Recently, National Public Radio’s All Things Considered presented a news story entitled, “New Data Shows That Patients On Ventilators Are Likely To Survive”. It was following up on a story from April, presenting an earlier dataset from New York City that indicated an alarmingly high death rate (88%) among COVID-19 patients who were placed on ventilators.
Listen to the story. If you prefer, or for later reference, relevant parts of the transcript are at the bottom of this question.
Optional extra source: If you wish for more details to answer any questions below, you may also refer to the original New York study.
A) (6 points) The first New York study aimed “to describe the clinical characteristics and outcomes of patients with COVID-19 hospitalized in a US health care system.” They obtained data on 5,700 “patients with COVID-19 admitted to 12 hospitals in New York City, Long Island, and Westchester County, New York, within the Northwell Health system. The study included all sequentially hospitalized patients between March 1, 2020, and April 4, 2020, inclusive of these dates.” They recorded “outcomes during hospitalization, such as invasive mechanical ventilation, kidney replacement therapy, and death. Demographics, baseline [conditions], vital signs, and test results were also collected.” What kind of study was this, and what kind of sampling scheme was used to select the sample? Briefly justify your answers citing details/characteristics from the study description.
B) (7 points) The original New York study said that “for patients requiring mechanical ventilation (n = 1151), 38 (3.3%) were discharged alive, 282 (24.5%) died, and 831 (72.2%) remained in hospital.” (This gives the mortality rate of 282 / (282 + 38) = 88%). For this study, what do you think are the population parameter of interest, the population of interest, a reasonable population about which we can actually make inferences based on these data? If you use any notation, define it explicitly.
C) (6 points) Compute a confidence interval for the mortality rate in the New York study. You may use a method of your choice, but show all work and report your results in context.
D) (7 points) How much might the study results have changed, if all 831 ventilator patients still in the hospital at the end of the study also survived? Re-compute and report a new (hypothetical) CI for this scenario.
E) (4 points) Why do you think the study authors chose to publish their results, with so much uncertainty remaining about the fate of patients still in the hospital? Explain why you think they made the right, or wrong, scientific and ethical decision.
F) (6 points) Give two or more reasons for the big differences between the mortality rates reported in the first New York study and those quoted in the May 14 NPR story. State which mortality rates you think are more reliable and likely to be right, and why.
You may upload up to 5 PDF(s) or image(s) of your work. (If you prefer, you may type in Moodle, but all calculations and/or R code used should be clearly shown.)
NPR TRANSCRIPT (relevant excerpts). Optional Reading/Reference. (from All Things Considered website)
Early studies have found high mortality rates among COVID-19 patients on ventilators. But the new data from some major medical centers shows that many of those patients are much likely to survive.
DR. COLIN COOKE (critical care specialist at the University of Michigan): It’s always disheartening to know that some people are out there saying, if you end up on a ventilator, it’s a death sentence, which is not what we’re experiencing. And I don’t think it’s what the data are showing.
HAMILTON (Journalist): Cooke is especially concerned by a study of some New York hospitals that found a mortality rate of 88%. He says that figure was misleading because the analysis included only patients who had either died or been discharged [but 72.2% of the patients were still in the hospital at the end of the study].
COOKE: And so folks who actually in the midst of fighting their illness were not being included in this statistic of patients who are still alive.
HAMILTON: Cooke says many of those patients are likely to survive.
COOKE: We think that mortality for folks that end up on the ventilator with COVID is going to end up being somewhere between probably 25% up to maybe 50%, somewhere in that range.
HAMILTON: Scary but hardly a death sentence.
[Dr. Todd Rice … directs the medical intensive care unit at Vanderbilt Hospital in Nashville.]
RICE: Our mortality is in the mid to high 20% range.
HAMILTON: That’s not much higher than the rate for other ICU patients with severe lung infections. And Rice says that suggests COVID-19 is less deadly than the early numbers suggested.
RICE: I think overall these mortality rates are going to be higher than we’re used to seeing but not dramatically higher.
HAMILTON: Preliminary data from Emory University in Atlanta supports that prediction. The mortality rate among the 165 COVID patients placed on a ventilator was less than 30%. And unlike the New York study, only a few patients were still on a ventilator when the data were collected.
Dr. Craig Coopersmith directs the Critical Care Center at Emory. He says they plan carefully for the arrival of COVID patients, but the care followed their usual protocols.
CRAIG COOPERSMITH: There’s no secret magic that can’t be replicated in other places. And I do believe that we will see a global trend towards better outcomes on the ventilator and in the intensive care unit.
Solutions will vary.