17

forms and reviewed them to assess institution-wide performance. While they did not use the information in realtime, the patient safety staff relied on accurate real-time documentation to support patient safety 10 initiatives. This use of documentation to support patient safety goals was a driving force for this research. The risk managers were the last set of stakeholders specifically identified for this research. Although furthest removed from direct patient care, the potential impact of their work on the overall care system made their information needs an important factor in assessing the current documentation form. By conducting a formal needs assessment of existing documentation tools and processes, anecdotal dissatisfaction was identified and measured. This needs assessment was based on triangulation between multiple complementary methods, including a medical records review, a series of field observations, a survey with qualitative and quantitative questions, and stakeholder focus group activities. This study employed all of these methods to compensate for weaknesses in any single data collection method, allowing for a comprehensive understanding of the documentation process. 23,24 Records review A medical records review was conducted to review every code blue record (n=161) during the 2008 calendar year. The entire 2008 calendar year was chosen to compensate for potential seasonal variations in documentation. The documentation forms were made available through the patient safety office, and the records review was conducted there to ensure the security of private patient information. To further safeguard private patient information, identifiers such as name and date of birth were examined but not transcribed. The presence or absence of data elements was transcribed onto a computer spreadsheet so that the completion rates could be tallied. Completion rates focused on the standardized checkbox elements recommended by the Utstein-style template. These checkbox elements were applicable to all documented cardiac arrests despite differences in patient condition and treatment. The checkbox 11 elements include administrative information, such as time, date, location, and patient identifiers. The majority of situational data elements were presented as binary choices, so either “yes” or “no” should have been marked during every event. These standardized data elements were tallied to determine completion rates. Some data elements were presented as fill-in-the blank responses, such as the time when the first electric shock was administered. Unlike the checkbox discussed above, these unstructured fill-in-theblank items were not applicable to every emergency event. To compensate, the results were normalized to measure completion rates only for applicable events when that information was available. For example, documentation about endotracheal tubes was only examined when the patient care algorithms suggested its use. When applicable for a particular patient, these data elements were tallied to determine completion rates. The code blue documentation forms also contained a semi-structured “timeline” area for recording medications and procedures. The timeline was labeled with suggested category headings, such as a patient vital signs and administered medications. Because each event was unique and dependent on patient responses, the completeness of the timeline could not be assessed based solely on the records review. Instead, the timeline was examined to see what types of information were consistently collected. In addition to examining the timeline for clinical content, the utilization of space on the forms was examined to detect patterns that might explain documentation workflow practices. This included information noted via free text entry in designated comments areas, as well as undesignated whitespace areas such as margins. The analysis of timeline elements was qualitatively noted and used as the basis for questions during the survey and focus group activities. 12 Field observation Field observations were conducted to provide familiarity with the clinical setting and obtain first-hand experience with the workflow processes governing the documentation of patient care. The literature indicated that the majority of code blue emergencies took place during the daytime hours and were equally likely to occur on weekdays and weekends. 25 During the observation activities, efforts were made to blend in with the hospital staff to avoid attracting attention or otherwise disrupting the care process. 26,27 However, it was important to avoid conveying expectations of being able to administer direct care. To balance these needs, the observations were conducted while dressed in a similar manner to hospital administrative staff, and not wearing hospital scrubs or other clothing associated with bedside care providers. A field observation form was developed and used to collect observation notes, including data about interactions between the care providers and the designated event recorder in charge of documentation. The observation data was used to inform the survey design and prompt questions during the focus group activities. Survey To supplement the records review and field observations, a survey was employed to collect data about care provider motivations and attitudes 

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