Background/Purpose:

Patient-reported outcomes (PRO) on Health Related Quality of Life (HRQoL) are increasingly used to assess responsiveness after treatment.  This study aimed to determine the ability of different PRO to asses responsiveness in psoriatic arthritis (PsA) after the initiation of anti-TNF therapy.

Methods:

A retrospective analysis was conducted on all PsA patients followed at the PsA clinic who had started and continued anti-TNF therapy and had at least 1 year of follow up.

PRO questionnaires administered at baseline and follow up visits included: patient global assessment (PGA) [very good, good, fair, poor and very poor], Short-Form 36 (SF-36), Health Assessment Questionnaire (HAQ), Functional Assessment of Chronic Illness Therapy (FACIT), Fatigue Severity Scale (FSS), EuroQol-5D (EQ-5D)and Dermatology Life Quality Index (DLQI). PGA after treatment was used as an external anchor to group the patients as: Improved, same or worse.

Ceiling and floor effects were checked at baseline and follow up visits. Distribution-based methods (effect size [ES], standardized response mean [SRM], Guyatt’s responsiveness statistic [GRS]) and anchor-based methods (Spearman correlation of the change scores of PRO) were used to assess responsiveness.

A priori hypothesis was that moderate to large responsiveness scores would be observed for the patients who improved or worsened.

Results:

223 patients with active PsA (62% male and 90% Caucasian) were identified. Disease duration at baseline visit was 12.9 (9.2). At follow up visit, 70 patients improved, 126 remained the same and 27 worsened based on PGA.

Ceiling effect:  1) At baseline, a ceiling effect was identified for HAQ, and 3 domains of SF-36 (Role Physical, Social Functioning and Role Emotional) and 2) At follow up, a ceiling effect was identified for HAQ, 4 domains of SF-36 (Physical Functioning, Role Physical, Social Functioning and Role Emotional) and DLQI. 

Floor effect: At baseline and follow up, a floor effect was identified for 2 domains of SF-36 (Role Physical and Role Emotional).

In particular SF-36, FACIT, DLQI and FSS indicated moderate to large responsiveness in the patients who improved and worsened (SRM 0.51-1.07 and Guyatt 0.64-3.06). A small SRM (0.25) was observed for HAQ in patients who improved. HAQ and FACIT were not responsive for patients who worsened.

At follow up, strong correlation was observed between the change scores: 1) Δ FACIT and Δ FSS (r= -0.82, p= 0.007) in patients who improved and 2) Δ FACIT and Δ HAQ (r= -0.91, p= 0.004) and Δ EQ-5D and Δ SF-36 MCS (0.80, p= 0.02) in patients who worsened.

Conclusion:

SF-36 and DLQI are responsive questionnaires able to capture improvement and worsening in response to treatment based on patient judgment.

HAQ is extremely nonresponsive and EQ-5D was the worst responsive compared to the other questionnaires.  FACIT was responsive and superior to FSS for measuring fatigue in patients who improved on biologics but not for measuring worsening. A larger sample size is needed to decide on the exclusion of either FACIT or FSS.

---

ACR Meeting Abstracts - https://acrabstracts.org/abstract/responsiveness-of-health-related-quality-of-life-questionnaires-to-treatment-with-anti-tnf-therapy-in-psoriatic-arthritis/