Predictive indexes of ventilatory weaning in people living with HIV/AIDS

Abstract

Mechanical ventilation (MV) is a form of life support which, once it is instituted, should
be removed as soon as possible. In order to do that, physiological indexes have been
used to predict a successful extubation. Our objective was to evaluate the predictive rates
of weaning from ventilation such as MIP (maximum inspiratory pressure), RSBI (rapid
shallow breathing index) and oxygenation through the PaO2/FiO2 ratio in PLHA under
orotracheal intubation (OTI). A prospective study was carried out at the ICU of the Emilio
Ribas Institute of Infectious Diseases with 75 patients of both genders, older than 18
years old, submitted to MV for a period longer than 24 hours. The sample was divided
into two groups: 50 HIV- positive patients (HIV group-HIVG) and 25 non-HIV patients
(control group-CG). Both groups met the criteria for weaning from ventilation in the
institutional protocol and were submitted to spontaneous breathing trials (SBT), and the
following indexes were calculated, namely, the PaO2/FiO2 ratio, MIP and RSBI. Patients
who presented predictive weaning rates recommended in the literature and who did not
show signs of respiratory distress in the SBT were extubated and observed for 48 hours,
in order to predict success and failure in the weaning from ventilation. In the HIVG, 42%
(21 of the patients) presented extubation failure versus. only 8% (2 of the patients) of the
CG (p=0.003); HIVG patients presented higher severity at the ICU admission with a mean
SAPS III of 63.28 versus. 51.36 in the CG. (p≤0.001); the HIVG remained hospitalized
in the ICU for longer with a mean of 20.6 days versus. 13.8 of the CG (p=0.004); 68%
of the HIVG patients were malnourished versus 16% of the CG. The main reason of the
HIVG reintubation was acute respiratory failure with 24% (12 of the cases) and in the
CG, 100% (2 of the cases), was the decrease in the level of consciousness. Regarding the
predictive indexes, only the PaO2/FiO2 ratio presented statistical significance (p=0.045),
being higher in the HIVG (mean=344.78) than in the CG (mean=305.08). Significant
correlations were found between the VL (viral load) X MIP (p=0.003), and CD4 X MIP
(p=0.044), and the period of MV X RSBI (p=0.0049). In relation to mortality, 20% (10)
of the total HIVG sample died, whereas the CG did not present any deaths. We conclude
that the predictive rates of weaning from ventilation were not sufficient to differentiate the
outcomes of success and failure in the present population.