M A J O R A R T I C L E
Largest Measles Epidemic in North America in
a Decade—Quebec, Canada, 2011: Contribution
of Susceptibility, Serendipity, and
Gaston De Serres,1,2France Markowski,3Eveline Toth,3Monique Landry,3Danielle Auger,3Marlène Mercier,3
Philippe Bélanger,3Bruno Turmel,3Horacio Arruda,3Nicole Boulianne,1Brian J. Ward,4and Danuta M. Skowronski5
1Institut National de Santé Publique du Québec,2Department of Social and Preventive Medicine, Laval University, and3Ministère de la Santé et des
Services Sociaux du Québec, Quebec City,4Research Institute of the McGill University Health Centre, Montréal, Quebec; and5British Columbia Center
for Disease Control, Vancouver, Canada
Canada, where rates of 1- and 2-dose vaccine coverage among children 3 years of age were 95%–97% and 90%,
respectively, with 3%–5% unvaccinated.
Methods.Case patients identified through passive surveillance and outbreak investigation were contacted to
determine clinical course, vaccination status, and possible source of infection.
Results.There were 21 measles importations and 725 cases. A superspreading event triggered by 1 importa-
tion resulted in sustained transmission and 678 cases. The overall incidence was 9.1 per 100 000; the highest
incidence was in adolescents 12–17 years old (75.6 per 100 000), who comprised 56% of case patients. Among
adolescents, 22% had received 2 vaccine doses. Outbreak investigation showed this proportion to have been an
underestimate; active case finding identified 130% more cases among 2-dose recipients. Two-dose recipients had
milder illness and a significantly lower risk of hospitalization than those who were unvaccinated or single-dose
Conclusions.A chance superspreading event revealed an overall level of immunity barely above the elimina-
tion threshold when unexpected vulnerability in 2-dose recipients was taken into account. Unvaccinated individu-
als remain the immunization priority, but a better understanding of susceptibility in 2-dose recipients is needed
to define effective interventions if elimination is to be achieved.
The largest measles epidemic in North America in the last decade, occurred in 2011 in Quebec,
Keywords.measles; epidemic; underreporting; surveillance; outbreak; elimination.
In 1995, Canada and other countries of the Americas
adopted a measles elimination goal [1, 2]. Toward this
goal, mass immunization campaigns targeting children
aged 1–17 years old were conducted in Canada in
1996–1997 to provide a second dose of a measles-
containingvaccine.A routine pediatric2-dose
measles-mumps-rubella (MMR) schedule was also in-
troduced at that time [1, 2]. The incidence of measles
per 100000 rapidly declined in Canada, from 8.03 in
1995 to 0.65 in 2000 . Between 2001 and the end of
2010, 360 cases were reported, with an annual inci-
dence varying between 0.02 and 0.31 per 100000 .
Three outbreaks involving ≥10 cases were reported
during that period: 2007 (Quebec 94 cases), 2008
(Ontario 56 cases), and 2010 (British Columbia 82
In 2011, Canada experienced several measles impor-
tations—mostly from Europe, where >30 000 cases
had been reported . Although these imported cases
were generally followed by limited or no secondary
transmission, 1 triggered the largest epidemic in
North America since 2001. This epidemic occurred in
Received 4 June 2012; accepted 23 October 2012; electronically published 21
Correspondence: Gaston De Serres, MD, PhD, Institut National de Santé Publi-
que du Québec, 2400 d’Estimauville, Quebec City, QC, Canada, G1E 7G9 (gaston.
The Journal of InfectiousDiseases2013;207:990–8
© The Author 2013. Published by Oxford University Press on behalf of the Infectious
Diseases Society of America. All rights reserved. For Permissions, please e-mail:
990 • JID 2013:207 (15 March) • De Serres et al
by guest on December 28, 2015
the province of Quebec (total population 7.9 million) where
the last major outbreak (>10 000 cases) had occurred in 1989
when the program included only 1 dose of vaccine adminis-
tered at 12 months of age . In 1996, a school-based mass
immunization campaign was conducted during which 89% of
children 5–17 years old received a second measles dose. That
same year, a routine 2-dose program was introduced with a
schedule of administration at 12 and 18 months of age. Al-
though immunization is voluntary and there are no vaccine
requirements for school enrolment in Quebec, provincial
vaccine coverage surveys conducted in 2006, 2008, and 2010
consistently showed that by 24 months of age, approximately
96% of children had received 1 dose and approximately 85%
had received 2 doses of measles vaccine, increasing to 97%
and 90%, respectively, by 28 months of age [8–10]. With addi-
tional first and second doses administered between 28 and 59
months of age, population measles vaccine coverage is even
higher by school entry. Against this background of high im-
munization coverage, we describe key epidemiological features
of the 2011 epidemic in Quebec and highlight observations
that may be applicable in the broad context of measles
Measles is notifiable by Quebec physicians and laboratories.
According to the national surveillance case definition, measles
cases are either laboratory confirmed or epidemiologically
linked . Laboratory confirmation is defined virologically
by culture or polymerase chain reaction or serologically by
measles-specific immunoglobulin M (IgM). Cases in patients
with fever ≥38.3°C (101°F) and cough or coryza or conjuncti-
vitis and a generalized maculopapular rash for ≥3 days with
an epidemiological link to a laboratory-confirmed case are
considered epidemiologically linked. Reported cases were in-
vestigated by a public health practitioner for information on
the clinical course of disease, vaccination status, and possible
source of infection.
Since vaccine coverage rates were thought to be high
enough to block sustained transmission , public health in-
terventions in this outbreak focused on administering immu-
noglobulin to persons at higher risk of measles complications
(unprotected pregnant women, infants, and immunocompro-
mised patients). Unvaccinated contacts were offered immuni-
zation but large-scale supplementary vaccination efforts were
not deployed, and there was no school exclusion policy for
incompletely immunized children.
Case vaccination status was determined by written record.
Products and schedules were per the provincially determined
Quebec program for which MMR II (Merck Canada) was the
only vaccine used from 1996–2008, replaced in 2008 by the
GSK vaccines Priorix (MMR) and Priorix Tetra (MMR-
varicella). A monovalent measles vaccine (Connaught Laborato-
ries) was used in the 1996 school-based, second-dose campaign.
Underreporting was estimated in the school at the origin of
this epidemic by dividing the number of cases determined by
active investigation by the number previously reported by
passive surveillance alone. Active investigation included a
questionnaire that was distributed and answered by students
in school in a single day regarding prior measles symptoms,
review of the school absenteeism registry for potential measles
attribution, calls to parents to confirm student information,
and serological testing of potential case patients whose symp-
toms did not meet national case criteria.
Proportions were compared statistically using the χ2test and
means by Student’s t test.
In 2011, 776 measles cases were passively reported in Quebec:
295 (38%) were laboratory confirmed, 430 (55%) were epide-
miologically linked cases and 51 (7%) were clinical cases
without an epidemiological link. These latter cases were ex-
cluded, leaving 725 that met the national measles case defini-
tion (including those subsequently identified through active
investigation). Of these 725 confirmed cases, 678 were the
result of sustained transmission after a superspreading event;
the remaining 47 cases occurred as a result of importations
and/or discrete, limited clusters. Among the 118 strains char-
acterized, 116 were genotype D4, and 2 were genotype B3.
Initial Importations and Ensuing Outbreaks
There were 22 documented measles imported cases in 2011
corresponding to 19 episodes of importation; 5 cases belonged
to 2 families whose members had traveled together in France
and developed symptoms within 14 days upon returning.
Among these episodes, 12 occurred between January and June
and 7 between July and December. Measles was imported
after travel to France (12 episodes), other European countries
(4 episodes, including 2 where travelers transited through
Paris airports), and Pakistan (1 case). As described below, the
imported case at the origin of the small cluster detected in
September was never identified. The index case patient for the
superspreading event had traveled to the Caribbean, but, as
described below, had most likely acquired measles at a Cana-
2011 Measles Epidemic in Quebec • JID 2013:207 (15 March) • 991
by guest on December 28, 2015
Only 4 of the 21 importations resulted in outbreaks, defined
as ≥2 confirmed cases, linked epidemiologically, virologically,
or both. The first outbreak occurred after an affected traveler
returned from France on 2 February and developed rash on 5
February. This case triggered an outbreak of 20 secondary
cases with protracted transmission over 11 weeks (≥6 genera-
tions). The second importation in April caused the large out-
break (678 cases), described below. The third importation in
September caused only 1 secondary case. The fourth importa-
tion occurred in late September, causing 4 secondary cases. It
involved a B3 genotype virus, isolated in an adolescent from a
region unaffected by the large outbreak thus distinguishing it
as a separate cluster from the larger epidemic. This patient’s
rash began on 24 September, and the patient had neither a
travel history nor reported contact with a traveler. That B3
virus was also found in another adolescent from the same
school, whose rash began 6 days later. These 2 patients infect-
ed 1 schoolmate and 2 household contacts, but no further
transmission was detected.
The index patient for the large outbreak was a 30–39-year-
old adult who became feverish on 2 April, 9 days after return-
ing to Canada from a 1-week holiday in the Caribbean .
When returning to Canada on 24 March, this individual spent
several hours at the Montreal airport with thousands of fellow
travelers returning from spring break. Because of the 9-day in-
terval between the return flight and the patient’s fever onset
and the absence of a recognized measles contact while the
patient was abroad, the Montreal airport was considered the
most likely source of infection. However, no other measles
cases associated with this airport were identified, and this at-
tribution is speculative; unrecognized exposure in the Caribbe-
an remains possible. This index case patient had received 1
dose of measles vaccine in childhood and was employed in a
high school. This patient remained in contact with several
groups of students for 3 days after developing nonspecific
symptoms but stopped working when rash appeared (7 April).
The patient sought care that day, and measles was confirmed
by polymerase chain reaction. The virus was isolated from sec-
ondary cases only and identified as genotype D4. Between 17
and 23 April, 10 secondary case patients became sick in this
high school, followed 7–14 days later by 61 additional cases
(Figure 1). Although this school outbreak, affecting 110 stu-
dents and 4 other staff members, was almost over by 20 May,
further substantial transmission then ensued in the adjacent
rural town (approximately 70000 inhabitants) and surround-
ing area. High transmission continued until the summer
recess (23 June) and decreased thereafter.
Ultimately this outbreak included 678 cases and lasted 26
weeks (7 April to 4 October). The region of the “index” high
school accounted for 72% of all cases (499 of 678), the neighbor-
ing region reported 20% (135 of 678 cases), and 6 other regions
in the province reported the remaining cases. Montreal was
almost entirely spared, with only 4 cases linked to this outbreak.
For 2011, the overall Quebec measles incidence was 9.1 per
100 000 population. The mean age of case patients was 15
years (range, 3 months to 53 years). The incidence was highest
Numbers of all measles cases, imported cases, and cases associated with the initial school outbreak, by week of rash onset in 2011.
992 • JID 2013:207 (15 March) • De Serres et al
by guest on December 28, 2015
in adolescents 12–17 years (75.6 per 100000), and this age
band comprised 56% of all case patients (Table 1; Figure 2).
The incidence was 3 times lower in infants <12 months old
(24.1 per 100000) and in that same range for children up to
12 years old. Case distribution by vaccination status varied by
age (Figure 3). Although only 4.1% of case patients aged 1–9
years had received 2 doses, in case patients aged 10–19 and
20–29 years, the proportions of 2-dose recipients were signifi-
cantly higher, at 20% and 11%, respectively (P<.001). Among
the 56 case patients aged 30–39 years, 48% had written proof
of 1 dose (Table 1; Figure 3). All but 1 of the 13 case patients
≥40 years old were unvaccinated. According to the Canada’s
National Advisory Committee on Immunization, individuals
born before 1970 are considered naturally immune and there-
fore not vaccine eligible.
Among the 21 infant case patients, 4 (19%) were hospitalized,
but none had pneumonia or serious complications (Table 1).
Among the 598 pediatric case patients 1–19 years old, 57 (9.5%)
were hospitalized. The risk was 3% in 2-dose recipients (3 of 98
patients) but significantly higher in both unvaccinated children
(10.8%; 43 of 398; P=.02) and 1-dose recipients (15.6%; 5 of
32; P=.01). In adults (≥20 years old), the overall risk of
Table 1. Hospitalization by Age and Vaccination Status
Hospitalizations/Cases by Vaccination Status, No. (%)
Age, y Cases, No. Incidence per 100000Unvaccinated 1 Dose
≥2 Doses Unknown or No Proofa
000 4/21 (19)
aUnknown vaccination status or no written proof of vaccination.
Number and incidence of measles cases by patient age.
2011 Measles Epidemic in Quebec • JID 2013:207 (15 March) • 993
by guest on December 28, 2015
hospitalization was 24%, similar to that in unvaccinated patients
(26%; 12 of 46) and single-dose recipients (26%; 8 of 31).
Exposures and Sources of Infection
Among the 725 case patients overall for the province, 84%
(611 patients) reported a known source of infection; daycare
or school was the most frequent source (48%), followed by
household exposure (19%).(Table 2) In children <1 year old,
community and daycare centers were the main sources of in-
fection (in 29% and 24%, respectively), and household expo-
sures accounted for only 14%. Household exposure was
reported by 20% of the 106 adult case patients aged ≥20 years.
Exposure in healthcare settings was reported by 3% (16 pa-
tients and 3 healthcare workers). Eight additional healthcare
workers were infected; no source could be identified in 5, and
3 had community exposure.
Estimation of Underreporting
Because of documented transmission in healthcare settings, case
patients in the neighboring rural town and surrounding area
were instructed as of 6 May to consult a physician only if their
condition was deteriorating and otherwise to avoid emergency
roomsor medicalclinics. This strategyresultedin
underreporting, which was assessed in the outbreak school. In
that school before the investigation, 77 cases had been reported
through passive surveillance (Table 3). The investigation identi-
fied 21 more cases meeting the national surveillance case defini-
tion and 12 additional attenuated cases . The patients with
attenuated cases had symptoms of measles but did not fully
meet the national case definition: blood samples collected 8
weeks after symptom onset showed no detectable IgM but im-
munoglobulin G (IgG) levels too high to be attributed to child-
hood immunization. As a result of active case finding and
ascertainment of vaccination status, the number of unvaccinated
case patients increased by 21% but more than doubled (130%
increase) in 2-dose recipients (P<.001). The proportion of cases
with written proof of 2-dose measles immunization shifted from
30% before the investigation (similar to the 22% observed
among the 404 cases reported provincially in the 12–17-year age
group) to 48%.
Given the large epidemic in Europe and the popularity of travel
between Quebec and France, multiple measles importations to
Vaccination status of case patients, by age group. Unknown/no proof indicates unknown vaccination status or no written proof of
994 • JID 2013:207 (15 March) • De Serres et al
by guest on December 28, 2015
Quebec in 2011 were to be expected. However, the explosive
school outbreak, followed by intense transmission that culmi-
nated in the largest epidemic in North America since 2001,
was unexpected given the province’s documented immuniza-
tion coverage. Based on Pan American Health Organization
criteria for elimination, defined as the absence of continuous
transmission for 12 months, this epidemic is not considered
evidence of reestablished endemic transmission . The out-
break did eventually cease without aggressive interventions
aimed at halting transmission. However, the substantial com-
munity spillover and sustained spread through the summer
months revealed a population susceptibility precariously close
to the epidemic threshold. Immunization coverage in Quebec
is similar to that reported in other North American jurisdic-
tions. As such, the Quebec experience raises critical questions
related to measles surveillance, vaccination strategies, and the
combined roles of susceptibility, serendipity, and superspread-
ing events in determining the success of measles elimination
Surveillance of measles was enhanced in the outbreak
context but this report probably underestimates the number of
measles cases. The proportion of hospitalized case patients
(11%) was lower than those reported in 2011 in the United
States (30%) and Europe (28%) or in 2012 in the United
Kingdom (18%) [15–17]. Although this may be partly attribut-
ed to healthcare system differences, patient age, or vaccination
status, it also suggests greater sensitivity in the detection of
milder cases and a lower level of underreporting in Quebec
than elsewhere. We also excluded 51 clinical cases that oc-
curred during the epidemic period but lacked a recognized ep-
idemiological link. A proportion of these clinical cases would
also have been confirmed had the patients’ physicians request-
ed laboratory testing. Even if vaccinated case patients are rec-
ognized by clinicians and notified, they may not meet all
clinical orserological (IgM-based)
[18–20]. Although paired IgG testing on acute and convales-
cent samples could contribute to confirmation, it is infre-
quently requested and cumbersome. Thus, passive surveillance
approaches based on classic clinical case definitions or detec-
tion of IgM may systematically exclude 2-dose recipients from
official tallies . If >90% of adolescents have received 2
MMR/measles vaccine doses, <5% are unvaccinated, and the
Table 2. Reported Source of Infection by Age Group
Cases by Infection Source, No. (%)
Age, y Cases, No.Household Daycare or School CommunityHealthcare Setting Travel Unknown
veillance, and Percentage Increase
Student Cases in Outbreak School Reported to Public Health Before Active Investigation, Cases Identified With Active Sur-
Cases by Vaccination Status, No. (% Increase)
Identification of Casesa
Unvaccinated1 Dose2 DosesUnknown Status or No ProofTotal
Before active investigation
After active investigation
Notifiable plus attenuated cases
434 237 77
aNonstudent cases were not included because active surveillance investigation was conducted only in students. Notifiable cases met the national surveillance
case definition. Attenuated cases included symptoms of measles but did not meet the national surveillance case definition; serum samples collected from these
patients 8 weeks after symptom onset showed no detectable immunoglobulin M but immunoglobulin G levels too high to be attributed to childhood
immunization, most likely confirming a recent infection.
2011 Measles Epidemic in Quebec • JID 2013:207 (15 March) • 995
by guest on December 28, 2015
2-dose vaccine effectiveness is approximately 95%, we would
expect about 50% of the case patients to have received 2 doses,
if all cases were reported. However, underreporting was much
greater among 2-dose recipients than among unvaccinated
case patients in our study. Active surveillance in the most af-
fected high school detected more than twice (130% increase)
the number of cases in 2-dose recipients than initially report-
ed, increasing their contribution to the total from 30% to 48%
(Table 1). Milder illness in 2-dose recipients (ie, lower hospi-
talization rates and attenuated presentation) in combination
with advice to avoid physician consultation for mild illness
probably led to a disproportionate reduction in notification by
passive surveillance and overestimation of the contribution of
unvaccinated individuals. Although this underestimation may
have been greatest in the most affected school, a significant
level of underreporting probably also pertains outside this
setting. The significance of attenuated cases in vaccinated pa-
tients to the overall goal of elimination is unknown but will
depend on the extent to which they contribute to total trans-
mission; this contribution is limited now relative to that of un-
vaccinated case patients but may become more relevant as the
elimination target is approached.
This outbreak raises other important questions concerning
the relative contributions of vaccine failure versus failure to vac-
cinate. As previously reported for the high school where the
large epidemic started, the vaccine effectiveness in 2-dose recipi-
ents was 95.5% (95% confidence interval, 93.8%–96.7%) without
attenuated cases and 94.2% (95% confidence interval, 92.9%–
95.6%) when attenuated cases were included . This estimate
is similar to the median value (94.1%) reported in a synthesis of
2-dose vaccine effectiveness studies . In the initial Quebec
outbreak school, protection in 2-dose recipients did not change
with greater intervals between the first and the second doses but
increased from 93% to 97.5% when the first dose was adminis-
tered at ≥15 months of age rather than at 12 months . A
follow-up study including Quebec cases in 2-dose recipients
outside this outbreak school confirmed the greater risk of
measles when the first dose was administered at 12–13 months
vs ≥15 months of age . These finding are consistent with
results from a large serological study of twice-vaccinated
German children aged 1–17 years (approximately 7000 children)
in which the proportion lacking measles antibody decreased
with older age at first vaccination up to 17 months of age. .
Two long-term prospective serological studies have shown per-
sistence of measles antibodies in twice-vaccinated children, but
these studies included mostly  or exclusively  children
whose first dose was administered at ≥14 months of age. Find-
ings from our epidemiological study [12, 22] and the German
serological study  pertain mostly to children born to previ-
ously infected mothers. Caution is required in extrapolating to
children born to vaccinated mothers, who are expected to have
less maternal antibody to interfere with vaccines .
No extensive vaccine coverage data were available for our
analysis among adolescents (aged 12–17 years), who com-
prised 56% of the cases and were the driving force of this epi-
demic. Indeed, the high school at the center of this epidemic,
which contributed nearly one-quarter of all cases in this age
group, probably represents the “worst-case scenario,” with
4.7% of students unvaccinated. However, nearly half (48%) of
the cases in this school outbreak were due to vaccine failures
in 2-dose recipients. Overall, 85% of the students in this
school had written proof of receiving 2 doses, slightly less
than the 91%–92% reported provincially in 14–15-year-old
students after the routine grade 9 immunization update
(Ministère de la santé et des Services Sociaux du Québec un-
published data). In the surrounding region, where several
hundreds of cases occurred, a 2008 vaccine coverage survey in
28-month-old children found that 5.4% were unvaccinated
and 94.6% and 89.1% had received 1 or 2 doses, respectively,
marginally less than the provincial estimates cited above (97%
and 90%) [8–10]. With the addition of children receiving their
first measles dose between age 28 months and school entry,
the proportion of school-age children unvaccinated in that
region was probably <4%. Although some heterogeneity in
vaccine coverage may exist, this region is not known for clus-
tered vaccine opposition. Whereas a school outbreak may
reflect local conditions, sustained transmission involving an
entire region probably provides a robust and widely applicable
signal. It suggests that with 3%–5% of the population unvacci-
nated, the additional susceptibility in 1- and 2-dose vaccine
recipients can push the population toward a critical tipping
point for epidemic risk.
Several other jurisdictions in North America have similar
immunization coverage, with approximately 90% 2-dose vacci-
nated and ≥3% unvaccinated [27–30]. In Canada, 5%–8% of
children have not yet received a first dose of measles vaccine
by 24 months [27, 28], compared with 7.7%–10% by 19–35
months in the United States  In kindergarten, rates of ex-
emptions, including medical, religious, and philosophical ex-
emptions, range from <1% to 6.2%, with a total exemption
rate of ≥3.0% in 15 of 48 US jurisdictions . In US adoles-
cents, about 90% have received 2 MMR doses . The
absence of measles in jurisdictions with the highest proportion
unvaccinated is unlikely to be due to a greater level of popula-
tion immunity. It more likely reflects variability in the chance
occurrence of superspreading opportunities. Serendipity prob-
ably played a role in the 2011 Quebec outbreak through the
index case, including initial acquisition (chance exposure at an
airport) and a social context (school worker) favoring broad
and swift transmission. The initial superspread happened in
the context of accumulated susceptibility in students who
themselves had ample transmission possibilities through their
own rich social networks. Absent this combination of propi-
tious conditions, there may have been no onward or sustained
996 • JID 2013:207 (15 March) • De Serres et al
by guest on December 28, 2015
transmission. Montreal, the largest city in the province, is
located just100 km from the most-affected region, has compa-
rable immunization coverage, but was mostly unaffected
(4 cases), probably reflecting a combination of luck, seasonal-
ity, and social networking. Had this outbreak begun in early
winter rather than spring, it could have been much greater in
intensity and geographic involvement.
The susceptibility targets of the WHO European Region for
measles elimination—<15% in children aged 2–4 years, <10%
in those aged 5–9 years, and <5% in those aged ≥10 years—
are based on deterministic models, assuming homogeneous
distribution of susceptibility [32, 33]. With the typical cluster-
ing of unvaccinated individuals because of shared socioeco-
nomic conditions, religious, or philosophical beliefs, the true
level of compensatory immunity required among the vaccinat-
ed may be higher. Models with heterogeneity in transmission
(some superspreaders and many low transmitters) have shown
that measles has difficulty penetrating when the population is
under the elimination threshold, but epidemics can happen if
initial circumstances (randomness) are favorable . If, as
our data suggest, 5%–6% of 2-dose recipients become or
remain susceptible in adolescence, even 100% 2-dose coverage
would barely meet the level of immunity required for elimina-
tion. The large proportion of cases in adolescents and the
small proportion in preschoolers and elementary school chil-
dren suggest a problem of waning immunity. Although the
German seroprevalence study suggests increasing vulnerability
with time since the second dose , US data suggest that the
proportion of children potentially susceptible 10 years after
the second MMR vaccine, is not significantly different from
that before the second dose . On balance, secondary
vaccine failure seems the most plausible explanation for ado-
lescent susceptibility, but this warrants further evaluation. Our
findings underscore the importance of incorporating residual
susceptibility among twice-vaccinated individuals into the as-
sessment of elimination targets and feasibility. In outbreaks,
aggressive public health control measures can be effective in
limiting spread. Ultimately, however, if immunization pro-
grams cannot induce the required level of population immuni-
ty to stay below the epidemic threshold, then elimination will
not be achievable.
In conclusion, the 2011 Quebec epidemic provides an op-
portunity for jurisdictions with a stated goal of measles elimi-
nation to reflect both on the very tight levels of population
immunity required and on how tenuous these levels may be.
Our data suggest that even with 3%–5% of children unvacci-
nated, the residual vulnerability in 2-dose recipients can lead
to a precarious balance with respect to the epidemic threshold,
exposed only through chance superspreading events. Although
such events may be infrequent, they signal the challenge that
may lie ahead for elimination. In this regard, unvaccinated in-
dividuals remain the top priority for immunization programs,
but elimination goals require that we also anticipate, assess,
and address residual vulnerability in 2-dose vaccine recipients.
et des Services Sociaux du Québec.
Potential conflicts of interest.
have received research funds from GlaxoSmithKline and Sanofi Pasteur.
Brian J. Ward is medical officer for Medicago, has received research funds
from GlaxoSmithKline, Sanofi Pasteur, and Pfizer, and has served as a
member of ad hoc advisory boards for GlaxoSmithKline, Sanofi Pasteur,
Pfizer, Merck, and Novartis. All other authors report no potential
All authors have submitted the ICMJE Form for Disclosure of Potential
Conflicts of Interest. Conflicts that the editors consider relevant to the
content of the manuscript have been disclosed.
This work was funded by the Ministère de la Santé
Gaston De Serres and Nicole Boulianne
1. National Advisory Committee on Immunization. Canadian immuni-
zation guide. 5th ed. Ottawa: Health Canada, 1998:118–24.
2. King A, Varughese P, De Serres G, Tipples GA, Waters J. Measles
elimination in Canada. J Infect Dis 2004; 189(Suppl 1):S236–42.
3. Notifiable disease incidence by year 1989–2004. http://dsol-smed.phac-
aspc.gc.ca/dsol-smed/ndis/c_time-eng.php. Accessed 3 January 2013.
4. Vaccine preventable diseases. Measles 2005–2011. http://www.phac-
aspc.gc.ca/im/vpd-mev/measles-rougeole-eng.php. Accessed 3 January
5. Dallaire F, De Serres G, Tremblay F-W, Markowski F, Tipples G. Long
lasting measles outbreak affecting several unrelated networks of un-
vaccinated persons. J Infect Dis 2009; 200:1602–5.
6. European Centre for Disease Prevention and Control. European
measles monthly monitoring. Issue 8. 2012. http://ecdc.europa.eu/en/
monitoring-February-2012.pdf. Accessed 4 January 2013.
7. McLean ME, Walsh PJ, Carter AO, Lavigne PM. Measles in Canada—
1989. Can Dis Wkly Rep 1990; 16:213–8.
8. Boulianne N, Audet D, Ouakki M. Enquête sur la couverture vaccinale
des enfants de 1 an et 2 ans au Québec en 2006. Quebec City, Canada:
Institut National de Santé Publique du Québec, 2007. http://www.
3 January 2013
9. Boulianne N, Bradet R, Audet D, Deceuninck G. Enquête sur la cou-
verture vaccinale des enfants de 1 an et 2 ans au Québec en 2008.
Quebec City, Canada: Institut National de Santé Publique du Québec,
Enfants1et2Ans2008.pdf. Accessed 3 January 2013.
10. Boulianne N, Bradet R, Audet D, et al. Enquête sur la couverture vac-
cinale des enfants de 1 an et 2 ans au Québec en 2010. Quebec City,
Canada: Institut National de Santé Publique du Québec, 2011 http://
AnsQc2010.pdf. Accessed 3 January 2013.
11. Public Health Agency of Canada (PHAC). Case definitions for diseas-
es under national surveillance, Can Com Dis Rep 2009; 35(S2): 71–2.
12. De Serres G, Gay NJ, Paddy C, Farrington CP. Epidemiology of trans-
missible diseases after elimination. Am J Epidemiol 2000; 151:
13. De Serres G, Boulianne N, Defay F, et al. Higher risk of measles when
the first dose of a two-dose schedule is given at 12–14 versus 15
months of age. Clin Infect Dis 2012; 55:394–402.
14. Castillo-Solorzano CC, Matus CR, Flannery B, Marsigli C, Tambini G,
Andrus JK. The Americas: paving the road toward global measles
eradication. J Infect Dis 2011; 204 (Suppl 1):S270–8.
15. Centers for Disease Control and Prevention. Measles—United States
2011. MMWR 2012; 61:253–7.
2011 Measles Epidemic in Quebec • JID 2013:207 (15 March) • 997
by guest on December 28, 2015
16. Increased transmission and outbreaks of measles—European region,
2011. MMWR 2011; 60:1605–10.
17. Vivancos R, Keenan A, Farmer S, et al. An ongoing large outbreak of
measles in Merseyside, England, January to June 2012. Eurosurveil-
lance 2012; 17 (29). doi:pii: 20226.
18. Edmonson BM, Addiss DG, McPherson JT, Berg JL, Circo RS,
Davis JP. Mild measles and secondary vaccine failure during a sus-
tained outbreak in a highly vaccinated population. JAMA 1990; 263:
19. Hickman CJ, Hyde TB, Sowers SB, et al. Laboratory characterization
of measles virus infection in previously vaccinated and unvaccinated
individuals. J Infect Dis 2011; 204(Suppl 1):S549–58.
20. Paunio M, Hedman K, Davidkin I, et al. Secondary measles vaccine
failures identified by measurement of IgG avidity: high occurrence
among teenagers vaccinated at a young age. Epidemiol Infect 2000;
21. Uzicanin A, Zimmerman L. Field effectiveness of live attenuated
measles-containing vaccines: a review of published literature. J Infect
Dis 2011; 204(Suppl 1):S133–48.
22. De Serres G, Defay F, Boulianne N, et al. Greater risk of measles with
earlier age of 1st dose in adolescents who received two pediatric doses
of vaccine beginning from 12 months of age: further evidence from
Quebec. Presented at: Interscience Conference on Antimicrobial
Agent and Chemotherapy 2012; 9–12 September 2012; San Francisco,
Calif. Poster G876-b.
23. Poethko-Müller C, Mankertz A. Sero-epidemiology of measles-specific
IgG antibodies and predictive factors for low or missing titres in a
German population-based cross-sectional study in children and ado-
lescents (KiGGS). Vaccine 2009; 29:7949–59.
24. LeBaron CW, Beeler J, Sullivan BJ, et al. Persistence of measles anti-
bodies after 2 doses of measles vaccine in a postelimination environ-
ment. Arch Pediatr Adolesc Med 2007; 161:294–301.
25. Davidkin I, Jokinen S, Broman M, Leinikki P, Peltola H. Persistence of
measles, mumps, and rubella antibodies in an MMR-vaccinated
cohort: a 20-year follow-up. J Infect Dis 2008; 197:950–6.
26. Leuridan E, Van Damme P. Passive transmission and persistence of
naturally acquired or vaccine-induced maternal antibodies against
measles in newborns. Vaccine 2007; 25:6296–304.
27. Public Health Agency of Canada. Measuring up: results from the Na-
tional Immunization Coverage Survey, 2002. http://www.phac-aspc.gc.
ca/publicat/ccdr-rmtc/04vol30/dr3005a-eng.php. Accessed 25 May
28. Laroche J, Frescura A, Belzak L. Results from the 2006 and 2009
Childhood National Immunization Coverage Surveys. Can J Infect Dis
Med Microbiol 2010; 21:205.
29. Centers for Disease Control and Prevention. National and state vacci-
nation coverage among children aged 19–35 months—United States,
2010. MMWR 2011; 60.
30. Centers for Disease Control and Prevention. Vaccination coverage
among children in kindergarten—United States, 2009–10 school year.
MMWR 2011; 60:700–4.
31. Centers for Disease Control and Prevention. National and state vacci-
nation coverage among adolescents aged 13 through 17 years—United
States 2010. MMWR 2011; 60:1117–23.
32. Ramsay M. Measles: a strategic framework for the elimination of
measles in European Region. The Expanded Programme on Immuni-
zation in the European Region of WHO. Copenhagen, Denmark:
World Health Organization, 1999. http://www.euro.who.int/__data/
assets/pdf_file/0003/119802/E68405.pdf. Accessed 3 January 2013.
33. Gay NJ. Modeling measles, mumps, and rubella: implications for the
design of vaccination programs. Infect Control Hosp Epidemiol 1998;
34. Garske T, Rhodes CJ. The effect of superspreading on epidemic out-
break size distributions. J Theoretical Biol 2008; 253:228–37.
998 • JID 2013:207 (15 March) • De Serres et al
by guest on December 28, 2015