Eagle Benjamin, Williams David J, Dingley John
From the *Morriston Hospital, Department of Anaesthetics, ABM University Health Board, Swansea, United Kingdom; †Welsh Centre for Burns, ABM University Health Board, Swansea, United Kingdom; and ‡Swansea University College of Medicine, Swansea, United Kingdom.
Anesth Analg. 2017 Aug;125(2):458-466. doi: 10.1213/ANE.0000000000002172.
An ideal electronic anesthesia recording system would be capable of not only recording physiological data but also injectable drug doses given, including those given incrementally from one syringe, without recourse to manual data entry. We compared 2 prototype devices which wirelessly recognized individual syringes and measured changes in their plunger positions via 2 different optical noncontact means, allowing calculation of incremental drug doses given.
Both devices incorporated a radio-frequency identification reader, which wirelessly read a unique code from a radio-frequency identification tag within syringe drug labels. A custom-designed cradle oriented any inserted 1-mL to 20-mL syringe in a repeatable position. The "laser" device had a moving laser beam broken by the end of the syringe plunger. The infrared (IR) device measured time of travel of IR light from a sender to a syringe plunger and back to a receiver. Both devices could therefore determine the drug and volume administered since the previous occasion when any syringe had been used. For each syringe size of 1, 2, 5, 10, and 20 mL, 121 plunger-length measurements were made over their full range, with each machine against a reference method of water filling and weighing using a randomized de Bruijn sequence.
For every syringe size, the laser device showed greater accuracy and precision, lower bias, and narrower limits of agreement (95% confidence intervals = bias ± 1.96 SD) than the IR device when compared to the reference method. For all syringe sizes, the range of bias was -0.05 to 0.32 mL for the laser and -2.42 to 1.38 mL for the IR. Lin concordance correlation coefficient values for the IR versus reference methods ranged from 0.6259 to 0.9255, with the lowest coefficients seen in syringes with the shortest distance of plunger travel (2 and 5 mL), while in laser versus reference comparisons, these coefficients were similar (0.9641-0.9981) over all syringe lengths.
Both devices measured syringe volume changes, demonstrating potential for measuring incremental drug doses, recording these, and also the time of each measurement. The IR device had no moving parts, which would be advantageous in a clinical situation. However, the current embodiment was not deemed accurate enough for clinical use, potentially remediable through improvements in hardware and software design. The laser device showed high accuracy and precision over all syringe sizes and contained volumes, and was considered potentially accurate enough for clinical use with suitable development.
理想的电子麻醉记录系统不仅应能够记录生理数据，还应能够记录所给予的可注射药物剂量，包括从一个注射器逐步给予的剂量，而无需手动输入数据。我们比较了两种原型设备，它们通过两种不同的光学非接触方式无线识别单个注射器并测量其柱塞位置的变化，从而能够计算所给予的递增药物剂量。
两种设备都集成了一个射频识别读取器，该读取器从注射器药物标签内的射频识别标签无线读取唯一代码。一个定制设计的支架将任何插入的1毫升至20毫升注射器置于可重复的位置。“激光”设备有一个移动的激光束被注射器柱塞的末端打断。红外（IR）设备测量红外光从发射器到注射器柱塞再回到接收器的传播时间。因此，两种设备都可以确定自上次使用任何注射器以来所给予的药物和体积。对于1、2、5、10和20毫升的每种注射器尺寸，在其整个范围内进行了121次柱塞长度测量，每台机器都与使用随机德布鲁因序列的水填充和称重参考方法进行比较。
与参考方法相比，对于每种注射器尺寸，激光设备在准确性、精密度、偏差更低以及一致性界限（95%置信区间 = 偏差 ± 1.96标准差）更窄方面均优于红外设备。对于所有注射器尺寸，激光设备的偏差范围为 -0.05至0.32毫升，红外设备的偏差范围为 -2.42至1.38毫升。红外设备与参考方法的林一致性相关系数值范围为0.6259至0.9255，在柱塞行程距离最短的注射器（2毫升和5毫升）中系数最低，而在激光设备与参考方法的比较中，所有注射器长度的这些系数相似（0.9641 - 0.9981）。
两种设备都测量了注射器体积变化，显示出测量递增药物剂量、记录这些剂量以及每次测量时间的潜力。红外设备没有移动部件，这在临床情况下将是有利的。然而，当前的实施方案被认为在临床上不够准确，可能通过硬件和软件设计的改进来补救。激光设备在所有注射器尺寸和容量范围内都显示出高精度和精密度，并且被认为经过适当开发后在临床上可能足够准确以用于临床。
