Annex 4: Application of Participatory Methods to Rinderpest epidemiology

Overview of Rinderpest Surveillance Programmes
Principals of Active Rinderpest Surveillance
Rinderpest Participatory Disease Search Methodology
Participatory Epidemiological Intelligence and Serosurveillance.

Rinderpest is a disease that can be investigated using participatory methods. The occurrence of classic rinderpest is a clinically recognisable and dramatic event that is sure to be recorded in the community memory. For instance, traditional Arab pastoralists in Chad described rinderpest as the most important disease of cattle (Mariner, 1995) despite the fact that it had not occurred in their herds for over 14 years. Rinderpest is a disease for which most pastoral communities have specific terminology and regard with a high level of concern. Traditional rinderpest terms from a number of communities in East and Central Africa are presented in Table 1.

Table 1: Traditional Terms for Rinderpest

Community	Term	Source
Afar	dugahabe	Mariner and others, 1995
Amhara	desta
Somali (Gedo Region)	shifow, madobaye, daadi	Mariner and Flanagan, 1996
Somali (Somaliland)	dabakaruub	Catley and Mohamed, 1995
Gabra	daadi	Mariner and Flanagan, 1996;ITDG and IIRR, 1996
Toposa	ata na choke	Mariner, unpublished report
Turkana	loikyo	Ohta, 1984
Karamajong	loleoo	Akabwai and others, 1994
Pokot	molmoloy, kiplok	Mariner and others, 1995; ITDG and IIRR, 1996
Fulani (Bamenda, Cameroon)	petu	Akabwai and others,1994
Messeriya pastoralist (Chad)	djediri	Mariner, 1994
Kikuyu	gatema ka ng’ombe, gacenge	ITDG and IIRR, 1996
Iteso (Kenya)	edeke	ITDG and IIRR, 1996
Kipsigis	kipkeita	ITDG and IIRR, 1996
Luidakho	shipakuli	ITDG and IIRR, 1996
Luo	nidira	ITDG and IIRR, 1996
Maasai	lodutua, olodoa	ITDG and IIRR, 1996
Maragoli	kibakuli, nyabora	ITDG and IIRR, 1996
Nandi	kip kaita	ITDG and IIRR, 1996
Samburu	lodua	ITDG and IIRR, 1996
Swahili	sotoka	ITDG and IIRR, 1996

As was mentioned in the introduction, traditional disease terminology often has literal meanings that refer to the symptoms or effects of the disease. For example, dugahabe means empty coral in Afar, and petu refers to cluster of fruit falling from a tree limb in the wind. Both terms are suggestive of the heavy mortality experienced in classic rinderpest. Shifow and loikyo are terms that refer to tearing. The Amharic term for rinderpest, desta, is striking. The general meaning of desta is happiness. The Amhara believe that one should not speak badly of undesirable events.

At the present time, mild rinderpest is still a major concern in East Africa. Mild rinderpest presents greater diagnostic difficulty for pastoralists, as it does for veterinarians. Several pastoral groups have been astute observers of mild disease over the decades and can provide significant epidemiological intelligence (Plowright 1998; Mariner and Flanagan, 1996).

There are several diseases of cattle that resemble rinderpest. Bovine viral diarrhoea, infectious bovine rhinotracheitis and malignant catarrhal fever are three major differential diagnoses. However, it is clear that the terms listed in Table 1 are meant to specifically refer to rinderpest. Where MCF is common, the disease has its own specific name. It may occur that individual herders are mistaken in the use of rinderpest terminology when referring to specific clinical cases or events in the field, especially where mild rinderpest is suspected. To a large extent, the processes of triangulation, probing and epidemiological analysis of results will result in the discounting of individual misdiagnosis. Further, clinical surveillance and disease searching in rinderpest surveillance programmes track clinical stomatitis-enteritis outbreaks rather than rinderpest specifically. This is to assure that a maximum number of rinderpest compatible events are detected and investigated.

Given the high level of community interest and existing veterinary knowledge concerning rinderpest, it is appropriate that rinderpest surveillance programmes make use of participatory epidemiology as a complementary tool to more conventional forms of epidemiology. In this Chapter, rinderpest surveillance strategies will be briefly reviewed with a view to determining where participatory enquiry can best contribute. Thereafter, specific approaches will be suggested on how to integrate participatory techniques within on-going surveillance programmes.

Overview of Rinderpest Surveillance Programmes

The Six Components of Comprehensive Rinderpest Surveillance
The Role of Participatory Epidemiology in Rinderpest Surveillance

The Six Components of Comprehensive Rinderpest Surveillance

The Office International des Epizooties have published Recommended Standards for Epidemiological Surveillance Systems for Rinderpest (1998). This document is commonly known as the ‘OIE Pathway’ to freedom from rinderpest. The document outlines a series surveillance standards for the epidemiological classification of countries into three categories on the road to freedom from rinderpest infection. These categories are provisional freedom from rinderpest, freedom from rinderpest disease and freedom from rinderpest infection. This last category represents freedom from all forms of rinderpest disease as well as freedom from the presence of the virus. This pathway provides that countries must ‘investigate all clinical evidence suggestive of rinderpest,’ have ‘an effective veterinary service,’ conduct two years of randomised serosurveillance and investigate ‘into infection in wild susceptible species... where these species occur in significant numbers’ in order to be recognised as definitively free of rinderpest infection. Thus, the pathway requires a comprehensive surveillance system be in place that would detect rinderpest if it were present. Comprehensive rinderpest surveillance programmes are systems made up of six main component activities. A brief description of each component follows. For a more detailed overview of rinderpest surveillance one is referred to Mariner among others (1998).

General Disease Surveillance and Reporting

General disease surveillance and reporting is a routine national disease reporting system which collates monthly reports on disease occurrence on all significant diseases. Normally, general disease reporting systems cover all OIE List A and List B diseases as well as any diseases of local significance. General disease surveillance and reporting is sometimes referred to as passive surveillance as the information largely consists of clinical diagnosis made by field veterinarians in the routine exercise of their functions. Where available, laboratory and meat inspection data should be incorporated into the general disease surveillance and reporting system. General disease surveillance and reporting is an integral part of ‘an effective veterinary service able to monitor the animal health situation’ and therefore is a requirement for the provisional freedom from rinderpest disease as defined by the OIE Pathway.

Active Disease Surveillance

The objective of active disease surveillance is to detect a high percentage of field events that are clinically compatible with a diagnosis of rinderpest for epidemiological and laboratory-based investigation. Active rinderpest disease surveillance is intended to detect the occurrence of infectious clinical syndromes associated with stomatitis, enteritis, ocular and nasal discharge, corneal opacity and death, or any subset of the above. The adjective active signifies that this surveillance system initiates and maintains field level surveillance with the sole purpose of detecting clinical syndromes of stomatitis-enteritis. Active disease surveillance is not a search for rinderpest disease; it is a search for clinical syndromes including stomatitis and enteritis (see Box 3). If active disease surveillance is functioning correctly, outbreaks of bovine viral diarrhoea, infectious bovine rhinotracheitis and malignant catarrhal fever should be detected, sampled and reported.

Active disease surveillance includes techniques such as report registries, questionnaire surveys, participatory epidemiology, and clinical surveillance. In questionnaire surveys and participatory epidemiology, the livestock owner becomes a key link in the surveillance system.

Stomatitis-Enteritis Outbreak Investigation

The objective of outbreak investigation is to collect descriptive field data and diagnostic samples on a high percentage of infectious stomatitis-enteritis outbreaks within a reasonable period of time from the initial report or recognition. This means that information and samples should be collected that would facilitate either a diagnosis of rinderpest or a definitive differential diagnosis (i.e. bovine viral diarrhoea (BVD), infectious bovine rhinotracheitis (IBR), malignant catarrhal fever (MCF), etc.).

Laboratory Confirmation

The objective of laboratory confirmation is to provide a definitive diagnosis on a high percentage of field outbreaks of stomatitis-enteritis within an appropriate period of time. The results of initial rinderpest tests should be available within a period of days. If rinderpest results are negative, laboratory investigation should continue until a differential diagnosis is confirmed in as many cases as possible.

Serosurveillance (Virus surveillance)

The objective of rinderpest serosurveillance is to detect the presence or confirm the absence of rinderpest virus in a population. Serosurveillance is a sensitive and relatively specific technique. Current tests do not allow one to distinguish between antibody responses to wild and vaccine viruses. This limits specificity of this component of the surveillance system in vaccinated populations. Dis-aggregated seroprevalence data combined and correlated with descriptive data from other surveillance activities and vaccination statistics is useful for understanding disease prevalence in endemic or recently vaccinated populations.

In the final stage of the OIE Pathway, freedom-from-rinderpest, serosurveillance will be indispensable to confidently establishing the absence of rinderpest virus and the total cessation of vaccination. A statistically valid serosurveillance programme is the key requirement of the OIE Pathway for a country to be recognised as definitively free from rinderpest.

Wildlife Surveillance

Where populations of highly and moderately susceptible wildlife coexist with vaccinated cattle populations, wildlife populations should be used as non-vaccinated sentinel populations in serosurveillance programs. Wildlife surveillance is a requirement of the OIE for certification of freedom-from-rinderpest.

The Role of Participatory Epidemiology in Rinderpest Surveillance

Participatory epidemiology has a role to play, directly or indirectly, in all components of a rinderpest surveillance programme.

The contribution which community-based projects and CAHW networks can make to general disease surveillance was described in Chapters 2 and 3.

Participatory disease searching has direct relevance to active disease search as well as outbreak investigation. A general approach to disease searching was presented in Chapter 4. Rinderpest disease searching will be discussed in detail in this chapter. Participatory disease searches and participatory epidemiology make indirect contributions to laboratory confirmation of SE outbreaks by providing clinical data and epidemiological intelligence which help orient the testing process. Participatory disease search will also assure that there is sufficient SE outbreak sample material for laboratory testing.

Participatory investigations are of direct relevance to serosurveillance. Accurate, decision-oriented, qualitative information is needed on husbandry systems, cattle movements and historical rinderpest prevalence in order to design sound surveys and to assist in the interpretation of serological results in populations where vaccination has been discontinued. The use of participatory epidemiology in relation to serology will be discussed at the end of this chapter.

Finally, the observations of farmers relative to wildlife mortality are key intelligence for wildlife epidemiologists. This has been known for decades. Wildlife surveillance is a specialised area that will not be covered in the present document.

Principals of Active Rinderpest Surveillance

Prevalence of stomatitis-enteritis agents
Clinical syndrome-based surveillance
Stomatitis-Enteritis Clinical Outbreak Definition
Stomatitis-Enteritis Outbreak Reporting and Performance Indicators
Participatory Epidemiology and Clinical Syndrome-Based Surveillance

Active disease surveillance is a dynamic and evolving activity. In recent years, several strategic improvements in active disease surveillance programmes have been formulated and are under implementation.

Prevalence of stomatitis-enteritis agents

It is accepted that a number of the stomatitis-enteritis agents are widely distributed throughout the world, if not ubiquitous. Among the differential diagnosis for rinderpest are bovine viral diarrhoea, infectious bovine rhinotracheitis, malignant catarrhal fever and East Coast fever. Occasional cases of other diseases such as foot and mouth disease may also present clinical similarities. Thus, outbreaks of disease that resemble rinderpest are occurring in almost all countries each year. For example, the United States and Australia carryout exotic disease investigations in response to clinical stomatitis-enteritis cases in order to rule out possible introductions of rinderpest. Such investigations are part of routine, functional disease surveillance and emergency preparedness. This leads one to the conclusion that all countries with substantial cattle populations and functional surveillance programmes should be detecting and investigating rinderpest compatible events each year.

Unfortunately, the status quo for rinderpest surveillance in many countries has been national statistics that indicate no rinderpest suspect events detected and no rinderpest investigations carried-out. As rinderpest compatible events appear to be ubiquitous, this suggests the absence of surveillance rather than the absence of disease. Perhaps not surprisingly, rinderpest has a propensity to reappear after a period of ‘apparent disease freedom’ in many countries that take this approach to surveillance.

If rinderpest eradication is to be successful at the national, regional and world-wide levels, active surveillance in infected and apparently disease free countries will need to detect the majority of rinderpest compatible disease outbreaks.

Clinical syndrome-based surveillance

Although co-ordinated, multilateral disease eradication efforts in the veterinary world are mainly limited to rinderpest at the world-wide level, a vast experience is available from human health programmes. It is significant that many programmes, including national and regional FMD eradication programmes, survey for a clinical syndrome rather than the target disease per say. Some examples are presented in Table 2. In all cases, several disease agents are capable of provoking the clinical syndrome. However, cases and outbreaks must be detected and differentiated from the target disease.

Table 2: Examples of surveillance targets in major eradication campaigns

Disease	Clinical syndrome
Small pox	Rash with fever
Polio	Acute flaccid paralysis
FMD	Vesicular disease
Measles	Fever with generalised rash and cough, coryza or conjunctivitis

Sources: WHO, 1979 and 1997; CDC, 1997

‘After it appeared that smallpox transmission has (sic) been interrupted in Somalia by the end of 1977, it became of even greater importance that every suspected case and, in fact all cases of rash-with-fever, should be detected, reported and carefully investigated. Correct diagnosis of rash-with-fever cases became very important, as misdiagnosis of mild smallpox cases, caused by variola minor, and consequent failure to take containment action could have caused a serious setback in the progress towards global smallpox eradication. From January 1978, a new element of surveillance was developed: rash-with-fever surveillance.’

‘Every outbreak of chickenpox, the majority of outbreaks diagnosed as due to measles and a random sample of persons suffering from “other skin diseases” were revisited by regional epidemiologists or WHO advisers who verified the diagnosis and entered the findings in the rumour register (WHO, 1979).’

The Global Rinderpest Eradication Programme has recently adopted clinical syndrome-based surveillance based on the search for stomatitis-enteritis.

Stomatitis-Enteritis Clinical Outbreak Definition

The Global Rinderpest Eradication Programme has defined a stomatitis-enteritis outbreak definition (Box 3). It is important to emphasis that the definition is not a rinderpest definition.

Box 3: Stomatitis-Enteritis Clinical Outbreak Definition Ocular Discharge Nasal Discharge And Fever Oral Erosions/Lesions Salivation Corneal Opacity Diarrhoea Death Outbreaks of contagious disease exhibiting discharges (ocular and nasal) and any two of the other above-mentioned symptoms should be reported as stomatitis-enteritis outbreaks. Note that it is the outbreak that must meet the criteria, not individual animals.

The stomatitis-enteritis outbreak definition is applied to the aggregate of cases in an outbreak. Ocular and nasal discharges are mandatory signs. Fever, oral lesions, salivation, corneal opacity, diarrhoea and death are variable signs. If an outbreak presents with ocular and nasal discharge and any two of the variable signs then it is considered a stomatitis - enteritis outbreak for reporting and investigation purposes.

As an example, an outbreak where discharges, oral lesions and one death were noted would qualify as a stomatitis-enteritis outbreak. Discharges and fever alone would not qualify. In such cases, a careful search of the herd, including the examination of mouths of all animals with discharges, should be made for evidence of other variable signs.

It should be clear that the stomatitis-enteritis outbreak definition is a broad net. The definition encompasses all forms of rinderpest and BVD as well as many clinical presentations of IBR, MCF. Occasionally other diseases such as FMD and bluetongue may meet the definition.

Stomatitis-Enteritis Outbreak Reporting and Performance Indicators

The Global Rinderpest Eradication Programme and PARC are now in the process of introducing performance indicators for the monitoring of rinderpest surveillance activities at the national level. Application of the stomatitis-enteritis outbreak definition will lead to the detection and reporting of SE outbreaks in most countries regardless of their rinderpest status. This is expected. In fact, the failure to detect, report and investigate SE outbreaks annually will not be received as evidence that the country is disease free. Failure to detect SE outbreaks will indicate that the national rinderpest surveillance programme is non-functional and that the rinderpest status of the country may not be confirmed by objective measures (Mariner among others, 1998).

Participatory Epidemiology and Clinical Syndrome-Based Surveillance

Concerning participatory epidemiology and syndrome-based surveillance, the following points are evident:

· Most, if not all, traditional terms for rinderpest meet the stomatitis-enteritis outbreak definition. Thus, herder reports of rinderpest should enter the surveillance pathway as stomatitis-enteritis reports and be investigated.

· Livestock owner reports, investigated or not, should be collated and analysed epidemiologically as part of the national surveillance database.

· Participatory disease searches are a form of clinical disease search for historical and current stomatitis-enteritis outbreaks. They are analogous to the rash-with-fever searches carried-out in the final stages of smallpox eradication.

· Participatory disease searches will enhance the sensitivity of the overall surveillance system and contribute to a positive assessment of surveillance performance.

Rinderpest Participatory Disease Search Methodology

Preparations
Interviewing
Topics for Probing
Mapping
Clinical Observation, Sample Collection and Transects
Analysis of Results

This section will describe PDS as it directly relates to rinderpest. The specific techniques that are most useful are open-ended questions, probing, time line construction, triangulation (cross-checking of reports), key informants (community leaders, traditional healers, veterinary staff, expatriate technical assistants and international campaign co-ordination offices), mapping as well as direct clinical observation and transects. Basic techniques of participatory epidemiology and PDS together with concerns regarding analysis and interpretation were discussed in Chapters 2 and 4, respectively. To obtain an adequate understanding of the subject, the reader should complete the earlier sections before proceeding.

Preparations

Rinderpest is a disease that depends on the mixing of livestock for survival. There are no known carrier states or examples of chronic infection. The virus survives through the continuous occurrence of new acute cases in susceptible animals. Thus, the virus must continuously find new susceptible animals for survival. A safe, effective and inexpensive vaccine exists that has been successfully applied in most areas where veterinary service delivery is reliable. These factors suggest, and indeed experience has shown, that remote, marginalised pastoral communities, where livestock contact rates are high and vaccination is sporadic, are often the reservoir for the endemic persistence of rinderpest (Mariner, 1996).

Rinderpest PDS should take these factor into account when designing reconnaissance missions. As a first step, an inventory of remote cultures and communities should be made and secondary sources of information should be obtained and researched. In PRA, one often looks at ‘proxy indicators’ in regard to sensitive subjects for which it may be difficult to obtain unbiased direct information. Remoteness, insecurity, and lack of services can serve as proxy indicators for the likely presence of infectious disease when prioritising search areas at the national level. It is especially important to avoid ‘tarmac’ or access bias in regard to rinderpest searching.

As with any disease investigations, the appraisal team should carry all necessary sampling materials in case a SE outbreak is encountered. In addition, the team should carry Clearview tests and a camera, if these are available. Photographs of affected animals and lesions are supporting data for diagnostic purposes and in some cases become invaluable training and communication materials. A list of recommended materials is provided in Annex 1.

In general, it is better to involve a representative of the veterinary services from the national level as members of the study team who is not known by community members. Local representatives of the veterinary services are included in the study as key informants. The presence of local veterinary authorities in the appraisal team can introduce bias as the respondents may wish to avoid sensitive subjects. Incorporation of local officials in the study team makes the process less anonymous and confidential.

Interviewing

As with other types of disease searches, the interviewer should be careful not to communicate their specific interest in rinderpest to the respondents. If possible avoid travel in vehicles used in rinderpest control activities or with individuals locally known to be specialised in rinderpest control. The interview should be introduced as a general study of animal health issues in cattle or a similarly broad theme.

A sample PRA checklist for rinderpest disease searching is presented in Box 4. Items 1 to 3 introduce the subject and establish the identity of the participants. Items 4 to 6 are an expanding enquiry into animal health problems. Item 4 investigates what animal health problems the respondents are personally and presently facing. Item 5 inquires into what animal health problems the immediate community is presently facing. Finally, Item 6 inquires into the most significant animal health problems the community has faced regardless of time. It is important to note at which stage in the interview process rinderpest is mentioned.

Rinderpest should not be mentioned in an interview by the appraisal team prior to the respondents listing of animal health problems. If a respondent introduces the subject of rinderpest, he/she should be asked to describe the disease as part of the verification process. If the respondent cannot accurately describe the disease, his report should be discounted. The weighting process described in Chapter 4 should be used for volunteered reports from individuals who can accurately describe the disease.

If rinderpest has been described accurately, the following topics should be probed:

1. Have they had personal experience with the disease or did they learn about it from others?

2. If they have had personal experience with a rinderpest compatible event, when, where and in whose cattle did they observe the disease?

3. How severe was the outbreak? What was the relative importance of symptoms?

4. What were the general circumstance at the time of the event (grazing conditions, water availability, security conditions, livestock contacts with other communities and wildlife, trade links, etc.)?

5. Has the disease occurred at any other time? Repeat questions 1 to 3, as time permits, for each previous occurrence of SE outbreaks.

Box 4: Sample PRA Checklist for a Participatory Rinderpest Disease Search Avoid mentioning rinderpest before the cattle owners do. 1. Introduce the appraisal team as an animal health appraisal. 2. Identify the respondents and establish if they cattle owners. 3. Establish their main herding locations (mapping). 4. What are the current cattle disease problems in their herd? If tearing or diarrhoea is mentioned, explore these syndromes in detail. 5. What are the current cattle disease problems in the area? 6. Historically, what are the most important disease problems of cattle? Invariably rinderpest is mentioned in the response to this question if the cattle owners have experienced outbreaks in the last two decades. Frequently it will be the first disease mentioned. 7. Have they personally seen rinderpest in the their lifetimes? What does it look like? 8. When was the last time their cattle where affected by rinderpest? Where? Where did it come from? As warranted, further probing questions can be added to cross-check reports made in other interviews, further define cattle movements which may affect the epidemiology of the disease, or to contrast current outbreaks with previous outbreaks in regard to the severity of disease.

Topics for Probing

Probing on specific subjects can provide very useful insights into community knowledge on disease epidemiology, pathology and diagnostic processes. Often it is best to reserve these probing sessions for especially knowledgeable key informants. These are usually more senior members of the community respected and consulted by the community for their livestock knowledge.

Diarrhoeal disease: In a subset of interviews, the community should be probed about the different terms used to describe diarrhoeal disease. Points to be investigated are:

· What terms are used to signify diarrhoea in general?

· Are different types of diarrhoea distinguished (i.e. bloody vs. non-bloody, acute vs. chronic, etc.)

· What specific diseases do they associate with diarrhoea?

· What indicators are used to differentiate between different diarrhoeal diseases?

Disease concepts and methods of rinderpest transmission: It is useful to attempt to understand local concepts regarding the cause of disease and methods of transmission. In regard to rinderpest, pastoralists can often accurately describe risk factors and types of contact that lead to transmission. Knowledge in these areas varies significantly between communities and it is important not to make assumptions.

Community responses to rinderpest: Understanding community responses to rinderpest, either endemic or epidemic, can contribute to risk analysis in regard to the spread of the disease or provide insight into the mechanisms of endemicity. Some responses to rinderpest that have been noted in the past are:

· ‘Quarantining’ affected herds by making them graze away from unaffected herds or water after other herds

· Avoidance of wildlife

· Running away from outbreaks

· Traditional vaccination

Mapping

Rinderpest is a disease that depends on cattle contact and movement for its very survival. Mapping of cattle movement and determination inter-community contact is a very important activity in regard to understanding local rinderpest ecology.

In order to initiate a mapping exercise, respondents should be asked to specify their primary grazing sites by season. Depending on the complexity of the bio-climatic system, it may be worth constructing a seasonal calendar as a companion exercise. Often, pastoralists will specify location names that are not evident on modern maps. These initial questions often turn to a discussion of just exactly where these key resource sites are which naturally leads to the sketching of maps on the ground.

A broad area about an arm and a half’s reach should be cleared and smoothed. Usually, participants will naturally gather round and equip themselves with the necessary tools: normally sticks and other objections to assist in drawing and act as land marks. The participants should first be asked to indicate key landmarks such rivers, market towns, major wells or watering sites. Then the grazing sites can be indicated. In addition to normal grazing sites, emergency-grazing areas used in time of drought or insecurity should be indicated. The participants can also indicate the territories occupied by other communities and points of contact as well as shared or contested resources. During the mapping process, information can be collected on the nature of the interaction with neighbouring communities (e.g. trade, competition, raiding, hostility, etc.). Eventually, the map will become crowded with information and it will be time to stop and transfer the information to note books.

If desired, the mapping exercise can be repeated in later sessions for the collections of other types of data such as the location of forests, bush and crop areas, or wildlife distribution. Later, the different types of information can be analysed conceptually as layers, much as in geographic information systems.

For rinderpest epidemiology, the data on movement, mixing (contact) and trade will be the most significant. These will be key factors in subsequent risk analysis and in disease control strategy design. Movement and contact data relative to the presence of the virus will determine where and when vaccination or surveillance is appropriate.

Clinical Observation, Sample Collection and Transects

Before or after an interview, it is always useful to walk the camp, herd and adjacent environment. In rinderpest disease searching, tearing is a sign that can be detected at a distance. If you are walking the herd prior to the interview and note tearing, it is best not to call attention to the sign. Proper clinical exams should be carried out after the interview. The only exception is when the livestock are on the move and you run the risk of losing the opportunity. If you are unfamiliar with the temperament of local breeds or they are know to be aggressive, due caution should be exercised.

If active rinderpest is reported or tearing is noted, a complete clinical exam of affected animals is essential. Detailed notes on the individual history and clinical presentation of each animal should be taken. After examination, clinical cases compatible with the SE outbreak definition should be sampled. Stomatitis-enteritis sampling procedures are presented in Annex 3.

In any event, the appraisal team should take a moment to investigate any examples of current health problems the livestock owners would like to present. This gives the study team an additional opportunity to cross check disease descriptions with actual clinical cases. Further, as the livestock owner has donated his valuable time to the study, it is only appropriate that the team take a moment to investigate the farmer’s concerns and provide useful advice. Bear in the mind the problem of creating false expectations. It is best not to promise or suggest any future assistance. The provision of free drugs, although it may seem helpful and even be enjoyable, contributes to the creation of false expectations and perpetuates the psychology of dependence. If drugs are required, the best solution is often to provide the farmer with a verbal or written prescription and advice on where they can be purchased.

Analysis of Results

Participatory disease searching is somewhat different from other types of PRA. Most PRA interviews start at a general level and work towards specifics. The interview technique used in PDS casts an ever broadening net until the respondents volunteer rinderpest as a problem. At this point the interview begins to focus down on EVK regarding rinderpest. This could happen at any of three levels: current personal experience, current personal observation or in the past. Part of the process of judging the quality and significance of reports relates to when the respondent introduces the subject of rinderpest. The earlier in the interview process that the subject is raised, the more significant the report. This feature incorporated in the ranking criteria presented in Chapter 4.

As described previously, reports of rinderpest or other SE events should be categorised, tabulated and examined for trends or unifying factors. This process should begin in the field and be carried on throughout the study. The lessons learnt during the PDS will lead to the reformulation of hypothesis, new questions and modification of the criteria of analysis. Always bear in mind that rinderpest intelligence may be sensitive information at any level of the system from the herder to head of veterinary services. The key is to remain flexible, patient and open-minded throughout the process.

The existence of the stomatitis-enteritis outbreak definition and guidelines for comprehensive rinderpest surveillance assists in the analysis of rinderpest PDS results. If the communities terminology fits the SE outbreak definition, then the reports collected as part of the PDS should enter in the rinderpest report registries. If the community consensus points to the circulation of rinderpest or another SE agent, then the investigation should continue until a definitive diagnosis is reached. The principals of comprehensive rinderpest surveillance and performance monitoring require that the PDS should continue until active cases are found for sampling. In the event that active SE cases cannot be found, then a purposive serosurvey is indicated.

The concepts of epidemiologically linked cases, epidemiologically characteristic outbreaks and suspected versus probable outbreaks presented in Chapter 4 are vary useful in the analysis of rinderpest epidemiology. As an example, the 1994 Tsavo outbreak in buffaloes was clinically and epidemiologically characteristic of rinderpest. This led to a very high index of suspicion in regard to rinderpest. Repeatedly laboratory investigations were necessary to confirm a rinderpest diagnosis. Although it was not overtly recognised at the time, the Tsavo outbreak was accepted as a probable rinderpest outbreak and the burden of proof had shifted from proving that the outbreak was rinderpest to proving that it was not. In fact, qualitative diagnostic assessments carried more weight than initial laboratory tests until repeated tests began to detect rinderpest. If this hadn’t been the case, the Tsavo outbreak might have been misdiagnosed as a non-rinderpest event. Considerable more work is required to assure that the concept of epidemiological diagnosis is incorporated into rinderpest control and the diagnostic decision tree.

In the event that a PDS detects a rinderpest focus and representative cases are confirmed, then all the cases that are epidemiologically linked to the confirmed case are themselves confirmed. The linked cases are all those that fit the same clinical description and were determined to be in contact or for which a chain of transmission can be reasonably assumed. In regard to rinderpest, a chain of transmission could be reasonably assumed where herds share a watering hole or grazing area.

In the event that a detailed PDS detects a rinderpest focus and representative cases are not confirmed, then the epidemiological intelligence gathered by the PDS should form a working hypothesis for future disease control efforts in the area. The PDS result can be used to formulate future surveillance and vaccination tactics. The PDS may also serve as a research hypothesis for further detailed epidemiological studies using both qualitative and quantitative techniques. The need to conduct further research must be balanced with action-oriented needs. This is especially the case in regard to vaccination activities that will preclude serological investigation or may obscure clinical disease without achieving eradication. It is better to have a full understanding of the epidemiological and ecological dynamics of the disease prior to embarking on extensive vaccination programmes with vague objectives.

Participatory Epidemiological Intelligence and Serosurveillance.

Seromonitoring
Participatory Epidemiology and the Design of Serosurveillance Programmes
Interpretation and Investigation of Serosurveillance Results

Serology is used in GREP in two ways. In countries or zones where vaccination is practised, seromonitoring to measure the effectiveness of vaccination programmes is used. In disease-free area where vaccination has ceased, serosurveillance is used to detect residual circulation of virus.

The OIE pathway for the eradication of rinderpest relies heavily on serosurveillance in its final stages. This is based on the assumption that negative national surveys will verify the eradication of the virus. At present, it is not known if fully negative surveys are feasible. Although the competitive ELISA test (Anderson among others, 1991) in use has been shown to highly specific in benchmarking studies, data on non-specific reactions in large field surveys from diverse areas have not been fully reviewed. Specificity is not a constant test property across geographical areas as local factors can alter the rate of non-specific reactions (Schwabe, 1984). The OIE Pathway provides that ‘Any positive result’ obtained through serosurveillance ‘be evaluated using epidemiological and laboratory methods to confirm or refute the suspicion of rinderpest virus activity’ (OIE, 1998).

To date, application of serology in vaccination monitoring and rinderpest surveillance programs has resulted in more questions than conclusive answers. This is not undesirable as the questions generated by serology lead member states to take a closer look and gather epidemiological intelligence.

Seromonitoring

In the case of vaccination seromonitoring in areas where there is significant risk of rinderpest infection, it cannot be said with certainty whether seroprevalence rates represent vaccination or vaccination and disease exposure. In fact, in some instances, ancillary data on vaccination returns and population estimates combined with qualitative intelligence have indicated that seromonitoring was detecting more than vaccination. The consistent use of participatory surveillance as part of general disease surveillance (Chapter 3) and participatory disease searches in high-risk areas make interpretation of seromonitoring results more reliable.

Participatory Epidemiology and the Design of Serosurveillance Programmes

The OIE recommends sampling age groups born after the cessation of vaccination but older than one year of age using a random sampling technique. The OIE pathway provides some guidelines on epidemiological methods for sampling to insure a valid sample of sufficient sensitivity (OIE 1998). However, selection of the options best suited to local conditions from within the sampling techniques suggested by the OIE requires a thorough knowledge of husbandry and management practices at the local level. As most of the practices are culturally based, participatory appraisal would be an appropriate method for establishing:

· Data on contact rate and mixing
· Appropriate herding units to be used as sampling units
· Inventorying bore holes, wells and water points for use as sampling units in nomadic areas
· Identification of populations and stratification criteria

Interpretation and Investigation of Serosurveillance Results

After vaccination is ceased, successive age groups of young stock should be fully seronegative by 9 months of age. Preliminary serosurveillance exercises in disease free areas have been difficult to interpret due to foci of seropositivity in ‘unexpected’ locations and age groups. As the competitive ELISA test in use is reported to be highly specific for rinderpest, these foci are taken to be instances of improper ageing at time of sampling, unauthorised vaccination, animal movement or residual foci of rinderpest infection. Instances of all four types are believed to have occurred in the past. In order to resolve these enigmas, the epidemiologist must resort to epidemiological intelligence on the following issues:

· Ageing methods used in serosampling and the reliability of their application
· Unauthorised or inadvertent vaccine distribution
· Unauthorised or inadvertent use of residual vaccine stocks remaining in the field
· Stock owner reports of vaccination
· Animal movement and transhumance patterns
· Reports or other evidence of rinderpest disease

In this case, it is essential that the study team contact all parties that may have been involved in unauthorised or inadvertent vaccination. As the checklist indicates, the vaccine production laboratory, all distribution points and all the links in the distribution cold chain (provincial, state, the district affected and neighbouring districts) should be contacted and interviewed. Inventory and distribution lists should be verified. In addition, all organisations (projects, international organisations and non-governmental organisations) and individuals such as private veterinarians who might have conducted vaccination should be contacted.

As the appraisal reaches field level, the team should contact key informants in the community about the history of animal health interventions in the area. At the community level, the appraisal should discuss general animal health interventions and problems. At all times, the team should be alert to reports or rumours of SE disease. The appraisal should map transhumance and trade animal movements to determine if movement may have been a factor. If no concrete and specific evidence of vaccination on a scale sufficient to explain the seropositive results is detected or, if reports of stomatitis-enteritis are detected in addition to evidence of vaccination, than a full participatory disease search of the area of unexplained antibody prevalence is essential.

Based on the intelligence collected during the participatory investigation, ancillary serosampling may be indicated. This can take the form of larger purposive surveys of selected cattle age groups or the sampling of small ruminants. However, more serology often proves to be an inconclusive and expensive operation. In any event, decision-makers should be prepared to use the intelligence gathered by the participatory investigation as a working hypothesis to guide disease control operations, either as a stand alone technique or while waiting for the results of ancillary surveys.

Box 5: Sample PRA Checklist for the Investigation of Seropositive Foci in Apparent Disease Free Areas 1. Interview serosampling teams as to ageing methods in use and the consistency of their application. Brain storm as to the possible causes of seroconversion. 2. Review the vaccine distribution chain for unauthorised or inadvertent distribution and the presence of unnecessary residual stocks in nonvaccinated areas. Interview, validate records and directly observe stocks at: Laboratory stores Vaccine depots All other links in the distribution chain 3. Interview key informants at all organisations and individuals who might have carried out rinderpest vaccination in cattle, used rinderpest vaccination in the control of PPR, or received reports of rinderpest or other stomatitis-enteritis disease outbreaks. Veterinary services Projects International organisations Non-governmental organisation Private veterinarians 4. Interview key informants at the community level regarding animal health interventions performed in the area. Do not focus on rinderpest. Discuss general animal health issues. Lastly, probe issues relevant to stomatitis-enteritis. 5. Map animal movement to determine if animals enter vaccinated zones or high-risk areas for infection. 6. Conduct a disease search in the seropositive communities (See Box 4). 7. Identify and collect ancillary serosamples for biological testing.

FAO ANIMAL HEALTH MANUALS

1 Manual on the diagnosis of rinderpest, 1996

2 Manual on bovine spongiform encephalopathy, 1998

3 Epidemiology, diagnosis and control of helminth parasites of swine, 1998

4 Epidemiology, diagnosis and control of poultry parasites, 1998

5 Recognizing peste des petits ruminants - A field manual, 1999

6 Manual on the preparation of national animal disease emergency preparedness plans, 1999

7 Manual on the preparation of national rinderpest contingency plans, 1999

8 Manual on livestock disease surveillance and information systems, 1999

9 Recognizing African swine fever - A field manual, 2001

10 Manual on participatory epidemiology - Methods for the collection of action-oriented epidemiological intelligence, 2001

The FAO Animal Health Manuals are available through FAO Sales Agents or directly from the Sales and Marketing Group, FAO, Viale delle Terme di Caracalla, 00100 Rome, Italy; fax (+39) 06 5705 3360; e-mail: [email protected]

For any comments or suggestions about this publication please contact: EMPRES (Livestock) Animal Health Service FAO Animal Production and Health Division Viale delle Terme di Caracalla 00100 Rome, Italy Tel.: +39 06 57054798/6772 Fax: +39 06 57053023 E-mail: [email protected] www.fao.org/empres

Acknowledgements Cover photo: Roger Paskin The initiative to compile and publish this manual came from Peter Roeder of the Global Rinderpest Eradication Programme (GREP).