Background The aim of this study was to determine whether patient characteristics were influencing clinicians’ decision making for “step 3” treatment options.

Methods Data between 1990–2011 were collected from the Clinical Practice Research Datalink and Optimum Patient Care Research Database. Children aged 5–12 years with asthma, on inhaled corticosteroids (ICS) ±reliever therapy only were stratified by the first step up option: increase ICS; add long acting beta agonist as a free combination (+LABA); adding LABA as a fixed dose combination (FDC); add leukotriene receptor antagonist (+LTRA). Exacerbation was defined as need for oral steroids and/or A&E attendance. Asthma control was determined from an algorithm including reliever medication use.

Results There were 14,566 increases to step 3 including 9,393 ICS increases (64%), 2934 +LABAs (20%), 1119 FDCs (8%) and 1120 +LTRAs (8%). There were subtle differences between groups. ICS increase was associated with lower doses of ICS pre step up and better asthma control. FDC step up was associated with older age, increased short acting beta agonist (SABA) use and exacerbations. LABA+ was associated with higher doses of ICS. There was evidence of changing practice where step up with FDC and LTRA+ became more commonplace after 2000.

Conclusions BTS/SIGN guidelines recommends +LABA as first-choice “step 3” treatment since 2005 but increasing ICS is the most common option chosen. A child’s age, current dose of ICS and year of step up may influence a clinician’s choice of step 3 therapy. Ongoing work will determine which step 3 option (if any) leads to better asthma outcome.