The patient acceptable symptom state for knee pain - a systematic literature review and meta-analysis

Georgopoulos, V.; Smith, S.; McWilliams, D.; Vincent, T.L.; Watt, F.E.; Walsh, D.A.

Abstract

Purpose: Knee pain is highly prevalent, most commonly attributed to osteoarthritis in older people, and in younger people often due to internal derangements. Knee pain can be measured using numerical patient-reported outcome measures (PROMs). Several pain measurement questionnaires have been used for OA pain. These questionnaires each purport to measure the participant’s experience of pain, but may address different pain characteristics (recollection over different time periods, pain impact on function, constant or intermittent or other qualitative aspects of pain). Pooling pain data between studies using different PROMs requires demonstration or transformation to ensure that each PROM would give the same value for pain in a single participant. The Patient Acceptable Symptoms State (PASS) indicates a clinical benchmark that permits comparison between PROMs. Current treatments might relieve but often do not eliminate pain, and PASS is the threshold representing pain which a patient would accept for the remainder of their life. We aimed to systematically review PASS thresholds for different pain PROMs used with people with knee pain, and to identify factors that might influence PASS heterogeneity.
Methods: We systematically reviewed literature for PASS scores in knee pain using searches of CENTRAL, MEDLINE, EMBASE, AMED, CINAHL, and SPORTDiscus databases from their inception date up to June 2020. PROMs of interest were pain-specific questionnaires (or their related domains). Title screening, data extraction, and methodological quality assessments were performed independently by 2 reviewers. Outcome scores were standardised and included in meta-analysis models as a 0-100 scale (0: no pain, 100: highest pain severity). Based on a-priori hypotheses (PROMs, diagnoses, interventions, follow-up timepoints and methodological quality) and following review of data from included studies (PASS score derivation methods), potential effects of study and patient characteristics on PASS were explored. Post-hoc meta-regression explored the relationship between baseline pain and PASS scores. The significance of differences observed between subgroups was evaluated via a Cochran's Q-test. Study heterogeneity was evaluated with the I2 statistic.
Results: Eighteen eligible studies (n=7766 participants) reported PASS from pain PROMs in people with knee pain. All studies were longitudinal and observational, undertaken within the context of a treatment for knee pain. Identified PROMS were the Knee Injury and Osteoarthritis Outcome Score (KOOS), Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC), the Numerical Rating Scale (NRS) and the Visual Analogue Scale (VAS). Most studies were of low risk of bias (8/18), with 6/18 of moderate and 4/18 of high risk of bias. Thirteen studies (n=6339 participants) reported data that allowed their inclusion in metaanalysis models. The pooled pain PASS score was 27 (95%CI: 21 to 35; n=6339 participants) with significant heterogeneity (I2 =96%, p <0.01) (Figure). No significant differences (Q=2.07, p = 0.36) were observed between PASS scores derived for the different knee pain PROMs (KOOS: 23, 95%CI: 16 to 30; WOMAC: 28, 95% CI: 23 to 32; NRS or VAS: 35, 95%CI: 24 to 45). Lower estimates of PASS were associated with lower baseline pain (β=0.60, p=0.02), longer time to follow up at which PASS was estimated (6-months 30, 95%CI: 20 to 40; 12-months: 24, 95%CI: 17 to 30; more than 12-months: 16, 95% CI: 9 to 22), and with surgical (24, 95%CI: 17 to 30) rather than non-surgical interventions (40, 95%CI: 29 to 52). PASS scores were similar between knee osteoarthritis (31, 95%CI: 26 to 36) and meniscal tear (27, 95%CI: 20 to 35) but lower for ligament tears (12, 95%CI: 11 to 13). Observed differences in estimates of PASS due to risk of bias (low: 23, 95%CI: 11 to 35; moderate: 34, 95%CI: 24 to 45; high: 26, 95%CI: 21 to 31) were not significant (Q=1.93, p = 0.38).
Conclusions: Standardised knee pain PROMs scores of approximately 30/100 are considered acceptable by people with knee pain. The level of pain that is acceptable might depend upon the baseline pain severity (higher with worse baseline pain), decrease with time from commencing an intervention and vary according to diagnostic or treatment group. However, different knee pain PROMs when transformed produce similar PASS scores, suggesting that standardised scores derived from multiple instruments might be validly combined in large multicentre studies using historically collected data.

Citation

Georgopoulos, V., Smith, S., McWilliams, D., Vincent, T., Watt, F., & Walsh, D. The patient acceptable symptom state for knee pain - a systematic literature review and meta-analysis. Presented at Osteoarthritis Society International World Congress, Virtual

• Presentation Conference Type: Conference Abstract
• Conference Name: Osteoarthritis Society International World Congress
• Acceptance Date: Apr 1, 2021
• Online Publication Date: Apr 20, 2021
• Publication Date: 2021-04
• Deposit Date: Oct 19, 2022
• Publicly Available Date: Oct 20, 2022
• Print ISSN: 1063-4584
• Publisher: Elsevier
• Peer Reviewed: Peer Reviewed
• Volume: 29
• Issue: Suppl. 1
• Pages: S52
• DOI: https://doi.org/10.1016/j.joca.2021.02.076
• Keywords: Rheumatology; Orthopedics and Sports Medicine; Biomedical Engineering
• Public URL: https://nottingham-repository.worktribe.com/output/5522214
• Publisher URL: https://www.oarsijournal.com/article/S1063-4584(21)00113-8/fulltext
• Additional Information: Abstract from the Virtual 2021 OARSI World Congress on Osteoarthritis, 29 April 2021 - 01 May 2021