Portal de Periódicos da CAPES

The epidemiology of meningococcal disease in India

2010; Wiley; Volume: 15; Issue: 12 Linguagem: Inglês

10.1111/j.1365-3156.2010.02660.x

1365-3156

David Sinclair, Marie‐Pierre Préziosi, Thomas John, Brian Greenwood,

Virology and Viral Diseases

Objective To undertake a review of the literature on the epidemiology of meningococcal disease in India, with a view to informing future control policies. Methods We searched the PUBMED, EMBASE, Global Health, IMSEAR and MedIND databases for observational studies relating to the burden of endemic meningococcal disease in India, the occurrence and epidemiological characteristics of epidemics, and the prevalence of individual meningococcal serogroups. Results The relatively few reports identified suggest that the incidence of endemic meningococcal disease in India is low, but that occasional epidemics of meningococcal disease have been recorded for at least 100 years. The larger epidemics have affected mainly the cities of northern India and have almost universally been caused by meningococci belonging to serogroup A. These epidemics have showed a few characteristics, including a marked seasonality, which are similar to those of epidemic meningococcal A disease in Africa. Conclusions New serogroup A-containing meningococcal conjugate vaccines are now being developed and reaching the market, including an affordable monovalent serogroup A vaccine manufactured in India, but intended primarily for use in Africa. These new tools may have a role in containing future Indian epidemics, but their usefulness is dependent on early identification of epidemics. This will require a functional disease surveillance system with adequate laboratory support throughout India. Objectif: Revue de la littérature sur l'épidémiologie des maladies à méningocoque en Inde, en vue d'informer les futures politiques de lutte. Méthodes: Nous avons effectué une recherche sur les bases de données PubMed, Embase, Global Health, IMSEAR et MedIND pour des études observationnelles relatives à la charge des maladies méningococales endémiques en Inde, la présence et les caractéristiques épidémiologiques des épidémies, ainsi que la prévalence des différents sérogroupes de méningocoques. Résultats: Les publications relativement peu nombreuses identifiées suggèrent que l'incidence des maladies méningococales endémiques en Inde est faible, mais que des épidémies occasionnelles de maladies méningococales ont été enregistrées depuis au moins 100 ans. Les épidémies les plus importantes ont touché principalement les villes du nord de l'Inde et ont été presque universellement causées par des méningocoques appartenant au sérogroupe A. Ces épidémies ont montré des caractéristiques comprenant une saisonnalité marquée, similaires à celles de l'épidémie des maladies à méningocoque A en Afrique. Conclusions: De nouveaux vaccins conjugués contenant le méningocoque du sérogroupe A sont en cours de développement et arrivent sur marché, y compris un vaccin monovalent du sérogroupe A abordable, fabriqué en Inde, mais principalement destinéà l'Afrique. Ces nouveaux outils peuvent jouer un rôle en limitant de futures épidémies en Inde, mais leur utilité dépend de l'identification précoce des épidémies. Il faudra pour cela un système fonctionnel de surveillance des maladies avec le soutien approprié des laboratoires dans toute l'Inde. Mots-clés: méningocoque, Inde, épidémiologie Objetivo: Realizar una revisión de la literatura sobre la epidemiología de la enfermedad meningocócica en la India con el fin de informar de cara a futuras políticas de control. Métodos: Se realizaron búsquedas en las bases de datos de PUBMED, EMBASE, Global Health, IMSEAR y MedIND de estudios observacionales relacionados con la carga de la enfermedad meningocócica en India, la incidencia y las características de epidemias, y la prevalencia de serogrupos meningocócicos individuales. Resultados: El número relativamente pequeño de artículos identificados sugiere que la incidencia de la enfermedad meningocócica en India es baja, aunque las epidemias ocasionales de enfermedad meningocócica han sido reportadas durante al menos los últimos 100 años. Las epidemias más grandes han afectado principalmente las ciudades del norte de la India, y han sido casi universalmente causadas por meningococos del serogrupo A. Estas epidemias han mostrado características, incluyendo una marcada estacionalidad, similares a aquellas de las epidemias por meningococo del grupo A en África. Conclusiones: Nuevas vacunas meningocócicas conjugadas que contienen el serogrupo A están siendo desarrolladas y van llegando al mercado, incluyendo una vacuna monovalente para el grupo A, asequible y manufacturada en India, pero principalmente intencionada para ser utilizada en África. Estas nuevas herramientas juegan un papel importante en la contención de futuras epidemias en la India, pero su utilidad depende de la identificación temprana de las epidemias. Para ello, se requiere un sistema de vigilancia de la enfermedad con un apoyo laboratorial adecuado alrededor de toda la India. Palabras clave: Meningococo, India, Epidemiología Infection with Neisseria meningitidis (the meningococcus) is common, although the majority of infections are transient and asymptomatic (Stephens 1999). The meningococcus can live harmlessly in the nasopharynx of up to 10% of healthy populations (Yazdankhah & Caugant 2004). This 'carrier' state facilitates the circulation of the bacteria within communities with transmission taking place via airborne respiratory secretions or through direct contact such as kissing (Stephens 1999). The factors that determine progression from harmless carriage to invasive disease are poorly understood, but they are influenced by both the virulence of the infecting strain and the susceptibility of the host (Stephens 1999; Yazdankhah & Caugant 2004). When disease occurs, its onset is often rapid and life-threatening, and the meningococcus remains one of the major causes of death and disability as a result of invasive bacterial disease (Schwartz et al. 1989; Rosenstein et al. 2001). Meningococci can be categorized into 13 distinct serogroups based on the composition of their capsular polysaccharide, although just six of these cause more than 90% of invasive disease worldwide (A, B, C, W135, X and Y) (Schwartz et al. 1989; Rosenstein et al. 2001). In the industrialized countries of Europe and North America, where serogroups B and C predominate, the disease is endemic with a low overall incidence and characterized by seasonal peaks and small clusters of cases (Harrison et al. 2009). In contrast, in an area of the African Sahel and sub-Sahel (known as the 'meningitis belt'), large epidemics of meningococcal disease occur frequently on a scale unparalleled in other regions. Most African epidemics have been caused by meningococci belonging to serogroup A, but outbreaks of serogroup C, W135 and X disease have also been recorded (Greenwood 1999, 2006; Molesworth et al. 2002; Girard et al. 2006; Harrison et al. 2009). The epidemiology of meningococcal disease in other parts of the developing world, particularly Asia, is not well described (Manchanda et al. 2006; Harrison et al. 2009). We have, therefore, undertaken a literature review of meningococcal epidemiology in India, where a new serogroup A meningococcal conjugate vaccine is being produced by the Serum Institute of India, with support from the Meningitis Vaccine Project (Das 2004). This vaccine has now been pre-qualified by the World Health Organization, and mass vaccination campaigns will start in Burkina Faso, Mali and Niger in the last quarter of 2010. This vaccine might also find a use in India as epidemics of meningococcal disease have occurred in New Delhi in 2005 and Meghalaya and Tripura in 2008/2009 (Sinha 2009). The MEDLINE, EMBASE and Global Health databases were reviewed (with no date restrictions) for published studies using 'mening*' OR 'septic?emia' AND 'India' as search terms. The IMSEAR and MedIND databases were also examined using combinations of India AND one of 'meningococcus', 'meningococcal', 'meningitis', 'septicaemia', 'septicemia' or 'sepsis'. The reference lists of all included papers were reviewed for additional back references relevant to this review. The search was conducted in July 2009 and identified 1536 articles. Screening of abstracts identified 253 for possible inclusion, of which 214 full-text copies (84.6%) were obtained. The reference lists identified a further 23 papers for possible inclusion. After this process, 237 papers were formally reviewed for relevant epidemiological data using the inclusion and exclusion criteria given in Table 1, and data from 99 papers have contributed to this review. Two additional unpublished reports were obtained from the Indian National Institute of Communicable Disease (NICD). Meningococcal meningitis is a mandatory notifiable disease in India. But reporting remains incomplete and inconsistent, partly because it is not enforced and partly because of the complex private–public mix of Indian healthcare providers (IDSP 2008). The majority of data presented here is therefore from observational hospital-based reports and likely to present an incomplete picture (Verghese 1990). During non-epidemic periods, pyogenic meningitis [defined by cerebrospinal fluid (CSF) changes suggestive of bacterial infection] has been reported as the main diagnosis in up to 3.3% of acute admissions to tertiary paediatric hospitals in India (Achar & Rao 1953; Paul 1963; Ahmed et al. 1964; Srivastava et al. 1968; Gandhi 1969; Reddi et al. 1973; Tamaskar & Bhandari 1976; Nalwa 1978; Ayyagari et al. 1980; Venkatesh et al. 1985; Kabra et al. 1991; Chinchankar et al. 2002). In the largest study reported, which covered six paediatric departments (all from northern India) for the year 1989, 1.5% of hospital admissions (852/56 338) of children aged <12 years were diagnosed as cases of acute bacterial meningitis (Kabra et al. 1991). Quantifying the proportion of these cases of meningitis that were attributable to N. meningitidis infection (or to any other pathogen) is difficult because of the large proportion of patients (often with significant mortality) in whom no organism was identified. This low diagnostic yield is likely to be due to a combination of antibiotic use prior to culture, lack of 24-h laboratory facilities and practical difficulties in sampling and culture techniques (Venkatesh et al. 1985; Kalra 1989; Pandit et al. 2005; Manchanda et al. 2006). The microbiological results obtained in 33 case series of pyogenic meningitis are presented in Table 2. Neisseria meningitidis was a rare isolate (median detection rate 1.9%, range 0.0–20.0%), and some of the higher rates may represent the extension of reported or unreported epidemics (Achar & Rao 1953; Ahmed et al. 1964; Ayyagari et al. 1980; Bhat et al. 1991; : Bhaumik 1998; Chinchankar et al. 2002; Deivanayagam et al. 1993; Gandhi 1969; Hemalatha et al. 2002; Jain et al. 2000; Javadekar et al. 1997; John et al. 2001; Kabra et al. 1991; Kalra & Dayal 1977; Kumar et al. 1980; Mani et al. 2007; Minz et al. 2008; Pal & Sant 1982; Panjarathinam & Shah 1993; Paul 1963; Reddi et al. 1973; Sahai et al. 2001; Shameem et al. 2008; Shivaprakash et al. 2004; Singhi et al. 2002a,b, 2004; Srivastava et al. 1968; Suvarna Devi et al. 1982; Tamaskar & Bhandari 1976; Taneja & Ghai 1955; Venkatesh et al. 1985; Vincent et al. 1987). We found only one study which provided an adequate denominator to define the incidence of bacterial meningitis (Minz et al. 2008). This was a 2-year prospective study, conducted in southern India, whose primary aim was to estimate the burden of invasive Haemophilus influenzae type b disease in children aged 100 WBC/μl) and of culture positive meningitis of any cause in children under 5 years of age were 37 per 100 000 per year (95% CI 27–50) and 15.9 per 100 000 per year (95% CI 8–23) respectively. These figures are low by international standards and may not be representative of the whole country. We found few data on the incidence of septicaemia. One study from Chandigarh suggested that the burden of septicaemia was similar to that of meningitis; 1222 (2.8%) and 1126 (2.6%) of 43 800 emergency department attendances in 1995–2000 were attributable to meningitis or septicaemia respectively (Singhi et al. 2003). However, in seven case series of childhood or adult septicaemia, N. meningitidis was not isolated from over 6000 samples (Bhakoo et al. 1968; Somu et al. 1976; Das et al. 1978; Yardi et al. 1984; Pathare et al. 1988; Arora & Devi 2007; Garg et al. 2007). There is little information on the serogroups responsible for endemic meningococcal disease in India. Ayyagari et al. (1987) reported that between 1980 and 1984 only 12 strains of meningococcus were isolated at the Postgraduate Institute of Medical Education and Research in Chandigarh, of which eight were serogroup A and four serogroup C. We found only a single case report of proven group B meningococcal disease in India (Suri et al. 1994). The earliest reports of cerebrospinal fever (an early term for meningococcal meningitis, not necessarily excluding other aetiologies) in India were summarized by Patel 1926: 'The first authentic report of a case is by Vandyke Carter, who records four cases in Bombay where he was working at famine fever in 1878. Lieut.-Colonel Dimmock recognized the disease in epidemic form amongst the convicts of the Shikarpur Jail during the cold season of 1883–1884. Since 1881, cases have been reported all over India from various jails, army barracks, etc., but in Bombay proper cases began to arrive in municipal hospitals in the year 1912. They were diagnosed as cerebro-spinal fever from clinical signs only until 1921, when the present observations began.' (Patel 1926). Patel (1926) then reported 170 consecutive cases from Bombay (Mumbai) during the years 1921–1924. He described cerebrospinal fever as 'a disease of winter and spring', 'of overcrowding and poverty', and as 'affecting mainly adult males over the age of 20 years'– astute observations that mimic the epidemiology of documented meningococcal epidemics in India today. The first records of widespread epidemic disease among the general population appear in the 1930s. This epidemic peaked between February and June, 1934, with 2953 reported cases and 1793 deaths (Ahuja & Singh 1935). The major towns of northern India – Delhi, Kolkata and Ahmadabad – were particularly affected; though, widespread outbreaks were recorded (Sen et al. 1934; Tyrell & Kapur 1935; Vengsarkar 1947). Since the 1930s, India has experienced three documented periods of increased meningococcal disease (with cases always reported first in New Delhi), with an apparent periodicity of approximately 20 years, and with small isolated outbreaks in between (Figure 1, Table 3) (Verghese 1990; NICD 2009). The distribution of Indian epidemics over time and place. Data are adapted from the studies presented in Table 3. Where multiple reports of the same epidemic exist, the most complete figures were used. Data were included for each year when the number of cases was stated to be significantly higher than that for a normal year. Epidemics are shown proportional to size. The vertical red columns represent the periods of generalized epidemics. During these epidemic periods, hospital-based reports record a leap in both the number of admissions caused by suspected bacterial meningitis, and in the proportion of positive N. meningitidis isolations (rising to 40% or 50% of all suspected cases) (Pattanayak et al. 1967; Ghosh & Ghosh Roy 1970; Basu 1985). The majority of reports related to clinically suspected meningitis, but three reports (of small case-series) suggested that meningococcal septicaemia also occurred in significant numbers during these epidemics (Seghal et al. 1986; Paul et al. 1988; Saha et al. 2006). The largest Indian epidemic on record began in New Delhi in the winter of 1984 (Basu 1985, 1986a,b; Gulati & Mehta 1985; Prasad et al. 1985; Taneja 1985; Bhave & Gaikwad 1986; Seghal et al. 1986; Ayyagari et al. 1987; Deorari et al. 1987; Sarkar et al. 1987; Singh et al. 1987, 1988; Paul et al. 1988; Talukdar et al. 1988; Annapurna et al. 1989; Verghese 1990; Saha et al. 2006). Data from this epidemic are presented graphically in Figure 2. In 1985, at the peak of the epidemic, 6133 cases and 799 deaths were reported. These figures, though widely quoted, represent only those cases who presented to the seven major hospitals in New Delhi (Basu 1986a,b). The true scale of this epidemic in New Delhi, or indeed in the wider Indian community, is unknown. A conservative estimate of the magnitude of the epidemic, based on the proportion of the population covered by the seven sentinel hospitals, suggests that the figure for New Delhi could have been at least double the 6133 recorded11 In 1987, the population of New Delhi was approximately 6.8 million, and the bed-to-population ratio was 2.29/1000 (source http://delhiplanning.nic.in/Economic%20Survey/chapter_16.htm). The total number of beds at the seven hospitals (using 2009 figures) estimates the population coverage at 3.4 million. However, (i) the number of beds was likely to be considerably lower in 1985 than today (this would suggest that 6133 cases is an even smaller fraction of the true total), and(ii) this calculation assumes an even spread of cases throughout Delhi and that they did not preferentially attend the seven sentinel hospitals. . There are published reports of smaller outbreaks during 1985 from Chandigarh, Mumbai and Agra, and the NICD received reports of cases throughout the country in this year (Basu 1985, 1986a,b; Gulati & Mehta 1985; Kalra et al. 1985; Prasad 1985; Prasad et al. 1985; Taneja 1985; Bhave & Gaikwad 1986; Dubey et al. 1986; Ayyagari et al. 1987; Deorari et al. 1987; Sarkar et al. 1987; Sehgal 1987, 1989; Singh et al. 1987, 1988; Paul et al. 1988; Talukdar et al. 1988; Bhavsar et al. 1989; Verghese 1990; Vadher et al. 1991). The effect of season on reported pyogenic meningitis in New Delhi 1980–1989 and 1960–1969 (Basu 1986a; Pattanayak et al. 1967; NICD 2009; Verghese 1990). Meteorological data is from the New Delhi Safdarjun station and was accessed through NNDC Climatic Data Online at: http://cdo.ncdc.noaa.gov. An increase in the number of cases of meningococcal disease was again reported from New Delhi in the winter of 2004/2005 (with a peak of 867 reported cases in 2006), and more recently from Meghalaya and Tripura in 2008/2009 (Gupta 2005; Malhotra et al. 2005; Sachdeva et al. 2005; Manchanda & Bhalla 2006; Saha et al. 2006; Duggal et al. 2007; Mehta & Goyal 2007; Singhal et al. 2007; Kumar et al. 2008; Nair et al. 2009). The most complete data available for this period are taken from an unpublished NICD report (NICD 2009). The figures quoted for all the generalised epidemics lack the denominator necessary to define attack rates which would allow comparison between Indian epidemics, or with epidemics elsewhere. An unpublished report from New Delhi in early 2006 gives attack rates as high as six per 100 000 per week for specific city zones but without a clear description of how these data were collected or calculated (NICD 2006). The geographical distribution of reported outbreaks in India, shown in Figure 3, suggests that there has been relative sparing of central and southern areas with the majority of outbreaks being reported from the large cities of northern and coastal areas including New Delhi (population 18.6 million), Mumbai (21 million) and Kolkata (15.4 million). The distribution of Indian epidemics compared to climatic zones and population density. Climatic map courtesy of Saravask, based on work by Planemad and Nichalp (2007), accessed via http://en.wikipedia.org/wiki/file:India_climatic_zone_map_en.svg. Population density map courtesy of PlaneMad, accessed via http://en.wikipedia.org/wiki/file:India_population_density_map_en.svg. In a case series of 112 admissions with suspected meningococcal disease seen in Delhi in 1985, the majority of cases came from crowded areas of the city (81.6%), and from large families where many people shared a single room (59.8%) (Gulati & Mehta 1985). However, no control group is described, so it is impossible to say whether this pattern differed from the normal socioeconomic profile of this hospital's clientele. Other papers have observed similar links to overcrowding and poverty, but suffer the same methodological problems (Lal & Yacob 1935; Ahmad & Roy 1940; Vengsarkar 1947; Lal et al. 1963; Pattanayak et al. 1967; Bhalla et al. 1972; Basu 1986a; Talukdar et al. 1988; Annapurna et al. 1989; Bhavsar et al. 1989; Nagraj et al. 1998). In India, as elsewhere, the timing of epidemics of meningococcal disease seems inextricably linked to season (Greenwood et al. 1985; Molesworth et al. 2003; Harrison et al. 2009). The data presented in Figure 2 demonstrate the peaks of suspected meningococcal disease in New Delhi between February and April 1966, and in 1985, 1986 and 1987. In New Delhi, the epidemic period corresponds with the end of winter (November–March) and the beginning of summer (April–July). The rise in monthly case reports consistently begins during the driest period of the year, peaks before the temperature reaches its maximum and falls with the onset of the monsoon, usually in July or August. During the monsoon, the number of cases does not return to pre-epidemic levels, but remains moderately raised, implying continuing transmission, with a renewed epidemic occurring once the rains cease and climatic conditions again become permissive (Schwartz et al. 1989). All-cause bacterial meningitis in India is predominantly a disease of young children with a male preponderance. During epidemics, there is a marked shift in the age profile of cases of meningitis, with an increasing proportion of cases occurring in adults. This shift is best demonstrated by data collated from five hospitals in New Delhi between 1963 and 1966 (Figure 4) (Pattanayak et al. 1967). With the onset of the epidemic in 1966, the proportion of cases in the 10–25 age group increased from 18.4% to 30.4%. A similar shift in age profile was seen in the 1985 and 2005/2006 outbreaks (Basu 1985; Duggal et al. 2007). The demographic shift in pyogenic meningitis presentations associated with epidemics (Basu 1985; Duggal et al. 1985; Pattanayak et al. 1967). Group A meningococcus has caused all major, investigated Indian epidemics (Table 4) (Ahuja & Singh 1935; Bhatia et al. 1968; Prakash & Ghosh Ray 1968; Ghosh & Ghosh Roy 1970; Dubey et al. 1986; Deorari et al. 1987; Sarkar et al. 1987; Sehgal 1987; Singh et al. 1988; Talukdar et al. 1988; Annapurna et al. 1989; Bhavsar et al. 1989; Mirdha et al. 1991; Saha et al. 2006; Duggal et al. 2007; Kumar et al. 2008; Nair et al. 2009). However, group C meningococcus was also isolated in small numbers during the 1985 epidemic in Chandigarh (Dubey et al. 1986; Ayyagari et al. 1987). It is postulated that the 1985 New Delhi epidemic was an extension of an epidemic which began in Nepal in 1983 and spread via the 1987 Haj pilgrimage to Saudi Arabia and Chad (Moore et al. 1989). This was never proven as molecular typing of Indian isolates was not performed. Multi Locus Sequence Typing of isolated strains from the 2005 New Delhi epidemic is reported as showing similar strains to those responsible for epidemics in Bangladesh in 2002 and Nigeria in 2003 (Nair et al. 2009). Meningococcal carriage in Indian populations appears to be lower than that reported in other settings, even during epidemics, and without obvious age or sex predilections (Table 5). During the 1985 epidemic in New Delhi, the prevalence of carriage in healthy subjects aged 6–20 years was below 2%, consistent with the level of carriage seen in new army recruits and new prisoners (proxy indicators for the general population) in older studies (Ardeshir 1935; Lal & Yacob1935; Ichhpujani et al. 1990). Higher carriage rates have been found in close family contacts of cases (Bhalla et al. 1972; Dubey et al. 1986; Ichhpujani et al. 1986; Paul et al. 1987; Nagraj et al. 1998). These findings suggest that targeted chemoprophylaxis for household and other close contacts could be useful. The limited evidence available suggests that the background incidence of sporadic meningococcal disease in India is low, and almost certainly less than that owing to Streptococcus pneumoniae and H. influenzae type b infections. Routine childhood vaccination with meningococcal vaccines is, therefore, unlikely to be a priority. However, there is strong evidence that India (especially the northern region) is susceptible to epidemic meningococcal disease. Several major epidemics have been reported, pre-dominantly from the major cities, and particularly from New Delhi. This distribution may be real, because of overcrowding or vulnerability to the importation of new strains, or to suitable climatic conditions. On the other hand, it is possible that there may be a form of publication bias, with epidemics elsewhere going unreported. However, had outbreaks occurred in southern cities with many reputed healthcare institutions, it seems unlikely that they would have gone unnoticed or unreported. Group A meningococcus appears to be the pre-dominant cause of meningococcal epidemics in India, as it is in the meningitis belt of sub-Sahelian Africa (Schwartz et al. 1989; Girard et al. 2006; Harrison et al. 2009). The pattern of epidemic meningococcal disease in India shows many similarities to that of epidemic meningococcal disease in Africa, although epidemics appear to have been much smaller than the majority of those reported in Africa. Some suggestive evidence was found for an epidemic cycle in New Delhi, with epidemics occurring every 15–20 years. This is less frequent than the 5- to 10-year cycle generally seen in the meningitis belt of Africa, but similar to the pattern that has been observed in African countries on the fringes of the belt (Molesworth et al. 2002; Cuevas et al. 2007). A strong seasonal effect was observed, with epidemics peaking during the dry period of the year and decreasing but not disappearing during the monsoon before returning the following dry season. This seasonal cycle is similar to that seen in Africa where outbreaks peak during the hot dry season and subside during the rainy season to reappear, on some occasions, during the following dry season (Greenwood et al. 1985; Schwartz et al. 1989; Greenwood 1999; Molesworth et al. 2003). The mechanism for this seasonal association is not known, but suggested theories include damage to the natural mucosal barrier of the naso-pharynx by the drying effect of prolonged low humidity (increasing the chance of invasion), an increased incidence of viral infections during the dry period of the year, and behavioural changes associated with season (Schwartz et al. 1989). Some evidence was found to support the view that Indian epidemics are more likely to occur (or at least be reported), in the drier northern parts of the country than in the more humid south. This is also observed in the African countries which span the dry (and meningitis prone) Sahelian areas to southern forested regions where epidemics are less common. During Indian epidemics, the average age of cases of bacterial meningitis increased. This is not unexpected as meningococcal disease is known to typically affect age groups older than either S. pneumoniae or H. influenzae type b. An almost identical age distribution is characteristic of African epidemics (Greenwood 2006). This finding may also represent the result of waning levels of herd immunity following 20 years of low transmission or widespread susceptibility to a new and immunologically distinct strain. We found very little information on the response by health authorities to meningococcal epidemics in India. In 1985, vaccination was reportedly provided for hospital personnel at high risk of infection, and in 2005 chemoprophylaxis was given to nearly 3000 close contacts of proven meningococcal cases (Basu 1986b; NICD 2006). In 2009, a mass vaccination campaign was planned for the affected area in Tripura using polysaccharide vaccine (Sinha 2009). Mass vaccination is the approach endorsed by the WHO for epidemic control, once set thresholds have been exceeded, but the relative success of this approach is critically dependent on the speed of the response (WHO 1997, 2000, Woods et al. 2000). The establishment of sensitive and specific Indian epidemic thresholds would require the collection and analysis of reliable longitudinal surveillance data. A full evaluation of the Tripura vaccination campaign would be highly informative for future epidemic control. Our review has established that only limited data are available on the epidemiology of meningococcal infections in India. Based on the limited data available, it appears that meningococcal disease in India is characterized by a low background incidence with occasional, but large, epidemics. In the absence of adequate surveillance for meningococcal disease, many additional smaller outbreaks are likely to have gone unreported, and the size of reported ones to have been underestimated. Establishment of functional disease surveillance with adequate laboratory support throughout India would facilitate the detection of epidemics of meningococcal disease and those of other infectious diseases. Vaccination appears to have been little used in the control of epidemics of meningococcal disease in India. The protection following polysaccharide vaccination is short-lived, especially in children. The development of meningococcal conjugate vaccines, including a new highly immunogenic group A vaccine developed in India, provides new tools that could be used to contain future epidemics of group A meningococcal disease in India provided that these are detected early through improved surveillance. The authors thank Lionel Martellet and Monique Berlier from PATH, and Tomas Allen, Katya Fernandez-Vegas and William Perea from WHO for their advice, support and help in acquiring relevant evidence and articles. The authors alone are responsible for the views expressed in this publication, and they do not necessarily represent the decisions, policy or views of the SACC, CMC, WHO or LSHTM.