To estimate transition probabilities between mild, moderate and severe asthma states based on asthma drug-use and identify determinants of these transition probabilities.

Retrospective data were used to identify all patients in the three major asthma states based on therapy as recommended by the NHLBI guidelines. Monthly transition probabilities between the three asthma states were then calculated based on ER visits, medication step up or down, rescue medication use and hospitalizations. A logistic regression was formulated that calculated the odd ratio of transitions between asthma states as function of predetermined exacerbation/non-exacerbation variables.

Calculated odd ratios showed that transitions between asthma states were significantly affected by seasonality with worse exacerbations in winter and spring months. Transitions also responded significantly to the type of treating physician (PCP, Allergist, or Pulmonologist), age, gender, allergic rhinitis status, short acting beta2-agonist, anti-histamine, Immunotherapy and anti-acid use as well as to preceding upper respiratory infection, and history of intubation.

Monthly transitions between asthma states calculated from retrospective data were shown to significantly respond to well known clinical practice parameters.