Full Text Review
Please note: In 2003, the CTF updated its Grades of
Recommendations to include an "I Recommendation" for situations where
insufficient evidence exists to allow a recommendation to be made.
(Formerly, these situations were captured under a "C
Recommendation".) This change is not retroactive, and all
"C Recommendations" made prior to 2003 have not been
reevaluated in light of the new "I" recommendation grade. For a
discussion of these recommendation grades, please link to the 2003 article in
the Canadian Medical Association Journal here.
Screening and Isoniazid Prophylactic Therapy for Tuberculosis
Prepared by Sharon L. Walmsley, MD, FRCPC, Assistant Professor, Department
of Medicine and Microbiology, University of Toronto
These recommendations were finalized by the Task Force in January 1994
Contents
Overview
Tuberculosis remains a major worldwide health problem
with at least 10
million new cases and 3 million deaths estimated to occur annually. In
1979,
the Canadian Task Force on the Periodic Health Exam< 1>
recommended Bacille Calmette-GuŽ erin (BCG) immunization and chemoprophylaxis
of unimmunized tuberculin-positive individuals from high risk-groups (A
Recommendation) but not for the general population (E Recommendation).
With the changing epidemiology of tuberculosis in North America and concerns
about the risk of drug-induced hepatitis, the value of screening for tuberculosis
and of isoniazid (INH) chemoprophylactic therapy as part of the periodic
health examination was re-examined. The Task Force concludes that screening
by tuberculin skin testing should be offered to persons at high risk of
infection, but is not recommended for the general population. INH chemoprophylactic
therapy is recommended for household contacts of active cases of tuberculosis.
It is also recommended for persons with positive tuberculin skin tests
who have documented skin test conversion, radiographic lesions suggestive
of inactive but previously untreated tuberculosis, who are under 35 years
of age, or those with an underlying medical condition which predisposes
them to reactivation (especially HIV infection) regardless of age.
Burden
of Suffering
The number of cases of active tuberculosis in Canada
has been decreasing gradually over the past few decades, and in the past
few years has plateaued, with 1,995
cases (7.5 per 100,000)
reported in 1990.<2>
For individuals born in Canada, most reported cases occur in the very young
and in those of advancing age. Among the Canadian-born, rates are ten times
higher for the aboriginal group. Proportionally, more active cases occur
in immigrants to Canada, with the highest rates reported in those immigrating
from areas where tuberculosis is endemic, including Africa, Asia, Central
America and certain countries in South America and the Caribbean.
Death rates from tuberculosis have also declined.
In Canada in 1990,
129
persons with tuberculosis died (0.6 per 100,000
population) and in 60 cases, tuberculosis was the cause of, or a significant
contributor to death.
Most new cases of tuberculosis are pulmonary. Persons
infected with large inocula (i.e. following household exposure to a cavitary
case) and those with increased susceptibility to infection (e.g. children
less than 5 years) are at higher risk of acquiring infection. In 90% of
exposed persons, host defenses contain the primary infection, but the individual
develops a positive tuberculin skin test. In the absence of preventive
measures 5-15%
will develop reactivation tuberculosis during their lifetime, the risk
being greatest during the first two years following exposure. From observational
studies, certain groups have been identified to be at an increased risk
of reactivation, although in many cases the exact level of risk is not
well defined. In patients with silicosis, head and neck cancer, jejunoileal
bypass or in those who require hemodialysis, the relative risk is reported
to be increased 10-30
times over that of the baseline population. The relative risk for patients
with low body weight or nutritional deficiency, diabetes mellitus, or gastrectomy
is reported to be increased 2-5 times. Persons infected with HIV are now
recognized to be at the highest risk of reactivation with rates reported
at 8-9% per year in recent studies. Other groups identified to be at increased
risk are those with other immunosuppressive disorders (hematologic malignancies),
those requiring immunosuppressive therapy for malignant or non-malignant
conditions, patients requiring high dosages of corticosteroids over prolonged
periods, the urban poor, persons living in shelters, intravenous drug users
and alcoholics.
Maneuver
The Mantoux tuberculin skin test is the gold standard
screening test to identify infection with the human tubercle bacillus.
For testing, 0.1 ml
of PPD-T (standardized purified protein derivative, Connaught Laboratories)
containing 5 TU (tuberculin units) is injected intracutaneously into the
dorsal (volar) surface of the forearm. The 1
TU
and 250 TU tuberculins are not biologically standardized and are not recommended
for routine use. The injection is made intradermally and a discrete wheal
6-10 mm in
diameter should be produced in all cases. Tuberculin tests should be read
at 48-72 hours. The basis of reading is the transverse diameter in millimetres
of the induration (not the erythema) which can be determined by inspection
and palpation or by the ball point pen method.
The cut-off point for the interpretation of a positive
test was derived from multiple population surveys comparing the reaction
size in patients with or without presumed exposure and development of subsequent
active disease. There is no clear point of separation between a positive
and a negative test. Defining a positive reaction becomes increasingly
difficult as the prevalence of cross-reactions with atypical mycobacteria
increases and true infections due to Mycobacterium tuberculosis become
less frequent. These observations have led experts to recommend use of
different cut-off values for different population groups.<3> For individuals
at very high risk of reactivation, including contacts of infectious cases,
and persons with HIV infection, the recommended cut point is 5 mm. For
persons not likely to be exposed, who reside in geographic areas where
infection with atypical mycobacterium is common (i.e. the United States),
the cut off is increased to 15
mm. For those with an intermediate risk, such as those living in or immigrating
from endemic areas, or those areas where cross reactions with atypical
mycobacteria are less common (i.e. Canada) an intermediate level of 10
mm is used. It should be noted that most patients who receive BCG vaccination
as children, lose their cutaneous hypersensitivity reaction to tuberculin
within 5 years. Therefore, a significant reaction more likely represents
true exposure to tuberculosis especially in the setting of a recent exposure.
Another problem with tuberculin skin testing is inter-
and intra-observer variation in the interpretation of test results.<4>
Agreement is highest when the results are clearly positive (>15
mm) or clearly negative (<5 mm), with more inter-observer variation
(mean 6 mm) when more readings are in the 5-14
mm range. This degree of variation can result in incorrect labelling. Some
patients with compromised immune systems (especially those infected with
HIV who have low CD4 counts) may be anergic. These individuals are unable
to mount a skin test response to injected antigens (including tuberculin)
despite previous exposure, and may be incorrectly labelled as negative.
The tuberculin skin test is a safe procedure. Purified
protein derivative (PPD) does not sensitize nor activate a latent reaction.
A few (1-2%)
exquisitively sensitive persons with positive tests may respond with a
severe local reaction with ulceration or vesiculation, regional adenopathy
and fever. These reactions are typically self-limited, but painful, and
may be reduced with intralesional steroids. Localized allergic reactions
including wheal and flare and localized rash may occur in 2-3% of patients
particularly those with a history of atopy. These reactions tend to occur
in the first few hours, are self-limited, and do not correlate with the
tuberculin response. Anaphylaxis or death have not been reported.
For persons with positive reactions it is imperative
to rule out active disease by clinical history, chest radiograph and the
appropriate laboratory investigations. Once active infection is ruled out,
it is important to explain to the patient that a positive reaction implies
a subclinical infection with tuberculosis and the risk for subsequent reactivation
to active disease, but that the individual does not have active disease
nor are they infectious to others.
Effectiveness
of Prevention
The main purpose of chemoprophylaxis is to prevent reactivation
of a latent infection to clinical disease. It is important prior to the
initiation of chemoprophylaxis that active TB is excluded. The only drug
which has been studied extensively for chemoprophylaxis is Isoniazid (INH).
More than 20 clinical trials evaluating the role of chemoprophylaxis in
tuberculosis are reported in the literature. Despite the multitude of methodological
flaws of these early studies, including 1)
non-random allocation; 2) inclusion of previously treated cases; 3) use
of other therapies in combination with INH; 4) historical controls; 5)
randomization by groups (families, hospital wards), but analysis by individuals;
and 6) the inclusion of active cases, all but one of these trials has shown
a significant benefit for INH prophylaxis. The best studies demonstrating
effectiveness of INH prophylaxis in different patient groups are outlined
below:
Positive Tuberculin
Skin Test <35 Years
During a routine skin testing program (1958-1966)
in San Francisco public school students with a large immigrant population,<5>
a non-randomized comparative study evaluated INH prophylaxis. Over a 1-2
year follow up, 1 of
2,910 children
in the INH-treated group (0.34 per 1,000)
and 25 of 1,192
on no treatment (20.9 per 1,000)
developed active tuberculosis. Although demonstrating a highly statistical
and clinical benefit of INH prophylaxis, there is concern about the comparability
of the two groups, with racial differences noted between the treatment
and control groups. Other reasons for not choosing treatment may also have
been associated with increased risk of tuberculosis in controls. There
is fair evidence to support prophylaxis in this setting, especially for
children.
Positive Tuberculin
Skin Test > 35 Years
27,924 patients from 39 U.S. Mental Institutions were
randomly allocated by ward to receive INH or placebo.<6> During the
medication year, 21 of
12,326
(1.7 per
1,000)
patients assigned placebo and 4 of 12,884
(0.3 per 1,000)
assigned INH developed active tuberculosis, with 17
of the 25 patients who developed tuberculosis having had an abnormal chest
x-ray on entry. Although this is a statistically significant reduction
in cases, the clinical significance and generalizability of the results
are uncertain. The crowding and institutionalization would likely have
increased the risk of disease. Other biases in the study weaken the strength
of the observations. The Task Force feels there is insufficient evidence
to recommend routine prophylaxis for all adults >35 years with a positive
tuberculin skin test.
Although no controlled studies have been performed,
INH prophylaxis is strongly recommended (by expert opinion) for patients
with medical conditions that increase the risk of reactivation of tuberculosis,
regardless of age.<16>
HIV Infected
Patients
From 1986-89,
118
asymptomatic HIV positive persons in Haiti were randomized to receive INH
prophylaxis with vitamin B6 or B6 alone for 12
months.<7> Eleven of sixty (18%)
patients in the B6 group developed active tuberculosis over the study period
in contrast to four of 62 who received INH and B6, p=0.03. This is consistent
with a reduction from 7.5 to 2.2 cases per 100-person
years. There is good evidence to support the use of INH prophylaxis in
patients with HIV infection.
Household Contacts
of Active Cases
The Tuberculosis Program of the U.S. Public Health Service
studied INH prophylaxis in household contacts of new active pulmonary cases
in the U.S. and Puerto Rico between 1957
and 1960.<8>
The groups were similar in age and race distribution. Two thirds of all
participants <20 years were randomized to receive INH or placebo for
one year. Pulmonary or extrapulmonary tuberculosis developed in 78 of 12,594
(6.5 per 1,000)
placebo recipients versus 18
of 12,439
(1.5 per
1,000)
INH recipients.
A similar household-based study in Japan failed to
show a benefit of INH prophylaxis<9> although there was a trend toward
decreased tuberculosis in the INH treated group (8 per 1,142)
compared to the control group (11 per
1,096).
Poor compliance and the small sample size were the most likely reasons
for the lack of significant effect.
The Task Force concludes there is good evidence to
support the role of INH chemoprophylaxis in household contacts of active
cases of tuberculosis.
Tuberculin Skin
Test Converters
Recruits of the Royal Netherlands Navy whose skin tests
converted to positive following exposure to an active case of pulmonary
tuberculosis were randomized to receive INH (n=133)
or placebo (n=128)
for one year.<10>
During the treatment year, 9 cases of tuberculosis developed in the placebo
recipients (70 per 1,000)
in contrast to 1 case
in the INH group (7.5 per 1,000),
a significant difference.
INH prophylaxis was evaluated in a cohort of nursing
home patients with tuberculin skin test conversion.<11>
Tuberculosis developed in 1 of
605 converters receiving INH (1.6
per 1,000)
and in 45 of 757 (59 per 1,000)
receiving no treatment. This was clinically and statistically significant.
The Task Force concludes that there is good evidence
to recommend INH prophylaxis in skin test converters
regardless of age.
PPD Positive,
Old Fibrotic Scar on Chest Radiograph
The International Union Against Tuberculosis (IUAT)
performed a European multi-center, double-blind, placebo-controlled trial
of INH prophylaxis in patients with a positive PPD and an old fibrotic
scar on chest x-ray.<12>
Of the 27,830 patients enrolled (median age 50 years), 6,953 were randomized
to receive INH for 12
weeks, 6,965 INH for 24 weeks, 6,919
INH for 52 weeks and 6,990 patients were randomized to placebo. Over the
5 year study, active pulmonary tuberculosis developed in 97 placebo recipients
(14.3 per
1,000),
76 patients on INH for 12
weeks (11.3
per 1,000),
34 patients on 24 weeks of INH (5.0 per 1,000)
and 24 patients on 52 weeks of INH (3.6 per 1,000).
This represents a 60% reduction in cases for patients receiving a 6 month
course and approximately 90% reduction for patients receiving a 12
month course of INH. An additional 181
patients
were removed from the trial as suspected cases of tuberculosis (either
extra-pulmonary tuberculosis, or pulmonary tuberculosis without positive
culture) but it is not reported how many of these patients received INH.
Based on this study the Task Force feels there is fair evidence to support
the use of INH prophylaxis in this setting.
Duration of Therapy
Most studies evaluating INH chemoprophylaxis have utilized
a one year course of therapy, and this is the duration recommended by the
Task Force. A cost analysis study has suggested that a 24 week regime was
more cost effective.<13>
Alternative Regimes
There have been case reports of failed INH chemoprophylaxis
for patients infected with INH-resistant organisms. There is no study available
to recommend alternative agents in this setting. Expert opinion suggests
that Rifampin and Ethambutol, Rifampin and INH or Pyrazinamide containing
regimes should be considered.
Compliance
Compliance remains an important barrier to effective
chemoprophylaxis. In the clinical trials cited above, compliance ranged
from 50-75% through one year of preventive therapy. In a study involving
Canadian aboriginals, compliance was improved by a program of daily observed
prophylaxis and education but was not sustained after observation was discontinued.<14>
Adverse Events
of Screening and Treatment
Prophylaxis with INH has been associated with many adverse
events, including hypersensitivity reactions, INH-induced lupus-like syndrome,
peripheral neuropathy, gastrointestinal distress, and central nervous system
abnormalities ranging from memory loss to psychosis or seizures. The most
significant and potentially dose-limiting side effect is hepatitis. Fifteen
percent of patients will experience a transient asymptomatic increase in
their liver transaminases on treatment. Clinical hepatitis is much less
common and is rarely fatal, particularly when recommendations for surveillance
are followed.
The best prospective study to determine the incidence
of INH hepatitis documented 236 suspected cases among 13,838
receiving prophylaxis.<15>
Of these cases, 92 were classified as probably INH-related and 82 possibly
INH-related by an expert panel. The case rates appeared to increase with
age, with rates for probable cases <20 years (0 per 1,000),
20-34 years (3 per 1,000),
35-49 (12
per 1,000),
50-64 years (23 per 1,000)
and >64 years (8 per 1,000).
The rates also varied with race, with the highest rates occurring in Orientals
(18 per 1,000)
in contrast to 11.4
per 1,000
in whites and 7.1 per
1,000
for blacks. The rate for those who drank alcohol (14.1
per
1,000)
was double that of non-drinkers and the rate for daily drinkers (26.5 per
1,000)
was even higher. In the IUAT trial, the risk of INH-related hepatitis was
0.5%, with a mortality rate of 14
per 100,000.<12>
Other disadvantages of chemoprophylaxis include the
inconvenience of daily medication over a one year period and the need to
attend regular clinic visits. The cost of the program, and demands on staff
for screening and follow-up must also be considered.
Recommendations
of Others
The American Thoracic Society (ATS) has recently published
new guidelines.<7> They propose the same groups for tuberculin testing
as the Canadian Task Force but use "recent contacts" rather than "household
contacts" of new cases. The high risk group for reactivation infection,
also includes medically underserviced populations, persons resident in
long-term care facilities, nursing homes, mental institutions and correctional
institutions. They do not recommend prophylaxis for patients with positive
tuberculin skin test and fibrotic lesions on chest radiographs, but rather
recommend that these patients be treated with multi-drug regimes as for
active disease. They recommend INH prophylaxis for all patients with known
or suspected HIV infection and a positive PPD.
The Centers For Disease Control in Atlanta also recommend
routine skin testing of all persons at high risk, including the homeless.<17>
The Canadian Thoracic Society has also developed
new guidelines (personal communication, Dr. M. Fitzgerald). They recommend
that given the low frequency of infection with atypical mycobacteria, that
the 10 mm
cut-off be used for the general Canadian population.
In 1989,
the U.S. Preventive Services Task Force recommended skin testing for patients
at increased risk for tuberculosis.<18>
In addition, they recommend following the ATS guidelines for prophylaxis.
The American Academy of Pediatrics Redbook recommends
annual testing of high-risk children, including those who are black, Hispanic,
Asian, American Indian, Native Alaskan, the socioeconomically deprived,
those living in neighbourhoods where case rates are above the national
average, children or their parents who themselves have immigrated from
high risk areas of Asia, Africa, the Middle East, Latin America or the
Caribbean, households where there have been active cases and those with
underlying medical risk factors.<19>
Annual testing is not recommended for low-risk groups but skin testing
in these low-risk children is recommended at 12-15
months, 4-6 years, and 14-16
years. INH is recommended for all infants, children and young adults up
to 35 years of age with positive PPD for 9 months. No justification is
given for the duration
Conclusions
and Recommendations
Tuberculosis screening should be offered to persons
in Canada at high risk of infection with the tubercle bacillus, including
immigrants from endemic areas, Canadian-born aboriginals, close contacts
of active cases, persons with abnormal chest radiographs consistent with
healed tuberculosis, and persons with underlying medical conditions which
increase their likelihood of reactivation of tuberculosis if infected,
including those with HIV infection, silicosis, hemodialysis patients, those
with immunosuppressive conditions or therapy, intravenous drug users, diabetes,
gastrectomy patients or those with gastrointestinal bypass surgery, and
the nutritionally deficient (A
Recommendation). Routine screening is not recommended for the general
population (E
Recommendation). INH prophylaxis for twelve months is recommended for
household contacts of active cases of tuberculosis (A
Recommendation) and for persons with positive tuberculin skin tests
who have documented skin test conversion (A
Recommendation), chest radiographic lesions suggestive of inactive
tuberculosis (B
Recommendation), HIV infection (A
Recommendation) and for patients under 35 years of age with a positive
tuberculin test (B
Recommendation). Prophylactic therapy is not recommended for persons
with positive skin tests over the age of 35 years (C
Recommendation) unless they have a medical condition associated with
an increased risk of reactivation where prophylaxis is recommended (A
Recommendation).
Unanswered
Questions (Research Agenda)
A number of areas continue to be of concern regarding
the detection of infected persons and the use of prophylaxis. Priorities
that need to be addressed include the development of: 1)
more effective, less toxic, shorter duration preventive treatments; 2)
ways to improve compliance with chemoprophylaxis; 3) effective vaccines;
4) better studies to determine the true incidence of reactivation tuberculosis
in patients with isolated positive tuberculin skin tests with no other
risk factors; 5) the true relative risks for reactivation in those patients
with various underlying medical disorders; and 6) chemoprophylactic measures
for patients infected with INH resistant strains.
Evidence
The literature was identified with a MEDLINE search
with the use of the main MESH heading "tuberculosis" and the subheadings
"tuberculin skin test", "prevention and control" and "isoniazid" for articles
presented from 1966
through 1992.
Indices of the American Review of Respiratory Diseases and Tubercle were
screened from 1960-1992
for articles with emphasis on screening and treatment.
This review was initiated in August 1992
and the recommendations were finalized by the Task Force in January 1994.
Acknowledgements
The Task Force thanks Sam Akor, BSc, MB, ChB, MCommH,
MA, Director/DMOH, Rural Health Training School, Kintempo, Ghana for his
work on the preliminary report. Helen Holden MD, FRCPC, FCCP, Valley Regional
Hospital, Kentville, Nova Scotia, Richard Menzies, MD, MSc, FRCPC, Assistant
Professor, McGill University and Montreal Chest Hospital, Montreal, Quebec
and Mark Fitzgerald, MD, FRCPC, Assistant Professor, University of British
Columbia and Department of Respiratory Medicine, Vancouver General Hospital,
Vancouver, British Columbia are thanked for reviewing the draft report.
Full Citation
Link to Structured Abstract of
this review
Link to Summary Table of this
review
Link to Selected References list of this review
Top of Page
CTFPHC
Home Page
Reprinted in modified format by the Canadian
Task Force on Preventive Health Care
with permission.
For any technical issues please contact: webmaster@ctfphc.org
Original Copyright
© 1994 Minister of Supply and Services Canada.
Last modified March 27, 1998.