Data are actively obtained by recruiting individuals or healthcare facilities to participate in testing and providing data. 

The determination of the susceptibility of a bacterial isolate to antimicrobials designed to kill or inhibit it.

The quantity of antimicrobial drugs prescribed or administered to an individual person or animal, or group of animals (e.g. herd, flock), measured at the level of healthcare facility, farm, region or country.

A pure culture of a bacterial strain isolated from a mixed population of bacteria in an animal/human or environment. In the context of AMR surveillance this will be a strain we are interested in because we want to know if has AMR.

Errors that represent a consistent deviation from the true value – for example, measurements that are consistently higher or lower than the true value. It can arise due to problems with the design, data collection, data management or analysis stages of surveillance, or research studies.

Collecting data from every single unit (e.g. human, animal, farm) of a population or category of objects.

Bacteria that normally live in the human body without doing the host any harm. In fact, some of these are beneficial to the host in different ways, which includes competing with other potentially pathogenic bacteria.

The probability of getting the results observed by chance alone: the higher the confidence, the less likely the results observed are due to chance. Larger sample sizes result in higher levels of confidence. Confidence levels of 90% or 95% are commonly used.

A statistical measure of drug consumption, defined by the World Health Organization Collaborating Centre for Drug Statistics Methodology.

The loss of collected samples due to issues with sample transport, failure to obtain a result from the sample analysis in the laboratory, and in human health instances where people stop participating in a study. Sample size calculations should take this into consideration.

Deviation from the true value that occurs due to chance alone. In general, the smaller the sample size, the larger the random error.

The degree to which the sample is representative of the target population. It is required to extrapolate findings from the sampled population to the target population. It is sometimes referred to as ‘generalisability’.

The degree to which the sample is representative of the source population.

A sampling and analysis method where a population is classified as having high or low prevalence of AMR depending on a classification of a sample of isolates.

Bacteria (or, infrequently, other micro-organisms) that are resistant to a range of different antibiotics.

Making use of existing activities to collect data, such as using antimicrobial susceptibility testing (AST) data used to guide patient treatment for surveillance purposes.

The power of a study is the probability of detecting an effect, such as a difference between two groups, if it is true. Larger sample sizes result in more power. In general, powers of 80% or higher are considered acceptable.

In the context of sample size calculations, precision describes the margin of random error around an estimate. Precisions of 5% or lower are commonly used.

The proportion of a population with a particular condition at a particular time, such as the number of Salmonella isolates in the population that are resistant to a particular antibiotic at the time of reading.

The extent to which an event is likely to occur, expressed quantitatively as a value between 0 and 1 (or between 0% and 100%).

Single units from a category (e.g. human, animal, farm) selected to be representative of the category as a whole.

A list of all sampling units in the source population.

The basic element(s) of the population we are sampling. This is often the case when sampling is conducted in multiple steps, or stages. For example, we may have: 

• hospitals (primary sampling unit) 
• wards (secondary sampling unit) 
• individual patients (tertiary sampling unit).

The actual population from which eligible study subjects are drawn, such as hospitalised patients at a single hospital in the capital city. Ideally, the source population should be the same as the target (reference) population, but this is often not practical.

The units selected from the source population; for example, 40 patients randomly selected from inpatients on a surgical ward over a defined time period. The study sample should be similar to the source population.

The population to which the results will be extrapolated to; for example, all hospitalised inpatients in the country.

Disease-causing bacteria that can be transferred from animals to people.