We performed a literature search by using PubMed with no date restrictions, applying the following search terms in various combinations to identify episodes of facility-associated strains infecting humans or being isolated from environmental samples in nonendemic and nonoutbreak reporting areas to complement the information known to us: poliovirus or polio, contamination, accidental environmental contamination, accidental release, and laboratory accidents. This search found a total of 29 references, all of which are described in this article.

The laboratory accident in April 2017 involving the exposure of a worker who subsequently excreted WPV2 into the sewage system (12) serves as a stark reminder that the achievements of the GPEI are fragile and can be reversed if remaining sources of polioviruses are not contained to reduce the likelihood and consequences of virus reintroduction after eradication. Even more recently, a suspected contamination of ≈150,000 bivalent OPV vials with type 2 Sabin virus strains has been reported in India (38), highlighting the importance of completeness of containment-related activities for type 2 OPV at all stages of vaccine manufacturing and release. Our review demonstrates that the known risk for poliovirus release from a facility into the community appears to be small, based on only a handful of reported incidents, primarily from vaccine manufacturing sites. However, these reported incidents might represent the proverbial tip of the iceberg, given that the reporting requirements in the past were not very stringent.

The release of polioviruses into the community in the posteradication era is a major public health concern for GPEI as it implements the planned, sequential cessation of OPV use during the polio endgame. With the global discontinuation of type 2 OPV for routine and supplemental immunization in April 2016 and the planned cessation of all OPV use in the next 4–5 years, such release of polioviruses from facilities into the community will become a greater public health concern because population immunity wanes in settings of high population density, poor hygiene, and suboptimal immunity (e.g., tropical developing countries) after OPV withdrawal, increasing the potential for transmission (39).

We should note that in the prevaccine era most of the reported incidents occurred in research settings and the exposed persons did not have vaccine-induced immunity for protection against paralysis or virus transmission. In sharp contrast, the incidents reported in the past 3 decades were related to containment failures at vaccine production plants where the staff are expected to be fully vaccinated. Incidents in the more recent period are likely to be more representative of the public health impact of such containment failures for the current phase of the polio endgame and the near future.

The paucity of reports of laboratory-associated poliomyelitis during the past 3 decades testifies to the effectiveness of vaccines and to improved laboratory facilities and biorisk management. However, laboratory breaches in the past might not have been recognized in the absence of clinical cases, and environmental surveillance was less extensive than it is today. Environmental surveillance for polio, for example, appears to have played a major role in such incidents primarily in the past 2 decades, and more so in recent times, with the expansion and enhancement of scope and methodologies that were introduced as a component of the polio endgame. Furthermore, reporting requirements were less stringent in the past, and we assume that not all facility incidents were recorded. The lack of documented incidents of laboratory accidents during 1955–1991 is difficult to explain but might also have been affected by these factors, in addition to the fact that the formation of GPEI in 1988 led to a more concerted, globally synchronized effort to track and report polio cases and outbreaks.

As seen in the Netherlands incident in 2017, previously vaccinated persons, although probably protected from paralysis, can excrete poliovirus after accidental exposure from containment failures and put the community at risk for virus transmission. This risk underscores the importance of stringent containment measures at the vaccine production sites and preparedness for deployment or enhancement of surveillance activities, such as environmental monitoring, as a public health response strategy.

The smallpox experience illustrates the importance of containment for an eradicated pathogen. Within 1 year of detection of the last known natural case of smallpox in 1977, a case linked to laboratory transmission was reported (40). We have a historic opportunity to benefit from an additional 40 years of experience in risk management and disease-control measures to ensure containment measures are designed, implemented, and maintained to provide a world free from all risks for polio-related paralysis. The first step is a uniform, global awareness of the importance of implementing the GAPIII guidelines to minimize facility-associated risks for poliovirus reintroduction. The risk for accidental release can be minimized by retaining poliovirus in a limited number of poliovirus-essential facilities. GAPIII also proposes further risk reduction by establishing international standards of biorisk management for facility containment (i.e., primary safeguards), population immunity (i.e., secondary safeguards), and facility location (i.e., tertiary safeguards) with assurance by national (National Authority for Containment) and international (GCC) oversight that such standards are met. Timeliness and completeness in implementing these measures through well-defined risk management systems and an effective National Authority for Containment are key for success in sustaining a polio-free world.

Equally important is the identification of materials that are potentially infectious for polioviruses in all facilities that store human stool specimens, respiratory samples, or environmental sewage for any purpose. Depending on the place and time of collection, such materials might harbor infectious polioviruses that have been eradicated (WPV2) or are nearly eradicated (WPV1 and WPV3) in the wild. Identifying the risk, eliminating the risk through destruction, or mitigating the risk of handling such materials is essential (13).

Finally, the risk management approach to containment might have to be optimized and balanced to enable other risk mitigation efforts for the endgame and beyond, such as novel vaccine and drug development to further reduce any risk for vaccine-derived circulation. Timely availability of antiviral drugs and effective novel OPVs that have less risk for reversion to neurovirulence compared with current Sabin vaccines might strengthen outbreak response strategies and mitigate medical risks for inadvertent poliovirus exposure and improve reporting of accidental safety breaches by employees (41–43). The World Health Assembly resolution on poliovirus containment (WHA71.16) urges intense efforts to accelerate progress toward poliovirus containment globally (44).