Purpose: Current treatment guidelines are based on objective clinical outcomes, however there is an ongoing trend to integrate information from patient self-reporting instruments in treatment decisions. This is illustrated when patient self-report instruments prompt clinicians to deviate from guidelines that are based on clinical objective measures. The purpose of this study was to investigate the equivalence of an objective clinical outcome and a patient assessed outcome. Additionally the influence of a patient assessed outcome on treatment decisions was investigated.

Method: The baseline and three month follow-up data from a daily clinical practice cohort of patients with rheumatoid arthritis who started on a TNFα blocking agent was used. Patients were classified as responders or non-responders according to an objective, disease-specific clinical outcome measure. A visual analogue scale (range 0-100mm) for general health (GH) was used as a patient assessed outcome measure. Differences were calculated between baseline and three months values. In order to assess equivalency of objective and subjective measures, a student's t-test was performed, identifying significant differences in subjective outcome between responders and non-responders. Furthermore percentage of guideline adherence was determined. Differences in subjective outcome were investigated between patients where guidelines were, and were not, followed.

Results: At three months 45.5% of the patients was classified as responder and 54.5% as non-responder. Responders reported a 31 [95% CI: 27 ; 36] GH improvement compared to 5 [95% CI: 1 ; 9] for non-responders. Of the non-responding patients 64.6% continued initial treatment. These clinically non-responding patients had reported a 10 [95% CI: 6 ;15] GH improvement, compared to a non-significant difference in GH [mean: -5 95% CI: -12 ; 2] of non-responding patients for whom guidelines were followed.

Conclusion: On average both clinical responders and non-responders reported an improvement on self-perceived general health. Apparently clinical objective outcomes and patient self-reported outcomes are not substitutable in regard to treatment decisions, however they may be complimentary. Furthermore in most patients whose self-reports coincided with objective measures, clinicians adhered to objective guidelines, but guidelines where not followed for non-responding patients who predominantly perceived improvement in subjective outcome. This demonstrates that integration of patient self-report measures in treatment decisions can lead to contra-indicated treatment continuation.