This double blind, stratified, randomized placebo control trial was conducted between November 2007 and April 2009.
Eligible participants were patients aged 3 months to 15 yrs who required orotracheal or nasotracheal intubation and mechanical ventilation. Patients who had been mechanically ventilated for over 24 hrs prior to PICU admission, with tracheostomies, with inaccessible oral cavities, and with known hypersensitivity to chlorhexidine were excluded. We included patients with preexisting pneumonia as these patients constitute a large (almost 2/3 in our study group) proportion of our patient population, continue to remain susceptible to VAP, and contribute to the burden of hospital-acquired infections. The study was stopped from July 2008 to August 2008, as an independent institutional review of the methodology of all ongoing studies in the department and potential adverse events during the study period was being carried out. The review did not observe any significant adverse events and the study was continued after the review; no interim analysis of efficacy was done during these interim reviews.
The study took place at the medical PICU of the Department of Pediatrics, All India Institute of Medical Sciences, New Delhi, India, and the study protocol was approved by the institutional ethics committee. All India Institute of Medical Sciences is an academic tertiary care hospital with a six-bed PICU. The PICU services are provided by five physicians (providing 24-hr services), two to four nurses in each of the three shifts, and a technician available round the clock for equipment maintenance. Hand hygiene measures are implemented and sterile gowns are worn. Alcohol-based hand rub solutions (chlorhexidine free) were used for hand hygiene. Suctioning is performed with new suction catheters each time by open suctioning. We used heated wire humidifiers with reusable ventilator circuits. Reusable ventilator circuits were changed and sterilized at 48- to 72-hr intervals (typically these develop significant condensate in the expiratory limb and at times, the tubing gets soiled with secretions, as most of the patients that are admitted to PICU are with lower respiratory tract infections). Heat moisture exchangers for humidification and continuous subglottic suctioning are not available. Patients are routinely positioned in the semirecumbent position unless contraindicated for other reasons.
After obtaining written, informed consent from the parents, eligible patients were stratified into one of four groups—age <1 yr with no preexisting pneumonia, age <1 yr with preexisting pneumonia, age ≥1 yr with no preexisting pneumonia, and age ≥1 yr with preexisting pneumonia. Within each stratum, consecutive patients were randomized (1:1) to receive either 1% chlorhexidine gel (ICPA Health Products Ltd, Andheri, Mumbai, India) or placebo application and were allotted a numbered bag containing the assigned gel. The placebo applied was constituted of an inert base; this was the same as used in the chlorhexidine gel and was identical to the medicated gel with regards to appearance, consistency, taste, and smell. The random sequence was generated for each stratum using the STATA 9.0 program (Stata Inc., College Station, TX) in blocks of 6. The unlabelled tubes of placebo and 1% chlorhexidine were serially numbered and bagged for each patient according to the randomization sequence. Randomization and numbering of the tubes were done by personnel not involved in the study, and the allocation sequence remained concealed through the entire length of the study. The patients, physicians, outcome assessors, and data analysts were kept blinded to the interventions. The placebo or medicated gel was applied at 8-hr intervals with a standardized disposable applicator over the buccal mucosa. At each application 0.5 g (1.5 cm) of the gel was used. The oral cavity was first suctioned to remove any accumulated secretions, and a saline-soaked gauze was used to clean the mucosal surfaces subsequently. The allotted gel was then measured out (a total of 1.5 cm [0.5 g] with 0.75 cm being used for each side) on a standardized, disposable applicator and spread out on the right and left buccal mucosal surfaces of the child. Strict hand hygiene was used during the procedure. The nursing staff was trained on the method of application in order to ensure uniformity. The period of application was from the day of intubation till extubation, for a maximum period of 3 wks. Baseline data collected included age, sex, underlying diagnosis and its severity as assessed by the Pediatric Index of Mortality 2 score, presence of pneumonia at enrollment, presence of other infections at enrollment, antibiotics used at enrollment, chest radiograph, and bronchoalveolar lavage (BAL) cultures collected on days 1 and 3. An admission diagnosis of pneumonia was based on clinico-radiologic features. In addition, BAL samples for cultures were obtained on days 1 and 3 of ventilation. As VAP was defined as pneumonia occurring >48 hrs after endotracheal intubation and mechanical ventilation, this was the time frame after which a change in the clinical and radiological findings was used to suspect VAP, and the presence of a different microbial growth on BAL cultures was used to confirm the diagnosis.
The primary outcome studied was the incidence rate of VAP in each group. The secondary outcomes analyzed were duration of ICU stay, duration of hospital stay, in hospital mortality rate, and type and antibiotic sensitivity of organisms cultured. The patients were followed up for a period of 3 wks or till discharge, whichever was earlier. VAP was defined as pneumonia occurring >48 hrs after endotracheal intubation and mechanical ventilation. Children who were extubated were followed up till 48 hrs after extubation. The age appropriate diagnostic criteria given by the Centers for Disease Control and Prevention was used to diagnose VAP in patients with no evidence of preexisting pneumonia. In patients with underlying pneumonia, worsening of the clinical and radiological findings given by the Centers for Disease Control and Prevention was used to suspect VAP. The BAL samples cultured on days 1 and 3 were considered as the baseline. Any subsequent change in BAL flora was then used to diagnose a case of VAP. BAL samples were collected using the protected nonbronchoscopic mini BAL technique. Semi-quantitative cultures were used, and the presence of >10,000 colony forming units/mL was deemed significant.
Based on the estimated prevalence of VAP in our PICU of 40% (unpublished data) and an expected 50% fall in the rates of VAP, for an 80% power of the study and 5% α error, the calculated sample size was found to be 91 in each group. The trial was designed to be time driven, and the final analysis was scheduled in April 2009. The final analysis was performed as planned with the analyst remaining blinded to the intervention, and since no trend toward a significant difference between the two interventions was found in any of the analyzed parameters, further enrollment was stopped in April 2009.
The data analysis was performed using STATA 9.0 software, on an intention to treat basis. The baseline variables were compared, and the primary and secondary outcomes were analyzed in two groups—chlorhexidine group and placebo group. Chi-square or Fisher's exact test were used for categorical variables, and Wilcoxon rank-sum test was used for continuous variables to evaluate statistical significance.