The statewide stay at home order has more than served its intended purpose.
We now have more than two weeks of official data to assess the effectiveness of the extraordinary statewide stay at home order issued by Governor Newsom. The express purpose of the order was to flatten the daily hospital admission curve at a point where our healthcare system could continuously provide the highest quality care to all COVID-19 and other patients. Have we done it?
Yes. On April 8th, I wrote that the official State data pointed to a bent curve – hospitalizations were in decline. On April 11th, I wrote that the data confirmed my prior view – the curve had clearly been bent. At that time, I wrote that I expected to see a continuous, but bumpy, deceleration in the virus. How have my two prior analyses fared in light of another week of State data? They were “spot on.” Here is the updated statewide hospitalization curve, which has continued its dramatic downward trajectory:

The California COVID-19 Hospitalization Curve Remains Bent. Source: California Department of Public Health.
Let’s hope the “third time is the charm” for a demonstrably bent curve to make a difference in the course being charted by our elected officials. The data clearly reveal that the statewide stay at home order has more than achieved its intended purpose. If we are, as we claim to be, a State that is guided by data and science, then now is the time to transition into the next phase of the pandemic, during which “localism” will be “determinative” according to our Governor. The data show that the State is clearly ready to pass the baton. Our Counties need to take it.
Be safe. Stay healthy.
April 18, 2020
So when the California death stats didn’t go your way you jumped over to the hospitalization rate. Smells fishy Brian.
April 21, 2020
As I’ve said from the beginning, the best way to measure the success of the stay at home order is to track the the number of daily hospitalizations. Given the lag in the data (infection to symptoms to hospitalization to State reporting), the hospitalization data relevant to the impact of the stay at home order may have started appearing as early as March 25th, but the full impact of the stay at home order would not be seen in the data until sometime after April 2nd. What we’ve seen since the State started reporting hospitalizations on March 31st is a dramatic decrease in daily hospitalizations. On April 7th, the State reported almost 5800 hospitalizations. Yesterday, it was less than 4700. In evaluating the rate of decline, it is important to recognize that net hospitalizations change based on two factors – daily admissions and daily discharges. The State does not report either. I believe that admissions are dropping dramatically (as evidenced by the almost 60% reduction in suspected COVID-19 hospitalizations over the past three weeks) but that discharges are being spread out over a longer period because of the length of time patients are spending in intensive care. Given the lag in the data and the ongoing stay at home order, I am confident that the hospitalization curve will continue on its downward trajectory for quite some time before finding its floor. Lastly, it is important to recognize that the stay at home order was never intended to drive hospitalizations to zero. It was designed to ensure that we could manage the pandemic with 50,000 hospital beds (and perhaps 20,000 ICU beds). As I’ve noted, we needed about 5800. In light of what we know from the serology testing in California and the CDC’s analysis of the risk factors driving hospitalizations (age and pre-existing conditions), I am confident that we could devise far more carefully-tailored public health measures to protect the most vulnerable and ensure social distancing in public without seeing a true surge in hospitalizations. As for the comment that I am ignoring fatality data that “didn’t go my way,” I must disagree. As I’ve said from the beginning, fatality data is relevant for assessing the trajectory of the pandemic, but it is a lagging indicator. I initially estimated that the lag was 15-25 days from infection to fatality. Based on more recent reporting, I believe the lag could be as long as 28-31 days from infection to fatality in many cases. It appears the lag from hospital admission to fatality could be 12-17 days in many cases. Accordingly, I’ll be monitoring fatality data very closely over the next few days to see if it departs in any meaningful respect from the hospitalization data (which clearly demonstrates that hospitalizations peaked by April 7th) and, if so, what conclusions we can draw from the different trajectories. Lastly, it is worth noting that the recent fatality data in California is largely being driven by LA County. Over a recent 6-day period, LA County accounted for almost 60% of the State’s reported fatalities despite accounting for only 26% of the State’s population. It appears we may have two fatality curves in California – one that covers 75% of the State and is tracking the hospitalization curve and one for LA County that is highly erratic (the County reported 24 and 81 fatalities on consecutive days in the last week). I’ll be following this as well in the days ahead.
April 21, 2020
We absolutely need a lot more testing to get a better idea on the scope of this pandemic. The progression here in Ca. been so much slower than most of the rest of the country that I am beginning to suspect there may be a significant amount of people who have gotten a small dose infection of the virus and their immune system has been able to stifle it before they even had any symptoms. In this case we may be better off than we think but we need so many more tests, and not just for symptomatic people, to determine that and then act accordingly.
All the damage the stalled economy has caused and will cause in the future and whether that is worse than the pandemic itself is another thorny issue that is a tough call. Hundreds of thousands more horrible choking deaths vs. poverty, depression, increased domestic violence, possibly higher rates of suicide, drug and alcohol abuse, etc. Who wants to make that call?
April 19, 2020
I, too, question the use of hospitalizations as the parameter that should be used to determine State policy on the extent to which social distancing should continue to be implemented. But today I’m writing to make another point.
The slope of the linear curve shown in the graph of hospitalizations vs. time is -0.4% /day. In other words, it’s barely trending downward. And this is with full social distancing policies in force by the state! Relaxing those standards will inevitably cause that curve to at least decrease (be less steep) or even go up (more hospitalizations vs. time instead of less). Not the result we want!
If, in fact, the current social distancing policies were to continue and the hospitalization curve were to remain at -0.4%/day, it would take 250 days for the curve to reach zero hospitalizations per day. Relaxing social distancing policies would result in it taking even longer!
April 21, 2020
Thank you for your comment. Please see my reply to Lars Partman, which addresses your concerns.