Aligning monitoring with risk: what the new British Society for Rheumatology guidance means for practice

The 2025 British Society for Rheumatology (BSR) guideline for the prescription and monitoring of conventional synthetic DMARDs (csDMARDs) introduces risk-stratified monitoring. This marks a shift from the 2017 guideline, which established a largely standardized monitoring framework applied uniformly to all patients. The rationale for this change is simple. Under uniform monitoring, a patient with normal renal function and years of stable therapy receives the same surveillance schedule as someone with stage 3 chronic kidney disease on combination csDMARD therapy. This one-size-fits-all approach, while simple to implement, does not reflect the heterogeneity of patients prescribed these medications. Risk-stratified monitoring tailors testing intervals to a patient’s likelihood of toxicity and reflects the wider move toward personalized medicine that is reshaping many areas of rheumatology practice. There is substantial evidence supporting this change. Randomized trials comparing DMARD-monitoring strategies are still lacking, a limitation noted in the 2017 guideline and yet to be addressed. However, prognostic models using primary care electronic health record data from the Clinical Practice Research Datalink have now been developed. Nakafero, Abhishek and colleagues have validated prediction models for MTX, LEF, thiopurines and SSZ that estimate the probability of treatment discontinuation due to abnormal monitoring blood tests [1–4]. The MTX model, developed and validated in >37 000 patients, performed consistently across age groups, disease types and dose ranges. Their analyses indicate that risk stratification could allow reduced monitoring for roughly two-thirds of those prescribed MTX while identifying a subgroup who may benefit from more intensive surveillance [1]. The COVID-19 pandemic offered an unplanned natural experiment in relaxed monitoring. When the BSR recommended a temporary extension of monitoring intervals for stable patients, analysis of 24.2 million patient records through OpenSAFELY found that while missed monitoring increased during the first lockdown, it recovered rapidly as lockdown measures were eased [5]. Audits of delayed MTX monitoring found no evidence of excess harm. These observations were reinforced by a Dutch cohort study published in 2025, which analysed 59 555 routine monitoring test sets from 4774 patients over 18 383 patient-years and found that very abnormal results occurred in only 0.15% of tests, of which 36% did not prompt any change in management. The authors concluded that the value of frequent, non-targeted monitoring in stable patients warrants reconsideration [6]. It was the triangulation of these data that informed the working group’s decision. Substantial input from stakeholders shaped the final recommendations; primary care voices were essential given that much csDMARD monitoring occurs in general practice, and patients contributed important perspectives on the burden of frequent testing vs the reassurance it can provide. The guideline recommends that someone starting a csDMARD such as MTX, who has no risk factors for toxicity, may be monitored with blood tests once or twice a year once on stable treatment. Several factors warrant increased surveillance. Chronic kidney disease stage 3 or worse is a consistent predictor of MTX toxicity, reflecting the drug’s renal excretion. Older age, obesity, hazardous alcohol consumption (>21 units/week) and pre-existing liver disease may contribute to risk, but no single factor in isolation is sufficient, and risk should be assessed on an individual basis. Combination therapy with MTX and LEF is associated with higher rates of liver enzyme abnormalities than either agent alone, and the presence of other medicines such as antiepileptics may raise the risk further. For people with these risk factors, the guideline recommends intensifying monitoring frequency. Reduced monitoring for patients at low risk means fewer phlebotomy appointments, less time away from work and lower personal costs. For the National Health Service, potential savings include reduced pressure on phlebotomy services, reduced laboratory processing and less clinician time spent reviewing results. These benefits must be balanced against the need to preserve capacity for enhanced monitoring where it is indicated. The guideline recommends annual reassessment of risk status. Some may view this as additional burden that will fall on general practice; however, this is not a new expectation. Reviewing whether a patient on csDMARD has developed new risk factors for toxicity such as declining renal function or interacting medicines is a routine part of good prescribing practice. Structured medication reviews, increasingly undertaken by clinical pharmacists in primary care, provide a practical and well-established setting for this reassessment. In effect, the guideline formalizes what should already be standard practice. Effective risk stratification depends on an accurate baseline assessment, and a person’s risk profile may evolve over time. The recommendation to relax monitoring assumes stable disease and consistent adherence. Clinicians and patients should remain alert to intercurrent illness or changes in circumstance that warrant more intensive surveillance. The 2025 BSR guideline applies risk stratification to csDMARD monitoring, drawing on prognostic models and observational evidence accumulated over the past decade. It recognizes that patients differ and that monitoring should reflect this. Implementation requires clinicians to assess individual risk and adjust monitoring accordingly. Whether this approach improves outcomes or resource use compared with past uniform schedules remains to be formally tested, but it represents a move toward care that is better aligned with individual need. No new data were generated or analysed in support of this research. No specific funding was received from any bodies in the public, commercial or not-for-profit sectors to carry out the work described in this article. Disclosure statement: L.M. has received support to attend conferences from Nordic Pharma and UCB. K.B. has received institutional research grants from Pfizer, conference registration fees from UCB Pharma, and honoraria from UCB, Vifor Pharma and Galapagos. J.G. has received fees for advisory work from AbbVie, Galapagos, Janssen, Eli Lilly, Novartis, Pfizer and UCB. J.G. also holds the role of Rheumatology Co-Editor and has not peer reviewed or made any editorial decisions for this article. We would like to acknowledge the contributions of the csDMARD Guideline Working Group, whose expertise, discussion and careful review informed the development of this guidance.