International1–3 and Spanish4–7 guidelines for the treatment of tuberculosis (TB) routinely recommend the standard administration of 4 drugs (isoniazid, rifampicin, pyrazinamide, and ethambutol) in the initial or intensive treatment phase of the disease. However, there is some debate as to whether the administration of the fourth drug (ethambutol) in the intensive phase can be avoided, depending on the level of resistance to antituberculous drugs in that region or the type of TB the patient presents (disease site, microbiological data).

The main reason for including ethambutol in the intensive phase is to initiate treatment with a sufficient number of effective drugs, in order to offset any existing resistance to first line drugs and to avoid the development of new resistances. Hence, the most important consideration when prescribing a 4-drug regimen is the national or regional resistance pattern. Nevertheless, in today's globalized world in which people constantly migrate from country to country, it is difficult to predict if a specific patient will be resistant or not. Aside from widespread population movements, TB has widely varying mechanisms for transmission, meaning that individuals with highly resistant strains can transmit the disease to groups traditionally associated with drug sensitive strains, and vice versa. For this reason, international and Spanish guidelines recommend the systematic use of 4 drugs in the intensive treatment of pulmonary TB, at least until the antibiotic resistance profiles have been established.

A recent national multicenter study in tuberculosis resistance in Spain,8 found 6.7% cases of isoniazid-resistant and 1.9% cases of multi-drug resistant (MDR) TB (resistant to at least hydrazides and rifampicin). Moreover, a preliminary analysis of the use of 3 or 4 drugs in the intensive treatment of TB suggested that a high percentage of cases in Spain were treated with 3 drugs,9 despite the indications given in different guidelines.

In view of this situation, we identified the need for a study to determine if the guideline recommendations for intensive TB treatment with 4 drugs were being followed in Spain, and to identify the factors associated with prescribing only 3 drugs (standard regimen without ethambutol) in the intensive phase.

A total of 3189 patients were included in this study. The population was predominantly male (1940; 60.8%), and 1122 (35.2%) were immigrants. Disease site was the lung in 2476 (77.6%) cases. A total of 1669 (52.3%) had positive sputum smear, 907 (28.4%) had negative sputum smear with positive culture, and 524 (16.4%) had negative sputum smear and negative culture. A total of 2981 (93.5%) had no resistance to any drug, and 208 (6.5%) had resistance to at least 1 drug. A total of 1776 (55.7%) were treated with 4 drugs and 1413 (44.3%) received 3 (Table 1).

The results of the statistical analysis of the characteristics of the cases treated with 3 or 4 drugs and the factors associated with the use of only 3 drugs are shown in Tables 1 and 2, respectively. It is interesting to note that the 4-drug regimen was administered to 982 (58.8%) patients with a positive sputum smear, to 844 (75.2%) immigrants, to 94 (68.6%) HIV-infected individuals, to 293 (77.8%) subjects who lived in a group, and to 133 (63.9%) patients with resistance to at least 1 drug. The 3-drug combination was administered to 687 (41.2%) patients who had a positive sputum smear, 278 (24.8%) immigrant patients, 43 (31.4%) HIV-infected individuals, and 75 (36.1%) patients who were resistant to at least 1 drug.

The multivariate analysis confirmed that administration of the 3-drug combination was associated with being female; non-immigrant; retired; homeless, living alone or in a family (using living in a group as reference); having been treated in specialist units (using treatment in the emergency room as reference); no HIV infection; and negative sputum smear and positive culture (Table 3).

Treatment trends (3 drugs vs 4 drugs) over the years (Fig. 2 and Table 1) show a marked increase in the proportion of cases treated with 4 drugs in 2008 and 2009. After that, rates fall steadily, and by 2012 the proportion of patients receiving 3 drugs was even higher that those receiving 4.

Fig. 2.

Number of patients treated annually with 3 or 4 drugs.

(0.07MB).

No significant differences were observed in the clinical progress or final treatment outcome between patients receiving 3 or 4 drugs. Treatment was considered satisfactory (cured or treatment completed) in 79.9% of patients who received 3 drugs and in 76.2% of those who received 4. Treatment outcome was satisfactory in 73% of patients who had at least 1 resistance and received 3 drugs, and in 81% of patients with resistances who received 4 drugs. The difference was not significant (P=.18).

Finally, more patients in the drug-resistant group received prolonged treatment than in the group of patients with sensitive strains (13% vs 3.1%; P<.001).

This study in TB patients from practically all the Autonomous Communities of Spain showed that a significant proportion of subjects, some of whom are at risk of developing resistance, were not treated in accordance with the Spanish and international guidelines for intensive 4-drug therapy.

A total of 58% of patients with positive sputum smear received 4 drugs. However, the remaining TB carriers were in a precarious situation, since they could well have had resistant strains that would not be determined until weeks later, at which stage the treatment would have to be changed. Moreover, new resistances may even have been activated and/or transmitted, with the risk of generating cases of MDR TB and extremely drug resistant (XDR) TB (defined as MDR TB associated with resistance to quinolones and at least 1 second-line injectable). This risk may be minimized, at least in part, by the generalized use of new molecular techniques for diagnosing resistance, such as Xpert MTB/RIF assay or GeneXpert®,10 which can detect rifampicin resistance, thus suggesting MDR TB, at the time of diagnosis and start of treatment, or GenoType MTBDRplus11 for the rapid detection of rifampicin and hydrazide resistance.

More patients with negative sputum smear (53.6%) received 3 drugs rather than 4 drugs. The reason for this may be purely practical: after receiving a positive result from microbiology, the treating clinician knows that the lab will shortly send the resistance profile. Therefore, if the patient does have a resistant strain, there will be insufficient time to develop new resistances due to inappropriate treatment.

It is interesting to note that in our study, only 75.2% of the immigrant population received the 4-drug regimen, leaving a quarter of them to receive 3 drugs. This practice is clinically inacceptable, since the resistance rates in immigrant patients are known to be higher than in Spanish nationals (11.9% isoniazid resistance in immigrants, compared to 4.1% in nationals, according to a recent Spanish study).8 Moreover, the percentage of foreign patients among the overall population of TB patients in Spain is 29.7%, according to a previous study,12 and 35% according to our study, underlining the inadvisability and high risk of initiating 3-drug treatment.

Spanish nationals received 3 drugs (54.9%) more often than 4, which may be a reflection of the physicians’ confidence in reports of low resistance rates in this country. However, recent data published by our group8 reveal a higher than expected rate of resistance in Spanish nationals (4.1% isoniazid resistance), suggesting that it is advisable to treat patients with 4 drugs right from the start.

Interestingly, 75 patients with resistance to first-line drugs (36.1% of the total) began treatment with 3 drugs. This may cause significant therapeutic problems for the afore-mentioned reasons and the treatment regimen may need to be amended if the resistance profile results are positive. This further highlights the need to start treatment with 4 drugs, since any patient can present drug-resistance that hampers treatment and can even extend the contagion period by delaying the effect of the treatment. The similar percentage of satisfactory treatment outcomes in patients with sensitive or resistant strains receiving 3 or 4 drugs may be explained by the fact that the Spanish TB physicians who participated in this study are specialists with extensive experience in managing this disease. Therefore, it is reasonable to assume that if resistances had developed, the physicians would have prescribed a new treatment regimen in accordance with the developing resistance, thus achieving a good final outcome for their patients. This is another reason why treatment was extended in more resistant cases (13%) than in cases with sensitive strains (3.1%).

Trends in the use of the 3- or 4-drug regimen over the years of the study show an increase in the use of 4 drugs in 2008 and 2009. This may have been due to the influence of the recent (at that time) publication of the Spanish guidelines. However, rates have subsequently diminished, and by 2012 the proportion of patients receiving 3 drugs was even higher that those receiving 4, showing a growing trend towards ignoring TB guidelines. These data are worrying when viewed alongside the recent increase (according to the latest 2 national multicenter studies)8,13 in both initial isoniazid resistance and MDR TB. Moreover, no specific studies have been performed in the Spanish Autonomous Communities which would justify the administration of 3- or 4-drug therapies on the basis of local resistance and MDR figures. The introduction of Xpert MTB/RIF and other nucleic acid amplification techniques, mentioned above, may change TB morbidity and mortality by improving diagnosis through early detection of resistance. Prompt treatment of new cases will have a significant impact on the dynamics of long-term transmission.11

Very few international studies have evaluated compliance with treatment recommendations in the intensive phase of TB. A Malaysian study reported that only 49.4% of patients completed the 2-month intensive phase, despite the recommendations of the World Health Organization (WHO).14 In France, a significant number of doctors (12.2%) responding to a national questionnaire reported that they systematically used a 3-drug regimen in the initial treatment, instead of the recommended 4 drugs, but the proportion of patients receiving the 3-drug regimen could not be retrieved from the data, since each doctor treated a different number of patients. The authors concluded that efforts needed to be made to achieve compliance with treatment guidelines.15 In India, the Revised National Tuberculosis Control Programme uses an initial regimen, consisting of rifampicin administered intermittently 3 times a week, including HIV-infected patients, while the WHO recommends daily treatment at least in the intensive disease phase: the mortality rate in these patients was high.16

One of the limitations of our study is that we could not show if the use of 3 drugs generated cases of resistance, since no clinical–epidemiological or molecular disease transmission studies were performed. Nor could we demonstrate a difference in the final classification or outcome of patients treated with 3 or 4 drugs, possibly because the level of resistance in Spain is low, and because treatment was administered by physicians with expertise in the treatment of TB: if resistance was determined in the initial tests, they may have decided to switch to an effective regimen. The strength of the study is that it reflects the standard practice of a large group of TB specialists representing practically all the Spanish Autonomous Communities.

Despite current TB treatment guidelines, a significant proportion of patients treated in the intensive phase receive 3 drugs (44.3%). It seems highly likely that a significant proportion of immigrants or non-Spanish natives, patients with positive sputum smears, and patients with initial resistance may have initially received the 3-drug regimen, exposing them to the risk of developing resistance. It is worrying to note that in the last year of the study, more patients received 3 drugs than 4.

For these reasons, we recommend raising awareness of the guidelines, and call on the Plan for Prevention of Control of Tuberculosis in Spain to put in place systems for monitoring compliance with the recommendations of national and international TB treatment guidelines. Four drugs must be prescribed in the intensive phase of the disease, unless results are available from rapid resistance tests in a matter of hours.

JMGG, TR, JAC: study conception. JMGG, TR, JAC: drafting the manuscript. MC: statistical analysis. JMGG, TR, JC, MC: analysis of results. JMGG, TR, MC, JAC, TPP, JRM: critical reading, review and final approval of the manuscript. PIITB Working Group: data collection and final approval of the manuscript.

The authors declare that they have no conflict of interests.