Global ratings of change do not accurately reflect functional change over time in clinical practice

Retrospective longitudinal cohort study.

To determine whether the global rating of change (GROC) is valid for assessing functional change over time in the clinical setting, and whether validity decreases as length of recall increases.

The validity of GROC instruments for representing change over time has been questioned. Our previous study showed that GROC did not adequately represent functional change over time in patients with knee disorders. This was a validation study in patients with hip or ankle/foot disorders.

We obtained complete data from the Focus On Therapeutic Outcomes clinical database on episodes of care for 7341 patients with hip or ankle/foot disorders. Functional status (FS) was assessed at the intake and discharge physical therapy visits via computerized adaptive testing. Each patient's GROC was assessed at the discharge visit. We assessed the extent to which GROC was associated with intake, discharge, and change in FS using correlation and linear regression analyses, stratified by recall period and anatomical location.

Correlations of GROC with hip and ankle FS change scores were modest for recall less than 30 days and deteriorated further as the recall period lengthened. Prediction of GROC from discharge and change in FS scores by linear regression modeling demonstrated a decline of standardized beta coefficients and R2 values as the recall period lengthened. Regression analyses indicated that GROC was biased toward discharge FS, with little or no influence of intake FS.

These results confirm that GROC does not adequately or consistently correlate with functional change across varying lengths of time, and is therefore not recommended as a substitute for change scores on valid functional outcome measures for use in routine clinical practice.