Interrupted Time Series Design

TL;DR

Compares outcome measurements before and after a clearly timed intervention to estimate its effect.
Useful when random assignment is impractical and when observing how effects evolve over time.
Vulnerable to other events occurring at the same time, which can confound attribution.

Definition

An interrupted time series design is a type of quasi-experimental research design in which a treatment or intervention is applied to a population at a specific time, and the effect of this intervention is measured by comparing the changes in the outcome variable(s) before and after the intervention. It is called “interrupted” because the time series of the outcome variable is “interrupted” by the application of the intervention.

Explanation

The design measures an outcome repeatedly over time, noting the point when an intervention is introduced and comparing the trajectory or level of the outcome before and after that point. This approach allows researchers to observe the intervention’s effect across a period rather than at a single time point, helping to reveal patterns such as immediate changes, gradual trends, or unintended consequences. It is particularly applicable when randomized controlled trials are not possible and when studying interventions that cannot be manipulated in a laboratory setting (for example, public health campaigns or policy changes). A key limitation is susceptibility to external factors that coincide with the intervention, which can make it difficult to attribute observed changes solely to the intervention.

Examples

Smoking cessation program and hospital admissions for heart attacks

A study could measure the number of hospital admissions for heart attacks over a period before a new smoking cessation program is introduced in a particular city, implement the program, then measure admissions over a similar period afterwards. Comparing the before-and-after measurements allows researchers to assess whether the program had a significant effect on hospital admissions for heart attacks.

Road safety campaign and traffic accidents

A study could measure the number of traffic accidents in a particular region before a new road safety campaign is launched, implement the campaign, then measure accidents over a similar period afterwards. Comparing the before-and-after measurements allows researchers to assess whether the campaign had a significant effect on the number of traffic accidents.

Use cases

When it is not possible or practical to randomly assign participants to control and intervention groups (unlike randomized controlled trials).
To study interventions that are difficult or impossible to manipulate in a laboratory setting, such as public health campaigns or policy changes.
To evaluate the effects of interventions in real-world settings over time.

Notes or pitfalls

Vulnerable to external factors occurring at the same time as the intervention, which can confound results.
Example pitfall: if a smoking cessation program is introduced simultaneously with a major public health campaign promoting healthy eating, it may be difficult to disentangle the effects of the two interventions on hospital admissions for heart attacks.

Quasi-experimental research design
Randomized controlled trials