Two ways to measure frailty in one site-completed form: the clinician-rated, nine-level Clinical Frailty Scale and the performance-based Fried phenotype. This page covers structure, scoring and licensing, and how to record both consistently.
Free sandbox · No credit card · 21 CFR Part 11 aligned
Clinical Frailty Scale
CFS version
CFSVERCFS level
CFSLVLFried phenotype
Unintentional weight loss
Exhaustion (self-report)
Weak grip
Slow walking speed
Low physical activity
2 of 5: pre-frail (derived in analysis)
Frailty measurement at a glance
Licensing
Free, with conditionsRights holder: Dalhousie University (Clinical Frailty Scale, Rockwood)The Clinical Frailty Scale is copyrighted; Dalhousie University reviews permission requests. Dalhousie's technology office says non-commercial educational, clinical and research uses usually do not need a licence agreement, but commercially sponsored use may need one, so confirm in writing. Version 2.0 is current. The Fried phenotype is a published research method and is not itself licensed. This was checked through search summaries because the Dalhousie page could not be opened, so verify directly.
In Capture: Capture doesn't supply this instrument or a licence for it. Once you have both, enter the official wording yourself and keep the licence with your study documents. Capture's licence controls apply to patient questionnaires, not to site-completed forms like this one.
Instrument names belong to their rights holders, and Capture is not affiliated with or endorsed by them. Terms change, so confirm them with the rights holder before your study starts. This is general information, not legal advice.
The instruments
Frailty is a state of reduced reserve that makes people vulnerable to a small stressor. The CFS, developed by Rockwood and colleagues, summarises it in one judgement after a clinical assessment and history. It is quick, works in hospital and clinic and predicts outcomes, but it is a rating, so training and a consistent rater matter. The Fried phenotype, from the Cardiovascular Health Study, is a performance-based definition: the participant is measured, not judged. It takes more time and equipment, and it ties directly to physical function endpoints such as the SPPB and grip strength.
Many ageing trials record both: the CFS as an enrolment or stratification variable and the Fried phenotype as a stricter characterisation or endpoint. For the design side of this, see EDC for frailty and healthy ageing trials.
The five criteria are unintentional weight loss over the past year, self-reported exhaustion, weakness measured by grip strength, slow walking speed over a short course, and low physical activity. In the original definition, weakness, slowness and low activity are set against the lowest fifth of a reference population, adjusted for sex and body size, so studies publish or adopt their own cut-points. Because adaptations vary, specify them in the protocol and record the raw measurements, not just yes or no, so you can re-classify later.
Field by field
| Field | Type | Notes |
|---|---|---|
| Assessment date, rater | Date, text | Rater initials support consistency review |
| CFS version | Single choice | Version 1.3 or 2.0, fixed per study |
| CFS level | Single choice, 1 to 9 | One level; use the licensed scale, not a local paraphrase |
| CFS basis | Single choice | In person, record review or informant, per protocol |
| Weight loss criterion | Single choice, number | Weight now and a year ago in kg, so the loss is checkable |
| Exhaustion criterion | Single choice | Self-report using the protocol's question |
| Grip strength | Number, kg | Best of three, hand and dynamometer noted |
| Walking speed | Number, seconds | Time over a fixed distance; aid used |
| Physical activity | Number or category | Per the protocol's activity instrument |
| Criteria met | Number | Entered by site or derived in analysis from the five answers |
| Phenotype category | Single choice | Robust, pre-frail or frail, derived in analysis |
This page does not reproduce CFS level descriptions. Use the scale as licensed by Dalhousie University.
Licensing
The CFS is a clinician-rated scale, so in Capture it lives on a site-completed eCRF. Capture's licence controls (licence number, holder, expiry, a "copyright approval pending" watermark) exist for participant questionnaires only and do not apply to eCRFs. The permission email or agreement with Dalhousie stays in the trial master file. Capture does not verify licences with rights holders and does not supply the scale.
Copyright & licence
Approval pendingLicence number
Not recorded
Expiry date
Not recorded
Watermark shown to staff only. Participants never see it.
Setup
Request use of the CFS from its rights holder, state that the study is research, and say whether it is commercially sponsored. Fix the version.
Add CFS version and level as single-choice fields, then the five Fried raw measurements.
Range checks on grip strength and walking time catch a unit slip. A custom-value check can prompt the site when weight now and weight a year ago imply a loss of 5 percent or more, so the weight-loss criterion is not left at No by mistake.
If the participant cannot walk, show a reason field instead of the walking time. See calculated fields and skip logic.
Set source verification on the CFS level and grip strength.
Approve the form and place it at baseline and the visits your endpoints need.
Worked example
A participant has a CFS level of 5. Grip strength is below the study's sex-adjusted cut-point, and the participant reports exhaustion. Weight is stable, walking speed is within range, and activity is above the cut-point. Two Fried criteria are met, so the participant is pre-frail. The CFS and Fried results disagree on how to label the person, which is normal: they are not the same construct. The protocol states which one defines eligibility and which is descriptive.
Because raw grip, walk time and weights are stored, the analyst can re-run the classification if the cut-points change, which a stored yes/no would not allow.
Criteria met of 5
2 / 5
Before go-live
Written permission or confirmation from the rights holder for your use.
Raters score example cases and agree on how to handle informant history.
The protocol states reference population and adjustments for sex and body size.
Equipment and walking distance standardised across sites.
Measurements are kept so classification can be rerun.
Add the fields, set the checks and enter a sample participant in the free sandbox.
Dalhousie University holds the copyright asserted by the scale's author. Dalhousie's technology office says non-commercial educational, clinical and research uses usually do not need a licence agreement, but it reviews all permission requests. Commercially sponsored use may need a licence, so confirm with Dalhousie in writing.
Version 2.0 is the current form. Whatever you pick, record the version on the form and keep it constant in the study.
Count the criteria met out of five: weight loss, exhaustion, weakness, slowness and low physical activity. Zero is robust, one or two is pre-frail, three or more is frail.
No. Capture's calculated fields are a fixed set of clinical derivations. Store the raw measurements and derive the count and category in analysis.
Cut-points differ between populations and publications. Raw values let you reclassify later without recontacting participants.
It is a clinician rating based on assessment and history. Build it as a site-completed eCRF.
No. The scale is copyrighted, so use the official version obtained under Dalhousie's terms.
Keep exploring
Build the form, set the edit checks and test a participant. Free sandbox, no credit card.