Using Participatry Epidemilgy t Assess the Impact f Livestck Diseases Andy Catley and Berhanu Admassu Cmmunity-based Animal Health and Participatry Epidemilgy (CAPE) Unit, Pan African Prgramme fr the Cntrl f Epiztics, African Unin s Interafrican Bureau fr Animal Resurces, PO Bx 3786, 1 Nairbi, Kenya. E-mail: andy.catley@au-ibar.rg Website: http://www.cape-ibar.rg 1. Intrductin In an era f declining public sectr veterinary services in Africa, pririty setting and ratinal allcatin f resurces is becming increasingly imprtant. Regarding livestck disease cntrl, many cuntries lack the basic epidemilgical and ecnmic infrmatin that enables disease prblems t be priritised at lcal r natinal levels. Furthermre, infrmatin deficits are ften mst evident in thse areas characterised by large livestck ppulatins and high levels f pverty. In recent years the methds f participatry rural appraisal have been adapted by epidemilgists t imprve understanding f livestck diseases in resurce-pr settings and in areas where cnventinal methds are difficult t use. The value f this apprach is apparent frm the emergence f participatry epidemilgy (PE) as a distinct branch f veterinary epidemilgy, and the applicatin f PE by prgrammes such as the Pan African Prgramme fr the Cntrl f Epiztics (AU-IBAR) and the Glbal Rinderpest Eradicatin Prgramme (FAO). This paper prvides an verview f PE, utlines hw PE has been used in impact assessment t date and prpses hw PE can be adapted t understand hw and why livestck keepers priritise diseases. 2. What is participatry epidemilgy? Participatry epidemilgy is the use f participatry methds t imprve understanding f animal health issues. Key features are summarised belw: Attitudes and behaviur Practitiners are required t assess their wn prfessinal and cultural biases. Essentially, they needed t be genuinely willing t learn frm lcal peple, nt lecture t them but actively and patiently listen. This requires respect fr lcal knwledge and culture. Cmbined methds and triangulatin Participatry epidemilgy uses interviewing, scring and ranking, and visualisatin methds (Table 1). Of these, interviews are the mst imprtant grup f methds because they are used alne but als cmplement and frm the basis fr ther methds. The visualisatin methds include mapping (natural resurce maps, scial maps, service maps), seasnal calendars, time-lines, transects, Venn diagrams, flw diagrams. Scring methds include matrix scring and prprtinal piling. These methds are cmbined with cnventinal veterinary investigatin and epidemilgical tls. The use f key infrmants Althugh pastral cmmunities generally are recgnised as knwledgeable abut animal health matters, certain peple are knwn t pssess special livestck knwledge and skills. These lcal experts are imprtant key infrmants fr participatry epidemilgists. Actin-rientated Participatry epidemilgy aims t generate infrmatin that can be verified with cmmunities and leads t agreement n apprpriate actin. Initially, the aims f a particular study r investigatin shuld be clearly explained t avid raising expectatins. In sme situatins, further labratry results
will be required and the mechanism fr transferring these results back t the cmmunity shuld be defined. Methdlgical flexibility, adaptatin and develpment Participatry epidemilgy is a relatively new branch f epidemilgy that is still develping. The apprach is based n qualitative inquiry and cmplements the qualitative nature f standard veterinary investigatin prcedures. Accrding t the needs f a given cmmunity r rganisatin, participatry epidemilgy can als cmbine the benefits f participatry appraches and methds with quantitative inquiry. Methdlgical adaptatin is encuraged. Table 1. Examples f participatry epidemilgy methds Infrmatin required Backgrund infrmatin: System bundary Scial rganisatin Wealth grups Relative livestck wnership Preferred types f livestck reared Fd, incme and ther benefits frm livestck Marketing systems Veterinary services Resurces available t rear livestck Disease-specific infrmatin: Pririty livestck diseases, with reasns Lcal characterisatin f diseases accrding t disease signs and causes Estimates f incidence and mrtality Tempral infrmatin: - histry f livestck diseases - seasnal variatins in livestck disease, vectrs and livestck-wildlife interactins Spatial infrmatin: - cntact with neighburing herds, wildlife, disease vectrs - areas f disease events - preferred cntrl ptins, with reasns PE methds a Natural resurce maps, scial maps. Scial mapping, Venn diagram Wealth ranking Prprtinal piling Livestck species scring Prprtinal piling Flw diagrams, service maps Service map, Venn diagrams, ranking and scring Natural resurce maps, transects. Disease scring Matrix scring Prprtinal piling; prgeny histry Timelines Seasnal calendars Mapping; mbility maps Mapping Matrix scring a Semi-structured interviews can prvide infrmatin n all tpics 3. Uses f participatry epidemilgy Uses f PE t date are summarised in Figure 1. Experiences f particular relevance t impact assessment are: Basic epidemilgical research, including estimates f disease incidence and mrtality Methds used in the impact assessment f cmmunity-based animal health prgrammes
Figure 1. Current uses f participatry epidemilgy in pastral areas f the Hrn f Africa Basic research n epidemilgy f endemics & epiztics* Studies n new diseases Prblem analysis & prgramme design Mnitring & impact assessment* Cmmunitybased Animal Health Prgrammes Research Epiztic Disease Cntrl Disease mdeling* Reliability & validity studies n PE methds Participatry disease searching Uses marked * are particularly relevant t impact assessment f livestck diseases 3.1 Basic epidemilgical research: estimates f disease incidence and mrtality Participatry epidemilgy studies have included estimatin f disease incidence and mrtality using methds such as prprtinal piling. Sme f the benefits f the methd include: Ppulatin data in terms f numbers f animals is nt required. A ppulatin r herd is defined using spatial and tempral criteria. This avids sensitive questins n herd size and means that the methd can be used in areas with limited r n baseline data n ppulatin. Lcal definitins f herd structure and age grups are used, tgether with lcal disease names. This reduces translatin errrs and specifically, nndifferential misclassificatin bias (cf. questinnaires) The methd is cmparative and assesses up t 1 diseases simultaneusly. If the researcher has an interest in a particular disease (e.g. CBPP), infrmants shuld nt be aware f this interest.
Sme f the difficulties r limitatins f the methd include: Very careful explanatin f the methd and therefre gd training f researchers is required Mst applicatin s far has been with pastral r agrpastral infrmants, with strng diagnstic ability. The methd may be less useful with ther types f livestck keeper. Crsscheck diagnstic skills with ther methds e.g. matrix scring. Recall is an issue. Pastralists seem able t accurately recall disease events ver many years and in specific animals, but what abut ther livestck keepers? Crss-check with timelines and secndary data n disease utbreaks. Examples f the type f data that can be prduced by prprtinal piling are shwn in Figures 2 and 3. 7 Mean herd incidence and mrtality (%) 5 3 1 mean herd incidence Figure 2. Mean herd incidence and mrtality estimates fr three cattle diseases in Maasai herds, Mrgr regin, Tanzania, -1 (n=5 herds) using prprtinal piling calves yung stck adult mean herd mrtality Age grup a. Olukuluku 8 8 Mean herd incidence and mrtality (%) - calves yung stck adult mean herd incidence mean herd mrtality Mean herd incidence and mrtality (%) 7 5 3 1 calves yung stck adult mean herd incidence mean herd mrtality Age grup Age grup b. Endrb c. Oltikana
healthy 42.4% Figure 3. Estimates f cattle disease incidence and healthy cattle in Orma herds, Tana River District, Kenya, 1999- (n=5 herds) gandi 17.9% thers 6.3% buku 6.2% madbesa.7% hyale 14.6% s mba 11.9% 3.2 Methds used in the impact assessment f cmmunity-based animal health prgrammes Impact assessment f cmmunity-based animal healthy prgrammes has included the use f lcallydefined indicatrs f the impact f diseases. One f the principles here is that livestck keepers determine impact using sme indicatrs that veterinarians might verlk. Fr example, an impact assessment in Ethipia revealed that Afar herders regarded varius scial payments such as alms giving and dwry payments as imprtant benefits derived frm cattle. In these cmmunities, marriage requires payment f cattle t the bride s father and alms giving includes gifts f livestck t the pr. Figure 4. Relative imprtance f benefits derived frm cattle in Afar cmmunities, Ethipia (n=1 infrmant grups, prprtinal piling) Alms giving 6.6% Calves fr the herd 24.5% Incme frm sales 7.6% Dwry payments 6.2% Use f skins 7.6% Milk sales 5.% Milk fr cnsumptin 31.% Meat fr cnsumptin 6.% Other 5.3%
In the Afar example, standardisatin f the methd and repetitin with different infrmants (r infrmant grups) allwed a statistical assessment f data reliability. When livestck keeper perceptins f benefit are knwn, it is then pssible t cmpare methds such as prprtinal piling t shw the impact f different diseases n each f these benefits. An example is prvided in Figure 5. Nte that depending n the specific questins asked, this methd can capture perceptins f incidence, mrtality and duratin f impact as an verall reductin in benefit indicatr. Figure 5. Relative impact f six cattle diseases in Afar cmmunities, Ethipia 8 Benefits Reductin in benefit as a prprtin (%) f ttal benefit abeeb dahan gubl halab kiribi migda Alms giving Calves fr herd Incme frm sales Dwry Skins Other Meat cnsumptin Milk cnsumptin Milk sales Diseases 4. An utline PE-based methdlgy fr assessing the impact f CBPP Based n the PE methds utlined abve, a draft methdlgy fr the cmparative assessment f cattle diseases is presented in Table 2. This invlves initial stages f defining a systems bundary and cmmunity identificatin f the 1 mst imprtant cattle diseases. In the event that CBPP is nt mentined during this initial stage, the research team can chse t add CBPP as an additinal disease. Hwever, this risks biasing the research because infrmants may suspect that the researchers have a particular interest in CBPP.
Table 2. Outline minimum methdlgy fr PE-based impact assessment f cattle diseases Infrmatin required (per study lcatin) Participatry appraisal methds: Methd Sample size per lcatin Cnventinal methds/surces f secndary data 1. System bundaries: - spatial - tempral Mapping Timelines 1 key infrmant grup per methd Cnventinal maps DVO recrds 2. Livelihd surces by wealth grup - surces f fd - surces f incme - cntributin f livestck, by species, t livelihd 3. Identificatin f the 1 mst imprtant cattle diseases a Wealth ranking; prprtinal piling Simple disease ranking crsschecked with pair-wise ranking 4. Analysis f impact f the 1 mst imprtant cattle diseases - identify lcal impact indicatrs b - relate impact indicatrs t diseases c SSI Matrix scring 5 infrmants/ wealth grup 5 infrmants/ wealth grup 5 infrmants/ wealth grup 5. Incidence & mrtality estimates Prprtinal piling 5 infrmants/ wealth grup Sci-ecnmic reprts (if any) DVO recrds; previus research studies Market recrds fr value f livestck & livestck prducts Previus studies 6. Optins fr preventing r treating the 1 mst imprtant diseases - identify cntrl ptins used fr each disease - rank/analyse preferences - identify & rank main cnstraints t cntrl fr each disease SSI Ranking/SSI SSI/ranking 7. Market pprtunities & cnstraints Service maps, SSI, ranking 5 infrmants/ wealth grup 3 infrmant grups per wealth grup Optins/ntes a. This can be separated ut by livestck species, but dramatically increases time inputs b. Requires breakdwn f general impact indicatrs e.g. General indicatr = cash Specific indicatrs = uses f cash (fd, schl fees, clthes, medical etc) c. Includes impact in relatin t acute r chrnic nature f the diseases