18

towards the documentation process. The survey included qualitative components in the form of short-answer questions. The survey was constructed to gauge user attitudes towards the current documentation process using a series of short-answer questions. The questions were designed to measure the perceived utility and usability of the current paper documentation form, to identify which aspects might be candidates for redesign. In addition to the short-answer questions, the survey included a series of Likert-scale questions. These questions were used to measure stakeholder attitudes towards the importance of accurate and timely hospital documentation (Figure 1). The survey was intended to be completed quickly, to minimize the impact of 13 this research activity on a subject’s work schedule. Therefore the initial draft of the survey was limited to a single demographic question and seven short-answer questions. Figure 1: Needs assessment survey consisting of seven short-answer questions, a series of Likert-scale questions about documentation usability and effectiveness, and a series of questions about technology preferences. The survey was administered to a convenience sample of nurses determined to be most likely users of the documentation forms at the bedside. This included the 12 highly trained STAT nurses, who are registered nurses with extensive critical care training. Rather than being assigned to a single location within the hospital, these STAT nurses respond to emergency events in a roaming capacity. This includes seeing at-risk patients as part of a proactive rapid response team, as well as administering emergency care during cardiac arrest and respiratory arrest events. 28 Because of this, STAT nurses have extensive knowledge and experience dealing with code blue emergencies. The survey was also targeted towards the 300 ICU nurses. Because ICU patients were at greater risk of cardiac arrest, the ICU nurses were 14 more likely to have experienced a code blue event. Consequently, ICU nurses were likely to have experience using the code blue documentation form. The survey was pilot tested with two nurses while attending code blue events. The purpose of the pilot surveys was to test the feasibility of conducting surveys within the hospital setting. Specifically, the pilot survey was implemented to examine recorder accessibility, as well as survey length and format. Although the pilot surveys were administered in-person, the survey was converted into an online survey. The reasons for this will be covered in the discussion section. The online version was administered using an access-controlled University of Washington website for security and privacy, and to prevent duplicate entries. The qualitative survey questions were modeled on the validated Post-Study System Usability Questionnaire (PSSUQ). 29 Originally developed for evaluating the usability of computer-based systems, the PSSUQ survey tool was modified by removing computer-specific questions and rewording them for paper-based documentation. The survey asked participants about attitudes towards the documentation forms and process, to investigate whether socio-technical reasons might explain why the forms were not thoroughly completed. This was accomplished through a combination of open-ended qualitative questions with an accompanying set of Likert-scale questions. Results were collected as typewritten responses through an online survey tool. Once collected, a basic thematic analysis was conducted for the responses associated with each of the survey questions. This analysis was initially conducted as an open coding exercise. Representative keywords were extracted verbatim from the text responses. These keywords formed the basis of an initial coding scheme. The responses were then axially coded, and related keywords were combined into larger coding categories so that thematically related responses could be clustered. 15 To supplement the qualitative results, a set of quantitative Likert-scale questions were used to assess user attitudes towards documentation. The questions were formulated to determine how user attitudes and information availability impact the completion rate of the documentation forms. Questions also assessed receptiveness towards technology-based documentation tools. The technologies selected were chosen from previous examples in the literature as well as technologies being tested at other local hospitals. Once the surveys were completed, the results were tallied, and basic statistical analysis was conducted to assess the mean, standard deviation, and confidence interval. Focus groups A set of three focus group meetings was held with the STAT nurses (experienced nurses), the nursing practice council group (floor nurses), and a combination of patient safety and risk management staff. The purpose of these meetings was to assess the validity of findings from the records review and field observations by discussing the findings and asking for verbal clarification. This was accompanied by an open solicitation for stories about rare, unusual, or undocumented practices otherwise not captured through the observations or surveys. Focus groups were chosen to maximize participation among care providers, who are busy and difficult to recruit. The group format also encouraged 

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