| Grade | Description | KPS Equivalent |
|---|---|---|
| 0 | Fully active, no restriction | 100% |
| 1 | Strenuous activity restricted; ambulatory; light work possible | 80–90% |
| 2 | Ambulatory; all self-care; cannot work; up >50% of waking hours | 60–70% |
| 3 | Limited self-care; in bed/chair >50% of waking hours | 40–50% |
| 4 | Completely disabled; no self-care; totally bed/chair-bound | 10–30% |
| 5 | Dead | 0% |
| Grade | Trial Eligibility | Treatment Approach |
|---|---|---|
| 0–1 | Eligible for most clinical trials | Full chemotherapy doses; curative intent |
| 2 | Selected trial eligibility; study-specific | Reduced doses; less aggressive regimens; assess benefit vs. risk |
| 3–4 | Generally ineligible for trials | Palliative intent; best supportive care; chemotherapy likely harmful |
| 5 | N/A | Death — document cause |
The ECOG and Karnofsky scales measure the same underlying construct — functional status — but with different granularity. The Karnofsky scale (0–100 in steps of 10) offers finer discrimination, particularly at higher performance levels, while ECOG (0–5) is faster to apply and more reproducible between raters in busy clinical settings.
Neither scale is superior; selection depends on context. ECOG is the standard for most oncology clinical trials and NICE assessments; Karnofsky remains widely used in transplant medicine, bone marrow transplant eligibility, and some European centres.
ECOG PS is one of the strongest independent prognostic factors across virtually all solid tumours. It integrates multiple dimensions of health — nutritional status, comorbidity burden, disease activity, and psychological reserve — into a single clinically accessible number.
Chemotherapy decisions: Most phase III trials restrict enrolment to ECOG 0–1 or 0–2. In routine practice, ECOG 3–4 represents a population where the toxicity of standard chemotherapy is likely to outweigh any benefit. This should trigger a goals-of-care conversation and palliative referral.
Longitudinal monitoring: Worsening ECOG PS during treatment is a key early warning sign — it often precedes radiological progression and should prompt prompt reassessment of the treatment plan.
Reproducibility caveat: Inter-observer variability exists, particularly at grade 2. Use structured descriptors rather than numeric estimates alone to improve consistency.