Cancer and Palliative Care Assessment Instruments

Three kinds of instrument are used here. Symptom burden measures such as the Edmonton Symptom Assessment System and the Memorial Symptom Assessment Scale record several symptoms at once, because people with advanced disease rarely have only one. Performance status measures such as the Karnofsky Performance Status Scale grade overall functional capacity and are used in eligibility criteria and treatment decisions as much as in outcome measurement. Cancer-specific quality of life questionnaires such as the EORTC QLQ-C30 measure the burden of disease and treatment across domains, with disease-specific modules added to the core questionnaire.

Two further instruments sit here for specific reasons: the Functional Assessment of Chronic Illness Therapy fatigue measure, because fatigue is the most commonly reported and least well managed symptom in this population, and the Distress Thermometer, a single-item screen designed to make distress screening feasible in a busy clinic.

Symptom burden instruments are usually collected repeatedly and frequently, sometimes daily, which makes response burden a real design constraint in a population that is unwell. The Edmonton system is short for that reason. Longer instruments give more detail and are completed less often and less completely, and in this population the missing responses are systematically from the people who are sickest.

Performance status carries a specific caution because it is used to decide eligibility for treatment and for trials. It is a clinician judgement with broad bands and known inter-rater variation, and small differences in rating change who is offered treatment. Where it is used as an eligibility criterion, the rating should be documented rather than asserted.

Check the recall period and the administration schedule together. Daily symptom collection and weekly collection answer different questions and cannot be pooled.

Check which EORTC module is being used alongside the core questionnaire, and report both. Module scores are not comparable across modules.

Plan explicitly for missing data caused by deterioration and death. In this population, missingness is strongly informative, and analysing only completed questionnaires produces a picture of a cohort that is systematically healthier than the one enrolled. Pre-specify how it will be handled. Finally, screening instruments such as the Distress Thermometer need a defined response pathway before they are deployed, because screening without a route to support is not a neutral act.

Which symptom burden instrument should I use?

It depends on how often it will be collected. Short multi-symptom instruments such as the Edmonton Symptom Assessment System suit frequent repeated collection. Longer instruments give more detail but are completed less reliably in unwell populations, and the people who stop completing are the ones whose data matters most.

How reliable is performance status rating?

Moderately. The bands are broad and raters disagree, particularly in the middle of the range, yet the rating is used to decide treatment and trial eligibility. Document the basis for the rating rather than recording the grade alone.

How should missing quality of life data be handled in advanced cancer studies?

As informative. Deterioration and death are the main causes of missing responses, so complete-case analysis systematically overstates quality of life at later timepoints. Pre-specify the approach and report completion rates at every timepoint.

Is a single-item distress screen enough?

It is enough to identify who needs a conversation, which is what it was designed for. It is not a severity measure and not a diagnosis, and it should only be deployed where there is a defined pathway for responding to a positive result.