Tuberculosis (TB) is one of the major infectious diseases in severity and morbidity despite coverage of the National Tuberculosis Program and improved social and economic conditions¹⁾. Recently, epidemic of acquired immune deficiency syndromes, poor isolation of drug-resistant subjects and increased numbers of poor and crowded residences have caused increased proportion of drug resistance (DR), especially multi-drug resistance (MDR)^2–7). In Korea, although the overall prevalence of tuberculosis diminished to less than 1.0%, the rate of drug resistance, especially MDR in previously treated groups was still high^{8, 9)}. It is important to define the risk factors, including the socioeconomic status of DR for the prompt isolation and proper management of drug-resistant tuberculosis. We examined current characteristics of DR in a university hospital setting and compared it to national studies.

This study presented findings related to drug resistance in a TB endemic country not associated with HIV epidemic. The drug resistance in TB is one of the most important problems in the world and a major cause of morbidity and mortality²⁾. TB was one of the most common 10 causes of Korean adults’ deaths until 1996⁸⁾. DR and severity of infection are important factors determining the mortality in TB patients. The national goal of TB control is closely related to control of DR. According to a national TB survey of Korea, the rate of DR was the highest in 1980 and then it tended to decline^{8, 9)}.

The rate of isoniazid-resistant TB was the highest among the antituberculous drugs throughout the world and the rate of rifampin-resistant TB was high in an endemic area^5,11–14). This study showed that nearly 20 percent of the patients with active TB could transmit isoniazid-resistant organisms. The use of rifampin with or without pyrazinamide in chemoprophylaxis for the case contact in a prevalent area of isoniazid-resistant TB is issued¹⁵⁾. The increasing resistance to rifampin is particularly troubling, since rifampin is essential to short-course antituberculosis therapy¹⁶⁾. In this study, 10 percent of the total cases could transmit organisms resistant to both isoniazid and rifampin. Consensus on preventive therapy for MDR TB was not reached and the current recommendation was a regimen of pyrazinamide and fluoroquinolone for 4 months¹⁷⁾.

Because previous use of antituberculous medication was the most important risk factor of DR^{4,6,13,14,18,19)} as shown in this study, and accompanied medical illness due to decreased immunity was high risk group of DR^{3, 20, 21)}, DR should be suspected in cases of delayed response to therapy or continued positive AFB smear despite treatment. The ability to complete antituberculosis therapy can be affected by housing status, employment status, level of literacy and psychological aspects^22–25). In this study, alcohol abuse, marital status or employment status did not influence the development of DR. Because patients with DR are usually treated with 2^nd line antituberculosis drugs for a long duration, which are more toxic and costlier than first line drugs, they are apt to have poor adherence to medication compared with drug-sensitive patients. The other factors to develop DR are drug intolerance during therapy²⁵⁾, contact with DR TB²⁶⁾, birth and residence in an endemic area^{5,6,13,14,19)}, HIV infection^{6, 13)} and cavitary pulmonary tuberculosis^{18, 19)}.

The first large scale sample survey in Korea in 1994, cooperation with WHO and IUATLD^{9, 11)}, presented lower rates in overall DR and primary DR and higher rate in acquired DR compared with this study (Table 6). The discrepancies between this study and the nationwide study could originate from differences in the size and nature of the study population, study period and study area. The population of previous treated TB cases out of all cases in this study was higher than that in the national study (84 of 308 cases, 27.3% vs. 189 of 2675 cases, 7.1%). This could contribute to raise the overall rate of DR in this study. Although the rate of acquired DR was lower in this study than the national survey of 1994, the rate of MDR is about 3 times that of the national survey of 1994. The population of this study was derived mainly from residents in Seoul. Residents in other areas were only 52 (13.6%) out of 308 patients. Therefore, this study could represent drug resistance in Seoul as an urban area. Generally, an urban community is more crowded and has more chance to transmit an infectious disease than a rural community. As another reason to be considered, more advanced cases among newly diagnosed subjects might prefer to be treated in 3^rd referral hospitals rather than in a public TB care center as in rural areas. Actually, in this study, patients with primary DR showed a higher rate of hospitalization despite a younger age than those with acquired DR, and the rate of primary MDR in this study was far higher than that in the national survey of 1994. Our data had comparable results with the report from Chungnam University of Korea²⁷⁾.

In this study, patients with acquired DR showed chronic reluctant course, low education level and low economic status compared with those with primary DR. Traditionally, patients with DR TB are classified as having acquired DR or primary DR on the basis of a history of previous TB treatment^{28, 29)}. Only cases of primary DR are assumed to be due to transmission of DR strains. Recent studies based on restriction-fragment length polymorphisms, useful in distinguishing different strains of M. tuberculosis, revealed that recent transmission as well as true acquisition of DR during therapy was an important cause of DR, especially MDR, even in previously treated cases in certain urban populations^{30, 31)}. Therefore, traditional clinical classification based on a history of previous treatment may result in misinterpretation and the underestimation of transmission. Improper isolation and poor ventilation due to crowding are important reasons for spread of TB particularly in the low socioeconomic status of urban areas. The resistant cases, especially previously treated, remain sources of infection for prolonged periods and they are likely to infect others^31–35). Finally, the high rate of DR and MDR in acquired cases contribute to increasing rates of DR and MDR in primary cases.

When dealing with a patient who has resistant TB, it is very important that the physician evaluate the situation fully. When seeing a patient who has already been treated for TB, susceptibility tests should be ordered. The only way to ensure that patients take their medication properly is to give it to them, i.e. direct observed therapy (DOT)³⁴⁾. Although routine DOT can not be afforded in many countries with a high prevalence of MDR TB, application of DOT is requisite in drug resistant TB, especially for poor adherence to treatment, to prevent the development of further drug resistance³⁶⁾.

Previously treated tuberculosis is a most important risk factor for DR. DR group, especially ADR, showed less compliance with treatment. More proper education and attention to prevent self-interruption should be given to a previously treated group. This study showed a higher rate of primary DR and primary MDR compared to the national study, suggesting more severe cases among newly diagnosed patients treated in tertiary referral hospitals rather than public TB care centers. High rate of acquired DR contributes to increasing the rate of primary DR. Initial drug susceptibility testing is necessary to guide optimal treatment to a culture positive case with the high risk of DR, especially in TB prevalent areas.