• Home
• About Us
• Articles
• Pandemic Preparedness Store
• Resources
• Feedback
• Links

Epidemology Report in The Lancet Underestimates Pandemic Deaths

By Grattan Woodson, MD, FACP
January 11, 2007

A Review, Analysis, and Criticism of Recent Study Predicting Deaths During the Coming Influenza Pandemic: An article published in December 2006 in The Lancet provides a clear link between national socioeconomic status and death rates during the 1918 Spanish Flu.¹ This may be the best work done so far documenting the number of deaths occurring during the 1918 pandemic for which the authors deserve our appreciation. In their article, the authors used the information developed for the 1918 pandemic and extrapolated it to the present in order to project the effect a pandemic of similar magnitude would have today. For a variety of reasons discussed below, this exercise is flawed and the results could lead some to discount the severity of a major influenza pandemic and its impact on both the rich and poor nations alike.

A review of the article

Epidemiologist’s Murray etal published their article in the prestigious medical journal, The Lancet, on 23Dec2006. It provided an estimate of the number of deaths worldwide during the next pandemic based upon their work on the 1918 Spanish Flu. They assumed the coming pandemic would be severe like the 1918 event rather than moderate or mild as was seen in 1957 and 1968. Their principal finding was 50% of the variability in mortality seen in 1918 was due to socioeconomic status, the higher the status, the lower the number of excess deaths due to flu. An excess death is one that was unexpected and for the purposes of this article is one that was due directly or indirectly to influenza. Using data from 27 countries, these investigators showed that excess death rates were up to 30 times higher in the poor nations compared with the well to do. They project the next severe pandemic will cause between 51 and 157 million excess deaths worldwide with 95% of these occurring in the poorest nations.

Using age and sex adjusted mortality data from 13 countries, Murray etal reconfirmed the work of others that excess death rates among young health adults aged 15 to 40 was higher than for any other group. The chart below, reproduced from their article clearly shows the infamous “W” shaped mortality curve observed during the 1918 Spanish Flu.

While during the 1918 pandemic, the author’s finding that socioeconomic status explained 50% of the excess mortality seen then is very interesting result. The authors proposed that factors affecting the remaining 50% included “nutritional status, comorbidity, community characteristics associated with poverty, and the effect of supportive care, since therapeutic interventions had little or no effect on mortality in 1918–20”.

An analysis of the article’s methods

Past epidemiologic studies of the 1918 pandemic in the US estimated the number dying from influenza was between 500,000 and 625,000 persons. In order to compare these reports with the work of Murray etal, I applied a variation on the method they used by comparing crude death rates provided by the Insurance Information Institute for a period of years before and after the pandemic years as a baseline for expected mortality.² The excess mortality seen during the pandemic years is calculated by subtracting the average annual baseline mortality from the average annual pandemic mortality.

Crude death rates in US before and after 1918 pandemic

Year	1912	1913	1914	1915	1916	1917	1918	1919	1920	1921	1922	1923
Rate	10.0	12.3	8.9	15.7	27.0	17.4	301.8	99.0	70.5	11.3	31.0	44.1

These calculations for the US indicate that during the 6 years before and 3 years following the pandemic years of 1918-20, the average crude death rate per 100,000 was 21. During the three pandemic years it averaged 157. Using the methodology of Murray etal, the number of excess deaths attributable to influenza would equal (157-21 = 136) 136 per 100,000 US citizens for 3 years.

The US population at the start of the pandemic was 100 million and this number divided by 100,000 equals 1,000. To calculate the excess US deaths attributable to influenza would be (1000 x 136 x 3 = 408,000) 480,000 excess influenza deaths. This number is similar to the estimates made previously and supports the validity of the method used by the Murray etal

Criticism of the article’s projections for the coming pandemic

To calculate their predictions for the coming pandemic, they adjusted their study findings from 1918 for the higher socioeconomic status enjoyed by most people today, even those in the poorest countries compared and extrapolated them to the present. Despite the improved circumstances for the 87% of humankind that live in the developing and poor nations, the authors conclude that fully 95% of the average estimate of 62 million deaths for the next pandemic will occur in these nations. Approximately 858 people live in the developed world today. Applying Murray etal’s estimate for a major pandemic results in 3.1 million excess deaths. If a clinical attack rate of 40% is used, the calculated case fatality rate for this number of deaths is 0.9%. Lowering the clinical attack rate to 30% implies a case fatality rate of 1.2%. These are quite low estimates and well below those made by the US Department of Health and Human Services who’s predictions have been criticized for their optimism and many sanguine assumptions concerning the positive effect of public health measures used to slow the pandemic and the sustainability of the healthcare system under pandemic conditions.³

In their discussion, the authors state, “individual factors such as current immune function, nutritional status, acquired immunity through previous influenza infection, co-morbidity, access to health care, quality of care, and the physical environment could all have a role”. Despite the recognition of the importance of these issues, the author’s elected to dismiss these from their model, detracting, in my opinion, from its robustness.

Murray etal point out that mortality from the coming pandemic could be “larger if the 1918–20 pattern of low older adult mortality were in fact due to some acquired immunity from the pandemics of the mid-19th century” which is something that will not be a factor protecting today’s frail elderly from H5N1. If those 45 years of age and older are affected by the coming pandemic severely, the number of deaths seen will rise significantly.

They dismiss “concerns about increased travel and mixing, which lead to larger epidemics, might not alter our extrapolations” because “the historical record suggests that nearly all human populations were eventually exposed to the 1918–20 influenza virus”. While it is true that past and future pandemic influenza will reach everyone, the speed with which the pandemic moves through a population is an important factor affecting both clinical attack rates and death rates. The above factors discounted by the article’s authors, are all well known to increase the transmission rate (R0) of influenza virus.

A high transmission rate is a prescription for overwhelming healthcare service capacity everywhere including the rich countries, which the authors are counting on to hold down mortality in those nations. A high transmission rate will exacerbate worker absenteeism, something that threatens a systematic failure of other critical infrastructure including the electric power grid, water service, and food production, processing, and distribution.

A breakdown of this nature will translate into an increase the case fatality rate in patients who would otherwise survived had they had access to a set-up and staffed hospital bed rather than being treated in a poorly staffed and supplied temporary hospital or at home.

The author’s explanation for why they predict mortality will be so low in the rich nations are based upon “symptomatic medical management is better now than in 1918–20”, use “of antivirals such as zanamivir and oseltamivir phosphate might have a positive effect on the transmission and case-fatality rates”, the availability of 500 million effective doses of an efficacious pandemic vaccine “with a lag of 4-6 months from the onset of a pandemic”, strictly enforced quarantine measures in combination with antiviral drug prophylaxis, and efficacious “antibiotics for pneumonia” post-influenza “could have a substantial effect on case-fatality rates”.

A discussion of the article’s principal findings

While I agree that the poor will suffer disproportionately compared with the rich nations, the implication that world’s rich will be so little impacted by the coming pandemic is unwarrented. Specifically, the author’s extrapolation of their 1918 finding to the coming pandemic is inappropriate for several reasons. The first is the somewhat implied assumption that a stressed but nonetheless functional advanced healthcare system would be available to provide for the sick in the rich countries for the duration of the pandemic. That very good healthcare will be provided to many of those first infected with the virus in these nations, it is pretty likely that soon hospitals and clinics will be awash with patients. Given the high worker absenteeism expected with a major pandemic and its effect on critical infrastructure and healthcare, medical supplies of everything from drugs to diapers are likely to become extremely scarce.⁴

While quarantines combined with prophylactic antiviral treatment might be an effective public health measure slowing transmission rate of pandemic flu, almost no one, including the US Department of Homeland Security, thinks they can implement a quarantine that is tight enough to be efficacious in the US or elsewhere.⁵ What’s more, there is not going to be a sufficient stockpile of antiviral drugs for use as prophylaxis within the quarantined areas; rather the WHO and others have recommended that all available stores be reserved for treatment only.⁶ Yes, antibiotics are effective for treatment of the likely causes of post-influenza bacterial pneumonia, if they were available which is a prospect that is unlikely given the tremendous increase in demand expected for these agents during the pandemic and the very real possibility that manufacture of them will be interrupted worker absenteeism and critical supply shortages.

For these reasons I am not reassured or convinced by the author’s conclusions that the coming pandemic will have so relative impact on the rich nations compared with the poor. Both are likely to be severely affected. I agree with the authors that poverty and the unavailability of any healthcare will be accompanied by a higher case fatality rate in the poor nations compared with the rich. Where I differ is in my estimate for the magnitude of the coming pandemic and how it will impact the rich and poor alike.

In my opinion, only 1 in 3 critically ill people within the developed nations seeking advanced medical care in a hospital will be accommodated.⁷ In this alternative model I assume a worldwide clinical attack rate of 40% and case fatality rate of 8% in the developed nations and 12.5% in the poor countries. This results in excess deaths amounting to 287 million in the poor nations and 27 million in rich ones for a total average estimate of 314 million excess deaths. This projection call for excess deaths in the developed nations that is 9 times greater that the 3.1 million deaths predicted by Murray etal. In the poor nations, this model estimates the number of excess deaths will be 5 times greater than Murray and colleagues’ figure of 58.9 million. Using this alternative model, 9.4% of the excess deaths are predicted to occur in the rich countries, almost double the percentage predicted in the article.

Dangerous unintended consequences

While the authors deserve great credit for providing us with a well-researched study of death due to the 1918 Spanish Flu, the predictions made for the coming pandemic could have several dangerous unintended consequences. The very low estimates provided for deaths expected in the developed nations could help quell the growing sense of urgency to prepare properly for a major pandemic. For instance, it is certain some within and outside government will use this article to bolster their arguments for less pandemic preparedness spending.⁸ Citizens living in the wealthy nations might conclude from the results of this article, which were widely published in the world press, that while pandemic influenza is a serious problem, it will be mainly “over there, just another problem those poor countries will have to cope with”.⁹ This view might cause some to ignore discussions about the risk of pandemic in the same way they tune out the most heartrending pictures of starving children or seemingly endless war atrocities coming out of Africa. It is my hope that this otherwise fine article will not be used to stem the ardor to prepare properly and appropriately for the coming pandemic for those in government or for people everywhere. I am sure that this was not the author’s intent but may become the regretful result of this work.

References

¹

Christopher J L Murray, Alan D Lopez, Brian Chin, Dennis Feehan, Kenneth H Hill., Estimation of potential global pandemic influenza mortality on the basis of vital registry data from the 1918–20 pandemic: a quantitative analysis. The Lancet 2006;368;2211-18

²

Steven Weisbart., PANDEMIC: CAN THE LIFE INSURANCE INDUSTRY SURVIVE THE AVIAN FLU? Insurance Information Institute 17Jan2006

³

The US Department of Health and Human Services Pandemic Influenza Plan. November 2, 2005 (Official Pandemic Plans)

⁴

Michael T. Osterholm, PhD, MPH., Avian Flu: Addressing the Global Threat, Testimony Before the House Committee On International Relations December 7, 2005

⁵

US National Strategy for Pandemic Influenza., US Department of Homeland Security 2May2006

⁶

Treatment of Avian Flu, WHO May2006

⁷

Woodson GC., Estimates of Illness and Death During the Great Bird Flu Pandemic., published on the www.birdflumanual.com website 30May2006.

⁸

The Chicken Littles Were Wrong The bird flu threat flew the coop by Michael Fumento The Weekly Standard 25Dec2006

⁹

Next flu pandemic forecast to kill about 62 million., The Washington Times 25Dec2006