An Outbreak of Wild Poliovirus in the Republic of Congo, 2010–2011

METHODS

RESULTS

DISCUSSION

This explosive poliomyelitis outbreak in ROC, centered in PN and originating from Angola, was unusual in its scope. In addition, the incidence of clinical polio was higher among young adults than among children, the CFR was much higher than reported in other WPV1 outbreaks, and the epidemiology of the outbreak varied based on residence. Also, in this outbreak, most cases were clinically confirmed rather than laboratory-confirmed, which is the global standard. Finally, in response to the right-shifted age distribution, the response effort had added challenges beyond those seen in outbreaks among younger age cohorts. A national campaign targeting the entire population had to be planned and executed quickly; the campaign’s effectiveness is apparent in the deceleration in cases soon after the SIAs started. The quick deceleration can be attributed to the fact that the entire population was targeted for vaccination, with added contribution from environmental secondary immunization through shed Sabin OPV strains.

Children usually comprise a majority of cases in poliomyelitis outbreaks; outbreaks among adults are unusual but have been described. In the 1996 WPV1 outbreak in Albania, 96 of 138 (70%) clinical and laboratory-confirmed cases were among persons aged ≥ 15 years [15]. The unusual age distribution in the Albanian outbreak was attributed to multiple factors. First, the adults were unlikely to be exposed to WPV1 and thus had little natural immunity; second, adults were recommended to receive only 2 doses of vaccine in infancy according to the country’s strategy. Finally, mismanagement of cold chain during the years that these adults were eligible for vaccination could have resulted in vaccination with nonimmunogenic vaccine. In a 2000 poliomyelitis outbreak in Namibia, >90% of 45 cases were among persons aged ≥15 years (N. Yusuf, personal written communication, February 2012); persons with poliomyelitis were suspected of being either unvaccinated or incompletely vaccinated due to civil conflict. Furthermore, these age cohorts had not been eligible for catch-up SIAs and were not exposed to circulating WPV.

In ROC, poor immunity is the likely reason that this outbreak affected predominantly 15–29-year-olds (birth years: 1981–1995). This cohort likely had little natural immunity due to low exposure to WPV1, as well as suboptimal vaccine-derived immunity due to disrupted routine services in the 1990s and ineligibility for the catch-up SIAs. The SIAs appear to have protected cohorts of children born from 1996 to 2006, accounting for the low incidence in 5–14-year-olds. The findings of a 2010 vaccine coverage survey conducted in one PN neighborhood are consistent with poor immunity in adults; 1-dose tOPV coverage was 87% among children aged <5 years, 74% among 5–14-year-olds, 50% among 15–29-year-olds, and 34% among adults aged ≥30 years [9]. Although there are limitations with these data in that they only cover one neighborhood in PN and much of the vaccination data are based on recall, they are consistent with an immunity gap in adults in PN.

Another unusual feature is that the mortality associated with this outbreak is much higher than the 5%–10% CFR typically seen in WPV1 outbreaks. Most polio outbreaks primarily affect children, and children are less likely to die as a result of WPV1 infection [16, 17]. However, even among children aged <15 years in ROC, the CFR was 32%. Furthermore, the CFR among those aged ≥15 years was 48%, and while a higher CFR is expected in adults, CFRs this high have not been observed in other recent outbreaks in low-income countries. In the 2010 WPV1 outbreak in Tajikistan, the CFR was 6% overall, with those aged ≥15 years having a CFR of 6% [3]. In a 1996 WPV1 outbreak in Albania, the CFR was 12% overall and 14% among adults [15]. An epidemiologic study of risk factors for death is presented separately [18].

The epidemiology differed between cases in PN and those from the rest of ROC. The outbreak spread broadly throughout the PN region, affecting older age groups. In the rest of ROC, fewer cases were reported, with most cases among children. One contributing factor could have been the higher population density in PN as well as the slum-like conditions in parts of the city, both of which can help to propagate the virus. However, this does not fully explain the burden in PN as Brazzaville is approximately 20 times as dense with similar slum-like conditions and only reported 13 cases of polio. Furthermore, children with polio in PN were more likely to die than those in the rest of ROC. Why the outbreak affected cases more severely in PN is unknown, although a variety of reasons were investigated. There were no significant genetic differences among the virus strains, including the Angolan parental virus, the viruses from different locations within ROC, and the related cases in neighboring countries; thus, viral differences were not driving the differential epidemiology. There was no evidence of a detection bias in PN with persons having less severe polio failing to seek care [19]. No yearly provincial immunization data were available to see if there was a larger immunity gap in PN than the rest of ROC.

A limitation of this investigation was that only 39% of survivors had follow-up; thus, the NPEC had to rely on presenting clinical data when classifying those without positive specimens. Additionally, a high number of cases had to be confirmed clinically due to the lack of specimens; adequate stool samples were collected from only 12% of persons with polio compared with the global benchmark of 80% [10]. Clinical confirmation is unreliable in areas where poliomyelitis is now rare; therefore, laboratory confirmation is the gold standard. Laboratory surveillance was challenging in PN, particularly at the peak of the outbreak when the 2 main hospitals in PN were caring for >200 paralyzed individuals, straining clinical and public health staff. However, because this was a confirmed WPV outbreak, the specificity of the definition increases. Because of the increased specificity and the similarities between the laboratory-confirmed and clinical cases, we are confident that the vast majority of clinical cases represent true WPV paralysis. A limitation of accepting the clinical case definition is that some cases might have been misclassified.

Another limitation of this investigation was surveillance. While country-wide active surveillance was conducted, surveillance was facility-based. However, once the outbreak was confirmed and vaccination campaigns commenced, media coverage stressed the need to visit health facilities, and medical care for polio was free, helping to decrease the impact of this surveillance bias. Additionally, surveillance only focused on paralytic polio, and it is estimated that for every case of paralytic polio there are 200 more cases that are nonparalytic and thus not reported. Finally, we could not obtain more details about the cases, especially among the deceased; we are unable to account for bulbar paralysis, living conditions, socioeconomic status, and vaccination history.

Several lessons from this outbreak are applicable to other countries. First, ROC had periods of civil conflict with low routine vaccination coverage which created age-specific immunity gaps. Although these gaps were mitigated to some extent through SIAs, older children were not vaccinated, leaving a vulnerable cohort. Health authorities in other polio-free countries, especially those with a history of armed conflict or other ruptures of immunization service delivery, should carefully evaluate polio immunity level by age cohorts and mitigate the risk for an outbreak both among young children and among older cohorts. Second, countries in close proximity to countries with active or recent polio circulation are at high risk for importation. ROC is adjacent to Angola and DRC, both of which had ongoing, reestablished transmission that should have led to the anticipation of an importation. Appropriate measures such as increased surveillance and vaccination of nonimmune populations should have been taken prophylactically. Third, a high index of suspicion for polio is required in the evaluation of all AFP cases, even among adults. Appropriate specimen collection is vital to rapidly identifying the first cases of poliomyelitis in previously polio-free countries. Delays in processing stool specimens from cases reported during September and October resulted in delayed identification of this outbreak until November, causing avoidable delays in response activities which prolonged and magnified this outbreak. Fourth, WPV transmission was stopped <6 months after onset of the first case with multiple coordinated SIAs targeting the entire population. Finally, routine immunization coverage needs to be strengthened dramatically in order to mitigate the risk of future importations.

Outbreaks such as the one in the ROC undermine confidence and drain scarce resources. At this critical juncture in polio eradication, preventing importations by increasing population immunity in vulnerable countries is imperative. Heightened vigilance for WPV importation is critical to ending an outbreak quickly.