Skip to content

Chapter 06 / 06

From a clue to a supported explanation

What does a positive test actually tell us?

The previous chapters looked at possible mechanisms. Now ask how we decide which explanation fits—connecting observations and staying open to information that could change the conclusion.

Follow the mechanism
  1. Describe a precise observation
  2. Consider possible explanations
  3. Collect relevant evidence
  4. Revise the explanation if needed

A simplified concept map. The body has many interacting pathways, not a single straight line.

A measurement needs its circumstances

A laboratory result describes a sample taken at a particular time. Stress can temporarily raise feline blood glucose, and hydration or muscle mass can affect how kidney markers are interpreted. A reference interval compares results with a defined reference population; it is not a standalone diagnosis. History, physical examination, related tests and trends over time provide context that a single number cannot supply.

Sources and further readingFeline DiabetesChronic Kidney DiseaseReference intervals

The starting population matters

Sensitivity asks how often a test comes back positive among animals with a condition. Specificity asks how often it comes back negative among those without it. Neither is the chance that a positive result means disease—that also depends on how common the condition is in the population being tested. Try the fictional example below to see the distinction.

Sources and further readingBasic Principles of Epidemiology

The same test, a different starting point

Imagine a fictional group of 1,000 animals. This test comes back positive for 90% of animals that actually have the condition, and negative for 90% of animals that do not. Only the share of animals with the condition changes.

Proportion with the condition: 1%
Condition presentCondition absent
Positive result999
Negative result1891

Of 108 positive results, 9 are true positives.

Proportion with the condition: 10%
Condition presentCondition absent
Positive result9090
Negative result10810

Of 180 positive results, 90 are true positives.

Proportion with the condition: 50%
Condition presentCondition absent
Positive result45050
Negative result50450

Of 500 positive results, 450 are true positives.

These invented numbers do not describe any real disease or test. They show why a positive result needs context, not your cat’s probability of illness.

An observation can suggest more than it proves

A case report describes an experience; an observational study can reveal associations; a well-designed controlled comparison can better examine an intervention’s effects. Imagine a simple example: cats given a new food also receive more playtime. Weight improvement alone cannot separate these two influences. When reading a study, ask who was studied, what was compared, which outcome was measured and what alternative explanations remain.

Sources and further readingBasic Principles of Epidemiology

Three terms to keep

Sensitivity
The fraction of truly affected animals that test positive.
Specificity
The fraction of truly unaffected animals that test negative.
Base rate
How common a condition is in the population considered, before the test result.

Pause and reason it through

Does 90% sensitivity and 90% specificity mean that every positive result implies a 90% chance of disease?

Read the explanation

No. With 1% affected in our fictional population of 1,000 animals, there are 9 true positives and 99 false positives—so only 9 of the 108 positive results are true positives. This arithmetic explains why context matters; it is not a clinical estimate.

Connect this to the everyday guide

For curious readers, with no medical background required. These chapters explain concepts; decisions about a particular cat belong in a veterinary consultation.

Sources and further reading

We wrote these explanations independently and checked them against the references below. Citing a source does not imply institutional endorsement or veterinary review of PurrJung.

Sources checked · · References in English

  1. 04
  2. 10
  3. 14
  4. 15