Chen Xin, Wang Yu, Schoenfeld Elinor, Saltz Mary, Saltz Joel, Wang Fusheng
Stony Brook University, Stony Brook, NY.
AMIA Jt Summits Transl Sci Proc. 2017 Jul 26;2017:483-492. eCollection 2017.
Increased accessibility of health data provides unique opportunities to discover spatio-temporal patterns of diseases. For example, New York State SPARCS (Statewide Planning and Research Cooperative System) data collects patient level detail on patient demographics, diagnoses, services, and charges for each hospital inpatient stay and outpatient visit. Such data also provides home addresses for each patient. This paper presents our preliminary work on spatial, temporal, and spatial-temporal analysis of disease patterns for New York State using SPARCS data. We analyzed spatial distribution patterns of typical diseases at ZIP code level. We performed temporal analysis of common diseases based on 12 years' historical data. We then compared the spatial variations for diseases with different levels of clustering tendency, and studied the evolution history of such spatial patterns. Case studies based on asthma demonstrated that the discovered spatial clusters are consistent with prior studies. We visualized our spatial-temporal patterns as animations through videos.
健康数据可及性的提高为发现疾病的时空模式提供了独特的机会。例如，纽约州SPARCS（全州规划与研究合作系统）数据收集了每位住院患者住院和门诊就诊的患者人口统计学、诊断、服务及费用等患者层面的详细信息。这些数据还提供了每位患者的家庭住址。本文展示了我们利用SPARCS数据对纽约州疾病模式进行空间、时间和时空分析的初步工作。我们在邮政编码层面分析了典型疾病的空间分布模式。我们基于12年的历史数据对常见疾病进行了时间分析。然后，我们比较了具有不同聚集倾向水平的疾病的空间变化，并研究了此类空间模式的演变历史。基于哮喘的案例研究表明，所发现的空间集群与先前的研究一致。我们通过视频将时空模式可视化为动画。
