Department of Emergency Medicine and Critical Care, Japanese Red Cross Society Kyoto Daini Hospital, 602-8026, Haruobicho 355-5, Kamigyo, Kyoto, Japan.
Department of Obstetrics and Gynecology, Japanese Red Cross Society Kyoto Daini Hospital, Kyoto, Japan.
BMC Emerg Med. 2020 May 6;20(1):35. doi: 10.1186/s12873-020-00332-z.
Obstetric and gynecological (OBGY) diseases are among the most important differential diagnoses for young women with acute abdominal pain. However, there are few established clinical prediction rules for screening OBGY diseases in emergency departments (EDs). This study aimed to develop a prediction model for diagnosing OBGY diseases in the ED.
This single-center retrospective cohort study included female patients with acute abdominal pain who presented to our ED. We developed a logistic regression model for predicting OBGY diseases and assessed its diagnostic ability. This study included young female patients aged between 16 and 49 years who had abdominal pain and were examined at the ED between April 2017 and March 2018. Trauma patients and patients who were referred from other hospitals or from the OBGY department of our hospital were excluded.
Out of 27,991 patients, 740 were included. Sixty-five patients were diagnosed with OBGY diseases (8.8%). The "POP" scoring system (past history of OBGY diseases + 1, no other symptoms + 1, and peritoneal irritation signs + 1) was developed. Cut-off values set between 0 and 1 points, sensitivity at 0.97, specificity at 0.39, and negative likelihood ratio (LR-) of 0.1 (95% CI: 0.02-0.31) were considered to rule-out, while cut-off values set between 2 and 3 points, sensitivity at 0.23 (95% CI 0.13-0.33), specificity at 0.99 (95% CI 0.98-1.00), and positive likelihood ratio (LR+) of 17.30 (95% CI: 7.88-37.99) were considered to rule-in.
Our "POP" scoring system may be useful for screening OBGY diseases in the ED. Further research is necessary to assess the predictive performance and external validity of different data sets.
妇产科疾病是年轻女性急性腹痛最重要的鉴别诊断之一。然而，在急诊科（ED）中，用于筛查妇产科疾病的临床预测规则很少。本研究旨在建立一种用于诊断 ED 中妇产科疾病的预测模型。
这是一项单中心回顾性队列研究，纳入了我院 ED 就诊的急性腹痛女性患者。我们建立了预测妇产科疾病的逻辑回归模型，并评估了其诊断能力。本研究纳入了年龄在 16 至 49 岁之间、有腹痛且在 2017 年 4 月至 2018 年 3 月在 ED 接受检查的年轻女性患者。创伤患者和从其他医院或我院妇产科转来的患者被排除在外。
在 27991 名患者中，纳入了 740 名患者。65 名患者被诊断为妇产科疾病（8.8%）。开发了“POP”评分系统（过去妇产科病史+1、无其他症状+1、腹膜刺激征+1）。将 0 至 1 分的切点值设为阴性预测值（NPV）为 0.97、特异性为 0.39、阴性似然比（LR-）为 0.1（95%CI：0.02-0.31），将 2 至 3 分的切点值设为阳性预测值（PPV）为 0.23（95%CI 0.13-0.33）、特异性为 0.99（95%CI 0.98-1.00）、阳性似然比（LR+）为 17.30（95%CI：7.88-37.99），以排除或纳入妇产科疾病。
我们的“POP”评分系统可能有助于在 ED 中筛查妇产科疾病。需要进一步的研究来评估不同数据集的预测性能和外部有效性。
