Risk/benefit analysis of hydroxychloroquine sulfate treatment in rheumatoid arthritis.

Controlled studies have their limitations in assessing the clinical effectiveness of drugs, particularly in patients requiring years of treatment. Retrospective analysis of long-term therapy can be useful and can be summarized easily by life-table analysis. Our first analysis revealed that side effects were more common in patients treated with gold than in those receiving hydroxychloroquine sulfate. Relapses in patients who were treated with gold occurred at a continuous rate over time, whereas in patients treated with hydroxychloroquine this rate leveled off after 20 months of treatment. Our second analysis revealed that the rate of drug discontinuation was lowest with hydroxychloroquine, followed--in order of discontinuation--by gold, D-penicillamine, and levamisole. In both studies, no definite retinopathy was seen in any patient. Since hydroxychloroquine is better tolerated and easier to use than D-penicillamine, gold, or levamisole, it is a reasonable choice for the first trial of a slow-acting antirheumatic drug.