Current issues of ACP Journal Club are published in Annals of Internal Medicine


Separate drug formulations were more effective than combined isoniazid, rifampin, and pyrazinamide for pulmonary tuberculosis

ACP J Club. 1991 Sept-Oct;115:37. doi:10.7326/ACPJC-1991-115-2-037

Source Citation

Singapore Tuberculosis Service/British Medical Research Council. Assessment of a daily combined preparation of isoniazid, rifampin, and pyrazinamide in a controlled trial of three 6-month regimens for smear-positive pulmonary tuberculosis. Am Rev Respir Dis. 1991 April;143:707-12.



To compare the efficacy and acceptability of combined with separate formulations of isoniazid, rifampin, and pyrazinamide for the treatment of sputum smear-positive pulmonary tuberculosis.


Randomized controlled trial.


Singapore Tuberculosis Service.


310 patients met the inclusion criteria of age ≥ 15 years, pulmonary tuberculosis with sputum positive for acid-fast bacilli and yielding Mycobacterium tuberculosis on culture, and no previous antituberculosis chemotherapy. 271 of the patients had fully drug-susceptible bacilli.


Patients were allocated to 1 of 3 treatments and to combined or separate formulations. Initial phases all included isoniazid (200 to 300 mg/d), rifampin (480 to 780 mg/d), and pyrazinamide (1200 to 2000 mg/d). These were given for 2 months with streptomycin (750 mg/d) or for 1 month with streptomycin (750 mg/d) or for 2 months without streptomycin. All continuation phases consisted of isoniazid (600 to 1000 mg) and rifampin (600 mg) 3 times/wk for 6 months.

Main outcome measures

Negative culture for tuberculosis from sputum samples taken each month for 18 months, then every 3 months; and complaints made spontaneously during treatment visits or elicited at assessments done monthly or as needed.

Main results

22 patients were not included in the analysis because of adverse reactions (n = 16), noncompliance (n = 3), or other reasons (n = 3). Spontaneous complaints were made in the first month by 13 of 155 patients (8%) and 15 of 155 patients (10%) who received combined and separate formulations, respectively. {Absolute risk difference 2%, 95% CI -5% to 8%, P = 0.7}.* The frequency of adverse reactions was similar for treatments without streptomycin, and for both formulations. After 3 months, all but 1 fully drug-susceptible patient were culture negative. No bacteriologic failures occurred during chemotherapy. During 18 months of follow-up, more bacteriologic relapses occurred among patients receiving the combined formulation (8 of 128 patients, 6%) compared with those receiving the separate formulation (2 of 137 patients, 1%, P = 0.04). {This absolute risk difference of 5% means that 1 additional relapse occurred for every 21 patients who received the combined formulation (compared with the separate formulation), CI 9 to 426; the relative risk increase was 328%, CI 5% to 1662%}.* Patients receiving 1 or 2 months of streptomycin initially experienced similar relapse rates (P > 0.5).


6 months of treatment including an initial 1 or 2 months of treatment with a combined formulation of isoniazid, rifampin, and pyrazinamide was less effective than similar 6-month regimens using separate drug formulations for the treatment of pulmonary tuberculosis. Acceptability was no different between formulations.

Source of funding: Singapore Ministry of Health.

Address for article reprint: Dr. D.J. Girling, MRC Cardiothoracic Epidemiology Group, Brompton Hospital, Fulham Road, London SW3 6HP England, UK.

*Numbers calculated from data in article.


Isoniazid, rifampin, and streptomycin for 6 months plus pyrazinamide for 2 months increased culture negativity in pulmonary tuberculosis

These 2 studies are part of an ongoing research program of the Medical Research Council Cardiothoracic Epidemiology Group (British) evaluating effectiveness of short-course, multiple-drug combination chemotherapy in active pulmonary tuberculosis. An additional aim of these studies was to assess patient acceptance of a combined formulation of isoniazid, rifampin, and pyrazinamide compared with the individual drugs taken separately. The settings for the studies were Hong Kong and Singapore. Patient demographics were similar for the 2 studies. As would be expected from studies coordinated by a single group, the general design and methods were also similar.

In the Hong Kong study, all mycobacteriologic studies were done in London, England. In Singapore, entry laboratory studies for tuberculosis were done by both the London laboratory and by a facility in Singapore. After chemotherapy, any sputum smear or culture with a positive result in Singapore was repeated by the London laboratory. It would have been illustrative for the authors to have given the interlaboratory agreement between London and Singapore and the manner in which discrepancies were dealt with. If the agreement was not 100%, how confident were the authors in the bacteriologically defined relapse rate (especially if the Singapore laboratory produced false-negative results)? It is also not clear from either article how patients with postive sputum smears who did not subsequently have a positive culture for Mycobacterium tuberculosis (either on entry or during follow-up) were handled in the analysis. This information would be important for the specimens that were transported from Hong Kong to London, because the organism may have died in transit, yet still be seen on staining of the sputum smear but not be recoverable on culture. The last point concerning the laboratory component of both studies is the definition of bacteriologic relapse after chemotherapy " a culture growing 10 or more colonies in 2 different months during any consecutive 3 months up to 24 months or 6 months thereafter (Hong Kong) or to 18 months or 4 months thereafter (Singapore)." The authors did not comment on this breakpoint of colonies and time frame with respect to the clinical significance of any positive Mycobacterium tuberculosis culture after chemotherapy. The distribution of all the quantified positive cultures would be an interesting feature in the effectiveness analysis of these studies.

Although these laboratory points may affect the magnitude of the observed effectiveness, they would be unlikely to affect the intertreatment group comparisons because the laboratory effects should be evenly distributed. Hence, these studies, and presumably the studies to follow from this group, add to the overall knowledge base for the treatment of active pulmonary tuberculosis in southeast Asia and elsewhere, including North America, where the combined issues of drug cost and availability, compliance and tolerance with prescribed therapy, and clinical follow-up must be balanced with treatment efficacy, emergence of bacterial resistance to single and multiple agents, and the public health issues related to spread of tuberculosis.

The results related to the 2 combined formulations of isoniazid, rifampin, and pyrazinamide used in these studies (Rifater 2 in Singapore and Rifater 3 in Hong Kong) are encouraging in that they were subjectively well accepted by the patients. It is unfortunate that the former product appeared to be less effective than the agents used individually because the latter agent is reported by the authors to be no longer available.

Douglas MacPherson, MD
McMaster UniversityHamilton, Ontario, Canada