Analysis of spirometry benefits in children with asthma using propensity score matching

OBJECTIVES: To assess the benefits of spirometry for children with asthma using propensity score method. METHODS: A retrospective cohort study of children with asthma aged 5-18 was conducted with 2005-2007 MarketScan data. The main outcome was the number of hospitalizations or ER visits for asthma (acute asthma events) in 2007. We used propensity score matching (PSM) to control for confounding between the groups with and without spirometry tests. Four PSM strategies were used: one nearest neighbor (0.02), five nearest neighbors (0.02), local regression and kernel matching (0.8). An exposed group was defined as children who had at least one spirometry test in 2005-2006. A separate unexposed group without spirometry was generated based on one of four PSM strategies. RESULTS: Children with spirometry had similar demographic, geographic, health status and plan type characteristics to those without; however children with a spirometry test had more asthma medications, specialist visits and acute asthma events (AAEs) compared to children without spirometry. The average number of AAEs in 2007 in the exposed group was 0.105, while that number was 0.117, 0.119, 0.115 and 0.112 in the unexposed groups, as previously defined. Thus without adjustment children with spirometry had about 25% more AAEs than did children without spirometry. After PSM adjustment having spirometry was associated with 6-11% fewer AAEs depending on the PSM strategies used, suggesting that this correction is likely due to PSM adjustments for possible confounding effect. CONCLUSIONS: Children with asthma who had spirometry tests had significantly fewer AAEs, suggesting important benefits of spirometry in pediatric care.