Generating a Comprehensive Disease Pattern of Obstructive Sleep Apnea: A Data Mining Approach

Huang Qi,Zhen Qin,Shichao Zhang,Chin Moi Chow
2007-01-01
Abstract:Obstructive sleep apnea (OSA) is a sleep disorder with a wide range of comorbid conditions. Despite its known aetiology and manifestations, the strength of its association with comorbid conditions has not been fully explored. This research explored a large dataset (1999-2004) derived from the New South Wales Inpatient Data Collection, using a data mining approach, to provide a comprehensive disease pattern of OSA. All diagnoses were coded with the International Classification of Diseases (ICD)-Australian Modification codes. A total of 60,197 OSA cases (4% of total records) were identified. Their onset of occurrence was age-related with a peak occurrence at age 50-59y for both genders with a higher incidence in males. Comorbid conditions, in order of strength of association with OSA, were essential hypertension, obesity, hyper cholesterolemia, type 2 diabetes, and ischemic heart disease. When categorised according to disease systems, all cormorbid conditions clustered mainly into endocrine/ metabolic, cardiovascular and respiratory systems. The data mining approach unveiled the full OSA disease pattern and illustrated the progression of comorbid conditions.
What problem does this paper attempt to address?