Factors Associated With Receiving Longer Than Recommended Therapy Among Culture-Negative Pulmonary Tuberculosis Patients

Abstract

Background

US tuberculosis (TB) guidelines recommend treatment ≥6 months with a regimen composed of multiple effective anti-TB drugs. Since 2003, a 4-month regimen for a specific subset of TB patients has also been recommended.

Methods

We used 2011–2018 US National Tuberculosis Surveillance System data to characterize factors associated with 4-month (111–140 days) therapy among adult patients who had completed treatment and were potentially eligible at that time for 4-month therapy (culture-negative pulmonary-only TB, absence of certain risk factors, and initial treatment that included pyrazinamide). We used modified Poisson regression with backward elimination of main effect variables to calculate adjusted relative risks (aRRs).

Results

During 2011–2018, 63 393 adults completed TB treatment: 5560 (8.8%) were potentially eligible for 4-month therapy; of these, 5560 patients (79%) received >4-month therapy (median, 193 days or ∼6 months). Patients with cavitary disease were more likely to receive >4-month therapy (aRR, 1.10; 95% CI, 1.07–1.14) vs patients without cavitary disease. Patients more likely to receive 4-month therapy included patients treated by health departments vs private providers only (aRR, 0.94; 95% CI, 0.91–0.98), those in the South and West vs the Midwest, non-US-born persons (aRR, 0.95; 95% CI, 0.91–0.99) vs US-born persons, and aged 25–64 years vs 15–24 years.

Conclusions

Most patients potentially eligible for 4-month therapy were treated with standard 6-month courses. Beyond clinical eligibility criteria, other patient- and program-related factors might be more critical determinants of treatment duration.

Most culture-negative pulmonary tuberculosis patients potentially eligible for 4-month therapy in the United States were treated with longer courses. These analyses indicate that certain patient- and program-related factors might affect treatment duration.

Advances in reduced duration of tuberculosis (TB) treatment have the potential to dramatically change TB treatment in the United States by improving quality of life, increasing treatment completion, and reducing costs. While only recently have studies been published broadly demonstrating the efficacy of shortened anti-TB regimens for patients with drug-susceptible and multidrug-resistant TB [1–5], TB treatment guidelines from the American Thoracic Society (ATS), Centers for Disease Control and Prevention (CDC), and the Infectious Diseases Society of America (IDSA) have recommended a shortened 4-month regimen since 2003 for a specific subset of patients with TB [6]. This subset, TB patients ≥15 years old with culture-negative pulmonary disease and certain clinical characteristics treated with a 4-month regimen, was shown, >2 decades ago, to achieve clinical outcomes comparable to those of patients treated with a full 6-month regimen [7]. This 4-month regimen consists of an initial phase of 2 months of isoniazid, rifampin, pyrazinamide, and ethambutol, followed by a shortened continuation phase of 2 months of isoniazid and rifampin. In a study evaluating the efficacy of 4 months of treatment (regimens inclusive of isoniazid, rifampicin, and pyrazinamide) among 314 smear-negative pulmonary TB patients, patients were randomized to receive a daily or combined (daily and intermittent) regimen for a total of 4 months if culture-negative and 6 months if culture-positive [8]. There was no relapse at 30 months among 96 culture-negative patients receiving daily treatment nor among 68 patients assessed at 60 months; there was only 1 relapse at 30 months among 100 culture-negative patients receiving combined treatment and among 74 patients assessed at 60 months, demonstrating the effectiveness of a 4-month regimen in the treatment of nonimmunocompromised patients with smear- and culture-negative TB [8]. Based on these and other clinical data, the 2016 ATS/CDC/IDSA guidelines continue to recommend the 4-month regimen for HIV-negative patients with a negative sputum culture who have shown symptomatic or radiographic improvement on TB therapy [9].

Factors associated with receipt of longer than recommended therapy among those eligible for shortened 4-month therapy on a national scale in the United States have not been well described. Identifying these factors may highlight areas for further evaluation and potentially intervention to increase awareness about TB patients who are eligible for shortened therapy, particularly as broader initiatives to shorten TB therapy move forward. Using TB national surveillance data, this study assesses factors associated with completion of a longer treatment course among culture-negative patients eligible for a shortened 4-month treatment course.

METHODS

Data Source

We analyzed TB cases reported to the CDC through the National TB Surveillance System (NTSS) from 2011 to 2018. TB surveillance data are reported by 60 US reporting regions—the 50 US states, District of Columbia, New York City, American Samoa, Guam, Puerto Rico, US Virgin Islands, Commonwealth of the Northern Mariana Islands, Federated States of Micronesia, Republic of the Marshall Islands, and Republic of Palau—and include information on demographics, risk factors, clinical characteristics, laboratory results, sites of disease, and drug treatments.

Inclusion and Exclusion Criteria

Patients were included if they met the suggested criteria to receive a shortened 4-month regimen based upon guidelines by the ATS, CDC, and IDSA: those with pulmonary TB only, negative cultures, initial treatment that included pyrazinamide, a negative HIV test result, and the absence of other causes of immunocompromise, including prior organ transplantation, tumor necrosis factor–alpha antagonist therapy, and end-stage renal disease [6]. Patients who did not complete treatment, were aged <15 years, had miliary TB, or were contacts of multidrug-resistant TB patients within the past 2 years were excluded from analysis (Figure 1).

Figure 1.

Determination of study cohort. We selected patients from the National Tuberculosis Surveillance System from 2011–2018 and excluded patients if they did not meet the criteria to receive a shortened 4-month regimen based upon guidelines by the American Thoracic Society, Centers for Disease Control and Prevention, and Infectious Diseases Society of America [6]. Numbers presented for those excluded are mutually exclusive, extracted at each step of exclusion (progressing from top to bottom). Abbreviation: TB, tuberculosis.

Outcome and Predictor Variables

The outcome in this analysis was receipt of longer than recommended 4-month therapy among patients eligible for a shorter regimen. Consistent with prior studies, our study defined short therapy duration as 4 months [10]; after examining a natural break in the data set and to be consistent with the programmatic setting due to missed doses or intolerance, we included cases with specific therapy duration 111–140 days (∼±2 weeks) to be interpreted as 4 months of therapy. Patients who were eligible for shortened therapy but took >140 days to complete therapy were classified as receiving >4-month therapy.

Based on the literature and NTSS data availability, a combination of patient- and program-related factors were considered as predictors for treatment completion. These variables included type of TB health care provider (public health department, which includes at least 1 visit with a public health department vs private provider only); race/ethnicity using standard census categories; age group (15–24 years, 25–44 years, 45–64 years, and ≥65 years); sex; origin of birth (US-born and non-US-born); US Census region of residence (Midwest, Northeast, South, West, US-affiliated Pacific Islands [USAPIs], and Puerto Rico/US Virgin Islands); clinical variables (diabetes mellitus, abnormal chest radiograph, evidence of cavitary disease, and previous TB); and social risk factors (past-year injection or noninjection drug use, residence in a long-term care or correctional facility at the time of diagnosis, experiencing homelessness within the last 12 months, and unemployment).

Data Analysis

Bivariate and adjusted analyses were conducted to determine associations between potential predictors and our outcome. We used a modified Poisson regression model with backward elimination of all main effect variables. We evaluated 2-level interactions for all the main effects. Observations with missing data were excluded. All statistical analyses were conducted using an alpha level of .05 (SAS, version 9.4; SAS Institute, Cary, NC, USA). This project was determined not to be human subjects research by the CDC and did not require approval by an institutional review board because data were collected and analyzed as part of routine public health surveillance.

RESULTS

Descriptive Statistics

During 2011–2018, 63 393 adults with TB completed treatment. Of the 5560 (8.8%) patients who were eligible for 4-month therapy, 1182 (21.3%) received 4-month therapy and 4378 (78.7%) received >4-month therapy, with a median therapy duration of 193 days (ie, >6 months) (Table 1). Most eligible patients received outpatient treatment for TB at a public health department (82.9%). Among all eligible patients, 5092 cases were reported in the United States and 468 were reported in the USAPIs or Puerto Rico/US Virgin Islands. The majority of eligible patients were male (62.3%), non-US-born persons (72.0%), and aged 25–64 years (69.6%). Approximately one-third of eligible patients were Asian, (30.5%), 26.8% were Hispanic, and 20.6% were Black or African American. Almost 40% were from the South (39.4%), compared with only 13.5% from the Northeast. Almost all eligible patients had an abnormal chest radiograph (97.2%), and 24.9% had evidence of cavitary disease. Diabetes mellitus was reported in 13.5% of eligible patients, and 28.4% were unemployed (Table 1).

Table 1.

Characteristics of Adult TB Patients who Completed TB Therapy and Were Eligible for a 4-Month Course, 2011–2018

Characteristic	Completed Therapy
	(n = 5560) No. (%)
Therapy duration, median (IQR), d	193 (183–222)
Treatment regimen
Completed 4-mo therapya	1182 (21.3)
Completed >4-mo therapy	4378 (78.7)
Race/ethnicityb
White	769 (13.8)
American Indian or Alaskan Native	47 (0.8)
Asian	1697 (30.5)
Black or African American	1145 (20.6)
Hispanic	1489 (26.8)
Multiple race	10 (0.2)
Native Hawaiian or Pacific Islander	380 (6.8)
Missing or unknown	23 (0.4)
Age group
15–24 y	771 (13.9)
25–44 y	1839 (33.1)
45–64 y	2028 (36.5)
65+ y	922 (16.6)
Sex
Female	2093 (37.6)
Male	3466 (62.3)
Missing or unknown	1 (0.0)
Origin of birth
US-born	1553 (28.0)
Non-US-born	4004 (72.0)
Missing or unknown	3 (0.1)
US Census region
Midwest	727 (13.1)
Northeast	751 (13.5)
South	2193 (39.4)
West	1421 (25.6)
US-affiliated Pacific Islands	432 (7.8)
Puerto Rico and US Virgin Islands	36 (0.6)
Type of provider
Health department	4609 (82.9)
Private/other	800 (14.4)
Missing	151 (2.7)
Medical risk factors
Abnormal chest radiograph	5407 (97.2)
Cavitary disease	1382 (24.9)
Previous TB	510 (9.2)
Diabetes mellitus	749 (13.5)
Social risk factors
Resident of long-term care facilityc	68 (1.2)
Resident of correctional facilityc	380 (6.8)
Drug use	361 (6.5)
Homelessnessd	308 (5.5)
Unemployedd	1582 (28.4)

Abbreviations: IQR, interquartile range; TB, tuberculosis.

^aAppropriate therapy defined as 111–140 days.

^bAll races are non-Hispanic.

^cAt the time of TB diagnosis.

^dWithin the last 12 months.

Adjusted Associations

For the multivariable model, we started with all variables assessed in the unadjusted analysis and conducted backward elimination. A modified Poisson regression model was fit, resulting in a final model consisting of race/ethnicity, age, origin of birth, cavitary disease, experiencing homelessness within the year preceding TB diagnosis, being a resident of a correctional facility at the time of TB diagnosis, type of outpatient health care provider, and region of residence, and included 4571 patients (Table 2). Patients seen by health departments vs private providers only (adjusted risk ratio [aRR], 0.94; 95% CI, 0.91–0.98), those in the South (aRR, 0.88; 95% CI, 0.84–0.92) and West (aRR, 0.92; 95% CI, 0.88–0.96) vs Midwest, non-US-born persons (aRR, 0.95; 95% CI, 0.91–0.99) vs US-born persons, and those aged 25–44 (aRR, 0.93; 95% CI, 0.88–0.97) and 45–64 years (aRR, 0.94; 95% CI, 0.89–0.99) vs 15–24 years were more likely to receive 4-month therapy (Table 2). Persons with cavitary disease were more likely to receive >4-month therapy (aRR, 1.10; 95% CI, 1.07–1.14) than persons without cavitary disease. Patients who experienced homelessness in the past year and were also residents of a correctional facility at the time of diagnosis were more likely to receive >4-month therapy (aRR, 1.21; 95% CI, 1.06–1.39) than persons who did not have both of these conditions (Table 2).

Table 2.

Adjusted Associations Between >4-Month Therapy and Predictors, United States

Characteristic	Total	4-Month Therapya	>4-Month Therapy	aRR	95% CI		P Value
	(n = 4571)	(n = 1060) No. (%)	(n = 3511) No. (%)
Health care provider type
Health department	3867	924 (87.2)	2943 (83.8)	0.94	0.91	0.98	.007
Private/other	704	136 (12.8)	568 (16.2)	Ref.
Race/ethnicityb
White	708	159 (15.0)	549 (15.6)	Ref.
American Indian or Alaskan Native	40	6 (0.6)	34 (1.0)	1.06	0.92	1.21	.443
Asian	1457	308 (29.1)	1149 (32.7)	1.05	0.99	1.11	.102
Black or African American	1021	286 (27.0)	735 (20.9)	0.95	0.90	1.01	.107
Hispanic	1286	288 (27.2)	998 (28.4)	1.04	0.99	1.10	.135
Multiple race	8	1 (0.1)	7 (0.2)	1.12	0.87	1.44	.386
Native Hawaiian or Pacific Islander	51	12 (1.1)	39 (1.1)	1.04	0.89	1.22	.601
Age group
15–24 y	621	118 (11.1)	503 (14.3)	Ref.
25–44 y	1498	383 (36.1)	1115 (31.8)	0.93	0.88	0.97	.003
45–64 y	1683	405 (38.2)	1278 (36.4)	0.94	0.89	0.99	.011
65+ y	769	154 (14.5)	615 (17.5)	0.98	0.93	1.04	.562
Origin of birth
US-born	1342	301 (28.4)	1041 (29.6)	Ref.
Non-US-born	3229	759 (71.6)	2470 (70.4)	0.95	0.91	0.99	.026
US Census region
Midwest	699	118 (11.1)	581 (16.5)	Ref.
Northeast	428	90 (8.5)	338 (9.6)	0.95	0.90	1.01	.119
South	2066	556 (52.5)	1510 (43.0)	0.88	0.84	0.92	<.001
West	1378	296 (27.9)	1082 (30.8)	0.92	0.88	0.96	<.001
Cavitary disease
No	3391	849 (80.1)	2542 (72.4)	Ref.
Yes	1180	211 (19.9)	969 (27.6)	1.10	1.07	1.14	<.001
Resident of correctional facilityc
No	4234	977 (92.2)	3257 (92.8)	N/A	N/A	N/A	N/A
Yes	337	83 (7.8)	254 (7.2)	N/A	N/A	N/A	N/A
Homelessnessd
No	4295	992 (93.6)	3303 (94.1)	N/A	N/A	N/A	N/A
Yes	276	68 (6.4)	208 (5.9)	N/A	N/A	N/A	N/A
Homelessnessd
Resident of correctional facilityc - no	233	63 (5.9)	170 (4.8)	0.94	0.87	1.02	.157
Resident of correctional facilityc - yes	43	5 (0.5)	38 (1.1)	1.21	1.06	1.39	.005

Abbreviations: aRR, adjusted relative risk; N/A, not applicable due to effect modification of homelessness with resident of correctional facility; TB, tuberculosis.

^aAppropriate therapy defined as 111–140 days.

^bAll races are non-Hispanic.

^cAt the time of diagnosis.

^dWithin the last 12 months.

DISCUSSION

Prior studies have shown the efficacy of a shortened 4-month regimen among some nonimmunocompromised patients diagnosed with culture-negative TB [8]. Despite ATS, CDC, and IDSA recommendations for shortened regimens, our study found that only 21% of TB patients eligible for shortened therapy in the United States received shortened therapy during 2011–2018.

Provider Type and Region

In the modified Poisson model, culture-negative TB patients who had been seen by health departments at least once were more likely to receive 4-month therapy than those who were seen by private providers only. In a Chicago study conducted by Fletcher and colleagues, patients of private providers had treatment durations 48 days longer than patients of health department providers [11]. In the United States, 77% of US TB patients receive most or all of their clinical care through health departments [12], and many US private sector clinicians have never seen or treated a person with TB disease [13]. This could be another reason why health departments may have more awareness of TB treatment nuances. In the adjusted analyses, we found US region to be significantly associated with treatment duration among eligible patients. After adjusting for race/ethnicity, age, origin of birth, diabetes, cavitary disease, homelessness, residency in a correctional facility at the time of TB diagnosis, and type of provider, we found that the likelihood of receiving 4-month therapy was higher among eligible patients in the South and West compared with eligible patients in the Midwest. In 2019, rates in the West South Central and Pacific regions were above the US national average of 2.7 cases per 100 000 population (3.4 cases per 100 000 population and 4.8 cases per 100 000 population, respectively) [14]. These areas include 2 states, Texas and California, that have some of the highest TB rates in the nation [15]. With high incidence of TB in these areas, local and state TB programs may have more experience and practical expertise in evidence-based practices, fostering promotion of shorter treatment regimens among eligible patients in their regions.

Origin of Birth and Age

Non-US-born persons with TB were more likely to receive 4-month therapy compared with US-born persons. In a study done by Mitruka and colleagues, movement and loss to follow-up were important factors in distinguishing treatment durations and completion between US born and non-US-born persons [16], which were not examined in our analysis of patients who completed treatment. Our data show that 71.4% of TB patients who moved outside the United States and were eligible for 4-month therapy had a transnational referral. Factors such as involvement with health care providers and programs outside the United States, as well as extended travel to visit family outside the country, could influence prescribing patterns.

Our study showed that patients aged 15–24 years were more likely to receive >4-month therapy than those aged 25–64 years. Because the majority of TB cases occur among the age group of 25–64 years [15], clinicians may be more comfortable prescribing a shortened regimen among this group. In addition, because the ATS/CDC/IDSA guidelines recommend a 4-month regimen in culture-negative adults [6], the younger group of eligible 15–24-year-olds might be less likely to be prescribed shortened therapy.

Clinical Conditions and Social Risk Factors

Our study also found that eligible patients with cavitary disease were significantly more likely to complete >4-month therapy. Previous studies have shown that relapse was significantly associated with initial presence of cavitary lesions among sputum culture–positive patients compared with patients with no cavitary disease and sputum culture–negative patients [6]. In addition, drug resistance is more likely to occur in patients with cavitary disease because of the abundant number of rapidly multiplying bacilli in cavities [17]. For these reasons, clinicians might be hesitant to provide shortened therapy among eligible patients with cavitary disease.

TB patients who were incarcerated at the time of diagnosis with an experience of homelessness within the past year were more likely to complete >4-month therapy. Mitruka and colleagues found that persons who were incarcerated at the time of diagnosis and had experienced homelessness in the past year were less likely to complete standard 6-month TB therapy within 12 months. Among subpopulations found to be at highest risk for failure to complete therapy within 12 months, residents of a correctional facility at TB diagnosis were less likely to complete treatment; importantly, risk of no documented completion of therapy was 4 times higher for residents of a correctional facility at the time of TB diagnosis than for those who did not reside in a correctional facility. In an analysis of persons with TB who had recently experienced homelessness, being incarcerated at the time of TB diagnosis increased the risk of not completing treatment [18].

Our study provides insight into understanding factors associated with >4-month therapy among culture-negative pulmonary TB patients and can be used as a model for future evaluations of shortened TB regimens during real-world implementation. Our study also determines gaps in awareness among TB patients who are eligible for 4-month therapy and the potential need to enhance educational efforts about shortened TB regimens to assist TB patients in improving quality of life, enhancing treatment completion, and reducing costs. Of note, a recent study by the CDC and the National Institutes of Health published results from a randomized controlled trial demonstrating the efficacy and safety of a 4-month anti-TB regimen among culture-positive TB patients [1]. Another recent study among children aged <16 years with smear-negative TB suggested noninferiority of a 4-month regimen compared with the standard 6-month treatment, with no statistical difference in relation to treatment failure, TB recurrence, death of any cause, or loss to follow-up [2]. Comparable strides are taking place in treatment shortening for multidrug-resistant TB, with publication of results supporting a regimen one-third the duration of standard treatment for extensively drug-resistant TB [3], and a recent presentation of results supporting shortened regimens for all forms of multidrug-resistant TB [4].

Limitations

This study has several limitations. The NTSS does not capture a full clinical history or treatment course, such as any protocols used to ensure or any clinical concerns about adequate sputum culture collection, length of therapy prescribed, all comorbidities, adverse drug events, treatment interruptions, and follow-up clinical and radiographic data. As a result, these surveillance data cannot be used for clinical inference. Furthermore, while all patients who were reported to have completed treatment were included in our study, it is unknown if completion of treatment was patient- or provider-driven. The generalizability of the surveillance data used in this analysis is also limited because this study only included patients who met specific criteria for eligibility for shortened treatment. Another limitation is that the number of baseline negative cultures for a patient could influence provider recommendations around duration of treatment for an individual patient, but this cannot be assessed with the NTSS. As recommended in prior studies and guidelines, eligibility for a shortened 4-month treatment regimen was restricted to those with pulmonary TB with a negative culture, HIV-negative status, and absence of any other immunosuppression.

CONCLUSIONS

Shortened treatment among some patients has been shown to yield outcomes comparable to longer treatments. A shortened treatment regimen following current guidelines holds promise for increasing adherence, improving treatment completion, and reducing health care provider time and costs for TB treatments. However, multiple factors might need to be considered in a programmatic setting before setting expectations that all patients potentially eligible for shortened therapy will ultimately complete therapy in a shortened time frame. Results from our study suggest that certain patient- and program-related factors were associated with shorter and longer TB treatment durations and should be considered in the larger context by patients, providers, and policy-makers. Furthermore, more education and awareness may be needed about TB patients who are eligible for shortened therapy. Other system-based initiatives and factors not collected in the NTSS (eg, clinician awareness of the 4-month therapy recommendations) might also be important influences on TB patient treatment durations; additional data collection specific to this issue might be warranted.