Volume 213 Issue 2

COVID‐19 ARDS is a predictable serious complication of COVID‐19 that requires early recognition and comprehensive management “This disease is still too strange to us, and there are too many doubts”, says Dr Ling Qin (LQ), after reviewing more than 400 patients with coronavirus disease 2019 (COVID‐19) pneumonia in Wuhan Union Hospital, China. COVID‐19 is a novel disease. We are familiar with acute respiratory distress syndrome (ARDS); however, when it occurs as part of COVID‐19, it has different features and there remain unanswered questions. So if someone has COVID‐19 ARDS, how does it compare and contrast with ARDS from other causes? To answer this question we provide a summary of the published literature (based on a PubMed search using the terms “COVID‐19” and “ARDS”, 17 April 2020) and current clinical experience from managing patients with COVID‐19 ARDS in Singapore (SHP) and Wuhan (LQ). Severe COVID‐19 represents viral pneumonia from severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2) infection leading to ARDS. Its manifestations can be viewed as a combination of the two processes, namely viral pneumonia and ARDS. COVID‐19 is a novel disease recognised initially in Wuhan, China, in December 2019, and is now pandemic. It is likely caused by zoonotic spillover of a β‐coronavirus type 2b that is now transmitted between humans. Along with the other serious coronavirus infections of severe acute respiratory syndrome and Middle East respiratory syndrome, which also cause ARDS, COVID‐19 represents an ongoing global threat as this virus family has the potential to mutate and infect non‐immune populations. Australia's living guidelines provide the latest recommendations and evidence.1 Diagnosis SARS‐CoV‐2 infection can be confirmed by positive detection of viral RNA in nasopharyngeal secretions using a specific PCR test. COVID‐19 illness can be confirmed by a consistent clinical history, epidemiological contact, and a positive SARS‐CoV‐2 test. COVID‐19 ARDS is diagnosed when someone with confirmed COVID‐19 infection meets the Berlin 2012 ARDS diagnostic criteria2 of (i) acute hypoxaemic respiratory failure; (ii) presentation within 1 week of worsening respiratory symptoms; (iii) bilateral airspace disease on chest x‐ray, computed tomography (CT) or ultrasound that is not fully explained by effusions, lobar or lung collapse, or nodules; and (iv) cardiac failure is not the primary cause of acute hypoxaemic respiratory failure. ARDS is underdiagnosed in intensive care settings.3 ARDS develops in 42% of patients presenting with COVID‐19 pneumonia, and 61–81% of those requiring intensive care.4 COVID‐19 ARDS follows a predictable time course over days, with median time to intubation of 8.5 days after symptom onset in Singaporean patients.5 This is similar to previous reports where ARDS developed at day 8 or 9 after symptom onset. It is therefore important to monitor patients for the development of ARDS as their COVID‐19 infection progresses. Respiratory rate and SpO2 are two important parameters for judging patients’ clinical condition and allowing early recognition of ARDS. A patient who fits any one of the following conditions may have severe disease and require further evaluation: respiratory rate ≥ 30 breaths/min; SpO2 ≤ 92%; and PaO2/FiO2 ≤ 300 mmHg. Blood tests can also be helpful. In Singapore, it was noted that raised C‐reactive protein levels and blood neutrophil counts along with lymphopenia were more common in patients requiring invasive mechanical ventilation for COVID‐19 ARDS.5 Lung pathology ARDS causes diffuse alveolar damage in the lung. There is hyaline membrane formation in the alveoli in the acute stage, and this is followed by interstitial widening and by oedema and then fibroblast proliferation in the organising stage. COVID‐19 ARDS causes the typical ARDS pathological changes of diffuse alveolar damage in the lung.6,7 As patients move through the course of their illness, the longer term outcomes of ARDS are being reported, with lung fibrosis appearing as part of COVID‐19 ARDS.8,9 A study reported that 17% of patients had fibrous stripes in chest CT scans,9 and considered that the fibrous lesions may form during the healing of pulmonary chronic inflammation or proliferative diseases, with gradual replacement of cellular components by scar tissues. Thrombosis Pulmonary thrombosis is common in sepsis‐induced ARDS. Coagulation dysfunction appears to be common in COVID‐19, and is detected by elevated D‐dimer levels. In fatal cases there is diffuse microvascular thrombosis, suggesting a thrombotic microangiopathy, and most deaths from COVID‐19 ARDS have evidence of thrombotic disseminated intravascular coagulation.10 This may explain some of the atypical or unexpected manifestations seen in the lung, such as dilated pulmonary vessels on chest CT, and episodes of pleuritic pain. Vascular enlargement is rarely reported in typical ARDS, yet was seen in most cases of COVID‐19 ARDS.9 Mortality COVID‐19 ARDS appears to have worse outcomes than ARDS from other causes. The intensive care unit and hospital mortality from typical ARDS are 35.3% (95% CI, 33.3–37.2%) and 40.0% (95% CI, 38.1–42.1%), respectively.3 For COVID‐19 ARDS, mortality ranged between 26% and 61.5% if ever admitted into a critical care setting, and in patients who received mechanical ventilation, the mortality can range between 65.7% to 94%.4 Risk factors for poor outcomes include older age; presence of comorbidities such as hypertension, cardiovascular disease and diabetes mellitus; lower lymphocyte counts; kidney injury; and raised D‐dimer levels. Death from COVID‐19 ARDS is due to respiratory failure (53%), respiratory failure combined with cardiac failure (33%), myocardial damage and circulatory failure (7%), or death from an unknown cause.4 Radiology The radiology of ARDS is distinctive, yet COVID‐19 pneumonia appears to have unique features. This likely results from the co‐occurrence of viral pneumonia and ARDS, and allows radiologists to be fairly specific in diagnosing COVID‐19 pneumonia. The most discriminating features for COVID‐19 pneumonia in China compared with viral pneumonia in the United States included a peripheral distribution of opacification (80% v 57%; P < 0.001), frosted glass opacities (91% v 68%; P < 0.001), and vascular thickening or enlargement (58% v 22%; P < 0.001).11 These imaging features appear to be typical for COVID‐19 pneumonia and can be helpful in early screening of highly suspected cases and in evaluation of the severity and extent of disease. As COVID‐19 lung disease progresses, the lesions are more likely to be bilateral, lower lung predominant and multifocal. They often have the appearance of rounded opacities, termed “COVID balls”. With the development of ARDS, the extent of lung involvement increases, and there is a consolidative component.12 The opacities resolve with recovery from COVID‐19;13 however, with ARDS, the lesions increase in their extent and density, and evolve to fibrotic bands. Ventilation The strategy of breathing support is very important in treating COVID‐19 ARDS, as is the case with typical ARDS caused by other pathogens.14 The key elements are: use oxygen by nasal cannulae to achieve SpO2 > 92%; use of high flow nasal oxygen is controversial and highly dependent on the treatment location; avoid non‐invasive ventilation; prone ventilation appears to be beneficial; and consider extracorporeal membrane oxygenation for rescue. Because of concerns about viral transmission to other patients and health care workers,15 the use of high flow nasal oxygen and non‐invasive ventilation (such as bi‐level positive pressure ventilation) for COVID‐19 ARDS is highly dependent on the health care setting. Australian COVID‐19 guidelines1 strongly recommend against the use of high flow nasal oxygen in emergency departments, but provide a strong recommendation for its use in negative pressure single rooms. Non‐invasive ventilation may be used in negative pressure rooms with appropriate viral transmission precautions.1 Clinical experience has found inconsistent benefit from non‐invasive ventilation and there is concern about aerosol generation and increased risk of viral transmission. Prone ventilation appears to be beneficial for COVID‐19 ARDS.1 Placing a person in prone position promotes more homogenous aeration of the lung in ARDS and can improve oxygenation. While prone ventilation is used in only about 16% of patients with typical ARDS,3,16 in COVID‐19 it is being used successfully earlier in the course of ARDS, and suggested use is for > 12 hours per day.16 Venovenous extracorporeal membrane oxygenation can be used as rescue for mechanically ventilated adults with COVID‐19 and hypoxaemia that persists despite optimised ventilation, use of rescue therapies and prone ventilation. Among critically ill patients treated in Wuhan, prone ventilation and extracorporeal membrane oxygenation treatment were not found to be as effective as for ARDS caused by other pathogens. Possible reasons include: COVID‐19 pneumonia was still progressing and was not under control; lung lesions were not completely gravity‐dependent under ultrasound, so the effect of the prone position was limited; the patient's immune status was not restored, and a secondary hospital‐acquired infection worsened the condition; and when case numbers are high from the epidemic, the management mode and human resource arrangement of the isolation wards still need to be discussed and strengthened. Anecdotal observations in Singapore (SHP) and investigations in the Netherlands17 suggested that patients ventilated for COVID‐19 ARDS tended to have plateau pressures < 30 cmH20 and driving pressures < 15 cmH20 despite high oxygen requirements. The lung protective ventilation strategy used in typical ARDS involves a low tidal volume (6 mL/kg) and higher positive end expiratory pressure targets. For COVID‐19 ARDS, a change to more generous tidal volume targets allowing up to 8 mL/kg and lower positive end expiratory pressure levels is suggested to prevent patient self‐inflicted lung injury. Adjunct treatment In typical ARDS, continuous neuromuscular blocking agents, high dose corticosteroids and recruitment manoeuvers were the most frequently used adjunctive therapies. In COVID‐19 ARDS, the evidence for systemic steroids is still scarce and they are only recommended in patients with concomitant shock which has been unresponsive to vasopressors. There are concerns that steroids may increase viral shedding and possibly lead to a higher mortality rate. Antiviral therapy Many patients with COVID‐19 receive antiviral or immunosuppressive therapy. In Australia, the National COVID‐19 Clinical Evidence Taskforce1 recommends administering antiviral medications or other disease‐modifying treatments in the context of clinical trials. Singapore was using empiric lopinavir–ritonavir plus subcutaneous interferon‐β 1b initially, but is now randomising patients to receive remdesivir. In Wuhan, a broad range of antiviral and immune therapies are being used. All patients also received treatment with Chinese medicine. COVID‐19 ARDS is a predictable serious complication of COVID‐19 that requires early recognition and comprehensive management. Research programs such as the Medical Research Future Fund 2020 Respiratory Medicine Clinical Trials Research on COVID‐19 grant opportunity are required to answer the important questions that remain about therapies for COVID‐19 ARDS.

Peter G Gibson · Ling Qin · Ser Hon Puah

How to deal with diabetes and COVID‐19 — do we just dial in? The emergence of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), and the subsequent announcement by the World Health Organization of a global pandemic, has altered health care across the public and private sectors. Clearly, coronavirus disease 2019 (COVID‐19) is having a huge impact on general practitioners, emergency physicians, respiratory physicians, intensivists and related staff, and is also impacting the day‐to‐day delivery of chronic health care. Conditions such as type 1 and type 2 diabetes require regular, usually face‐to‐face contact with GPs, endocrinologists, diabetes educators, dieticians and podiatrists to optimise glycaemic control. However, with recommendations regarding social isolation to minimise spread of COVID‐19,1 the delivery of such health care is increasingly being delivered using telehealth. This has been greatly facilitated in Australia with the announcement of temporary Medical Benefits Schedule telehealth (videoconference and telephone) item numbers during the COVID‐19 pandemic.2 The eligibility criteria for bulk‐billing include individuals who are immunosuppressed or with chronic health conditions, which include diabetes mellitus. Significantly, the eligibility criteria also include pregnancy. At a time when individuals are being asked to socially isolate to minimise exposure to SARS‐CoV‐2, many who greatly need ongoing health care are avoiding these appointments out of fear of infection from those also attending the clinic and perhaps from clinicians, who are recognised to be at high risk of infection. It is therefore crucial that telehealth care services are offered (and if not available, fast tracked). This may be via videoconference, or a phone call for those who are less able to use newer technologies or if videoconferencing services are not available. Services traditionally offered in a group setting, such as group education for those with gestational diabetes or type 2 diabetes, may also be delivered via this technology. The use of diabetes management technology can benefit health care and telehealth consultation. Continuous or flash glucose monitoring can be used with both type 1 and 2 diabetes with therapeutic benefit. Many of these devices upload automatically to cloud‐based software. Other devices such as insulin pumps and blood glucose meters can be uploaded by the user before the consultation, which places a greater burden on the individual. Involvement of parents, carers and diabetes educators before the appointment may be of benefit. This may lead to improvements in individual diabetes self‐management. The electronic medical record is vital in updating changes to medications, results and scheduled appointments. Many pathology providers facilitate electronic access to test results, but anecdotally some patients may avoid having pathology tests taken in order to minimise travel and exposure to others. The individuals who are at greatest risk of complications from COVID‐19 are still being determined. It seems clear that those with diabetes and other chronic comorbidities are at increased risk of morbidity. A number of small observational studies have analysed comorbidities in individuals with severe SARS‐CoV‐2 infection from China. Rates of diabetes (type unspecified, but most likely to be type 2 diabetes given the low incidence of type 1 diabetes in China3) are reported between 7.4% in confirmed cases infection,4 and up to 17% in cases with severe pneumonia.5 A recent meta‐analysis of six studies (including 1527 people) analysed the prevalence of comorbidities among individuals with severe and non‐severe COVID‐19.6 Diabetes complicated 11.7% of severe cases compared with 4% of cases of non‐severe COVID‐19. This did not reach statistical significance; however, this is clearly limited by a lack of statistical power and further analyses are required. Retrospective analysis of survivors and non‐survivors of the 2002–2003 SARS coronavirus outbreak suggested that diabetes was a predictor for mortality.7 This association seemed to be driven by glycaemic control, with an independent association with elevated fasting plasma glucose. This further highlights the need to maintain ongoing medical care to optimise glucose control throughout the current COVID‐19 pandemic. All people who currently smoke should be advised to cease smoking. In addition, receiving the recently available influenza vaccine, which is usually recommended for people with diabetes, would be advisable, although patients should be aware that this will not protect against COVID‐19. Individuals with both type 1 and type 2 diabetes need to have clear action and sick day plans in the eventuality that they become unwell, and should be encouraged to seek face‐to‐face care for complications such as myocardial infarction or high risk foot ulcer. This should be emphasised even for individuals with long standing diabetes, whose initial diabetes education may have occurred years ago. Individuals with type 1 diabetes should have ketone monitoring strips available (preferably blood continuous subcutaneous insulin infusion ketone test strips), know when to test for ketones, and be aware of the need for additional insulin doses (via injection or continuous subcutaneous insulin infusion) during an intercurrent illness. Excellent sick day management resources are available on the websites of the Australian Diabetes Educators Association (https://www.adea.com.au) and National Diabetes Services Scheme (https://www.ndss.com.au). Individuals using hybrid closed loop insulin pumps should be educated that during illness the wearer may need to exit automatic mode to enable more rapid correction of hyperglycaemia with manual correction boluses of insulin and a temporary increased basal rate.8 The algorithm within the hybrid closed pump may otherwise not adapt quickly enough to manage hyperglycaemia during acute illness. With the increased use of sodium–glucose cotransporter type 2 (SGLT2) inhibitors in Australia and internationally, all individuals treated with SGLT2 inhibitors should be educated on the need to withhold these drugs during illness to minimise the risk of ketoacidosis.9 This should be reiterated to GPs and emergency physicians to screen for SGLT2 inhibitor use in patients presenting with COVID‐19 or any illness. There have been mixed anecdotal reports as to whether non‐steroidal anti‐inflammatory drugs (NSAIDs) may predispose patients to COVID‐19. There has been suggestion that NSAIDs may upregulate angiotensin‐converting enzyme 2 (ACE2),10 and therefore potentially predispose by a similar mechanism suggested for angiotensin receptor blockers (ARBs). Currently there are no guidelines to avoid the use of NSAIDs. In relation to COVID‐19, there is also increasing interest in the use of ACE inhibitors and ARBs in individuals with type 1 or type 2 diabetes and other chronic care conditions, such as diabetes. SARS‐CoV‐2 binds to ACE2, allowing entrance into the host cells.11 ACE inhibitors and ARBs can result in upregulation of ACE2 in some tissues in both human and animal models.10,12,13 However, not all investigators have found a link between these antihypertensives and upregulation of ACE2.14 It has been proposed that ACE inhibitors and ARBs may theoretically increase susceptibility to COVID‐19 by increasing ACE2 levels. However, there is currently no evidence to link the use of these agents to increased risk or severity of COVID‐19. Indeed, other research groups have hypothesised that the use of ARBs may be a potential therapeutic modality.15 Following SARS‐CoV‐2 binding to ACE2, there is downregulation of ACE2 with subsequent increased angiotensin levels and exacerbation of COVID‐19 related lung injury. It has been proposed that the downregulation of ACE2 by ARBs might protect against such injury.15 Other groups have suggested that ARBs may stabilise the binding of ACE2 to the type 1 angiotensin receptor and may therefore reduce available binding sites for SARS‐Cov‐2.16 Currently, there is no evidence to suggest changing antihypertensive therapy, and multiple national and international bodies including the Australian Diabetes Society, Australian and New Zealand Society of Cardiac and Thoracic Surgeons, and the American College of Cardiology and European Society of Hypertension have recommended that ACE inhibitors and ARBs should not be ceased.17 Trials are currently underway to assess the impact of these agents during COVID‐19 infection (https://clinicaltrials.gov). Patients should be encouraged to continue their ACE inhibitor or ARB drugs, and if not prepared to do so, be offered alternative drugs for blood pressure control. This is a time of great concern to all individuals, and perhaps more so to those who have been informed they are at greater risk of COVID‐19 and its complications. This may necessitate a greater state of preparedness. The current advice is that there will be no shortage of insulin supplies or consumables needed for insulin pump therapy or blood glucose monitoring equipment, yet anecdotally, local pharmacy shortages of insulins, ketone strips and oral hypoglycaemic drugs have been reported and are being addressed by government prescription limits. Should patients be unable to obtain their usual prescriptions, suitable alternatives can be recommended to them by their diabetes care clinicians. It is critical that individuals with diabetes and other chronic conditions do not hoard these medical supplies and inadvertently create a critical supply shortage. Continuation of health care to at‐risk individuals is crucial throughout the pandemic. Telehealth is the key for the delivery of such care. It is important that people with diabetes are educated regarding the management of their condition during acute illness, including medication changes. It is also critical that there is no deterioration in the medical management of glycaemia and other complications of diabetes, which, if neglected, may result in increased morbidity and mortality independent of COVID‐19.

Emma S Scott · Alicia J Jenkins · Gregory R Fulcher

The current reduced capacity of post‐release services may compound offender vulnerabilities, increasing their risk of harm to themselves and others The coronavirus disease 2019 (COVID‐19) pandemic has affected many countries internationally and has been implicated in more than 445 000 deaths worldwide.1 The speed at which this infectious disease is transmitted has led to calls to immediately release prisoners from custody in some countries, including Australia, and has already led to the release of some prisoners in others. The reasons for these calls to action are intuitively rational. Custodial environments are susceptible to a COVID‐19 outbreak given the confined conditions and potential for overcrowding.2 Moreover, prison populations are often vulnerable, having poorer physical and mental health and other social challenges (eg, substance misuse, homelessness) compared with the general population.3 At the time of writing, Australian governments have yet to immediately release select prisoners into the community as part of efforts to mitigate the spread of COVID‐19, despite recent advocacy to do so. Experts across a number of sectors have recommended the early release of prisoners from vulnerable groups if possible, including Indigenous Australians, women, children, older prisoners, victims of domestic violence, and those with chronic health issues.4 However, the proposed early release strategy requires a nuanced assessment of its potential societal consequences and, most importantly, its immediate impact on the health and wellbeing of candidate prisoners for release. Victoria and New South Wales — Australia's most populous states — provide a useful case in point. Stage 3 restrictions — home confinement except for restricted essential activities5 — have been in effect for approximately 2 months. Some restrictions were eased in late May, although limits on public and private gatherings remain in place. Prisons in both Victoria and NSW have yet to record a single case of COVID‐19 within the inmate population. In Victoria, a number of safety mechanisms have been introduced by correctional centres to help manage the potential transmission of the virus.6 These include: temperature testing of all staff before entry to the facility; sending staff home who present with high temperatures and other flu‐like symptoms, and requiring them to undertake a COVID‐19 test before returning; isolating all new prison admissions for up to 14 days; isolating prisoners who display cold or flu‐like symptoms; adjusting programs to abide by physical distancing; and enabling prisoners to connect with family members via video calls on tablet devices (“video visits”) as an alternative to in‐prison visitation. Moreover, conventional medical, psychological and cultural support services continue to be available to prisoners. Similar precautions have been implemented in NSW correctional centres,7 which hold the largest proportion of prisoners in Australia. With no confirmed cases in both Victorian and NSW correctional centres and ongoing efforts to restrict the materialisation of COVID‐19 in custodial settings, the potential costs of releasing vulnerable prisoners into the community necessitates scrutiny. Any prisoners released under anti‐COVID‐19 preventive measures will return to a general community enduring social restrictions and society‐wide economic contraction. The post‐release community support services ordinarily available to released offenders are currently compromised or are experiencing significant delays.8,9,10,11 Moreover, government social security services (ie, Centrelink), which are heavily relied upon by individuals after release, are currently overwhelmed as they service thousands of newly unemployed clients.12 Mental health and crisis support services are also strained as they adjust to remote service delivery and contend with an elevated spike in community‐wide help seeking.13 The reduced capacity for intensive case management, monitoring and re‐entry assistance for released prisoners is a serious concern given their higher levels of complex mental and physical health needs, as well as histories of unemployment, addiction, social disadvantage and homelessness.3 Australian research points to high rates of mortality and self‐harm shortly after release from custody.14 Key predictors of post‐release mortality include mental disorder, suicide and substance misuse — concerns that transitional support programs and other post‐release interventions will be unable to optimally manage during the national lockdown. This scenario poses an increased health risk for released prisoners, compounding their vulnerability and increasing the likelihood of problem behaviour and recidivism. Without readily available coping strategies and assistance with pro‐social functioning, released prisoners with histories of violence, aggression, impulsivity and serious mental illness may put themselves and others (particularly cohabitants) at risk. Calls to release prisoners who are survivors of domestic violence must also consider the heightened risk of revictimisation after release. The social isolation, economic stress and reduced options for support during stage 3 restrictions may exacerbate unhealthy relationship dynamics, especially among those with complex needs. Moreover, physical distancing may not be adhered to by those whose behavioural and psychological needs are untreated. Proposals to immediately release vulnerable prisoners to avert the ostensible threat of COVID‐19 in Australian custodial environments warrant consideration. The potential for a COVID‐19 outbreak in custody is a genuine concern, notwithstanding proactive measures employed in Victorian and NSW correctional centres. However, this advocacy must consider the broader social context. A focus on early release cannot be uncoupled from the current compromised community environment prisoners will face after detainment. Community support services are increasingly strained at a time when released prisoners will have a greater need for them. As such, it is important to balance the relative health and safety trade‐offs of remaining in custody — in Victorian and NSW prisons, there are no confirmed cases of COVID‐19 and health supports remain available — with early release into a resource‐depleted community. The real prospect of harmful outcomes for immediately released vulnerable prisoners must be weighed heavily during this challenging period.

Stephane Shepherd · Benjamin L Spivak

Australia has managed the crisis well so far but we should now also plan for future waves and the recovery phase Coronavirus disease 2019 (COVID‐19) pandemic management is focused on prevention, case finding and survival. Australia and New Zealand have done well and the numbers in our intensive care units (ICUs) are currently manageable. Our subacute sector is presently able to deal with patients requiring rehabilitation. However, rehabilitation needs following COVID‐19 are broad, complex and include cognitive, motor and respiratory sequelae to the infection, acute respiratory distress syndrome, and the thromboembolic response. Planning and anticipatory action has been Australia's strength so far. In the same vein, an active planning approach is now required for the post‐acute and rehabilitation response. This pandemic will inevitably have its waves, and will continue to threaten until a vaccine is rolled out. Not having a plan for possible surges is unconscionable, particularly when the consequences of the relaxation of restrictions are unknown. Currently, Australian numbers are at a trickle; however, the challenge has been front and centre in countries such as Italy,1 Spain, the United Kingdom and North America.2 In Wuhan, China, 36% of those with severe COVID‐19 had neurological complications such as stroke, critical care neuropathy, and the complications of prolonged bed rest (eg, venous thromboembolism, disseminated intravascular coagulation, acute kidney injury, delirium anxiety, post‐traumatic stress disorder).3 In Italy, rehabilitation physicians have been treating post‐extubation dysphagia, impaired mobility, critical care myopathy and neurocognitive losses,1 while the British Society of Rehabilitation Medicine has established a framework of partnership with acute services to improve patient flow, outcomes and access to ventilators.4 In the United States, hospitals have had to rapidly transition acute patients to rehabilitation hospitals. In New Orleans, a 1000 bed post‐acute hospital was dedicated to post‐COVID‐19 disability, with rehabilitation teams treating patients battling persistent hypoxia, stroke and mental illness.5 The majority of patients who are ventilated for more than 7 days suffer complications that require rehabilitation, 60% are unable to walk, and 17% die within a year.7 One‐third suffer neurological complications, many require inpatient rehabilitation for over 3 weeks, and some take over 150 days to regain their capacity to walk independently.8 Others with stroke or cardiac complications of COVID‐19 will require rehabilitation for up to 6 weeks, with some requiring lifelong support. Australia needs to plan now, not just for survivors in the initial post‐acute stage, but also to manage individuals affected in subsequent waves. Such patients may require rehabilitation, along with those, fearful of infection, who present to hospital late with non‐COVID‐19 conditions like stroke, and those with deteriorating chronic diseases who have not had access to hospital based services. That means not only estimating the patient population but also ensuring that subacute health workers have sufficient access to personal protective equipment, staffing and training. In the Australian Government's emergency response plan,9 the recovery phase devolves to the states, but there is no mention of the post‐acute phase. In April, the New South Wales Ministry of Health established a rehabilitation community of practice to advise it. This follows international experience, as the International Society of Physical and Rehabilitation Medicine's disaster committee lead, Australia's Fary Khan states: “early rehabilitation reduces disability and improves clinical outcomes”.10 Currently, many rehabilitation units are not prepared. Inpatient rehabilitation units (public and private) are almost always working to capacity. COVID‐19 patients will be expect to be accommodated in addition to usual patients (eg, strokes, spinal injuries, amputations). The NSW Rehabilitation Community of Practice has developed a staged COVID‐19 escalation plan,11 but the plans turn on one integral point — contagion. When COVID‐19 survivors come to rehabilitation wards will they no longer be infected? While a national statement exists,12 local de‐isolation protocols are yet to be implemented or updated in many hospitals and local health districts. Many people with severe COVID‐19 have positive nasal swabs for up to 37 days13 but are not considered infectious once 10 days have elapsed from after first symptoms. Attention to this timetable is critical should we need to make ICU beds readily available by shifting patients to rehabilitation. In some US rehabilitation hospitals, patients are assumed to always be infectious, which has a significant impact on personal protective equipment usage. To ensure de‐isolation, moderate and severe COVID‐19 patients transferring to rehabilitation must have negative swabs on 2 consecutive days, be symptom‐free for 2–3 days and be at least 10 days from symptom onset. In NSW, these criteria are currently being put in place and such a protocol will require discussion, review of the evidence, and leadership to execute. Once we have a de‐isolation protocol, we can confidently activate a staged escalation plan. While our ICUs are coping with current numbers, our subacute sector has been managing with innovative models of care, such as mobile rehabilitation teams.14 The NSW Rehabilitation Community of Practice's COVID‐19 response principles11 refer to mobile rehabilitation teams, variously called ART (acute care rehabilitation team) or SMART (specialist management with acute rehabilitation treatment) teams. They provide rehabilitation and discharge planning services to patients in the acute hospitals. It is a parallel care model in partnership with acute care that has been successful in decreasing length of stay and facilitating early discharge or transfer to inpatient rehabilitation facilities. These teams have discharged almost 50% of their patients directly home, avoiding inpatient rehabilitation admissions, and have been active in many NSW hospitals since 2009. Once home, tele‐rehabilitation physician consultations, supported by community‐based allied health practitioners, can be delivered, although additional resources are still being sought. Similar models exist or are under development in other jurisdictions as well. In the event that our acute hospitals start to face challenges in accommodating those needing COVID‐19 or ICU beds, the subacute sector will need to escalate to the next stage to create access. Options include decanting non‐COVID‐19 patients to the private sector, increasing resources to acute or mobile rehabilitation teams, scaled up tele‐rehabilitation services, and preparation for public hospital rehabilitation inpatient units to manage COVID‐19 patients. However, in order to decant to the private sector we need completed agreements with private hospitals, as flagged on 31 March by the Minister of Health.15 Many private hospitals have facilities that are well suited for rehabilitation patients. This would require delineating private hospitals as COVID‐19‐free facilities and would be dependent on appropriate triage and testing facilities. Managing the logistics will be a challenge in the subacute sector, particularly if planning is left as an afterthought. The efficient flow of disabled COVID‐19 patients from acute to rehabilitation care will likely produce better patient outcomes and improve safety. Egress from acute hospitals means access to intensive care and ventilation for the community. If Australia and New Zealand's success at flattening the curve continues, our existing subacute sector will manage. If not, mobile rehabilitation teams will need to be expanded, systems for patient flow to the private sector will need to be operational, and enhanced tele‐rehabilitation services will need to be working. This will require the same vision and leadership that made our acute COVID‐19 response world leading, collaborative and publicly supported. In the UK and the US, we see the brutality of this pandemic, with mass burials and the tragic toll on health care workers. Australia and New Zealand have avoided this so far, but it is because we have planned well. We now need to prepare for the recovery phase because surviving may not be the same as living.

Steven G Faux · Kathy Eagar · Ian D Cameron · Christopher J Poulos

Australia needs to improve vigilance in the global endeavour to eradicate poliomyelitis In 1988, there were over 350 000 cases of paralytic poliomyelitis globally.1 In 2018, there were 29 cases and in 2019 there were 112 cases2 — all in the only two remaining countries in the world where wild poliovirus (WPV) is endemic (Afghanistan and Pakistan). We are tantalisingly close to global eradication. What is poliomyelitis? Poliovirus is an enterovirus and exists as three serotypes: WPV types 1, 2 and 3. Spread via the faecal–oral route, poliomyelitis results in subclinical or self‐limited infection in most patients, but causes acute flaccid paralysis (AFP) due to anterior horn cell damage in about one in 200 cases.3 Ubiquitous distribution of polioviruses and epidemics of paralysis caused widespread panic throughout the world in the early 20th century. With an ambitious and unprecedented level of international public and private collaboration and funding, the Global Polio Eradication Initiative (GPEI) was launched in 1988.2 In recent years, significant achievements have been recorded (Box 1), but the target of global eradication is yet to be reached. The GPEI currently faces two main global issues. Firstly, addressing the eradication of WPV1 in Afghanistan and Pakistan, and secondly, dealing with the growing issue of vaccine‐derived poliovirus (VDPV).1 Vaccine strain virus can slowly accumulate mutations over time, which eventually result in reversion to neurovirulence — these strains are known as VDPV. Although extremely uncommon, this phenomenon becomes increasingly prominent in areas where there are long term low vaccination rates, allowing continued circulation of the attenuated poliovirus contained in the Sabin vaccine. Ironically, the modern prominence of VDPVs is a consequence of the GPEI's successful endeavours to reduce WPV. Poliomyelitis close to home VDPVs are appearing in areas with low immunisation rates in Africa, and recent emergence in closer neighbours puts poliomyelitis back on our doorstep. In 2018, there was an outbreak in Papua New Guinea involving 26 VDPV type 1 AFP cases, including a death.4 In late 2019, the Philippines reported 15 VDPV cases, and Malaysia reported three cases in 2019 and one in 2020.5 These countries had previously been declared poliomyelitis‐free.6 Between 2012–13 and 2017–18, the median number of annual arrivals for Philippine citizens to Australia was 141 813, with 8% of these arrivals being children younger than 15 years.7 Screening individuals at our borders is not an economically viable option to prevent poliomyelitis, thus highlighting the importance of optimal immunisation and high quality surveillance. Australia's commitment to World Health Organization targets Australia and all other Western Pacific region countries were certified as poliomyelitis‐free on 29 October 2000.6 As a signatory to the World Health Organization's International Health Regulations (2005),8 Australia reports annually on its compliance, with obligations to prevent and respond to acute public health risks of international consequence. This includes observing temporary recommendations issued when the WHO declared the risk of international spread of poliovirus a public health emergency of international concern in 2014, poliovirus containment activities, and reporting to the WHO Regional Certification Commission providing evidence that Australia's poliomyelitis‐free status has been maintained. This evidence requires Australia to meet WHO‐specified surveillance standards. The Australian National Enterovirus Reference Laboratory plays an important role in providing enterovirus testing and environmental surveillance for Australia and the Western Pacific region to meet these requirements. Environmental surveillance for polioviruses is costly and labour‐intensive and involves sampling sewage for detection and then characterisation of enteroviruses. There is currently inadequate capacity to routinely conduct environmental surveillance throughout Australia. Therefore, this capacity is currently directed at monitoring during high risk episodes; for example, when there is a cluster of AFP cases or after the importation of a confirmed case. The detection of any poliovirus in Australia is considered a likely importation event, as Australia stopped the use of the oral polio vaccine in 2005. Adequate clinical surveillance is based on two key WHO indicators. Firstly, achieving an AFP detection rate of at least one case per 100 000 children younger than 15 years. Secondly, the WHO requires enterovirus culture on two stool samples collected at least 24 hours apart, both within 14 days of onset of paralysis, for at least 80% of reported AFP cases.9 Submission of two samples ensures adequate sensitivity, required due to intermittent viral shedding.10 Meeting these targets provides national and international reassurance that there is timely investigation that excludes poliomyelitis as the cause of AFP. However, for Australian clinicians, awareness of this surveillance and its purpose is often not well understood. An overview of the AFP surveillance structure is provided in Box 2. Importantly, AFP cases need to be notified and investigated even if another diagnosis (eg, Guillain–Barré syndrome) is likely. Australia's performance in meeting World Health Organization targets While Australia has met the surveillance target for AFP notification for the past 11 years, we consistently fail to reach the WHO benchmark for stool submissions (Box 3).11 This is in marked contrast to many of our closest neighbours. Only New Zealand, the small Pacific Island countries and Papua New Guinea have a similarly low performance over recent years. In 2018, adequate stool collection was achieved in only 44% of Australian AFP cases and 2019 results are currently at 65%.12 The most populous states of New South Wales and Victoria consistently underperform, with rates of 33% and 42% respectively for 2018 (Bruce Thorley, Head of Victorian Infectious Diseases Reference Laboratory, Australia, personal communication, September 2019). In 2018, three cases of AFP and anterior horn cell abnormality on magnetic resonance imaging in young children were reported to WHO by Australia as “poliomyelitis compatible” because of a lack of adequate clinical information and appropriate stool sample collection (David Isaacs, Chair of Polio Expert Panel, Australia, personal communication, September 2019). In addition to providing robust public health surveillance, ensuring adequate investigation of AFP can produce relevant diagnostic information for an individual. A 3‐year‐old child with permanent significant disability following AFP in 2018 had the neuropathic enterovirus D68 (EV‐D68) in faeces sent for AFP surveillance purposes.13,14 The converse may also apply. Detection of a non‐polio enterovirus by polymerase chain reaction (PCR) in a clinical sample does not preclude the possibility of dual infection with poliovirus. Co‐infection and subsequent recombination of species C non‐polio enteroviruses with Sabin‐like poliovirus is an important precursor event in the development of VDPVs.15 Barriers to improvement A number of logistical issues affect successful stool sample collection; for example, late presentation of patients, discharge before sample collection, and constipation may all have an impact on stool collection rates.16 In some instances, pre‐examination by microbiology laboratories using enterovirus reverse transcriptase PCR (RT‐PCR) may occur. This does not exclude poliovirus infection and testing at the WHO reference laboratory is still required. Due to the extended viral shedding in the gastrointestinal tract, stool samples are the specimen type most likely to facilitate enterovirus identification. The collection of rectal or throat swabs is discouraged by WHO due to reduced sensitivity compared with faeces samples. Pragmatism may dictate that the former may be preferable to no testing at all in a particular child if barriers to faeces collection exist. Recognising poliomyelitis in a low prevalence community Cases of poliomyelitis present as acute and often painful weakness in affected limbs. The weakness is often asymmetrical, affecting lower limbs more frequently than upper limbs, with rapid onset and usually no further progression after 48 hours. Sometimes patients may present atypically, reinforcing the need for any AFP to be reported and investigated. Alternative presentations may include dyspnoea or dysphagia due to weakness of bulbar or respiratory muscles. Cerebrospinal fluid findings are suggestive of viral meningitis. There are usually no systemic symptoms, although a recent history of a mild upper respiratory tract infection with or without headache may be elicited.3 A history of exposure to a high risk area (eg, Central Africa, Pakistan, Papua New Guinea or Afghanistan) and/or lack of previous immunisation is important. A history of distant past immunisation will not exclude the diagnosis, particularly if this was received overseas.10 Call to action There appears to be a level of complacency among physicians due to the rarity of clinical poliomyelitis in Australia. In addition, there is a lack of awareness in the diagnostic chain regarding the importance of laboratory surveillance. In 2020, a comprehensive action plan was implemented by the Paediatric Active Enhanced Disease Surveillance (PAEDS) network to improve faeces collections across the country. Clinicians should not fear that they are being alarmist in notifying AFP cases that they believe have negligible risk of poliomyelitis. The emphasis on detection and investigation of AFP cases despite an alternative diagnosis may seem pointless for an individual case, but at a national level, it allows confidence in the integrity of surveillance and, ultimately, achievement of poliomyelitis eradication. Conclusion The recent VDPV outbreaks in Papua New Guinea and the Philippines and the ongoing WPV1 circulation in Pakistan and Afghanistan emphasise the possibility of poliomyelitis re‐introduction into Australia. Clinical acumen is unlikely to provide a timely diagnosis. Clinicians are reminded that poliomyelitis as a diagnosis should be excluded in all cases of AFP; faeces collection from all AFP cases independent of age should be viewed as a priority to ensure the country remains poliomyelitis‐free and as an opportunity to maintain surveillance, even when another diagnosis is confirmed or highly likely. Box 1 – Selected achievements relevant for Australia in the history of the Global Polio Eradication Initiative2 Year Milestone 2000 Australia declared poliomyelitis‐free 2005 Inactivated polio vaccine replaces oral polio vaccine in Australia 2014 South‐East Asia declared poliomyelitis‐free 2015 Wild poliovirus type 2 declared eradicated 2017 99% of poliomyelitis eradicated globally 2019 Wild poliovirus type 3 declared eradicated Box 2 – Schematic overview of acute flaccid paralysis surveillance structure in Australia APSU = Australian Paediatric Surveillance Unit (www.apsu.org.au); PEP = Polio Expert Panel; PAEDS = Paediatric Active Enhanced Disease Surveillance Network (www.paeds.org.au); VIDRL = Victorian Infectious Disease Reference Laboratory (https://www/vidrl.org.au/surveillance/afp-surveillance1); WHO = World Health Organization. Box 3 – Percentage of acute flaccid paralysis notification with adequate stool sample collection, Australia, 1995–2018* WHO = World Health Organization. * Data reproduced, with permission, from Roberts et al.11

Meryta May · David Durrheim · Jason A Roberts · Rhonda Owen