Department of Acute Internal Medicine, University Medical Center Utrecht, Utrecht, The Netherlands.
Department of Anesthesiology, University Medical Center Utrecht, Utrecht, The Netherlands.
PLoS One. 2023 May 25;18(5):e0286080. doi: 10.1371/journal.pone.0286080. eCollection 2023.
Continuous monitoring of vital signs is introduced at general hospital wards to detect patient deterioration. Interpretation and response currently rely on experience and expert opinion. This study aims to determine whether consensus exist among hospital professionals regarding the interpretation of vital signs of COVID-19 patients. In addition, we assessed the ability to recognise respiratory insufficiency and evaluated the interpretation process.
We performed a mixed methods study including 24 hospital professionals (6 nurses, 6 junior physicians, 6 internal medicine specialists, 6 ICU nurses). Each participant was presented with 20 cases of COVID-19 patients, including 4 or 8 hours of continuously measured vital signs data. Participants estimated the patient's situation ('improving', 'stable', or 'deteriorating') and the possibility of developing respiratory insufficiency. Subsequently, a semi-structured interview was held focussing on the interpretation process. Consensus was assessed using Krippendorff's alpha. For the estimation of respiratory insufficiency, we calculated the mean positive/negative predictive value. Interviews were analysed using inductive thematic analysis.
We found no consensus regarding the patient's situation (α 0.41, 95%CI 0.29-0.52). The mean positive predictive value for respiratory insufficiency was high (0.91, 95%CI 0.86-0.97), but the negative predictive value was 0.66 (95%CI 0.44-0.88). In the interviews, two themes regarding the interpretation process emerged. "Interpretation of deviations" included the strategies participants use to determine stability, focused on finding deviations in data. "Inability to see the patient" entailed the need of hospital professionals to perform a patient evaluation when estimating a patient's situation.
The interpretation of continuously measured vital signs by hospital professionals, and recognition of respiratory insufficiency using these data, is variable, which might be the result of different interpretation strategies, uncertainty regarding deviations, and not being able to see the patient. Protocols and training could help to uniform interpretation, but decision support systems might be necessary to find signs of deterioration that might otherwise go unnoticed.
连续监测生命体征被引入综合医院病房，以检测患者病情恶化。目前的解释和反应依赖于经验和专家意见。本研究旨在确定医院专业人员在解释 COVID-19 患者生命体征方面是否存在共识。此外，我们评估了识别呼吸功能不全的能力，并评估了解释过程。
我们进行了一项混合方法研究，包括 24 名医院专业人员（6 名护士、6 名初级医生、6 名内科专家、6 名 ICU 护士）。每位参与者都收到了 20 例 COVID-19 患者的病例，包括 4 或 8 小时的连续测量生命体征数据。参与者估计患者的情况（“改善”、“稳定”或“恶化”）和发展呼吸功能不全的可能性。随后，进行了一次半结构化访谈，重点关注解释过程。使用 Krippendorff 的 alpha 评估共识。对于呼吸功能不全的估计，我们计算了平均阳性/阴性预测值。使用归纳主题分析对访谈进行分析。
我们没有发现关于患者情况的共识（α 0.41，95%CI 0.29-0.52）。呼吸功能不全的阳性预测值较高（0.91，95%CI 0.86-0.97），但阴性预测值为 0.66（95%CI 0.44-0.88）。在访谈中，出现了两个关于解释过程的主题。“偏差的解释”包括参与者用来确定稳定性的策略，重点是发现数据中的偏差。“无法看到患者”意味着医院专业人员在估计患者情况时需要进行患者评估。
医院专业人员对连续测量的生命体征的解释，以及使用这些数据识别呼吸功能不全，是可变的，这可能是由于不同的解释策略、对偏差的不确定性以及无法看到患者造成的。协议和培训可以帮助统一解释，但决策支持系统可能是必要的，以发现可能被忽视的恶化迹象。
