Indian Pediatrics 2003; 40:933-938 

Investigation of Outbreaks in India. How good are we at it?

The nation periodically goes through outbreaks of mostly known and some unknown diseases. A few outbreaks are investigated, fewer still are published in the medical press and investigation reports never reach the public domain. In the absence of peer reviewed medical articles, the public is left with the speculations of the press, which invariably creates panic. Failure to publish (as evidenced by the paucity of publications from State Public Health service providers) may be due to censorship by higher authorities, lack of interest to publish, or an inability to publish since most journals will not accept one more outbreak. Among the recent prominent outbreaks that have caught public attention nationally are the ‘plague’ outbreak in Surat (1994), the ‘mystery killer disease’ of Siliguri (2001) and the more recent ‘killer brain disease’ in Andhra Pradesh and Maharashtra (2003).

Public health actions/inactions, interven-tions or its failures, invariably invite criticism from fellow professionals and the public. However, in keeping alive our professional-ism, we have a commitment to examine ever so often, how well we have performed in handling our outbreaks, what ails the system and the way we should go forward.

Encephalitis or encephalopathy?

A recent outbreak that has triggered articles recently in Indian Pediatrics is the Andhra Pradesh and Maharashtra outbreak referred to in the World Health Organization’s Outbreak Verification List, "Unknown, India, Location: Andhra Pradesh, Maharashtra, Onset: 02-07-2003, update: 10-09-2003", reads thus: "329 cases including 183 deaths in Andhra Pradesh, 287 cases including 115 deaths reported from Maharashtra. In Andhra Pradesh there has been only 1 death reported since 20th August. The number of new cases for Andhra Pradesh continues to decline(1).

In the last issue of Indian Pediatrics Dr. Jacob John, compared the differences between encephalitis and encephalopathy because of professional disagreement on the etiology of the Andhra Pradesh outbreak(2). His article deals with the clinical picture and pathological findings of both conditions and retraces the origins of the terminologies in India and urges the Indian Academy of Pediatrics to educate its members about Reye's syndrome.

As a nation, we are used to outbreaks of Japanese encephalitis (JE) in different parts of the country(3-8). Between 1994 and 1999, about 1200 to 2500 cases of JE were reported annually to the national Government and about 25 to 30% of these cases died(9). Familiarity with JE seems to have heightened our index of suspicion and one tends to label CNS related outbreaks in the community as probably JE, until proven otherwise.

Isolation of a virus from cerebrospinal fluid (CSF) is often implicated in the causation of disease, rather than being considered as incidental (since CSF is never tested in healthy normal individuals for comparison). As early as 1953, Bang and Bang(10) identified CSF pleocytosis in patients presenting with parotitis but without central nervous system (CNS) symptoms during an outbreak of mumps, showing that CNS invasion by mumps virus in the CSF need not necessarily cause CNS symptoms (they did not need control samples to prove their point). Recent studies in India have implicated measles in outbreaks with CNS symptoms. Some have shown measles virus in the CSF in a small proportion of cases, but the majority of evidence has relied solely on serology from cases, without the benefit of comparative serology from controls and now there is a danger that measles may soon be another common label that is used, just as with JE.

Wairagkar(11) and Ghosh(12) reported the results of serology (and a few CSF measles virus isolations) among children presenting with acute encephalopathy from Punjab, Haryana and Uttar Pradesh. The former study implicated measles in the causation of acute encephalopathy and the latter implicated measles, varicella and salicylates in the causation of Reye’s syndrome. The former study was hospital based while the latter was a case-control community based study. Neither study considered the need for control sera to establish background rates for IgM/IgG positivity and hence had no comparative figures to establish an association. The findings in both studies do not justify the conclusions reached by the authors. Vashishtha(13) reported, an ‘epidemic’ of acute encephalopathy in ‘most of the towns of Western Uttar Pradesh’ in 2002. Measles has again been implicated in triggering Reye’s syndrome. Once again, this speculation is not supported with evidence from control samples. The authors did identify the lack of laboratory facilities as an important obstacle in investigation of the outbreak without first establishing an association, one cannot start thinking of casuality.

There now appears to be reasonable confusion about the cause and nature of these recent outbreaks. Both public and medical professionals have been given conflicting views and the lack of a systematic approach to outbreak investigations and its interpretation is self-evident.

What is the need to investigate an outbreak?

The most compelling reasons are to pre-vent additional cases (since exposure to source of infection may be continuing) by identifying the source. Additionally, an epidemiological (and environmental) investigation can increase our knowledge of the disease and prevent future outbreaks. More academically, outbreak investigation can provide epidemio-logical training and foster cooperation between clinical and public health communi-ties and recommendation of strategies may prevent future outbreaks(14) .

What types of investigations are needed?

There are well-defined systematic approaches to investigating an outbreak that most textbooks would offer. If these are not followed, the uncertainness we have encountered with current outbreaks inevitably follows. One or more hypotheses are generated after an exploratory visit. A broad based questionnaire (sometimes called a trawling questionnaire) is used to capture a breadth of information from cases. Commonly, pre-tested questionnaires are available for use and trained persons in the team (and those in training); carry out the investigation using standard procedures. Hypotheses generated from the preliminary results are tested using analytical studies. Analytical studies by virtue of their design will include the selection of controls using standard methodologies. Specimen collection is concurrent and is often done using the existing health delivery system. Its results are interpreted along with the results of the analytical study. It is often prudent to look beyond the numbers collected and look at the magnitude of exposure and disease and the biological plausibility while interpreting results.

What are the special skills needed?

The science of epidemiology, communica-tion skills with the public, the press and an overview of legal implications and public health laws are the most obvious requirements. The skill to resist pressure to conclude investigation quickly and to avoid taking hasty decisions is hardly ever taught in any training program. Critical appraisals of journal articles, particularly in interpretation of results are as important as skills in designing and carrying out investigations.

Associations are often interpreted as causality. A commonly used set of standards advanced by Bradford Hill(15) helps to dis-tinguish causal from non-causal associations. Some tend to use these as a checklist for inference although this was not what they were meant for(16). The Institute of Medicine has recommended a causality scale that helps to decide if there is sufficient evidence for causality(17) – (i) No evidence bearing on causality; (ii) Evidence insufficient to indicate a causal relation; (iii) Evidence does not indicate a causal relation; (iv) Evidence is consistent with a causal relation; (v) Evidence indicates a causal relation.

Are there major problems with outbreak investigations in India?

Investigations are not well coordinated (delay in the report reaching, delay in start up, travel, inadequate skills in the local regions and inadequate laboratory support and perhaps little or no skills in writing the report). Investigatory bodies are often not adequately respected for their expertise or pro-fessionalism, partly because professional disagreement spills out into the press. The mysterious disease in Siliguri (February 2001) caused major panic, resulting in the public and health personnel leaving Siliguri(18,19). Even today, all one has to speculate with was what the press had reported as ‘localised viral epidemic’ or Hantavirus Pulmonary Syndrome; 36 of 59 cases are reported to have died, many were health care personnel who attended sick patients(9). If this was indeed caused by Hantavirus (for which we have not seen any evidence), this would have been the first time this infection had been identified in India. Appropriate education and preventive measures could then have followed.

The 1994 ‘plague’ outbreak in Surat has as many versions as the teams that investigated it. ‘Plague that never was’(20) was one version that came too late, four years after the damage was done and national pride was hurt.

When a new disease emerges, the discrepancy between clinical and laboratory is discernable only by review of detailed clinical findings(18). Detailed findings are the product of diligent search and data collection early enough in the outbreak. The SARS outbreak is an obvious example of how concerted international effort resulted in very rapid identification of this emerging disease. To quote Knudsen(21), "the concerted efforts of a globally united scientific community have led to the independent isolation and identification of a novel coronavirus from SARS patients by several groups. The extraordinarily rapid isolation of a causative agent of this newly emerged infectious disease constitutes an unprecedented scientific achievement.

Who should be investigating?

The National Institute of Communicable Diseases (NICD), under the Ministry of Health and Family welfare is the national body established for this function. Centrally funded laboratories are expected to support the work of NICD to establish etiology. The Table summarizes outbreaks investigated by NICD during 2001(22).

TABLE

Investigation of Outbreaks during 2001

Month in	Type of outbreak and location	Period	Cases (deaths)
Jan	Public Health Situation, Bhuj, Gujarat (post-earthquake)	28 Jan
Feb	Disease Surveillance Cell, Bhuj, Gujarat	10 Feb - 3 Mar
	Encephalitis Siliguri, W Bengal	22 Feb - 2 Mar	66 (45)
March	Viral Hepatitis E, Palam, Delhi	16 Mar	32 (1)
	Disease Surveillance Cell, Bhuj, Gujarat	17 Mar
April	Acute Febrile Illness Siliguri, W. Bengal	1 - 4 Apr
	Viral Hepatitis E, Super Bazaar, N. Delhi	20 Apr	50-70 (1)
May	Diarrhea, Faridabad, Haryana	29 May	50
June	Investigation Of Reported Cholera Deaths, Ghaziabad, UP	14 Jun	512 (5)
July	Rapid Health Assessment, Orissa Flood	21-24 Jul
August	Flood Situation and Diseases, Orissa	29 Jul - 2 Aug
	Fever 22 Bn, Jalandhar, Punjab	10-11 Aug	126 (1)
	Viral Hepatitis B, Saharanpur, UP	6-11 Aug	52
September	Dengue Fever, Alwar, Rajasthan	Aug-Sep
	Japanese Encephalitis, Gorakhpur, UP	02-08 Sep	76 (17)
	Kala Azar, Trichur, Kerala	13-16 Sep	2
October	Dengue/Dengue Hemorrhagic Fever, Rajasthan	06-09 Oct	1820 (30)
November	No investigation was carried out
December	No investigation was carried out

 

Obligatory versus elective responsibility and role conflicts

National Institutes are centrally funded and are expected to serve as reference centers by virtue of their position, coordinating surveillance, investigation of reports/ out-breaks, providing professional advice, training, carrying out research and reassuring to the public. Given the numbers and locations where NICD was involved in fulfilling its role in 2001, there is much scope for establishment of many more decentralized centrally funded expert-led centers. These centers would no doubt require a reporting network that provides information sufficiently early, as well as adequate laboratory backup.

State funded bodies that provide health care also have responsibilities to investigate outbreaks in order to undertake control measures. Their service delivery role often conflicts with an investigatory role result- ing in suppression of information locally (personal experience as an NGO working with the local Government over 25 years are filled with examples of this conflict of roles).

Successful international models in developed countries provide a service that could be accessed by both government and private agencies. Epidemiologists discuss reports that originate from calls or through other electronic means, prioritize its status, develop rapid action plans and if needed, carry out investigations. All investigations end with a report accessible to fellow professionals if needed. Annual reports detailing the numbers and nature of investigations done are made available in the public domain. Laboratory support of high quality is mostly assured in these situations. Multi-specialty ‘outbreak control meetings’ are the norm. Generic investigation plans and questionnaires are restructured to suit current needs. Often, a ‘trawling questionnaire’ is used to study a sample of cases, followed by an analytical study design (most often a case-control design) to bring out risk factors. Laboratory work up is often easier, since virtually all cases that are ill are hospitalized and mandatory investigations are the rule. Special tests may follow, depending both on the findings of the questionnaire survey and the basic laboratory work up. An important feature of the entire exercise is that there is a ‘team’ that works towards a common goal, mostly, unmindful of the hours of work put in, or who takes the credit for it. People with poor communication skills would probably have never made it into this team. The team has its press releases ready and only the designated press person talks to the press. Help lines are publicized and manned by adequately briefed health care personnel. Non-government organizations (NGOs) often elect to work with the government organization, pursuing common goals.

Could the Government use the NGOs more effectively?

India has a number of non-government academic and research centers with a good reputation that could be supportive of a Government led initiative. Non-government organizations with sufficient skills are currently used for training Government health staff and for carrying out evaluations of special projects on behalf of the Government. Often, these tasks are contracted out because it is a requirement by funding agencies. Using an unbiased third party to provide an evaluation report is seen as a fairer way to assess work done. Role conflicts are often eliminated by this method. Outbreak investigations should be reviewed in the same light. Public confidence is determined by the perceived validity of statements made, competence of investigators and the track record of the organization.

This editorial

This editorial is meant to highlight that all is not well with the existing system of managing outbreaks in India. So how could the investigation of the recent outbreaks in India have been handled more effectively? I asked my teaching colleagues on what they found lacking in the outbreaks the nation has witnessed in the past few years. I summarize their comments in the following paragraphs.

"The lack of properly coordinated epidemiological investigation with detailed analytical studies, using robust methodologies that followed preliminary investigation and adequate laboratory results to support claims, has been sorely lacking in the country. Most information about outbreaks is now read about only in newspapers and magazines and not in peer reviewed medical journals. This makes one suspect that there has either been a cover up, or that the investigation report was so shoddy that no journal would accept it".

Pretending that problems do not exist or hoping they would go away or censoring information is unlikely to work in future since the nation is no longer isolated from the world. SARS and China are a good example of why we need to be proactive in our approach. We have to earn credibility in international circles as a nation that is in control, and also learn to use international expertise when it is offered.

Funding: None.
Competing interests: None.

V. Balraj,
Professor in Community Medicine
Community Health & Training Center,
Christian Medical College,
Vellore 632 002, India.
E-mail: [email protected]

 References