Comparative epidemiological study in patients with rheumatic diseases illustrated in an example of a treatment with non-steroidal anti-inflammatory drugs versus an oral enzyme combination

To establish the safety and efficacy of an oral enzyme-combination product (OE; Phlogenzym(R), containing trypsin, bromelain and rutin) in the treatme nt of rheumatic diseases a retrolective cohort study with parallel groups w as undertaken as an epidemiological study, in which the enzyme combination was compared with non steroidal anti-inflammatory drugs (NSAID). Data of 33 26 patients treated for rheumatic diseases between January 1993 and the end of March 1995 were registered by 380 physicians. From the patient file age , gender, indication for treatment (diagnostic group), anamnestic data (esp ecially pre-treatment), complaints at the beginning and end of treatment, t reatment duration, prescribed drugs (OE, NSAID), additional treatment and a dverse events were transferred into case report forms (CRFs). The quality o f the data was monitored and additionally checked by internal and external quality audits. Included in the efficacy analysis were 2139 patients which were treated either only with OE or only with NSAID and could be classified unambiguously into one of the following diagnostic groups: joint diseases, spinal diseases, rheumatic soft tissue diseases. As clinically relevant an d reliably evaluable criterion freedom from rheumatic complaints at the end of treatment was considered. For evaluation of safety the documented adver se events of all patients were considered. Two thirds of the OE patients re ceived the recommended dose of 6 tablets/day, taken for 23 to 35 days. The respective mean values for NSAID patients were 16 to 25 days, and the patie nts were treated with the recommended symptomatically effective doses of NS AID. As the allocation of the compared treatment options (OE or NSAID) to the pa tients was not randomized, a mixing of treatment effects with other factors cannot be excluded. For adjustment of these confounding factors two method s were applied: a) logistic regression of the relative ratio of the main cr iterion and all confounding factors and b) stratification of data according to the propensity score i.e. the probability of a treatment with OE. Both methods yielded similar results: A 50 % higher success rate can be expected in total for OE than for NSAID at comparable initial and treatment situati ons (95 % confidence interval with logistic regression = 1,218-1,96, with s tratification according to propensity score 1,16-1,84). As significant nega tive indicators for response age over 50 years, pre-treatment with antirheu matic or analgetic drugs, treatment duration more than 30 days and joint di seases or fibromyalgias could be revealed. Since there was no interaction b etween these indicators and the type of treatment also in patients presenti ng with these indicators a treatment success (freedom from symptoms) with O E can be expected with a higher probability than with NSAID. OE were well tolerated showing much less adverse events when compared with conventional doses of NSAID.