View report page

16
Contribution
to a study of
some of the
factors
governing
the form of
the influenza
wave.
tically equal extent in the early nineties, and again in 1916-18). Second—It is important to note that
the constitutions extend over more years in the later centuries than in the earlier, and arising out of
this it will be clearly discerned that the state of the population concerned, as regards freedom of communication
with the world at large and also as regards immunisation by previous attack, has momentous
influence in determining duration of the constitution and the behaviour generally of the epidemic
disease.
It is assumed in the following argument that influenza spreads directly from person to person,
and hence, as Parsons and Clemow have shown, follows lines of communication and traffic. It is further
assumed that Creighton's conclusion that there is "something more than accident in the association
between epidemics of influenza and epidemics of ague'' is now accepted, on the understanding that for
epidemics of "ague" the pulmonary, gastro-intestinal or cerebrospinal complications of influenza
must be substituted. There is in fact change of type of the epidemic disease and the cause of the change
must be sought in the degree and extent to which the particular population concerned is exposed to
infection on the one hand and susceptible or immunised by previous recent attack on the other.
On a review of the whole matter it would seem, therefore, that the following contributions to a
theory of the surend of influenza mav lie formnlatprl from the eniHeminlnoripnl data
(a) In the first place it must be recognised that the closer the aggregation of population in a
particular area the more likely is the disease to maintain itself for a series of years round about a pandemic
prevalence; the more sparse the population and the more removed from the great lines of
communication and trarlic the area, the more likely is an epidemic when introduced to die out there quite
speedily. Thus, in the years before and after 1890, London and New York presented a smouldering fire
of prevalence, an epidemic constitution of ten or more years' duration; in the London of the 17th and
18th centuries the constitution only lasted from one to four or five years; and in the London of the
15th and 16th centuries there were single isolated years of epidemic prevalence. In 15th century London
continental traffic was practically in abeyance in winter and all the English sweats occurred in summer
time; it will hence be appreciated that in 15th century London, or again in what is fairly comparable
with it, in the present day, say, some little-visited island in the high seas, or Greenland, or the
Murman coast, influenza is only found to prevail when infection is newly introduced. Prevalences in
such places as these are likely when they do occur to be severe ones; in contrast to the prevalences
in the large and partially immunised centres of population, in which, while influenza may smoulder
for years a great epidemic is only possible a long time after the last pandemic prevalence.
(b) Then, secondly, certain rules as to the intervals between prevalences in thickly populated
areas, the extent of these prevalences and their duration may be formulated. The prevalence of the
pandemic year is, subject to some exceptions, likely to be a maximum prevalence, it extends over the
shortest time, and the interval between it and its successor will be short. Succeeding "trailers "will
gradually diminish in height, but will last longer and recur at longer intervals.*
The form of the epidemic curve has been studied by Ross, Brownlee, Greenwood and others
by aid of mathematical methods. It must suffice here to illustrate the subj ect by referring to the comparatively
simple case of measles. The curve of measles in London was explained (in the Milroy
Lectures of f906) as resulting in large degree from interplay between the deflecting energy due to new
infections, and the restoring inertia due to resistance to spread of disease caused by acquirement of
immunity and consequent increase in the amount of insusceptible material.
In London, the measles wave has a 1½—2 years' length from crest to crest; about 180,000 children
are attacked during that period and the number of susceptibles varies from a maximum of from 200,000
—300,000 to a minimum of about 50,000.
Between the comparatively simple epidemic curve of measles and the complex curve of influenza,
the disparities are great. The incubation periods in the two diseases of course differ widely; secondly,
measles attacks everybody, some people appear to be immune to influenza; thirdly, measles protects,
some people suffer again and again from influenza, and the period of protection afforded by a previous
attack may be quite short; fourthly, there is the question of infecting power, and this, while fairly
constant in measles, varies at the differing phases of influenza prevalence. It will be clear that the rate
of reaping will be far more rapid in influenza, but, on the other hand, the harvest to be reaped is greater.
But now comes the most noteworthy disparity. In measles, a succeeding prevalence does not
occur until 180,000 recently born and comparativelv insusceptible babies have grown up to be, say, from
18 months to three years old. In influenza the babies arc of relatively small importance, for a much
more appreciable effect is exerted by loss of immunity on the part of older persons. The following
points will, therefore, be noted on contrasting an original (pandemic) prevalence of influenza with
the next succeeding prevalence. In the former, persons who mix freely with their fellows are specially
likely to be attacked ; in the latter young children and old persons will suffer in larger proportion,
in the former, the prevalence will be widespread, the mortality comparatively low; in the latter, the
*And here it may be parenthetically remarked that difficulty is frequently experienced in understanding how
it comes about that susceptibles are not all attacked at the time of the pandemic prevalence. This difficulty presumably
arises from the ingrained habit of comparing an epidemic to a conflagration. The necessary antecedent to a prairie
fire is a spell of dry weather during which the grass can become like so much tinder; then by accident or design, it
is set ablaze and the fire proceeds apace until the limit of the dry grass is reached, or until it is checked by interposing
a bare interval which the fire cannot overleap. It should be observed, however, that the fire burns only for hours
or days, the growing and drying of the grass occupies months or years. In a fire all the dry grass within reach is
consumed, at any rate there is practically no replacement of it, but in an epidemic, susceptibles are continually
being added to the population: moreover, spread of the epidemic only progresses slowly, for the appropriate
incubation period must intervene between a person being infected and his becoming in turn infectious. Hence, provided
the prairie is sufficiently extensive, i.e., provided the epidemic affects a sufficiently thickly populated area, it
will never go out; and yet it will never have consumed at any one time, any thing like the whole of the grass.