<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by John Spry on Medium]]></title>
        <description><![CDATA[Stories by John Spry on Medium]]></description>
        <link>https://medium.com/@johnspry?source=rss-7972ec239467------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*96LpdwesBctkdsqT.</url>
            <title>Stories by John Spry on Medium</title>
            <link>https://medium.com/@johnspry?source=rss-7972ec239467------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Mon, 16 Nov 2026 12:45:09 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@johnspry/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Disregard the Claims of “Association of Stay-at-Home Orders with COVID-19 Hospitalizations in 4…]]></title>
            <link>https://medium.com/@johnspry/disregard-the-claims-of-association-of-stay-at-home-orders-with-covid-19-hospitalizations-in-4-bd5247c56b3?source=rss-7972ec239467------2</link>
            <guid isPermaLink="false">https://medium.com/p/bd5247c56b3</guid>
            <category><![CDATA[econometrics]]></category>
            <category><![CDATA[statistics]]></category>
            <category><![CDATA[stay-at-home]]></category>
            <category><![CDATA[covid19]]></category>
            <dc:creator><![CDATA[John Spry]]></dc:creator>
            <pubDate>Mon, 27 Jul 2020 19:16:57 GMT</pubDate>
            <atom:updated>2020-07-27T19:16:57.768Z</atom:updated>
            <content:encoded><![CDATA[<h3><strong>Disregard the Claims of “Association of Stay-at-Home Orders with COVID-19 Hospitalizations in 4 States” Because Its Methodology is Flawed</strong></h3><p>By Muffadal Baxamusa, Ph.D., Department of Finance, Opus College of Business, University of St. Thomas, <a href="/service/mailto:BAXA0428@stthomas.edu">BAXA0428@stthomas.edu</a>; Paul Novosad, Ph.D., Department of Economics, Dartmouth College, <a href="/service/mailto:paul.novosad@dartmouth.edu">paul.novosad@dartmouth.edu</a>; and John A. Spry, Ph.D., Department of Finance, Opus College of Business, University of St. Thomas., <a href="/service/mailto:jaspry@stthomas.edu">jaspry@stthomas.edu</a></p><p>The JAMA research letter “Association of Stay-at-Home Orders with COVID-19 Hospitalizations in 4 States”<a href="#_edn1">[i]</a> (referred to as SKG) should not be used to make any claims about the effects of stay-at-home orders. It showed that actual hospitalizations were well below the forecast from a univariate exponential trend after the effective date of stay-at-home orders in four states. Unfortunately, it did not have a placebo test, which would examine whether actual hospitalizations were also lower than a forecast of exponential growth in states without stay-at-home orders.</p><p><strong>Methods</strong></p><p>We study whether cumulative COVID-19 hospitalizations or cases continue to grow at the initial estimated exponential rate in three additional states without stay-at-home orders. We add data from three states in the middle of the United States (like three of the four states in SKG) that did not impose statewide or local stay-at-home orders at any time.</p><p>Cumulative hospitalization data was downloaded from the COVID-19 Tracking Project website<a href="#_edn2">[ii]</a> for Iowa and North Dakota. These states were studied because of data availability. Cumulative COVID-19 case data for Arkansas was downloaded from the Johns Hopkins Coronavirus Resource Center<a href="#_edn3">[iii]</a> because Arkansas did not have cumulative hospitalization data available until April 28.</p><p>We implement the methodology of SKG by estimating a univariate exponential model where either COVID-19 hospitalizations or cases are a function of time during the curve fitting period shown in the Table. We use a counterfactual stay-at-home placebo date of March 30, the same day Virginia issued its stay-at-home order, to determine the curve fitting period.<a href="#_edn4">[iv]</a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*evnbqIKdHyos5CJVJ-9eRg.png" /><figcaption><strong>Table: Cumulative Hospitalizations or Infections Due to COVID-19 in Iowa, North Dakota, and Arkansas</strong></figcaption></figure><p><strong>Results</strong></p><p>Actual COVID-19 cumulative hospitalizations or cases are substantially lower than the forecasted exponential growth rates in Iowa, North Dakota, and Arkansas, even without stay-at-home orders. Actual hospitalizations and infections did not continue to grow exponentially in these states, contrary to the predictions of the SKM model.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/674/1*7sEteWkj79f71urNiMdetw.png" /><figcaption><strong>Figure: Projected vs. Observed COVID-19 Events Before and After Placebo Stay-At-Home Dates</strong></figcaption></figure><p><strong>Discussion</strong></p><p>If the SKG methodology was valid, states without stay-at-home models should have experienced exponential growth in COVID-19 hospitalizations and cases. Yet, they also experienced COVID-19 events that were far below those forecasted.</p><p>The exponential functional form ignores Farr’s Law.<a href="#_edn5">[v]</a> Farr observed in 1840 that the flow (infections, hospitalizations, or deaths) of an infectious disease could be approximated by a normal bell curve.<a href="#_edn6">[vi]</a> Cumulative counts of disease events, like hospitalizations, may be approximated by a sigmoid function, such as the logistic function. In contrast, the exponential functional form assumes constant exponential growth. For instance, the exponential SKG model forecasts that Colorado would have had 1,549,000 hospitalizations without its stay-at-home order, roughly twenty-seven percent of its population by the April 28, instead of its actual 2,723 cumulative hospitalizations. This projection from an exponential model is implausible.</p><p>This does not mean that stay-at-home orders do not have an effect on COVID-19 hospitalizations. Instead, it means that the SKG methodology for assessing an association between stay-at-home orders and COVID-19 hospitalizations is incorrect. Whether or not stay-at-home orders play a role in reducing hospitalization is impossible to determine using this flawed methodology.[vii]</p><p><a href="#_ednref1">[i]</a> Sen S, Karaca-Mandic P, Georgiou A. Association of Stay-at-Home Orders With COVID-19 Hospitalizations in 4 States. <em>JAMA.</em> Published online May 27, 2020. doi:10.1001/jama.2020.9176</p><p><a href="#_ednref2">[ii]</a> <a href="https://carlsonschool.umn.edu/mili-misrc-covid19-tracking-project">https://carlsonschool.umn.edu/mili-misrc-covid19-tracking-project</a></p><p><a href="https://carlsonschool.umn.edu/mili-misrc-covid19-tracking-project">COVID-19 Hospitalization Tracking Project</a></p><p><a href="#_ednref3">[iii]</a> <a href="https://coronavirus.jhu.edu/">https://coronavirus.jhu.edu/</a></p><p><a href="https://coronavirus.jhu.edu/">Home - Johns Hopkins Coronavirus Resource Center</a></p><p><a href="#_ednref4">[iv]</a> Data on Iowa hospitalizations starts on April 1, resulting in the April 1-April 10 fitting period. Data on North Dakota hospitalizations starts on March 21, resulting in the March 21-April 10 fitting period. Data on Arkansas cases starts on March 13, resulting in the March 13-April 3 fitting period.</p><p><a href="#_ednref5">[v]</a> Farr, William (1840). <a href="https://babel.hathitrust.org/cgi/pt?id=njp.32101064041955&amp;view=1up&amp;seq=71">“Causes of death in England and Wales”</a>. <em>Second Annual Report of the Registrar General of Births, Deaths and Marriages in England</em>. 2: 69–98.</p><p><a href="#_ednref6">[vi]</a> Bregman DJ, Langmuir AD. Farr’s Law Applied to AIDS Projections. <em>JAMA.</em> 1990;263(11):1522–1525. doi:10.1001/jama.1990.03440110088033</p><p>[vii] A preliminary plot of Colorado’s COVID-19 cases and concerns about the methodology used in “Disregard the Claims of ‘Association of Stay-at-Home Orders with COVID-19 Hospitalizations in 4 States’ Because Its Methodology is Flawed” was shared by the authors on Twitter on May 27, 2020.</p><p><a href="https://threadreaderapp.com/thread/1265793481387249679.html">Thread by @paulnovosad: How low is the bar for publishing a study showing stay-at-home orders work? Lower than you think. JAMA edition: fit an absurd exponential mo...</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bd5247c56b3" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>