The survival rate for people suffering from coronavirus has improved but doctors still want more effective therapies.
Tending to Covid-19 patients in the early days of the pandemic, Leora Horwitz felt like a doctor from the 18th century, desperately trying to discover more about a new disease to learn how to stop people from dying.
"We couldn't tell how quickly people were likely to deteriorate, what kinds of deterioration they would have . . . or when they were out of the woods," says the clinician-researcher at NYU Langone, an academic medical centre in New York. "We had no idea."
Last week, the coronavirus pandemic hit a haunting milestone, with an official death toll of 1 million worldwide — almost half of those in the US, India and Brazil. The rapid spread of the virus has triggered fears among some scientific and medical experts that millions more will die.
On Friday it was announced that Donald Trump, US president, had tested positive for the virus. Yet the death rate — how many people who have contracted coronavirus die — may be falling because of improved care, according to the World Health Organisation. Countries grappling with a new surge in infections hope that doctors have grasped how to keep more patients alive. The global survival rate for people hospitalised with Covid-19 has increased from 66 per cent in March to 84 per cent in August, according to the International Severe Acute Respiratory and Emerging Infection Consortium.
Dr Horwitz says the contrast between New York's wards in late March and today is like "night and day". Even accounting for the demographic differences between the patients being treated at her hospital, she found their chances of survival were 22 percentage points higher in August than in March in research that has yet to be peer-reviewed.
But some scientists remain sceptical about whether the death rate is actually falling, questioning the quality of the data. They argue that if it is coming down it has more to do with the rise in younger people getting sick.
"The biggest question that's out there right now — or at least one of them is [whether] the mortality rate from Covid-19 is actually dropping, or is it just apparently dropping," says Jonathan Slotkin, chief medical officer at Contigo Health, part of Premier, a group of over 4,000 US hospitals.
No 'silver bullet'
David Battinelli, chief medical officer of Northwell Health, New York State's largest healthcare provider, says it is easy for people to forget how rapidly the city's hospitals were flooded with patients.
"We went from zero in our health system to thousands hospitalised within 30 days [with] just about 1,000 patients in our intensive care units on ventilators. We had no effective treatment and almost an overwhelming number of patients," he says, adding that many other systems were swamped.
The hospital staff tried using what drugs they could: antivirals, plasma from recovered patients and steroids. If a patient improved, he says doctors could be "easily convinced" that this was a cause and effect, though subsequent trials may show otherwise.
The months since then have brought no blockbuster breakthroughs in drug treatments. Remdesivir is the only antiviral drug authorised in the US to treat Covid-19 yet its benefits are moderate: it can speed up a patient's recovery but there is no evidence it has reduced deaths. The Recovery Trial, run by the University of Oxford, discovered the generic steroid dexamethasone lowered deaths in patients receiving respiratory support and has now been widely adopted globally.
"I think dexamethasone seems to [provide] the biggest benefit that we have right now," says Amesh Adalja, senior scholar at the Johns Hopkins Center for Health Security in Baltimore.
Even without a "silver bullet" drug, Dr Horwitz believes there are "dramatic" differences in our understanding of Covid-19, especially the sheer havoc it can cause across the body, from the heart to the toes.
Patients can get diagnosed earlier and put on protocols based on what has been learned: when to give blood thinners, how to turn patients in a manoeuvre called "proning" — putting them on their stomach to improve oxygen intake — and when to hold off from pushing people on to ventilators too early.
Heather Pierce, senior director at the Association of American Medical Colleges, says these are "major steps forward". "We don't have a vaccine. And we don't have a cure for Covid. But the more we understand . . . be it using novel therapies, techniques, or supportive [care], the more we can help save lives," she says.
Outside hospital treatments, public health measures could be playing a major role in bringing the death rate down. In the US, about 40 per cent of deaths were linked to nursing homes, so measures to stop the spread inside those facilities are helping, says Dr Adalja.
The New England Journal of Medicine in September, published a commentary which suggested that mask-wearing could be reducing the severity of the disease. "For me that stands to reason," says Dr Slotkin, "but I don't think anybody can prove that it's true."
Others including Professor Sunetra Gupta, a theoretical epidemiologist at Oxford university, suggest that some regions may be reaching herd immunity at a much lower rate of infection than previously predicted. The theory is based on the idea that the virus appears to be much more likely to infect a subset of susceptible people — and that previous exposures to other coronaviruses may be protective for some. But Dr Battinelli dismisses this as "pure speculation".
Testing difference
Many dispute the idea that the death rate is falling, putting any apparent improvement down to more testing uncovering more cases and the better underlying health of a new wave of younger patients.
Laureen Hill, chief operating officer of New York Presbyterian Hospital, believes it is very hard to separate the factors that could contribute to lower death rates. She has practised intensive care medicine for 30 years and says her staff followed the usual playbook for Covid-19, so there has not been dramatic changes in how they look after people.
Instead, the difference between the rates of death in the early months and now may be down to a better understanding of how many people are infected. A death rate is the proportion of confirmed cases who die — so the count of confirmed cases is key. "Early on we weren't testing nearly as many patients as we were today. So when you look at rates, that's very much affected by the denominator, how many tests are there being done," she says.
The data is patchy and poorly monitored in some countries and hard to compare across regions and borders. Even counting a death is sometimes not simple, especially if the patient had other diseases or died at home. Fatalities also occur weeks after an initial infection.
In the US, the Centers for Disease Control and Prevention has changed how it reports the Covid-19 death rate. In July the overall rate was 0.65 per cent. Now it breaks down the data by age with the most likely to die — the over 70s — having a 5.4 per cent death rate. Douglas Rothman, a professor at Yale School of Medicine, is adamant that, when accounting for the age of patients, death rates have not fallen at all. He calculates that the September US death rate was about 0.69 per cent.
In his own study in Arizona, one of the sunbelt states hit by a wave of Covid-19 infections in summer, he found that, when adjusted for age, the death rate for the population up until the end of July was about the same as the national estimate for the spring of 2020.
He accuses some doctors of "false optimism" and a "sunshine effect", where they are inclined to focus on their successes, and argues there needs to be more independent research.
The people getting infected and hospitalised now tend to be younger, as the older or more vulnerable are more likely to take protective measures like social distancing. Yoko Furuya, New York Presbyterian's medical director of infection prevention, says the younger population may obscure any medical advances.
"We're seeing a major shift in epidemiology that may drown out some of the smaller changes that come from the treatments that we now learn are effective," she says.
'We don't have drugs for viruses'
While Covid-19 cases are again rising in many European countries, and in more than half of US states, some doctors are hoping that new treatments will help save more lives.
Monoclonal antibody treatments could help patients even before they get to the hospital. Eli Lilly recently released positive phase 2 results, showing the treatment — artificially developed from the best performing antibodies from recovering Covid-19 patients — reduced the rate of hospitalisations.
"I'm cautiously excited about the promise of monoclonal antibodies, because preliminarily it looks like they may show some promise treating people who just have mild symptoms and it may help prevent them from getting sicker," said Dr Furuya.
Dr Horwitz is less hopeful that researchers will discover an antiviral that could stop the virus from replicating.
"We don't have drugs for viruses," she says, adding there had only been two viruses that have been tackled successfully: Hepatitis C and HIV, which has been kept under control but not cured. "After 100 years, we don't have a treatment for influenza that's effective."
Surges risk bringing back the threat of higher death rates because of overwhelmed hospitals, especially if they are trying to cope with Covid-19 at the same time as seasonal flu — and possibly even in the same patient, as it is not yet known if the conditions can coexist.
A vaccine could make a significant dent in the Covid-19 infection rate — and, therefore, the absolute death figures. But even if a vaccine proves successful and is approved on an emergency basis in the next few weeks, it is unlikely to be widely available until next year.
It was unfathomable this time last year that a new virus could make doctors feel as lost as their ancestors were when wrangling with long subdued infectious diseases.
And yet Sars-Cov-2 has not only taken at least 1m lives, but is forecast to take many more. Models struggle to predict future deaths more than four weeks ahead — but the CDC says the average of about 44 models expects between 2,700 and 8,600 deaths in the week ending October 24 in the US.
One model does try to foresee as far as the start of next year. The Institute of Health Metrics and Evaluation at the University of Washington predicts the death toll could hit 2.5 illionm by January 1, a number that could be brought down to 1.8m with universal masking — or rise to 3.3 million if restrictions are further eased, it says.
"This virus is not done killing people yet," says Dr Adalja.
Written by: Hannah Kuchler
© Financial Times
