Predicting patient-reported stroke outcomes: a validation of the six simple variable prognostic model

Area of Science:

• Neurology
• Clinical Epidemiology
• Health Services Research

Background:

• Case-mix adjustment is crucial for comparing healthcare outcomes across diverse patient populations.
• The Six Simple Variable (SSV) model is established for adjusting hard stroke endpoints.
• Patient-reported outcome measures (PROMs) lack robust case-mix adjustment models.

Purpose of the Study:

• To assess the utility of the SSV model for case-mix adjustment of a PROM after stroke.
• To evaluate the SSV model's performance in predicting Subjective Index of Physical and Social Outcome (SIPSO) scores.

Main Methods:

• Utilized data from 176 acute stroke patients in a prospective cohort study.
• Applied SSV model beta coefficients to calculate probabilities of 'good' SIPSO outcome.
• Assessed model discrimination using C statistics and visualized predictions with calibration plots.

Main Results:

• The SSV model achieved C statistics of 0.73 for physical and 0.66 for social SIPSO subscales.
• Including deceased patients improved discrimination to 0.76 and 0.70, respectively.
• Calibration plots indicated a tendency towards over-optimistic predictions with wide confidence limits.

Conclusions:

• The SSV model adequately predicts the physical component of the SIPSO PROM post-stroke.
• The SSV model shows potential for case-mix adjustment of SIPSO in stroke survivors.
• This approach could benefit observational studies, clinical trials, and inter-institutional comparisons.