Department of Emergency Medicine, Alpert Medical School of Brown University, Providence, Rhode Island.
Philips Research Eindhoven, Eindhoven, The Netherlands.
J Emerg Med. 2022 Jul;63(1):115-129. doi: 10.1016/j.jemermed.2022.06.001. Epub 2022 Aug 6.
Contactless vital signs (VS) measurement with video photoplethysmography (vPPG), motion analysis (MA), and passive infrared thermometry (pIR) has shown promise.
To compare conventional (contact-based) and experimental contactless VS measurement approaches for emergency department (ED) walk-in triage in pandemic conditions.
Patients' heart rates (HR), respiratory rates (RR), and temperatures were measured with cardiorespiratory monitor and vPPG, manual count and MA, and contact thermometers and pIR, respectively.
There were 475 walk-in ED patients studied (95% of eligible). Subjects were 35.2 ± 20.8 years old (range 4 days‒95 years); 52% female, 0.2% transgender; had Fitzpatrick skin type of 2.3 ± 1.4 (range 1‒6), Emergency Severity Index of 3.0 ± 0.6 (range 2‒5), and contact temperature of 36.83°C (range 35.89-39.4°C) (98.3°F [96.6‒103°F]). Pediatric HR and RR data were excluded from analysis due to research challenges associated with pandemic workflow. For a 30-s, unprimed "Triage" window in 377 adult patients, vPPG-MA acquired 377 (100%) HR measurements featuring a mean difference with cardiorespiratory monitor HR of 5.9 ± 12.8 beats/min (R = 0.6833) and 252 (66.8%) RR measurements featuring a mean difference with manual RR of -0.4 ± 2.6 beats/min (R = 0.8128). Subjects' Emergency Severity Index components based on conventional VS and contactless VS matched for 83.8% (HR) and 89.3% (RR). Filtering out vPPG-MA measurements with low algorithmic confidence reduced VS acquired while improving correlation with conventional measurements. The mean difference between contact and pIR temperatures was 0.83 ± 0.67°C (range -1.16-3.5°C) (1.5 ± 1.2°F [range -2.1-6.3°F]); pIR fever detection improved with post hoc adjustment for mean bias.
Contactless VS acquisition demonstrated good agreement with contact methods during adult walk-in ED patient triage in pandemic conditions; clinical applications will need further study.
无接触式生命体征(VS)测量技术,如视频光体积描记法(vPPG)、运动分析(MA)和被动红外测温法(pIR),已经显示出了良好的应用前景。
在大流行期间,比较常规(接触式)和实验性无接触式 VS 测量方法在急诊部门(ED)门诊分诊中的应用。
使用心肺监测仪和 vPPG、手动计数和 MA 以及接触式温度计和 pIR 分别测量患者的心率(HR)、呼吸频率(RR)和体温。
共纳入 475 例 ED 门诊患者(符合条件者的 95%)。患者年龄为 35.2±20.8 岁(范围 4 天至 95 岁);52%为女性,0.2%为跨性别者;皮肤类型为 Fitzpatrick 2.3±1.4(范围 1-6),紧急严重程度指数为 3.0±0.6(范围 2-5),接触温度为 36.83°C(范围 35.89-39.4°C)(98.3°F[96.6-103°F])。由于与大流行工作流程相关的研究挑战,儿科 HR 和 RR 数据未纳入分析。在 377 例成年患者中,对无预触发的“分诊”窗口 30 秒进行分析,vPPG-MA 获得了 377(100%)的 HR 测量值,与心肺监测仪 HR 的平均差异为 5.9±12.8 次/分(R=0.6833),获得了 252(66.8%)的 RR 测量值,与手动 RR 的平均差异为-0.4±2.6 次/分(R=0.8128)。基于常规 VS 和无接触 VS 的患者紧急严重程度指数成分的匹配度为 83.8%(HR)和 89.3%(RR)。过滤掉算法可信度低的 vPPG-MA 测量值可以减少 VS 的获取,同时提高与常规测量值的相关性。接触式和 pIR 温度之间的平均差异为 0.83±0.67°C(范围-1.16-3.5°C)(1.5±1.2°F[范围-2.1-6.3°F]);通过事后调整平均偏差,提高了 pIR 发热检测的准确性。
在大流行期间,无接触式 VS 采集在成年 ED 门诊患者分诊中与接触式方法具有良好的一致性;临床应用还需要进一步研究。