RISK ASSESSMENT FOR THE TRANSMISSION OF MIDDLE EAST RESPIRATORY SYNDROME CORONAVIRUS (MERS-CoV) ON AIRCRAFT: A SYSTEMATIC REVIEW

(1)

RISK ASSESSMENT FOR THE

TRANSMISSION OF MIDDLE EAST RESPIRATORY SYNDROME

CORONAVIRUS (MERS-CoV) ON

AIRCRAFT: A SYSTEMATIC REVIEW

Talía Berruga Fernández

__________________________________________

Master Degree Project in Infection Biology, 30 credits. Spring 2019

Department: Medical Biochemistry and Microbiology

Supervisors: Emmanuel Robesyn, Teija Korhonen, Josep Jansa

(2)

ABSTRACT

Introduction: MERS-CoV causes a respiratory disease that can be fatal. Although it is most common in countries in the Arabian Peninsula, through travel it has been exported to 17 countries outside Middle East. Most of these exportations have occurred by people travelling by air. The Risk Assessment Guidelines for Infectious Disease transmitted on Aircraft (RAGIDA) produced by the European Centre of Disease Prevention and Control, advises European countries on the measures to take when an infected individual travels by air.

Aim: To gather all available information on the MERS-CoV cases that have travelled by air and produce an update of the RAGIDA project.

Methods: A systematic review was conducted. The online health databases used were PubMed, Embase, Scopus and Global Index Medicus. As additional information sources, Google was searched for grey literature and hand searching was performed on the Early Warning and Response System of the EU, as well as the Disease Outbreak News page of WHO.

Results: A total of 48 records were identified, describing 22 cases of MERS that travelled on board a total of 31 flights. No cases of in-flight transmission were observed in any of the flights. Contact tracing (CT) was performed for 18 of the 22 cases. Most countries defined

“contacts” as the passengers sitting in the same row and the two rows in front and behind the case. Only one country decided to trace all passengers and crew on board the aircraft.

Conclusion: A conservative approach might be more adequate when performing CT of an aircraft where a case of MERS has travelled, because of the lack of in-flight transmission observed and the great amount of resources needed for a CT investigation.

POPULAR SCIENTIFIC SUMMARY

In September of 2012 a 61-year-old-man died in Saudi Arabia of a mysterious lung disease.

Analysis of his sputum lead to the discovery of a previously unknown virus. It turned out this virus was related to a group of viruses known as “coronaviruses”, which in humans are the second most common cause of the common cold, and can also be found in bats, birds, cats, dogs, pigs, mice and other mammals.

Initially, the disease was only observed in countries inside the Arabian Peninsula, and consequently, the virus was named Middle East respiratory syndrome coronavirus, or MERS- CoV, to simplify it. The virus was soon exported to other countries, mainly through air travel.

Currently, the disease has been reported from 27 countries around the world, however, all cases have either visited or been in contact with someone who visited countries of the Arabian Peninsula.

The most common symptoms caused by the virus are fever, cough and myalgia, but some people show no symptoms; others show gastrointestinal symptoms, or can develop severe pneumonia.

In 2007, the European Centre of Disease Prevention and Control (ECDC) initiated a project called Risk Assessment Guidelines for Infectious Disease transmitted on Aircraft (RAGIDA).

What this project aims to achieve, is to produce a series of documents with advice for public health authorities of countries of the European Union (EU) on what to do when a person infected with a transmissible disease (including MERS-CoV) travels by air. The most common action taken in these cases, is to identify people inside the aircraft who are at risk of

(4)

acquiring the disease and contact them to offer diagnostic tests. This intervention is known as

“contact tracing”, and the people at risk are called “contacts”.

In the case of MERS-CoV infection, it is still debated whether all passengers on board the airplane should be considered contacts, or if only passengers sitting in the rows near to the case should be traced.

In this study, we searched for all available published or unpublished articles in selected online healthcare databases (PubMed, Embase, Scopus and Global Index Medicus). We also looked for additional sources of information (news reports, notifications from countries) in Google, the Disease Outbreak News page of the World Health Organization, and in the Early Warning and Response System (EWRS), a communication tool between health authorities within countries of the EU.

From the records found, we aimed to retrieve all information regarding MERS-CoV cases that had travelled by air, such as country of origin and destination, date of travel, age, sex, symptoms during the flight, and outcome of the disease. We also looked for information on what health authorities had done after the MERS case had travelled by air and, when the intervention was contact tracing, how they had defined the contacts (i.e. entire airplane or in selected rows around the case), how many contacts were identified, with how many of them were they able to communicate, how many were tested, and how many developed symptoms and gave a positive result on the tests.

We found 43 records concerning a case of MERS travelling by air, and they described a total of 22 cases. Most countries performed CT shortly after identifying the case, and we found out that in all the 31 flights boarded by the 22 cases, no other people were infected by MERS- CoV.

KEYWORDS

Middle East respiratory syndrome coronavirus, MERS-CoV, coronavirus infection, in-flight transmission, aircraft, travel, RAGIDA.

INTRODUCTION

A novel coronavirus was discovered in 2012 after a patient in Jeddah, Saudi Arabia, died from a severe respiratory disease¹. The virus, now known as Middle East respiratory syndrome coronavirus (MERS-CoV), has currently been detected in to 27 countries and a total of 2,428 cases have been reported to the WHO as of April 2019, along with 838 deaths (35% crude case fatality rate [CFR])². With increasing numbers of commercial flights, this mass transport represents a significant risk for the spread of the disease.

Coronavirus characteristics

The family Coronaviridae comprises spherical enveloped viruses with surface projections resembling a crown (figure 1)³, a feature that gave the name to the original group of coronaviruses, now classified in the subfamily Coronavirinae⁴. They all have a single

Figure 1: MERS-CoV micrograph.

Source: Hui et al, 2015.

(5)

stranded, positive-sense RNA genome of 27 to 32 kb, making them the viruses with the largest RNA genome known^4–6. Their host range includes mammals, birds, and fish. The Coronavirus family is the largest in the order of the Nidovirales, and consists of four genera:

alpha-, beta-, gamma- and delta-coronaviruses⁷. MERS-CoV belongs to the genus Betacoronavirus, which includes only virus with mammalian hosts, and is divided into four lineages: A, B, C, and D. MERS-CoV belongs to the C lineage, and is the first virus of this lineage known to infect humans⁸. The only distinct characteristic that sets the Betacoronavirus genus apart from other coronaviruses is their non-structural protein 1 (nsp1), which differs in size and sequence⁴.

Virus proteins

The virion of coronaviruses contains four major structural proteins: the spike (S), membrane (M) and envelope (E) proteins located in the virus envelope, and the nucleocapsid (N) protein that encloses the genome (figure 2)⁵. Some coronaviruses (but not MERS-CoV) also have a haemagglutinin protein.

The S glycoprotein is the largest transmembrane protein of the viral envelope and has a club and stalk morphology. Each monomer consists of an S1 (club) and an S2 (stalk) subunits that form trimers. The S1 subunit corresponds to the amino-terminal domain of the protein and is extremely variable⁹, as it is responsible for virus entry and, thus, determines the host range of the virus⁵. In MERS-CoV, it has been shown to bind to dipeptidyl peptidase 4 (DPP4, also known as CD26), a transmembrane glycoprotein which cleaves peptides such as hormones and chemokines and regulates their activity. This protein is abundant in epithelial and endothelial tissues, and in humans is present in the lungs, kidneys, small intestine, liver, prostate and activated leukocytes¹⁰.

The S2 subunit corresponds to the carboxyl-terminal domain of the protein and has a transmembrane portion. This subunit facilitates fusion of the viral and cellular membranes and aids in entering the host cells⁵. The S protein is the main inducer of neutralizing antibodies⁴.

The M protein is the most abundant protein in the coronavirus envelope. It is highly conserved and gives the virion envelope its shape. It has a small external domain and a large internal domain that binds to the nucleocapsid protein, acting as a matrix protein that mediates the association between the helical nucleocapsid and the viral envelope⁶.

The envelope protein is a small protein present in limited numbers in the virus envelope. It aids in virion formation, budding, and infectivity, although knockout of the E gene doesn’t result in the virus not being viable^5,9.

The N protein is the only constituent of the nucleocapsid, binding along the RNA in a configuration analogous to beads on a string, and forming a helical structure, an uncommon feature for positive-strand RNA viruses, which usually have icosahedral capsids⁹. Unlike N proteins of other viruses, the coronavirus N protein doesn’t provide protection against ribonucleases⁹.

Figure 2: Coronavirus virion.

Source: Acheson, 2011 S

M N

E

(6)

The virion particles are sensitive to heat, lipid solvents, non-ionic detergents, formaldehyde, oxidizing agents and UV radiation⁴.

Genome organisation

The single stranded RNA genome of coronaviruses is capped at its 5’ end and polyadenylated at its 3’ end, enabling it to be translated upon its release into the cytoplasm⁶. It can be infectious by itself⁴. The non-structural proteins involved in replication, transcription and protein processing are encoded in the 5’ proximal segment of the virus genome, within the open reading frame 1 (ORF1, a replicase), which comprises two ORFs that partially overlap (ORF1a and ORF1b). The complete expression of this protein complex is achieved by a mechanism known as ribosomal frameshifting. The major structural proteins come after the nsps, always in the order S, E, M and N⁶, (i.e., 5`-ORF1a/b-S-E-M-N-poly(A)- 3`).

The coronavirus genome contains seven to eight translation initiation sites that correspond to ORFs. In order to translate all their proteins, they generate 7-8 individual mRNAs, the largest of them corresponding to the whole genomic RNA. Each mRNA has the same 3’ end, meaning each mRNA contains all the gene sequences of the next smaller mRNA, plus one additional gene at its 5’1 end, which will be the only one translated. This is known as a nested set of mRNA, where they are structurally polycistrionic but functionally monocistrionic⁶. The viral replicase possesses the RNA-dependent RNA polymerase (RdRp), RNA helicase and protease activities, as well as processing activities that include sequence-specific endo- and exoribonucleases, methyltransferase, phosphatase and phosphodiesterase, required for translation and replication⁶.

The clinical disease

The clinical presentation of MERS extends from a mild respiratory disease to a fatal lower respiratory tract infection. The incubation period ranges between 2 to a maximum of 14 days (average 5 days), and the most commonly reported symptoms and signs are fever, cough (usually non-productive) and dyspnoea (shortness of breath). Approximately one third will also have pneumonia with alterations in chest radiography, myalgia, diarrhoea, vomiting and abdominal pain^11–13. Atypical presentations described include people presenting only gastrointestinal symptoms, or no symptoms at all¹⁴. Renal or pulmonary failure have been commonly described in fatal cases⁸. The disease outcome is influenced by age and comorbidities, with CFR increasing as age increases⁸. The median time for symptom onset to death is 11.5 days (range: 4-298 days) and the median length of hospital stay is 41 days (range: 8-96 days)¹³.

Viral excretion from the respiratory tract has been documented through the first month of illness, with a higher viral load detected in lower respiratory tract samples. The virus is also excreted in urine and stool⁸. Several studies have shown that patients are not infectious during the incubation period, and only become infectious once symptoms appear^14,15. Cowling et al.¹⁵ suggest that infectiousness might begin 0.4 days before symptom onset. Most asymptomatic infections have been identified through contact tracing (CT) investigations⁸. The diagnosis of the disease should be performed by using an initial screening test, followed by a confirmatory test. The current standard screening test used is reverse transcription PCR, which can be real time or conventional, targeting either the upstream of the gene E (UpE gene) or the gene N (UpN gene). For confirmation, ORF1b or ORF1a are used^8,16. Lower respiratory tract specimens are preferred over upper respiratory tract ones, since they contain higher viral loads¹⁶. There are also sequencing tests available, commonly used for strain

(7)

differentiation, that target regions of the RdRp or the N genes⁸. Serologic samples can be screened using enzyme linked immunosorbent assay (ELISA) and confirmed by immunofluorescence assay (IFA), however, these are usually preferred for screening of asymptomatic carriers, in whom the virus cannot be detected by PCR in respiratory samples¹⁶. The only current available treatment is symptomatic care. The efficacy of antivirals such as ribavirin and interferon, remains controversial: this combination was given to five patients in a study, but all died. In other studies, it appeared to increase the short-term survival (14 days), however, one study only included one subject and in the other only 10/44 patients were still alive by day 28⁸.

Epidemiology of MERS-CoV

Coronaviruses are widely distributed among mammals and birds⁹, and they are the second most common cause of the common cold⁶. This group of viruses was previously thought to only cause mild respiratory diseases in humans until 2003, when the multi-state outbreak of a respiratory disease led to the identification of the Severe Acute Respiratory Syndrome (SARS) coronavirus¹⁷, a betacoronavirus closely related to MERS-CoV. It is now known that other coronaviruses can sporadically cause severe lower respiratory infections in infants and the elderly^4,6.

The origin and the transmission of MERS is poorly understood; however, coronaviruses have been detected in bats worldwide and MERS-CoV has been detected in camel respiratory secretions and milk throughout Middle East and Africa. Identical genomic sequences of MERS-CoV isolated from humans and their camel contacts have been documented¹⁸, with this animal being the only documented zoonotic source of infection to humans⁸. Human-to- human transmission appears to be limited, but when it does occur, it is through respiratory secretions and close contact (living with or caring for an infected individual). Most secondary cases occur in family members of the diseased person or in the healthcare setting^8,12.

Although MERS-CoV was initially identified in Saudi Arabia, the first cases occurred in Jordan, and were confirmed retrospectively¹⁹. Around 84% of all MERS cases have been reported from Saudi Arabia²⁰, and all reported cases have been linked to countries in or near the Arabian Peninsula. Seventeen countries outside the Arabian Peninsula have reported travel-associated cases²¹ (table 1). One of these importations culminated in a multi-hospital outbreak in South Korea in 2015, with a total of 186 confirmed cases^22,23.

Table 1: countries with lab-confirmed MERS cases Countries in or near the Arabian Peninsula

Bahrain, Iran, Jordan, Kuwait, Lebanon, Oman, Qatar, Saudi Arabia, United Arab Emirates (UAE), and Yemen.

Countries outside the Arabian Peninsula with travel-associated cases

Algeria, Austria, China, Egypt, France, Germany, Greece, Italy, Malaysia, Netherlands, Philippines, South Korea, Thailand, Tunisia, United Kingdom (UK), United States of America (USA).

Most MERS cases have occurred in adults, with an average age of 50 years and male predominance¹².

Most studies about the basic reproductive number (R0) of MERS-CoV, that is, the number of individuals that can become infected from one index case (the index case is the person with a disease who infects other individuals), have estimated it to be below 1.0^8,23,24. A R0 above 1 is

(8)

required for an outbreak to be self-sustaining²⁵, but the precedent of the SARS-CoV outbreak from 2003 still raises concerns over MERS-CoV pandemic potential. In addition, studies of localised MERS-CoV outbreaks (South Korea, Saudi Arabia) have reported R0 of up to 8.0^26,27, likely related to superspreaders^8,28, individuals who spread the disease to a larger amount of people than what is usually expected for the disease.

Transmission of diseases in airplanes

As the number of passengers travelling internationally by air increases, so does the risk of introduction and spread of infectious diseases between countries. In 2018, the Airports Council International reported 8.3 billion passengers travelling from 175 countries worldwide, with 41.4% accounted for by international travellers²⁹.

In-flight transmission has been documented for several diseases. The systematic reviews conducted in 2009 by experts from the European Centre for Disease Prevention and Control (ECDC) identified 18 cases of on-board transmission from tuberculosis infected persons that had travelled by air; 81 infections resulted from influenza infected travellers; 26 passengers were infected during flight by persons with SARS; 1 person resulted infected with meningococcal disease; and 6 people had been infected with measles^30–32.

Secondary cases of tuberculosis occurred in passengers sitting within two rows of the index case, secondary influenza cases in passengers from two rows away from the index case up to ten, secondary SARS cases were sitting in the same row up to seven rows away, and secondary cases of measles occurred in rows 2 and 8 from the index case.

It is important to note that one of the in-flight transmission events of influenza³³ occurred during ground delay, when 54 passengers on board an airliner stayed inside the aircraft for three hours while the engines were off due to a failure during a take-off attempt. This incident resulted in 39 (72%) of passengers infected, and highlights the importance of air circulation on aircraft. Modern airplanes control the pressurisation, oxygen level, humidity and filtration of air in the passenger cabin. During flight, fresh air is usually supplied to the cabin from the outside, and 50% of that air is recirculated inside the cabin, after passing through high efficiency particulate arresting (HEPA) filters, which can remove 99.97% of particles larger than 0.3μm in diameter (which includes the majority of pathogens) from the cabin air. Viruses smaller than 0.3μm that tend to adhere to particles or form clumps will also be eliminated (MERS-CoV is 0.11-0.14μm). Cabin air is usually exchanged 10-20 times per hour. Usually, airplanes built before 1980 and for fewer than 100 passengers do not have HEPA filters.³⁰ The most important public health intervention carried out after a case of any infectious disease travels on board aircraft is contact tracing (CT). CT is the process of identifying people who may have come into contact with an infected individual, referred to as “contacts”, to alert them about the possibility of infection, offer testing for diagnosis and provide prophylactic care, when available. The goal of CT is to interrupt the transmission of the disease and reduce the spread of the infection.

The RAGIDA project

In 2007, the RAGIDA project (Risk Assessment Guidance for Infectious Diseases transmitted on Aircraft) was initiated by ECDC, to assist national public health authorities in the EU on the evaluation of the risks associated with the transmission of infectious agents on board aircraft and to advise on measures for containment³⁰. This project consists of two parts: the first is a set of guidelines published in 2009³⁰, based on a systematic review, expert opinions and established disease-specific parameters, which provide a basis for countries to assess in-

(9)

flight transmission events. The second part³⁴ includes a set of disease-specific guidance documents. Originally, the project was agreed to cover 12 diseases: tuberculosis, influenza, SARS, meningococcal disease, measles, rubella, diphtheria, Ebola fever, Marburg fever, Lassa fever, smallpox and anthrax. However, due to the outbreak in the Middle East and its potential introduction to Europe, it was decided that a guidance document for Middle East Respiratory Syndrome coronavirus (MERS-CoV) would also be produced.

OBJECTIVES

This review aims to gather the available evidence needed to guide health interventions, such as contact tracing, in the event of a case of MERS travelling on aircraft, and to provide a thorough description of cases on aircraft, interventions undertaken, and, if any, in-flight transmission events.

We aim to answer the following primary and secondary questions:

1. Is there any evidence that MERS-CoV has been transmitted to passengers and/or crew on aircraft?

2. Have any interventions been taken after a case of MERS-CoV travelled on aircraft? What were they and what were the outcomes?

METHODS

This systematic review was conducted following the Preferred Reporting Items for Systematic reviews and Meta-analyses (PRISMA) guidelines³⁵, and the protocol was registered in advance at the International Prospective Register of Systematic Reviews (PROSPERO), an international database created by the Centre of Reviews and Dissemination of The University of York and funded by the National Health Service (NHS)³⁶.

For the primary question, the study population was defined as all aircraft passengers and crew exposed to a case of MERS on board aircraft during a flight. The outcome sought was whether any in-flight transmission to passengers or crew had occurred.

For the secondary question, the study population was defined as all passengers and crew after exposure to a case of MERS on board aircraft, with intervention defined as contact tracing, laboratory testing or authorities informing the study population about the exposure to a MERS case. The outcome sought was whether any of these interventions may have altered the risk of onwards transmission of the disease.

For this review, a “case” refers to a passenger on board an aircraft who was diagnosed with MERS-CoV infection (using molecular methods as stated in the WHO case definitions) shortly after (less than a week) or before the flight, and who had signs or symptoms compatible with the disease that initiated before or during the flight.

A possible event of in-flight transmission was defined as any person on board an aircraft where a case of MERS was present, and who was diagnosed with MERS (using the Polymerase Chain Reaction [PCR] or a viral culture) after the flight, who had no other known

(10)

previous exposure to the virus or risk factors (i.e. direct or indirect contact with camels), and with symptom onset within 14 days of the flight.

Eligibility criteria

All reports describing a case of MERS travelling on aircraft while symptomatic were included in this review. Reports where no public health measures are described after a case of MERS has travelled by aircraft were also included. A report was excluded if the case was asymptomatic during the flight, or if it described an importation or exportation of MERS-CoV between countries, but without specifying that the case had travelled by aircraft. Other exclusion criteria included the setting not being an aircraft, the population not human or the disease not MERS.

Information sources

The electronic literature databases used for retrieval of peer-reviewed articles were Medline (PubMed), Embase, Scopus, and Global Index Medicus. Google was used to search for grey literature (i.e. literature not yet published in a peer-reviewed article). Additionally, hand searching was performed on relevant events reported in the EU Early Warning and Response System (EWRS), a communication tool between health authorities within countries of the EU;

and on the WHO Disease Outbreak News (DON) section for MERS-CoV. Experts were contacted to complete information missing from sources when needed.

Search strategy

Keywords from natural and controlled vocabulary (MeSH terms and Emtree terms) were identified for each component of the study questions to be used in the electronic literature search. The specific search strategy was created with input from all authors and with support of a Medical Library Specialist, and subsequently peer-reviewed by a second Medical Library Specialist.

No limits were set regarding time coverage, language, type of study design, or publication status.

Forward and backward reference checking of the articles selected and all systematic reviews, literature reviews and modelling studies was performed to ensure the identification of all relevant studies.

An alert was set up for all database searches, so notifications of new search results were received up until the final analyses were complete and further eligible studies were retrieved for inclusion.

The search was broadened by including terms related to SARS-CoV on board aircraft to ensure misclassified articles about MERS-CoV were captured, or to gather indirect information on transmission if the number of MERS-CoV related articles was too low to perform an analysis.

An example of the search strategy used for PubMed is included in the box.

Data management

All results from the database search were uploaded to Endnote v7.8; this tool was used for de- duplication of citations, for the initial title and abstract screening process, and for full text retrieval of relevant articles.

(11)

Selection process

For each of the citations resulting from the search strategy, an initial title and abstract screening was performed to identify articles relevant to answer the review questions. Titles and abstracts were assessed independently by two reviewers to increase the reliability of the inclusion and exclusion process. For records lacking abstracts, the full text of articles with relevant titles were considered.

At this step, all systematic reviews, literature reviews and modelling studies were included to perform a reference cross-check in order to validate our search and prevent omissions, even if they were later on excluded while evaluating the full text.

#1 "Coronavirus"[Mesh] OR "Coronavirus Infections"[Mesh] OR "Middle East Respiratory Syndrome Coronavirus"[Mesh] OR coronavirus*[TW] OR cov[TW]

OR hcov[TW] OR ncov[TW] OR middle east respiratory syndrome[TW] OR

"hcov-emc"[TW] OR mers virus*[TW] OR "SARS Virus"[Mesh] OR "Severe Acute Respiratory Syndrome"[Mesh] OR sars[TW] OR "Severe Acute Respiratory Syndrome"[TW] OR "Severe Acute Respiratory infection"[TW] OR "sudden acute respiratory syndrome"[TW]

#2 "Aerospace Medicine"[Mesh] OR "Aircraft"[Mesh] OR "Aviation"[Mesh] OR

"Airports"[Mesh] OR aircraft*[TW] OR aeroplane*[TW] OR airplane*[TW] OR helicopter*[TW] OR airline*[TW] OR flight*[TW] OR aircrew[TW] OR airflight*[TW] OR aviation[TW] OR airport*[TW] OR aeroport*[TW] OR heliport*[TW] OR "aero transport"[TW] OR "air port"[TW] OR steward[TW] OR stewardess[TW] OR inflight[TW] OR "in-flight"[TW] OR cabin[TW] OR cabins[TW] OR (("Travel"[Mesh] OR travel*[TW] OR transport*[TW] OR transport hub*[TW] OR journey*[TW] OR trip[TW] OR trips[TW]) AND air[TW]) OR ((plane[TW] OR planes[TW]) AND (air[TW] OR travel*[TW] OR

"Travel"[Mesh] OR transport*[TW] OR journey*[TW] OR trip[TW] OR trips[TW])) OR ((passenger*[TW] OR crew[TW] OR traveller*[TW] OR traveler*[TW] OR personnel[TW] OR staff[TW] OR pilot*[TW]) AND (flying[TW] OR fly[TW] OR air[TW]))

#3 (#1 AND #2)

#4 ("time of flight"[TW] AND spectrometry[TW])

#5 (#3 NOT #4) Limits: no limits Results: 192

Date of search: 18 February 2019

Box: Example of search strategy, used for PubMed. The first two search strings (#1 and #2) were combined so that the databases would search for papers that included terms from both sections. Any papers with the combination “Time of flight” AND “spectrometry” were excluded from the search to avoid identifying studies related to the analytical technique MALDI-TOF, which was not relevant for this study.

(12)

For all sources passing the initial abstract screening process, a copy of the full text was retrieved and evaluated by the two reviewers.

Articles were included when both reviewers determined that the article met the inclusion criteria. All disagreements between the reviewers were resolved by consensus.

Data collection process

Data about the flight, the confirmed MERS case on board, the contact tracing investigation of the flights, and any additional public health care interventions implemented was extracted using pre-designed tables, in order to systematise the data collection and evaluation. Explicit descriptions of each item were formulated before the extraction. All specific items retrieved are listed in table 2.

The data extraction process was performed by one author in duplicate, to avoid omissions. Six experts on the field were contacted through email for clarification of data or to request information missing from records, and a questionnaire was produced for them to fill in.

When information between different sources regarding the same case conflicted, the information on the source with the highest quality of evidence score (using the bias assessment tool) was considered.

WHO DON reports that only provided an update of a previous mentioned case, but did not report a new case, were regarded as duplicates of the first.

Flight times were estimated when not reported, by using Google Maps. When no city of departure/destination was available, capital cities were used. For all MERS cases that travelled by flight while symptomatic, country of flight origin was considered the country of probable exposure, in order to facilitate data analysis.

Risk of bias in individual studies

Risk of bias assessments were performed on all included records using a modified version of the Bias Assessment Tool (table 3) developed by Leitmeyer and Adlhoch, 2016³¹. No studies were excluded based on the score obtained.

Data synthesis

All data was collected, summarised and analysed using Excel.

Table 2: data extracted from included records Flight characteristics

Flight origin and destination, type of aircraft, date of flight, flight duration, ground delay, number of passengers and crew members on board, HEPA filter function.

Confirmed case on board

Country of residence, nationality, age, sex, date of symptom, onset, symptoms/signs during flight, date of diagnosis, sample taken and method of diagnosis, seating characteristics on aircraft, family traveling with case, outcome.

Contact tracing investigation of flight

Country initiating CT, other countries involved, starting date, duration, definition of contacts, methods used to identify and reach contacts, number of contacts, successfully traced contacts, contacts followed for 14 days, contacts with respiratory symptoms, contacts tested, contacts with positive results.

Other

Alternative or additional measures taken by authorities, starting date of intervention, duration of intervention, other comments.

(13)

RESULTS

Identified records

A total of 48 records (18 peer-reviewed articles, 10 EWRS notifications and 20 WHO DON reports) describing 22 cases of MERS who had travelled on aircraft were identified for inclusion in this review. The search of online medical databases provided a total of 729 citations, plus 49 additional records obtained by the set-up alerts. Another 239 records were identified through hand searching, Google searching and reference checking, for a total of 1,017 records. After de-duplications, 727 records remained for title and abstract screening. Of these, 635 records were discarded because they did not meet the inclusion criteria. The full text of the 92 remaining records was evaluated in more detail, and 48 were deemed relevant for this review. Reasons for exclusion are described in figure 3.

Table 3: Bias Assessment Tool

Criteria

Points Awarded

or Withdrawn Index case classification

Laboratory confirmation 1

Unspecific clinical presentation or data not provided 0 Secondary case ascertainment

Laboratory testing of possible cases on flight 2

Syndromic (i.e. MERS-CoV-like illness) or no comprehensive

confirmation of all possible cases 1

Not provided 0

Public health interventions

Contact tracing of flight 2

Other 1

No intervention conducted or not mentioned 0

Timeliness of contact tracing of flight

Within 1 week 2

Within 2 weeks 1

3 weeks or more 0

Proportion of aircraft contacts followed up

More than 80% followed up 2

Between 80% and 50% were followed up 1

Less than 50% were followed up or retrospective identification 0 Limitations

Alternative exposure before flight possible/alternative exposure not

addressed -1

Resulting evidence levels: 0-4 low, 5-7 medium, 8-9 high.

(14)

None of the records included described all items sought (table 4). All records reported country of flight origin and destination, although 28 (58%) did not report city of departure and 17 (35%) did not report city of destination. None of the records mention any ground delay or the HEPA filter function.

One article mentioned the type of aircraft, one mentioned the exact total flight duration, two mentioned the total number of passengers and crew members on board, and one mentioned the sitting characteristics of the case on board. Half or more of the records mentioned date of flight, country of residence of the case, nationality of the case, age, sex, date of symptom onset, symptoms present during the flight, date of diagnosis and additional measures taken by health authorities. Seven records did not mention the case’s exact age, only the decade.

Twenty-eight records mentioned a contact tracing investigation was done on the case’s flight.

Six experts were contacted for clarification or to request additional information on cases, none of whom responded before the writing of this report.

Figure 3: flow diagram of study selection process.

(15)

Of the 48 records, 32 were classified as having a low evidence level, which included all WHO DON reports and most of the EWRS, due to the limited amount of information contained in them. Eleven records obtained a medium score and 5 got a high score. Four of the five records with a high score were peer-reviewed articles, the other one was an EWRS notification. Most of the peer-reviewed articles obtained a medium score (Appendix 1).

Table 4: reported items in 48 sources

Item No. %

Flight origin and destination 48 100.0

Type of aircraft 1 2.1

Date of flight 45 93.8

Flight duration 1 2.1

Ground delay 0 0.0

Number of passengers and crew members on board 2 4.2

HEPA filter function 0 0.0

Country of residence 38 79.2

Nationality 25 52.1

Age 29 60.4

Sex 41 85.4

Date of symptom onset 40 83.3

Symptoms/signs present during flight 28 58.3

Date of diagnosis 29 60.4

Sample taken 14 29.2

Method of diagnosis 17 35.4

Seating characteristics on aircraft 1 2.1

Family travelling with case 9 18.8

Outcome 15 31.3

Country initiating CT 29 60.4

Other countries involved in CT 19 39.6

Starting date of CT 10 20.8

Duration of CT 15 31.3

Definition of contacts 20 41.7

Method(s) used to identify and reach contacts 13 27.1 Total number of identified aircraft contacts 17 35.4 Number of successfully traced aircraft contacts 15 31.3 Number of aircraft contacts followed for 14 days 11 22.9 Aircraft contacts with respiratory symptoms/fever

within 14 days of flight 12 25.0

Number of aircraft contacts tested 11 22.9 Aircraft contacts with positive results 10 20.8 Additional/alternative measures taken by health

authorities 26 54.2

(16)

Cases of MERS that travelled by aircraft

The 22 cases described boarded a total of 31 flights while symptomatic. Three of the cases were transported by air ambulances to another country. No secondary cases of in-flight transmission were identified among all flights.

Seventeen cases were male (77%).

The average age was 55.4 years (interval: 18-85) (figure 4).

Fifteen cases boarded only one flight during their trip, 6 boarded two, and one took four. Most cases travelled during 2014 (figure 5), and the average total flight time was 5.2 hours (range 1.3-9.5 hours). Most cases were probably exposed in Saudi Arabia (country of flight origin) (figure 6).

Eight cases were residents of Saudi Arabia. Other countries of residence included Qatar, United Arab Emirates (UAE), United Kingdom (UK), Italy, Spain, Netherlands, South Korea, Oman, Kuwait and Malaysia.

On average, date of symptom onset preceded the date of flight by 6.2 days (range 0-19). The most common symptom present during flight was fever (figure 7).

The three cases transported by air ambulance had a severe disease and were intubated.

Seven cases travelled with family members.

The outcome of the disease was not mentioned for all cases (68.9% had no information), however, at least 3 died and 11 were discharged from the hospital as cured. CFR among the 22 cases was 14%, although this could be underestimated since the disease outcome for all cases is not known.

Figure 4: Number of MERS cases per age group.

Figure 5: Number of MERS cases that travelled by flight per year.

Figure 6: Country of probable exposure (flight origins).

(17)

Contact tracing investigations

Out of the 48 records included in the review, 28 stated that a contact tracing (CT) investigation had been carried out on the flight’s case, and of those, 23 described the CT results. It was not mentioned if CT was performed for 4 of the 22 cases. The total number of flights traced was 25, out of the 31 flights included in this review. At least 19 countries are mentioned to have been involved in one of the CT investigations: Germany, the UK, Italy, Spain, Greece, the USA, Qatar, Saudi Arabia, Netherlands, Austria, Turkey, China, Poland, Philippines, South Korea, Taiwan, Oman, Thailand and Malaysia.

CT investigations were initiated an average of 0.4 days after the case was diagnosed with MERS, that is, either on the same day of diagnosis or before, when MERS-CoV infection was suspected.

The cases were diagnosed an average of 6.1 days after they had travelled on aircraft. The average reported duration of CT investigations was 13.5 days. “Contacts” for CT investigations were most commonly defined as the passengers sitting in the same row as the case and in the two rows in front and behind (known as the two-row rule). Contact definitions for passengers and crew are described in more detail in table 5.

Table 5: Contact definition per country

Country Year Definition for passengers Definition for crew Germany (air

ambulance) 2012

Contact tracing of healthcare workers included those during transport to the hospital

.

UK 2013 Two-row rule .

Italy 2013 Two-row rule All crew

Spain 2013 Two-seat radius Crew attending case

Greece 2014 Two-row rule .

Figure 7: Symptoms presented during flight.

(18)

UK 2014

Two-seat radius prioritised, two rows in front and behind only followed if symptomatic

.

USA 2014 All passengers and crew (for all

emerging diseases) All crew

Netherlands 2014 Within three rows of case .

China 2014 Close contacts: two-row rule; other

contacts: rest of passengers .

Philippines 2015

Category A: sitting in surrounding 3 rows; category B: in surrounding 3 rows but only transited in Philippines;

category E: all other passengers

.

Thailand 2015 Passengers in the two rows surrounding case. Low risk: contact further than 1m

High risk: interaction closer than 1m (crew); low risk:

contact further than 1m

UK 2018 Three rows in front and behind .

The most widely used method for identifying contacts was requesting the passenger manifest and contact details from the airline. Contacts were most commonly reached by phone (table 6). Some countries set up a hotline so passengers in the flight could call if they developed symptoms.

Table 6: Methods used to identify and reach contacts by country

Country Methods used to identify contacts Methods used to reach contacts

Germany (air ambulance)

Requested healthcare workers to report contact with the patient during transport to hospital

Questionnaire given to fill in

Italy Requested contact details from airline . Greece Requested contact details from airline Phone

UK

Requested contact details from airline;

requested through press release that passengers of the flight communicated to a health telephone service

Phone

Austria Requested contact details from airline Crew data sent to WHO to communicate to Qatar

(19)

China

Passenger list provided by WHO through IHR; airline provided seating plan and contact details; travel agency provided tour member list; hotline set up and case's travel details published

Some contacts had an initial interview in person, and were monitored by phone; others called hotline

Philippines . Interviewed in person

USA

Order passenger manifest from airline, use of federal databases (i.e. Border Patrol), custom declaration forms, contacted PHE for details of Riyadh- London passengers

Phone, email, letter, interviewed in person.

Crew contacted by airline

Thailand Requested passenger details from airline

Phone, located at address, voluntarily reported and interviewed in person

The average number of contacts identified per flight were 97. This widely varied between cases, depending on the contact definition used (two or three rows, whole plane), and on the number of flights boarded by each case. In most CT investigations, more than 50% of the identified contacts were reached and followed for the 14-day incubation period of MERS- CoV. CT investigations are described in detail in table 7.

The most important factor referred by authors that delayed the possibility of conducting an adequate CT of the flight was the unavailability of all passengers’ contact details, since many airlines do not demand these details for booking tickets.

Additional or alternative health interventions carried out

The most common additional health intervention described in the reports included CT of healthcare workers, family members and other contacts of the case, which was mentioned in 19 records.

Other additional or alternative interventions carried out by national public health authorities included: press releases to alert passengers and crew about a possible exposure to MERS-CoV and to inform on what measures to take, such as going to a health care professional if they develop any symptoms (mentioned in 9 reports); setting up a hotline for passengers on the flight to be able to reach authorities (1 report); and communication with international public health authorities of countries whose nationals had been on board the flight (6 reports).

One of the countries also decided to evaluate the psychological stress generated by the CT investigation in the passengers of the flight, and reports that CT seems to be a stressful event for passengers.

Other findings

During the screening process, 16 records were found describing 8 cases of MERS that travelled by aircraft before symptom onset. For two of these cases, CT of the flights was reported, and no secondary cases related to these flights have been reported to date by any of the involved countries.

(20)

Table 7: Contact tracing investigations of flights Positive . . . 0 . 0 0 . 0 0 0 0 0 . 0 0 0 . . . . . 0 0 0

Contacts tested . . . 1 . 17 18 . 3 230 17 17 1 . 85 6 . . . . . . 395 39.5 17

Contacts that developed symptoms . . . 2 0 . . . 16 35 2 2 1 . 0 0 0 . . . . . 58 5.8 2

Contacts followed 14 days (%) . . . 11 (55) 9 (100) 17 (100) 18 (100) . 42 (15) 3 (0.5) 17 (100) 17 (100) 43 (100) . . 6 (22) 26 (29) . . . . . 209 19 17

Contacts reached (%) . . . 11 (55) 9 (100) 17 (100) 18 (100) . 144 (53)* 450 (80)* 17 (100) 17 (100) 43 (100) . 85 (35) 27 (100) 89 (100) . . . . 17 (94) 944.0 72.6 18

Identified contacts . . . 20 9 17 18 12 269 561 17 17 43 . 237 27 89 . . . . 18 1354 96.7 19

Flights boarded 1 1 1 1 2 1 1 2 2 4 2 2 2 1 1 1 1 1 1 1 1 1 Total Mean Median

Case 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

Sources EWRS 2012/09/2337 , Bermingham38 , Pebody39 , DON 2012/09/2340 EWRS 2012/11/2337 , Buchholz41 EWRS 2012/11/2337, Reuss42, DON 2013/03/2643 EWRS 2013/02/0837, HPA44, DON 2013/02/1145 EWRS 2013/05/3137, Puzelli46, DON 2013/06/0147 EWRS 2013/11/0537, DON 2013/11/0748 EWRS 2013/11/0537 EWRS 2014/04/1837, Tsiodras49, DON 2014/04/2050 EWRS 2014/05/0237, Bialek51, Parry-Ford52, Regan53, Lippold54, DON 2014/05/0555 EWRS 2014/05/0237, Bialek51, Parry-Ford52, Regan53, Lippold54, DON 2014/05/1456 EWRS 2014/05/1437, Kraaij57, Mollers58, DON 2014/05/1559 EWRS 2014/05/1437, Kraaij57, Mollers58, DON 2014/05/1660 EWRS 2014/09/3037, Kwok-ming61, DON 2014/10/0262 DON 2014/10/2463 Racelis64, DON 2015/02/1365 Wu66, Kang67, DON 2015/05/3068 Plipat69, DON 2015/06/2070 DON 2016/01/2671 DON 2016/01/2972 DON 2016/08/2673 DON 2018/01/0874 EWRS 2018/08/2337 , DON 2018/08/3175 *Includes people who rejected interview

(21)

DISCUSSION

In this review, no evidence of secondary transmission of MERS-CoV on board aircraft was found among the 22 cases and 31 flights included in the study. To the authors’ knowledge, there have been no secondary cases reported from flights worldwide, an interesting finding considering there were at least 26 secondary in-flight transmission events during the multi- state outbreak of SARS-CoV³⁰, a close relative of MERS-CoV. MERS-CoV seems not to have acquired the ability of sustained transmission between humans. Although SARS-CoV was around for only a year (no new cases have been reported since 2004)⁷⁶, it infected 8,096 people, almost 4 times the amount of people infected with MERS-CoV in the 7 years that have passed since its discovery.

The demographics of our results (male predominance and average age of 55 years) were as expected compared to the global epidemiology of the disease. Most cases were probably exposed in Saudi Arabia, the country where a majority of MERS-CoV infections have been reported. The low CRF observed could be due to the missing data on the disease outcome for most of the cases, or simply because severely ill individuals are less likely to travel.

Cases 1 to 3 in this review (see table 7) were transported via air ambulance, and information about the people present during the transport, as well as whether CT was carried out for those flights, is lacking.

Although it is not expected that every record identified will report all items sought in the review, none of the sources mentioned whether there was any ground delay of the flights, or whether it was known if the HEPA filter was functioning adequately. Since these two factors have proven important in previous in-flight transmission events³³, more attention should be paid to them when evaluating in-flight transmission events. Similarly, only two records mentioned the total number of passengers and crew members on board, one mentioned the exact duration of the flight, and one mentioned the seating characteristics of the case. All these items are important for the CT investigations and should be mentioned in the description of CT investigations for a more complete analysis of events.

The aim of this review was to gather all available information on MERS-CoV cases that have travelled by aircraft and any evidence of in-flight transmission events, with the ultimate goal of producing an update of the RAGIDA project that will include a chapter on MERS-CoV, nonetheless, having no transmission events during any flight hinders the possibility of making an assertive recommendation on whether to trace a limited number of rows or the whole plane when a case of MERS-CoV travels by aircraft. Since no transmission has been observed either on rows around the case or on further rows, we cannot know for certain which passengers are more likely to be infected. However, the lack of in-flight transmission observed in this study might suggest that a more conservative approach with fewer rows traced could be more adequate, considering that MERS-CoV is transmitted by droplets or close contact. Following the completion of this systematic review, an expert meeting will be held to discuss the aforementioned points and to determine the most adequate course of action to take whenever a case of MERS-CoV has travelled by aircraft.

An important factor to consider when deciding who to trace, is the movement of passengers around the aircraft. Any passenger that changed seats might therefore have to be included (or excluded) in the CT investigation. The case may also have interacted with passengers sitting on further rows during the flight.

(22)

It was encouraging to find that most CT investigations were initiated soon after the cases were diagnosed with MERS-CoV. This highlights that most countries have appropriate guidelines in place and are able to respond timely to these events.

Limitations

The most significant limitation was the scarce amount of data regarding the flights and the CT investigations from many of the records included. The review aimed to gather the information available of all known MERS-CoV cases to have travelled by air, so records were not excluded based on the quality or quantity of the information given.

For 18 of the 22 cases included more than one source of information was identified. Two cases were mentioned in six different records. Incorporating all the information from several sources, which was presented differently in each, into one set of results, while avoiding making incorrect assumptions, was also challenging.

Although experts were contacted to request the missing information and the initial response was positive, finding long-ago archived data and retrieving the requested items from it takes time, and no response was received by the time of writing of this report, which could have improved the analysis.

CONCLUSION

In this review evidence is provided to show no in-flight transmission has been observed after cases of MERS travelled on board aircraft. It is also shown that the most common public health intervention carried out after such events is contact tracing.

These results may have important implications for public health action in similar situations, when it is important for health authorities to decide whether to conduct CT on all passengers of the flight, or only on passengers sitting on selected rows. The lack of in-flight transmission observed in this study might suggest that a more conservative approach with fewer rows traced may be appropriate, considering the large economical and human resources needed for conducting CT investigations^30,52.

ACKNOWLEDGEMENTS

We would like to thank everyone who supported and advised us throughout the development of this review. Special thanks are due to both of ECDC´s library specialists Anna Belen Escriva and Helena Simanova; to Helena de Carvalho Gomes, for her support on the methodology during the initial stages of the review; and to Pasi Penttinen and Maria van Kerkhove for her expert advice.

RISK ASSESSMENT FOR THE TRANSMISSION OF MIDDLE EAST RESPIRATORY SYNDROME CORONAVIRUS (MERS-CoV) ON AIRCRAFT: A SYSTEMATIC REVIEW