Table 4

Summary of evidence for disease-related and psychological predictors of adherence to MTX

PredictorStudyNAdherence outcomeUnadjusted effect size (95% CI)/univariate analysesp ValueAdjusted effect size (95% CI)p Value
RA duration
YearsSalt and Frazier29*108Non-adherent (MARS ≤38)NPNPOR 1.00 (1.01 to 1.00)0.83
Waimann et al30107Adherence (MEMS)r 0.08>0.20NPNP
1–5de Thurah et al26941Non-adherence (CMG)β 0.03 (0.01 to 0.06)Sigβ 0.01 (−0.01 to 0.04)NS
>5de Thurah et al26941Non-adherence (CMG)β 0.02 (−0.01 to 0.04)NSβ −0.04 (−0.07 to −0.02)Sig
>5de Thurah et al2785Non-adherence (CQR ≤25th centile) at BLPR 1.7 (0.7 to 4.1)NSPR 1.5 (0.5 to 4.7)NS
>5de Thurah et al2765Non-adherence (CQR ≤25th centile) at 9 moPR 1.5 (0.6 to 3.6)NSPR 1.2 (0.4 to 3.1)NS
Inflammatory biomarkers
CRPContreras-Yanez et al24*93Adherent (CQ ≥9) vs non-adherent (CQ ≤8)2.4±2.6 vs 2.6±2.40.68NPNP
CRP 8–32de Thurah et al26941Non-adherence (CMG)β 0.00 (−0.02 to 0.02)NSβ −0.02 (−0.04 to 0.01)NS
CRP >32de Thurah et al26941Non-adherence (CMG)β −0.02 (−0.05 to 0.01)NSβ −0.04 (−0.07 to −0.02)Sig
Erythrocyte sedimentation rateContreras-Yanez et al24*93Adherent (CQ ≥9) vs non-adherent (CQ ≤8)24.1±17.4 37.5±23.80.003NPNP
Disease Activity Score-28
Contreras-Yanez et al24*93Adherent (CQ ≥9) vs non-adherent (CQ ≤8)3.6±1.3 vs 5.1±1.9≤0.001NPNP
Waimann et al3090Per cent of adherent (MEMS)r −0.270.01NPNP
Sharp scoreWaimann et al3079Per cent of adherent (MEMS)r −0.06>0.20NPNP
Functional ability
HAQde Klerk et al22*,‡127Per cent of adherence (MEMS)ANOVA (no data)NSNPNP
Contreras-Yanez et al24*93Adherent (CQ ≥9) vs non-adherent (CQ ≤8)0.2±0.4 vs 0.4±0.50.04NPNP
HAQ >1.75de Thurah et al2785Nonadherence (CQR ≤25th centile)PR 1.2 (0.5 to 2.5)NSPR 1.4 (0.6 to 3.1)NS
0.75–1.75de Thurah et al2765Non-adherence (CQR ≤25th centile) at BLPR 1.5 (0.5 to 4.9)NSPR 0.8 (0.2 to 3.3)NS
HAQ >1.75de Thurah et al2765Non-adherence (CQR ≤25th centile) at 9 moPR 0.8 (0.3 to 2.5)NSPR 1.0 (0.2 to 3.4)NS
Modified—HAQWaimann et al30107Per cent of adherence (MEMS)r −0.200.04NPNP
Comorbidity
Number of comorbiditiesWaimann et al30107Per cent of adherence (MEMS)r −0.06>0.20NPNP
Per cent with comorbidityContreras-Yanez et al 24*93Adherent (CQ ≥9) vs non-adherent (CQ ≤8)40% (85) vs 36% (78.3)0.43NPNP
Any vs nonede Thurah et al2785Non-adherence (CQR ≤25th centile) at BLPR 1.3 (0.4 to 3.9)NSPR 1.1 (0.4 to 3.3)NS
de Thurah et al2765Non-adherence (CQR ≤25th centile) at 9 moPR 1.3 (0.4 to 3.8)NSPR 2.2 (0.5 to 9.7)NS
COPDde Thurah et al26941Non-adherence (CMG)β 0.00 (−0.04 to 0.04)NSβ 0.04 (0.00 to 0.07)Sig
Diabetesde Thurah et al26941Non-adherence (CMG)β −0.04 (−0.1 to 0.02)NSβ 0.00 (−0.05 to 0.05)NS
Liver diseasede Thurah et al26941Non-adherence (CMG)β 0.06 (0.02 to 0.10)Sigβ 0.04 (0.00 to 0.08)Sig
BMQ low concern about MTXde Thurah et al2785Non-adherence (CQR ≤25th centile) at BLPR 0.8 (0.4 to 1.8)NSPR 0.7 (0.3 to 1.8)NS
de Thurah et al2765Non-adherence (CQR ≤25th centile) at 9 moPR 0.5 (0.2 to 1.2)NSPR 0.5 (0.2 to 1.3)NS
BMQ high perceptions of MTX necessityde Thurah et al2785Non-adherence (CQR ≤25th centile) at BLPR 0.4 (0.1 to 0.8)SigPR 0.3 (0.1 to 0.8)Sig
de Thurah et al2765Non-adherence (CQR ≤25th centile) at 9 moPR 0.2 (0.1 to 0.6)SigPR 0.4 (0.1 to 1.1)NS
LTMBS (self-efficacy)de Klerk et al22127Per cent of adherence (MEMS)F 5.90.02NPNP
UCL avoidant copingde Klerk et al22*,‡127Per cent of adherence (MEMS)NPNSβ −0.41< 0.05
Passive reactive coping
UCL passive copingde Klerk et al22*,‡127Per cent of adherence (MEMS)NPNSβ 0.79<0.05
UCL reactive copingde Klerk et al22*,‡127Per cent of adherence (MEMS)NPNSβ 0.4<0.05
UCL active copingde Klerk et al22*,‡127Per cent of adherence (MEMS)NPNSNPNP
UCL reassuring thoughtsde Klerk et al22*,‡125Per cent of adherence (MEMS)NPNSNPNP
Mental health
CES-D10Waimann et al30107Per cent of adherence (MEMS)r −0.190.05NPNP
MOS SF-12 MCSWaimann et al30107Per cent of adherence (MEMS)r 0.34<0.01NPNP
MOS social supportWaimann et al30107Per cent of adherence (MEMS)r 0.170.08NPNP
Health-related quality of life
European Quality of Life Measurede Klerk et al22*,‡127Per cent of adherence (MEMS)F 5.42<0.01NPNP
RA Quality of Life Measurede Klerk et al22*,‡81Per cent of adherence (MEMS)F 0.210.65NPNP
Nottingham Health Profilede Klerk et al22*,‡127Per cent of adherence (MEMS)NPNSβ −0.62<0.05
MOS SF-12 Physical Component SummaryWaimann et al30107Per cent of adherence (MEMS)r 0.07>0.20NPNP
  • *Studies judged low quality.

  • †MTX adherence.

  • ‡Includes RA, PMR and gout.

  • 9 mo, 9 months; ANOVA, analysis of variance; BL, baseline; BMQ, Beliefs in Medicines Questionnaire; CES-D10, Centre of Epidemiologic Studies Depression Scale; CMG, continuous medication gap; COPD, chronic obstructive pulmonary disease; CQ, Compliance Questionnaire; CQR, Compliance Questionnaire—Rheumatology; CRP, C reactive protein; F, ANOVA test statistic; HAQ, Health Assessment Questionnaire; MARS, Medication Adherence Revised Scale; MEMS, Medicine Event Monitoring System; MOS, Medical Outcomes Study; MTX, methotrexate; NP, not presented; NS, non-significant; PR, prevalence ratio; r, Pearson correlation coefficient; RA, rheumatoid arthritis; SF-12 MCS, Mood Component Summary of MOS 12-item Short Form Health Survey; Sig, significant; UCL, Utrecht Coping List; β, regression coefficient.