The Value of Registries, As Demonstrated by the Scotland Childhood Asthma Study
Posted Sep 15 2010 6:59pm
Electronic health records? Bleh, not ready, says the Disease Management Care Blog. Clinical decision support? Pfft, overrated, scoffs the DMCB. But mention registries and now you've got the DMCB's lusty attention. These medical-claims-demographic databases promise to offer up easy access to huge swaths of clinical and economic correlations that are just not possible in a single clinical trial.
Case in point? The Smoke-free Legislation and Hospitalizations for Childhood Asthma study authored by Daniel MacKay and colleagues appearing in the Sept. 16 New England Journal. These researchers had access to the "Scottish Morbidity Record," a national data warehouse that links information on all individual admissions among the 5.1 million inhabitants of Scotland. They used this data base to examine the impact of the "Smoking, Health and Social Care Act of Scotland," which banned public and workplace tobacco smoking. Before the legislation went into effect on March 26, 2006, hospital admissions among children age 15 years or less for asthma was increasing at a rate of 5.2% per year. After it was passed, the asthma admission rate decreased by 13% per year. The authors went on to mathematically neutralize (statistically control for) the impact of gender, age, urban vs. rural residence, socioeconomic status and number of prior admissions for asthma and the pre-post relationship still held up: there was a 4.4% increase per year before vs. a 20% decrease per year after. If the child population was split into pre-school and school age groups, it was found that the preschoolers were experiencing an annual increase while the school age children were not. When the smoking ban went into effect: afterwards, both groups dropped by 18% and 21%, respectively.
This study is important because it demonstrates how public health measures to control tobacco use in public places has an actionable link to the scourge of childhood asthma. This is probably because the smoking ban was accompanied by decreased tobacco use in the home, perhaps from greater adult quit-rates or voluntary restrictions.
Yet, this study is also important because it's another demonstration of the use of large clinical databases to study population health. This contrasts with the usual gold-standard clinical research approach, involving a two-armed randomized control trial comparing an intervention group (not allowed to use tobacco in public) vs. a control group. If MacKay et al had decided to answer the question that way, we'd still be waiting for an answer. Good thing they and others are demonstrating the value of this alternate approach.