Then and now: Reasons why death rates are improving
SAN JOSE – The sickest coronavirus patients can live for weeks with a gripping headache, profound nausea, burning lungs, malaise, cough and waves of pain in their bones. They may be tethered to a breathing machine.
But eight months into the pandemic, fewer are dying.
New data reveals that while patients are still being rushed to intensive care units, a greater proportion are coming out alive. Since the pandemic began, the cumulative death rate for Californians with COVID-19 has fallen by more than half in the past three months. In early June, it was 5.87%; by Sept. 13, it was down to 2.14%.
What’s going on?
Some of the decline simply reflects a shift in testing, as infections in younger and healthier people are diagnosed. But that doesn’t explain all of it. There also have been fundamental improvements in how we prepare and care for the sickest patients, according to interviews with top medical experts.
“These declines in the case fatality ratios are striking,” said Dr. George Lemp, an epidemiologist and former director of the HIV/AIDS Research Program at UC’S Office of the President, who analyzed death rates using state data.
“We should applaud and appreciate the medical community for being able to find rapid ways to improve the outcome of this lifethreatening illness,” he said.
Here are six major reasons why the death rate is falling:
When the pandemic first hit, only people with severe symptoms were tested. Now expanded testing is detecting milder and earlier cases, so the prognosis is better, said UCSF epidemiologist Dr.
George Rutherford.
We’re also diagnosing more infections in younger people, who fare better.
Early on, we focused a lot of testing on outbreaks in nursing care facilities, where the sick and elderly face slimmer odds of survival. Now, fewer of this vulnerable population is getting sick.
As the patient mix has changed, so has the math, explained Rutherford. The denominator – the total number of cases – has grown faster than the numerator – total deaths. So the overall mortality rate is falling.
The declines have occurred across all age groups, during the four-month period between May and August 2020, according to Lemp’s analysis. This means everyone is doing better.
The phenomenon is driven by more than statistics, said Rutherford. “As we gain greater experience with novel infections, mortality goes down.”
Hospitals cite “the four S’s” needed for effective “surge” planning: staff, supplies, space and systems. Managing a patient on a ventilator, in particular, is a labor-intensive and delicate task.
During a surge of cases, hospitals in Southern California fell short on all four of these metrics, nearly hitting capacity. Some patients were intubated in emergency rooms instead of the intensive care units. Hospitals were forced to use older equipment, as well as doctors and nurses from outside hospitals who were less familiar with procedures and life-saving devices.
In the San Francisco Bay Area, hospitals were better prepared. And this planning has improved, over time.
The death rates of the regions are a study in contrast. For example, Los Angeles reports 57 deaths per 100,000 residents, while San Francisco has only onetenth as many, with about 6 deaths per 100,000 residents, according to data presented at UCSF’S Grand Rounds last week.
“None of our hospitals were flooded,” said UCSF infectious disease expert Dr. Monica Gandhi. Preparation
“allows an organized response. You have enough nurses. You’re only doing your job, not other people’s jobs. You’re not running around. Rooms are ready. You have PPE.”
“It is the chaos that can occur when hospitals are not ready that absolutely contributes to mortality,” she said.
Doctors now have a better understanding of how to manage breathing in severely ill patients, said Dr. Andra Blomkalns, chair of the Department of Emergency Medicine at Stanford Medicine.
Initially seen as a lastditch measure – and a sign of impending death – doctors now recognize the value of putting people on mechanical ventilation early, if needed, she said.
“We used to say ‘someone doesn’t quite need it yet, let’s see how they do in the hospital.’ That hasn’t worked well,” she said. “We’ve resolved that if they have to be on it, it’s better to put them on it earlier rather than wait until too late.”
We’ve also gotten better at fine-tuning ventilation, understanding the optimal amount of oxygen, pressure and time between breaths, she said. We’ve learned to be very gentle on the lungs.
Clinicians are also more skilled at deploying other tactics.
We’ve enlisted “proning,” where a team of caregivers gently roll a patient from their back onto their abdomen, said Dr.
Alan Chausow, chair of Pulmonary Medicine at Palo Alto Medical Foundation and medical director of the Critical Care Unit at Mountain View’s El Camino Hospital. When the patient is lying face down, it’s easier for the back of their lungs to expand.
“It’s not comfortable to lay on your tummy. But it definitely helps,” he said. “We’ve been proning people for other diseases for 10 years. We’re just taking that experience and adding onto it.
We’ve also learned to be especially vigilant in the prevention of other infections, Chausow said. In the ICU, for instance, use of a urinary catheter boosts the risk of deadly infection.
And now that research shows COVID-19 boosts the risk of lethal blood clots, doctors monitor blood more closely, and increase the use of preventive blood thinners, experts said.
“With every novel infection, as we gain greater experience and as we use better and better tools, the mortality goes done,” said UCSF’S Rutherford. “It was true of Ebola, HIV and a bunch of others. Now it’s also true of COVID-19.”
Use of drugs such as remdesivir and the steroid dexamethasone may be helping.
Remdesivir, authorized on May 1, shortens the recovery time for some of the sickest patients. While it blocks the virus from replicating, it’s unclear whether it’s actually keeping more people alive. In clinical trials, death rates are slightly lower – 7.1% vs. 11.9% – but this difference was not statistically significant.