Department of Plastic & Reconstructive Surgery UZ Gent, Corneel Heymanslaan 10, Gent, Belgium.
Crunch Analytics, Rodelijvekensstraat 28 bus 002, Gent, Belgium.
Sci Rep. 2019 Sep 9;9(1):12899. doi: 10.1038/s41598-019-49125-w.
With every hospital admission, a vast amount of data is collected from every patient. Big data can help in data mining and processing of this volume of data. The goal of this study is to investigate the potential of big data analyses by analyzing clinically relevant data from the immediate postoperative phase using big data mining techniques. A second aim is to understand the importance of different postoperative parameters. We analyzed all data generated during the admission of 739 women undergoing a free DIEAP flap breast reconstruction. The patients' complete midcare nursing report, laboratory data, operative reports and drug schedule were examined (7,405,359 data points). The duration of anesthesia does not predict the need for revision. Low Red Blood cell Counts (3.53 × 10/µL versus 3.79 × 10/µL, p < 0.001) and a low MAP (MAP = 73.37 versus 76.62; p < 0.001) postoperatively are correlated with significantly more revisions. Different drugs (asthma/COPD medication, Butyrophenones) can also play a significant role in the success of the free flap. In a world that is becoming more data driven, there is a clear need for electronic medical records which are easy to use for the practitioner, nursing staff, and the researcher. Very large datasets can be used, and big data analysis allows a relatively easy and fast interpretation all this information.
每一次住院，都会从每位患者身上收集大量数据。大数据可以帮助进行数据挖掘和处理这些数据量。本研究的目的是通过使用大数据挖掘技术分析术后即刻的临床相关数据来研究大数据分析的潜力。第二个目的是了解不同术后参数的重要性。我们分析了 739 名接受游离 DIEAP 皮瓣乳房再造的女性住院期间产生的所有数据。检查了患者的完整中期护理报告、实验室数据、手术报告和药物计划（7,405,359 个数据点）。麻醉持续时间并不能预测是否需要修复。低红细胞计数（3.53×10/µL 与 3.79×10/µL，p<0.001）和术后低 MAP（MAP=73.37 与 76.62；p<0.001）与更多修复明显相关。不同的药物（哮喘/COPD 药物、丁酰苯类）也可能在游离皮瓣的成功中发挥重要作用。在一个日益数据驱动的世界中，电子病历显然需要易于医生、护理人员和研究人员使用。可以使用非常大的数据集，并且大数据分析允许相对容易和快速地解释所有这些信息。
