resourced, poorly trained and undersupported; and poor workforce planning and tenuous future job security, compelling junior doctors to pursue increased non-clinical commitments without an allocated time to do so. A consequence of these factors is burnout, which disproportionately affects intensivists and is an increasingly significant risk in trainees.2 Changes in work practices, severity of illness, increased demand for limited resources and increasing intensive care unit size — problems that have recently been addressed by Corke3 — have all played a part in burnout rates. The Trainee Committee welcomes the approach of the CICM to tackle these issues.4 The college is steadfast in its zero tolerance to bullying, discrimination and harassment, and remains committed to fair and equitable access to training. The college also values the need for a reasonable balance between provision of clinical services and time for professional development, and recognises the importance of work–life balance, including part-time training and the provision of parental and other forms of leave. However, the CICM has limited ability to enforce these needs at hospital level. Finally, the CICM embraces rapid response teams, recognising the importance of an appropriately supported service, but hospitals need to respond to this need, and lack of planning and matching training with lifetime workforce demands have to be a priority for the government at all levels. Our specialty will change significantly in the next decade or so. We chose intensive care because we enjoy the work we do and find the challenge it provides rewarding. We must ensure that the next generation of intensivists can meet this challenge too.

Alun T Ellis · Sandra Lussier · Sarah A Yong

To the Editor: I found the article by Kiely and colleagues1 very interesting. The authors concluded that “it should be noted that a relatively small number of imported ZIKV [Zika virus] infections have been reported in Australia, there have been no reported cases of local ZIKV transmission, and the geographical distribution of the potential ZIKV mosquito vector in Australia (Aedes aegypti) is limited to northern Queensland,” and that “at present, ZIKV represents a low risk to blood safety in Australia.”1 Indeed, Kiely and colleagues may be correct in their statement. Nevertheless, without supportive evidence, it seems too soon to draw a conclusion. The estimation of the risk or possibility of transfusion-transmitted ZIKV infection in each setting may be based on mathematical modelling with reference to risk of transmission of other arboviruses in that setting. A good example is the previous report on the estimated risk of transfusion-transmitted ZIKV infection in Thailand.2 However, the lack of reports on ZIKV transmission does not mean that the problem does not exist; many patients with ZIKV infection are asymptomatic and can be easily missed.3

Viroj Wiwanitkit

In reply

Philip Kiely · Erica M Wood · Allen C Cheng

To the Editor: Retinal screening is crucial to the prevention of vision loss from diabetic retinopathy. The recent National Eye Health Survey highlighted a gap in screening rates between Indigenous Australians (aged ≥ 40 years) and non-Indigenous Australians (aged ≥ 50 years),1 providing a foundation on which to target future eye health services. While acknowledging the budgetary and logistical constraints of such a large-scale study, we are concerned that another at-risk group — younger adults with type 2 diabetes who are aged 18–49 years — was not included in the National Eye Health Survey. The past two decades have seen a global increase in the incidence of type 2 diabetes in youth and young adults, with younger age being an independent risk factor for the development of diabetic retinopathy.2 In Australia, about 130 000 people with type 2 diabetes who are aged < 50 years are registered on the National Diabetes Services Scheme.3 However, there are no population-level data available regarding retinal screening rates in this age group. A decade ago, a survey of young Australian adults with type 2 diabetes (aged 16–35 years) reported a 55% retinal screening rate.4 Given the small, self-selected sample, this is likely an overestimate and compares unfavourably with the non-Indigenous rate of 78% reported for adults aged ≥ 50 years in the National Eye Health Survey.1Moreover, our qualitative research has shown that young adults aged 18–39 years who have type 2 diabetes face different psychosocial challenges and barriers to retinal screening compared with their older counterparts aged ≥ 40 years.5 Younger adults with type 2 diabetes require targeted, tailored intervention, which can only be provided if accurate, population-level data are available for this group. As Foreman and colleagues1 acknowledge, diabetic retinopathy is the leading cause of vision loss in working age adults — a situation with potential for considerable social and economic burden. The absence of current eye examination data for younger Australians with type 2 diabetes increases the risk that their needs will be neglected in future evidence-based policy and program delivery initiatives. We urge the Australian Government and other stakeholders to extend future population-level surveys (and other research and policy initiatives) to include all adults with diabetes.

Amelia J Lake · Jessica L Browne · Jane Speight

In reply

Joshua Foreman · Stuart Keel · Mohamed Dirani

To the Editor: We read with interest the recent article by Bonevski and colleagues1 calling for targeting of tobacco cessation interventions to high-risk populations, including prisoners. People who cycle through prisons in Australia smoke tobacco at a rate five times that of the general population,2 and suffer disproportionately from smoking-related morbidity and mortality.3 However, the suggestion by Bonevski and colleagues1 that smoke-free policies in prisons “impact on reducing smoking” is unfortunately a case of misplaced optimism: although these policies reduce smoking in prisons, they have almost no effect on long term smoking behaviour in people who cycle through prisons. Research in the United States shows that about 60% of people released from smoke-free prisons resume smoking on the day of release,4 and 97% relapse within 6 months of release.5 Preliminary findings from a cross-sectional survey we have conducted with 114 ex-smokers released from smoke-free prisons in Queensland paint a similar picture, with 72% of participants reporting relapse on the day of release. Smoke-free policies in Australian prisons are an important public health initiative and should be supported. However, alone they are insufficient to reduce the remarkably high rates of smoking, and of related morbidity and mortality, in the vulnerable populations who cycle through these institutions. There is an urgent need for development and rigorous evaluation of smoking cessation and relapse prevention interventions targeting people released from prison in Australia. Building on the findings of a recent trial in the United States,4 we have recently been awarded funding from the Victorian Health Promotion Foundation to undertake a double-blinded, randomised controlled trial of an intervention designed to reduce relapse to smoking among people released from smoke-free prisons in Victoria. We hope that our study will provide new evidence to guide future efforts to reduce tobacco-related harm in this population. We echo Bonevski and colleagues’1 call for a comprehensive policy shift aimed at reducing tobacco use among disadvantaged populations in Australia. However, prison smoking bans alone are insufficient. Investment in evidence-based efforts to prevent smoking relapse after release from prison will be critical to reduce tobacco-related health disparities in this profoundly vulnerable, marginalised population and realise this important public health opportunity.

Cheneal Puljević · Stuart A Kinner

In reply

Billie Bonevski · David P Thomas · Robyn L Richmond