Download presentation
Presentation is loading. Please wait.
Published byNicholas Hamilton Modified over 9 years ago
1
Principles of AAD surveillance Should be tailored designed wide variety of species cultured, the pathogens and management systems Support to domestic production Aid to international trade (international disease reporting, OIE standards) Moving toward Output based approaches
2
Sabina Šerić-Haračić TCP consusltant
3
Principles of AAD surveillance Major determinants of good AAD surveillance strategy Planning AAD surveillance – resources, sampling, analysis Presentation outline
4
Principles of AAD surveillance Different certification level Farm accreditation National/regional disease free status Monitoring of diseases in environment May target specific disease May include all diseases (even previously unknown/unseen)
5
Principles of AAD surveillance AAD surveillance objectives: Early detection of disease Demonstrating freedom of disease Control/eradication of disease AAD monitoring – detection of disease trends DEALS with endemic diseases DEALS with Exotic diseases
6
Principles of AAD surveillance
7
Principles of AAD surveillance Surveillance to demonstrate freedom from disease On-going evidence Certification for zones, regions and compartments Threshold set by design prevalence No single survey is enough
8
Principles of AAD surveillance Surveillance for distribution and occurrence of disease prevalence and incidence Morbidity and mortality Risk factors Differences between epi units Days from confirmation to control actions
9
Principles of AAD surveillance Surveillance for distribution and occurrence of disease On going evidence Definition of population Structured survey/s +other: sentinel, production records, biosecurity etc. Representative sampling
10
Principles of AAD surveillance
13
Diagnostics More than having good test Surveillance More than having right sample size
14
Principles of AAD surveillance Major determinants of good AAD surveillance strategy the definition of population documentation of methodology/sampling diagnostics design prevalence quality assurance systems
15
Principles of AAD surveillance Data collection Targeted and non targeted Disease focus Selection of units
16
Principles of AAD surveillance Methods of data collection Active primary purposes of surveillance activities Data tailored to surveillance needs Population based surveys (at slaughter or live animals) Expensive Passive Surveillance uses data from other sources (drug use, farm records, etc.) May lack representativeness, completeness, timing Collect all data you need, use all data you collected!!!
17
Principles of AAD surveillance In addition to disease diagnosis data Epidemiology of disease Movement of animals (cultured and wild) History of trade/import Compliance with health regulation ALL DATA SOURCES SHOULD BE DESCRIBED!!!
18
Principles of AAD surveillance Sources of data for AAD surveillance Laboratory databases Field reporting system Negative reporting system Production records STRUCTURED SURVEYS!!!
19
Principles of AAD surveillance Structured surveys for both exotic and endemic diseases Designed based on hypothesis testing (i.e. Disease frequency = 0, or < designed prevalence) Designed base on estimation of population parameters Surveillance ≠ survey
20
Principles of AAD surveillance Single stage survey (individual animals) Certification of batches of animals for export Certification of single establishment (one pond/cage) Stratified surveys By species By region By production type Allows multiple sampling methods to account for differences
21
Current status of disease Purpose of surveillance Data type and sources Population, coverage, representativeness Quality Approach/methodology Planning AAD surveillance
22
Disease present Reliably measure disease frequency/trends Make corrective actions Monitor effectiveness Disease absent Demonstrate disease freedom Early diagnosis Purpose of surveillance Dependent from disease presence/absence Dependent from certification level (farm/region/country) Planning AAD surveillance
23
Presentation outline Challenges of AAD surveillance comparing to surveillance of terrestrial animal diseases (P9 17.7.2013.) Principles of AAD surveillance Major determinants of good AAD surveillance strategy (P13 17.7.2013.) Planning AAD surveillance – resources, sampling, analysis (P13 17.7.2013.) Draft surveillance design (P16 18.7.2013.)
24
Types of data collected Diagnosis disease presence Surveys, laboratory confirmation Classification – disease stage Syndrome, signs (morts, swimmers) Indirect indicators (growth, feed consumption, reproductive performance) Risk factors (monitoring of environment) Planning AAD surveillance
25
Proving disease absence – never with 100% Probabilistic approach i.e. Acceptable probability that surveillance system will detect disease at designed prevalence Different techniques – same output – Confidence measure that diseases will be detected Statistical basis – hypothesis testing Null hypothesis – i.e. Disease is present at equal or greater than designed prevalence Evidence to support null hypothesis Planning AAD surveillance - Survey design -
26
Design prevalence at animal level (i.e. prevalence of infected animals in a cage): between 1% and 5% for infections that are present in a small part of the population over 5% for highly transmissible infections first level of clustering (i.e. proportion of infected farms in a zone) is not greater than 2% Planning AAD surveillance - Survey design -
27
Probability of rejecting true null hypothesis = alfa (α) 1-α = strength of evidence confirming null hypothesis – measure of confidence ≥95% (account for test characteristics) Reject Ho = disease freedom Accept false Ho - no international standards Consequence – more samples Planning AAD surveillance - Hypothesis testing-
28
Estimation of disease prevalence Confidence interval that includes true prevalence Representative sample Allowed error Test characteristics New techniques still developing Use currently available tools, technically valid and scientifically based Planning AAD surveillance - Survey design -
29
Analysis of results Account for survey design Account for diagnostic test imperfection Account for prevalence Planning AAD surveillance
30
Sampling method Test entire population – census Test sample Representative from population Non representative Large populations - sampling frame not available - multi-stage sampling Planning AAD surveillance
31
Representative sampling – each individual in population has same and equal probability being selected into sample Simple random sapling Systemic random sampling Stratified sampling Proportionally stratified sampling Cluster sampling Planning AAD surveillance
32
simple systemic stratified cluster
33
Planning AAD surveillance Random sampling requires having sample frame i.e. All individuals/units accessible and identified NOTHING IS RANDOM IN RANDOM SAMPLING!!! Alternative use systemic/spatial sampling Used for wilages/farms/cages For individual aquatic animals – no sampling frame Use any method to achieve random selection – documented and described Convenience sampling never acceptable
34
Non representative sampling Sampling of available animals – convenience sampling Targeted sampling Moribund /with lesions animals Planning AAD surveillance
35
Why different sampling methods? Farm management Size of groups/pools? With/without broodstock? Feasibility Ensure randomness Identify all animals/groups/farms Access to all animals/groups/farms Disease biology Infectious vs. Noninfectious diseases
36
Sample size factors into account: Imperfection of diagnostic test/s (Se, Sp) the design prevalence/s the level of confidence Other factors: Population size (acceptable to assume infinitely large population) power of the survey Before – standardized tables Now – tailored calculation based on above factors Planning AAD surveillance
37
Any procedure that help determining diseases better than by chance alone Screening and diagnostic Gold Standard Planning AAD surveillance - diagnosistic tests -
38
12345678910 1 2 3 4 5 6 7 8 9
39
Planning AAD surveillance - diagnosistic tests - 12345678910 1 2 3 4 5 6 7 8 9
40
Planning AAD surveillance - diagnosistic tests - 12345678910 1 2 3 4 5 6 7 8 9
41
Planning AAD surveillance - diagnosistic tests - No perfect test!!! How much the test can be wrong Overall Diagnosing disease Diagnosing health Imperfection of test/s for interpretation of surveillance data
42
Planning AAD surveillance - diagnosistic tests - 12345678910 1 2 3 4 5 6 7 8 9
44
Planning AAD surveillance
45
DnoD T+45348 T-54752 50 100 Planning AAD surveillance
46
DnoD T+45348 T-54752 50 100 Planning AAD surveillance True Prevalence = 50/100 Apparent Prevalence = 48/100 Ability of test to diagnose disease Test sensitivity=45/50 Ability of test to diagnose health Test specificity=47/50
47
Planning AAD surveillance Se/Sp must be known for test used to demonstrate freedom Ideally determined for same population where test is applied Most use more than one test
48
Planning AAD surveillance Combination of tests increases specificity on the expense of sensitivity and visa verse Independency of test results – use biologically independent tests For pooled sample testing use relevant Se/Sp
49
Quality of data Disease present Prevalence (precision and accuracy) Disease absent Sensitivity - probability that surveillance sys. Will detect at least one case if disease is present at ≥ design prevalence Planning AAD surveillance
53
Approaches/methodologies of surveillance Passive disease reporting Structured surveys - population based surveillance Sentinel surveillance Risk factor surveillance Syndromic surveillance Participatory disease surveillance Post harvest processing Planning AAD surveillance
54
Sentinel surveillance Alternative for population based surveillance Selected individuals/establishments Fewer resources - restricted number of samples Regular complete reports One or more diseases NONREPRESENTATIVE for entire population Suitable for high risk groups – exotic diseases, rare diseases Planning AAD surveillance
55
Risk base surveillance Theory – EASY! Looks where you expect disease to occur More sensitivity with less samples Efficient but cheaper Practice – LITTLE COMPLICATED!? What is risk? Where does it apply? How to calculate sample size? Planning AAD surveillance
56
Relative risk 8 High risk: 10% animals, 80% prevalence Low risk: 90% animals, 20% prevalence True prevalence 17% Apparent prevalence 80% Planning AAD surveillance - risk based surveillance -
57
Risk - likelihood of adverse event Likelihood and consequences – result of risk analysis Risk based surveillance Risk factors (water temperature, age, moribund,...) Difference in risk (with and without risk factor) – relative risk Sampling contribution of high risk subpopulation Planning AAD surveillance - risk based surveillance -
58
Relative Risk Risk of event (disease) relative to exposure (risk factor) Planning AAD surveillance - risk based surveillance - DnoD Expos ed 104050 Non expos ed 54550 1585100
59
Samplin g type High risk populati on % Low risk populati on % Sample size %saving Represe ntative 20803310 Risk based (RR=3) 50 23130 Risk based (RR=3) 901016550 Biased (non represent ative) 1090387-17 Planning AAD surveillance - risk based surveillance - Prevalence 1%, Test sensitivity 90%
60
Planning AAD surveillance - risk based surveillance -
61
Syndromic surveillance Early detection of outbreaks a threshold number of early symptomatic cases Well-defined disease or clinical syndromes Indicates unusual clustering or sentinel cases Trends – size, spread and tempo Use existing health data Planning AAD surveillance
62
Participatory surveillance Give stakeholders greater role overcomes limitations of conventional epidemiology developed in small-scale applied to major international disease control efforts – OIE – rinder pest Provides: observations, semi-quantitative scores, quantitative data Planning AAD surveillance
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.