Swapping salt for a heart healthier option

ABC Listen. Podcasts, radio, news, music and more. What if living offline isn't a choice anymore? I'm Rae Johnston and in Opt Out, I'm finding out just how much of our lives are online. and whether we can ever really unplug from privacy experts to hackers, we explore what it means to disappear in a world that's always watching.

How do you feel about adding salt to your food? Are you one of those people who sprinkles it on every meal? God, you've got to ask me these questions where this is about do as I say, but not as I do. You're sprinkling. I'm terrible with salt. I am truly terrible with salt. This is making people feel better, I suppose, about themselves. I have a salt pig in this kitchen. Do you know what a salt pig is? Yes, I do.

it's not designed for salt restriction and the other thing it's not designed for salt restriction is I'm addicted to the New York Times online cookbook. And there's a lot of salt in there. I'm telling you, these American recipes, one reason why they taste so good is that they're a complete shite for your health. Butter.

But delicious. And food is about more than just nutrition. It is. But I do feel guilty about my salt and I do try and cut it back. And I don't tend to add salt. I don't add salt at the table. It's cooking where I add it. Well, that's because, and we're both very aware of the fact that salt can increase the risk of high blood pressure, contribute to heart disease, kidney disease, and now there's talk about salt substitutions. Yeah, and maybe autoimmune disease as well. But...

Today we're going to be covering on the Health Report salt substitutions. Is there any merit there and is that what should be in your salt pig instead? I'm increasingly thinking so, but that's coming up. On The Health Report, I'm Norman Swan on Gadigal land. And I'm Priya Alexander on Wurundjeri land. Also on the show, genetic carrier screening is a test that should be offered to everybody, either preconception or in early pregnancy, but there's still many people who don't even know about it.

Artificial intelligence is everywhere. Some doctors think it might help people make decisions for their loved ones at their end of life. Would you trust it? In this summer series of The Health Report, we are reflecting on the year that has been and...

We are covering today the Make America Healthy Again or the Maha movement. And it's very hard to know where to begin. It's a very complicated movement because you'd think that it's all hardcore Republican voters. It's not. It's a group of people with kind of a mixed view of the world, where they're into natural foods, organic farming. One of their influencers does a lot with how she lives her life in a very natural way. And that's kind of what they are pushing for from Robert F. Kennedy Jr.

But because it's a mixed movement, he's kind of picking and choosing what he takes from it. And there's been a lot that has happened this year. So I think it's probably appropriate to go all the way back to February when RFK Jr. was. confirmed as the Department of Health and Human Services head. That was in itself quite the journey that we went on together.

We did. And with that came lots of changes in February. And perhaps one of the biggest ones was the dismantling of USAID, which is the US Agency for International Development. Which wasn't done by him. It was done by Trump. Yes, but the dismantling of US aid has had significant consequences globally. And USA, just to put it in perspective, has prevented more than 75 million deaths worldwide from conditions like malaria, tuberculosis.

There was a study published late this year in the Lancet Child and Adolescent Health, which actually tried to estimate what the consequences would be to the US cuts in funding to global health. And it actually has projected an additional 2.5 million paediatric TB cases, tuberculosis cases in low and middle income countries.

and estimated there's going to be an additional 340,000 paediatric tuberculosis deaths. And the other program they're concerned about amongst many is PEPFAR, which is the program that gave... either cheap or free antiretroviral AIDS drugs to low-income countries. Huge consequences. In February, around the same time, there was also the government announcement to slash funding from the National Institutes of Health. So they cut 6.4 Australian...

billion worth of funding in research overhead costs, which had consequences for not just researchers in the country, but also here in Australia. So there were suspended projects on malaria, tuberculosis and women's health as a result. They've reinstated some of the funding, so a lot of the funding just got stopped from the DEI, the whole thing. They didn't want funding for woke projects.

And a lot of funding got stopped or screened. A lot of that's been reinstated, but it's been a very disruptive year for both international and U.S. research. And on that note, earlier in the year, there was actually the removal, if you recall, we talked about it on the health report at the time, of federal resources in the spaces of contraception, gender-affirming care, STIs and HIV.

where suddenly resources on federal government health pages went either missing, they vanished completely, people were trying to screenshot the information so that it wasn't lost forever. And it also had implications for tracking of infectious diseases. It has significant global impacts. Staff reductions at the Centre for Disease Control and the sacking of their Vaccine Advisory Committee. Yes, in June.

Do you know I did a bit of digging, Norman, because things were coming back to me. But if you recall back in February when RFK Jr. was being put forward as the candidate, Senator Bill Cassidy. Republican. yes, had expressed some caution about RFK Jr. and had said, I'm not going to pass the vote through. At the time, RFK Jr. actually said to Bill Cassidy that he would not touch... the Advisory Committee for Immunisation Practices.

panel of 17 sitting members, immunologists, paediatricians, physicians, these expert group who make recommendations on vaccines. And in fact, in June, RFK Jr. completely unprecedented move sacked all of them. citing that they were all biased, that they all had conflicts of interest, and replaced them with some interesting characters. Some of whom have a track record, as indeed does RFK Jr., as anti-vaxxers. One of their first actions was to limit the access to COVID-19 vaccines.

and create a lot of uncertainty, which is more recent, of childhood vaccines, because they're obsessed with additives such as aluminium, which, by the way, does not exist in childhood vaccines in Australia. and separating out vaccines rather than giving mixed vaccines. And again, there's no evidence for that. And there aren't products on the market or easily obtainable products on the market which separate out these vaccines. So America is going to see plummeting vaccination rates.

And in fact, they have. They've had a huge outbreak of measles in Texas. They've also had an outbreak now in South Carolina. So it is having implications on the ground. It's fueling a lot of vaccine hesitancy. As if that wasn't enough, though. In August, the CDC director, Susan Menares, was fired after she refused to resign. Yeah, because she refused to rub her stamp.

what she felt were non-evidence-based policies. And when she was sacked, nine former CDC directors published an op-ed in the New York Times and it was a scathing criticism of the... public health undermining that was happening by the Trump administration. And it was something that had not been seen before. But you can just get a vibe of what is happening in the US. And it has had ripple effects, I think, even here in Australia. Would you agree?

Yes, I think that the sovereign nation movement has taken on a lot of the flavor of Maha, which is about control. government control, telling people what to do about their health, insisting on vaccinations and so on. So there is leakage. And the more something's said, the more it must be true. But some people would argue... Priya that he's doing some good like for instance food diets removing food diets

Yes, and the original push of the Maha movement was really around processed foods and going back to whole foods and these lifestyle. And he talked a lot about chronic disease. Yes, it sounds good. It does sound... Except that chronic disease... This concentrates in disadvantaged populations who are black. And that's what's being cut in terms of research. So they give with one hand and take away with the other. That's right. And I think...

RFK Jr is not going as hard on the food industry as many had expected. And so there is a push from him to voluntarily phase out synthetic dyes. He's asked the food industry to do that voluntarily. Let's see. But yes, I agree with you. It's the people who probably need these interventions the most who are now lacking them. So the divide will probably worsen in terms of health outcomes. On ABC Radio National, you're with the Health Report.

So salt this year was a big topic of conversation on the program. And we've covered before this issue of salt substitutes, where you substitute sodium chloride for potassium chloride. And preliminary research showed that this really has got an effect. And earlier this year, there was a big study from the George Institute in Sydney, which is called the Salt Substitute and Stroke Study, which looked at what substituting potassium for sodium...

does to the risk of recurrent stroke. And the person I spoke to was Professor Bruce Neal, who's the executive director of the George Institute, and about some of the do's and don'ts of salt substitution. What this trial did, or what this analysis did, was looked at people who'd had a... a prior stroke, they had people who were at high risk of having another stroke.

We know already that if you give them drug-based blood pressure lowering, you can reduce that risk. And what we showed very clearly here was that if you gave them potassium-enriched salt, which lowers blood pressure, and that's almost certainly how it had its benefits, then it also protects against the risk of having another stroke. Okay, let's just take a couple of the themes in terms of the response we've had to the health report. One is...

I'm on a diuretic which spares potassium and it tells me to watch my potassium intake. So in other words, can everyone substitute potassium for sodium or are there groups of people where they're taking a risk? Yes. So this is obviously a really important point because if you have too much potassium in your blood, that causes you problems. And if you have too low potassium in your blood, that also causes your problems. And these problems can be severe because...

Too high or too low blood potassium levels can cause arrhythmias of the heart, so the heart beats irregularly, and in the worst case, it could lead to death. So if you're...

already doing something that causes you to have high blood potassium levels, then you would want to think twice about switching to a potassium enriched salt. And the people who are at risk of having too high blood potassium levels are people who are taking... potassium supplements, people who are taking a particular type of diuretic that spares potassium.

And also people who've got serious kidney disease, because if you've got serious kidney disease, your kidneys can't maintain blood potassium levels as effectively as they normally would. And people taking potassium supplements are also people on diuretics to replace the potassium they might lose in their wee. Absolutely. And look, one of the interesting observations about this debate is that everyone focuses on the high blood potassium levels.

we know that actually one of the most common side effects of taking blood pressure lowering drug therapy is to get low blood potassium levels because most people include a diuretic in that therapy and mostly that isn't a potassium sparing diuretic and so one of the areas that we're looking to explore is actually whether co-administration of

potassium-enriched salt with a regular blood pressure lowering therapy might actually be an effective way of preventing people from getting low blood potassium. So let's assume you're in a regular situation and you're interested in low your blood pressure by manipulating your salt intake the common refrain we've had from the audience is where the hell do you find this stuff yeah and that is definitely a problem

You can actually get potassium-enriched salts in 50 or more countries around the world, but in no country would potassium-enriched salt constitute more than 1% of sales of salt. And that's because people don't know about it and because it's not on the supermarket shelves. You can get it in Australia, probably only the bigger supermarket shelves, and probably in the health food aisle rather than in the salt aisle. And you can also buy it online by searching for low-sodium salt or cassian salt.

So what's the strategy somewhere like Australia? Obviously, you want to see some degree of implementation of your findings in the Australian context. Ideally, we'd like to see all salt in processed foods and all salt that you add at home or when cooking.

switched for potassium-rich salt. But in the first instance, our strategy is to go for people who are being managed for hypertension and try and get a recommendation into the guidelines which says everyone who's being managed by a clinician for hypertension should

switch to potassium rich salt unless they've got a good reason not to and in parallel we want to get potassium rich salt on the supermarket shelves next to regular salt so that those people can find it and the reason that we've gone with this clinical hypertension management approach in the first instance is that we can deal with that risk side of things. Because if you're seeing a GP or indeed another health professional,

they're not going to recommend that you switch to potassium-enriched salt if you're using a potassium-sparing diuretic or if you've got kidney disease. And so that's the sort of compromised position in the meantime. Professor Bruce Neill, who's Executive Director of the George Institute for Global Health. I do tell a lot of patients that salt sneaks into your diet in very interesting ways and it's worth going onto a site like Live Lighter, checking out their food label chart and looking...

at everything, marinades, sauces, whatever you're cooking with. And it often accompanies sugar as well. Win-win. Flash the sriracha sauce on your food. I love sriracha. So do I. We can be friends. Let's not go too far. You're on The Health Report here on Radio National.

Preconception consults, Norman, are one of my absolute favourites. I will declare that. I love when a patient comes in and says they're planning a pregnancy because it means we can do lots of things to optimise health for both mum and baby. And one of the things that lots of people are not aware of is genetic carrier screening. A test that's available, either a blood test or a saliva swab.

before pregnancy or in early pregnancy to see what the chances are of them passing on inherited genetic conditions like cystic fibrosis onto their baby. So what's the purpose in early pregnancy when you're already pregnant? So it just means that if you want to pursue further diagnostic testing of your baby, you can if you want to consider all of your options or potentially plan.

Ideally, we like to do it preconception so that if we pick it up before you're even pregnant, it opens up the door for IVF and other options and extra testing so that you can potentially avoid inherited genetic conditions if you want to, if both you and your partner are carried. And so for some of these, it's a one in four risk if it's something like cystic fibrosis or hemochromatosis.

But there are some which are called autosomal dominant, where it's a one and two risk. And that can make you decide what you might want to do in the future. That's right. And there are serious implications. So it's, you know, a lot of people haven't heard of this testing. It is available. There's three tests that are covered under Medicare. So it's bulk built. You can do an extended panel test for hundreds and hundreds of conditions at a cost.

But Rachel Casella is a genetic carrier screening and reproductive health advocate, and her daughter Mackenzie was diagnosed with one of these inherited genetic conditions, spinal muscle atrophy, at 10 weeks of age, and she died tragically at 7 months and 11 days old.

due to the condition and Rachel has shared her story here. At 10 weeks old, our daughter Mackenzie was... diagnosed with a genetic condition called spinal muscular atrophy, which meant that she was going to lose the ability to move her arms and her legs and then eventually lose the ability to swallow and to breathe.

We were told that we would likely have her to around about eight months of age and that happened overnight. We went from being in this beautiful bubble of... complete perfection and then all of a sudden we were told that we were going to lose our baby. The world just completely crumbled for us and we just wanted to know why because we found out that a simple saliva swab could have told us that we were at high risk because both my husband and I are carriers of SMA.

And that meant that any children that we had would have a 25% chance of being affected. And we were just so mad that this simple test was available, yet it wasn't regularly offered to people. It was only offered at that time to people who had a family history. Usually it was through private testing. It was very expensive and we just thought this needs to change.

So this led Rachel and her husband to set up Mackenzie's mission, which advocates for genetic carrier screening and has undertaken research into this. Professor Edwin Kirk took part in this as well. He's a genetic pathologist at New South Wales Health Pathology and joined... and Rachel in the studio.

The reason it's really important for people to be offered carrier screening is there are many, many genetic conditions that people first learn about when they have an affected child because you can be a healthy carrier and in fact all of us are healthy. carriers generally for at least one genetic condition. And the purpose of carrier screening is to give people information about their carrier status. It gives you options that you wouldn't otherwise have.

And those options are a bit different if you find out before you're pregnant than if you find out during a pregnancy. But either way, the knowledge is power. and it's power that you don't have without the test. It also gives options of early interventions, like early treatments. It's not solely about family planning, about IVF, about unfortunately saying goodbye to a pregnancy, which we had to do.

two times after we lost Mackenzie, but there are actually early interventions. There are treatments for some of these conditions. The conditions that we're talking about are severe and life-threatening. But what everyone does with that information, it could be different based on their circumstances. But like what Edwin said, it's all about choice.

Edwin, do you just mind explaining what genetic carrier screening actually looks like for someone who's undergoing it and what the options are and the potential costs? It can be done mostly either by a cheek swab or by a blood test. I guess the key point is being able to access screening. At the moment in Australia, we have funding for screening for three conditions, spinal muscular atrophy, fragile X syndrome and cystic fibrosis.

And that test's been available for about a year and a half now, and it's running at more than 100,000 women a year accessing the test. So that's huge. But it's still fewer than half of all the people who might be eligible and for whom it might be useful information. So there is still a bit of a gap there. And that's a free test. That's Medicare funded, no out-of-pocket cost. That's right, yes.

Then there are other options that vary quite a lot, but involve testing for a larger number of genes, typically a few hundred, but in some cases more than a thousand genes. Most of the tests involve testing one partner first and then only testing the other partner if the first one who's tested, usually the woman, is found to be a carrier.

But there are some options that test both at once. And that's the model that we used in the McKenzie's Mission Project. So, Rachel, you and your partner were both detected as being carriers for the same problem. What can people expect if they do this test, both them and their partner come back as carriers? What does the road look like?

can be really quite varied depending on what the condition is and their beliefs and their situation. So for us, SMA is a life-threatening condition, although in recent years there have been advancements in gene therapies. That's why it's important to get the information from the right person being a genetic counsellor.

We sat down with the genetic counsellor, spoken through our options, given a bit of an understanding of what genetics is and how both of us being carriers meant that we had an affected child. For future children, we were told that we could either get pregnant naturally, and then we could test in pregnancy through a CVS or an amniocentesis to see whether that baby was affected. Then we could make a decision based on that, whether we were going to...

terminate the pregnancy or whether we would continue with the pregnancy and try early interventions. We were also told that we could go through the path of IVF so we could get embryos and then test them to check whether they had the... genetic condition and only transfer healthy embryos. But for others, it can be having that baby. If there's early treatments, it can be based on whether they're different beliefs of quality of life because severity is very different.

some people's eyes versus others can be based on their family support, financial factors. So there are a few different avenues available. So I think that's really important, you know, that there is hope. Edwin, you just mentioned before that there's still a lot of people who are potentially not accessing this service. Why do we think that's the case? Is it that people aren't being offered it? Is that people aren't...

Aware of it is an issue with health professionals. Where's the problem, do we think? We don't know that there's a problem. It's possible that everyone is getting information and some people are choosing not to access screening. I think it's more likely that there's still a gap, that there are quite a lot of people who just aren't hearing about it or hear about it but don't understand the importance.

From my experience, nearly every person that comes to me that hasn't done the test didn't know about it. They weren't offered it. There were quite regular instances of going to their GP and their GP not knowing. about carrier screening or still believing the old guidelines that it was only supposed to be recommended or offered to people who had a family history. There's a bit of confusion that the public have when they confuse it with the NIP.

test that's done at 10 weeks. They think that that's the genetic test that we're talking about, which it isn't. It's sort of a wide variety of reasons, and that's why we try to educate not just the public, but also medical professionals, making people realise that even if you've had one healthy child, that doesn't mean that your second might not be affected. There's just...

a lot of avenues for misinformation out there. And is the hope that more conditions will be included in a Medicare funded test? Does that seem likely? Is that what's going to happen in the future, do you think? The three gene screen is a really great start. And I think it's fantastic that the government have had the vision to do that when, as I've said, hardly any countries around the world have done anything similar.

And obviously we think the next step is to expand carrier screening to include many more conditions. And we hope it might be on the model that McKenzie's mission has set because we've built up the evidence for that. That was Professor Edwin Kirk, who's a genetic pathologist at New South Wales Health Pathology, and Rachel Casella, genetic carrier screening and reproductive health advocate. On Radio National, you're with the health support. End of life care.

can be really quite tricky. And people may have heard of advanced care directives, but only 14% of older Australians actually have one in place. Whilst they can help guide end-of-life care, there can be some complexities and often all situations aren't.

covered. It's impossible to predict. Like, are you going to accept antibiotics if you've got sore ear? You can't necessarily think of all of it. And this is where artificial intelligence may have a role. Who knows? Our producer, Shelby Traynor, has looked into this, the role of AI in this space. And as you alluded to earlier, Norman, there are some convincing arguments here. I think it's really important.

Having someone that can speak on your behalf that knows how you'd like to live your life and what sort of things you would or wouldn't want at the end of your life. Nicola Champion is more familiar than most people with the end of life. She's a palliative care nurse who also took care of her dad in his final weeks in their hometown of Port Pirie. Dad was an interesting character, very independent, very strong views on things, great sense of humour.

Nicola's dad Charlie had prostate cancer, but he'd lived a relatively healthy and independent life until two years after his diagnosis. when he arrived on Nicola's doorstep out of sorts. And I opened the door and he said, I thought you were going to look after me when I got sick. And I was really taken aback because I thought...

Well, yeah, I will. It turned out he was in a lot of pain. Charlie's cancer had spread. He would have six weeks left if he didn't do anything, six months if he did. And my dad just said, I'll take the six weeks. He was just so pragmatic and he was just like, well, that's it. That's my lot. I've got six weeks. Charlie had an advanced care directive that Nicola helped him write out.

An advanced care directive lets doctors know what you do or don't want at the end of life. Would you want to be given antibiotics? Would you want a blood transfusion if it could extend your time? I don't remember so much the conversation, but what I know is that he fully trusted me. So we didn't go down to every scenario that he might want to consent to or not consent to. He just...

appointed me as his medical decision maker. Surrogate decision makers are people you legally designate to make decisions for you in the event you become incapacitated. So for me, it was about really knowing my dad's values in life and the way he liked to live and the way I think he wanted to die. Not everyone will have a surrogate decision maker.

And even if they do, those surrogates aren't always available in an emergency. And they don't always feel equipped to make tough calls. It can be incredibly stressful to be the voice of someone who can't speak for themselves. It happens so frequently in the ICU where patients are critically ill and can't make decisions for themselves. And you see their family and their surrogate decision makers struggling with the burden of that choice.

This is Dr Teva Brenda, an internal medicine resident at the University of California, San Francisco. He saw this situation play out almost daily. The patient has a breathing tube, they can't speak. or they're so ill that they're confused. And every decision becomes this huge pivot point. Do we continue antibiotics? Do we pursue this procedure or this surgery? Dr. Brenda and his mentor wondered if there was a better way. a less stressful and more accurate way

to come to these decisions. And so we were just chatting together and AI has obviously been in the news since 2020 with ChatGPT. And we thought, well, how could we leverage artificial intelligence to help surrogates? Because this is... a common problem in the ICU. They decided to put an idea out there, drafting a paper that theorized how generative AI, or large language models like ChatGPT, might be used.

Eventually, they brought in a geriatrician and palliative care doctor to get his perspective. It's kind of funny. His first reaction to our proposal was, heck no, that's frightening. That's dystopian. Do we really want to be considering this? But Dr. Brenda and his colleagues are far from the only people thinking about this. My name is Brian Earp, and I'm an associate professor of biomedical ethics at the National University of Singapore.

At the same time AI was having its first of many moments in the sun, Dr. Earp was working as an editor on the Journal of Medical Ethics. In that journal, experts were trying to come up with better ways to deliver end-of-life care in line with a person's wishes.

One proposal that was gaining attention was called a patient preference predictor. The idea here is that you would do a big survey of the population, you would give people various scenarios that they might encounter, and you would ask them, what would you like to have happen if you found yourself in... this situation, this situation, this situation and so on.

And then you would also collect various demographic features about people, their age and their sex and their ethnic background, their religious affiliation, maybe their socioeconomic status. So instead of everyone filling out an advanced care directive, a sample of the population would fill one out.

an extensive one. It was assumed that people with the same demographic features would make similar decisions at the end of life. This seems to solve some of the problems, but it also is met with a lot of criticisms. The biggest one? People don't like being reduced to their demographic features. We like to think of ourselves as unique, not just in age, sex and ethnicity.

Dr. Erb was watching the debate about this preference predictor play out at the same time he and his colleagues were playing around with AI. My then housemate was also my friend and collaborator, Sebastian Porstam Mann. He realized there were these interfaces. you could gain access to or you can further train a general model in a process called fine-tuning. Basically, you can tweak it to fit a specific purpose. They ended up feeding their AI model dozens of their own research papers.

The purpose was to teach the model to recognise the relationship between a paper's abstract, which is a short summary of the research, And the research itself. And once it's learned that relationship, you can put in a new title and a new abstract, press go, and in a matter of seconds, it will just generate a draft in your voice using your style of reasoning, drawing on the kinds of arguments that you've used in the past.

past but applying it to this new topic. This proved that the AI was able to learn Brian's views and apply that knowledge to new situations. So he thought, what if it could be used to learn your views on things? For example, whether you'd want to be put on a ventilator if it gave you six more months of life. There was a problem though.

I work professionally in philosophical ethics, so I write dozens of papers explicitly stating what my views are on various topics. And so it's not a big surprise that this sort of model can come up with a reasonable guess about what I might say morally about some situation.

including a hypothetical situation in which I were incapacitated. So if you haven't written dozens of papers on medical ethics, who has? How would an AI model learn about your preferences? Well, one way could be through social media. There already are existing digital duplicates of us out there in the world that are owned by technology companies. So there's a digital twin of me that Amazon owns.

And there's a little twin of me that Facebook owns and that Twitter owns and so forth. And they use these to predict my preferences in particular domains, namely so that they can sell me stuff.

Another approach would be to train AI on your medical records. Some of these things really do seem like they might be far-fetched, and you might think that you'd have a much better go if you focused on...

specific prior treatment decisions people have made. So it's one thing to just be chitchatting with my doctor, but if I... train a model on my electronic health records, for example, which show all the various other decisions I've made, there's at least a question about the extent to which I can extrapolate from those kinds of decisions to novel cases that I won't have yet encountered.

Of course, this brings up a whole host of privacy issues. But imagine this AI could find out your vaccination status or whether you went for a more aggressive approach to cancer treatment. Dr. Earp has even suggested training the AI not passively, using data collected from the internet or from your medical records, but directly. You could take charge of building your own digital duplicate.

Whether an AI should therefore make a decision is a different question because it's an ethical question, it's not a technical question. And this raises another problem, one that AI may or may not solve. Nicola was familiar with the healthcare system. She trusted she could make the right choices for her dad. But even with her expertise and an advanced care directive in hand, she struggled to get doctors to listen.

I just couldn't get anyone to listen to me that dad was told he would have six weeks. He accepted that he was terminally ill and was going to have six weeks, that he wanted to be back in the country town where we were from. And I really had to fight hard to try and get him home.

It's not good enough to have an advance key plan. People have to respect it. And they have to believe that if someone's been named as a decision maker, that that person didn't do that lightly. Nicola thinks that if there had been a digital duplicate of her dad in the room that day...

it might have swayed the doctors. If they'd asked the AI, OK, what's important to you? He would have been able to say, I don't want to fuss. I want my daughter to look after me. But as Dr Earp points out, that all rests on AI being proven accurate. and being trusted. And for that to happen there needs to be a level of transparency around how the AI reaches the conclusions that it does.

Suppose the model says it's likely, maybe with 80% confidence, that John would want to have treatment withheld in this situation. And some of the major evidence I'm using for this claim is that John wrote an email on January 22nd of 2015, explicitly saying to his friend, Bob that if he ever was unable to feed himself, he would definitely not want to live under those conditions.

Well, that would be like explicit evidence that people could then evaluate. They could say, oh, well, let's go check with Bob and confirm that that's true. And oh, OK, so that seems like a pretty strong expression of his values. And I haven't ever heard him say anything otherwise. So maybe we should go with that.

You know, it might be that the person's spouse is there and asked to make the decision and the model comes up with some reason and the spouse says, look, that's not anything that actually reflects, you know, the John I know. And then, OK, now you have a difficult discussion to make and you have to do some further investigation.

There's a long way to go in proving AI decision makers are accurate and finding out whether people would ever actually trust them to help make life or death decisions, which is why I asked Brian what he would do.

I have thought about this. I mean, a lot depends on how predictively accurate these models turn out to be once we've trained them and tested them. And so right now I sort of shrug my shoulders and say, well, it kind of depends. If it turns out that they're far more accurate on average than human circuits.

then I'd be more likely to want to have it used in cases where I lost capacity. Whether I would want to use it if a loved one lost capacity and I was the proxy decision maker, I suppose that my sense of curiosity would be great enough that I would be very interested in what the prediction was. But then I would also want to know that I could trust myself enough to not be unduly swayed by the prediction if I know that I have good reason to trust my own judgments about the person.

Nicola looks back at her time as her dad's carer and is glad she could be his voice in his final weeks. If it weren't for her, it's possible Charlie would have died in hospital rather than back at home. I felt I knew what was important to my dad and I feel like top of his list was me to care for him and my understanding was for that to be in our home. He never specifically said that but I still think about him turning up to my front door.

You know, I thought you'd look after me when I was sick. And she only felt reassured about her decision when they arrived back in Port Piri. He had the biggest smile on his face as we wheeled him in that back door. And it's something that... I will always treasure because I just thought I know I made the right decision. That was Nicola Champion finishing off that story by Shelby Traynor. And earlier you heard from Dr. Brian Earp and Dr. Teva Brenda. On Radio National, you with the health report.

That's it for The Health Report this week. But don't forget our sister podcast, What's That Rash?, where we answer the health questions everyone is asking. And this week we're answering the questions about sunscreen. There's been a lot of them. controversy this year about sunscreens and their SPFs and a lot being taken off the market.

Find What's That Rash on the ABC Listen app for the comprehensive sunscreen spectacular that you need for the summer. Lots of alliteration for you. See you in your salt peak next week, Norman. Yeah, absolutely. Bring it along. before I brine the turkey. You've been listening to an ABC podcast. Discover more great ABC podcasts, live radio and exclusives on the ABC Listen app.