INT J TUBERC LUNG DIS 18(10):1245–1251
Q 2014 The Union
http://dx.doi.org/10.5588/ijtld.14.0155
Monitoring liver function among patients who initiated antituberculosis drugs in Taiwan, 2000–2011
H-C. Chou,* S-W. Lin,†‡§ W-W. Chen,* W-M. Ke,* P-H. Chao,* F-Y. Hsiao†‡§
*Taiwan Drug Relief Foundation, Taipei, †Graduate Institute of Clinical Pharmacy, and ‡School of Pharmacy,
College of Medicine, National Taiwan University, Taipei, §Department of Pharmacy, National Taiwan University
Hospital, Taipei, Taiwan
SUMMARY
O B J E C T I V E : To investigate adherence to liver function
monitoring as recommended in Taiwan’s tuberculosis
(TB) diagnosis and treatment guidelines for newly
diagnosed TB patients.
D E S I G N : Retrospective cohort study of the National
Health Insurance Research Database (NHIRD), Taiwan,
2000–2011.
M E T H O D S : From the NHIRD, we identified 11 397
newly diagnosed TB patients who initiated anti-tuberculosis treatment between 2000 and 2011 and categorised these into three groups: completely, partially and
non-adherent. Logistic regression was used to explore
potential factors associated with the adherence rate.
R E S U LT S : The completely adherent rate increased from
0.5% in 2000 to 9.2% in 2011, while the non-adherent
rate decreased from 17.5% to 1.2%. Compared to the
non-adherent group, patients with a history of liver
disease (OR 4.36, 95%CI 1.92–9.87) and viral hepatitis
(OR 9.39, 95%CI 1.47–60.19), as well as patients
whose prescribing physicians were specialists in chest
(OR 4.59, 95%CI 1.91–11.05), TB (OR 2.55, 95%CI
1.01–6.40) and infectious diseases (OR 3.93, 95%CI
1.08–14.31), had higher odds of being completely
adherent to the guidelines.
C O N C L U S I O N : Our findings could serve as an important reference for developing effective strategies to
improve adherence to guidelines and prevent patients
from developing anti-tuberculosis drug-associated hepatotoxicity.
K E Y W O R D S : adherence; tuberculosis; liver function
tests; hepatotoxicity
TUBERCULOSIS (TB) is a major global health
problem. According to the 2013 Global Tuberculosis
Report released by the World Health Organization
(WHO), it is estimated that there were 8.6 million
incident TB cases worldwide, equivalent to 122 cases
per 100 000 population, in 2012. Although the
mortality rate had fallen by 45% since 1990, an
estimated 1.3 million TB deaths were reported.1
Before 1985, TB was a major threat to public
health in Taiwan, and a principal cause of death in the
country for several decades.2 However, continued
efforts to control TB have reduced the prevalence of
and deaths due to TB in Taiwan. Since its launch in
2006, directly observed treatment (DOT) has been
used to reduce all-cause mortality among TB patients3 and TB-specific mortality by 55%.4
In addition to TB disease, hepatotoxicity caused by
anti-tuberculosis drugs cannot be ignored. Antituberculosis drugs are one of the most common
causes of drug-induced liver injury.5,6 The incidence
of impaired liver function tests (LFTs) varies from
10% to 25%, while symptomatic liver disease occurs
in 0.5–3% of patients.7 Several first-line anti-tuberculosis drugs, including isoniazid (INH), rifampicin
(RMP) and pyrazinamide (PZA), were reported to be
associated with potential hepatotoxicity.8 Anti-tuberculosis drug-induced hepatotoxicity (ATDH) may
also lead to unsuccessful anti-tuberculosis treatment
and prolongation of the intensive phase of treatment.
If ATDH is not recognised in time, hepatotoxicity
could be fatal.9
Although the clinical symptoms and signs of
hepatotoxicity are not specific enough to detect liver
injury and initial biochemical abnormalities are
predominant in the development of liver injury
induced by INH and PZA,10 the importance of
routine LFT monitoring during anti-tuberculosis
treatment is not highlighted in most international
guidelines, or it is limited to patients with pre-existing
liver disease.2,11,12 Routine LFT monitoring was
reported to be helpful in preventing severe hepatotoxicity and death;13,14 it was also effective in
Correspondence to: Fei-Yuan Hsiao, Graduate Institute of Clinical Pharmacy, College of Medicine, National Taiwan
University, Room 220, 2F, No. 33 Lin-Sen S. Road, Taipei 10050, Taiwan. Tel: (þ886) 2 3366 8787. Fax: (þ886) 2 2395
1113. e-mail: [email protected]
Article submitted 24 February 2014. Final version accepted 1 June 2014.
1246
The International Journal of Tuberculosis and Lung Disease
identifying hepatotoxicity and improving adherence
to anti-tuberculosis treatment.15,16 More effort
should thus be devoted to promoting routine LFT
monitoring during anti-tuberculosis treatment.
In Taiwan’s guidelines for TB diagnosis and
treatment, issued by the Center for Disease Control
of Taiwan (Taiwan CDC), LFT monitoring before
initiating anti-tuberculosis treatment and at weeks 2,
4 and 8 has been recommended.17 However, the rate
of LFT monitoring in Taiwan has never been
evaluated.
The present study aimed to investigate adherence
to LFT monitoring among newly diagnosed TB
patients initiating anti-tuberculosis treatment. Potential factors associated with the adherence rate were
also examined.
MATERIALS AND METHODS
Data source
This retrospective, population-based study used
Taiwan’s 2000–2011 National Health Insurance
Research Database (NHIRD) as its data source. The
NHIRD is a nationwide database comprising anonymous eligibility and enrolment information, as well
as claims for visits, procedures and prescription
medications for more than 99% of the entire
population (23 million) of Taiwan.
Two subsets of the NHIRD, the Longitudinal
Health Insurance Database (LHID) 2000 and 2005,
were also used as data sources. These two data sets
contain claims data on one million beneficiaries
randomly sampled from the Registry of Beneficiaries
of the NHIRD in years 2000 and 2005. A total of two
million individuals, accounting for 10% of the total
population in Taiwan, thus served as our original
cohort. Age and sex distributions were not significantly different between patients in LHID 2000,
LHID 2005 (2000, v2 ¼ 1.74, degrees of freedom [df]
¼ 1, P ¼ 0.187; 2005, v2 ¼ 0.008, df ¼ 1, P ¼ 0.931)
and the original NHIRD. The details of the NHIRD
are described on the NHIRD website and in our
previous publication.18
Identification numbers of all subjects in the
NHIRD were encrypted, as were all traceable
personal identifiers, to protect the privacy of the
individuals. The study was therefore exempt from full
review by the Institutional Review Board of the
National Taiwan University Hospital, Taipei, Taiwan, and the requirement for providing informed
consent was waived.
Study population
From LHID 2000 and 2005, we identified 38 529
patients diagnosed with TB (International Classification of Diseases-9-CM code: 010.x–018.x or Acode:
A020–A021) during in- or out-patient visits between
1998 and 2011. A washout period of 1998 and 1999
was generated to select new TB patients; this means
that patients whose first TB diagnosis was between
2000 and 2011 were defined as newly diagnosed TB.
Among these newly diagnosed TB patients, we
further identified those who initiated anti-tuberculosis treatment within 30 days after the day of TB
diagnosis as our study subjects. Anti-tuberculosis
treatment was defined as the use of two or more antituberculosis drugs for 728 cumulative days. The date
of first TB prescription was defined as the cohort
entry date (index date). Patients of unknown sex or
those with ,63 days of follow-up were excluded.
Measures
An LFT was defined as any test used to measure levels
of aspartate aminotransferase (AST), alanine aminotransferase (ALT) or total bilirubin. In accordance
with the Taiwan Guidelines,17 LFTs within 90 days of
the index date (baseline), as well as between days 7
and 21 (week 2), days 22 and 35 (week 4) and days 50
and 63 (week 8) following the index date in in- and
out-patient settings, were retrieved for each study
subject. Our study subjects were further categorised
into 1) completely adherent, 2) partially adherent,
and 3) non-adherent based on their physicians’
adherence to LFT monitoring, as suggested in the
guidelines. Patients in the completely adherent group
were defined as those whose prescribing physician
had issued baseline LFT and LFT in weeks 2, 4 and 8.
Patients in the partially adherent group were those
whose prescribing physician had issued LFTs in any
one, but not at all time points recommended by the
guidelines. Patients in the non-adherent group were
those with no record of LFT during anti-tuberculosis
treatment. We further categorised all study subjects
into two groups based on whether or not they had
received a baseline LFT, and examined the LFT
monitoring adherence rates between the two groups
Statistical analysis
The annual prevalence of adherence rates of LFT
monitoring suggested by the guidelines was reported
for each fiscal year from 2000 to 2011. We retrieved
each patient’s baseline characteristics, medical condition (liver diseases, viral hepatitis, alcohol-related
diseases) and use of medications (acetaminophen,
non-steroidal anti-inflammatory drugs [NSAIDs],
thiazolidinediones, anti-epileptic agents, anti-fungal
agents, antibiotics, 3-hydroxy-3-methyl-glutarylCoA reductase [HMG-CoA], reductase enzyme inhibitors and antiviral agents) potentially associated
with hepatotoxicity in the year before the initiation of
anti-tuberculosis treatment. Characteristics of health
providers, including the medical department and
medical facilities visited by the patient, were also
examined. Medical facilities were characterised by
accreditation level (medical centre, metropolitan
hospital, local community hospital, clinic) and
LFT monitoring in Taiwan, 2000–2011
Table 1
1247
Baseline characteristics of newly diagnosed TB patients who initiated anti-tuberculosis treatment
Adherent to the guidelines
Completely
(n ¼ 486)
n (%)
Sex, male
Age, years, mean 6 SD
0–17
18–64
765
Medical department visited
Chest diseases
Thoracic surgery
TB
Infectious diseases
Family medicine
Other department of general medicine
Other department of surgery
Others
Partially
(n ¼ 10 049)
n (%)
Non-adherent
(n ¼ 862)
n (%)
P value
318 (65.4)
6766 (67.3)
548 (63.6)
0.06
58.91 6 18.79
5 (1.0)
274 (56.4)
207 (42.6)
57.71 6 20.02
220 (2.2)
5454 (54.3)
4375 (43.5)
49.24 6 21.69
59 (6.8)
539 (62.5)
264 (30.6)
,0.0001
,0.0001
294
3
97
24
3
50
4
11
(60.5)
(0.6)
(20.0)
(4.9)
(0.6)
(10.3)
(0.8)
(2.3)
4674
150
1928
509
159
1838
137
654
(46.5)
(1.5)
(19.2)
(5.1)
(1.6)
(18.3)
(1.4)
(6.5)
254
7
186
14
65
202
12
122
(29.5)
(0.8)
(21.6)
(1.6)
(7.5)
(23.4)
(1.4)
(14.2)
Accreditation level
Medical centre
Metropolitan hospital
Local community hospital
Clinic
197
181
65
43
(40.5)
(37.2)
(13.4)
(8.8)
3697
3975
1828
549
(36.8)
(39.6)
(18.2)
(5.5)
254
209
222
177
(29.5)
(24.2)
(25.8)
(20.5)
Region
Taipei
Northern
Central
Southern
Kao-Ping
Eastern
218
36
70
83
76
3
(44.9)
(7.4)
(14.4)
(17.1)
(15.6)
(0.6)
2894
1228
1850
1564
2086
427
(28.8)
(12.2)
(18.4)
(15.6)
(20.8)
(4.2)
187
169
154
100
161
91
(21.7)
(19.6)
(17.9)
(11.6)
(18.7)
(10.6)
55
19
13
104
(11.3)
(3.9)
(2.7)
(21.4)
956
254
251
2750
(9.5)
(2.5)
(2.5)
(27.4)
24
2
7
65
(2.8)
(0.2)
(0.8)
(7.5)
,0.0001
,0.0001
,0.01
,0.0001
288
271
9
15
2
278
13
55
(59.3)
(55.8)
(1.9)
(3.1)
(0.4)
(57.2)
(2.7)
(11.3)
5382
5346
134
263
81
5242
312
740
(53.6)
(53.2)
(1.3)
(2.6)
(0.8)
(52.2)
(3.1)
(7.4)
347
343
4
7
7
324
7
61
(40.3)
(39.8)
(0.5)
(0.8)
(0.8)
(37.6)
(0.8)
(7.1)
,0.0001
,0.0001
0.05
,0.01
0.63
,0.0001
,0.01
,0.01
Medical facility
Medical condition in previous year
Liver disease
Viral hepatitis
Alcohol-related disease
Hospitalisation
Medication use in previous year
Acetaminophen
NSAIDs
Thiazolidinediones
Anti-epileptic agents
Anti-fungal agents
Antibiotics
HMG-CoA reductase enzyme inhibitors
Antiviral agents
,0.0001
,0.0001
TB ¼ tuberculosis; SD ¼ standard deviation; NSAID ¼ non-steroidal anti-inflammatory drug; HMG-CoA ¼ 3-hydroxy-3-methyl-glutaryl-CoA reductase.
regions. Prescribing physicians were grouped according to specialisaton (e.g., chest diseases).
Patient demographics and provider characteristics
were compared using the v2 test for categorical
variables or analysis of variance for continuous
variables. Logistic regression was used to explore
potential determinants associated with adherence
rate. P values were two-sided and a was set at 0.05.
All analyses were generated by SAS software, version
9.2 (Statistical Analysis System Institute Inc, Cary,
NC, USA).
RESULTS
We identified 11 397 newly diagnosed TB patients
who underwent anti-tuberculosis treatment between
2000 and 2011 (Table 1). Overall, the completely
adherent rate increased from 0.5% in 2000 to 9.2%
in 2011, while the non-adherent rate decreased from
17.5% to 1.2% (Figure 1).
Most (88.2%) of the newly diagnosed TB patients
were categorised as partially adherent based on their
physicians’ adherence to liver function monitoring as
recommended in the guidelines. The sex distribution
in the three groups was similar. Patients in the nonadherent group were 10 years younger than those
who had ever had LFTs. Approximately 60% of
patients in the completely adherent group, but only
30% in the non-adherent group, were treated by
physicians specialising in chest diseases. In contrast,
only 10% of patients in the completely adherent
group, but a quarter (23.4%) of patients in the non-
1248
The International Journal of Tuberculosis and Lung Disease
Figure 1
Trend in adherence rates for liver function monitoring, 2000–2011.
adherent group received their treatment from physicians specialising in general medicine. The distribution of medical facilities across the three groups was
significantly different. In the completely and partially
adherent groups, approximately 80% of patients
received anti-tuberculosis treatment from medical
centres and metropolitan hospitals compared to only
53.7% in the non-adherent group. Patients in the
non-adherent group mostly attended local community hospitals (25.8%) and clinics (20.5%) for antituberculosis treatment (Table 1).
Significant differences were observed in medical
conditions and the use of medications associated with
hepatotoxicity across the three groups. The proportion of patients with liver disease, viral hepatitis,
alcohol-related diseases or hospitalisation, as well as
patients who used acetaminophen, NSAIDs, antiepileptic agents, antibiotics, HMG-CoA, reductase
enzyme inhibitors and antiviral agents, was highest in
the completely adherent group and lowest in the nonadherent group.
Compared to the non-adherent group, patients
with a history of liver disease (odds ratio [OR] 4.36,
95% confidence interval [CI] 1.92–9.87), viral
hepatitis (OR 9.39, 95%CI 1.47–60.19) and hospitalisation (OR 5.24, 95%CI 3.13–8.80) had greater
odds of being completely adherent to guidelines. In
addition, patients whose prescribing physicians specialised in chest diseases (OR 4.59, 95%CI 1.91–
11.05), TB (OR 2.55, 95%CI 1.01–6.40) and
infectious diseases (OR 3.93, 95%CI 1.08–14.31)
had greater odds of being completely adherent to
guidelines. There were also regional differences in the
adherence to guidelines. Patients who lived outside
Taipei City had lower odds of being completely
adherent to guidelines (Table 2).
We further categorised all study subjects into two
groups based on whether or not they underwent
baseline LFT. We found that, after initiation of antituberculosis treatment, 32.2% of patients with
baseline LFT compared to 44.9% of patients without
baseline LFT did not undergo any subsequent LFT
(Figure 2).
DISCUSSION
In this population-based study on LFT monitoring
among newly diagnosed TB patients, a significant
increase in adherence to the CDC Taiwan guidelines
was observed in the 12 years from 2000 to 2011 in
Taiwan. The most striking change was the decrease in
the rate of non-adherence, from 17.5% to 1.2%,
indicating growing awareness of anti-tuberculosis
drug-related hepatotoxicity. However, the completely
adherent rate was suboptimal, increasing from 0.5%
in 2000 to 9.2% in 2011. Less than 10% of patients
underwent close LFT monitoring according to
guideline recommendations. Moreover, 44.9% of
patients with no baseline LFT received no subsequent
LFT during anti-tuberculosis treatment.
In addition to the overall trends in adherence rates,
our study examined both patient and provider
characteristics associated with greater odds of adherence to the guidelines. Patient characteristics, such as
increased age, female sex, concomitant medication
use and underlying disease states have been reported
LFT monitoring in Taiwan, 2000–2011
1249
Table 2 Analysis of potential determinants associated with adherence rate
Adherence to the guidelines
Completely vs. non-adherent
Sex (reference: male)
Age, years
Medical department visited (reference: Others)
Chest diseases
Thoracic surgery
Tuberculosis
Infectious diseases
Family medicine
Other department of general medicine
Other department of surgery
Medical facility
Partially vs. non-adherent
OR (95%CI)
P value
OR (95%CI)
P value
0.80 (0.55–1.15)
1.02 (1.01–1.03)
0.23
,0.0001
0.91 (0.78–1.07)
1.01 (1.01–1.02)
0.25
,0.0001
4.59
0.85
2.55
3.93
0.22
2.68
1.53
,0.01
0.88
0.05
,0.05
0.06
,0.05
0.66
1.86
1.57
1.40
2.65
0.54
1.61
1.20
(1.42–2.43)
(0.70–3.55)
(1.06–1.85)
(1.47–4.80)
(0.36–0.80)
(1.22–2.13)
(0.61–2.34)
,0.0001
0.28
,0.05
,0.01
,0.01
,0.01
0.60
1.93 (1.50–2.50)
2.30 (1.77–3.00)
1.47 (1.14–1.89)
,0.0001
,0.0001
,0.01
(1.91–11.05)
(0.10–6.99)
(1.01–6.40)
(1.08–14.31)
(0.05–1.03)
(1.03–6.95)
(0.24–9.77)
Level (reference: clinic)
Medical centre
Metropolitan hospital
Local community hospital
2.02 (0.99–4.13)
1.56 (0.76–3.20)
0.83 (0.39–1.74)
Region (reference: Taipei)
Northern
Central
Southern
Kao-Ping
Eastern
0.09
0.22
0.38
0.30
0.06
(0.05–0.16)
(0.13–0.38)
(0.21–0.66)
(0.17–0.51)
(0.02–0.26)
,0.0001
,0.0001
,0.01
,0.0001
,0.01
0.41
0.67
0.91
0.73
0.35
(0.32–0.52)
(0.53–0.85)
(0.69–1.19)
(0.58–0.93)
(0.26–0.47)
,0.0001
,0.01
0.48
,0.01
,0.0001
Medical condition in previous year
Liver disease
Viral hepatitis
Alcohol-related disease
Hospitalisation
4.36
9.39
1.84
5.24
(1.92–9.87)
(1.47–60.19)
(0.51–6.67)
(3.13–8.80)
,0.01
,0.05
0.35
,0.0001
2.47
3.70
1.98
4.90
(1.61–3.79)
(0.90–15.18)
(0.91–4.33)
(3.74–6.41)
,0.0001
0.07
0.09
,0.0001
Medication use in previous year
Acetaminophen
NSAIDs
Thiazolidinedione
Anti-epileptic agents
Anti-fungal agents
Antibiotics
HMG-CoA reductase enzyme inhibitors
Antiviral agents
1.34
0.92
1.61
1.58
0.21
1.27
2.01
1.83
(0.82–2.20)
(0.55–1.54)
(0.36–7.16)
(0.38–6.63)
(0.02–2.28)
(0.76–2.12)
(0.61–6.63)
(0.94–3.56)
0.24
0.75
0.53
0.53
0.20
0.37
0.25
0.08
1.21
1.27
1.26
2.13
0.92
1.25
1.56
0.94
(0.96–1.52)
(1.02–1.59)
(0.45–3.52)
(0.97–4.64)
(0.41–2.09)
(1.0–1.56)
(0.72–3.38)
(0.70–1.28)
0.11
,0.05
0.66
0.06
0.85
0.05
0.27
0.71
0.06
0.23
0.62
OR ¼ odds ratio; CI ¼ confidence interval; NSAID ¼ non-steroidal anti-inflammatory drug; HMG-CoA ¼ 3-hydroxy-3-methyl-glutaryl-CoA reductase.
to be associated with drug-induced liver injury.19
Prescribing physicians may therefore tend to monitor
liver function for patients with these risk factors. Our
study found that patients with pre-existing liver
disease, viral hepatitis and hospitalisation had greater
odds of being completely adherent to guidelines,
while this trend was not seen in age, sex and
medication use.
Patients whose prescribing physicians were specialists in chest diseases, TB or infectious diseases had
Figure 2 Liver function monitoring after the initiation of anti-tuberculosis drugs among newly
diagnosed patients with and without baseline liver function monitoring.
1250
The International Journal of Tuberculosis and Lung Disease
greater odds of being completely adherent to the
guidelines. In contrast, patients whose prescribing
physicians specialised in family medicine had lower
odds of being completely adherent to the guidelines.
Our analyses of patient and provider characteristics
provide additional insights for future strategies to
improve adherence rates. Specifically, we highlight
the need for additional efforts in educating practitioners specialising in family medicine to prescribe
LFTs for TB patients. Regional differences, as well as
differences in accreditation levels, in adherence rates
also indicate that greater educational efforts may be
required in regions other than Taipei.
Our pre-defined subgroup analysis served to
regroup study patients based on whether or not they
had received baseline LFT. Although the baseline
LFT results are not available, baseline LFT could be
a significant indicator of LFT monitoring after the
initiation of anti-tuberculosis treatment, as physicians may judge the need for subsequent monitoring
on the basis of baseline LFT results. Recommendations from the American Thoracic Society guidelines11 and British Thoracic Society guidelines12
recommend baseline LFT for all patients, and
regular monitoring thereafter only for those with
liver disease or abnormalities. Compared to patients
with baseline LFT, a lower proportion of patients
without baseline LFT underwent LFT monitoring. In
terms of the timing of LFT monitoring, hepatotoxicity occurs generally within weeks to months and
may exist with prolonged latency.11 Continuous
monitoring helps in the timely management of
hepatic abnormality. However, both groups had the
highest LFT rate at initiation of anti-tuberculosis
treatment, which subsequently decreased. Consistency in LFT monitoring is essential for ensuring
patient health.
Some potential limitations should be taken into
account when interpreting the results of the present
study. First, as our data were sourced from a claims
data set, we had information only on whether or not
the patient underwent LFT, but none on the LFT
results. We also did not have access to radiographs or
laboratory results to be able to distinguish between
the different types of TB. Second, although we were
able to include a wide range of patient and provider
characteristics, we could not include variables not
routinely captured in claims databases, such as
alcohol intake, which may have influenced the
physician’s decision to prescribe LFT.
Finally, as we found several significant patient and
provider characteristics associated with higher odds
of being completely adherent to the guidelines, our
findings may be reproducible and applicable in a
setting with different levels of pre-existing liver
diseases and a different medical infrastructure.
CONCLUSIONS
We observed increased, but not optimal, adherence to
liver function monitoring among newly diagnosed TB
patients during the decade from 2000 to 2011 in
Taiwan. Our findings on possible factors of complete
adherence to guidelines could serve as an important
reference for developing effective strategies to ensure
adherence and prevent TB patients from developing
drug-associated hepatotoxicity.
Acknowledgements
This work was supported by a research grant from the Food and
Drug Administration, Taiwan (DOH102-FDA-41100 and
MOHW103-FDA-41100). The funders had no role in study design,
data collection and analysis, decision to publish, or preparation of
the manuscript.
Conflict of interest: none declared.
References
1 World Health Organization. Global tuberculosis report, 2013.
WHO/HTM/TB/2013.11. Geneva, Switzerland: WHO, 2013.
2 Taiwan Ministry of Health and Welfare. Cause of Death
Statistics. Taipei, Taiwan: Ministry of Health and Welfare,
2014.
3 Yen Y, Rodwell T, Yen M, et al. DOT associated with reduced
all-cause mortality among tuberculosis patients in Taipei,
Taiwan, 2006–2008. Int J Tuberc Lung Dis 2012; 16: 178–184.
4 Yen Y-F, Yen M-Y, Lin Y-P, et al. Directly observed therapy
reduces tuberculosis-specific mortality: a population-based
follow-up study in Taipei, Taiwan. PLOS ONE 2013; 8:
e79644.
5 Meier Y, Cavallaro M, Roos M, et al. Incidence of drug-induced
liver injury in medical inpatients. Eur J Clin Pharmacol 2005;
61: 135–143.
6 Hartleb M, Biernat L, Kochel A. Drug-induced liver damage—a
three-year study of patients from one gastroenterological
department. Med Sci Monit 2002; 8: CR292–296.
7 Thompson N, Caplin M, Hamilton M, et al. Anti-tuberculosis
medication and the liver: dangers and recommendations in
management. Eur Respir J 1995; 8: 1384–1388.
8 Frieden T R, Sterling T R, Munsiff S S, Watt C J, Dye C.
Tuberculosis. Lancet 2003; 362: 887–899.
9 Tostmann A, Boeree M J, Aarnoutse R E, De Lange W, Van Der
Ven A J, Dekhuijzen R. Anti-tuberculosis drug-induced
hepatotoxicity: Concise up-to-date review. J Gastroenterol
Hepatol 2008; 23: 192–202.
10 Navarro V J, Senior J R. Drug-related hepatotoxicity. N Engl J
Med 2006; 354: 731–739.
11 Saukkonen J J, Cohn D L, Jasmer R M, et al. An official ATS
statement: hepatotoxicity of anti-tuberculosis therapy. Am J
Respir Crit Care Med 2006; 174: 935–952.
12 Chemotherapy and management of tuberculosis in the United
Kingdom: recommendations of the Joint Tuberculosis Committee of the British Thoracic Society. Thorax 1998; 53: 536–
548.
13 Byrd R B, Horn B R, Solomon D A, Griggs G A. Toxic effects of
isoniazid in tuberculosis chemoprophylaxis: role of biochemical
monitoring in 1,000 patients. JAMA 1979; 241: 1239–1241.
14 McNeill L, Allen M, Estrada C, Cook P. Pyrazinamide and
rifampin vs isoniazid for the treatment of latent tuberculosis:
improved completion rates but more hepatotoxicity. Chest
2003; 123: 102–106.
15 Agal S, Baijal R, Pramanik S, et al. Monitoring and
management of antituberculosis drug induced hepatotoxicity.
J Gastroenterol Hepatol 2005; 20: 1745–1752.
LFT monitoring in Taiwan, 2000–2011
16 Wu S, Xia Y, Lv X, et al. Effect of scheduled monitoring of liver
function during anti-tuberculosis treatment in a retrospective
cohort in China. BMC Public Health 2012; 12: 454.
17 Taiwan Centers for Disease Control, Department of Health,
Taiwan. Taiwan Guidelines for TB Diagnosis & Treatment. 5th
ed. Taipei, Taiwan: CDC, 2013. http://www.cdc.gov.tw/
infectionreportinfo.aspx?treeid¼075874dc882a5bfd&
1251
nowtreeid¼c6e4d08fdc49de51&tid¼ED3E98C81FC2D3C2.
Accessed June 2014. [Chinese]
18 Hsiao F, Yang C, Huang Y, Huang W. Using Taiwan’s national
health insurance research databases for pharmacoepidemiology
research. J Food Drug Anal 2007; 15: 99–108.
19 Bell L N, Chalasani N, eds. Epidemiology of idiosyncratic druginduced liver injury. Semin Liver Dis 2009; 29: 337–347.
LFT monitoring in Taiwan, 2000–2011
i
RESUME
O B J E C T I F : Etudier l’adhésion au suivi des fonctions
hépatiques préconisées par les directives de Taiwan pour
le diagnostic et le traitement d la tuberculose (TB) parmi
des patients tuberculeux récemment diagnostiqués.
S C H É M A : Etude rétrospective de cohorte dans la base
de données nationale de l’assurance santé (NHIRD) de
2000 à 2011, à Taiwan.
M É T H O D E S : A partir de la NHIRD, nous avons
identifié 11 397 patients tuberculeux r écemment
diagnostiqu és qui ont d ébut é leur traitement
antituberculeux entre 2000 et 2011 et qui ont été
classés en trois groupes : adhérents complets, partiels ou
non adh érents. La régression logistique a permis
d’explorer les déterminants potentiels associés au taux
d’adhérence.
R É S U LT A T S : Le taux d’adhérence complète a augmenté
de 0,5% en 2000 à 9,2% en 2011 tandis que le taux de
non-adhérence a diminué de 17,5% à 1,2%. Comparés
au groupe des non-adhérents, les patients ayant des
antécédents d’affection hépatique (OR 4,36 ; IC95%
1,92–9,87) et d’hépatite virale (OR 9,39 ; IC95% 1,47–
60,19) ainsi que les patients dont les médecins traitants
étaient spécialistes en TB pulmonaire (OR 4,59 ; IC95%
1,91–11,05), TB (OR 2,55 ; IC95% 1,01–6,40) et
maladies infectieuses (OR 3,93 ; IC95% 1,08–14,31)
avaient davantage de chances d’être complètement
adhérents aux directives.
C O N C L U S I O N : Nos résultats pourraient constituer une
importante référence pour élaborer des stratégies efficaces
pour améliorer l’adhérence aux directives et prévenir
l’hépatotoxicité des médicaments antituberculeux.
RESUMEN
Investigar el cumplimiento con la
supervisión de la función hepática de los pacientes con
diagnóstico reciente de tuberculosis (TB) que
recomiendan las directrices al diagn óstico y
tratamiento de la TB de Taiwán.
M É T O D O: Fue este un estudio retrospectivo de cohortes
a partir de la base de datos cientı́fica del sistema de
seguridad social en Taiwán del 2000 al 2011 (NHIRD).
M É T O D O S: En la base de datos del NHIRD se
encontraron 11 397 casos nuevos de TB que
comenzaron el tratamiento antituberculoso entre el
2000 y el 2011; en función de la observancia de las
directrices, se categorizaron los casos en tres grupos, a
saber cumplimiento total, parcial o incumplimiento. Se
aplicó un análisis de regresión logı́stica con el fin de
explorar los posibles factores determinantes que se
asociaban con la tasa de cumplimento de las directrices.
R E S U LT A D O S: La tasa de cumplimiento total aumentó
O B J E T I V O:
de 0,5% en el 2000 a 9,2% en el 2011 y la tasa de
incumplimiento disminuyó de 17,5% a 1,2%. En
comparaci ón con el grupo que no observó las
directrices, presentaron las mayores posibilidades de
un cumplimiento total de las directrices los pacientes con
antecedente de enfermedad hepática (OR 4,36; IC95%
1,92–9,87) y de hepatitis viral (OR 9,39; IC95% 1,47–
60,19) y los pacientes cuyo médico tratante era
especializado en neumologı́a (OR 4,59; IC95% 1,91–
11,05), en TB (OR 2,55; IC95% 1,01–6,40) y en
enfermedades infecciosas (OR 3,93; IC95% 1,08–
14,31).
C O N C L U S I Ó N: Estos resultados constituirán una
referencia importante al formular estrategias eficaces,
encaminadas a estimular el cumplimiento de las
directrices y prevenir la hepatotoxicidad asociada con
los medicamentos antituberculosos.