21

with documentation. This means that training-based efforts to improve documentation would only benefit the small percentage of the nurses who were eventually tasked with recording responsibilities. Policy changes to pre-assign recording duties or otherwise reduce the pool of potential recorders would make a training-based solution more effective. Otherwise, a redesign of the documentation form might be effective if it could reduce the need for specialized training. A redesign might accomplish this by providing more thorough instructions for these uncommonly encountered forms, or by improving workflow compatibility between the form layout and the typical sequences of clinical events during a cardiac arrest event. Discussion Taken together, the results of the various research activities combine to create an overall picture of the current state of in-hospital cardiac arrest documentation. The records review reveals that the existing cardiac arrest documentation forms are often missing important pieces of information, including physician signatures, patient survival, assisted ventilation, and endotracheal tube placement and confirmation. This prompted further questions about why those pieces of data were not recorded. 22 To begin answering this question, a series of field observations was conducted. The field observations highlighted some of the challenges with recording information during emergency events. In particular, the observations revealed that there was a delay between the start of care and the start of recording, which limited the ability of recording nurses to document the care process. This demonstrated that recording nurses must be able to start transcribing information without delay. This is one of the reasons why recording nurses complained that existing computerized documentation systems were too slow for real-time emergency documentation. Observations also highlighted problems with overcrowding, which created a noisy environment that interfered with data collection. Noise problems would limit the potential use of automated data capture systems that rely on audio-recording. This includes verbally announced medication orders, or verbal confirmation of medical procedures such as intubation or placement of a chest tube. Some types of data can be collected electronically, such as telemetry from monitoring equipment. For example, patient vital signs are monitored electronically for patients in the intensive care units. Some brands of readily available equipment can capture this data electronically right now, but it is limited to proprietary tools and data formats. In addition, there is no single system that currently captures all of the recommended types of data. While the automated collection of patient data would assist with documentation during a code blue event, existing systems are not comprehensive enough to collect all of the currently requested information. The records review and observations were supplemented with a survey of care providers who were involved with the documentation process. Initial pilot surveys were conducted in-person at emergency events. Unfortunately, a limiting factor was gaining access to subjects, because nurses and physicians were often in a hurry to return to their regularly scheduled clinical duties. This time pressure also explained difficulties in capturing signatures and other post-event information, since care providers 23 quickly dispersed once the emergency was concluded. The high-stress environment of an emergency situation was also emotionally draining, particularly when the patient could not be resuscitated. As a result, some recorders did not want to be interviewed. In addition, the exact timing of code blue events was unpredictable, and this presented problems with administering surveys in-person. These difficulties necessitated a change in survey methodology, prompting the development of an online survey in lieu of in-person surveys. While less dynamic, they allowed access to a greater number of respondents. The online survey also provided the additional benefit of allowing for lengthier and more detailed free-text responses. Survey results indicated that recorders treated documentation as an important task, and that they recognized the availability and usefulness of real-time clinical information. This showed that recorders were interested and motivated to capture real-time information, suggesting that problems with incomplete documentation records were likely due to problems transcribing information onto the code blue documentation form. To deal with this, participants suggested workflow improvements, such as moving sections around or rearranging the data elements to conform to expectations about information flow. These conclusions were supported by participant comments about workflow compatibility and ease of use. The survey and focus groups also asked participants to rate their preferences for a computer-based documentation solution. This was done in response to previous efforts in the literature to develop electronic replacements for pen and paper. Participants expressed interest in reducing the amount of work required to transfer records from paper into the electronic record. However, recorders expressed frustration with the speed and availability of existing desktop computer systems, and a 

1   2    3    4    5    6   7   8   9   10   11   12   13   14   15   16   17   18   19   20   21   22   23   24   25