Governor’s metrics still raise questions

Iowa’s social distancing policy appears to be hostage to an unexplained and backward-looking indicator for hospitalizations.

(UPDATED, APRIL 16)

The latest “metrics” from the Governor’s office once again raise serious questions. A few days ago it seemed clear that two or three of the state’s six regions would very soon reach the magic number 10, at which point shelter-in-place is considered justified by the Governor and the Iowa Department of Public Health, according to their guidelines. Instead, as of April 15, regions 1 and 6 remained stuck at 8 and 9, respectively, and region 5 had fallen to 8. Why? Because the hospitalization rate score, which by deduction must have been at 3 for all 6 regions just a week ago, was suddenly downgraded to 1 in two regions, and 2 in two others.

Today, April 16, things changed again. Lo and behold, Region 6 made it to 10. And in fact the Governor followed through with something akin to shelter in place, with most kinds of gatherings limited to families, not groups of 10 or fewer. No additional business closures were announced, however. Meanwhile, Region 5 jumped two points with new outbreaks at two more nursing homes, but then lost a point because the hospitalization score apparently was lowered again, without any explanation. So it remains at 9, even though it is maxed out on all criteria except hospitalizations.

No explanation of the hospitalization score has been forthcoming beyond the vague definition in the “Guidance” memo unearthed by Zachary Smith of the Iowa City Press-Citizen last week. That memo defines it thus: “Percent of identified cases requiring hospitalization.” Is the numerator the cumulative total of all cases in Iowa that required hospitalization at some point, or just cases in the last 14 days, or just current hospitalizations as of the most recent day? Is the numerator cumulative cases, cases in the last 14 days, or something else? We don’t know, and no hospitalization rate by this measure has been reported even statewide, nor does the newly launched dashboard contain any hospitalization data at the county or regional level.

Total cases of COVID-19 continue to rise, as do current hospitalizations. So in the face of a rising number of Iowans currently with a severe enough case of the virus to be hospitalized, why does the hospitalization score decline, lessening the supposed need for shelter in place? Why is the percent relevant in the first place? Surely the total number of persons hospitalized for the virus is the single most important indicator, since it signifies not only the number of Iowans seriously affected by the virus, but the potential strain on hospital resources.

A forecast of this number is the crucial indicator in the widely known forecasting models by epidemiologists at the University of Washington and elsewhere. But in Iowa, we still do not have a forecast, and social distancing policy appears now to be hostage to an unexplained and backward-looking indicator. If that percentage continues to be low, or to fall, despite daily increases in cases, deaths, and hospitalizations, we may not see another region get to the magic number of 10.

Peter Fisher is research director for the nonpartisan Iowa Policy Project in Iowa City.

pfisher@iowapolicyproject.org

Sheltering the data in place

One thing is clear: transparency has been sadly lacking, and for no apparent reason.

Governor Kim Reynolds over the past few weeks has moved incrementally to close more kinds of businesses, to the point where Iowa’s restrictions now resemble those of states that have a blanket statewide “shelter in place” order. Significant distinctions remain: a proper and comprehensive shelter in place order closes all businesses except those specified as essential, leaving no ambiguities and loopholes, and comes with clear and enforceable restrictions on travel and social activities.

The governor continues to assert that her recommendations are driven by the same four metrics that have guided her since the beginning and that only recently became partly public information due to efforts by the press. We provided a thorough analysis of that guidance several days ago. On Tuesday, we finally learned about one of those metrics: There are three long-term care facilities with a sufficient number of COVID-19 cases to be classified as a facility with an outbreak.

We now know enough to construct the point system in spite of stonewalling by the Governor’s Office.

The first of the four measures — percent of population age 65 or over — can be found from census data. The second — cases per 100,000 population — can be calculated because the number of cases has been released by IDPH by county. The third — outbreaks at care facilities — is now known, with locations, because of a question at a press conference.

That leaves the fourth — hospitalizations as a percent of cases — that is unknown by county or region because the governor still refuses to release the data. But we know the total score by region because it shows up on the maps that are intermittently released at press conferences (but remain unavailable on the IDPH website). Thus by subtraction we can determine that all four regions must be at the highest level, a 3, on the hospitalization rate score.

From here on out, the only thing that can change is the cases per 100,000 population and the number of care facility outbreaks. Region 5 is already at the maximum on the cases measure, and regions 1 and 6 will likely get there soon, leaving all three regions with a score of 9, 1 short of 10, the number that supposedly triggers shelter in place. So those regions, covering a large majority of the state’s population and COVID-19 cases, can get to 10 only with another outbreak at a care facility.

The governor on the one hand argues that we already have the equivalent of shelter in place, and at the same time the metric that she says still guides her decisions shows that shelter in place is not yet warranted anywhere in the state. Has that metric really been used thus far, and in what way? How do you get from the metrics to a list of particular additional businesses to close? What will happen when a region reaches 10? Will the governor order more stringent measures in just that region? Or will the whole thing be scrapped once a proper forecasting model is developed that meets with her approval?

One thing is clear: transparency has been sadly lacking, and for no apparent reason.

Peter Fisher is research director of the nonpartisan Iowa Policy Project.

pfisher@iowapolicyproject.org