Prognostic factors associated with changes in knee pain outcomes, identified from initial primary care consultation data. A systematic literature review

Collier, TS, Hughes, T, Chester, R, Callaghan, MJ ORCID: https://orcid.org/0000-0003-3540-2838 and Selfe, J ORCID: https://orcid.org/0000-0001-9931-4998 (2023) Prognostic factors associated with changes in knee pain outcomes, identified from initial primary care consultation data. A systematic literature review. Annals of Medicine, 55 (1). pp. 401-418. ISSN 0785-3890

Preview

Published Version Available under License Creative Commons Attribution. Download (1MB) | Preview

Abstract

Background: Data collected during initial primary care consultations could be a source of baseline prognostic factors associated with changes in outcome measures for patients with knee pain. Objectives: To identify, appraise and synthesize studies investigating prognostic factors associated with changes in outcome for people presenting with knee pain in primary care. Methods: EMBASE, CINAHL, AMED, MEDLINE and MedRxiv electronic databases were searched from inception to March 2021 and repeated in August 2022. Prospective cohort studies of adult participants with musculoskeletal knee pain assessing the association between putative prognostic factors and outcomes in primary care were included. The Quality in Prognostic Studies (QUIPS) tool and The Modified Grading of Recommendations Assessment, Development and Evaluation (GRADE) framework, specific to prognostic reviews were used to appraise and synthesize the evidence respectively. Results: Eight studies were included. Eight knee pain outcomes were identified. Methodological and statistical heterogeneity resulted in qualitative analysis. All evidence was judged to be of low to very low quality. Bilateral knee pain (multivariable odds ratio (OR) range 2.60–2.74; 95%CI range 0.90–8.10, p value = 0.09) and a lower educational level (multivariable (OR) range 1.74–5.6; 95%CI range 1.16–16.20, p value = <0.001) were synonymously associated with persisting knee pain at 12-month follow up. A total of 37 univariable and 63 multivariable prognostic factors were statistically associated with outcomes (p ≤ 0.05) in single studies. Conclusions: There was consensus from two independent studies that bilateral knee pain and lower educational level were associated with persistent knee pain. Many baseline factors were associated with outcome in individual studies but not consistently between studies. The current understanding, accuracy and reliability of the prognostic value of initial primary care consultation data for knee pain outcomes are limited. This review will provide an essential guide for candidate variable selection in future primary care prognostic confirmatory studies.Key messages Bilateral knee pain and lower educational level were associated with persistent knee pain. Many baseline factors were associated with outcome in individual studies but not consistently between studies. The current understanding, accuracy and reliability of the prognostic value of initial primary care consultation data for knee pain outcomes are limited.