Endemic measles was declared eliminated from the United States in 2000, but imported cases associated with international travel remain an important source of outbreaks.¹ Measles virus is highly contagious, and transmission during air travel has been well documented,^2,3 despite high rates of air exchange in modern commercial aircraft.⁴ Persons with measles are infectious for up to 4 days before rash onset, so informing contacts of their exposure increases awareness and can prevent ongoing community transmission by encouraging asymptomatic exposed persons to immediately isolate and seek care if they develop prodromal signs. Post-exposure prophylaxis (PEP) is available in the form of vaccination or immunoglobulin administration and is effective if given within 3 or 6 days (respectively) of exposure. Measles cases have increased worldwide in recent years, including in countries like the United States where measles has been eliminated.^5,6 This is expected to result in a corresponding increase in the number of exposure events on aircraft during domestic and international air travel that require public health intervention.

U.S. Centers for Disease Control and Prevention (CDC) coordinates aircraft contact investigations after measles exposure events on aircraft. Eligible flights include any domestic or inbound international flight carrying an infectious person, regardless of the length of the flight. Modern aircraft have high rates of air turnover, and airflow is directed from the center aisle(s) to the sides of the plane, thus limiting shared air between passengers seated more than a few rows apart.⁷ For this reason, a common starting point for infectious disease contact investigations is to consider travellers seated within two rows in front of or behind the index case to be exposed.⁴ However, because measle is so transmissible, the size of the contact zone CDC uses for measles has been expanded several times since the United States first began conducting measles aircraft contact investigations in 2005. In the current protocol, the contact definition for large aircraft includes: (i) all passengers seated in the same row as the index case and those seated within two rows in front and behind, (ii) all infants in arms on the plane, and (iii) all cabin crew that served the index case. For smaller aircraft (passenger capacity ≤ 50) or airlines with unassigned seating, all passengers and flight crew are considered contacts. The evaluation of individual contacts is conducted by state, local, and territorial health departments, and decisions about whether to offer PEP are determined by the individual jurisdiction’s protocol.

Other high-income countries where measles is uncommon also take public health action when persons with measles travel while infectious,^8–10 but the details of the protocols vary by country, and many grant health departments discretion to adjust their approach based on individual circumstances and resource availability. For example, the European Union’s RADIGA guidance states that all passengers and crew should be considered at risk but presents a prioritization scheme based on age, vaccinations status, and seating location if resources are limited.⁹ The United Kingdom Health Security Agency advocates a ‘warn and inform’ approach that focuses on mass communication to all persons potentially affected in lieu of more targeted contact tracing.⁸

The logistics of aircraft contact investigations are complicated, and challenges such as difficulty obtaining timely, accurate contact information from airlines, inability to verify contact vaccination history, and reluctance of contacts to accept PEP may compromise the public health benefit of this activity. There is also a dearth of evidence-based recommendations about how large the area of the aircraft used to define close contact (the contact zone) should be and which contacts should be prioritized for intervention.

Two recent systematic reviews of measles transmission on aircraft covered literature published from 1946–March 2012 (Edelson³ ) and 2001–2013 (Jost et al.² ). Jost et al. identified 132 index cases that led to 47 secondary cases and found that transmission was more likely if the index case was young or if there were multiple index cases on the same flight. Edelson’s review focused on the role of a contact’s location in the plane relative to the index case, with the goal of informing the selection of an appropriate contact zone. Among the studies included, the median distance between the index and secondary cases was six rows, with the farthest seated 17 rows away from the index case. Shortly after this study was published, the maximum passenger capacity for aircraft requiring a whole-plane contact investigation under CDC protocols was increased from 30 to 50.

This systematic review seeks to summarize and update available evidence for measles transmission risk during commercial air travel to inform best practices for conducting measles aircraft contact investigations.

A CDC librarian designed customized search strategies for eight databases (Medline, Embase, Global Health, Cochrane Library, CINAHL, Scopus, Epistemonikos, and Global Index Medicus), focusing on keywords related to measles transmission and air travel (Box 1, Supplemental Table 1). The search was limited to published articles and pre-prints from 2004–2023. No restrictions were placed on publication language. The resulting article lists were imported into a reference management software (EndNote, Clarivate, Philadelphia, PA) for deduplication. The reference sections of review articles identified in the initial search were reviewed for any additional articles that met the search criteria.

The deduplicated article list was imported into a cloud-based systematic review software (Covidence, Melbourne, Australia), where each article was reviewed and evaluated by two reviewers to determine whether inclusion criteria were met. To be included, articles needed to contain primary data on measles exposure events on commercial flights for which contact investigation was performed. Review articles, editorials without primary data, and modelling studies were not included. Initial screening comprised title and abstract review and, if the article met inclusion criteria or if eligibility could not be determined, full text review was performed. Discordance between the two reviewers about eligibility were resolved through discussion with the larger group until a final determination was reached. Articles that were deemed eligible after full text review were advanced to the data extraction phase.

Two reviewers used a data extraction template to collect information about each article, including flight characteristics, approach to contact investigation, and contact data (vaccination history, disease status, PEP). Some articles were primarily focused on community spread of measles after transmission on a plane; for these articles, we included only data clearly identified as coming from aircraft contacts. We also explored whether articles reported cases as co-travellers of the index case (e.g. household members, same tour group); if they did not, we assumed secondary cases had no association with the index case. If the article described an index case whose itinerary included several flights, but contact investigation was not performed for all, the first country in the index case’s itinerary was still listed as the origin. There were several instances where overlap between articles was suspected. If there was enough detail (e.g. flight itineraries, demographic information) to be certain that two articles were describing the same index and secondary cases, these cases were only counted once. If there was uncertainty, the cases were treated as separate.

Because most of the articles included in this review did not have a comparator group, the risk of bias was assessed using the Mixed Methods Appraisal Tool (MMAT).¹¹ If a study included a comparator (non-exposed group), it was considered quantitative non-randomized, and if it did not, it was considered quantitative descriptive. Studies were considered to include a comparator group if they defined a contact zone and included any outcome data for persons outside that contact zone (even if that outcome data was acquired passively). If a study did not define a contact zone, it was assumed that there was no comparator group. The MMAT questions were used as written, except for question 3.4 (‘Are the confounders accounted for in the design and analysis?’), where the answer choices were modified to include ‘partial yes’ if the study addressed some, but not all, pre-specified confounders. Specific confounders considered included the contact’s age, vaccination status, and travel history. Discrepancies between reviewers during the data extraction and risk of bias assessment phases were resolved through group discussion. For question 4.5 (‘Is the statistical analysis appropriate to answer the research question?’), if a study was not designed in a way that supported conducting more complex statistical methods, but the manuscript reported basic descriptive statistics (including the number of secondary cases among air travellers and the number of persons exposed), reviewers were instructed to respond ‘Yes.’

While endemic measles transmission has been eliminated from the United States, disease translocation and onward community transmission by persons travelling from endemic areas remain a threat to public health.¹ Transmission events that occur on aircraft have the potential for outsized impact because contacts are likely to disperse, potentially spreading the virus to communities across the country. This literature review identified relatively few published reports of secondary transmission on aircraft (70 secondary cases associated with 182 aircraft exposure events in the published literature over a 20-year period), but measles exposure on aircraft still represents a risk, especially for unvaccinated persons. Reports of secondary cases in airports^19–21,23 and at final destinations^{12–14,18,21,23–26} further highlight the impact that travel-associated cases can have on an under-vaccinated population.

Many articles included in this review did not explicitly define a contact zone. Among those that used a contact definition that was limited to persons seated within two rows in front of or behind the index case, 71% (5/7) reported secondary cases outside of the contact zone, suggesting that this approach may not be adequate to identify passengers at risk. However, no articles provided thorough, systematic follow-up for passengers seated throughout the plane, so it is difficult to determine whether a whole-plane contact investigation is warranted for every measles exposure or if some intermediate contact definition (>2 rows, but less than the entire plane) might be appropriate. Consideration should also be given to whether a different approach is needed if multiple cases are identified on the same flight.

At least 57% of the secondary cases documented in this review occurred among persons who were unvaccinated. International travellers are considered at high-risk for measles exposure and transmission, and they should be encouraged to consider vaccination before travel if they are not already fully vaccinated. U.S. CDC recommends that infants aged 6–11 months should receive one dose of MMR and persons aged ≥12 months should receive two doses of MMR separated by at least 28 days prior to international travel.²⁷ The availability of PEP options for susceptible contacts has the potential to amplify the benefits of contact tracing for measles, compared to many other infectious diseases. However, very few articles reported detailed data on the number of contacts offered PEP or the proportion that accepted. Among the four articles that reported PEP uptake,^14,16,17,19 the number of contacts that accepted PEP was very low, and public health officials encounter many implementation barriers, most notably difficulty identifying and reaching contacts within the required timeframe. While PEP may be a valuable tool in certain circumstances and improving the timeliness of the notification process may allow it to be offered more widely, there are logistical considerations that prevent it from being the primary strategy to address measles exposure on aircraft. For this reason, routine and pre-travel vaccination remain the most effective method to prevent measles transmission on aircraft.

Most articles about measles transmission during air travel report data collected during contact investigations, which carry inherent limitations. Most contact investigations only include persons considered at the highest risk of exposure, without a low-risk comparison group, making it impossible to calculate odds ratios or make inferences about the appropriateness of the contact zone the authors used. Aircraft contact investigations often have low response rates for a variety of reasons, including difficulty obtaining accurate contact information for passengers and availability of resources to conduct follow-up. Among studies that considered the whole plane exposed, follow-up often consisted of a notification sent to all passengers, with a request to respond only if they developed symptoms, leaving these studies vulnerable to non-response bias. Many authors took a narrative approach to describing the investigation and therefore did not include the methodological detail necessary to fully assess their methods, as reflected in this article by the high proportion of domains marked ‘Can’t tell’ in the MMAT. This lack of a standardized method of reporting made the data difficult to aggregate to calculate response rates or odds of infection among exposed contacts. In particular, the lack of consistent reporting of denominator data and the reliance in many cases on passive follow up makes it difficult to quantify risk. Because measles has a variable incubation period, it can be difficult to rule out the possibility that cases were exposed in the community or the household setting before the flight. Since most papers did not explicitly control for other exposure risk factors (high rates of community transmission or ongoing contact between index cases and secondary cases outside the flight), it is possible that some infections that were classified as secondary cases associated with air travel were actually contracted in other, non-aircraft settings. Information regarding the implementation of PEP was especially limited, making it impossible to quantify the relative importance of different barriers to PEP implementation.

Our study has several limitations. First, due to the date range restriction, we may have missed papers published prior to 2004. Second, we may have missed papers not listed in one of the eight databases searched. Third, because were not able to identify meaningful outcome variables that were consistently available across most manuscripts, we were not able to conduct a metanalysis. This also meant that we had to adapt the MMAT methodology, which is intended to be applied to analysis of specific outcome variables, to apply to the study methods more generally.

Based on the findings of this systematic review, CDC’s currently recommended approach to measles aircraft contact investigations is likely to capture a majority of secondary cases but may miss exposed passengers seated outside of the defined contact zone on larger flights. Many questions remain about the best way to manage measles exposure risk on aircraft, including questions about the best approach to contact investigation (size of contact zone, target vs broad outreach, etc.) and the most effective ways to prevent secondary cases. These questions are difficult to answer using the existing literature and may be better addressed through a prospective study specifically designed to collect the types of data necessary to directly answer specific epidemiological and operational questions.

Supplementary data

Supplementary data are available at JTM online.