James Wood Lectures

Interpretation of diagnostic tests

Purpose of diagnostic tests 

  • reduce uncertainty in diagnosis and management
  • diagnosis is based on combination of clinical signs and test results 
  • judgement re interpretation is critical - leads to (in)action

Determinants of test usefulness 

  • Validity
  • Reliability - reproducability 
  • Effect on decision of clinical outcome - quality of test given characteristics and clinical context 
1 of 12

Measures of tests

Measures of test characteristics

  • sensitivity - proportion with disease who test positive 
  • specificity - proportion without disease who test negative 
  • these are normally inversely related

Measures of test interpretation 

  • positive predictive value (PPV) - how likely is positive result to be real
  • negative predictive value (NPV) - how likely is negative result to be real 
  • test interpretation is very dependent on disease prevelance in population - both are highly impacted by prevalence 
2 of 12


3 of 12

Control Programmes

  • control case numbers per unit time  - important if management can be overwhelmed 
  • minimise total number of cases occuring in an epidemic - are under curve of infecteds 
  • minimise impact of disease and associated control programme - total no. culled farms in FMD outbreak 
  • need to lower infection rate, keep it less that 1:1
4 of 12

FMDV control

  • much transmission occurs prior to diagnosis 
  • key features: minimising infectious animals (early diagnosis), reduce transmission, reduce infectious animals by culling, tracing 
  • timing of controls is critical 
  • vaccination can be used to prevent or aid control 
5 of 12

Bovine TB controls

  • endemic infection, dealt with by test and slaughter of individual cattle with herd level controls/movement restrictions 
  • large geographic heterogeneity, - high incidence area extending from SW to S Wales, Scotland almost free
  • complicated by - significant wildlife reservoir, poorly sensitive cattle diagnostic test 
  • much transmission occurs prior to diagnosis - sometimes symptoms never present 
  • key features - minimise infectious animals by early diagnosis, reduce transmission
  • timing of controls critical
6 of 12

African Horse Sickness control

  • culicoides midge transmitted, as BTV and Schmallenberg
  • disease not found in Europe - BTV and disease modelling suggest disease could spread in summer 
  • major focus on excluding infection from Europe 
  • timing of control critical - need early diagnosis 
  • vaccination likely important if outbreak occured
7 of 12

Bovine Tuberculosis Control - Disease

  • disease caused by Mycobacterium bovis infection 
  • M. bovis can infect a wide range of mammalian hosts - incl. cattle, humans, badgers, sheep, camelids, wild boar, deer etc.
  • prior to pasteurisation M. bovis killed 2,500 people a year in the UK (GI disease) - now very few cases - most people are aged >50
  • disease has been subject to statutory controls in UK since 1950s
8 of 12

Bovine Tuberculosis Control - Diagnosis

  • routine slaughterhouse surveillance of all animals (meat inspection)
  • routine testing in cattle herds and slaughter of diagnosed animals - annual testing in high incidence areas, four yearly testing of breeding stock in low incidence areas 
  • pre-movement testing of cattle moved from/in high incidence areas
  • detailed follow up on herds, with movement restrictions, where infection diagnosed 
  • ante-mortem diagnostics (skin test/hypersensitivity reaction) - insensitive - leaves burden of infection, specificity ~99.98% - can be used in low incidence areas 
  • gamma interferon assay improves sensitivity but poor specificity - cost is high 
9 of 12

Bovine Tuberculosis Control - Epidemiology

  • transmission by cattle to cattle contact and contact with wildlife reservoir 
  • missed infection - herd level persistence, movement of undetected infected cattle between herds 
  • significant wildlife reservoirs - badgers GB/Ireland, different densities, different species in other countries 
  • evidence for wildlife infection - shared infections locally (genotyping strain), culling trials, experimental trials
  • very heterogeneous across UK - cattle to cattle transmission in low incidence areas, heavy burden of wildlife challenge in SW England and parts of Wales - badgers infection estimated to be associated with 50% of herd breakdowns in trial areas 
10 of 12

Bovine Tuberculosis Control - Control

  • still actively spreading despite controls - massive increase in disease when controls halted - eg 2001 FMDV outbreak when testing ceased 
  • very regionalised - different drivers in different regions 
  • RBCT - culling badgers can reduce incidence of new infections - ~25% in culled regions, incidence increased ~12% in peripheral regions, average rate is 16%
  • cattle vaccination slight effect on disease - massively hampered by need to continue testing vaccinated animals, lack of available DIVA, cost-benefit analysis suggests it might not be feasible to get farmer uptake 
11 of 12

Bovine Tuberculosis Control - Future Prevention

  • Badgers - culling part of England government policy, but may make things worse, vaccination part of Welsh and NI policies - unclear of impact on cattle disease 
  • vaccination of cattle - partial efficacy - previously slower than test and slaughter programme in 1950s, subject to complex rules with EC, currently illegal, unclear quite how much vaccination would interfere with testing 
12 of 12


No comments have yet been made

Similar Veterinary Science resources:

See all Veterinary Science resources »See all tests resources »