Jornal Brasileiro de Pneumologia 31(2) - Mar/Abr de 2005
Original Article
Analysis of treatment outcomes related to the
tuberculosis control program in the city of Campinas, in
the state of São Paulo, Brazil*
HELENICE BOSCO DE OLIVEIRA, LETICIA MARIN-LEÓN, JOVANA GARDINALI
Background: Tuberculosis cure is dependent upon treatment adherence.
Objective: To analyze the results of tuberculosis treatment in public health clinics in the city of Campinas, in the state
of São Paulo, Brazil, during 2002.
Method: From a cohort of 484 patients diagnosed with tuberculosis, we evaluated 436. Treatment outcomes
were described for all patients, whether new patients or patients in retreatment, including those presenting
the pulmonary form, with or without acquired immunodeficiency syndrome (AIDS) comorbidity.
Results: The success rate was 68.6% (72.3% among non-AIDS patients and 57.6% among AIDS patients).
Among new cases, the non-AIDS group presented a 2.2-times greater chance of presenting favorable results. In
the AIDS group, no differences were observed between new patients and those in retreatment. Among the
unfavorable outcomes, only lethality presented a difference (18.9% among AIDS patients and 8.0% among nonAIDS patients). In patients presenting the pulmonary form, the success rate was similar between those who were
initially acid-fast bacilli positive and those who were not.
Conclusion: The tuberculosis control program in Campinas presented low effectiveness. In comparison to
the 2001 national cohort, success rates were higher for non-AIDS patients but lower for AIDS patients. The
higher success rate among cases of tuberculosis without AIDS was primarily derived from the treatment of
new cases. The unfavorable profile of tuberculosis patients co-infected with AIDS, characterized by the
(18.9%) lethality and the (15.3%) noncompliance, were partially responsible for the lower success rate seen
among such patients. It is notable that such high proportions of noncompliant patients were seen in a city
providing easy access to treatment. In order to improve the program, medical teams should receive further
training in supervised treatment, health education and techniques for interacting with patients, all of
which will require considerable investment.
J Bras Pneumol 2005; 31(2): 133-8.
Key words: Tuberculosis. Treatment Refusal. Retreatment. Acquired Immunodeficiency Syndrome. Comorbidity
* Study carried out in the Department of Preventive and Social Medicine of the UNICAMP School of Medical Sciences.Endereço para correspondência:
Correspondence to: Helenice Bosco de Oliveira. Rua Waldyr Aparecido da Silva, 60, Residencial 2. Condomínio Barão do Café. CEP 13085-065
- Barão Geraldo. Campinas-SP. E-mail: helenice@unicamp.br
Submitted: 22 June 2004. Accepted, after review: 4 February 2005.
133
Oliveira, HB, et al.
Analysis of treatment outcomes related to the tuberculosis control
program in the city of Campinas, in the state of São Paulo, Brazil
INTRODUCTION
Although tuberculosis (TB) is a severe disease, it is
curable in virtually 100% of new cases if chemotherapy
principles are followed(1). Due to the existence of
different bacterial populations, in distinct metabolic
situations and reacting diversely to medication(2), that
easily become resistant when chemotherapy is
inappropriate(3), it is necessary to combine at least three
drugs at the beginning of the treatment.
In Brazil, the Programa de Controle da
Tuberculose (PCT, Tuberculosis Control Program)(4)
recommends breaking the chain of disease
transmission as one of the eradication strategies.
The effectiveness of the PCT depends on both the
percentage of patients who complete the treatment
and the efficacy of the medication (5).
Currently, the biggest problem encountered in
TB treatment is patient noncompliance(6). As a result,
the indicators of incidence, mortality and multidrug
resistance are rising(7), and this is currently the
principal cause for concern among health care
professionals worldwide. A patient who abandons
TB treatment becomes a major source of
transmission of the bacillus, especially to individuals
infected with human immunodeficiency virus (HIV).
In this context, an increasing understanding of
certain quantitative data regarding treatment results
has become relevant due to the consequences these
data have for individual and collective action. The
need to monitoring the PCT through periodic
evaluation of treatment outcomes, together with the
growing need for research into TB in the city of
Campinas (SP), motivated the present study, which
aims to analyze the results of TB treatment in public
health clinics in the city of Campinas during 2002.
METHODS
Treatment outcomes of the cohort of individuals
enrolled in the PCT in 2002 in the city of Campinas
were analyzed using data from the Epidemiological
Surveillance Database of the Universidade Estadual
de Campinas (UNICAMP, Campinas State University).
The information in this database came from the
Sistema Nacional de Agravos de Notificação (SINAN,
National Case Registry) through the Campinas
Municipal Department of Health.
The patients in the 2002 cohort began treatment
between 1 January and 31 December 2002.
In the variable “admission status”(4), patients were
classified as “new cases” (in which no specific treatment
134
had previously been administered) or “retreatment
cases”, (those who reentered the PCT due to relapse,
prior noncompliance or previous treatment failure).
Not all TB patients undergo HIV testing, which
appears on the SINAN form as “HIV testing not
performed”. Data on the number of patients with
acquired immunodeficiency syndrome (AIDS) were
obtained from the fields “associated diseases” and
“HIV test results”.
Treatment completion was determined as follows:
in the seventh month for patients presenting
pulmonary forms, extrapulmonary forms and AIDS
comorbidity; in the tenth month for special situations
such as TB and diabetes, chronic use of corticosteroids,
post-gastrectomy, chronic kidney disease, etc.; and in
the thirteenth month for patients under treatment
regimens for treatment failure or for meningitides.
Treatment outcomes were classified(4) as “cure/
treatment complete”, “noncompliance”, “transfer”,
“mortality” or “treatment failure”. “Cure/treatment
complete” was defined as two negative sputum smears,
one in the follow-up phase and one in the final
assessment. This group also included those patients
who presented sputum smears negative for acid-fast
bacilli (AFB) at the beginning of the treatment and
took the medication for the prescribed length of time
(treatment complete/cure not confirmed). The term
“noncompliance” was used to describe cases in which
the patient failed to appear in the clinic for more than
30 consecutive days after the scheduled appointment.
When the patient was transferred to another health
clinic, the outcome was defined as “transfer”. Outcomes
were classified as “mortality” if a patient death was
reported during the treatment, regardless of the cause.
“Treatment failure” was defined as positive sputum
smears throughout the treatment period or until the
fourth month.
The favorable outcomes, represented by the
cures and treatments completed, were compared
to the unfavorable outcomes (noncompliance,
mortality, transfer and treatment failure). The
dichotomous favorable/unfavorable-outcome
variable was determined based on exposure factors:
admission status (new case/retreatment), AIDS
comorbidity and (in patients presenting the
pulmonary form) sputum smear microscopy results.
Of the 484 patients enrolled in the PCT, 37
were excluded because their treatment outcomes
were not registered in the SINAN, 6 were excluded
because they were diagnosed as being infected
Jornal Brasileiro de Pneumologia 31(2) - Mar/Abr de 2005
with atypical mycobacteria, and 5 were excluded
because they were still under treatment. Therefore,
the final study sample consisted of 436 patients.
The database containing data related to the patients
who entered treatment in 2002 was analyzed using
proportions, comparison of proportions, odds ratio
(OR) with a 95% confidence interval (95% CI) and
the chi-square test with Yates’ correction. The software
Epi Info 6.04 d was used for this analysis. The level of
significance adopted was d” 5%.
TABLE 1
Effectiveness of the treatment provided through
the Tuberculosis Control Program for tuberculosis
patients with and without AIDS
Comorbidity
With AIDS
Without AIDS
Total
Success
235
64
299
Failure
90
47
137
Total
325
111
436
OR = 1.9; 95% CI: 1.20 - 3.08; p = 0.0059
RESULTS
Of the 436 TB patients, 111 (25.5%) presented
concomitant AIDS (Table 1). New patients numbered
360 (82.6%), compared to 76 (17.4%) who were
patients in retreatment (Table 2). AIDS concomitance
was seen in 23.6% (85/360) of the new patients and
34.2% (26/76) of the patients in retreatment (Table
2). A total of 364 (83.5%) of the patients presented
the pulmonary form and 73 (16.8%) presented the
extrapulmonary form (Table 3).
The success rate (cure/treatments complete) was
68.6% (299/436): 72.3% (235/325) among nonAIDS patients and 57.6% (64/111) among AIDS
patients. This finding shows that the PCT was more
effective for non-AIDS patients (OR = 1.92; 95% CI
= 1.20 - 3.08; p = 0.0059) (Table 1).
Among new patients and patients in retreatment,
the effectiveness profile of treatment provided by the
PCT differed between AIDS patients and non-AIDS
patients. The success rate was higher in the group of
new patients without AIDS than in the group of new
patients with AIDS (74.9% vs. 54.1% or 206/275 vs.
46/85; ÷2 = 13.37; p = 0.0002). The non-AIDS group
presented a 2.2-times greater chance of presenting
favorable results among new cases (OR = 2.2; 95% CI
= 1.11 - 4.22; p = 0.0222). In the AIDS group, no
differences were observed between new patients and
those in retreatment regarding treatment success (OR
= 0.52; 95% CI = 0.18 - 1.46; p = 0.2551) (Table 2).
Table 4 shows the analysis of the unfavorable
outcomes (noncompliance, mortality, treatment
failure and transfer) in the AIDS and non-AIDS
groups. Comparison of these outcomes revealed
that only mortality presented a significant
difference: 18.9% (21/111) among AIDS patients
and 8.0% (26/325) among non-AIDS patients (÷2
= 10.26; p = 0.0014). Noncompliance, transfer and
treatment failure presented no significant
differences.
In Table 4, we can also see that, when patients
were stratified into new cases and retreatment
cases, noncompliance was 8.9% (32/360) among
new patients and 22.4% (17/76) among patients
in retreatment. Patients in retreatment presented a
3-times greater chance of being noncompliant (OR
= 3; 95% CI = 1.46 - 5.92; p = 0.0015).
TABLE 2
Effectiveness of the treatment provided through the Tuberculosis Control Program for tuberculosis
patients with and without AIDS by admission status
Admission status
New cases
Retreatment cases
Total
With AIDS (n = 111)
Tt
Tt
Subtotal
success
failure
46
39
85
18
8
26
64
47
111
Without AIDS (n = 325)
Tt
Tt
success failure
206
69
29
21
235
90
Subtotal
275
50
325
Total
360
76
436
OR (with AIDS) = 0.52; 95% CI: 0.18 - 1.46; p = 0.2551
OR (without AIDS) = 2.2; 95% CI: 1.11 - 4.22; p = 0.0222
Tt: treatment.
135
Oliveira, HB, et al.
Analysis of treatment outcomes related to the tuberculosis control
program in the city of Campinas, in the state of São Paulo, Brazil
not undergo the exam (22/144) than among those
who were AFB positive (15/220) (15.3% vs. 6.8%;
÷2 = 6.82; p = 0.0090). The other outcome groups
did not present any significant differences
regarding the sputum smear microscopy results.
TABLE 3
Effectiveness of the treatment provided through the
Tuberculosis Control Program for pulmonary tuberculosis
patients, according to sputum smear microscopy
Pulmonary tuberculosis
Sputum smear positive
Sputum smear negative
or not performe
Total
Success Failure
153
67
Total
220
101
254
144
364
43
110
DISCUSSION
OR = 0.97; 95% CI: 0.60 - 1.58; p = 0.9969
The analysis of treatment outcomes, co-infection
with AIDS and admission status (Table 4) revealed
that, among new cases, noncompliance was 15.3%
(13/85) in the AIDS group and 6.9% (19/275) in the
non-AIDS group (÷2 = 5.64; p = 0.0176). In the same
group (new cases), mortality was 22.4% (19/85) in
the AIDS group and 8% (22/275) in the non-AIDS
group - a significant difference (÷2 = 13.25; p =
0.0003). Among patients in retreatment, no
differences regarding treatment outcomes were
observed between AIDS and non-AIDS patients.
Table 3 shows the effectiveness of the PCT in the
treatment of patients presenting the pulmonary forms.
There was no difference in the success rate between
those who were initially AFB positive and those who
were not (OR = 0.97; 95% CI = 0.60 - 1.58; p =
0.9969). Similarly, when AIDS and non-AIDS patients
were compared, no significant difference was found.
Table 5 shows the stratification of the treatment
outcomes in patients with the pulmonary forms
based on sputum smear microscopy results.
Mortality was significantly higher among patients
who were AFB negative and among those who did
The development of efficacious chemotherapy
for TB was one of the greatest medical advances
of the twentieth century. Rouillon et al.(1) state that
the use of antibacterial drugs completely changes
the natural evolution of TB. It not only increases
survival rates and makes cure possible but also
considerably reduces the period of infectiousness,
prevents recurrence and eliminates chronic cases.
However, this efficaciousness is partially reduced
since patients do not always complete the
treatment successfully. Campaigns to control TB
aim to detect 70% of the expected cases and cure
at least 85% of the diagnosed cases(4,5).
The World Health Organization, in its eighth
report (2004) (7) , in which the 2001 cohort is
analyzed, reports that the treatment success rate
was low in three countries (< 70%): Russia, Uganda
and Brazil (67%). The situation was similar in the
city of Campinas: in 2002 the PCT presented low
effectiveness, with a success rate of only 68.6%.
Campinas is a city where there is easy access to
treatment and there are public health services,
including two medical schools, providing
decentralized TB treatment. This leads us to expect
that the result would be different from what was
determined. Although the Directly Observed
Treatment, Short-course strategy has proved to be
effective in several parts of the world(8,9), it is still
in the implementation phase in the city.
TABLE 4
Distribution of tuberculosis patients by AIDS comorbidity and treatment outcome
Treatment outcome
Cure/treatment complete
Noncompliance
Mortality
Transfer
Treatment failure
Total
136
New
46
13
19
7
0
85
With AIDS
Retreat/ Subtotal
18
64
4
17
2
21
1
8
1
1
26
111
New
206
19
22
27
1
275
Without AIDS
Retreat/ Subtotal
29
235
13
32
4
26
4
31
0
1
50
325
TOTAL
299
49
47
39
2
436
Jornal Brasileiro de Pneumologia 31(2) - Mar/Abr de 2005
TABLE 5
Distribution of pulmonary tuberculosis patients by sputum smear microscopy results and treatment outcome
Treatment outcome
Cure/treatment complete
Noncompliance
Mortality
Transfer
Treatment failure
Total
Sputum smear
positive
153
30
15
20
2
220
In Campinas, success rates were different for
AIDS patients than for non-AIDS patients, showing
the importance of stratifying the cohort according
to this variable in the analysis of effectiveness.
Much higher treatment success rates were seen in
non-AIDS patients than in AIDS patients (72.3%
and 57.6%, respectively). In comparison to the
2001 national cohort(7), the effectiveness of the
treatment provided through the PCT was higher
for non-AIDS patients than for AIDS patients.
The success rate among non-AIDS TB cases was
primarily derived from the positive outcomes of
the treatment of new cases, in which the chance
of presenting favorable results was 2.2-times
greater than in the retreatment group.
In the 2001 Brazilian cohort(7) , the success rate
among patients in retreatment was 47%. Campos
et al.(10) observed the same proportion of favorable
results in the city of Recife (PE) in 1997. In the
present study, 47 (61.8%) of the 76 patients in
retreatment were cured or completed the treatment.
This rate was higher than the national rate, as well
as that recorded for Campinas in the 1993-1994
period, when the rate was lower than 50% (11). In
the comparison between TB patients with AIDS
comorbidity in retreatment between 1993 and 1994
and those evaluated in the present study, mortality
was seven times higher in the 1993-1994 period,
indicating that the increased rates of success may
be related to decreased mortality among AIDS
patients as a result of the use of antiretroviral drugs.
In the present study, the success rate among
patients who presented the pulmonary forms and
were AFB positive was 69.5%. This rate was higher
than the 54% seen in the 2001 Brazilian cohort (7)
but much lower than the 80.9% obtained by Diel
Sputum smear
negative
or not performed
101
13
22
8
0
144
Total
254
43
37
28
2
364
& Niemann in Hamburg, Germany(12) between 1997
and 2001 and the 77.2% obtained by El-Sony et
al. (13) in Sudan. The higher mortality observed
among patients presenting unconfirmed pulmonary
forms merits more in-depth study since these
patients would be expected to have better
prognoses. It is possible that, in the routine
practices of health services, other diseases have
been incorrectly treated as TB.
In the categorization of unsuccessful cases, the
unfavorable profile of TB patients with AIDS
comorbidity, characterized by high mortality
(18.9%) and noncompliance (15.3%), was partially
responsible for the lower success rate seen among
such patients. A study carried out in Sudan(13) did
not include critical patients but showed higher
mortality among new TB patients with comorbid
AIDS (12% among HIV-positive individuals and
1.8% among HIV-negative individuals).
A limitation to the present study was the
number of patients who were excluded due to the
lack of information regarding treatment outcomes.
The recording of information is not satisfactory,
and this may even worsen the statistics since
noncompliant cases would undoubtedly be
included among these patients about whom there
was no information. Another limitation was the
partial lack of HIV serology data, which may lead
to an underestimation of the rate of TB-AIDS
comorbidity. In view of the importance of
comorbidity data in devising appropriate treatment
strategies, attempts should be made to perform
HIV testing in 100% of the cases. The transfer
situation also deserves special care since the final
outcome for these patients (cure, death or
noncompliance) is unknown.
137
Oliveira, HB, et al.
Analysis of treatment outcomes related to the tuberculosis control
program in the city of Campinas, in the state of São Paulo, Brazil
The high rates of noncompliance, especially
among patients in retreatment, demand that health
clinics retool in order to achieve higher treatment
adherence(6), taking into consideration the fact that
compliance depends on several factors (6,14,15) .
According to Gonçalves et al., some individuals
may be led to believe that the disease is cured and
abandon the treatment when the symptoms
disappear. In addition, through association
mechanisms, some patients believe that taking the
medication would revive the disease and the
limitations imposed by it on their daily routine limitations that they would like to forget.
Interaction between physicians and patients has
also been identified as important. Patients who do
not trust the health system or the physician are
more likely not to adhere to the medication
regimen(16). It is important to bear in mind that
noncompliance is directly related not only to the
dissemination of the disease but also to the
development of multidrug-resistant strains(6). The
appropriate combination and supervised use of
correct doses for a sufficient length of time is the
best means of preventing bacterial persistence and
the development of drug resistance, thereby
ensuring the cure of the patient.
Health education(17) is one of the strategies used
to reduce the rate of noncompliance and,
consequently, the indicators cited. The importance
of organized and accessible health services in the
care of TB patients cannot be overemphasized.
During consultation and education activities,
personal patient issues must be taken into account,
bearing in mind that each individual is unique and
has his/her own history, values and beliefs.
Improving the effectiveness of the PCT in
Campinas and achieving the goal set by the World
Health Organization will require modifying not only
physician-patient interaction but also the
interaction of all members of the medical team
with the patient. In addition, health education
programs aimed at the patients and their families
should be implemented or refined. The teams
should perform sputum smear microscopy
whenever there is cough and expectoration. Finally,
in view of the lower success rate among TB patients
with AIDS comorbidity and in addition to the
considerations above, special care should be taken
to monitor interaction between the medical team
and these patients as well as to supervise treatment.
138
REFERENCES
1. Rouillon A, Perdrizet S, Parrot R, Waaler H. Métodos de
control de la tuberculosis. La transmissión del bacilo
tuberculoso. El efecto de la quimioterapia. WHO/
Tb,1977;346:1-30.
2. Grosset J. Bacteriological basis of chemotherapy of
tuberculosis. In: III Regional Seminar on Tuberculosis
Chemotherapy. Paho,1979;1-10.
3. Mitchison DA. The action of antituberculosis drugs in
short-course chemotherapy. Tubercle 1985;66:219-25.
4. Ministério da Saúde. Fundação Nacional de Saúde.
Controle da tuberculose: uma proposta de integração
ensino-serviço. FUNASA/CRPHF/SBPT,2002; 1-236.
5. World Health Organization. Treatment of tuberculosis:
Guidelines for national programmes. WHO/CDS/TB,
1997;1-78.
6. World Health Organization. Adherence to long-term
therapies:evidence for action. Tuberculosis. WHO,
2003;123-30
7. World Health Organization. WHO Report 2004. Global
Tuberculosis Control [cited 2004 may 02]. Available from:
www.who.int/tb/publications/global_report/2004/en/.
8. Khatri GR, Frieden TR. Controlling tuberculosis in India.
N Engl J Med 2002;347:1420-5.
9. Ruohonen RP, Goloubeva TM, Trnka L, Fomin MM,
Zhemkova GA, Sinitsyn AV, Lichachev AA, Koskela KG.
Implementation of the DOTS strategy for tuberculosis
in the Leningrad Region, Russian Federation (19981999). Int J Tuberc Lung Dis 2002;6:192-7.
10. Campos HMA, Albuquerque MFM, Campelo ARL, Souza
W, Brito AM. O retratamento da tuberculose no
município do Recife, 1997: uma abordagem
epidemiológica. J Pneumol 2000;26:235-40.
11. Oliveira HB, Moreira Filho DC. Abandono de tratamento
e recidiva da tuberculose: aspectos de episódios prévios,
Campinas,SP, Brasil, 1993-1994. Rev Saúde Pública
2000;34:437-43.
12. Diel R, Niemann S. Outcome of tuberculosis treatment
in Hamburg: a survey, 1997-2001. Int J Tuberc Lung
Dis 2003;7:124-31.
13. El Sony AL, Khamis AH, Enarson DA, Baraka O, Mustafa
SA, Bjuna G. Treatment results of DOTS in 1797
sudanese tuberculosis patients with or without HIV
co-infection. Int J Tuberc Lung Dis 2002;6:1058-66.
14. Costa JSD, Gonçalves H, Menezes AMB, Devens E, Piva
M, Gomes M, Vaz M. Controle epidemiológico da
tuberculose na cidade de Pelotas, Rio Grande do Sul,
Brasil: adesão ao tratamento. Cad. Saúde Pública
1998;14:409-15.
15. Sumartojo E. When tuberculosis treatment fails. A social
behavioral account of patient adherence. Am J Respir
Dis, 1992;68:49-59.
16. Gonçalves H, Costa JSD, Menezes AMB, Knauth D, Leal
OF. Adesão à terapêutica da tuberculose em Pelotas,
Rio Grande do Sul: na perspectiva do paciente. Cad.
Saúde Pública 1999;15:777-87.
17. Levy SN, Silva JJC, Cardoso IFR, Werberich PM, Moreira
LLS, Montiani H, Carneiro RM. Educação em Saúde:
Histórico, conceitos e propostos. Conferência Nacional
de Saúde. Ministério da Saúde. Diretoria de Programas
de Educação em Saúde. [cited 2004 May 02]. Available
from: http://www.datasus.gov.br/cns/datasus.htm.