[Durgical](https://durgical.com/en/) › [DurgiPedia](https://durgical.com/en/durgipedia/) › [Epidemiology](https://durgical.com/en/epidemiology/) › [Survival analysis](https://durgical.com/en/epidemiology/areas/survival/) › Stage migration (Will Rogers phenomenon)

# Stage migration (Will Rogers phenomenon)

- Bias

Better staging moves patients to a more advanced stage: both stages seem to improve although nobody lives longer.

Updated Nov 13, 2026

## Key points

- A new technique (CT, MRI, PET, sentinel node) finds hidden metastases and reclassifies patients.
- Both stages improve their survival; overall survival stays the same.
- It biases stage-by-stage comparisons between periods, hospitals or trials with different staging work-ups.
- It is controlled by comparing overall survival or by classifying everyone with the same criteria.
- It affects almost all solid tumours.

On this page

- The basics Key points
- What it is
- How it arises
- Worked example
- 5-year survival before and after reclassification (example)
- Markers
- How to prevent or control it
- Common pitfalls
- Memory aid
- Elsewhere in DurgiPedia

- Study Test yourself

- Connections Connections

## What it is

Stage migration happens when a new diagnostic technique reveals spread that could not be seen before, and some patients who would have been classed in an early stage move to an advanced one. Those who migrate are the worst-prognosis patients of the early stage but the best-prognosis patients of the advanced stage, so survival in each stage rises while overall survival does not change. It is called the Will Rogers phenomenon and it affects comparisons of stage-specific survival between eras or centres.

## How it arises

- Before: the early stage includes patients with hidden disease who live less than the rest of the group.
- After: those patients move to the advanced stage, where they live longer than the group's average.
- The average of both groups rises; the total number of deaths does not change.

## Worked example

- 200 patients, 5-year survival. Before: stage I 100 (80 alive, 80%); stage II 100 (40 alive, 40%).
- Overall: 120 of 200 = 60%. A new, more sensitive imaging test arrives.
- It finds hidden metastases in 20 stage I patients, of whom 10 live 5 years (50%).
- New stage I: 80 patients and 70 alive = 87.5% (before 80%).
- New stage II: 120 patients and 50 alive = 41.7% (before 40%).
- Overall: 120 of 200 = 60%, unchanged. Reading: both stages improve and nobody lives longer.

## 5-year survival before and after reclassification (example)

| Group | Before | After |
| --- | --- | --- |
| Stage I | 80/100 = 80% | 70/80 = 87.5% |
| Stage II | 40/100 = 40% | 50/120 = 41.7% |
| Total | 120/200 = 60% | 120/200 = 60% |

## Markers

- Will Rogers
- Stage
- TNM

## How to prevent or control it

- Compare eras or centres with overall survival or mortality, not only by stage.
- If comparing by stage, reclassify everyone with the same criteria and the same tests.
- In trials, require the same staging work-up in every arm.

## Common pitfalls

- Crediting a new treatment for the stage-specific improvement of a recent series against a historical one.
- Comparing stage-specific results of a hospital with more advanced imaging against one with less.
- Confusing it with lead-time bias: here the time of diagnosis does not change, the classification does.

## Memory aid

Moving the weakest of the strong group into the weak group raises both averages, yet nobody got any stronger.

## Test yourself

Think of the answer, then tap the question to see it.

## Why does survival improve in both stages with stage migration?

Those who migrate are the worst of the early stage and the best of the advanced stage, so each group's average rises.

## Which measure does stage migration leave unchanged?

The overall survival (or mortality) of all patients together.

## A 2025 series with PET against a 2005 series without it shows better survival in every stage. What do you suspect?

Stage migration; compare overall survival or reclassify with the same criteria.

[Add to my cards (3)](https://durgical.app/estudiar/?add=/en/epidemiology/stage-migration/)

## Elsewhere in DurgiPedia

- [Grading and staging: what TNM means](https://durgical.com/en/general-pathology/grading-and-staging-what-tnm-means/)

## Connections

## Related 6

- [Lead-time bias](https://durgical.com/en/epidemiology/lead-time-bias/)
- [Length-time bias](https://durgical.com/en/epidemiology/length-time-bias/)
- [Survival analysis and censoring](https://durgical.com/en/epidemiology/survival-analysis-and-censoring/)
- [Five-year and relative survival](https://durgical.com/en/epidemiology/five-year-and-relative-survival/)
- [Uncontrolled studies and historical controls](https://durgical.com/en/epidemiology/uncontrolled-studies-and-historical-controls/)
- [Age, period and cohort effects; reading trends](https://durgical.com/en/epidemiology/age-period-and-cohort-effects-reading-trends/)

[← All of epidemiology](https://durgical.com/en/epidemiology/)

Educational content on epidemiological and statistical methods; the examples use made-up numbers to teach the calculation. Not a clinical or legal guide: current Colombian rules are in the official sources.

Sources: [SEER Cancer Statistics Review 1975-2016: Introduction](https://seer.cancer.gov/archive/csr/1975_2016/results_merged/sect_01_overview.pdf) (National Cancer Institute, 2019); [Measures of Cancer Survival](https://surveillance.cancer.gov/survival/measures.html) (National Cancer Institute, Surveillance Research Program, 2026).

## **Spotted a mistake?** We are always improving for the community. Tell us what is wrong and our team will review it. Report a mistake

Representative model, for educational purposes only.

---
Source: https://durgical.com/en/epidemiology/stage-migration/
License: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Cite as: DurgiPedia (durgical.com), https://durgical.com/en/epidemiology/stage-migration/