Abstract
Risk of active TB after the first year should not be ignored: extended follow-up in adult TB contacts is needed http://ow.ly/I9t28
To the Editor:
To eliminate tuberculosis (TB) on a global scale, the identification and neutralisation of latently infected high-risk individuals is of paramount importance [1]. Active TB contacts are well documented with an increased risk for both latent TB infection (LTBI) and the development of active TB disease [2]. Contact investigation is an important and effective active case-finding strategy, but also requires abundant public health resources [3].
The current World Health Organization (WHO) recommendation for contact investigation suggests that follow-up screening should be considered particularly in the first year after exposure, such as after 6 or 12 months [4]. However, most of TB contacts studies have included both adults and children, or enrolled children only. It remains uncertain if the 1-year follow-up period is adequate for all TB contacts [5]. To optimise the contact investigation strategy for adult TB contacts, understanding the dynamic pattern of active TB development is important. Characterising the clinical features of adult TB contacts with a higher risk for active TB progression is helpful to prioritise medical resources more efficiently. It also enables researchers to select individuals who will benefit the most from LTBI treatment in resource-limited areas.
To elucidate the issue, we conducted a nationwide population-based cohort study using data retrieved from the National Health Insurance Research Database (NHIRD) in Taiwan between January 2000 and December 2011. TB contacts were identified by diagnostic codes (V01.1 in the International Classification of Diseases (ICD), 9th revision, and the clinical modification ICD-9-CM) in conjunction with chest plain film radiography and/or sputum acid-fast smear examination and/or tuberculin skin test (TST). We also matched each TB contact with four nonexposed subjects by age and sex, and presence of comorbidities on the same index date of diagnosis. The TB contacts cohort and the matched cohort were followed until the development of active TB disease, which was defined by compatible ICD-9-CM codes (010–018 in ICD-9-CM) [6, 7]. The diagnosis of TB was validated by the prescription of at least two anti-TB medications for more than 28 days. If no TB occurred, subjects were followed until death, or the end of the study period (December 2011). Incidence rates (per 100 000 person-years) of active TB were analysed, and hazard ratios were calculated as compared with matched controls. For multivariate adjustment, the Cox proportional hazards model was used to compute hazard ratios (HRs) and 95% confidence intervals. All statistical analyses were performed using SAS 9.2 software (SAS Institute Inc., Cary, NC, USA).
A total of 8659 TB contacts were identified during the study period. Their median age was 44 years old, 13.8% were ⩾65 years of age, and 44.2% of them were male. This cohort was matched with 34 636 individuals without TB contact history in terms of age, sex and major comorbidities. The median (interquartile range (IQR)) follow-up times for TB contacts and their matched controls were 2.25 (1.06–3.69) years and 2.27 (1.07–3.70) years, respectively. Overall, 72 TB contacts and 67 individuals from the matched cohort were documented with active TB during the follow-up period, and the mean annual incidences of TB were 312.8 and 72.2 cases per 100 000 person-years, respectively (adjusted HR 4.39, 95% CI 3.15–6.12; p<0.001). The Kaplan–Meier analysis and annual incidences of active TB in TB contacts and control participants are shown in figure 1. The annual TB incidence among TB contacts after exposure in the first, second, and third years were 567.1 cases, 106.4 cases, and 229.3 cases per 100 000 person-years, respectively. The first year after the exposure showed the highest risk for active TB (adjusted HR 12.30, 95% CI 6.73–22.48), and the adjusted HRs from year two to year four were 2.11 (95% CI 1.20–3.73). In multivariate analysis, the independent risk factors associated with active TB development among adult TB contacts were age ⩾65 years old (HR 2.15, 95% CI 1.22–3.81), male sex (HR 1.74, 95% CI 1.07–2.81), and having autoimmune diseases (HR 2.60, 95% CI 1.40–4.81), chronic obstructive pulmonary disease (COPD) (HR 1.95, 95% CI 1.16–3.27) or liver cirrhosis (HR 2.61, 95% CI 1.03–6.65).
Analysing the trend of active TB development after exposure helps to improve our contact investigation strategy. In a large meta-analysis that included more than 200 studies, Fox et al. [2] reported that the incidence rate ratio (IRR) among contacts during the first year was 46.6 (IQR 3.2–68.0) in high-income countries and 15.9 (IQR 2.6–21.4) in low- to middle-income countries. In their meta-analysis, the IRR dropped in the second year and rose again in the third year. Interestingly, a similar dynamic pattern was found in the present study. As the information used in this study was obtained from a health insurance database, many important factors associated with TB exposure were not available, such as the closeness of contact and smear positivity and disease severity of the index case. The limited number of cases with active TB after the first year also made it difficult to compare the TB incidence in each single year. However, when we combined the cases from year two to year four we still found that the incidence of active TB was significantly higher in adult TB contacts, with a HR of 2.11. Due to the limitations mentioned earlier, the dynamic pattern of active TB occurrence in the present study could come from reporting bias or chance. Nonetheless, we also speculate that it could be related to our current contact investigation strategy. This is based on current WHO recommendations [4] with active case-finding in the first year after exposure leading to early diagnosis of some TB cases, which in turn masks the incidence curve of active TB in the second year. As active case-finding stopped after the first year, the true impact of TB exposure on the TB incidence is shown in the third year. Our findings suggest the potential limitation of a 6–12-month follow-up for adult TB contacts. A well-designed, prospective study is required for verification.
Plenty of studies have identified several independent risk factors for LTBI and active TB among TB contacts [5, 8–12]. However, studies which focus on adult TB contacts, and especially on analysing underlying comorbidities that increase their susceptibility to active TB, are much more limited. To the best of our knowledge, the present study is the first to focus on adult TB contacts and include details of comorbidities for analysis. We found that older age, male sex, and having autoimmune diseases, COPD or liver cirrhosis were independent predictors for TB development among adult TB contacts. The depressed immunity related to autoimmune diseases and liver cirrhosis obviously increases the risk for active TB [13, 14]. Animal and cellular studies have demonstrated the negative impact of smoke exposure on the local immunity of the lungs [15]. These risk factors are of pivotal importance and should never be overlooked in our practice of contact investigation.
TB contacts are high-risk individuals for active TB development and contact investigation is an effective way of active case-finding. This population-based cohort study demonstrated that the annual incidence of active TB in adult TB contacts is highest during the first year following exposure, but it also remains significantly higher after the first year as compared with a matched cohort. An intensive and frequent follow-up strategy in contact investigation is definitely needed during the first year after exposure. However, the increased risk of active TB after the first year should not be ignored and a risk-stratified strategy to extend the follow-up period in targeted adult TB contacts is needed. The underlying comorbidities that increase the risk of active TB development could be included in contact investigation and evaluation.
Footnotes
Conflict of interest: None declared.
- Received July 9, 2014.
- Accepted January 3, 2015.
- Copyright ©ERS 2015