A short introduction to Quantitative Microbial Risk Assessment v0.2

Mon Nov 22 2021

About this project

This page is still in its beta version and aims at giving a 10 minutes overview of the Quantitative Microbial Risk Assessment (QMRA). It is meant for users that have no prior experience with the method but with some familiarity with key concepts of risk assessment. Please take a look and give us your feedback.


While epidemiological studies represent the gold standard to understand health risk associated with environmental exposure, Quantitative Microbial Risk Assessment (QMRA) is an alternative method that has gained popularity in recent years.


Let’s first understand what Quantitative Microbial Risk Assessment (QMRA) means. The method helps in quantifying the risk caused by a pathogen(s) in a given situation.

As an example, we will look at the risk of illness while swimming in a river.

Step 1: Hazard Identification

In this step a specific hazard is identified for the analysis. This is usually related to a situation, a target group and involves a pathway for transmission and one or more pathogens.

In the example, as a pathway for transmission we will consider involuntary ingestion of contaminated water during swimming. Our target group will be swimmers and the hazard identification aim to identify what may be the most relevant pathogen to study, in this case we will consider Adenovirus. The choice of the pathogen can be determined based on prevalence in society, environmental presence or specific interest in a particular pathogen.

Step 2: Exposure assessment

This step quantifies the exposure of the target population (the swimmers) to the hazard ( Adenoviruses) through the route of transmission (accidental ingestion of contaminated water). The result is the quantification of a dose of pathogen that is ingested by the target.

To quantify the dose, i.e. the number of pathogens, involuntary ingested by a swimmer two data are needed: the involuntary ingestion of water and the concentration of viruses in the wste. The involuntary ingestion per 45 min swim, is assumed based on a previous study to be 20 ml of water. The concentration is estimated using field data as 5 virus per 100ml

The dose is therefore estimated as:

dose = ingestion x concentration

In the example, the average ingestion would be 1 virus per swim.Try it yourself:

Involuntary ingestion of water (ml):


Average concentration of viruses (viruses / 100ml):


Average dose of viruses ingested in one swim:


Step 3: Dose response

Using dose-response curves we can estimate the probability of infection, or illness given a certain dose. These curves are based on experimental or epidemiological data and are determined for each specific pathogen. See below the dose-response curve for Adenovirus that shows the relationship between ingested dose (x-axis) and the probability of infection (y-axis).


See some dose-response curves at this page.

Calculate risk of infection:

Not every infection will turn into illness. This is another parameter that can be estimated by looking at existing published studies, or the dose response model may directly relate the dose to illness e.g. when based on outbreak data. For Adenovirus, an infection to illness factor of 0.3 will be used. This means that roughly one infection out of three will result in illness. Explore how different pathogen concentrations and ingestions can affect the estimated probability of infection.

Ingestion (ml): 20.00

Concentration (viruses / 100 ml): 1.00

Infection to illness ratio: 0.3

Dose (ingested viruses): 0.20

Risk of illness: 0.02

Step 4: Risk characterization

This step combines all the previous components to characterize the risk of infection over a defined period of time. In our example this refers to the risk of infection from ingesting river water containing Adenoviruses. Previously we have estimate the risk for infection from a single swimming event to be 2%.

With the previously calculated risk per swim event we could also have an initial estimate of how many people could get sick in a hot summer day where 500 swimmers are present. This would simply be 500 x 2% = 10.

For a person being exposed repeated times for example by going on a daily swim during a month (therefore 30 events), the risk of infection can be calculated. This exposure would result in a 45 % risk of being infected if swimming everyday during a month 45%.

Risk per event: 0.02

Number ot events: 1.00

Risk for multiple events: 0.02

Step 5: risk mitigation

One of the uses of QMRA is to evaluate mitigation strategies to lower the risk for exposed persons to a level that is considered acceptable. What will be an acceptable risk would be very context dependent. This means that we could estimate the risk of infection for a swimmer, over the period of one month, in case one of the following measures are taken:

Mitigation 1: a more stringent regulation of wastewater discharges into the river lowers the concentration of Adenoviruses by half.

Mitigation 2: allowing only for activities above water (ex. canoeing) would likely reduce the ingestion of contaminated water (let’s assume to 7ml) and therefore risk of illness.

Going back to the previous steps, we could estimate each mitigation and evaluate which one could be more effective.


We hope this quick introduction has given a general understanding of the QMRA methodology.

For a detailed explanatation of the method look at the book Quantitative Microbial Risk Assessment by Haas, Rose and Gerba. For more concise information, there is a QMRAwiki page

A review by Van Abel and Taylor on QMRA done in Sub-Saharian Africa gives an overview of limitations, approaches and differences that can be taken.

For a comparison with epidemiological data and for a clear straightforward application of the method, read this article by Mara, Sleigh, Blumenthal.

Lasly, keep in mind that, as any numerical modelling tool, the results are highly dependant on the quality of the data and assumptions.If you are curious to see how different input data can cahnge results, read this article where we run a QMRA using two models: one using field data and one using published literature.


Idea,coding and text: by G.Butte and A.Nordin

Illustration: by Ailadi

Text editing: by A.Aldinucci

Tools: Idyll