Emergency response to a smallpox attack: The case

for mass vaccination

Edward H. Kaplan*†, David L. Craft‡, and Lawrence M. Wein§
*Yale University School of Management, and Department of Epidemiology and Public Health, Yale School of Medicine, New Haven, CT 06520-8200; and
‡Operations Research Center and §Sloan School of Management, Massachusetts Institute of Technology, Cambridge, MA 02139-3407

Edited by Burton H. Singer, Princeton University, Princeton, NJ, and approved June 11, 2002 (received for review May 10, 2002)

In the event of a smallpox bioterrorist attack in a large U.S. city, the breaks. Surely strategies that contain smallpox epidemics under
interim response policy is to isolate symptomatic cases, trace and the free mixing assumption will also work under mixing patterns
vaccinate their contacts, quarantine febrile contacts, but vaccinate less favorable to the spread of disease. In this sense, our
more broadly if the outbreak cannot be contained by these mea- approach is to plan for the worst but hope for the best.
sures. We embed this traced vaccination policy in a smallpox Absent intervention, each infectious individual generates an
disease transmission model to estimate the number of cases and average of R0 ⫽ ␤S0(0)兾r3 infections early in the epidemic (8),
deaths that would result from an attack in a large urban area. where S0(0) is the number of susceptibles immediately after an
Comparing the results to mass vaccination from the moment an attack at time t ⫽ 0, and ␤ is the effective disease transmission
attack is recognized, we find that mass vaccination results in both rate per person per unit time. Although the transmission rate is
far fewer deaths and much faster epidemic eradication over a wide likely to decrease after the attack is detected and announced, we
range of disease and intervention policy parameters, including use a single value of R0 throughout the epidemic that represents
those believed most likely, and that mass vaccination similarly the postdetection scenario. Because the total number of deaths

SOCIAL SCIENCES
outperforms the existing policy of starting with traced vaccination is highly dependent on the number of infected persons when the
and switching to mass vaccination only if required. emergency response is initiated, an increase in the detection
delay (discussed below) or the predetection R0 has an effect
similar to increasing the initial attack size—an effect we inves-
A lthough smallpox was eradicated in 1979 by the World
Health Organization campaign (1), it remains a feared
bioterrorist threat (2). In the aftermath of the September 11
tigate later.
Newly detected symptomatic infections become ‘‘index cases,’’
terrorist attacks, the U.S. is stockpiling 286 million doses of isolated to prevent further transmission. Index cases are inter-
smallpox vaccine (3). The Centers for Disease Control and viewed to generate a list of c persons with whom they have had
Prevention (CDC) interim response plan (4) calls for targeted potentially infectious contacts in the recent past, but only a
vaccination and quarantine: symptomatic smallpox cases would fraction p of true contacts (and hence only a fraction p of
be isolated, contacts of cases vaccinated, asymptomatic contacts infectees) are named and located via contact tracing. The model
monitored but not isolated, and febrile contacts quarantined for correctly places infected contacts found in their appropriate
stage of disease while accounting for the probability that such
5 days. These guidelines also recommend a broader vaccination
persons were already traced from a prior index, and thus avoids
strategy if the initial number of cases or outbreak locations is
double counting. This crucial aspect of our analysis differs from
‘‘sufficiently large,’’ or if new cases fail to decline after two or
previous contact tracing models (9, 10) and reveals a ‘‘race to
more generations of cases have developed from those initially
trace:’’ even if a contact is identified immediately on detection
identified, or 30% of vaccine stores have been used. An expert
of the infecting index case, the time from infection of the contact
panel has also recommended a targeted vaccination strategy
to detection of the index (and tracing of the contact) can exceed
(ref. 5; http:兾兾www.bt.cdc.gov兾DocumentsApp兾Smallpox兾
the time during which the infected contact remains sensitive to
RPG兾index.asp).
the vaccine. Contact vaccination in such a circumstance is both
ineffective (the infection is not prevented) and inefficient (vac-
Model. To assess the CDC guidelines (and possible alternatives),
cine is wasted). Contact tracing can control a smallpox outbreak
we embed their key features in a disease transmission model of only if the race to trace is repeatedly won.
a smallpox attack in a large urban center. The model incorpo- Susceptible and asymptomatic infected persons located via
rates four stages of infection (1, 6), each exponentially distrib- contact tracing enter a tracing兾vaccination queue serviced by n
uted in duration with mean rj⫺1 days spent in stage j ⫽ 1, 2, 3, or vaccinators who can each process (locate and vaccinate) ␮
4: (1) asymptomatic, noninfectious and vaccine-sensitive; persons per day on average. Only susceptibles and stage 1
(2) asymptomatic, noninfectious, and vaccine-insensitive; infected individuals can be protected via the vaccine, which is
(3) asymptomatic and infectious; and (4) symptomatic and effective with probabilities ␯0 and ␯1, respectively. Unsuccess-
isolated. In our model, the transition from stage 3 to stage 4 fully vaccinated persons return to the freely mixing population
reflects response operations in addition to disease progression. either as susceptibles or infecteds in the appropriate stage of
When an infected recognizes his or her symptoms, which may disease. On vaccination, a fraction h of individuals in stage 3 of
occur some time after the symptoms first appear, (s)he seeks infection are found to be febrile and are quarantined for ␣⫺1 days
medical help and is immediately isolated. A fraction ␦ of on average. A fraction f of vaccinations result in fatalities.
symptomatic cases die of disease, whereas the remaining fraction Mathematically, this model is described by the 17 ordinary
eventually recover (and are immune to reinfection). differential equations discussed further in the Appendix.
Disease transmission derives from free (or homogeneous) We focus on two strategies: traced vaccination (TV), described
mixing among available susceptibles and asymptomatic infec- in the last paragraph, and mass vaccination (MV), where the
tious individuals in the population. Although we recognize that
Downloaded at Indonesia: PNAS Sponsored on March 8, 2020

free mixing is not an accurate depiction of actual population

interactions, free mixing usually leads to larger epidemics than This paper was submitted directly (Track II) to the PNAS office.
nonrandom mixing would admit (7). Given the potentially Abbreviations: CDC, Centers for Disease Control; TV, traced vaccination; MV, mass vacci-
disastrous consequences of a smallpox attack, we ask whether nation.
proposed control strategies can control such ‘‘worst-case’’ out- †To whom reprint requests should be addressed. E-mail: edward.kaplan@yale.edu.

www.pnas.org兾cgi兾doi兾10.1073兾pnas.162282799 PNAS 兩 August 6, 2002 兩 vol. 99 兩 no. 16 兩 10935–10940

 Table 1. Parameter values for the model
Parameter Description Value Reference

␤ Infection rate 10⫺7 person⫺1䡠day⫺1 See text

c Names generated per index 50 1, 9
p Fraction of infectees named by index 0.5 See text
N Population size 107 See text
r1 Disease stage 1 rate (3 days)⫺1 1
r2 Disease stage 2 rate (8 days)⫺1 1
r3 Disease stage 3 rate (3 days)⫺1 1
r4 Disease stage 4 rate (12 days)⫺1 1
n Number of vaccinators 5000 11
␮ Service rate 50兾day (TV), 200兾day (MV) 12
h Fraction febrile in stage 3 0.9 See text
␣ Quarantine rate (5 days)⫺1 4
v0 Vaccine efficacy, stage 0 0.975 13, 14
v1 Vaccine efficacy, stage 1 0.975 13, 14
␦ Smallpox death rate 0.3 1
f Vaccination fatality rate 10⫺6 1
I10(0) Initial number infected 103 See text

entire population simultaneously enters the vaccination queue. tining febriles in Fig. 1 is to reduce the number of TV deaths
The CDC policy, which can be viewed as TV followed by a switch from 112,000 to 110,000 and the number of MV deaths from
to MV if the former is unable to contain the epidemic, will also 585 to 560.
be assessed. We also consider preattack vaccination. To allow for
detection and ramp up, TV and MV are initiated in our model
when 10% of the initial number of infected people have exhibited
symptoms.

Parameter Values. Table 1 contains the parameter values for our

base case. About half of these values are taken from classic
references (e.g., ref. 1). The infection rate ␤ was chosen so that
R0 ⫽ 3, which is near the lower range of 3.5–6 estimated in ref.
9 (although we will vary R0 in sensitivity analyses discussed
below). We assume a large attack of 1,000 initial cases in a
population of 10 million persons in our base case but vary the
attack size between 1 and 100,000 in our sensitivity analyses
below. To estimate the number of vaccinators that could be
requisitioned in an emergency, we focus on public health nurses.
A Health Resources and Services Administration survey (11)
reports that 0.78% of the U.S. population is used in nursing,
whereas 18.3% of used nurses work in public or community
health. Applying these percentages to N ⫽ 10 million and
dividing by 3 to produce round-the-clock 8-hr shifts yields 4,758,
which we have rounded to 5,000. Vaccinators trace and vaccinate
the contacts named by index cases. In our model, it takes T and
V time units, respectively, to trace and vaccinate a named
contact. Hence, ␮⫺1 ⫽ T ⫹ V for TV and ␮⫺1 ⫽ V for MV. We
assume T ⫽ 3V, recognizing that the time required to locate
contacts is much greater than the time required for vaccination
(12), and V⫺1 ⫽ 200 per day based on the 1947 mass vaccination
of New York City (2). When we were in doubt regarding the
values of some parameter values (e.g., p and h), we erred on the
side favoring TV.

Base Case Results. Fig. 1 depicts the population dynamics of TV

and MV, both initiated on day 5, for our base case parameter
values (Table 1). TV leads to 367,000 cases and 110,000 deaths
over 350 days (Fig. 1 A), even though tracing and vaccination
resources are ample (that is, the maximum queue length in Fig.
1 A is less than the total number of available health workers)
Fig. 1. Population dynamics of a smallpox attack with the basic reproductive
Downloaded at Indonesia: PNAS Sponsored on March 8, 2020

and requires 59,000 isolation units for symptomatic cases

ratio R0 ⫽ 3. The number of infected people, regardless of disease stage, and
(but only 750 for febrile quarantine). MV takes only 10 days the number of people, whether susceptible or asymptomatically infected,
to complete, leads to 1,830 cases and 560 deaths over 115 days waiting in the vaccination queue during the aftermath of a smallpox attack.
(Fig. 1B), and requires only 690 isolation units for sympto- The (A) TV policy and the (B) MV policy are both initiated 5 days after an initial
matic cases (60 for febrile quarantine). The effect of quaran- attack of size 1,000 in a population of 107.

10936 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.162282799 Kaplan et al.

 SOCIAL SCIENCES
Fig. 2. Sensitivity analysis for uncertain model parameters. The number of deaths under TV and MV versus (A) the fraction of the population initially infected,
(B) the basic reproductive ratio R0, and (C) the number of vaccinators per capita, keeping the total population size fixed at 107. The number of deaths under TV
versus (D) the fraction of infectees named by an index.

Sensitivity Analysis and Preattack Vaccination. A sensitivity analysis 3 to 0.3 (because only 10% of the population is unvaccinated),
comparing the number of deaths incurred under TV and MV was well below the epidemic threshold. Because TV traces over 80%
performed for the three model parameters that possess the most of the population in the base case, complications due to vacci-
uncertainty (Fig. 2 A–C). TV is more sensitive than MV to the nating the immunosuppressed will be similar for TV and MV.
size of the initial attack (Fig. 2 A), and R0 (Fig. 2B). Although Also, MV eradicates the epidemic much quicker than TV,
deaths decline as the number of vaccinators increases for both making the quarantine of unvaccinated immunosuppresseds a
TV and MV (Fig. 2C, which uses a log scale), the number of viable alternative under MV.
deaths under TV is independent of the number of vaccinators Similarly, the effect of preattack vaccination (15) can be
when the latter exceeds about 2,670, which is the maximum inferred from our model. Vaccinating a fraction ␪ of the
length of the vaccination queue throughout the course of the population before an attack implies that initially a fraction
epidemic in Fig. 1 A. The TV deaths vs. R0 curve has two kinks ␪(1 ⫺ f )␯0 would be immune, a fraction ␪(1 ⫺ f )(1 ⫺ ␯0) would
in Fig. 2B: the curve increases at R0 ⫽ 1.36, which is the largest be susceptible and unsuccessfully vaccinated, a fraction 1 ⫺ ␪
value of R0 for which TV can contain the spread of disease, and would be susceptible and untraced, and a tiny fraction ␪f would
further steepens at R0 ⫽ 5, when the vaccination queue becomes die of vaccine complications, where ␯0 is the vaccine efficacy for
congested. The TV policy is also sensitive to three other TV susceptibles (␯0 ⫽ 0.975 in our base case). A slightly conservative
parameters: the fraction of infectees named by an index (Fig. (i.e., over-) estimate of the postattack fatalities using either TV
2D), the ratio of tracing time to vaccination time, and the number or MV follows by reducing R0 to R0(1 ⫺ ␪␯0) in our model.
of names generated per index (not shown). Consequently, the effect of preattack vaccination, which reduces
Under either TV or MV, there may be a number of people who the gap between TV and MV, can be gleaned from Fig. 2B and
either cannot be located or who should not be vaccinated (e.g., Fig. 3 (described below). For example, if 40% of the population
people who have eczema, are pregnant, or are immunosup- undergoes preattack vaccination, R0 effectively falls from 3 to
pressed). Because prevaccination screening will likely preclude 1.83, and the number of deaths under TV and MV are 40,000 and
the vaccination of the great majority of immunosuppressed 440, respectively.
Downloaded at Indonesia: PNAS Sponsored on March 8, 2020

people, we assess this effect under MV by reducing the vaccine

efficacy. As an illustration, even if the vaccine efficacy is Comparing TV, MV, and the Interim CDC Policy. Fig. 3 shows
decreased from 0.975 to 0.9, the number of deaths increases only breakeven curves, where the total number of deaths from TV
from 560 to 700 in our base case because, although one million and MV are equal, in terms of the initial attack size and R0. For
persons remain susceptible by day 15, R0 essentially drops from the base case in which index cases name half of their contacts

Kaplan et al. PNAS 兩 August 6, 2002 兩 vol. 99 兩 no. 16 兩 10937

 there exists an option to switch from TV to MV after R0 and the
initial attack size have been observed. In this setting, it would be
optimal to initiate TV on day 5 only if the decision maker were very
confident that R0 and the initial attack size are very small.

Discussion
The CDC policy is fashioned after the famed ‘‘surveillance-
containment’’ policy widely credited with the elimination of
smallpox worldwide (1, 4). Our results are not inconsistent with
the historical claims for surveillance containment, for in many of
the smallpox outbreaks the World Health Organization con-
trolled, the estimated initial number of infections was low (1, 6),
prior vaccination and recovered cases had already rendered a
high degree of herd immunity, effectively reducing R0 (8), and
contact tracing was very accurate (high p) because of the physical
layout of affected villages and the relative isolation of their
inhabitants from other areas of settlement (1). The parameters
describing many historical smallpox outbreaks would thus favor
TV over MV, as suggested by Fig. 3B. By the same token, our
breakeven curves (Fig. 3) are not inconsistent with the mass
smallpox vaccination of New York City residents in 1947 in
response to eight cases (2), because a large city has millions of
highly mobile residents and visitors that render contact tracing
difficult (low p), and because the U.S. populace may not tolerate
the length of time required for TV to eradicate a small outbreak.
In summary, compared with MV, TV slows the overall vac-
cination rate, which lengthens the time until herd immunity is
reached. As a consequence, the infection continues to spread,
potentially increasing the tracing兾vaccination burden to the
point at which significant congestion occurs. Such queuing delays
further reduce the effectiveness of vaccination, for contacts are
less likely to be in the vaccine-sensitive stage 1 of disease by the
time they are found. Worded differently, congestion makes
losing the race to trace more likely. This combination of ineffi-
cient and ineffective vaccination could also lead to quarantine
Fig. 3. Breakeven curves for the number of deaths under TV versus MV, as
overrides (because of insufficient quarantine capacity) as the
a function of the basic reproductive ratio R0 and the number of people initially
infected I (0). (A) An index names half of his or her contacts, and (B) an index
number of infections continues to spread.
successfully names all contacts. All other parameters are fixed at their base
level (Table 1).
Conclusion
We have formulated a mathematical model that, to our knowl-
edge, is the first to capture two crucial and interrelated features
(Fig. 3A), even if only several people are exposed in the initial for comparing TV and MV: a detailed analysis of the race to
attack, MV generates less deaths than TV if R0 ⬎ 1.3, which is trace and the queuing caused by limited tracing and vaccination
well below most estimates for smallpox R0 (9). Similarly, with the resources. Our results show that over a wide range of scenarios,
optimistic assumption that index cases successfully name all MV leads to many fewer deaths than TV. Moreover, the cost in
contacts (Fig. 3B), MV results in fewer deaths than TV if R0 ⬎ 2. both deaths and time of delaying the switch to MV—as suggested
Highly asymmetric consequences lurk behind Fig. 3: in the by the interim CDC policy—is very high, particularly in light of
region where TV is favored, the number of deaths under each the public clamoring for vaccination and the mass panic that
strategy is quite small, whereas in the region where MV is would likely ensue in the event of an attack. For these reasons,
favored, TV may generate many thousands of deaths over a given our current states of knowledge and response capabilities
number of months. Moreover, if the MV vs. TV decision ever has [e.g., there is no widely accepted screening test for smallpox (5)],
to be made, the value of R0 will not be known more precisely than we believe that unless preattack vaccination is used to signifi-
at present, and an accurate estimate of the initial attack size may cantly increase herd immunity in the population, serious con-
not be available (ref. 16; http:兾兾www.hopkins-biodefense.org兾 sideration should be given to replacing the existing CDC policy
lessons.html). Hence, a decision maker concerned with mini- by MV in the event of a smallpox attack in a large urban center.
mizing the number of deaths in the face of this uncertainty
should choose TV only if (s)he is quite certain that R0 and the Appendix
initial attack size fall in the TV-favorable region of Fig. 3. Throughout, the subscript j denotes stage of infection, whereas
We model the CDC guidelines by switching from TV to MV 28 the superscript ᐉ denotes whether a person has yet to be traced
days (i.e., two generations, where ¥3j ⫽ 1 rj⫺1 ⫽ 14 days) after the start (ᐉ ⫽ 0) or has already been traced and vaccinated (ᐉ ⫽ 1). All
of TV (e.g., on day 33 in the base case). Doing so leads to 4,680 parameters used are defined in Table 1. The state variables are:
deaths and 15,570 cases over 165 days in the base case, whereas the Sᐉ ⫽ number of type ᐉ susceptibles; Ijᐉ ⫽ number of type ᐉ
epidemic peaks at 38 days with 11,400 persons infected. The cost of infected persons in disease stage j (j ⫽ 1, 2, 3, 4); Qj ⫽ number
Downloaded at Indonesia: PNAS Sponsored on March 8, 2020

waiting to switch from TV to MV is very high—4,120 incremental in tracing兾vaccination queue at disease stage j (j ⫽ 0 means
deaths in our base case compared with responding with MV from susceptibles); H ⫽ number in febrile quarantine; Z ⫽ number
the start. Consider a strategy that has two decision points: on day immune (from vaccination) or recovered from smallpox; D ⫽
5, the TV vs. MV decision is made with imprecise information about number dead. We also denote the total number of freely mixing
R0 and the initial attack size, and (as in the CDC plan) on day 33 infectious individuals as I3 ⫽ I03 ⫹ Q3 ⫹ I13, and the total number

10938 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.162282799 Kaplan et al.

 3
of persons in the tracing兾vaccination queues as Q ⫽ ¥j⫽0 Q j. dH
Finally, define ␶ to be the time interval between the smallpox ⫽ 共1 ⫺ f兲h ␮ Q 3 min共1, n兾Q兲 ⫺ r3H ⫺ ␣H, [8]
dt
attack and the initiation of intervention. Whereas in the text the
attack occurs at time t ⫽ 0 and the response begins at time t ⫽
as quarantined individuals do not contribute to disease trans-
␶ (5 days in our base case), to ease the presentation in the
equations below, we let the attack occur at time t ⫽ ⫺␶ and mission but continue to progress to smallpox.
the response begin at time t ⫽ 0. These assumptions lead to the
Traced but Unsuccessfully Vaccinated States. All unsuccessfully
following ordinary differential equations.
vaccinated individuals (except those in quarantine or vaccine
Untraced States. Disease transmission occurs in accord with the fatalities) freely mix in the population, contributing to disease
usual mass action law, whereas disease progression occurs with transmission, while disease progression occurs in the usual way.
stage-dependent constant hazard rates. Previously untraced
contacts who are named by their true index of infection and dS 1
⫽ 共1 ⫺ f兲共1 ⫺ v 0 兲 ␮ Q 0 min共1, n兾Q兲 ⫺ ␤S1I3 [9]
found are traced by their appropriate stage of disease accounting dt
for their (random) time of infection by the index {[p␭j(t)] terms
in Eqs. 2 and 3}. In contrast, those previously untraced contacts dI 11
who are susceptible, or infected but named by someone other ⫽ ␤ S 1 I 3 ⫹ 共1 ⫺ f兲共1 ⫺ v 1 兲 ␮ Q 1 min共1, n兾Q兲 ⫺ r1I11
dt
than their true index of infection, trace proportionately over all [10]
untraced individuals {[c ⫺ pR0(t)] terms in Eqs. 1–3; both
functions R0(t) and ␭j(t) will be explained further shortly}. Note dI 12
that no tracing occurs before time t ⫽ 0. ⫽ r 1 I 11 ⫹ 共1 ⫺ f兲 ␮ Q 2 min共1, n兾Q兲 ⫺ r2I12 [11]
dt
0 0
dS S
⫽ ⫺␤I3S0 ⫺ 关c ⫺ pR0共t兲兴 r3I3 [1]

SOCIAL SCIENCES
dt N dI 13
⫽ r 2 I 12 ⫹ 共1 ⫺ f兲共1 ⫺ h兲 ␮ Q 3 min共1, n兾Q兲 ⫹ ␣H ⫺ r3I13

再 冎
dt
dI 01 I 01 [12]
⫽ ␤ I 3 S 0 ⫺ 关c ⫺ pR 0 共t兲兴 ⫹ p ␭ 1 共t兲 r 3 I 3 ⫺ r 1 I 01
dt N
[2] dI 14
⫽ r 3 共I 13 ⫹ Q 3 ⫹ H兲 ⫺ r 4 I 14 . [13]

再 冎
dI j0 I j0 dt
⫽ r j ⫺ 1 I 0j ⫺ 1 ⫺ 关c ⫺ pR 0 共t兲兴 ⫹ p ␭ j 共t兲 r 3 I 3 ⫺ r j I j0
dt N Immune兾Recovered and Death States. Successful vaccinations oc-
for j ⫽ 2, 3 [3] cur with probability ␯0 and ␯1 for susceptibles and those in stage
1 disease, respectively, whereas the fraction 1 ⫺ ␦ of those who
dI 04
⫽ r 3 I 03 ⫺ r 4 I 04 . [4] progress to symptomatic smallpox eventually recover (on aver-
dt age r⫺1 days after development of symptoms) and remain
4
immune. However, the fraction ␦ of those who develop smallpox
Queueing States. Until t ⫽ 0, there is no tracing and hence no die of the disease, whereas the fraction f of all those vaccinated
queuing. Once t ⬎ 0, note that because only n tracers兾vaccinators die of vaccine-related complications.
are available, the total flow out of the queuing states can never
exceed n␮ per day. If the system becomes congested (more than
dZ
n persons are in the queue), then those in queue in disease stage ⫽ 共1 ⫺ f兲共v 0 Q 0 ⫹ v 1 Q 1 兲 ␮ min共1, n兾Q兲 ⫹ 共1 ⫺ ␦兲r4共I04 ⫹ I14兲
j receive service at rate n␮Qj兾Q, that is, the service provided is dt
proportional to the relative numbers in queue. This explains the [14]
min(1, n兾Q) in Eqs. 5–7. Disease transmission and progression
continue unabated among those in the queue. dD
⫽ f ␮ Q min共1, n兾Q兲 ⫹ ␦r4共I04 ⫹ I14兲. [15]
dt
dQ 0 S0
⫽ 关c ⫺ pR 0 共t兲兴 r I ⫺ ␤ I 3 Q 0 ⫺ ␮ Q 0 min共1, n兾Q兲
dt N 3 3 The Functions R0(t) and ␭j (t). When a newly symptomatic smallpox
[5] case is discovered at time t, she will have infected on average

再 冎
R0(t) persons over her duration of infectiousness, where
dQ 1 I 01

冕
⫽ ␤ I 3 Q 0 ⫹ 关c ⫺ pR 0 共t兲兴 ⫹ p ␭ 1 共t兲 r 3 I 3
dt N t⫹␶
R 0 共t兲 ⫽ e⫺r3x␤关S0共t ⫺ x兲 ⫹ Q0共t ⫺ x兲 ⫹ S1共t ⫺ x兲兴 dx
⫺ ␮ Q 1 min共1, n兾Q兲 ⫺ r1Q1 [6]
0

dQ j
dt
再 I j0
⫽ r j ⫺ 1 Q j ⫺ 1 ⫹ 关c ⫺ pR 0 共t兲兴 ⫹ p ␭ j 共t兲 r 3 I 3
N
冎 ⬇
␤关S0共t兲 ⫹ Q0共t兲 ⫹ S1共t兲兴
r3
, [16]

⫺ ␮ Q j min共1, n兾Q兲 ⫺ rjQj for j ⫽ 2, 3 [7]

with the quasistationary approximation (i.e., the epidemic is
Stage 3 Quarantine (Holding) State. Under CDC policy, in addition viewed as fixed on a short time scale) following from the fact that
to symptomatic cases, only febrile individuals are subject to persons are infectious for only 3 days on average, which is very
short relative to the duration of the overall epidemic. Note that
Downloaded at Indonesia: PNAS Sponsored on March 8, 2020

quarantine. We assume that a fraction h of infectious (stage 3)

individuals are identified as febrile and are remanded to quar- R0(0) ⫽ ␤S0(0)兾r3 ⫽ R0, as defined in the text. Of the R0(t)
antine of mean duration ␣⫺1 days at the time they are processed persons previously infected by an index detected at time t, some
in the vaccination queue (except for vaccine fatalities), which will already have been named at random by other index cases
leads to even though no transmission occurred. The rate with which

Kaplan et al. PNAS 兩 August 6, 2002 兩 vol. 99 兩 no. 16 兩 10939

 anyone in the population is randomly traced at time u ⱖ 0 is given where qj(t) is (approximately) the conditional probability that a
by contact of an index detected at time t is in stage j of disease given
that the contact has not been traced by time t, and equals
关c ⫺ pR 0 共u兲兴r 3 I 3 共u兲
␬ 共u兲 ⫽ [17]

写
N j⫺1
rk r 3 ⫹ ␬ 共t兲
q j 共t兲 ⫽ ⫻ . [19]
[␬(u) ⫽ 0 otherwise]. Therefore, the expected number of un- r k ⫹ r 3 ⫹ ␬ 共t兲 r j ⫹ r 3 ⫹ ␬ 共t兲
k⫽1
traced contacts previously infected by an index detected at time
t who are in disease stage j when the index is detected is equal to
The justification for the approximation in Eqs. 18 and 19 is the

冕 t⫹␶ same quasistationary argument used in Eq. 16.

t
␭ j 共t兲 ⫽ e⫺ r 3x ␤ S 0 共t ⫺ x兲e⫺兰 t ⫺ x␬ 共u兲du
0
We thank Ellis McKenzie of the Fogarty International Center, National
Institutes of Health, for initiating this project. This work was supported
⫻ Pr兵Contact in stage j at t 兩 infected at t ⫺ x其 dx by the Societal Institute for the Mathematical Sciences, via Grant
DA-09351 from the National Institute on Drug Abuse (E.H.K.) and a
␤S0共t兲 grant from the Singapore–Massachusetts Institute of Technology Alli-
⬇ qj共t兲 , [18]
r3 ⫹ ␬共t兲 ance (L.M.W.).

1. Fenner, F., Henderson, D. A., Arita, I., Jezek Z. Ladnyi, I. D. (1988) Smallpox 10. Müller, J., Kretzschmar, M. & Dietz, K. (2000) Math. Biosci. 164,
and Its Eradication (World Health Organization, Geneva). 39 – 64.
2. Henderson, D. A. (1999) Science 283, 1279–1282. 11. Division of Nursing, Bureau of Health Professions, Health Resources and
3. Check, E. (2001) Nature (London) 414, 677. Services Administration (2001) The Registered Nurse Population (Health Re-
4. Centers for Disease Control (2001) CDC Interim Smallpox Response Plan and sources and Services Administration, Washington, DC).
Guidelines (Centers for Disease Control, Atlanta, GA), Draft 2.0. 12. Macke, B. A., Hennessy, M., McFarlane, M. M. & Bliss, M. J. (1998) Sex.
5. Henderson, D. A. (1999) J. Am. Med. Assoc. 281, 2127–2137. Transm. Dis. 25, 561–568.
6. Meltzer, M. I., Damon, I., LeDuc, J. W. & Millar, J. D. (2001) Emerg. Infect. 13. Mack, T. M. (1972) J. Infect. Dis. 125, 161–169.
Dis. 7, 959–969. 14. Mack, T. M., Thomas, D. B. & Khan, M. M. (1972) Am. J. Epidemiol. 95,
7. Kaplan, E. H. (1991) Math. Biosci. 105, 97–109. 157–168.
8. Anderson, R. M. & May, R. M. (1991) Infectious Diseases of Humans (Oxford 15. Bicknell, W. J. (2002) N. Engl. J. Med. 346, 1323–1325.
Univ. Press, Oxford). 16. O’Toole, T. & Inglesby, T. (2001) Shining Light on Dark Winter (Johns Hopkins
9. Gani, R. & Leach, S. (2001) Nature (London) 414, 748–751. Center for Civilian Biodefense Strategies, Baltimore).
Downloaded at Indonesia: PNAS Sponsored on March 8, 2020

10940 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.162282799 Kaplan et al.