1 The Data

A county launched a preventive care program in 2020. Preventive screening rates increased after the program started. But look at the trend before the program—rates were already rising at the same pace. (Data are simulated for illustration.)

Preventive Screening Rate Over Time

Observed Data

2 The Threat Explained

Outcomes can change for reasons that have nothing to do with your program. Ignoring these other causes leads to false conclusions.

History

External events that happen at the same time as the program can cause the outcome to change.

Examples:

  • A new state mandate required insurers to cover preventive care
  • COVID-19 changed healthcare utilization patterns
  • A major employer added health benefits

Maturation

Outcomes naturally change over time, regardless of any intervention.

Examples:

  • Public awareness of preventive care was already growing
  • Healthcare technology was steadily improving
  • Population demographics were shifting

Why This Matters

In this lab's data, screening rates were increasing by about 2.1% per year before the program even started. After the program, they continued at nearly the same rate (2.3%). The program may have added nothing—we're just seeing a trend that was already underway.

Pre-post designs (comparing before to after in a single group) cannot distinguish program effects from underlying trends or coincidental events.

For a full comparison of how different designs handle this threat, see Lab 4: Study Design Ladder.

3 What To Do

Given the pre-existing trend, how should we interpret the post-program improvement?

Choose Your Interpretation

The pre-program trend continued at nearly the same rate after the program launched:

Policy Memorandum

To:
Health Department Leadership
From:
Health Economics Analysis Unit
Re:
Preventive Care Program Evaluation
Date: