Prevention of influenza at Hajj: applications for mass gatherings

Summary

Outbreaks of infectious diseases that spread via respiratory route, e.g. influenza, are common amongst Hajj congregation in Mecca, Saudi Arabia. The Saudi Arabian authority successfully organized the Hajj 2009 amidst fear of pandemic influenza. While severe influenza A(H1N1)pdm09 was rare, the true burden of pandemic influenza at Hajj that year remains speculative. In this article we review the latest evidence on influenza control and discuss our experience of influenza and its prevention at Hajj and possible application to other mass gatherings. Depending on study design the attack rate of seasonal influenza at Hajj has ranged from 6% in polymerase chain reaction or culture confirmed studies to 38% in serological surveillance. No significant effect of influenza vaccine or the use of personal protective measures against influenza has been established from observational studies, although the uptake of the vaccine and adherence to face masks and hand hygiene has been low. In all, there is a relatively poor evidence base for control of influenza. Until better evidence is obtained, vaccination coupled with rapid antiviral treatment of symptomatic individuals remains the mainstay of prevention at Hajj and other mass gatherings. Hajj pilgrimage provides a unique opportunity to test the effectiveness of various preventive measures that require a large sample size, such as testing the efficacy of plain surgical masks against laboratory-confirmed influenza. After successful completion of a pilot trial conducted among Australian pilgrims at the 2011 Hajj, a large multinational cluster randomized controlled trial is being planned. This will require effective international collaboration.

Introduction

Since the influenza pandemic of 2009 was declared, Saudi Arabia has hosted the annual Hajj pilgrimage three times. Every year about two and half million practising Muslims from over 180 nations attend. For about 1400 years, the mass gathering at Hajj has been associated with the risk of communicable diseases, particularly respiratory infection. This risk is increased by travel (for some long and arduous), severe crowding at Hajj, the weather and the stress of Hajj rituals.^1–4

The weeklong Hajj consists of a number of intricate religious rituals during the final month of the Islamic lunar calendar. These start before reaching the boundary of Greater Mecca when men exchange their clothing for plain white sheets to signify a spiritual state. Immediately on reaching the city of Mecca, masses of pilgrims circumambulate seven times the Kaabah, signifying to Muslims the House of God (Figure 1). Later, moving to the unpopulated valley, Mina (Figure 2), at the outskirts of Mecca, pilgrims undertake a three-day period of contemplation and supplication in basic but clean tents before spending a night in the open at Muzdalifa. Completion of Hajj also requires a metaphorical denunciation of Satan by casting stones at three symbolic pillars and the sacrifice of quadruped animals (most commonly cattle, sheep, goats and camels) at designated abattoirs.

Figure 1.

Assembly around the Kaabah

Figure 2.

Assembly at Mina

Extreme crowding in these rituals increases the risk of transmission of airborne infections. Influenza remains a major current concern and continuing vigilance is needed to recognize emergence of new mutations and the threat of crossing the species barrier.⁵

This review covers knowledge of control of influenza at Hajj, some recent developments in its epidemiology and the evidence of effectiveness of a range of preventive measures. We make recommendations for future prevention and control of influenza and for research at Hajj with application to other mass gatherings.

Methods

A PubMed search of the medical literature dating from 1950 to mid-2012 on the epidemiology or prevention of influenza at Hajj and Umrah was undertaken to identify all relevant peer-reviewed publications. MeSH terms ‘influenza’ or ‘flu’ AND ‘Hajj’ (also alternative spelling Hadj and Haj) OR ‘Umrah’ were used as well as other relevant terms such as ‘pneumonia’, ‘respiratory infections’, ‘infectious diseases’, ‘vaccine’, ‘infection control’, ‘communicable diseases’ AND ‘Mecca’ (alternatively ‘Makkah’) OR ‘pilgrimage’ were used. A freehand Google engine search using the same search terms supplemented our search to identify articles which were not indexed. Excluding review articles and commentaries 125 research articles were identified which included case series, cross-sectional, case-control and cohort studies. Of the 125 articles, 27 (21.6%) dealt with respiratory tract infections including 15 (12%) papers on laboratory-proven influenza. In addition, published guidance from the World Health Organization (WHO), the US Centers for Disease Control and Prevention (CDC), the UK Health Protection Agency, the European Centre for Disease Prevention and Control, Saudi Arabian Ministry of Health (MoH) and Ministry of Hajj were searched manually. The search result is summarized in Figure 3 and the full list will be available from the authors on request. Though some relevant randomized controlled trials were identified, none was undertaken at Hajj. We examined all studies to identify overlap and excluded studies where overlap was possible. The steps for a systematic review were followed and as far as possible the measures of effectiveness compared. We relied on the latest systematic reviews for the evidence of protection against influenza. Whenever an up-to-date systematic review (including a Cochrane review) was identified, we did not look for earlier observational/controlled studies to support or refute the evidence.

Figure 3.

Summary of the identified articles according to years of publications (in this review only papers that dealt with respiratory infections have been included)

What do we know about influenza at Hajj?

Jeddi et al.⁶ first documented influenza at Hajj using serology when in 1975 4% (37 of 950) of Tunisian pilgrims were shown to have had recent influenza. Subsequent studies by virological, serological or polymerase chain reaction (PCR)-based surveillance reported attack rates ranging from 6% to 38% depending on study design.^1,2,7–9 The Hajj 2009 took place during pandemic influenza. Although the Saudi MoH recommended seasonal and the influenza A(H1N1)pdm09 (pH1N1) vaccination for all pilgrims to the Hajj 2009, it estimated that about 53% and 30% of pilgrims had received the respective vaccines.⁴ Seasonal influenza vaccine (which contained the pH1N1 strain) was recommended for the elderly and those with chronic diseases before both the 2010 and 2011 pilgrimages.

An estimate of the overall incidence of pH1N1 at Hajj is not available but airport-based surveillance of arriving and departing pilgrims at Jeddah airport showed a low prevalence of pH1N1 or seasonal influenza (each with a prevalence rate ≤0.2%) among the pilgrims.⁴ Memish et al. reported the absence of pH1N1 and seasonal influenza in nasal and throat swabs of healthcare workers (HCWs) before and after their deployment at Hajj.¹⁰ A study conducted at Cairo seaport and airport among 542 vaccinated Egyptian pilgrims showed 1% prevalence rate of seasonal influenza (H3N2) but zero prevalence of pH1N1 implying that the pandemic vaccine was effective against pH1N1.¹¹ The Saudi MoH recorded 73 cases of pH1N1 including five deaths (case fatality ratio 4.9%), during the 2009 Hajj.¹² This small number of cases is much less than that expected during such a mass gathering and suggests success from the strategy for pH1N1 influenza control during the 2009 Hajj season, agreed at pre-Hajj strategic and technical consultation with global public health counterparts known as the Jeddah Hajj Consultancy Group.¹³ This consisted of: (a) advice for high-risk pilgrims not to attend Hajj, (b) hospital based surveillance to confirm suspected cases, (c) strengthening of infection control measures and (d) early treatment of at risk patients with antiviral drugs. The low incidence of influenza is thus likely to be have been due to a combination of cross-protection due to exposure of older pilgrims to earlier H1N1 epidemics¹⁴ and the infection control methods employed, including advice to those with recognized influenza risk factors to avoid Hajj in 2009.^4,13 A recent study reports a fairly high attack rate of seasonal influenza (10%) among 275 Iranian pilgrims returning from the Hajj 2009 all of whom received seasonal influenza vaccine.¹⁵ Another Iranian study conducted at the same time and in the same city measured prevalence of seasonal influenza among highly vaccinated returning pilgrims at 2.6% and the prevalence of pH1N1 at 1.6%.¹⁶ The dominance of seasonal influenza is likely to be due to the large proportion of pilgrims 50 or older with cross-immunity to H1N1.¹⁵

Bacterial infections due to Streptococcus pneumoniae, Staphylococcus aureus, Streptococcus mitis and Haemophilus influenzae can be a complication of severe and fatal cases of pH1N1 infection. While S. pneumoniae is a known cause of sepsis at Hajj,¹⁷ the frequency of bacterial co-infection with pandemic or seasonal influenza at Hajj is not known. Previous studies report occasional dual infection with influenza and other respiratory viruses such as parainfluenza or rhinovirus at Hajj.^7,9

Who is at risk of influenza at Hajj?

While influenza can affect all age groups the established risk factors for complications/severe disease are ages <2 and ≥65 years, and at any age chronic heart, lung, kidney, liver, blood or metabolic diseases (such as diabetes), immunosuppression, or, for pH1N1 extreme obesity. At least one-fourth of the pilgrims suffer from one or more of these risk factors.^4,18 The prevalence of diabetes, bronchial asthma and chronic obstructive airway disease is 12%, 9% and 9%, respectively.

How can effective prevention be achieved?

Interventions to prevent influenza can be considered as pharmaceutical and non-pharmaceutical methods. Pharmaceutical measures are vaccinations against influenza and other concomitant infections such as pneumococcal disease, and the use of antivirals, both of which apply to pilgrims but with poor uptake. Non-pharmaceutical methods of prevention include surveillance, identification of cases by rapid viral testing, contact tracing, personal protective equipment such as surgical and N95 masks, hand hygiene, respiratory etiquette, isolation of the sick, quarantine of the exposed, social distancing, travel restrictions (international and domestic), and exit and entry screening.^19,20

Pharmaceutical interventions

Influenza vaccine

WHO recommends that the most effective way to prevent influenza or its severe outcomes is by vaccination. The commonly used trivalent influenza vaccines are directed against two strains of influenza A and one strain of B, and are most effective when circulating influenza viruses are well-matched with vaccine strains. Periodic antigenic drift (and, rarely, shift) pose problems for vaccine production and procurement, as a new vaccine closely matching the circulating antigenic configuration must be developed, produced and procured for the beginning of each new influenza season. This relies on advice from the WHO Global Influenza Surveillance and Response System, a partnership of National Influenza Centres around the world to which Saudi Arabia is a contributor. Some countries, but not Saudi Arabia, make independent decisions, based on their own surveillance data.

A recent systematic review by Osterholm et al. showed a pooled efficacy of seasonal influenza of 59% (95% confidence interval [CI] 51–67) in adults aged 18–65 years and a median pH1N1 vaccine effectiveness of 69% (range 60–93).²¹ In 2010/2011 and 2011/2012 the seasonal influenza vaccine included the pH1N1 strain so that only one shot was necessary and not separate monovalent vaccines.

Although Saudi Arabia has recommended influenza vaccine for all with risk factors including those aged ≥65 years since the year 2005, this vaccination is not yet an entry requirement to Saudi Arabia and the uptake by pilgrims to Hajj is low.²² In 2006, the rate of influenza among Iranian as well as Saudi pilgrims was 10% with widely varying reported vaccine uptakes of 86% and 4%, respectively.^7,9 A combined analysis of data for the 2005 and 2006 Hajj years indicated that 5% of vaccinated ‘at risk’ pilgrims compared with 14% of unvaccinated (relative risk 0.37, CI 0.1–1.4) had confirmed influenza while the rate of influenza in vaccinated and unvaccinated ‘not at risk’ pilgrims were similar (10% versus 11%).²³

A multi-pronged approach, including a strong recommendation backed up by health education and making the vaccine easily available and free, perhaps at pre-Hajj seminars, is needed to improve vaccine coverage among pilgrims.

Pneumococcal vaccine

Because influenza predisposes individuals to pneumococcal pneumonia, the 23-valent polysaccharide vaccine (PPV-23) that is currently recommended in many developed countries for use in all adults who are aged ≥65 years and also for persons aged ≥2 years at high risk for disease (e.g. sickle cell disease, HIV infection or other immunocompromising conditions) should be considered for pilgrims to Hajj. The recent development of widespread microbial resistance to essential antibiotics in many countries, including Saudi Arabia, has supported the urgency of better control by PPV-23 immunization, particularly in view of its established efficacy against invasive pneumococcal disease and to a small extent against pneumonia. Currently, most (>95%) pilgrims at risk of pneumococcal infection are unvaccinated.²⁴ Recently, 13-valent conjugate pneumococcal vaccine has been licensed in the USA for those aged ≥50, and economic analysis favours its use.²⁵ Data on serotype distribution of pneumococcal disease at Hajj are needed to recommend it for pilgrims.

Antivirals

Antiviral drugs against influenza are available in some countries for both its treatment and prophylaxis. The neuraminidase inhibitors, oseltamivir and zanamivir, are the drugs of choice. Though they do not shorten the period of illness by much more than a day, they attenuate the disease, reducing fever, viral shedding and infectivity and probably the development of pneumonia. The Kingdom of Saudi Arabia along with other countries have disease control programmes that follow WHO advice to target influenza infections in the vulnerable and high risk. The Jeddah Hajj Consultancy Group recommended treatment for all inpatients and outpatients at high risk of severe complications from pH1N1 or seasonal influenza, but not antiviral chemoprophylaxis because of the risk of resistance as well as logistic problems in provision.¹³ Efficacy of neuraminidase inhibitors in preventing influenza in a range of population subgroups has been supported by a systematic review, with oseltamivir reported to have a protective efficacy of 81% against symptomatic laboratory-confirmed influenza in household contacts of cases.²⁶ However, should a pandemic cause mass infection at Hajj or other mass gatherings, ring postexposure prophylaxis, in which only the close contacts of cases are targeted (e.g. pilgrims sharing a tent with a case), is probably the only workable strategy.

Non-pharmaceutical interventions

Disease surveillance and case reporting

Though there is no strong evidence of the effectiveness of influenza surveillance and case reporting, experience with Severe Acute Respiratory Syndrome (SARS) and other respiratory disease outbreaks indicates that these indirect measures are important in controlling epidemics, particularly an influenza pandemic. Influenza surveillance enables better understanding of the epidemiology of influenza in the community and this accurate up-to date information guides and targets appropriate interventions.

Rapid viral testing

A number of near-patient tests of pH1N1 influenza have been evaluated in the USA, all showing moderate sensitivity (38–53%) but high specificity (86–100%). So far only one named near-patient test for seasonal influenza has been evaluated at Hajj with low sensitivity (22%) but high specificity (99%).²⁷ A clinical case definition, comprising the triad of cough, subjective fever and sore throat, with acceptable sensitivity (67%) and specificity (64%) could be used instead, supported by confirmatory PCR testing at Hajj. For the Hajj 2009, the Kingdom of Saudi Arabia made viral PCR testing available in key hospitals to enable accurate and rapid diagnosis of influenza in suspected cases, surveillance/epidemiology of the disease and targeting control.¹³

Personal protective measures

Uncertainty about the mode of influenza transmission has influenced debate about when and whether to use masks or N95 respirators for pandemic influenza. Droplet transmission is thought to be the primary mode of transmission, and provides the basis for the CDC guidelines that healthcare personnel wear masks for close patient contact (i.e. within 3 feet) to control influenza transmission during the influenza season. Observational studies have failed to demonstrate any clear benefit of using facemasks among Hajj pilgrims, but no large trial has yet been conducted at Hajj.

A recent systematic review on the use of masks and respirators to prevent transmission of influenza concluded that there was no convincing evidence of protection against influenza infection and that most of the studies examined were ‘too small to detect reliably what would be anticipated to be moderate effects’. The authors suggested that a large multicentre trial conducted for multiple years is needed to provide an answer.²⁸ Hajj pilgrimage provides an excellent opportunity to conduct such a large trial in a single setting. This could build on a successful pilot study undertaken among Australian pilgrims during the Hajj 2011. This was a cluster randomized controlled trial to test the effectiveness of masks in preventing transmission of influenza-like illness among Australian pilgrims at Hajj. For five days 164 pilgrims aged 15 years and over were recruited from 24 tents in Mina. Currently, the data analysis is underway.

Hand hygiene and respiratory etiquette

A recent Cochrane review supports simple and low-cost interventions, particularly hand hygiene, reducing the transmission of epidemic respiratory viruses.²⁹ Experimental studies also support this as well as an observational study involving HCWs employed at Hajj which demonstrated that failure to use alcohol-based hand disinfection carries a high risk of acquiring acute respiratory infection (adjusted odds ratio 8.4, CI 2.2–32.2).³⁰

Respiratory etiquette, which includes not spitting and covering the mouth and nose when coughing or sneezing, as well as hand washing, are infection control measures which need improvement at Hajj.

Isolation of the symptomatic

For those infected with influenza, viral shedding may begin even before the onset of symptoms so that isolation may be of limited value. While quarantine of contacts of a confirmed case might be considered, the evidence on quarantine measures is poor and most infection control experts do not recommend mandatory isolation. However, the Jeddah Hajj Consultancy Group recommended that, in the absence of sustained community transmission, any pilgrim showing signs that meet the definition of influenza-like illness should be isolated on their arrival to Saudi Arabia, until isolation capacity is exceeded.¹³

What works to control the spread of influenza?

This review of the literature, which includes a number of systematic reviews, has exposed a relatively poor evidence base for the control of influenza. The mainstay of prevention is vaccination coupled, in an outbreak/epidemic, with antiviral treatment of cases as soon as possible after they become symptomatic (but within 48 hours of onset) and, possibly, post-exposure prophylaxis of those exposed. For this, good surveillance is essential with virological confirmation of cases, to accurately document the epidemiology of the disease and focus interventions. Hand hygiene should be an important intervention at Hajj particularly as it complements Muslims’ ablutions five times a day before prayers.

Information from the Hajj 2009 showing low rates of pH1N1 suggests that:

1. Older pilgrims had some cross-immunity to pH1N1 from previous H1N1 influenza A infection;

2. Many with known risk factors, including pregnant women, heeded the official advice from the Ministry of Hajj to avoid Hajj 2009;

3. Control measures worked;

4. Targeted information and advice achieved voluntary limitation of exposure in those at particular risk of influenza.

Future action

Disease prevention applies to a whole defined population and needs to be programmatic and well-coordinated. Preparation for Hajj or other mass gatherings, e.g. the Olympic Games, requires an understanding of health risks and the preventive measures necessary. Explicit supportive arrangements between the host country and the country of residence of participants is necessary, as at Hajj, where the Saudi MoH coordinates immigration and health issues with countries that send pilgrims.

Influenza control continues as a worldwide challenge which requires a better evidence base. Hajj offers a unique opportunity for good, well-coordinated large epidemiological studies and other jointly funded research over its short duration as well as year on year. Following the pilot cluster randomized controlled trial mentioned above, a group of international investigators is planning to undertake a large-scale trial to assess the role of plain surgical masks against influenza at Hajj 2013. The success of such an ambitious undertaking requires effective international collaboration. We are inviting expressions of interest in such a multinational collaboration from potential collaborators.

Meanwhile, the control influenza at Hajj and other mass gatherings depends largely on continued quality surveillance, including post-event surveillance, widespread implementation of vaccination, antiviral prophylaxis and behavioural measures.

DECLARATIONS

Competing interests

RB has received financial support by vaccine producers, including CSL, Sanofi, GSK, Roche and Wyeth, to attend/present at scientific meetings; if fees were offered, these were placed in a university research account. The other authors declare that they have no conflict of interest in relation to this work

Funding

None declared

Ethical approval

Not applicable

Guarantor

EH

Contributorship

All authors contributed equally

Acknowledgements

None

Provenance

This article was submitted by the authors and peer reviewed by Aziz Sheikh