Bridging the Gap: Conference Record [Abstract book, International Conference on AIDS (12th: 1998: Geneva, Switzerland)]

12th World AIDS Conference

Abstracts 22135-22139 289

22135 Tuberculosis (TB) in HIV-infected persons at an urban US
Medical Center, 1991-1996
Robert Horsburgh, Jr.', M.J. Sotir1, J.L. Lennox1, P.P. Moore2, H.M. Blumberg1.
'Emory University School of Medicine 69 Butler Street Atlanta, GA 30303;
'Grady Memorial Hospital Atlanta, GA, USA
Objectives: To describe the epidemic of TB in HIV-infected persons at an urban US medical center and to identify potential strategies for TB prevention in
HIV-infected persons.
Design: Case series.
Methods: All patients with TB diagnosed between 1991 and 1996 at Grady
Memorial Hospital (GMH) in Atlanta GA, USA, were examined. Demographic
information, CD4 counts, date of first HIV test and date of enrollment in the HIV
outpatient clinic were collected from hospital records.
Results: 1203 patients with TB were identified during the study period. Of these,
493 (41%) were HIV-infected. The annual number of admissions declined from
255 in 1990 to 175 in 1996, coincident with expanded use of directly observed
therapy, prevention of nosocomial transmission and increased clinician/public
health cooperation. The percentage of HIV+ patients did not change significantly
over time. 380 (77%) of HIV+ TB patients were male; 440 (89%) were black.
Median age was 35 years and median CD4 count was 73 cells/mm3. None of
these variables changed significantly over time. During the most recent two-year
period, 118 patients had 124 admissions. Of these, 48 (41%) had the diagnosis
of HIV made concurrently with the diagnosis of active TB. Of the 70 patients with
HIV status known before TB, only 14 (20%) had attended the HIV clinic before
the diagnosis of TB was made.
Conclusions: Many HIV/TB patients did not know their HIV status before
developing active TB, and a substantial additional number did not obtain care
for HIV disease that could have reduced their risk of acquiring TB, such as INH
prophylaxis and antiretroviral therapy. Increased HIV screening and incentives to
attend the HIV clinic could reduce morbidity and mortality of TB in HIV-infected
persons.
22136 Tuberculosis and AIDS in prison
Horacio Jauregui Rueda'. M. Di Lonardo2, R. Marino3, M. Ambroggi2,
A. Monticelli3, E. Santucho3, J.J. Alvarez Casado3. 1Junin 969 90 B 1113
Buenos Aires: 2'Laboratorio Micobacterias Hosp. Muhiz, Buenos Aires;
3Unidad21-Servicio Penitenciario Federal, Buenos Aires, Argentina
Background: since AIDS appeared the injection drug user (IDU) and tuberculosis
(TB) have been the risk behavior and the most frequent marker disease observed
in the UNIT 21 - Federal Penitenciary Service.
Having in consideration a prevalence survey done in Jan. 1992, we observed
that HIV positive patients (p) that were in hospital had antecedents of TB (77%),
active TB (63%) and disseminated forms (50%).
Objectives: 1) To evaluate according to prevalence surveys: a) Risk behavior
(RB) associated to TB. b) TB like a previous marker disease (PMD), initial marker
disease (IMD), and actual marker disease (AMD). c) Disseminated forms (DF).
2) Incidence rate of multidrug-resistant (MR) TB.
Methods: two prevalence surveys in UNIT 21 Jan. 97 (22p) and Jan. 98 (22p).
The diagnosis of TB was made by: a) Baciloscopy and culture. b) Compatible
pathological anatomy. c) Clinical manifestation attributable to TB and with irrefutable improvement related only to the specific treatment. Sensibility study to
antituberculosis drugs: proportions' methods by Canetti, Rist and Grosset in its
economic variate.
Results:

(AIP) for tuberculosis (TB) implemented in a Hospital with a prevalent HIV+
population; 2) to construct a hypothetical model of isolation policy (HIP) based on
clinical predictors.
Design: 2-years prospective study on HIV+ patients (pts) with fever and
or respiratory symptoms. According to AlP all pts with respiratory specimens
AFB-smear+, chest x-ray with cavities or hilar/mediastinal adenopathy, culture+
for mycobacteria, were placed in a respiratory isolation room (RIR). Engineering
was not implemented. Case-patient was defined as culture+ TB. Pts were appropriately isolated if placed in RIR within 24 hours from admission. Outcomes
were tuberculosis exposure episodes (TEE), days tuberculosis exposure (DTE),
and RFLP analysis (similarity coefficient SAB at digitized dendrogram) combined
with epidemiological investigation. Clinical predictors for HIP were selected after
testing for interaction in a log-linear model and exploring by multiple logistic
regression.
Results: 514 HIV+ pts were enrolled, 51 with TB (24 smear+) and 463
controls. Screening of clinical predictors revealed that foreign-born (OR 2.99:
95%CI 1.09-8.27), a not previous AIDS (OR 2.06; 95%CI 1.08-3.94); and a
hilar/mediastinal adenopathy (OR 13.73; 95%CI 6.80--27.80) was predictive of
TB and was employed to design the HIP. Sensitivity and positive predictive value
(PPV) were 64.7% and 39.8 for AlP, 82.4% and 15.8% for HIP. Sensitivity and
PPV for detecting AFB-smear+ TB was 75.0% and 27.3% for AlP, 95.8% and
9.4% for HIP. With AlP 41/51 (80%) of cases were isolated, 33 appropriately,
with 18 TEE (0.75 per month) and 260 DTE (10.83). With HIP only 9 TEE (0.37)
and 157 DTE (4.29) would have been obtained. RFLP showed a high grade of
polymorphism (SAB 0.2-0.5 for most isolates) with 4 small clusters (homology
>90%). Epidemiologic investigation did not revealed linking among patients with
clustered isolates related to nosocomial transmission.
Conclusions: A selective isolation policy for HIV+ assures low rates of TEE and
limiting nosocomial transmission. HIP model tested was sensitive and cost-effective for hospitals with high HIV prevalence. Supported by National Tuberculosis
Project; Istituto Superiore di Sanita - Ministero della Sanita, Roma, Italy.
122138 Prediction of mycobacterium tuberculosis infection in
HIV-1 infected patients with prolonged fever
Raquel Barba, J. Gomez-Rodrigo, J. Marco, G. Eroles, P. Rondon, S. Plaza,
M. Lopez-Varas. Hospital Severo Ochoa, Madrid, Spain
Aims and Methods: HIV-1 infected patients meeting criteria for in-hospital FUO
were included in a prospective study where demographic, analytical, clinical,
microbiological, radiological and immunological data were systematically gazed.
The case end point was either death or a definite diagnosis. At the end of the
study these parameters were analysed in order to identify predictive factors of
Mycobacterium tuberculosis infection. Study entry period: from February 1995 to
December 1996.
Results: 54 patients were included; 22 with the final diagnosis of Mycobacterium tuberculosis infection, 27 with other diagnosis and 5 who died before
diagnosis. After univariant analysis, the parameters associated with the diagnosis
of M. Tuberculosis with a level of significance of 0.05 were as follows: absence
of abdominal pain, presence of crackles or wheezing, a positive PPD test prior
or during the actual clinical course, a lower plasma triglyceride level and higher
counts of CD4, leukocytes, platelets and haemoglobin level. A model of logistic
regression using these significant variables was performed. With this multivariant
model, Mycobacterium tuberculosis infection was significantly associated with
the previous PPD positive status, lower triglyceride levels, higher CD4 count and
absence of abdominal pain at the beginning of the process. Application of this
logistic regression equation to the 54 patients in our series achieved a correct
diagnosis of Mycobacterium tuberculosis infection Vs other diagnosis in 90.57%
of cases.
Conclusions: 1) It is possible to predict the diagnosis of TB infection with
a high level of confidence in cases of in-hospital FUO in HIV infected patients
through easily obtainable parameters.2) All patients meeting these criteria should
receive antituberculous treatment while awaiting a final diagnosis.
22139 Evaluation of tuberculosis (TB) cases in AIDS patients in
an out-patient clinic in Belo Horizonte - June 1995 to July
1996
Unai Tupinambas1, D.B. Greco2, R.M. Ferraz2, F.A. Ribeiro2, F.G.F. Ferreira2,
L.M. Diniz2, S.S. Moll2. 1Dona Cecilia St 55-501 Serra Belo Horizonte
30220-070 MG; 2Federal University of Minas Gerais Belllo Horizonte, Brazil
Objective: To evaluate the diagnostic criteria for TB, its clinical presentation, tuberculin skin test, HIV transmission category, and the compliance with treatment
in AIDS patients followed from june 95 to july 96
Methods: 124 records of patient previously diagnosed with TB were evaluated. The data was taken from tuberculosis notification forms and from the first
appointment files.
Results: 124 tuberculosis cases were studied in this period of time. 75% of
the patient were between 20 and 40 years, only 4.9% were older than 50 and

5 patient (4.1%) were younger than 15. There were 101 (81.5%) male and 23
(18.5%) female patients. Sexual transmission was responsible for 63.7% (79) of
the HIV infections (25.8% homosexual, 30.6% heterosexual and 7.3% bisexual).
19.4% of the patients did not relate any risk factors for HIV transmission. As
for the diagnostic criteria for tuberculosis, clinical diagnostic (clinical findings,
chest radiograph) was reponsible for 56.4% (70) of the diagnosis, followed by
histopathology (presence of caseous necrosis and/or acid fast bacili) 17.7% (22);

Prevalence
RB
PMD
AMD
IMD
DF

Surveys
January 1997
IDU 85~0
64~0o
40%~
4000
50so

January 1998
IDU 89%0
390%
44~0
560o
43~0

Incidence rate
1996
Patients 35
Resistant 2 (6%)
MR 1 (3%)

1997
38
11 (29%)
5 (13%)

Conclusion: The combination and the high prevalence of IDU and TB in
prison plus the irregular adherence to the therapy and the concomitant growth
of MR organisms makes imperative to improve the programs of information and
prevention so as to make an early diagnosis and supervision of the treatment.
22137 An isolation respiratory policy for HIV-infected patients
with tuberculosis: analysis of clinical predictors and
molecular epidemiologic assessment
Andrea Antinorii, A. Cingolani2, G. Scoppettuolo2, M. Sanguinetti3,
B. Posteraro:, G. Fadda3, L. Ortona2. Department of Infectious Diseases
Catholic University-L. Go A. Gemelli-8-Rome; 2Depart. Infect. Dis.-Catholic
University Rome;, Depart. Microbiology-Catholic University Rome, Italy
Objective: 1) to validate the accuracy of an actual respiratory isolation policy