Original Research

MYCOBACTERIAL DISEASE

Pulmonary Impairment After Tuberculosis*

Jotam G. Pasipanodya, MBChB; Thaddeus L. Miller, MPH, DrPH; Mauricio Vecino, MD; Guadalupe Munguia, MD; Robert Garmon, DO; Sejong Bae, PhD; Gerry Drewyer, RN; and Stephen E. Weis, DO, FCCP

Background: Pulmonary impairment subsequent to a cure of pulmonary tuberculosis has been described only in selected populations. Methods: We compared pulmonary function in a case-control study of 107 prospectively identified patients with pulmonary tuberculosis who had completed at least 20 weeks of therapy and 210 patients with latent tuberculosis infection (LTBI). Results: Both groups had similar risk factors for pulmonary impairment. Impairment was present in 59% of tuberculosis subjects and 20% of LTBI control subjects. FVC, FEV1, FEV1/FVC ratio, and the midexpiratory phase of forced expiratory flow were significantly lower in the treated pulmonary tuberculosis patients than in the comparison group. Ten patients with a history of pulmonary tuberculosis (9.4%) had less than half of their expected vital capacity vs one patient (0.53%) in the LTBI group. Another 42 patients (39%) with tuberculosis had between 20% and 50% of the expected vital capacity vs 36 patients with LTBI (17%). After adjusting for risk, survivors of tuberculosis were 5.4 times more likely to have abnormal pulmonary function test results than were LTBI patients (p > 0.001; 95% confidence interval, 2.98 to 9.68). Birth in the United States (odds ratio [OR], 2.64; p 0.003) and age (OR, 1.03; p 0.005) increased the odds of impairment. Pulmonary impairment was more common in cigarette smokers; however, after adjusting for demographic and other risk factors, the difference did not reach statistical significance (p 0.074). Conclusions: These findings indicate that pulmonary impairment after tuberculosis is associated with disability worldwide and support more aggressive case prevention strategies and posttreatment evaluation. For many persons with tuberculosis, a microbiological cure is the beginning not the end of their illness. (CHEST 2007; 131:18171824)
Key words: disability; impairment; pulmonary function; tuberculosis Abbreviations: BMI body mass index; CI confidence interval; FEF2575 midexpiratory phase of forced expiratory flow; LTBI latent tuberculosis infection; OR odds ratio; PFT pulmonary function test; TCPHD Tarrant County Public Health Department

is common, with an estimated 8.9 T uberculosis million new cases and 1.6 million deaths worldwide in 2004.1 It is unknown how many survivors of
*From the Department of Medicine (Drs. Pasipanodya, Miller, Vecino, Munguia, and Weis), and the School of Public Health (Dr. Bae), University of North Texas Health Science Center at Fort Worth, Fort Worth, TX; the Tarrant County Health Department (Mr. Drewyer), Fort Worth, TX; and private practice (Dr. Garmon), Weatherford, TX. This work was performed at the University of North Texas Health Science Center at Fort Worth, and was made possible by the cooperation of both the Centers for Disease Control and Preventions Tuberculosis Epidemiologic Studies Consortium and the Tuberculosis Trials Consortium (which provided salary support for Drs. Bae, Pasipanodya, Munguia, Vecino, Weis, Miller, and Ms. Drewyer), as well as the Tarrant County Public Health Department and the University of North Texas Health Science Center at Fort
www.chestjournal.org

tuberculosis are living today; however, when the incidence of tuberculosis and the success of therapy are considered, the numbers of living tuberculosis survivors are substantial and increasing.1
Worth. The study did not receive direct funding from these or other sources, however. The authors have reported to the ACCP that no significant conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article. Manuscript received December 7, 2006; revision accepted February 12, 2007. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (www.chestjournal. org/misc/reprints.shtml). Correspondence to: Stephen E. Weis, DO, FCCP, University of North Texas Health Science Center at Fort Worth, 3500 Camp Bowie Blvd, Fort Worth, TX 76107; e-mail: sweis@hsc.unt.edu DOI: 10.1378/chest.06-2949
CHEST / 131 / 6 / JUNE, 2007

1817

Downloaded From: http://journal.publications.chestnet.org/ on 12/15/2013

Histopathologic findings resulting from tuberculosis include the formation of caseating granuloma, tissue liquefaction, and cavity formation.2 When these occur in the lung, many survivors experience permanent anatomic changes. These result in pulmonary sequelae that are characterized by bronchial and parenchymal structural changes, including bronchovascular distortion, bronchiectasis, emphysematous changes, and fibrotic bands. While these changes remain after a microbiological cure, the further evaluation of persons after tuberculosis is not recommended.3 Current tuberculosis treatment guidelines suggest that performing a chest radiograph at the completion of therapy may be useful but is not essential.3 Additional evaluation for patients after tuberculosis has been cured is currently recommended only for those patients who have suggestions of disease recurrence.3 Studies of pulmonary function in individuals with pulmonary tuberculosis demonstrated variable patterns and severity of impairment.4 9 Pulmonary function studies4 9 can show restrictive, obstructive, or mixed patterns and range from normal to severe impairment. These findings are currently incompletely characterized. The studies performed to date4 9 have been of highly selected populations, as follows: patients receiving only inpatient tuberculosis treatment; patients who have been referred for symptom or preoperative evaluation; patients with an absence of other lung disease; and persons sufficiently well to be currently employed in mining.4 9 These patients do not completely represent the populations affected by tuberculosis. Accurate estimates of the frequency and extent of pulmonary impairment from tuberculosis are important to patients and clinicians as well as to those weighing the potential benefits of the treatment of latent tuberculosis infection (LTBI). Pulmonary function tests (PFTs) objectively quantify lung function and impairment, and are used to evaluate persons with chronic lung disease.10 To determine the pattern and extent of pulmonary function abnormalities that are attributable to tuberculosis, we conducted a prospective case-control study using a PFT of patients with culture-confirmed pulmonary tuberculosis and a comparison group with LTBI. Materials and Methods
A case-control study was conducted at Tarrant County Public Health Department (TCPHD) in Tarrant County, TX. The TCPHD serves approximately 1.5 million residents in north central Texas.11 Eligible case patients were all persons with culture-confirmed pulmonary tuberculosis or pulmonary tuberculosis with extrapulmonary tuberculosis who had completed at least 20 weeks of therapy between July 7, 2005, and November 7, 2006. It is
1818

uncertain when lung function stabilizes after tuberculosis, but one serial study4 showed no statistically significant change in lung mechanics and attenuation between the onset of therapy and after 5 weeks of therapy; another study9 suggested that a peak function loss occurs within 6 months after diagnosis. We therefore assumed, by enrolling persons after 20 weeks of directly observed therapy, that subjects would be microbiologically cured and that such timing would allow data collection to be convenient for a maximal number of subjects. Comparison subjects were persons who were at least 18 years of age who had LTBI. Comparison subjects were prospectively selected by requesting participation from every fifth person presenting to the clinic for the evaluation of LTBI. All persons with HIV-LTBI coinfection who had been referred to the TCPHD were eligible for inclusion in the comparison group. By state statute, all tuberculosis-positive cultures are reported to the local health authority; by state policy, all persons with tuberculosis are treated with directly observed therapy.12,13 In Tarrant County, all patients with confirmed tuberculosis are treated by the TCPHD. LTBI is not a reportable condition in Texas; however, the TCPHD offers the evaluation and treatment of persons with LTBI.14 Study evaluations were incorporated into clinical care visits. Evaluations included PFTs and the completion of a standardized questionnaire of demographic and occupational information, and smoking, medical, and other histories. Occupational information was used to ascertain pulmonary risk, which was defined by us as exposure to particulates or asbestos for 2 years. Other studies15,16 have defined the term high-risk occupation as exposure to asbestos for at least 5 years, or exposure to dust for at least 6 years. The evaluation of comparison subjects was performed during their initial clinic visit. Patients with pulmonary tuberculosis were evaluated after at least 20 weeks of treatment. PFTs were conducted according to the American Thoracic Society recommendations15,16 for subject maneuver, techniques, and quality control using a device (Spirotouch Spirometry System 086578; Spacelabs Burdick; Deerfield, WI) that was calibrated daily before use. PFTs were performed by trained technicians in the patients preferred posture with the support of a translator when needed. Three tests with values within 5% were defined as being acceptable, and the best of three values was used for comparisons. When the variation was 5%, the best three results from eight tests were selected. The FVC maneuver was performed without a nose clip. FEV1, FVC, FEV1/FVC ratio and their respective percent predicted values were outcomes. Persons with a history of intermittent bronchodilator use received albuterol nebulization 30 min prior to testing. FVC, FEV1, and FEV1/FVC ratio values for case patients were compared with gender-specific and race-specific adult predicted normative population values and the control group.10 Airway obstruction was defined as an FEV1/FVC ratio of 70% and an FVC of 80% predicted, restrictive defects as an FEV1/FVC ratio of 70% with an FVC of 80% predicted, and combined defects were FVC of 80% predicted and an FEV1/FVC ratio of 70%.17 Interpretive algorithms were used in determining restrictive or obstructive patterns.17,18 Data were double entered using a computer spreadsheet program (Microsoft Office 2003 ACCESS; Microsoft Corporation; Redmond, WA). Discrepancies were queried and corrected. Data analysis was performed using a statistical software package (SPSS, version 12 for Windows; SPSS Inc; Chicago, IL). Significant differences between groups were determined with the Student t test and the 2 test. Logistic regressions were used to identify predictors for outcomes after controlling for patient variables. The Institutional Review Board of the University of North Texas Health Science Center at Fort Worth approved the study, and written informed consent was obtained from all participants.
Original Research

Downloaded From: http://journal.publications.chestnet.org/ on 12/15/2013

Results Between July 7, 2005, and November 7, 2006, 121 persons who were potentially eligible for the case cohort were reported to the TCPHD. Of these, 14 patients did not meet the study eligibility criteria; all eligible case patients consented to participate (Fig 1). Ninety-two of the remaining 107 subjects (86%) had pulmonary tuberculosis and 15 subjects (14%) had pulmonary tuberculosis plus extrapulmonary tuberculosis. Patients with pulmonary and pulmonary plus extrapulmonary disease were considered to be a single group. Two hundred eighteen LTBI patients were selected as control subjects, and 210 of them agreed to participate in the study. All persons with known HIV infection who had been referred to the TCPHD with LTBI agreed to participate. The mean values for FVC, FEV1, and the midexpiratory phase of forced expiratory flow (FEF2575) for patients with a history of pulmonary tuberculosis, for comparison subjects with LTBI, and the predicted values for the general population are shown in Figure 2. The characteristics of study groups are similar. There were no significant differences in ethnicity, tobacco smoking, crack cocaine use, foreign birth, occupational risk, or HIV status (Tables 1-3). Pulmonary tuberculosis patients were more likely to be male (69% vs 53%, respectively; p 0.005) and older (mean age, 47 years [SD, 15 years] vs 41 years

[SD, 15 years], respectively; p 0.003) than control subjects. Mean body mass index (BMI) was lower in pulmonary tuberculosis patients than in control subjects (mean BMI, 23.8 [SD, 5.0] vs 27.7 [SD, 5.8], respectively; p 0.001) [Table 1]. Whites, Hispanics, and African Americans comprised 23%, 27%, and 22%, respectively, of the case patient population, and 22%, 26%, and 26%, respectively, of the comparison subject population. The majority of persons were born in the United States, with 49% of case patients and 47% of persons in the comparison groups were foreign born. The prevalence of HIV was 14% and 7%, respectively, among the case patient and comparison groups, respectively (Table 1). Ninety-nine percent of subjects (316 of 317 subjects) had reproducible spirometry results, and 100% of subjects (317 of 317 subjects) had acceptable spirometry results. Values for FVC, FEV1, FEV1/FVC, and FEF2575 were significantly lower in patients who had been treated for pulmonary tuberculosis than in the comparison group (Table 1). After adjusting for age, BMI, country of birth, gender, and smoking, survivors of pulmonary tuberculosis were 5.4 times more likely to have abnormal PFT results than patients with LTBI (95% confidence interval [CI], 2.98 to 9.68) [Table 4]. Except for a history of pulmonary tuberculosis, only birth in the United States and age significantly in-

121 culture confirmed subjects eligible between 7th July 2005 and 7th November 2006

203 LTBI subjects asked to participate

+15 consecutive subjects, HIV infected, with LTBI

-2 transferred

- 1 myopathy - 1 incarcerated -10 died 2 from TB 8 other causes

-8 declined participation

Cases 107 PTB subjects evaluated 92 PTB 15 PTB & EPTB

Comparison Group 210 LTBI comparisons evaluated

Figure 1. Study enrollment. TB tuberculosis; PTB pulmonary tuberculosis; EPTB extrapulmonary tuberculosis.
www.chestjournal.org CHEST / 131 / 6 / JUNE, 2007

1819

Downloaded From: http://journal.publications.chestnet.org/ on 12/15/2013

A
FV C (L)

FVC (L)
10

B
FEV1 (L)

FEV1 (L)
7 6

C
FEF 25-75 (L)

FEF 25-75
10

8 5

4 4

3 2 2

2 1 0 PTB LTBI Normal 0 PTB LTBI Normal

-2 PTB LTBI Normal

Category

Category

Category

Figure 2. Interquartile ranges for FVC (left, A), FEV1 (middle, B), and FEF2575 (right, C) for patients with pulmonary tuberculosis (PTB) and LTBI, and for the expected healthy population.

creased the odds of pulmonary impairment (odds ratio [OR], 2.64 [p 0.003] vs 1.03, respectively; p 0.005) [Table 4]. Table 2 gives the frequency and severity of pulmonary dysfunction in the case patient and control subject groups. Severe pulmonary dysfunction was more frequent among patients with a history of tuberculosis; 10 patients (10%) had 50% of expected vital capacity compared to 1 control subject (0.48%) in the comparison group. Another 42 patients (39%) with a history of tuberculosis had between 20% and 50% less expected vital capacity, compared to 36 patients (17%) in the comparison group; 14 tuberculosis patients (13%) vs 2 compari-

son subjects (1%) had obstruction with low vital capacity. Restrictive pulmonary function was the most frequent impairment pattern observed in both groups; it was identified in 31% of case and 15% of comparison subjects. Fifty-seven percent of case patients and 51% of subjects in the comparison group had a history of ever smoking (Table 1). Smoking volume was significantly higher in the case group (12.01 vs 7.24 pack-years in the comparison group; p 0.011) [Table 1]. Smoking volume (in pack-years) had a significant correlation with the decline in lung function, and FEV1/FVC ratio was significantly lower in smokers than in nonsmokers (mean, 79% vs 84% pre-

Table 1Characteristics of Patients Treated for Pulmonary Tuberculosis (Case Patients) and LTBI (Comparison Subjects)*
Characteristics Male gender Ethnic group White Hispanic African-American AAIO Ever-smokers Ever used crack cocaine Occupational pulmonary risk Born in United States HIV-positive Not done Pulmonary impairment Age, yr BMI Height, inches Smoking history, pack-yr FVC, % predicted FEV1, % predicted FEV1/FVC ratio, % predicted FEF2575, % predicted *AAIO Asians, Africans, Indians, others. Values given as No. (%). p 0.05 (Pearson 2 test). Values are given as the mean (SD). p 0.05 (t test). 1820
Original Research

Case Patients (n 107) 74 (69) 25 (23) 29 (27) 23 (22) 30 (28) 61 (57) 21 (20) 6 (6) 55 (51) 15 (14) 63 (59) 47 (15) 23.75 (4.98) 67 (4) 12.01 (18.24) 82.77 (24.89) 77.77 (27.37) 76.07 (13.34) 72.68 (41.45)

Comparison Subjects (n 210) 111 (53) 47 (22) 55 (26) 55 (26) 53 (25) 107 (51) 36 (17) 14 (7) 111 (53) 15 (7) 24 (11) 41 (20) 41 (15) 27.66 (5.78) 66 (4) 7.24 (14.36) 95.51 (19.49) 95.43 (19.48) 83.97 (7.85) 98.61 (31.79)

p Value 0.005 0.889 0.943 0.407 0.523 0.217 0.395 0.834 0.831 0.186 0.001 0.003 0.001 0.163 0.011 0.001 0.001 0.001 0.001

Downloaded From: http://journal.publications.chestnet.org/ on 12/15/2013

Table 2Pulmonary Impairment After Tuberculosis in Patients Treated for Pulmonary Tuberculosis (Case Patients) and LTBI (Comparison Subjects)*
Case Patients (n 107) 3 (3) 7 (7) 9 (8) 17 (16) 16 (15) 55 (51) 82.77 (24.89) 7 (7) 12 (11) 9 (8) 14 (13) 14 (13) 51 (48) 77.77 (27.37) 2 (2) 2 (2) 10 (9) 12 (11) 81 (76) 76.07 (13.34) 26 (24) 12 (11) 10 (9) 7 (7) 8 (8) 44 (41) 72.68 (41.45) 44 (41) 16 (15) 33 (31) 14 (13) Comparison Subjects (n 210) 0 1 (1) 3 (1) 10 (5) 23 (11) 173 (82) 95.51 (19.47) 0 1 (1) 3 (1) 10 (4) 22 (11) 174 (83) 95.43 (19.48) 0 0 0 6 (3) 204 (97) 83.97 (7.85) 2 (1) 6 (3) 17 (8) 15 (7) 23 (11) 147 (70) 98.61 (31.79) 169 (80) 6 (3) 33 (16) 2 (1)

Variables FVC, % predicted 039 4049 5059 6069 7079 80 Mean (SD) FEV1, % predicted 039 4049 5059 6069 7079 80 Mean (SD) FEV1/FVC ratio, % predicted 039 4049 5059 6069 70 Mean (SD) FEF2575, % predicted 039 4049 5059 6069 7079 80 Mean (SD) Type of impairment None Obstructive Restrictive Combined

*Values are given as No. (%), unless otherwise indicated. p 0.05 (Pearson 2 test). None FVC 80% predicted, FEV1/FVC ratio 70% predicted; Obstructive FVC 80% predicted, and FEV1/FVC ratio 70% predicted; Restrictive FVC 80% predicted, FEV1/FVC ratio 70% predicted; Combined FVC 80% predicted, FEV1/FVC ratio 70% predicted

dicted, respectively; p 0.001 [data not shown]). Even though pulmonary impairment was more common in cigarette smokers, it was not statistically significant (p 0.074) after adjusting for demographics and other risk factors (Tables 3, 4). Discussion Over half of all recently treated pulmonary tuberculosis patients in Tarrant County had significantly
www.chestjournal.org

impaired pulmonary function (Tables 1, 2). Patients receiving directly observed treatment for tuberculosis were more than five times more likely than LTBI patients to have pulmonary impairment (OR, 5.37; 95% CI, 2.98 to 9.68). The degree of impairment was variable, ranging from mild to severe, and was not uniformly distributed among patient groups; subjects born in the United States and older subjects were more likely to have impairment than were younger or foreign-born subjects (Table 4). Pulmonary restriction and obstruction with low vital capacity were the predominant patterns observed. This study is the first in which data have been collected from a nonselected population and then controlled for background risk that was not related to tuberculosis. These data suggest that pulmonary impairment after tuberculosis is an important, mostly unrecognized, cause of chronic lung disease worldwide. Our findings differ from those of earlier studies, which have occurred within highly selected populations. In a study6 of impairment performed prior to the adoption of short-course therapy, 60% of hospitalized tuberculosis patients had an FEV1 of 80% predicted. Fifteen percent of patients had a severe reduction in FEV1 ( 50% predicted).6 The authors found evenly distributed patterns of dysfunction among impaired patients, as follows: restrictive dysfunction, 24%; obstructive dysfunction, 23%; and mixed dysfunction, 19%.6 Studies of patients who have been treated with short-course regimens found variable levels of impairment.4,7,9,19 Two studies from South Africa showed abnormal lung function after the completion of antituberculosis therapy. One of these studies,9 which was limited to hospitalized patients, found that 53% had pulmonary function abnormalities. The second study9 evaluated current miners with a history of tuberculosis who had been identified through screening up to 31 years prior to their evaluation. The investigators found that the proportion of impaired subjects increased with episodes of tuberculosis; impairment was identified in 18%, 27%, and 35% of subjects, respectively, after one, two, or three episodes.9 A study from Canada4 found that 36% of patients with pulmonary tuberculosis had abnormal pulmonary function. This study employed stringent entry criteria and enrolled 25 of 84 subjects over 17 months.4 Of the 15 patients with cavitary disease, progressive obstruction developed in 1 patient with a final FEV1 of 0.73 L; 7 of 11 patients (64%) had bronchiectasis.4 These studies each demonstrated pulmonary impairment after tuberculosisis, but none measured the full spectrum of impairment, and subjects were drawn from selected populations not controlled for pulmonary risks extant in populations of subjects who were predisposed to tuberculosis. Previous studies4 9 of pulmonary impairment resulting from tuberculosis all used adult normative
CHEST / 131 / 6 / JUNE, 2007

1821

Downloaded From: http://journal.publications.chestnet.org/ on 12/15/2013

Table 3Relationship Between Volume of Cigarettes Smoked and Pulmonary Impairment*

Impairment Variables Case patients Never-smokers Mild smokers ( 30 pack-yr) Heavy smokers ( 30 pack-yr) Total Comparison subjects Never-smokers Mild smokers ( 30 pack-yr) Heavy smokers ( 30 pack-yr) Total None 17 (39) 24 (55) 3 (7) 44 (100) 87 (52) 73 (43) 9 (5) 169 (100) Some 29 (46) 24 (38) 10 (16) 63 (100) 16 (39) 20 (49) 5 (12) 41 (100) p Value 0.442 0.092 0.158

0.152 0.518 0.114

*Values are given as No. (%), unless otherwise indicated. Variables are controlled for demographics, BMI, crack cocaine use, occupational exposure, and pulmonary tuberculosis.

population pulmonary function values for comparison. Normative pulmonary function data may not be representative of the population of persons in whom tuberculosis is destined to develop and is not an ideal comparator (Fig 2). LTBI and tuberculosis are not distributed homogeneously throughout society. Both are more likely to occur in persons with unstable housing, foreign birth, low socioeconomic status, HIV coinfection, substance abuse, and other risk factors.20 There are no data on the effects of LTBI on lung function. In this study, the control group of persons in whom LTBI was diagnosed was chosen to identify the contribution to pulmonary impairment from pulmonary tuberculosis independent of these factors. To avoid case-selection bias, the study population was derived from a geographic area rather than from select groups that may not fully represent the population with tuberculosis.4 9 We believe that these methodologies give the most representative estimate currently available for pulmonary impairment due to pulmonary tuberculosis. Symptoms of pulmonary impairment generally do

Table 4 Adjusted ORs for Pulmonary Impairment After Tuberculosisis by Risk Factors Relative to The Comparison Group of LTBI Patients*
Variables Having PTB US born vs non-US born Age BMI Male vs female White vs nonwhite Ever-smoker vs never-smoker OR 5.37 2.64 1.03 0.97 1.3 1.56 1.8 95% CI 2.989.68 1.404.96 1.011.05 0.921.02 0.713.38 0.783.18 0.943.42 p Value 0.001 0.003 0.005 0.204 0.392 0.207 0.074

*PTB pulmonary tuberculosis. ORs are expressed per SD increase in each continuous risk factor for dichotomous reference groups. 1822

not occur in persons with chronic lung disease until FEV1 has fallen to 50% of normal values.21 Symptom screens alone are unlikely to detect pulmonary impairment after tuberculosisis. The routine identification of persons with pulmonary impairment after tuberculosisis will require the revision of recommendations for performing PFTs for tuberculosis patients.22 Patients with pulmonary impairment after tuberculosisis may benefit from the methods used to ameliorate other respiratory illness, such as immunizations against respiratory illness and advice on preventing further impairment. Moderate-to-severe lung disease has also been associated with increased mortality.23 Persons with impairment from chronic obstructive lung disease similar to that associated with lung disease after tuberculosis benefit from pulmonary rehabilitation.24,25 The improved exercise capacity and quality of life, and reduced use of medical resources associated with rehabilitation may warrant rehabilitation in select patients after treatment for pulmonary tuberculosis.24,25 The identification of impairment after tuberculosisis may therefore be an important strategy to limit tuberculosis morbidity and its attendant costs. A relationship between tobacco smoking and pulmonary tuberculosis is suspected, but has not been well defined.26 Tobacco smoking serves as a marker of risk for pulmonary tuberculosis; many physical and social factors associated with cigarette smoking are shared with tuberculosis.26 Tobacco smoking causes pulmonary impairment, and smoking would be expected to increase the risk of pulmonary impairment from tuberculosis.27 FEV1/FVC ratio was significantly lower in smokers than in nonsmokers (mean FEV1/FVC ratio difference, 79% vs 84% predicted, respectively; p 0.001 [data not shown]). A larger sample size would allow a more powerful analysis of the interaction between smoking and tuberculosis and its effect on pulmonary function. However, we did not find that persons with a history of smoking had an increased risk of pulmonary impairment when pulmonary tuberculosis developed, and 7 of 11 subjects (64%) with 50% of expected FVC had no history of tobacco use (data not shown). This study has limitations. The design did not allow the determination of causality, as the sequence of exposure to the outcome cannot be proved. Our population had other risks for chronic lung impairment. Nearly half of the study participants were current smokers, and lifetime smoking histories of 7.24 and 12.01 pack-years, respectively, were reported in comparison subjects and case patients. Chronic exposure to smoke from cooking fires or other sources is another contributing risk for both tuberculosis and other lung disease in immigrant participants.28 31 A history of such biomass smoke
Original Research

Downloaded From: http://journal.publications.chestnet.org/ on 12/15/2013

exposure may be an important risk factor for pulmonary tuberculosis and abnormal pulmonary function following its microbiological cure, but we found that subjects born in the United States who are unlikely to have biomass smoke exposure were more likely to have pulmonary impairment than were foreign-born subjects. We estimated the odds of impairment with and without pulmonary tuberculosis using a comparison group with similar pulmonary and other risk factors from a single geographic area. Our data reflect the epidemiologic mix of Tarrant County, which may be different from other areas. Spirometry using FVC and FEV1/FVC ratio values has a very low positive predictive value but a high negative predictive value for restrictive pulmonary dysfunction. Although spirometry can exclude 97% of those persons without a restrictive pattern, a more positive prediction of restriction can be obtained from total lung capacity.32 All case patients and control subjects were not given bronchodilators, and it is possible that bronchospasm may not be equally distributed between the case and control groups. The natural history of pulmonary impairment after tuberculosis is unknown, and it is possible that the impairment identified may worsen or improve. Finally, we used prediction equations for the interpretation of the PFT results that included adjustments for race but did not include the full range of races in our population.7 Despite these limitations, pulmonary tuberculosis remains the most likely etiology for the identified pulmonary impairment. Pulmonary tuberculosis is associated with frequent pulmonary damage despite microbiological cure. Impairment is variable, ranging from none to severe. Pulmonary impairment contributes important, previously unmeasured burdens of tuberculosis in microbiologically cured patients that may include chronic impairment and excess mortality.3336 PFTs and other measures may need consideration after a cure. When earlier data are considered with our findings, it is apparent that a microbiological cure is not protective against substantial pulmonary sequelae of pulmonary tuberculosis. These findings support the aggressive treatment of LTBI or other case preventing strategies worldwide, and indicate that, for many persons with tuberculosis, microbiological cure is the beginning, not the end, of their illness.
ACKNOWLEDGMENT: This study could not have been completed without the TCPHD supplying study resources. We are indebted to the study participants whose participation made this study possible, and to Francesca Sanchez, MD, who provided a critical review. We also are indebted to the Tuberculosis Epidemiologic Studies Consortium (TBESC) at the Centers for Disease Control and Prevention and to the Tuberculosis Trials Consortium (TBTC), which provided salary support for Drs. Bae, Pasipanodya, Munguia, Vecino, Weis, Miller, and Ms. Drewyer, although neither consortium directly funded this study, nor had
www.chestjournal.org

any role in study design, data collection, data analysis, data interpretation, or writing of the report.

References
1 World Health Organization. Estimated incidence, prevalence and TB mortality, 2004. Available at: http://www.who.int/ mediacentre/factsheets/fs104/en/. Accessed May 21, 2007 2 Sodeman WA, Sodeman, TM. Pathologic physiology: mechanisms of disease. 7th ed. Philadelphia: PA. WB Saunders, 1985; 467 3 American Thoracic Society. Treatment of tuberculosis. MMWR Morb Moral Wkly Rep 2003; 52:177 4 Long R, Maycher B, Dhar A, et al. Pulmonary tuberculosis treated with directly observed therapy: serial changes in lung structure and function. Chest 1998; 113:933943 5 Gaensler EA, Lindgren I. Chronic bronchitis as an etiologic factor in obstructive emphysema, preliminary report. Am Rev Respir Dis 1959; 80:185193 6 Snider GL, Doctor L, Demas TA, et al. Obstructive airways disease in patients with treated pulmonary tuberculosis. Am Rev Respir Dis 1971; 103:625 640 7 Wilcox PA, Ferguson AD. Chronic obstructive airways disease following treated pulmonary tuberculosis. Respir Med 1989; 83:195198 8 Weiner H. Changes in employment status associated with hospitalization for tuberculosis: analysis of 163 consecutively admitted males. Am Rev Respir Dis 1963; 87:1722 9 Hnzido E, Singh T, Churchyard G. Chronic pulmonary function impairment by initial and recurrent pulmonary tuberculosis following treatment. Thorax 2000; 55:3238 10 Crapo RO, Morris AH, Gardner RM. Reference spirometric values using techniques and equipment that meet ATS recommendations. Am Rev Respir Dis 1981; 123:650 664 11 Texas Association of Counties, Austin, Texas. County information project: Tarrant County profile. Available at: http://www.txcip.org/tac/census/profile.php?FIPS48439. Accessed May 4, 2006 12 Office of the Secretary of State, State of Texas, Austin, Texas. Texas administrative code title 25, part 1, chapter 97. Available at: http://info.sos.state.tx.us/pls/pub/readtac$ext.View TAC?tac view4&ti25&pt1&ch97. Accessed May 21, 2007 13 Texas Department of State Health Services, Austin, Texas, Infectious Disease Control Unit. Tuberculosis control programs. Available at: http://www.dshs.state.tx.us/idcu/disease/ tb/programs/default.asp. Accessed August 11, 2006 14 Tarrant County Public Health, Fort Worth, Texas. Client services. Available at: http://www.tarrantcounty.com/ehealth/ cwp/view.asp?A763&Q430433. Accessed May 21, 2007 15 American Thoracic Society. The diagnosis of non-malignant disease related to asbestos. Am Rev Respir Dis 1991; 144: 477 478 16 American Lung Association of San Diego. Taking occupational history. Ann Intern Med 1983; 99:641 651 17 American Thoracic Society. Evaluation of impairment/disability secondary to respiratory disorders. Am Rev Respir Dis 1986; 133:12051209 18 Enright PI, Hyatt RE. Office spirometry: a practical guide to selection and use of spirometers. Philadelphia, PA: Lea & Febiger, 1987 19 Plitt MI, Anderson R, Van Rensburg CEJ, et al. Influence of antimicrobial chemotherapy and pro-inflammatory indices in severe pulmonary tuberculosis. Eur Respir J 1998; 12:351356 20 American Thoracic Society. Controlling tuberculosis in the United States. MMWR Morb Moral Wkly Rep 2005; 54:1 81
CHEST / 131 / 6 / JUNE, 2007

1823

Downloaded From: http://journal.publications.chestnet.org/ on 12/15/2013

21 Sutherland ER, Cherniack RM. Current concepts: management of chronic obstructive pulmonary disease. N Engl J Med 2004; 350:2689 2697 22 American Thoracic Society. Task group on screening for adult respiratory disease: screening for adult respiratory disease. Am Rev Respir Dis 1983; 128:768 774 23 Mannino DM, Buist SA, Petty LT, et al. Lung function and mortality in the United States: data from the First National Health and Nutrition Examination Survey follow up study. Thorax 2003; 58:388 393 24 Lacasse Y, Wong E, Guyatt GH, et al. Meta-analysis of respiratory rehabilitation in chronic obstructive pulmonary disease. Lancet 1996; 348:11151119 25 Griffiths TL, Burr ML, Campbell IA, et al. Results at 1 year of outpatient multidisciplinary pulmonary rehabilitation: a randomized controlled trial. Lancet 2000; 355:362368 26 Bothamley GH. Smoking and tuberculosis: a chance or causal association. Thorax 2005; 60:527528 27 Pelkonena M, Tukiainenb H, Tervahautaa M, et al. Pulmonary function, smoking cessation and 30 year mortality in middle aged Finnish men. Thorax 2000; 55:746 750 28 Smith KR, Mehta S, Maeusezahl-Feuz M. Indoor smoke from solid fuels. In: Ezzati M, Rodgers AD, Lopez AD, et al, eds. Comparative quantification of health risks: global burden of disease due to selected major risk factors. Geneva, Switzerland: World Health Organization, 2004; 14371495 29 Zuber PL, McKenna MT, Binkin NJ, et al. Long-term risk of tuberculosis among foreign-born persons in the United

States. JAMA 1997; 278:304 307 30 Talbot EA, Moore M, McCray E, et al. Tuberculosis among foreign-born persons in the United States, 19931998. JAMA 2000; 284:2894 2900 31 Mishra VK, Retherford RD, Smith KR. Biomass cooking fuels and prevalence of tuberculosis in India. Int J Infect Dis 1999; 3:119 129 32 Shawn D, Dales R, Cardinal P. How accurate is spirometry at predicting restrictive pulmonary impairment. Chest 1999; 115:869 873 33 Connolly C, Reid A, Davies G, et al. Relapse and mortality among HIV-infected and uninfected patients with tuberculosis successfully treated with twice weekly directly observed therapy in rural South Africa. AIDS 1999; 13:15431547 34 Perriens JH, St. Louis ME, Mukadi YB, et al. Pulmonary tuberculosis in HIV-infected patients in Zaire: a controlled trial of treatment for either 6 or 12 months. N Engl J Med 1995; 332:779 784 35 Connolly C, Davies GR, Wilkinson D. Impact of the human immunodeficiency virus epidemic on mortality among adults with tuberculosis in rural South Africa, 19911998. Int J Tuberc Lung Dis 1998; 2:919 925 36 Garc a-Garc a M, Ponce-de-Leo n A, Garc a-Sancho MC, et al. Tuberculosis-related deaths within a well-functioning DOTS control program. Emerg Infect Dis [serial online] Available at: http://www.cdc.gov/ncidod/EID/vol8no11/02 0021.htm. Accessed February 1, 2007

1824

Original Research

Downloaded From: http://journal.publications.chestnet.org/ on 12/15/2013