Are slow-acting anti-rheumatic drugs monitored too often? An audit of current clinical practice.

Rheumatologists usually recommend monthly blood monitoring when patients with rheumatoid arthritis (RA) are treated with slow-acting anti-rheumatic drugs (SAARDs). Is monthly monitoring needed or could its frequency be reduced? We audited the opinions of UK rheumatologists and reviewed clinical experience at three centres. To ascertain the interval at which patients are monitored and the determinants of monitoring policy we sent a questionnaire to 193 consultant rheumatologists; 143 (74%) replied. The majority use monthly monitoring for most SAARDs except sulphasalazine, chloroquine and hydroxychloroquine. There is extensive variation, which is not related to the type of rheumatology unit or whether a shared scheme with general practitioners is used. Reviewing experience in 390 patients treated with SAARDs at three adjacent rheumatology units in London showed that haematological adverse reactions were infrequent. During 1560 patient-years of treatment involving 18,720 monthly monitoring visits there were 13 haematological adverse reactions (11 thrombocytopenias and two leucopenias). Five thrombocytopenias developed after 6 months of treatment; five occurred gradually over 5 months or more and one borderline low platelet count was seen once. The two leucopenias were borderline low white cell counts occurring gradually over 3-6 months. Such frequent monitoring is expensive. The total cost of monitoring 390 patients for 1560 patient-years was 420,000 pounds. The cost of detecting each adverse reaction was 32,000 pounds. Three-monthly monitoring when therapy is established after an initial stabilizing period would have identified seven out of eight late adverse reactions. Monitoring policies are mainly based on clinical consensus with few prospective studies of their value; they need re-evaluation.