Does directly observed therapy (DOT) reduce drug resistant tuberculosis?

Abstract

Abstract Background Directly observed therapy (DOT) is a widely recommended and promoted strategy to manage tuberculosis (TB), however, there is still disagreement about the role of DOT in TB control and the impact it has on reducing the acquisition and transmission of drug resistant TB. This study compares the portion of drug resistant genotype clusters, representing recent transmission, within and between communities implementing programs differing only in their directly observed therapy (DOT) practices. Methods Genotype clusters were defined as 2 or more patient members with matching IS6110 restriction fragment length polymorphism (RFLP) and spoligotype patterns from all culture-positive tuberculosis cases diagnosed between January 1, 1995 and December 31, 2001. Logistic regression was used to compute maximum-likelihood estimates of odds ratios (ORs) and 95% confidence intervals (CIs) comparing cluster members with and without drug resistant isolates. In the universal DOT county, all patients received doses under direct observation of health department staff; whereas in selective DOT county, the majority of received patients doses under direct observation of health department staff, while some were able to self-administer doses. Results Isolates from 1,706 persons collected during 1,721 episodes of tuberculosis were genotyped. Cluster members from the selective DOT county were more than twice as likely than cluster members from the universal DOT county to have at least one isolate resistant to isoniazid, rifampin, and/or ethambutol (OR = 2.3, 95% CI: 1.7, 3.1). Selective DOT county isolates were nearly 5 times more likely than universal DOT county isolates to belong to clusters with at least 2 resistant isolates having identical resistance patterns (OR = 4.7, 95% CI: 2.9, 7.6). Conclusions Universal DOT for tuberculosis is associated with a decrease in the acquisition and transmission of resistant tuberculosis.

Full text

Page 1

Non-Hispanic Whites have higher risk for pulmonary impairment from pulmonary
tuberculosis.
Jotam Pasipanodya1, 2, Edgar Vecino2, Thaddeus Miller2, Guadalupe Munguia2, Gerry
Drewyer4, Michel Fernandez4, Philip Slocum 3 and Stephen E. Weis,2,4*
1. Department of Internal Medicine, UNT- Health Science Center at Fort Worth, Fort
Worth, TX, USA. 2. UT Southwestern Medical Center at Dallas, Dallas. Texas, USA.
3. Department of Internal Medicine, A.T. Still University of Health Sciences Kirksville, MO
4. Tarrant County Public Health Department, Fort Worth, Texas. weistephen@me.com
Abstract

Page 2

Background
Methods
Results
Conclusions

Page 3

Disparities in clinical outcomes associated with race and ethnicity are well
documented for many diseases and patient populations. While there are a variety of
explanations for these effects they are not fully understood (1,30,40). Socio-economic,
biological, cultural, demographic, differences as well as other factors may all play a part in
the diseases and medical treatments that impact an individual patient’s health(1,30,34).
Tuberculosis disproportionately affects economically disadvantaged, racial and ethnic
minority populations (7,13,22-24,31-33). In developed countries the rate of decline in
incident tuberculosis (TB) has slowed, raising concerns about progress toward TB
elimination goals (5,32). It is plausible that the health impacts of slowing rates of TB
incidence may be concentrated in these vulnerable populations.
It is known that some contributors to health disparities can be mitigated by
improvements in health systems (1,40). However, the contribution of personal behavior,
value choices, biological and gender differences, and race/ethnicity on specific diseases and
their clinical outcomes remains vague. Better understanding of contributors to disparate
health impacts yields many benefits in cases of diseases of public health importance, such as
tuberculosis (1,22-24,30,40). Potential limited health care resources can be better directed
to controlled and eliminated if causes of disparities can be identified and remediated.
In a prior study, we measured the frequency and degree of pulmonary impairment in
TB patients who were treated with standard regimes delivered by directly observed therapy
(DOT) (28). Pulmonary tuberculosis after treatment (PIAT) was found in a majority of the
cohort and was more common in native born and older patients (27,28). The study’s sample
size did not allow an analysis of the impact of racial or other groups on impairment. Most
studies have shown that relative to other patients, Blacks and Hispanics fare poorly for most
measured TB outcomes (7,13,31-33). We expanded our sample to allow a comparison of
PIAT frequency across self-identified race/ethnicity groups and determine whether
socioeconomic status predicted PIAT.
Methods

Page 4

Patients and Setting.
Consecutive patients 18 years of age and older, with culture-confirmed M.
Tuberculosis, who had completed at least 20 weeks of DOT for TB through Tarrant County
Public Heath (TCPH) From July 2005 to December 2009, were asked to participate in a
study of their pulmonary function. By state statue, all TB cases diagnosed in the State of
Texas are required to be reported to the local public health authorities. Tarrant County
Public Health has jurisdiction as the health authority for Tarrant County, Texas, an urban
county with a 2010 population of 1,789,900 (37). TCPH provides treatment to all persons
with TB within this jurisdiction, using universal DOT delivered to the patient’s preferred
location (41). All patients were treated with standard 4 drug American Thoracic Society
(ATS) and Centers for Diseases and Prevention Control (CDC) recommended anti-TB
regimes (4). TB patients with no pulmonary involvement were included in the study to
assess differences that may be unrelated to the lung. The Institutional Review Board of the
University of North Texas Health Science Center at Fort Worth approved the study. All
subjects gave written informed consent.
Pulmonary Function Testing
Pulmonary function tests (PFTs) by spirometry were performed on each consenting
patient. Spirometry was conducted according to ATS guidelines for maneuver, techniques
and quality control using the Spirotouch device (Spirotouch Spirometry System 086578;
Spacelabs Burdick; Deerfield, WI) (11,21). Patients with a history of bronchodilator use
were provided inhaled albuterol 15 minutes before the test. Consistent results were
considered 5% or less variation between results for three separate tests. Once consistency
was established, the best of three consistent results was used for grading pulmonary function
impairment.
Impairment was defined and graded using American Medical Association (AMA)
guides to evaluation of permanent impairments (8). Forced Expiratory Volume in 1 minute
(FEV1) >=80%, Forced Vital Capacity (FVC) >=80% and FEV1/FVC>70% of predicted
were considered normal. An interpretive algorithm from the AMA was used to categorize
the degree of pulmonary impairment associated with non-normal PFT values (8).
Impairment was categorized as none, mild (FEV1 values >60% but <80%), moderate (41%
to 59%) or severe (FEV1 values <40%) (8,11). Patients with mild, moderate and severe

Page 5

impairment were combined to form impaired group. Those with moderate and severe
formed the moderate/severe group.
Variables.
Patient demographic data were prospectively obtained using standardized instrument
by trained research personnel. All data were double entered into Microsoft Office 2003
ACCESS database (Microsoft Corporation, Redmond, WA. 98052). Demographic data were
collected from the patients at the time of enrolment. Race/ethnicity was self-identified.
Patients were given an option to identify themselves as Hispanic in accordance with US
federal definitions. The United States’ Office of Management defines 5 non-exclusive racial
categories; (i.e. American Indian or Alaskan Native or Native Hawaiian or Pacific Islander,
Asian, Black and White), and 2 ethnic categories (of Hispanic Origin and not Hispanic
Origin) (26). The race/ethnicity groups identified in this study were; Non-Hispanic White,
Non-Hispanic Black, Asians and Hispanics. Because of their small numbers we combined
into one group Pacific Islander, Native American Indian and Arab and examined separately.
Socioeconomic variables assessed included educational attainment, employment
status at diagnosis and self-identified occupation. Employment status was defined as
employed, unemployed, retired, disabled and student. For those unemployed, retired,
disabled or in school, previous occupation was identified and used for occupations’
classification. Occupations were coded based on the Standard Occupation Classification
(6,9) and for comparability purposes were collapsed into 4 prestige rank ordered major
categories (6,9). Household income was estimated from average incomes within census-tract
ZIP codes of the patient’s home address (39). Education was categorized into quartiles of
years <12, 12, 12 – 15 and >16 years. Similarly, income was divided into quartiles of area
median household income of <$27,250, $27,251 -$37 180, $37,180 – 52,777 and >$52,778
and these were also comparable to previously published data from US TB patients (7,39).
We scored the reported occupations using standard methods and correlated them to levels of
education (6,9,35). Education was used as a proxy for estimating socioeconomic status (9).
The duration of untreated active TB was assessed via a time-to-TB-diagnosis
variable, defined as the time from self-reported onset of symptoms to beginning tuberculosis
therapy. The time-to-TB-diagnosis variable gives insight not only into severity of illness but
also patient-related factors associated with accessing healthcare (1). Ever smokers were

Page 6

patients who gave a history of current or past cigarettes smoking. Lifetime volume of
cigarette exposure was estimated using pack-years. Similarly, proportions exposed to solid
fuel smoke (biomass exposure) and duration of biomass exposure were compared between
groups. Cavitary disease denotes patients with single or multiple cavities visible on regular
posterioanterior chest radiographs (16). Patients in whom mycobacterium tuberculosis
disease was demonstrated outside of the lung parenchyma were defined as extra-pulmonary.
Statistical Analysis
We examined the influence of demographic and clinical characteristics and
socioeconomic status on PIAT. Parsimonious multivariate logistic regression models were
constructed and analyzed for the full sample and separately for US-born and foreign-born
persons. Impairment, the primary outcome variable, was dichotomized by combining mild,
moderate and severe impairment versus no impairment. The independent variables
examined were age, smoking in pack-years and body mass index (as continuous variables),
race (Non Hispanic Whites versus non-Whites) and socioeconomic status. Bivariate
Pearson’s correlation analysis between variables of interest together with the variance
inflation factors were used to examine for multicolinearity.
Because time-to-TB-diagnosis and treatment was not normally distributed, median
and interquartile values are reported and the Kruskal-Wallis test was used to compare values
across groups. Pearson’s Chi-Square test was used to compare proportions; analysis of
variance was used to compare continuous variables and the Mantel-Haenszel test was used
to examine for trend across pulmonary impairment levels. Both age and smoking were
included in all multivariate models as they have been shown to independently exacerbate
pulmonary function decline (2,12,14). The median age at which impairment and
moderate/severe impairment occurred among the racial/ethnic groups were compared using
Kaplan-Meir methods.
Analysis was performed using SPSS version 12 for Windows (SPSS Inc; Chicago,
IL) and Graphpad Prism version 5 (GraphPad Software; La Jolla, CA).

Page 7

Results
Between July 2005 and December 2009, XXX persons with culture-confirmed were
reported to Tarrant County Health (Figure 1). Of these, 320 were enrolled in the study and
318 (99 %) had reproducible PFT results. One subject failed to perform an acceptable PFT
procedure and the results were not included. All patients enrolled in the study had received
>80% of the recommended treatment regimen by DOT.
Of the 320 enrolled; 69 (22%) self-identified as White, 85 (27%) as Black, 81 (25%)
as Asian, 82 (26%) as Hispanic and 3 (0.9%) as other racial/ethnic group. Forty-three
percent of Asians were from Vietnam. The 3 subjects in the “other” racial/ethnic group were
all male and included; one non-impaired foreign-born who self-identified as an Arab, one
US-born Pacific-Islander with mild impairment and one US-born Native Indian who also
had mild impairment. These three were excluded from the data because there were too few
subjects in the ‘other’ racial/ethnic group for meaningful comparison purposes. One foreign-
born woman who self-identified as both White and Hispanic was categorized for the purpose
of this study as Hispanic to make the groups mutually exclusive.
Table 1 show that in terms of TB disease type and pattern and patients’ access to TB
care the study population was similar between races; however, there were significantly
different demographic and clinical characteristics. HIV infection was significantly higher
among Blacks, while level of education was significantly lower among Hispanics compared
to Whites. In addition, the clinical and demographic characteristics of US-born persons were
significantly different from those who were foreign-born. US-born patients were older than
foreign-born persons; 50 years (standard deviation [SD]) 14 versus 44 (17), (p=0.002).
Additionally, US- born persons smoked more cigarettes (4 fold higher, p < 0.001) compared
to foreign-born persons. The mean values were 21 (32) versus 5 (10) pack-years. Proportion
of ever-smokers was higher among Whites, p<0.001 than other groups, and they smoked
significantly more as measured by pack years than the other racial/ethnic groups, p<0.001
(table 1).
Pulmonary function Impairment After Tuberculosis. The distribution of pulmonary
impairment (PIAT) and its severity among the 4 racial/ethnic groups is shown in table 1 and
figure 1 and figure 2. PIAT was more frequently encountered among Whites compared to
other racial groups (p<0.001), and when encountered was more likely to be severe

Page 8

(p=0.001) (Figure 1). Pulmonary impairment was identified in 71% of Whites 71% versus
58% of Blacks, 49% of Asians and 32% of Hispanics. Figure 4 demonstrates that PIAT
frequency was significantly higher among Whites compared to other racial/ethnic groups in
both ever-smokers and never-smokers, (p<0.0001).
The distribution of employment, income, occupation, and education data within our
study population was similar to other prior published data obtained from TB patients in the
US (9 – 11). Education and income were significantly correlated; Pearson’s correlation
coefficient (r) 0.21, p <0.001. When the occupational status was ranked according to
prestige, it also significantly correlated with both education and income; r=0.33, p<0.001
and r=0.15, p=0.005, respectively. PIAT was evenly distributed across all levels of
socioeconomic status suggesting that PIAT and socioeconomic status are not related (Figure
3).
The median time-to-TB-diagnosis and treatment (inter-quartile range [IQR]) for not
impaired persons was 62 days (IQR 12 - 110) and that for mildly impaired persons was 93
days (IQR 61 - 110). For moderately impaired the median time-to-TB-diagnosis and
treatment was 138 days (IQR 32 - 271) and that for severely impaired was 37 days (IQR 12
-60). There was no significant association between race and time-to-TB-diagnosis and
treatment, p=0.978 (table 1). Similarly, no association between time-to-TB-diagnosis and
treatment and PIAT was observed, p= 0.058 (data not shown).
In univariate analysis race, age and US-born were significantly associated with PIAT
(table 2). The likelihood for having PIAT increased by 2% (95% confidence interval [CI] 1,
3) for each 1 year increase in age. PIAT was 2.3 fold (95% CI 1.46, 3.61) higher in US-born
compared to foreign-born subjects. The measures of socioeconomic status (table 2) used
including occupation and employment status (data not shown) were not associated with
PIAT. In a multivariate analysis that controlled for potential demographic and clinical
confounders; the only significant predictor for PIAT was White race for whom the risk for
PIAT was increased 3 folds (1.18, 8.40). However, the full model explained only 7% of the
data variance observed, suggesting that there are other unmeasured parameters that could
account for these findings.
As race and country of birth were significantly associated with impairment separate
multivariate regression models stratified by country of birth were constructed. Table 3

Page 9

shows the multivariate regression model containing age, smoking and race of 144 US-born
persons, where R2 doubled to 18% and White race and age independently predicts PIAT. The
age-related risk for PIAT doubles to 5% per year compared to the full model given in table
2. Combined these data demonstrated the different risks for PIAT between US-born and
foreign-born. White race and age are significantly associated with PIAT in US-born persons.
Onset of age-related lung function decline starts is variable (19 - 21); however, for
this study cohort onset of impairment was related to time of acquiring tuberculosis. Figure 5
shows that the time-to-developing some impairment (panel A) was significantly different
among the racial/ethnic groups. The median age was 51 years for Blacks, 59 for Whites, 56
for Asians and 71 years for Hispanics. Similarly, the probability for developing moderate to
severe impairment was higher in Blacks of younger age groups compared to other racial
groups (figure 5, panel B). The median age for Blacks was 63 and that for Whites was 72,
p=0.0239. The hazard ratio [HR] was 0.45 (0.22, 0.90).
Discussion
It has been established that PIAT is common and substantial in spite of adequate
treatment (27,28). A full understanding of the degree and distribution of the impairment may
be useful to direct efforts to reduce its health impacts (22-24). In the U.S., risks for TB
infection and measures of TB disease outcomes including mortality have consistently found

Page 10

to be higher among racial/ethnic minorities and foreign-born persons (7,13,31-33). We
analyzed the relationship between race/ethnicity and PIAT in a cohort with culture-
confirmed pulmonary tuberculosis that had completed a minimum of 20 weeks of therapy.
We found that self-identified White TB patients had disproportionately more pulmonary
impairment relative to other races/ethnicities. We found pulmonary impairment to be more
prevalent (72% vs. 48%), odds ratio for (OR 3.15), and of greater severity of among whites
compared to nonwhites. These differences did not change when age, body mass index,
smoking, country of birth and socio-economic status variables were controlled. The quality
of TB treatment was not different as medical regimen was standardized and all patients were
treated with directly observed therapy. Access to care, as estimated by time to diagnosis and
treatment (days) was also not different between groups. Among possibly explanatory
variables analyzed, only age and race were significant predictors for impairment in US born
persons. These data demonstrate an unrecognized negative health impacts to specific
populations of TB patients.
Currently older adults are not considered high-priority candidates for testing and
treatment of LTBI unless they have accompanying conditions increasing their risk of
developing TB disease such as immunosuppression, fibrotic lesions on chest X-ray or recent
contact with infectious TB (5). These recommendations are based on the potential for
adverse drug events associated with LTBI treatment, and the perceived low morbidity and
mortality risk for treated TB. The likelihood for PIAT increase by an average 2% for each
additional increase in age (table 3; figure 5). PIAT once developed is irreversible. In
addition, NHANES data has also shown that poorer lung function is also association with
poor clinical outcomes including premature death (19,25). This suggests that moderate to
severe PIAT could also be associated with earlier mortality. There are readily available
methods including LTBI treatment that can prevent TB and its sequelae PIAT. If these
findings of greater risk for PIAT in older persons with tuberculosis are confirmed then future
versions of LTBI treatment guidelines need to consider as an additional treatment benefit
PIAT reduction for reducing tuberculosis burden.
Smoking paragraph-Dr Slocum is working on this. Cigarette smocking is an
established cause of pulmonary impairment. The smoking prevalence was significantly
higher among whites compared to other racial groups. The proportion of Whites impaired

Page 11

among never-smokers was 70% compared to 78% among ever-smokers. In multivariate
analysis smoking was not a significant predictor for pulmonary impairment (table 2 &3).
Previous studies have investigated pulmonary sequelae of TB from a number of
perspectives (10,18,20,29,42). Poh et al evaluated patients hospitalized for treatment and
identified older age, disease severity at presentation and heavy smoking as predictors for
pulmonary impairment. That population had a high smoking prevalence and was treated
non-rifampin chemotherapy regimes (29). A population-based study from Latin America
demonstrated that older age and repeated TB disease were associated with pulmonary
impairment (20). Two South African studies of patients receiving inpatient treatment (10,42)
similarly demonstrated that repeated TB disease significantly increased risks for pulmonary
impairment. Each of these studies was included subjects from high TB burden areas with
limited resources and whose population structure and socioeconomic status is different to
that found in the US. Additionally factors such as smoking, aging or race/ethnicity, were not
fully explored in these studies (10,20,42). We examined patients who could be argued had
received current best therapy for tuberculosis; by definition the population included persons
successfully treated by directly observed therapy using modern standard chemotherapy.
Despite this treatment we identified PIAT can be prevented using with currently available
tools in medical practice. Our data combined with prior data (25, 27 - 30) indicates that
PIAT is independent of smoking and is more likely to occur in elderly population.
Our study failed to detect association between socioeconomic status and pulmonary
impairment. This was an unexpected and novel finding because multiple studies have
consistently associated poorer health outcomes with low socioeconomic status (1,7,9). A
leveling of access to health care allowed by the public treatment of TB within the study area
may explain this lack of disparity. Even though Hispanics had lower socioeconomic status in
this cohort and have higher TB rates that other racial groups in the US, they were protected
against pulmonary impairment compared to other racial/ethnic groups. This finding supports
what has been called the “healthy Hispanic Paradox”. This concept developed from
multiple sources that have shown that Hispanics enjoyed longer life expectancy compared
to other racial/ethnic groups (3,17). It is plausible to suggest a biological determinant for
more frequent pulmonary impairment among whites, or a complex interaction between other
unmeasured host, pathogen and environment factors. As an example of this type of

Page 12

interaction is the outcome of oropharyngeal squamous cell carcinoma (34). Outcome of this
cancer varies by race with a worse overall survival rates for black patients with compared to
whites. Further analysis of this issue demonstrated that survival was worse because of a
lower prevalence of HPV related oropharyngeal squamous cell carcinoma in blacks. As
treatment was same for all groups we suspect that there may be a similar complex
interaction that explains the variable outcome by race of our patients (34).
This study’s principle limitation is race and ethnicity are contextual, mutually
contradictory and usually assumes socially defined constructs including cultural, linguistic
and geopolitical factors which change with time and as such are subject to misclassification
bias (26,30,38,39). As an example racial/ethnicity definitions used by the federal agencies
change every 10 years with each US census (39). Further, while mixed race/ethnicity is rare
among self-identified Whites, about 30% of self –identified US-born Blacks and up to 50%
of Hispanics also consider themselves to be of mixed race (15,36). Regardless, the federally
designated classification system has been well validated and shown to precisely stratify
studied complex human diseases (30). Despite these limitations it is likely that the principal
finding of this study i.e. that PIAT distribution among populations is significantly variable
by race is generalizable. However, these data need to be confirmed other locations.
In summary we performed PFTs in patients successfully completing directly
observed therapy for pulmonary TB and demonstrated that patient’s who self-identified as
Whites were at increased risk for and more severe pulmonary impairment relative to other
racial groups. Additionally we found that pulmonary impairment was more frequent and
more severe in older persons irrespective of race. The risk for pulmonary impairment
remained after smoking and socioeconomic status were controlled. Further studies are
needed to define the etiology of these disparites.

Page 13

Page 14

Figure Legend
Figure 1. Study enrolment.
Figure 2. Comparisons of frequency and severity of pulmonary impairment between 317
self-identified racial and ethnic groups comprising 69 Whites, 85 Blacks, 82 Asians and 81
Hispanics. Figure 2 demonstrates that proportions impaired and the severity of impairment
significantly varies between racial and ethnic groups.
Figure 3. Comparisons of the frequency and severity of pulmonary impairment between 317
self-identified racial and ethnic groups comprising 69 Whites, 85 Blacks, 82 Asians and 81
Hispanics. Figure 3 shows that proportions impaired and the severity of impairment does not
varies with increase in socioeconomic status.
Figure 4. Comparison of the frequency of pulmonary impairment among all self-identified
racial groups by country of birth and smoking status.

Page 15

Figure 5.
Comparison of the probability for developing some pulmonary impairment (Panel A.) to that
of moderate/severe pulmonary impairment (Panel B) among all 317 patients who self-
identified as Whites, Blacks, Asians and Hispanics for ages 15 to 90 years. The median ages
for panel A are; Whites 58 years, Blacks 51 years, Asians 57 years and Hispanics 68 years.
For panel B the median age for Whites is 72 and that for Blacks is 63.
Tables and Figures
Figure 2.
Figure 3
WHITES
BLACKS
ASIANS
HISPANICS
0
20
40
60
80
100
None
Mild
Moderate
Severe
Non White
p=0.001
p<0.001
Impairment Levels
Race/Ethnicity
Percent