University of Pittsburgh Public Health Interventions & Epidemic Intensity During 1918 Influenza Reflection

University of Pittsburgh

Question Description

I’m studying for my Health & Medical class and need an explanation.


Attached is an article for the Proceedings of the National Academy of Science entitled “Public health interventions and epidemic intensity during the 1918 influenza pandemic” It is certainly timely considering the current Covid-19 pandemic. Your assignment is to read the paper and write a 200 word (approximate) reflection on how the conclusions of this article impact your understanding of the current pandemic, including what you would consider the most important take away/conclusion.

Unformatted Attachment Preview

Public health interventions and epidemic intensity during the 1918 influenza pandemic Richard J. Hatchett*†, Carter E. Mecher‡§, and Marc Lipsitch¶ *Division of Allergy, Immunology, and Transplantation, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD 20892; ‡Department of Veterans Affairs, VA Southeast Network, 3700 Crestwood Parkway, Duluth, GA 30096; §Homeland Security Council, Executive Office of the President, EEOB, 1650 Pennsylvania Avenue NW, Washington, DC 20502; and ¶Department of Epidemiology and Department of Immunology and Infectious Diseases, Harvard School of Public Health, 677 Huntington Avenue, Boston, MA 02115 Edited by Burton H. Singer, Princeton University, Princeton, NJ, and approved February 14, 2007 (received for review December 9, 2006) Nonpharmaceutical interventions (NPIs) intended to reduce infectious contacts between persons form an integral part of plans to mitigate the impact of the next influenza pandemic. Although the potential benefits of NPIs are supported by mathematical models, the historical evidence for the impact of such interventions in past pandemics has not been systematically examined. We obtained data on the timing of 19 classes of NPI in 17 U.S. cities during the 1918 pandemic and tested the hypothesis that early implementation of multiple interventions was associated with reduced disease transmission. Consistent with this hypothesis, cities in which multiple interventions were implemented at an early phase of the epidemic had peak death rates ⬇50% lower than those that did not and had less-steep epidemic curves. Cities in which multiple interventions were implemented at an early phase of the epidemic also showed a trend toward lower cumulative excess mortality, but the difference was smaller (⬇20%) and less statistically significant than that for peak death rates. This finding was not unexpected, given that few cities maintained NPIs longer than 6 weeks in 1918. Early implementation of certain interventions, including closure of schools, churches, and theaters, was associated with lower peak death rates, but no single intervention showed an association with improved aggregate outcomes for the 1918 phase of the pandemic. These findings support the hypothesis that rapid implementation of multiple NPIs can significantly reduce influenza transmission, but that viral spread will be renewed upon relaxation of such measures. mitigation 兩 nonpharmaceutical interventions 兩 closures Downloaded by guest on March 14, 2020 I nfluenza pandemics have occurred periodically in human populations, with three pandemics in the 20th century. The 1918 influenza pandemic resulted in unprecedented mortality, with an estimated 500,000–675,000 deaths in the U.S. and 50–100 million deaths worldwide (1–3). The spread of H5N1 avian influenza has provoked public concern (4) and accelerated efforts to plan for the next pandemic. Because antiviral medications and effective vaccines may not be widely available at the beginning of a pandemic, many authorities have suggested using nonpharmaceutical interventions (NPIs; i.e., voluntary quarantine of infected households, closure of schools, bans on public gatherings, and other measures) to decrease disease transmission. This approach is supported by mathematical models, which suggest that multiple simultaneous NPIs applied early in an epidemic may significantly reduce disease transmission (5). A recent review, however, concluded that the evidence base for recommending such interventions is limited, consisting primarily of historical and contemporary observations, rather than controlled studies (6). The intensity of the 1918 pandemic, whether assessed as total excess deaths, the rate of increase in the epidemic curve, or peak death rates, varied widely among U.S. cities. Cities also varied widely in their choice and timing of implementation of NPIs designed to reduce disease spread. Many cities closed schools, churches, theaters, dance halls, or other public accommodations; made influenza a notifiable disease; banned funerals or other public 7582–7587 兩 PNAS 兩 May 1, 2007 兩 vol. 104 兩 no. 18 gatherings; or introduced isolation of sick persons. In some cases, these NPIs were put in place in the first days of epidemic spread in a city, whereas in other cases, they were introduced late or not at all (Table 1). We noted that, in some cases, outcomes appear to have correlated with the quality and timing of the public health response. The contrast of mortality outcomes between Philadelphia and St. Louis is particularly striking (Fig. 1). The first cases of disease among civilians in Philadelphia were reported on September 17, 1918, but authorities downplayed their significance and allowed large public gatherings, notably a city-wide parade on September 28, 1918, to continue. School closures, bans on public gatherings, and other social distancing interventions were not implemented until October 3, when disease spread had already begun to overwhelm local medical and public health resources. In contrast, the first cases of disease among civilians in St. Louis were reported on October 5, and authorities moved rapidly to introduce a broad series of measures designed to promote social distancing, implementing these on October 7. The difference in response times between the two cities (⬇14 days, when measured from the first reported cases) represents approximately three to five doubling times for an influenza epidemic. The costs of this delay appear to have been significant; by the time Philadelphia responded, it faced an epidemic considerably larger than the epidemic St. Louis faced. Philadelphia ultimately experienced a peak weekly excess pneumonia and influenza (P&I) death rate of 257/100,000 and a cumulative excess P&I death rate (CEPID) during the period September 8–December 28, 1918 (the study period) of 719/100,000. St. Louis, on the other hand, experienced a peak P&I death rate, while NPIs were in place, of 31/100,000 and had a CEPID during the study period of 347/100,000. Consistent with the predictions of modeling, the effect of the NPIs in St. Louis appear to have had a lesspronounced effect on CEPID than on peak death rates, and death rates were observed to climb after the NPIs were lifted in midNovember (7–9). To investigate whether early implementation of individual interventions or of multiple measures reduces disease transmission during influenza pandemics, we analyzed the NPIs used in a collection of U.S. cities during the fall wave of the 1918 pandemic, identifying the NPIs used in each city as well as the timing of their implementation [details of individual city outcomes and intervenAuthor contributions: R.J.H., C.E.M., and M.L. designed research; R.J.H., C.E.M., and M.L. performed research; M.L. analyzed data; and R.J.H. and M.L. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. Freely available online through the PNAS open access option. Abbreviations: P&I, pneumonia and influenza; CEPID, cumulative excess P&I deaths; NPI, nonpharmaceutical intervention; CFP, case-fatality proportion. See Commentary on page 7313. †To whom correspondence should be addressed. E-mail: This article contains supporting information online at 0610941104/DC1. © 2007 by The National Academy of Sciences of the USA www.pnas.org兾cgi兾doi兾10.1073兾pnas.0610941104 SEE COMMENTARY 50 28-Dec 21-Dec 7-Dec 14-Dec 30-Nov 23-Nov 9-Nov 16-Nov 2-Nov 26-Oct 19-Oct 0 5-Oct — 15.7 (7.6, 30.8) — 100 12-Oct 4 14 5 150 28-Sep 5.6 (3.1, 25.9) 200 21-Sep 15 Philadelphia St. Louis 250 14-Sep Making influenza a notifiable disease Emergency declarations Isolation policies Quarantine of households where infection identified School closures Church closures Theater closures Dance hall closures Other closures Staggered business hours to reduce congestion in stores and on transit systems Mask ordinances Rules forbidding crowding on streetcars Private funerals Bans on door-to-door sales Interventions designed to reduce transmission in the workplace Protective sequestration of children Bans on public gatherings No-crowding rules in locations other than transit systems Community-wide business closures 300 Date 14 15 15 11 13 8 30.8 (15.1, 96.3) 29.9 (12.4, 130.6) 29.9 (10.3, 66.9) 44.7 (12.4, –) 84.7 (29.9, 322.0) — 2 6 — — 11 1 0 92.1 (30.8, –) — — 3 — 15 3 30.8 (12.4, 118.1) — Fig. 1. Excess P&I mortality over 1913–1917 baseline in Philadelphia and St. Louis, September 8 –December 28, 1918. Data are derived from ref. 10. the date on which the intervention was announced. Thus, early interventions in a given city were those that were implemented when relatively few individuals had died, whereas later ones a peak weekly excess P&I death rate (median for group) Intervention Number of 17 cities implementing Median (interquartile range) epidemic stage (CEPID) at time of implementation* Death Rate / 100,000 Population Table 1. Summary of interventions and their timing across 17 cities 160 ** * ** * ** * 140 120 100 80 60 40 20 0 1 5+ 4+ schools churches theaters public closed measures measures closed closed gatherings banned CEPID < 20 CEPID < 30 CEPID<30 CEPID<30 CEPID<30 CEPID<30 — *Shown only for interventions implemented in at least nine cities (⬎50%); 75th percentile not shown for interventions implemented in ⬍13 cities. b 600 * Results Downloaded by guest on March 14, 2020 Effect of Early Interventions on Epidemic Spread. We assessed the relationship between the timing of NPIs and three measures of epidemic outcome: (i) the peak weekly rate of excess P&I deaths per 100,000 population (peak death rate) during the study period; (ii) the ‘‘normalized’’ peak weekly excess P&I death rate (peak weekly death rate during the study period divided by the median weekly rate during the period); and (iii) the CEPID per 100,000 population during the study period. The stage of the epidemic at the time of each intervention was defined as the CEPID from the start of the study period until Hatchett et al. 400 300 200 100 0 4+ measures CEPID < 20 5+ schools churches theaters public measures closed closed closed gatherings CEPID < CEPID<30 CEPID<30 CEPID<30 banned 30 CEPID<30 Fig. 2. Relationship of (a) peak weekly excess P&I death rate and (b) total excess P&I death rate during the study period to the timing of various NPIs. Cities were divided evenly into those intervening early (black bars) vs. late or not at all (gray bars), and the median outcome for the early and late groups was plotted. The first two groups of bars assess overall timing of intervention, comparing those cities that announced four or more NPIs before experiencing 20/100,000 CEPID with those with three or fewer and those that announced five or more NPIs before experiencing 30/100,000 CEPID with those with four or fewer. The remaining groups compare those cities that announced particular measures before experiencing 30/100,000 CEPID with those that did not. Significance by Mann–Whitney U test: *, P ⬍ 0.05; **, P ⬍ 0.01. PNAS 兩 May 1, 2007 兩 vol. 104 兩 no. 18 兩 7583 MEDICAL SCIENCES tions are included in supporting information (SI) Appendix]. We then related this information to the observed outcomes of the peak weekly death rate and CEPID during the period September– December, 1918. Excess death rates were used as a proxy for case incidence because of the more accurate reporting of deaths than cases. We hypothesized that early implementation of multiple NPIs in an immunologically naı̈ve population would slow the progression of the epidemic, resulting in a flatter epidemic curve, but that over time aggregate outcomes would approach those observed in cities not implementing such measures, until roughly comparable levels of herd immunity were achieved. study period total excess P&I death rate (median for group) 500 Table 2. Correlation between influenza epidemic outcomes and timing of interventions in 17 U.S. cities in 1918 Outcome: Excess weekly P&I deaths Measure of interventions Number of interventions before: 10/100,000 CEPID 20/100,000 CEPID 30/100,000 CEPID 40/100,000 CEPID CEPID at time of intervention: First Second Third Fourth Fifth Sixth CEPID at time of: Closing schools Closing theaters Closing churches Closing dance halls Other closures Making influenza notifiable Bans on public gatherings Imposing case isolation Bans on public funerals Normalized peak Peak 1918 total ⴚ0.53, P ⴝ 0.03 ⴚ0.68, P ⴝ 0.002 ⴚ0.51, P ⴝ 0.04 ⫺0.32, P ⫽ 0.21 ⴚ0.53, P ⴝ 0.03 ⴚ0.64, P ⴝ 0.005 ⴚ0.55, P ⴝ 0.02 ⫺0.40, P ⫽ 0.11 ⫺0.31, P ⫽ 0.22 ⴚ0.52, P ⴝ 0.03 ⫺0.29, P ⫽ 0.27 ⫺0.07, P ⫽ 0.80 0.08, P ⫽ 0.76 0.54, P ⴝ 0.02 0.54, P ⴝ 0.02 0.66, P ⴝ 0.004 0.55, P ⴝ 0.02 0.26, P ⫽ 0.31 0.004, P ⫽ 0.87 0.47, P ⫽ 0.06 0.52, P ⴝ 0.03 0.70, P ⴝ 0.002 0.67, P ⴝ 0.003 0.44, P ⫽ 0.08 0.07, P ⫽ 0.79 0.39, P ⫽ 0.12 0.31, P ⫽ 0.22 0.38, P ⫽ 0.13 0.27, P ⫽ 0.30 0.05, P ⫽ 0.84 0.63, P ⴝ 0.007 0.72, P ⴝ 0.001 0.70, P ⴝ 0.002 0.04, P ⫽ 0.87 0.34, P ⫽ 0.18 ⫺0.07, P ⫽ 0.79 0.56, P ⴝ 0.02 0.14, P ⫽ 0.59 0.09, P ⫽ 0.72 0.25, P ⫽ 0.34 0.17, P ⫽ 0.52 0.17, P ⫽ 0.53 0.15, P ⫽ 0.57 0.24, P ⫽ 0.35 0.11, P ⫽ 0.67 0.27, P ⫽ 0.30 0.13, P ⫽ 0.62 ⫺0.41, P ⫽ 0.10 0.54, P ⴝ 0.02 0.56, P ⴝ 0.02 0.56, P ⴝ 0.02 0.03, P ⫽ 0.90 0.33, P ⫽ 0.19 0.01, P ⫽ 0.97 0.46, P ⫽ 0.06 0.16, P ⫽ 0.53 ⫺0.09, P ⫽ 0.75 Three measures of epidemic intensity. Peak weekly excess P&I death rate, normalized peak weekly excess P&I death rate (peak divided by median weekly rate during the study period), and 1918 study period total excess P&I death rate are related to number of interventions before reaching a specified CEPID, CEPID at time when specified numbers of interventions had been imposed, and CEPID at time when specific interventions had been imposed. Spearman rank correlations and associated P values are shown, with bold type for P ⬍ 0.05. Downloaded by guest on March 14, 2020 were those implemented after more excess P&I deaths had occurred. In comparisons across cities (Fig. 2a, Table 2), we found that aggressive early intervention was significantly associated with a lower peak of excess mortality (Spearman ␳ ⫽ ⫺0.49 to ⫺0.68, P ⫽ 0.002–0.047; see Table 2, Number of interventions before, for the number of NPIs before a given CEPID cutoff vs. peak mortality). Cities that implemented three or fewer NPIs before 20/100,000 CEPID had a median peak weekly death rate of 146/100,000, compared with 65/100,000 in those implementing four or more NPIs by that time (Fig. 2a, P ⫽ 0.005). The relationship was similar for normalized peak death rates and for a range of possible cutoffs (see Table 2, CEPID at time of intervention), although the relationship became weaker as later interventions were included. Cities with more early NPIs also had fewer total excess deaths during the study period (Fig. 2b, Table 2, 1918 total), but this association was weaker: cities with three or fewer NPIs before CEPID ⫽ 20/100,000 experienced a median total excess death rate of 551/100,000, compared with a median rate of 405/100,000 in cities with four or more NPIs (P ⫽ 0.03). The association of early intervention and lower peak death rates was also observed when cities were ranked according to the CEPID in each city at the time of the second, third, fourth, or fifth intervention (Table 2, CEPID at time of intervention). Similar relationships were again detected for the normalized peak death rate [Table 2, CEPID at time of intervention/Normalized peak]. Again, the relationship with total death rate was weaker and in this case not statistically significant. Effects of Individual Interventions. To assess whether particular NPIs were associated with better outcomes, we calculated a Spearman rank correlation coefficient between outcome measures and the stage at which individual NPIs were implemented in each city (cities that never implemented a given intervention 7584 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0610941104 were ranked last in each analysis). Results are shown in Table 2, CEPID at time of. Early school, church, or theater closure was associated with lower peak excess death rates (Spearman ␳ ⫽ 0.54–0.56, P ⫽ 0.02). Cities that made each of these interventions before they reached 30/100,000 CEPID had a median peak death rate of 65–68/100,000, compared with median peaks of 127–146/100,000 for cities that made these interventions later or not at all (Fig. 2a, P ⫽ 0.005–0.01). Announcements of school, church, and theater closures were linked in most cities, occurring within a span of ⱕ6 days in the majority, and this near simultaneity of implementation precludes multivariate analysis or strong inference about the relative importance of the individual NPIs. Early bans on public gatherings were also associated with lower peak excess death rates, but the statistical significance of this result depended on the test used [Table 2, CEPID at time of, and Fig. 2a]. Of the other NPIs considered (closure of dance halls, other closures, isolation of cases, bans on public funerals, and making influenza notifiable), none showed a statistically significant association between the stage of implementation and the peak or cumulative excess death rates (Table 2, CEPID at time of, and Fig. 2). Other Predictors of Epidemic Severity. We assessed the correlation between peak mortality rate and each of the following variables: latitude, longitude, 1910 population density, 1920 population density, 1918 population size, and epidemic start week, defined as the first week in which excess P&I mortality exceeded 10/100,000. Of these variables, only longitude (Spearman ␳ ⫽ ⫺0.61, P ⫽ 0.009) and epidemic start week (Spearman ␳ ⫽ ⫺0.55, P ⫽ 0.02) were significantly associated with the peak weekly excess P&I mortality rate, and these two variables were strongly associated with one another (Spearman ␳ ⫽ 0.66, P ⫽ 0.004), indicating that eastern U.S. cities were hit earlier in our data set. In addition, cities whose epidemics began later tended to intervene at an earlier stage of their Hatchett et al. Sensitivity Analyses. Similar results were obtained when the inter- vention date was defined as the date public health orders were promulgated (Table 2) or the last date a particular type of gathering was permitted (e.g., Sunday church service; SI Table 4). Results were identical or improved when 7- and 10-day lags in assessing CEPID were introduced to account for the lag between infection and death (SI Tables 5 and 6). Relationship Between Interventions and Subsequent Waves. Al- Downloaded by guest on March 14, 2020 though it was not the primary intent of this paper to analyze pandemic wave dynamics, it is possible to formulate descriptive observations from the data at hand (SI Table 7). In offering these observations, it is important to underscore that in some cities, information about the dates of relaxation of the interventions used was incomplete. All cities showed some fluctuation in mortality rates after the main wave of the 1918 pandemic subsided. The peak weekly mortality rates observed in ‘‘second waves’’ in the cities we studied ranged from 13.60 to 79.69/100,000, as compared with 31.29– 256.96/100,000 during the first wave. There was a statistically significant inverse correlation of the height of the first and second peaks (Spearman ␳ ⫽ ⫺0.53, P ⫽ 0.03), so that cities that had low peaks during the first wave were at greater risk of a large second wave. Cities that had lower peak mortality rates during ...
Purchase answer to see full attachment
Student has agreed that all tutoring, explanations, and answers provided by the tutor will be used to help in the learning process and in accordance with Studypool's honor code & terms of service.

Final Answer


Pandemics are terrifying, including the one that is occurring now. It was interesting and
surprising to read the in the 20th century there were three large Influenza outbreaks. Although
these outbreaks were controlled eventually, they still cost many their lives...

nixiypixiy (2148)
UT Austin

Solid work, thanks.

The tutor was great. I’m satisfied with the service.

Goes above and beyond expectations !

Similar Questions
Related Tags