Soslow Jonathan H, Hall Matthew, Burnette W Bryan, Hor Kan, Chisolm Joanne, Spurney Christopher, Godown Justin, Xu Meng, Slaughter James C, Markham Larry W
1Thomas P Graham Division of Pediatric Cardiology,Department of Pediatrics,Vanderbilt University Medical Center,Nashville,TN,USA.
2Children's Hospital Association,Lenexa,KS,USA.
Cardiol Young. 2019 Mar;29(3):290-296. doi: 10.1017/S1047951118002226. Epub 2019 Jan 26.
Outcome analyses in large administrative databases are ideal for rare diseases such as Becker and Duchenne muscular dystrophy. Unfortunately, Becker and Duchenne do not yet have specific International Classification of Disease-9/-10 codes. We hypothesised that an algorithm could accurately identify these patients within administrative data and improve assessment of cardiovascular morbidity.
Hospital discharges (n=13,189) for patients with muscular dystrophy classified by International Classification of Disease-9 code: 359.1 were identified from the Pediatric Health Information System database. An identification algorithm was created and then validated at three institutions. Multi-variable generalised linear mixed-effects models were used to estimate the associations of length of stay, hospitalisation cost, and 14-day readmission with age, encounter severity, and respiratory disease accounting for clustering within the hospital.
The identification algorithm improved identification of patients with Becker and Duchenne from 55% (code 359.1 alone) to 77%. On bi-variate analysis, left ventricular dysfunction and arrhythmia were associated with increased cost of hospitalisation, length of stay, and mortality (p<0.001). After adjustment, Becker and Duchenne patients with left ventricular dysfunction and arrhythmia had increased length of stay with rate ratio 1.4 and 1.2 (p<0.001 and p=0.004) and increased cost of hospitalization with rate ratio 1.4 and 1.4 (both p<0.001).
Our algorithm accurately identifies patients with Becker and Duchenne and can be used for future analysis of administrative data. Our analysis demonstrates the significant effects of cardiovascular disease on length of stay and hospitalisation cost in patients with Becker and Duchenne. Better recognition of the contribution of cardiovascular disease during hospitalisation with earlier more intensive evaluation and therapy may help improve outcomes in this patient population.
大型管理数据库中的结果分析对于贝克尔型和杜氏型肌营养不良等罕见疾病来说是理想的。不幸的是，贝克尔型和杜氏型肌营养不良尚未有特定的国际疾病分类第9版/第10版编码。我们假设一种算法能够在管理数据中准确识别这些患者，并改善对心血管疾病发病率的评估。
从儿科健康信息系统数据库中识别出根据国际疾病分类第9版编码359.1分类的肌营养不良患者的医院出院记录（n = 13,189）。创建了一种识别算法，然后在三个机构进行验证。使用多变量广义线性混合效应模型来估计住院时间、住院费用和14天再入院与年龄、就诊严重程度以及呼吸系统疾病之间的关联，并考虑医院内部的聚类情况。
识别算法将贝克尔型和杜氏型患者的识别率从55%（仅编码359.1）提高到了77%。在双变量分析中，左心室功能障碍和心律失常与住院费用增加、住院时间延长和死亡率增加相关（p<0.001）。调整后，患有左心室功能障碍和心律失常的贝克尔型和杜氏型患者住院时间延长，率比分别为1.4和1.2（p<0.001和p = 0.004），住院费用增加，率比分别为1.4和1.4（均p<0.001）。
我们的算法能够准确识别贝克尔型和杜氏型患者，可用于未来对管理数据的分析。我们的分析表明心血管疾病对贝克尔型和杜氏型患者的住院时间和住院费用有显著影响。在住院期间更好地认识心血管疾病的作用，进行更早、更强化的评估和治疗，可能有助于改善这一患者群体的预后。
