ORIGINAL ARTICLE
INFECTIOUS DISEASES
Mycobacterium tuberculosis infection of end-stage renal disease patients
in Taiwan: a nationwide longitudinal study
S.-Y. Li1, T.-J. Chen2, K.-W. Chung3, L.-W. Tsai4, W.-C. Yang5, J.-Y. Chen5 and T.-W. Chen6,7
1) Division of Nephrology, Department of Medicine, Taipei Veterans General Hospital and Institute of Clinical Medicine, 2) Department of Family Medicine,
Taipei Veterans General Hospital and School of Medicine, National Yang-Ming University, Taipei, 3) Software Development Director, Rhematech Inc.,
4) Evidence Based Medicine Centre, Taipei Medical University Hospital, 5) Division of Nephrology, Department of Medicine, Taipei Veterans General
Hospital and School of Medicine, National Yang-Ming University, 6) Division of Nephrology, Department of Medicine, Taipei Medical University Hospital and
7) Department of Medicine, School of Medicine, College of Medicine, Taipei Medical University Hospital, Taipei, Taiwan
Abstract
End-stage renal disease (ESRD) patients are vulnerable to Mycobacterium tuberculosis (MTB) infection, but the magnitude of the risk is
uncertain. In addition, there is no reliable information on the MTB infection rate of patients undergoing different types of renal replacement therapy (RRT). We used the Taiwan National Health Insurance Research Database to conduct a 9-year nationwide longitudinal
study. Among 49 983 ESRD patients who received three renal replacement modalities, there were 562 cases of MTB infection, corresponding to an incidence rate of 3.0 per 1000 patient-years. The risk of MTB infection relative to the general population was 4.5. Multivariate Cox regression analysis indicated that the independent risk factors for MTB infection in ESRD patients are old age (hazard ratio
(HR) 1.17 per 10 years, p <0.001), male gender (HR 1.37, p <0.001), silicosis (HR 5.82, p <0.001), and chronic obstructive pulmonary
disease (HR 1.68, p 0.012). Hyperlipidaemia (HR 0.71, p <0.001) and hypertension (HR 0.81, p 0.05) are associated with a lower infection rate. There was no effect of RRT modality on MTB infection rate.
Keywords: End-stage renal disease, Mycobacterium tuberculosis, renal replacement therapy, Taiwan National Health Insurance Research
Database
Original Submission: 2 November 2010; Revised Submission: 22 December 2010; Accepted: 13 January 2011
Editor: M. Drancourt
Article published online: 24 January 2011
Clin Microbiol Infect 2011; 17: 1646–1652
10.1111/j.1469-0691.2011.03473.x
Corresponding author: T.-W. Chen, Nephrology, Department of
Internal Medicine, Taipei Medical University Hospital, Taipei, Taiwan
and
Department of Medicine, School of Medicine, College of Medicine,
Taipei Medical University, No. 252, Wu Hsing St, Sinyi District,
Taipei 110, Taiwan
E-mail: twchen@tmu.edu.tw
Introduction
Mycobacterium tuberculosis (MTB) infection is the deadliest
infectious disease worldwide, and is responsible for 1.7 million deaths per year [1]. Although antimicrobial treatment
for MTB infection was developed more than 50 years ago,
the numbers of MTB cases and deaths continue to increase.
Risk factors for MTB infection are often considered to result
from impaired host immunity or increased exposure to infection. Age [2], male gender [3], malnutrition [4,5], human
immunodeficiency virus (HIV) infection [6], silicosis [7,8], diabetes mellitus (DM) [9,10], glucocorticoids [11], receipt of
transplants [12–14], malignancy [15], low socio-economic
status, substance abuse [16] and renal insufficiency [17,18]
are known risk factors for MTB infection. Previous studies
indicated that dialysis patients have a 6.9-fold to 52.5-fold
higher risk of MTB infection than the general population
[17]. Most of these previous studies were cross-sectional
single-centre studies that considered only small numbers of
patients, and this is presumably the reason for the wide
range in estimated relative risk (RR).
Patients with end-stage renal disease (ESRD) are treated
with renal replacement therapy (RRT): haemodialysis (HD),
peritoneal dialysis (PD), or kidney transplantation. Each of
these therapies has the potential to increase the probability
of MTB infection. In particular, long-term use of immunosuppressant agents by transplant recipients may increase the risk
of infection, long-term attendance at hospital three times a
week may increase the MTB transmission rate among HD
ª2011 The Authors
Clinical Microbiology and Infection ª2011 European Society of Clinical Microbiology and Infectious Diseases
Li et al.
CMI
patients, and peritoneal infection may occur during PD.
However, no previous studies have rigorously examined the
infection rates of patients given different RRT modalities.
Knowledge of the risk of MTB infection following different
RRT modalities is clearly very important, especially in MTBepidemic areas. Taiwan has the highest ESRD incidence
worldwide, and is also an MTB-epidemic area.
In the current study, we performed a longitudinal cohort
study in Taiwan to evaluate the MTB infection rate among
ESRD patients. We used the National Health Insurance
Research Database (NHIRD) [19] to evaluate the RR of
MTB infection in ESRD patients, and compared the hazard
ratios for different RRT modalities.
MTB infection among ESRD patients
1647
health insurance dataset of 1 000 000 beneficiaries (randomly
sampled from the original NHIRD) was used as a control group.
The NHIRD published a dataset entitled the Longitudinal Health
Insurance Database (LHID) in 2005. The LHID-2005 contains
the entire original claim data of 1 000 000 beneficiaries who
enrolled in 2005, randomly sampled from the 2005 Registry for
Beneficiaries (ID) of the NHIRD. Registration data of all
beneficiaries of the National Health Insurance programme from
1 January to 31 December 2005 were drawn for random
sampling. There were no significant differences in the age and
gender distribution between patients in the LHID-2005 and
those in the original NHIRD [19].
Definition of RTTs, MTB infection and comorbidity
Methods
Data source
This study used data from the NHIRD of Taiwan. In 1995,
universal health insurance was initiated in Taiwan, and this
covers 99% of the population of 23 million people [20]. In
1999, the Bureau of National Health Insurance began to
release all patient data in electronic form under the NHIRD
project [19]. Various extracted datasets are available to
researchers, and hundreds of published papers have used the
NHIRD as the basis for their studies. The NHIRD consists
of de-identified secondary data released to the public for
research purposes, so the present study was exempt from
full review by the Internal Review Board. In this study, we
excerpted data from a specially ordered dataset that includes
all claim information from patients under the registry of catastrophic illness from January 1996 to December 2006. In Taiwan, patients with ESRD who need long-term RRT can apply
for catastrophic illness registration cards from the Bureau of
National Health Insurance, so that co-payments are not
required for treatment.
Study design
In the current cohort study, all incident ESRD patients in
Taiwan from 1 January 1998 to 31 December 2004 were
identified from the catastrophic illness dataset. Age, gender,
RRT modality, comorbid diseases and the socio-economic
status of patients were retrieved from the database. Each
patient was followed from the initiation of RRT to the diagnosis of MTB infection, death, withdrawal from the original
RRT, or 31 December 2006. For patients who switched dialysis modality or had kidney transplantation during the study
period, the two RRT periods were analysed separately.
For comparison of the incidence of MTB infection of our
ESRD patients and that of the general population, a longitudinal
PD patients are those with catastrophic illness registration
cards for ESRD and insurance claims for PD dialysate. HD
patients are those with catastrophic illness registration cards
for ESRD and insurance claims for HD treatment for more
than 3 months. Kidney transplant recipients are those with
catastrophic illness registration cards for kidney transplantation (V42.0) and insurance claims for immunosuppressant
agents for more than 3 months. In Taiwan, MTB infection is
a legal infectious disease, and the anti-MTB drugs are controlled by a central government health institute. When a
doctor diagnoses an incident MTB patient and prescribes
anti-tuberculosis (TB) drugs, he or she needs to declare the
case to the Taiwan CDC within 1 week. Because it takes
2 months to obtain the TB culture report in clinical practice,
we defined an MTB-positive patient as one who had an ICD9 code (010–018) in at least three ambulatory visits and
insurance claims for anti-TB drugs (isoniazid, rifampicin, rifamycin, rifabutin, pyrazinamide, or ethambutol) for more than
90 days. To validate this definition, we use the same criteria
in a randomized sample of 1 million, and compared the MTB
infection incidence rate with the Taiwan CDC annual report.
Hypertension, diabetes and hyperlipidaemia were defined by
ICD-9 codes and insurance claims for medication for more
than 90 days per year. HIV infection (ICD-9 code 042), silicosis (ICD-9 codes 501–504), chronic obstructive pulmonary
disease (COPD) (ICD-9 codes 403, 416, 491–493, 495–495,
508, 515, 516, and 518), connective tissue disease (ICD-9
codes 710, 712, 714, 715, 716, 719, and 728) and malignancy
(ICD-9 codes 140–239) were defined by disease codes in at
least three ambulatory visits. The monthly salary of beneficiaries was used to indicate economic status.
Statistical analysis
Normally distributed continuous data are expressed as
means ± standard deviations. Numerical data with non-normal
distributions are expressed as medians and interquartile
ª2011 The Authors
Clinical Microbiology and Infection ª2011 European Society of Clinical Microbiology and Infectious Diseases, CMI, 17, 1646–1652
1648
Clinical Microbiology and Infection, Volume 17 Number 11, November 2011
ranges. For comparison of parameters between different dialysis modalities, the chi-square test was used for categorical
variables, and ANOVA and the Mann–Whitney U-test were
used for parametric and non-parametric continuous variables,
respectively. The incidence of MTB infection was calculated
by the Kaplan–Meier method, and compared by use of the
log-rank test. Multivariate Cox proportional hazard regression was performed with backward elimination to analyse
independent risk factors for MTB infection. Risk factors with
p-values <0.1 in univariate Cox analysis were entered into
the multivariate analysis. All probabilities were two-tailed. A
p-value <0.05 was considered to be significant.
Results
Basic characteristics of the study population
Fig. 1 illustrates the patient selection flow. During the 7-year
enrolment period, there were 51 443 new cases of ESRD.
After exclusion of patients with coding errors, those who
received RRT for <30 days, those who were <20 years in
age, those with a history of MTB infection before dialysis and
those with age coding errors (seven subjects’ birthday are
coded as 1800/1/1, or age >120 years), we used the records
of 49 983 patients (186 793 patient-year follow-up period)
for analysis.
Table 1 shows the distribution of gender, comorbid diseases, RRT modality and other characteristics. During the 7year study period, there were slight increases in the annual
number of new ESRD cases, mean patient age, and cases of
DM, hypertension, and hyperlipidaemia. Table 2 shows the
baseline characteristics of patients stratified by RRT modality.
Our cohort had 43 495 HD patients (87.0%), 4139 PD
patients (8.3%), and 2249 kidney transplantion patients
(4.7%). HD patients were the oldest, followed by PD
patients, and then kidney transplantion patients (ANOVA,
p <0.001). The PD group had the highest percentage of
females (p <0.001). The transplantation group had the lowest
rate of DM and the highest economic status (p <0.001).
Comparison of MTB infection in the general population and
ESRD patients
With use of the same criteria for diagnosis, the MTB infection incidence rate in LHID-2005 was 0.65/1000 patientyears, which is comparable to the 0.67/1000 patient-years
that was reported by the Taiwan Centres for Disease [21].
In our ESRD cohort, 562 patients had MTB infections during
the study period, corresponding to an incidence rate of 3.0/
1000 patient-years. Thus, the RR of MTB infection for our
ESRD patients is 4.5.
CMI
From 1998 to 2004
51 443 incident ESRD
patients
373 patients were
under age 20 years
51 070 patients
624 patients had MTB
infection before RRT
50 446 patients
56 patients had MTB infection
within 30 days after RRT
50 390 patients
400 patients had undergone
RRT for less than 30 days
49 990 patients
7 patients with coding error
49 983 patients were analysed
FIG. 1. Patient selection flow chart. ESRD, end-stage renal disease;
MTB, Mycobacterium tuberculosis; RRT, renal replacement therapy.
Independent predictors of MTB infection in ESRD patients
During the study period, we identified MTB infection in 562
patients (499 HD patients, 44 PD patients, and 19 kidney
transplantion patients). Univariate Cox regression analysis
indicated that nine risk factors were significantly associated
with MTB infection rate (Table 3). In particular, old age, male
gender, low socio-economic status, COPD, silicosis and
development of ESRD before the year 2001 were associated
with an increased MTB infection rate. In contrast, hypertension, hyperlipidaemia and kidney transplantation (Fig. 2) were
associated with a lower rate of MTB infection. The Kaplan–
Meier plots for each risk factor for MTB infection are
illustrated in Fig. 2.
Multivariate Cox regression analysis indicated that the
following variables were independent risk factors for
MTB infection: old age, male gender, COPD, silicosis, and
ª2011 The Authors
Clinical Microbiology and Infection ª2011 European Society of Clinical Microbiology and Infectious Diseases, CMI, 17, 1646–1652
Li et al.
CMI
MTB infection among ESRD patients
1649
TABLE 1. Patient characteristics stratified by calendar year
Characteristics
1998
1999
2000
2001
2002
2003
2004
Overall
New ESRD case, no.
Male gender (%)
Mean age (years)
Diabetes (%)
Hypertension (%)
Hyperlipidaemia (%)
Haemodialysis (%)
Patients in cohort, no.
6442
47.1
57.7
40.9
79.8
30.1
86.1
6442
6326
46.6
58.9
40.3
78.8
34.0
87.3
12 560
6905
46.3
59.4
42.5
82.8
37.8
87.3
18 371
7183
46.7
59.3
43.2
85.2
39.8
86.3
23 751
7594
47.4
60.2
44.9
87.1
42.0
87.3
29 012
7691
47.5
60.6
42.6
88.1
45.3
87.1
33 894
7842
48.1
61.4
44.0
88.6
47.2
87.5
38 461
49 983
47.1
59.7
43.8
83.6
39.9
87.0
ESRD, end-stage renal disease.
TABLE 2. Baseline patient characteristics stratified by renal
Discussion
replacement therapy (RRT) modality
RRT modality
Demographic data
Age (years), mean ± SD
Male gender, % (no.)
Monthly Ssalary ($US),
mean ± SD
Follow-up days (IQR)
Comorbid disease, % (no.)
Diabetes
Hypertension
Hyperlipidaemia
COPD
Silicosis
Malignancy
Connective tissue
disease
Medication, % (no.)
Systemic steroids >12
months
HD
N = 43 495
PD
N = 4139
61.2 ± 13.5
53.7 ± 15.1
47.7 (20 731) 39.2 (1624)
339 ± 429
458 ± 592
Transplant
N = 2349
p
43.9 ± 13.1
51.5 (1200)
765 ± 721
<0.001
<0.001
<0.001
1190
(702–1284)
1230
2031
<0.001
(770–1868) (1453–2648)
45.6 (19 849)
82.9 (36 046)
37.4 (16 272)
3.1 (1336)
0.11 (50)
4.7 (2033)
1.3 (583)
33.9 (1405)
88.4 (3659)
58.5 (2422)
2.9 (120)
0.02 (1)
3.2 (133)
3.5 (145)
27.7 (651)
89.1 (2092)
52.8 (1241)
0.8 (19)
0.04 (1)
1.7 (40)
3.4 (79)
<0.001
<0.001
<0.001
<0.001
0.143
<0.001
<0.001
8.0 (3499)
12.1 (502)
87.4 (2052)
<0.001
COPD, chronic obstructive pulmonary disease; HD, haemodialysis; IQR, interquartile range; PD, peritoneal dialysis; SD, standard deviation.
TABLE 3. Cox regression analysis for Mycobacterium tuberculosis infection risk factors in end-stage renal disease
(ESRD) patients
Univariate
Multivariate
Parameters
HR
Male gender
Age (per 10 years)
ESRD after year 2001
RRT modality
Haemodialysis (as
reference)
Peritoneal dialysis
Kidney transplantation
Diabetes
Hypertension
Hyperlipidaemia
Silicosis
Steroid use
COPD
Malignancy
Connective tissue disease
Monthly salary >$US800
1.39 <0.001 1.18–1.65 1.37 <0.001 1.16–1.62
1.17 <0.001 1.10–1.24 1.17 <0.001 1.10–1.24
0.52 <0.001 0.44–0.61 0.53 <0.001 0.45–0.63
0.033
1.00
1.000 1.00–1.00
p
95% CI
HR
p
95% CI
0.90
0.502 0.66–1.22
0.55
0.011 0.35–0.87
0.87
0.107 0.73–1.03
0.71
0.001 0.58–0.87 0.81
0.050 0.66–1.00
0.62 <0.001 0.52–0.75 0.71 <0.001 0.59–0.86
7.32 <0.001 2.74–19.5 5.82 <0.001 2.17–15.6
1.14
0.264 0.90–1.44
1.76
0.006 1.18–2.63 1.68
0.012 1.12–2.51
1.18
0.224 0.91–1.53
1.29
0.357 0.75–2.24
0.71
0.032 0.52–0.97
COPD, chronic obstructive pulmonary disease; HR, hazard ratio; RRT, renal
replacement therapy.
development of ESRD before the year 2001. Hyperlipidaemia
and hypertension were significant protective factors against
MTB infection.
Over the past decade, the Taiwan government has initiated
numerous programmes for prevention of MTB infection,
including the three-phase 5-year plan, the Directly Observed
Treatment Short-course (DOTS) plan, the DOTS-Plus, and
the Mobilization Plan for Halving TB Incidence in 10 Years
[21]. In fact, the incidence rate of MTB infection in Taiwan
has gradually declined over the past decade [22], but MTB
infection is still the most lethal, notifiable infectious disease.
Among all notifiable infectious diseases, MTB infection has
the greatest number of confirmed cases per year, with more
than 15 000 new infections per year. ESRD patients are vulnerable to MTB infection, but previous studies have reported
very different RR rates [23–26], mainly because of the small
sizes of these studies. An exception is the cross-sectional
study conducted by Chou et al. [27], which examined Taiwan
CDC MTB registration data and indicated an annual MTB
infection rate of 24 533 Taiwanese HD patients of 0.5 per
1000 patients in 2001. The other studies only examined 7–
38 MTB infection cases from single centres [17]. Thus, the
present study is valuable because of the large number of
cases and the long follow-up period. The NHIRD contains all
medical claim data for Taiwan, allowing a nationwide longitudinal study.
In agreement with previous studies of general populations,
we identified old age, male gender, silicosis and COPD as
risk factors for MTB infection in ESRD patients. Malnutrition
can lead to immune system dysregulation, and is a wellknown risk factor for MTB infection [4]. In the current
study, we were unable to measure nutritional parameters,
such as body mass index and serum albumin, but we found
that hyperlipidaemia and hypertension were protective factors in ESRD patients. Hypertension is common in ESRD
patients, unless the patient has heart failure or insufficient
effective blood volume. In fact, low blood pressure and low
cholesterol are often associated with unfavourable outcomes
in ESRD patients [28,29]. Thus, we believe that hyperlipidaemia and hypertension are indicators of nutritional status in
ESRD patients. Diabetes increases the risk of MTB infection
ª2011 The Authors
Clinical Microbiology and Infection ª2011 European Society of Clinical Microbiology and Infectious Diseases, CMI, 17, 1646–1652
1650
Clinical Microbiology and Infection, Volume 17 Number 11, November 2011
ª2011 The Authors
Clinical Microbiology and Infection ª2011 European Society of Clinical Microbiology and Infectious Diseases, CMI, 17, 1646–1652
CMI
Li et al.
CMI
MTB infection among ESRD patients
1651
FIG. 2. Kaplan–Meier plots of the risk factors for Mycobacterium tuberculosis (MTB) infection among end-stage renal disease (ESRD) patients.
Kaplan–Meier plots of cumulative MTB infection rate among ESRD patients, stratified by gender (a), age (b), hyperlipidaemia (c), hypertension
(d), chronic obstructive pulmonary disease (COPD) (e), silicosis (f), monthly salary (g), and renal replacement modality (h). HD, haemodialysis;
PD, peritoneal dialysis.
in the general population; however, the odds ratio is modest.
A large-scale study that compared 500 000 diabetes patients
and 500 000 non-diabetes subjects indicated that the odds
ratio of MTB infection was 1.12 [30]. In our analysis, diabetes
was not a risk factor for MTB infection in ESRD patients. It
is possible that ESRD was such a strong risk factor for MTB
infection that it diminished the impact of other risk factors
with modest odds ratios.
For patients with ESRD, kidney transplantation is considered the best RRT modality in terms of survival and life quality.
In the current study, we found that kidney transplantation was
associated with a lower MTB infection rate than PD or HD
(Fig. 2), but the transplantation group was also 20 years younger and had better socio-economic status. After multivariate
adjustment, we found that kidney transplantation was no
longer associated with a lower MTB infection rate.
Since HIV was discovered in the 1980s, the MTB infection
rate has increased worldwide, especially in Africa [31].
According to the WHO annual report of 2009, 15% of MTB
patients were HIV-positive, one-third of MTB-related deaths
were related to concomitant HIV infection, and MTB contributed to 26% of the estimated deaths resulting from HIV
infection [1]. In contrast to Western and African countries,
the HIV infection rate among MTB patients in Taiwan was
only 0.6% in 2007 [22]. In the current study, we only identified eight patients with HIV infection (all in the HD group),
but none of them developed MTB infection during the observation period. Thus, HIV infection was not considered in our
data analysis.
The current study had several limitations. First, because of
the retrospective nature of the study, we did not use the
standard procedure for diagnosis of MTB infection (acid-fast
stain and culturing). Instead, we diagnosed MTB infection if a
patient had taken anti-MTB medication for more than
3 months. It is likely that some patients with MTB infections
did not submit claims for medication, or died before
3 months of treatment. Second, it has been reported that
half of the MTB infections of ESRD patients are extrapulmonary, especially among PD patients [32,33]. We were unable
to study this issue, because these data were not available in
the NHIRD. Third, some possibly important risk factors,
such as smoking, nutritional status, intravenous drug abuse,
and occupational exposure, are also not available in the
NHIRD.
In conclusion, the results of this nationwide longitudinal
study indicate that the incidence rate of MTB infection is 3.0/
1000 patient-years for Taiwanese ESRD patients. As compared with the general population, the RR of MTB infection
for ESRD patients is 4.5. The independent risk factors for
MTB infection in ESRD patients are older age, male gender,
malnutrition, silicosis, and COPD. RRT modality had no
affect on the MTB infection rate.
Acknowledgements
This study is based in part on data from the NHRID provided by the Bureau of National Health Insurance, Department of Health, and managed by national health research
institutes. The interpretation and conclusions contained
herein do not represent those of Bureau of National Health
Insurance, Department of Health or national health research
institutes. This study is supported, in part, VGH research
grant 10ODMA0100220
Transparency Declaration
The authors declare no conflict of interest.
References
1. Global tuberculosis control: epidemiology, strategy, financing. World
Health Organization report 2009. 2009. Available at: http://www.who.
int/tb/publications/global_report/2009/pdf/full_report.pdf.
2. Donald PR, Marais BJ, Barry CE 3rd. Age and the epidemiology and
pathogenesis of tuberculosis. Lancet 2010; 375: 1852–1854.
3. Allotey P, Gyapong M. Gender in tuberculosis research. Int J Tuberc
Lung Dis 2008; 12: 831–836.
4. Cegielski JP, McMurray DN. The relationship between malnutrition
and tuberculosis: evidence from studies in humans and experimental
animals. Int J Tuberc Lung Dis 2004; 8: 286–298.
5. Leung CC, Lam TH, Chan WM et al. Lower risk of tuberculosis in
obesity. Arch Intern Med 2007; 167: 1297–1304.
6. Antonucci G, Girardi E, Raviglione MC, Ippolito G. Risk factors for
tuberculosis in HIV-infected persons. A prospective cohort study.
The Gruppo Italiano di Studio Tubercolosi e Aids (GISTA). JAMA
1995; 274: 143–148.
7. Cowie RL. The epidemiology of tuberculosis in gold miners with silicosis. Am J Respir Crit Care Med 1994; 150: 1460–1462.
ª2011 The Authors
Clinical Microbiology and Infection ª2011 European Society of Clinical Microbiology and Infectious Diseases, CMI, 17, 1646–1652
1652
Clinical Microbiology and Infection, Volume 17 Number 11, November 2011
8. Corbett EL, Churchyard GJ, Clayton TC et al. HIV infection and
silicosis: the impact of two potent risk factors on the incidence of
mycobacterial disease in South African miners. AIDS 2000; 14: 2759–
2768.
9. Dooley KE, Chaisson RE. Tuberculosis and diabetes mellitus: convergence of two epidemics. Lancet Infect Dis 2009; 9: 737–746.
10. Restrepo BI, Fisher-Hoch SP, Pino PA et al. Tuberculosis in poorly
controlled type 2 diabetes: altered cytokine expression in peripheral
white blood cells. Clin Infect Dis 2008; 47: 634–641.
11. Jick SS, Lieberman ES, Rahman MU, Choi HK. Glucocorticoid use,
other associated factors, and the risk of tuberculosis. Arthritis Rheum
2006; 55: 19–26.
12. Munoz P, Palomo J, Munoz R, Rodriguez-Creixems M, Pelaez T, Bouza
E. Tuberculosis in heart transplant recipients. Clin Infect Dis 1995; 21:
398–402.
13. Meyers BR, Halpern M, Sheiner P, Mendelson MH, Neibart E, Miller
C. Tuberculosis in liver transplant patients. Transplantation 1994; 58:
301–306.
14. Akan H, Arslan O, Akan OA. Tuberculosis in stem cell transplant
patients. J Hosp Infect 2006; 62: 421–426.
15. Kamboj M, Sepkowitz KA. The risk of tuberculosis in patients with
cancer. Clin Infect Dis 2006; 42: 1592–1595.
16. Oeltmann JE, Kammerer JS, Pevzner ES, Moonan PK. Tuberculosis
and substance abuse in the United States, 1997–2006. Arch Intern
Med 2009; 169: 189–197.
17. Hussein MM, Mooij JM, Roujouleh H. Tuberculosis and chronic renal
disease. Semin Dial 2003; 16: 38–44.
18. Pien FD, Younoszai BG, Pien BC. Mycobacterial infections in
patients with chronic renal disease. Infect Dis Clin North Am 2001; 15:
851–876.
19. National Health Insurance Research Database, Taiwan. Available at:
http://www.nhri.org.tw/nhird/en/index.htm.
20. Cheng TM. Taiwan’s new national health insurance program: genesis
and experience so far. Health Aff (Millwood) 2003; 22: 61–76.
CMI
21. Mobilization Plan for Having TB Incidence in Ten Years. Taiwan CDC
Annual Report 2008. 2008. Available at: http://www.cdc.gov.tw/public/
data/89116495971.pdf.
22. Center for Disease Control. Taiwan tuberculosis control report 2009.
ROC (Taiwan): Center for Disease Control, Department of Health,
2009.
23. Ram R, Swarnalatha G, Prasad N, Dakshinamurty KV. Tuberculosis in
renal transplant recipients. Transpl Infect Dis 2007; 9: 97–101.
24. Sen N, Turunc T, Karatasli M, Sezer S, Demiroglu YZ, Oner Eyuboglu F.
Tuberculosis in patients with end-stage renal disease undergoing dialysis
in an endemic region of Turkey. Transplant Proc 2008; 40: 81–84.
25. Kursat S, Ozgur B. Increased incidence of tuberculosis in chronic
hemodialysis patients. Am J Nephrol 2001; 21: 490–493.
26. Lui SL, Tang S, Li FK et al. Tuberculosis infection in Chinese patients
undergoing continuous ambulatory peritoneal dialysis. Am J Kidney Dis
2001; 38: 1055–1060.
27. Chou KJ, Fang HC, Bai KJ, Hwang SJ, Yang WC, Chung HM. Tuberculosis in maintenance dialysis patients. Nephron 2001; 88: 138–143.
28. Port FK, Hulbert-Shearon TE, Wolfe RA et al. Predialysis blood pressure and mortality risk in a national sample of maintenance hemodialysis patients. Am J Kidney Dis 1999; 33: 507–517.
29. Iseki K, Yamazato M, Tozawa M, Takishita S. Hypocholesterolemia is
a significant predictor of death in a cohort of chronic hemodialysis
patients. Kidney Int 2002; 61: 1887–1893.
30. Shah BR, Hux JE. Quantifying the risk of infectious diseases for
people with diabetes. Diabetes Care 2003; 26: 510–513.
31. Harries AD, Zachariah R, Corbett EL et al. The HIV-associated tuberculosis epidemic—when will we act? Lancet 2010; 375: 1906–1919.
32. Quantrill SJ, Woodhead MA, Bell CE, Hutchison AJ, Gokal R.
Peritoneal tuberculosis in patients receiving continuous ambulatory
peritoneal dialysis. Nephrol Dial Transplant 2001; 16: 1024–1027.
33. Chau TN, Leung VK, Wong S et al. Diagnostic challenges of tuberculosis peritonitis in patients with and without end-stage renal failure.
Clin Infect Dis 2007; 45: e141–e146.
ª2011 The Authors
Clinical Microbiology and Infection ª2011 European Society of Clinical Microbiology and Infectious Diseases, CMI, 17, 1646–1652