Doctors are planning to use special software to give coronavirus patients a score and help decide who gets an intensive-care bed if there aren't enough. Nicholas Jones reports.
The first warnings emerged from China. Then Italian doctors sent texts, social media posts and emails. Now they come from America and the United Kingdom.
"We have colleagues all over the world, and they have been saying - weeks ago - 'This is really bad - this isn't flu, you need to plan now for it, please let your governments know'," said Dr Andrew Stapleton of the Australian and NZ Intensive Care Society.
"ICU has really been the canary in the coalmine."
The speed at which coronavirus overwhelmed a world-class health system in northern Italy jolted much of the Western world to attention.
At the epicentre was Papa Giovanni XXIII Hospital, a state-of-the-art facility in the city of Bergamo, Lombardy. Its 48 intensive-care beds and other high dependency wards weren't nearly enough for the influx of mostly elderly patients with Covid in their lungs.
Waiting rooms, meeting spaces and corridors filled with patients on gurneys. Intensive care beds were given to those judged to have a reasonable chance of survival. Others weren't resuscitated and died alone.
"The family is notified over the phone, often by a well-intentioned, exhausted, and emotionally depleted physician with no prior contact," doctors reported in the New England Journal of Medicine. "We are far beyond the tipping point."
Mirco Nacoti, a physician who wrote that letter with colleagues, now has a message for New Zealand: do everything possible to contain and corner the disease.
"Covid is a sneaky virus," he told the Herald by email.
"Asymptomatic/paucisymptomatic patients, most of them young, are the main carrier.
"When you find a severe case or a death you'll have at least 1000 people contaminated.
"Restriction measures are very important... but they are not enough - you need proactive surveillance and to implement a step plan of treatment."
Rampant infection means thousands need hospital care at the same time, he said.
"That's what happened in Bergamo."
Anxiety and preparations
New Zealand has two coronavirus patients in intensive care, has had one death so far and nearly 650 overall cases. The hope is level 4 restrictions will contain community spread.
"There is a reasonable degree of anxiety," said Stapleton, of the mood amongst the country's intensivists. "We have seen what happens if you don't plan for it. In New Zealand, we have many advantages - we have a good head start, we are far away so it's been slow to get here, and the Government has listened to the scientific advice."
New Zealand is not Italy; most infections are linked to overseas travel or known cases. However, Stapleton said it would be "extraordinary" if we don't face a situation where there aren't enough intensive-care beds needed for Covid-19 patients.
He's part of a group charged with working out how patients will be triaged, if and when that situation arises.
A key part of that process will be world-leading computer software, built here and which will give each Covid-19 patient a score, calculated from criteria including underlying medical conditions, and extremes of weight and age.
That number will then help ICU doctors decide who should get the next free bed, if there aren't enough for every patient needing intensive care.
"It is a diagnostic aid. It is absolutely not a computer making the decision between life and death," Stapleton stressed.
"You have one bed and four patients - who are you going to give the bed to? The tool will aid that decision, it will not make that decision."
Software to 'codify knowledge'
The "1000minds" software already has a huge and largely hidden influence on New Zealand life. Its roots stretch back to a momentous change in how people needing electives (medical or surgical services that aren't required immediately) are prioritised.
Limited health funding and a growing, ageing and sickening population made the size of waiting lists a feature of every election cycle. That changed from July 1, 1998, when a new clinical priority assessment criteria (Cpac) booking system began.
Patients were scored from 0 to 100 (lowest to highest priority), according to clinical and social need. Those over a certain threshold - set by each DHB according to capacity and demand - get surgery within months. Those below are sent away and told to come back if things get worse.
Waiting lists vanished, but controversy remained in egregious examples of desperately needy patients being declined, some of whom later died.
Those failures interested economist Paul Hansen, who eventually diagnosed the problem: the criteria being used to score patients was sound, but the weight put on each criterion wasn't.
Hansen and computer scientist Franz Ombler invented software that aimed to solve the problem, by presenting medical experts with a series of fictional patient scenarios.
When they decided who most needed treatment, the programme worked out how much weight should be applied to each criterion - perhaps having underlying health conditions was more problematic than old age for someone needing heart surgery, for example.
"It's basically codifying their expert knowledge," Hansen said. "The software is being trained by the experts, as to what they think is important."
Health bosses realised the benefit, and for almost 20 years the "1000minds" software has calibrated points systems used to decide who gets elective surgery, from cataract removal to hip replacement.
'They did this stuff like I would choose beers at the pub.'
That list will soon include Covid-19, thanks to public health doctor Melyssa Roy. She was working on research with Hansen and Trudy Sullivan, both University of Otago colleagues, before a new virus emerged in Wuhan. Her subject: resource allocation during pandemics.
"When the Covid situation emerged we changed the focus," Roy said.
Dr Craig Carr, the Kiwi chairman of the Australian and NZ Intensive Care Society, linked them with a group figuring out ICU triage for Covid-19.
Last Thursday a dozen top intensivists met online, opened 1000minds and ran through about 60 different exercises showing two different "patients", each with a few information points about their health. One might be older and underweight, for example, and the other at risk of organ failure. Who should get the next free bed, assuming all other factors are equal?
"That's a mental step," said Hansen. "But actually the clinicians are amazing at it - they did this stuff like I would choose beers at the pub.
"They answer, and the software throws up another question. They're basically voting. And the level of consistency was remarkable. Occasionally there were differences of opinion, which is great, because some of these things are very subjective, and the evidence with Covid is emerging."
The 1000minds Covid-19 tool is now "ready to fly", and will be sent free-of-charge to overseas clinicians, including in Italy, Australia and the UK. They can change and hone criteria according to their own clinical knowledge and local populations.
What factors will count?
Stapleton declined to go into detail about what factors will help score Kiwi Covid patients, but said some were obvious: "extremes of age, extremes of weight, extremes of chronic disease".
Māori and Pacific New Zealanders suffer worse rates of chronic disease, obesity and other measures of poor health. Could that mean they would be more likely to miss out?
Ensuring that isn't the case "has been forefront in our discussions the whole way through", Roy said. The working group includes ethicist and Māori representation, and wider consultation with Māori health leaders is planned. The ministry has the final sign-off.
"Because it is a consensus of expert opinion, it means that one person is less responsible for fully making that decision. In normal situations, there may be one doctor making that decision. And they may have some bias. It's intended to make it much more transparent and robust."
Meanwhile, work is under way to treble the number of ICU beds, which stood at fewer than 180 nationwide. Anaesthetists will be redeployed as "resuscitationists" on intubation teams, working under supervision and with indemnity coverage.
Stapleton said it was important to know if a Covid-19 patient couldn't get an ICU bed, it wouldn't necessarily be a death sentence.
"There are lots of work-arounds happening in hospitals right now... there will be access to hospital care and advanced hospital care, even if there isn't access to intensive care."
Putting someone on life support wouldn't guarantee their survival, Stapleton said, and must be carefully considered, given what a gruelling experience it is (about 7 per cent muscle mass is lost for every day spent on a ventilator, for example).
"It is not a decision we take lightly at the best of times, and it's slightly unfortunate the way it is being portrayed - 'Only ventilators will save you and if there aren't enough you are all going to die'. I think that's an unsophisticated way of looking at it."
Preparing for the worst, hoping for the best
Italian deaths approach 11,000, and the hospitals are still full. Exhausted staff drink a minimum of water, so as to not waste time removing protective gear for toilet breaks.
New Zealand's lockdown is on day six. Its effectiveness won't be known until week three, at the earliest. Elective surgeries have been postponed or sent private and hospitals sit half-empty, with tents put up outside to assess potential Covid-19 patients.
"We hope this isn't used," Hansen said of the triage software. "This is for a crisis; this isn't business as usual. This is for when patients are stacking up outside the door and there are only 10 beds available.
"Fingers crossed there is no resource constraint. But suppose that isn't the case."