Volume 212 Issue 6

We need informative citation metrics that will be less prone to misuse and gaming

John PA Ioannidis · Kevin W Boyack

Living guidelines that incorporate new evidence as it becomes available have the potential to overcome some of the limitations inherent in static guidelines Diabetes is a complex chronic condition that affects about 1.7 million Australians and represents an estimated $15 billion per annum in direct and indirect costs to the Australian economy.1 Almost $215 million of subsidies were delivered during the 2015–16 financial year to the 1.32 million registrants of the National Diabetes Services Scheme, an Australian Government initiative that provides support to Australians living with diabetes. In 2019, an additional $100 million was announced for funding the Continuous Glucose Monitoring Initiative, which provides fully subsidised continuous glucose monitoring products to patients with diabetes who meet certain criteria.2 In 2017, almost 1.2 million hospitalisations and 11% of all deaths in Australia listed diabetes as the principal or associated cause.3 In addition to the costs associated with diabetes management and prevention, significant funding has been directed towards research into this key priority area, with the National Health and Medical Research Council (NHMRC) providing $375 million from 2013–2018 towards efforts to improve the prevention, diagnosis and management of diabetes.4 With the objective of strengthening diabetes policy and practice, the Australian Government developed the Australian National Diabetes Strategy 2016–2020, which outlines an integrated and coordinated approach for reducing the social, human and economic impact of diabetes.5 One of the key goals within this strategy involves strengthening prevention and care through the use of research, evidence and data. Indeed, developing a nationally endorsed set of diabetes guidelines, assessed against the clinical practice guidelines criteria, was a key recommendation of the Australian National Diabetes Strategy to improve complications and outcomes associated with the disease. Producing new clinical guidelines and implementing a system by which recommendations can be updated and adopted rapidly represents an important means by which this recommendation can be achieved. Clinical guidelines: is there a better way? High quality, evidence‐based clinical guidelines are integral to ensuring that health care decisions are based on the best available evidence. Unfortunately, evidence‐based clinical guideline development is an expensive and laborious undertaking in which several years can pass between inception and publication. In Australia, guidelines approved by the NHMRC are valid for 5 years from publication before they are considered outdated, following which they must be either updated or developed anew.6 These delays can result in several potential problems. First, new research is continually being generated throughout the development period, which may mean that a guideline is outdated before it is even published. Indeed, it has been demonstrated that one in five guideline recommendations are outdated within 3 years of guideline publication.7 Second, institutional memory of the decision‐making processes through which recommendations are derived can be lost, particularly if a significant period of time has transpired since the original guideline was developed. Third, changes in the policy and practice environment can shift priorities or raise new questions that were not considered when defining the original scope, resulting in the guideline failing to address some of the key current issues relating to the topic of interest (eg, the development of a new therapeutic or withdrawal of a technology from the Australian Register of Therapeutic Goods). Currently, all but one of the NHMRC‐approved diabetes clinical guidelines are outdated and have been rescinded. As a result, there is no up‐to‐date Australian guidance for clinicians caring for people with diabetes, potentially resulting in the suboptimal management and significant variation in care of this condition.8 Living guidelines Living guidelines represent an approach to guideline development in which individual recommendations are continually updated as new, relevant evidence becomes available. This is achieved through monthly searches of key databases to identify recently published research. Following analysis of the new data, an impact assessment is conducted to determine whether the evidence is of sufficient relevance, reliability and importance to justify revising recommendations.9 Updated recommendations are then published within a real‐time digital dissemination platform, providing stakeholders with access to the most up‐to‐date version of the guideline. Although the concept of living guidelines is not new, many of the processes employed in developing living guidelines have been generated through Project Transform, an innovative platform established by Cochrane to address the critical issue of evidence currency within clinical guidelines (https://community.cochrane.org/help/tools-and-software/project-transform/about-project-transform). These processes are supported by the development and refinement of machine learning algorithms (eg, randomised controlled trial classifiers), citizen science initiatives (eg, Cochrane Crowd), new methods for updating statistical analyses,10,11 and the development of online collaborative platforms for systematic review and clinical practice guideline production (eg, Covidence, MAGICApp). The application of these tools significantly reduces the workload of systematic review and guideline authors, and appears to result in the production of updated recommendations at a fraction of the resource and time costs otherwise required. In addition, the establishment of a living guideline development group improves the retention of institutional memory throughout the process of updating, and the feedback mechanisms built into the process provide a means by which the underlying scope can be adapted to changes in policy and practice in Australia (Box 1). Living evidence for diabetes Embracing the inherent potential in living guidelines, the Living Evidence for Diabetes Consortium is developing living guidelines that address key priorities relating to diabetes prevention, diagnosis and management (https://livingevidence.org.au/new-index-3#Living-Guidelines-for-Diabetes). Consisting of the Australian Diabetes Society, Diabetes Australia, the Australasian Paediatric Endocrine Group, the Australian Diabetes Educators Association and Cochrane Australia, with representation from the Royal Australian College of General Practitioners and the Australian Government Department of Health, the consortium has selected two proof‐of‐concept topics that fulfil the criteria for living guidelines (Box 2). Two systematic reviews are currently under development to underpin these guidelines, focused on the comparative safety and effectiveness of therapeutics for blood glucose control in adults with type 2 diabetes and the use of technologies (such as insulin pumps and continuous glucose monitors) for the management of type 1 diabetes in adult and paediatric populations. The need for clear guidance relating to these topics is demonstrated by the ongoing uncertainty regarding the most appropriate choice of second line therapies13 and the inception of do‐it‐yourself closed loop systems.14 Although the methods and processes required to produce living guidelines are still evolving, the development of living guidelines for diabetes represents a paradigm shift in the way recommendations are updated and shared with decision makers. Access to this resource should improve the likelihood that patients will consistently receive the best evidence‐based care available, and also provide an avenue through which guideline developers can respond to changes in policy and practice, resulting in guidelines that evolve to keep up with the current practice. Box 1 – Static guideline development (A) versus living guideline development (B) Box 2 – Requirements for converting traditional to living recommendations12 Not all recommendations are suitable for a living evidence approach. Three key requirements should be fulfilled to justify transitioning a static guideline into a living guideline: the guideline should focus on a priority topic for patient, clinical or policy decision‐making; uncertainty should exist regarding the strength and/or direction of recommendations; and there should be a high likelihood of new evidence becoming available in the near future which could increase certainty.

Heath White · Britta Tendal · Julian Elliott · Tari Turner · Sofianos Andrikopoulos · Sophia Zoungas

As a leading general medical journal, the MJA recognises the importance of very rapidly sharing accurate health information

Nicholas J Talley

A 33-year-old male tour guide presented at our clinic with rash and swelling on the anterior upper arms

Minh Cuong Duong · Phuc VD Le · Oanh NK Pham · Hong Quang Huynh

Glasziou PP, Jones MA, Pathirana T, et al. Estimating the magnitude of cancer overdiagnosis in Australia. Med J Aust 2020; 212: 163‐168. https://doi.org/10.5694/mja2.50455 In Box 3 of this article, the overdiagnosis proportion for thyroid cancer should be 73% (as elsewhere in the article), not 33%.

Patient-centred technology, delivered via mobile phone apps, can dramatically improve patient outcomes

Roy G Beran

Some patients may not benefit from inpatient rehabilitation, but numerous factors must be considered

Ian D Cameron

It may soon be possible to reverse the increase in food allergy of the past few decades

Rachel L Peters · Kirsten P Perrett

The potential of modern mass communication techniques for managing chronic disease should be explored further

Yang Si · Xiaoqiang Xiao · Cai Xia · Jiang Guo · Qiukui Hao · Qianning Mo · Yulong Niu · Hongbin Sun

The impact of discharge destination on outcomes after treatment for orthopaedic trauma requires further investigation

Lara A Kimmel · Pamela M Simpson · Anne E Holland · Elton R Edwards · Peter A Cameron · Richard S Steiger · Richard S Page · Raphael Hau · Andrew Bucknill · Jessica Kasza · Belinda J Gabbe

SmartStartAllergy faccilitates monitoring of infant feeding practices in primary care and parent-reported reactions to food

Michael O'Sullivan · Sandra Vale · Richard KS Loh · Jessica Metcalfe · Karin Orlemann · Sandra Salter · Ian Peters · Alan Leeb

Colorectal cancer is an important cause of morbidity and mortality in Australia.1 The National Bowel Cancer Screening Program (NBCSP) aims to detect the disease early by offering faecal occult blood testing (faecal immunochemical test, FIT) to people aged 50–74 years.2 The expansion of the NBCSP has been paralleled by increased numbers of FITs outside the program (community‐initiated FITs) for a number of reasons, including the presence of symptoms. We investigated whether colonoscopy services should provide endoscopies to patients with positive FIT results with the same priority, regardless of whether the test was instigated by the NBCSP, by analysing data from the Newcastle Direct Access Colonoscopy Service (DACS) for the period 2014–18. The DACS manages all patients in the same manner: a positive FIT result leads to assessment for colonoscopy.3,4 Ethics approval was granted by the Hunter New England Human Research Ethics Committee (reference, AU201608‐01). All data were recorded prospectively. Findings were categorised according to surveillance categories endorsed by the Gastroenterological Society of Australia and the Colorectal Surgical Society of Australia and New Zealand.5 Data accuracy was confirmed by reviewing the primary sources for 10% of patients. We identified 2693 patients referred for screening colonoscopy between 1 July 2014 and 30 June 2018; 1439 (53%) had had community‐initiated FITs (Box 1). After excluding 318 patients who did not attend or were lost to follow‐up (community‐initiated, 200; NBCSP, 118) and ten patients with poor bowel preparation and no follow‐up colonoscopy during the study period, 2365 complete screening colonoscopy outcomes were analysed: 1233 following community‐initiated and 1132 following NBCSP testing. With these sample sizes, the study had 80% power to detect differences in colonic neoplasia rate ranging from 16 percentage points (assumed prevalence, 50%) to two percentage points (assumed prevalence, 3%). Z‐tests were used to calculate P values, and Wald tests (two‐tailed) for calculating confidence intervals (CIs) for the differences between the two groups. Differences between the two groups in the proportion of patients with each specific finding are presented with 99% asymptotic CIs to control for multiple testing. Colonoscopy quality was high: the completion rate (defined as either caecal intubation, reaching an ileocolic anastomosis, or reaching an obstructing mass lesion) was 97.1% (community‐initiated, 1193 of 1233, 96.8%; NBCSP, 1104 of 1132, 97.5%), and the adenoma detection rate was 49%, exceeding international benchmarks for either symptomatic or screening patients (for screening: at least 25% in men and 15% in women;6 for populations enriched with patients with positive FIT results: 35%7). The rate of colorectal neoplasia (malignant or pre‐malignant) was similar in the two groups. Importantly, the difference in the rates of adenocarcinoma was not statistically significant (community‐initiated, 4.0%; NBCSP, 2.7%; difference, 1.3 percentage points [99% CI, –0.6 to 3.3 percentage points]; P = 0.09). The only statistically significant difference by type was that the incidence of high risk adenoma was slightly higher in the NBCSP group (22.9% v 17.2%; difference, 5.7 percentage points [99% CI, 1.4–10 percentage points]; P < 0.001) (Box 2). We found that the incidence and detection rates of colorectal neoplasia in people aged 50–74 years were similar for people with positive results for NBCSP or community‐initiated FITs. The large population in our study means that it provides colonoscopy providers strong evidence that evaluation should be performed equally promptly for patients with positive results from NBSCP and community‐initiated FITs. Box 1 – Demographic characteristics of the 2693 patients with positive faecal immunochemical test results and referred to the Newcastle Direct Access Colonoscopy Service for colonoscopy, 2014–18 Faecal immunochemical test Total Community‐initiated NBCSP Number of patients 1439 1254 2693 Sex Women 675 559 1234 Men 764 695 1459 Age (years), mean (SD) 62.9 (6.8) 63.2 (7.3) 63.1 (7.0) Numbers of patients 50–54 years 213 147 360 55–59 years 280 271 551 60–64 years 312 212 524 65–69 years 330 288 618 70–74 years 304 336 640 NBCSP = National Bowel Cancer Screening Program; SD = standard deviation. Box 2 – Differences in colonoscopy outcomes for people who had community‐initiated (1233 patients) or NBCSP (1132 patients) faecal immunochemical tests CI = confidence interval; NBCSP = National Bowel Cancer Screening Program. *Large sessile polyps (> 2 cm) or malignant polyps. † Between values for community‐initiated and NBCSP groups.

Simon Whitcher · Monique Magnusson · Jon Gani · Christopher Oldmeadow · Peter G Pockney

Patients with ruptured HCC should be treated with the aim of long term survival

Natassia P Tan · Ammar Majeed · Stuart K Roberts · Paul J Gow · Penny Hey · Xianjun Mah · Mark Goodwin · Siddharth Sood · John Lubel · Amanda Nicoll · Anouk Dev · Sally J Bell · William W Kemp

Activity against antibiotic-resistant pathogens and a lack of serious side effects make phage therapy an attractive treatment option to combat refractory bacterial infections

Aleksandra Petrovic Fabijan · Ali Khalid · Susan Maddocks · Josephine Ho · Timothy Gilbey · Indy Sandaradura · Ruby CY Lin · Nouri Ben Zakour · Carola Venturini · Bethany Bowring · Jonathan R Iredell

To the Editor: Nearly all industries profit from today's 25 million slaves and 150 million child labourers.1 The results of their work, including medical disposables, are sold worldwide. Unfortunately, there is not a comprehensive analysis identifying exactly where slaves are involved in the medical products supply chain. From the hazardous work forging surgical instruments in Pakistan to the manufacture of gloves in Malaysia,2 slavery permeates the manufacture and supply chains of medical products. Slaves are involved in the direct manufacture of medical products and in the generation of raw materials used to make medical devices, including cotton, rubber and metals.3 Some companies, have made a public effort to review their supply chains;4 however, many organisations are failing to monitor human rights abuses in their supply chains.5 As countries legislate Modern Day Slavery Acts, a few companies are moving to eliminate slavery. As an act of radical transparency in the long‐discussed issue of child labour in the cocoa industry, in 2017, Nestlé published the number of child labourers aged 5–17 years known to be working on cocoa farms that supply their cocoa.6 Nothing similar to Nestlé's effort has been done in the medical industry. However, notable efforts to regulate procurement have been demonstrated in the United Kingdom and Sweden.2 Few health professionals are responsible for the direct sourcing of medical products. Nonetheless, when speaking with managers, executives, and representatives of medical suppliers, we have the opportunity to share our concerns for the origins of the products we use. While a conversation with a medical representative on this topic may demonstrate scant knowledge of the manufacturing processes of the goods they are selling, that initial conversation is an important first step towards transparency, and we know that the influence of health professionals on industry is significant.7 Modern slaves are forced to work under threat of harm or by coercion or deception. Unable to refuse or leave, they earn little to no pay for extensive working hours in unsafe conditions, which may cause injury, sickness and, at times, death. Reports of harassment are common. In the medical industry, we must do all we can to address modern slavery because, above all, we should “first do no harm”.

Sharon Sitters

To the Editor: In August 2019, the Australian Institute of Health and Welfare (AIHW) released a report on chronic musculoskeletal conditions in Australia.1 In the report, the prevalence of gout is estimated at 0.8% (equating to 187 000 people), based on self‐reported survey data. This rate is much smaller than previously reported in Australia. Recent South Australian population‐based studies using “self‐reported doctor diagnosed gout” as the case definition reported the prevalence of gout as being between 5.2% and 6.8%.2 Thus, the rate of self‐reported gout detailed in the AIHW report is incongruous with published data and appears to be unusually low. The reported prevalence of gout in the AIHW document is based on the 2017–18 National Health Survey, in which participants were asked to self‐report doctor‐ or nurse‐diagnosed gout and whether the gout was current and expected to last for 6 months or more. Participants who did not identify as having current or long term gout did not have their condition recorded in the survey. Given that for most patients gout manifests as an intermittently flaring disease, with most flares lasting 7–10 days, respondents would likely not report their gout to be “current” or “likely to last 6 months” unless they have a clear understanding that gout is a chronic disease of monosodium urate crystal deposition. Many people with gout have not received this information from their health care providers and view the disease as present only when they are experiencing a flare.3 The chronic nature of the disease is reflected in the current definition of gout as “current or prior clinically evident disease”, as agreed by international gout experts.4 Gout is a systemic disease and an established independent risk factor for renal and cardiovascular disease. Like many chronic diseases, flares of gout and their long term consequences can be prevented with daily medication. However, both international and Australian evidence demonstrates that gout is inadequately treated, that persistence to urate‐lowering therapies is low, with suboptimal outcomes for patients.2 Under‐reporting and under‐recognition of gout and its burden on society is likely to contribute to undertreating and failure to manage it as a chronic disease — in contrast to other chronic conditions such as diabetes, in which the need for optimal disease management is well accepted. It is important that the burden of gout in Australia is understood and accurately measured to allow optimal use of limited health resources, reduce burden on society, and improve outcomes for people with this condition. Quality data are required but must be generated with appropriate definitions. We suggest that a validated case definition, such as “self‐reported gout” or “urate‐lowering therapies use”, be used in future Australian epidemiological studies.5

Helen I Keen · Philip C Robinson · Nicola Dalbeth · Catherine Hill