Risk factors for pneumothorax in very low birth weight infants

To identify risk factors for pneumothorax in very low birth weight infants.

Retrospective case-control study.

Neonatal intensive care unit in a pediatric tertiary care center.

Very low birth weight infants.

All very low birth weight infants with pneumothorax born during the period January 1997 through December 2002 were identified. These infants were matched to infants without pneumothorax for gestational age, birth weight, and gender. Perinatal, neonatal, and treatment variables were collected for all infants.

Very low birth weight infants with pneumothorax were compared with those without. Univariate analysis was performed using the paired Student's t-test for continuous variables and the McNemar test for categorical variables. All variables were entered into a stepwise logistic regression model using a paired case-control design. Statistical significance was defined at p < .05. Seventy-four of 679 very low birth weight infants (10.9%) admitted to the neonatal intensive care unit developed a pneumothorax and were matched to 74 control infants. Multivariate analysis showed that only factors present on the day of pneumothorax were associated with pneumothorax. An increased risk of pneumothorax was associated with maximal, peak inspiratory pressure (odds ratio [OR] 1.33, 95% confidence interval [CI] 1.07, 1.66), minimal Fio2 (OR 2.18, 95% CI 1.14, 4.17), pulmonary hemorrhage (OR 27.5, 95% CI 2.3, 337), and maximal arterial CO2 (OR 1.94, 95% CI 1.13, 3.34), while a decreased risk was associated with maximal positive end-expiratory pressure (OR 0.71, 95% CI 0.56, 0.91).

Pneumothorax is associated with factors present on day of pneumothorax and not with initial ventilation variables or initial severity of lung disease. Decreasing the risk of pneumothorax requires rigorous control of ventilation, including optimizing positive end-expiratory pressure and minimizing peak inspiratory pressure.