Tuberculosis (TB) is the term given to a wide and varied array of clinical diseases caused by the acid-fast bacilli Mycobacterium tuberculosis (MTB) and, to a much lesser extent, Mycobacterium bovis. Although proof of TB can be traced as far back as to early Egyptian remains, it did not become a major health issue until the industrialization of nations led to crowded living conditions facilitating the spread of this contagious disease. As a consequence, because of the use of large portions of the population coupled with an almost total lack of native resistance and effective treatment for this bacterium, TB became the basis of nearly one quarter of all adult deaths in Europe throughout the 1800s. Currently, it is estimated that one third (more than 2 billion people) of the world's population harbor this Mycobacterium and remain asymptomatic, a condition collectively referred to as having latent tuberculosis infection (LTBI). After peaking in the year 2003, the global incidence of TB now seems to be slowly declining. Nevertheless, according to estimates made public by the World Health Organization (WHO) in 2006, 14.4 million persons are estimated to have active TB infection, corresponding to a prevalence rate of 219 per 100,000 individuals.1 Furthermore, the approximate 9 million new cases of TB that occur each year translate into an incidence rate of 139 per 100,000 persons per year. In addition, second only to human immunodeficiency virus (HIV) infection as a cause of mortality worldwide resulting from a single infectious agent, TB is responsible for a death rate of 25 per 100,000 persons, which converts into the death of nearly 4700 people each day, or 2 million lost lives per year. Not surprisingly, TB has its greatest impact in developing nations. According to a 2009 WHO report,2 12 countries accounted for 70% of all reported cases of TB. The largest number of TB cases is reported from India. However, several sub-Sahara African countries, as well as China and the islands of Southeast Asia, have equally or higher infection rates per 100,000 population. Overcrowding, extreme poverty, the HIV epidemic (7% of all TB cases reported in 2006 were coinfected with HIV), and the burgeoning problem of multidrug-resistant TB strains all serve to intensify the challenge and magnify the difficulties of controlling TB. Recognizing this problem, in the 1990s, the WHO intensified their initiatives aimed at addressing this major global public health issue beginning with the directly observed therapy short course strategy, and more recently, a plan entitled Stop TB Partnership: The Global Plan to Stop TB 2006-2015.3,4 The latter program was directed more at accurate diagnosis and patient-centered treatment adherence and included several additional control components collectively devised to prevent and control TB. These new Millennium Development Goals included (1) to arrest and reverse the rising incidence of TB by 2015, (2) to cut in half the 1990 TB prevalence and death rate by 2015, and (3) to reduce the worldwide incidence of TB to 1 per million or fewer individuals by 2050.
The landscape of TB in the United States is substantially different than that witnessed in poorer countries and the developing world. Up until a nadir in 1984, there was a steady decline in reported TB cases in the United States. Beginning in 1985, an unexpected amount of "excess cases" was observed and eventually peaked in 1992 at 10.5 cases per 100,000 persons. This resurgence in TB cases in the United States was believed to be related to several factors, some of which included neglected TB control programs, the HIV/acquired immunodeficiency syndrome (AIDS) epidemic, increased immigration of persons from countries with high rates of TB, and urban homelessness. As a result of the success of implemented concerted efforts by federal and public health agencies, a significant diminution in the number of new US cases followed, with the 2009 incidence of TB in the United States at 3.8 cases per 100,000 population being the lowest in recorded history.5 At the present time, most of the TB cases reported in the United States occur in foreign-born individuals emigrating from areas of the world with high rates of endemic TB. In 2006, approximately 57% of US-reported TB (case rate, 22 per 100,000 persons) occurred in this group of people. The states accounting for the greatest number of cases (California, New York, New Jersey, Texas, and Florida) are in direct correlation with the numbers of immigrants from countries with high TB prevalence. Of epidemiologic importance is that sophisticated laboratory analysis has demonstrated that most of these cases were the result of reactivation of LTBI rather than transmission within their current living areas.
Although those living with HIV/AIDS make up approximately less than a third of a percentage point (0.3%) of the US population, according to data derived from the US National TB Surveillance System for 1993-2005, approximately 9% of patients with TB were coinfected with HIV.6 Other pertinent risk factors for TB in the United States are those that are directly related to the host, including systemic illnesses (eg, diabetes mellitus and renal insufficiency), medications that impair the immune system (eg, tumor necrosis factor-alpha inhibitors, systemic corticosteroids, and postorgan transplantation drugs), lifestyle habits (eg, alcohol and tobacco), and age. Socioeconomic status and environmental conditions also play a role in the risk of developing TB.
Historically, the tuberculin skin test (TST) has been relied on for diagnosing TB in persons who have been sensitized by MTB and have LTBI.
In a TST survey conducted in 2000, an estimated 4.2% of the civilian, noninstitutionalized US population older than 1 year had LTBI. Although this represented a 60% decline7 from 1972, the decrease in prevalence was not uniform across all the segments of the population.
Furthermore, approximately 9.4% of the 153,555 persons with active TB during the 10-year period between 1998 and 2007 died either before treatment was started or during therapy before completing the anti-TB regimen. Because the rates of MTB infection and active TB vary considerably, targeted testing and the selection of those persons likely to benefit from treatment for latent infection (ie, persons who are at increased risk of poor clinical outcome if active infection ensues) are assigned a high priority.
In this issue of Infectious Diseases in Clinical Practice, Barragan et al report a retrospective review over a 12-year period of a major inner city hospital in the United States to identify cases of TB diagnosed at the time of death or postmortem. The authors identified 35 cases considered as missed TB diagnosis (from a total of 1608 TB cases), and of these, 33 (94%) were culture-confirmed cases, whereas 2 (6%) were presumed to be TB based on clinical and epidemiological factors. Included were (1) patients who died during their hospital stay (>72 hours of hospitalization) and in whom the diagnosis of TB was identified by postmortem culture results and autopsy findings, and (2) patients dying within 72 hours of their presentation to the hospital, with clinical manifestations compatible with miliary TB (progressive and widely disseminated TB) and in whom a diagnosis of TB was made postmortem. Six (17%) of the 35 cases were diagnosed at autopsy. In all these cases, there was no clinical suspicion of TB. Pulmonary TB accounted for 27 (77%) of the discovered missed cases, whereas the remaining 8 (23%) consisted of extrapulmonary TB including peritonitis (1 case), meningitis (1 case), and disseminated disease (6 cases). Interestingly, all of the patients were African American, 18 (51%) were found to be HIV coinfected (median CD4 T-cell lymphocyte count, 49 cells/μL), and 60% had a medical encounter within 90 days before their death, of which 11 (52%) were hospital admissions. Most (76%) of the study patients presented with symptoms that in retrospect were clinically consistent with TB. However, a diagnosis of TB was considered for only 14 patients, and only 12 received presumptive anti-TB therapy within 72 hours before dying. The authors emphasize that in this day and age, an unfortunate consequence of late TB diagnosis is an unacceptably high complication and mortality rate as well as potential major public health implications with regard to transmission of TB in the community and in the hospital.
Taking into account the great deal of attention given to the epidemiology, diagnosis, prevention, and treatment of TB throughout the past 2 decades, the question must be asked as to why TB still remains an elusive diagnosis. Plausible explanations can likely be divided into 2 main categories classified as clinical and laboratory derived. Nothing can substitute for a sharp clinical acumen. In general, establishing a clinical diagnosis of TB includes the assessment of various factors, not the least of which are systemic symptoms and physical findings. Clearly, as illustrated in Barragan's series, atypical presentations along with comorbid conditions can cloud the clinical picture, thereby delaying or rendering the diagnosis of TB obscure. In the era of HIV/AIDS, there is a greater than the usual proportion of patients with extrapulmonary TB as well as concurrent infections that may serve to sway the differential diagnosis away from TB and into the direction of other maladies. Knowledge of prior exposure to MTB, either by TST or interferon-gamma release assays (IGRAs) or by suspicious chest radiographic findings, should alert the physician to the possibility that in the appropriate setting, their patient's infectious syndrome may be TB. Epidemiologic factors, including past or present residence in or travel to a TB-endemic area or belonging to a group that, perhaps owing to socioeconomic disadvantages, places them at higher risk for TB, warrant important consideration.
Whereas the diagnosis of TB has classically relied on the isolation of MTB from a respiratory tract, tissue, or fluid samples, in some circumstances, confirming a definitive laboratory diagnosis of TB may not be achievable. Nearly 15% to 20% of clinically diagnosed TB cases lack bacteriologic isolation of MTB.8 However, all clinical specimens should be appropriately processed and submitted for culture, which is also important for drug susceptibility testing. DNA probes can give results within a few hours, with a sensitivity and specificity approaching nearly 100% but cannot be used directly on clinical specimens. Although lacking the sensitivity and specificity associated with mycobacterial cultures, the detection of acid fast bacilli on microscopic smears of stained specimens submitted to the laboratory remains the most rapid and inexpensive TB diagnostic tool. Nucleic acid amplification assays offer another more modern technique for the direct detection of MTB in clinical specimens including nonrespiratory secretions. They can provide a rapid diagnosis (24-48 hours) and are currently recommended in patients for whom the suspicion for TB is moderate to high. The sensitivity and specificity of a nucleic acid amplification assay exceeds 95% for acid fast bacilli smear-positive specimens but has been found to have a lower sensitivity, ranging from 40% to 70%, for smear-negative cases.9 The IGRAs are novel laboratory tools, which, when compared to the standard TST, have the potential advantages of heightened sensitivity and specificity, less interpreter bias, and test results in less than 24 hours.10 Besides their use in detecting LTBI, the newest of the IGRAs have also been approved for diagnosing active TB infection.
Unfortunately, the retrospective nature of the study by Barragan et al does not allow using any or all of the aforementioned diagnostic tests to determine whether or not TB could have or should have been accurately diagnosed in the 35 missed cases. The authors suggest that in treating patients with TB risk factors, in particular HIV/AIDS, in urban medical settings with high prevalence of TB, physicians need to have a low threshold of suspicion in diagnosing TB and perhaps should consider treating empirically for TB, pending the results of diagnostic tests. This practice is not new to physicians in countries with very high prevalence of TB infection and where access to advanced diagnostic methods is limited.
To quote John F. Kennedy, "there are risks and costs to action, but they are far less than the long range risks of comfortable inaction."