“CATCH-IT Reports” are Critically Appraised Topics in Communication, Health Informatics, and Technology, discussing recently published ehealth research. We hope these reports will draw attention to important work published in journals, provide a platform for discussion around results and methodological issues in eHealth research, and help to develop a framework for evidence-based eHealth. CATCH-IT Reports arise from “journal club” - like sessions founded in February 2003 by Gunther Eysenbach.

Monday, November 2, 2009

Nov 9: User-designed information tools to support communication and care coordination in a trauma hospital

Gurses AP, Xiao Y, Hu P. User-designed information tools to support communication and care coordination in a trauma hospital. J Biomed Inform. 2009 Aug;42(4):667-77.

Full-Text · Abstract Only · Slideshow
· Draft CATCH-IT

In response to inherent inadequacies in health information technologies, clinicians create their own tools for managing their information needs. Little is known about these clinician-designed information tools. With greater appreciation for why clinicians resort to these tools, health information technology designers can develop systems that better meet clinicians' needs and that can also support clinicians in design and use of their own information tools.

OBJECTIVE: To describe the design characteristics and use of a clinician-designed information tool in supporting information transfer and care coordination

DESIGN: Observations, semi-structured interviews, and photographing were used to collect data. Participants were six nurse coordinators in a high-volume trauma hospital. Content analysis was carried out and interactions with information tools were analyzed.

RESULTS: Nurse coordinators used a paper-based information tool (a nurse coordinator's clipboard) that consisted of the compilation of essential data from disparate information sources. The tool was assembled twice daily through (1) selecting and formatting key data from multiple information systems (such as the unit census and the EHR), (2) data reduction (e.g., by cutting and whitening out non-essential items from the print-outs of computerized information systems), (3) bundling (e.g., organizing pieces of information and taping them to each other), and (4) annotating (e.g., through the use of colored highlighters and shorthand symbols). It took nurse coordinators an average of 41min to assemble the clipboard. The design goals articulated by nurse coordinators to fit the tool into their tasks included (1) making information compatible with the mobile nature of their work, (2) enabling rapid information access and note-taking under time pressure, and (3) supporting rapid information processing and attention management through the effective use of layout design, shorthand symbols, and color-coding.

CONCLUSIONS: Clinicians design their own information tools based on the existing health information technologies to meet their information needs. The characteristics of these clinician-designed tools provide insights into the "realities" of how clinicians work with health information technologies. The findings suggest an often overlooked role for health information technologies: facilitating user creation of information tools that will best meet their needs.


  1. Interesting paper, although I'm not clear why, if, and how the authors used grounded theory as they stated in the paper (and not in the abstract)- it seems there might be a mis-match. The paper's purpose is to describe the characteristics of the clipboard, which is very different from developing a mid range theory when using GT. Additionally, the methodology section is missing a lot of detail necessary for describing and understanding the coding and construction process in GT.

  2. I found this paper interesting because I tend to associate information systems with technology and often forget that information tools do not necessarily have to be electronic.

    Anyways, my question is regarding the sampling strategy: it is not clear how the authors decided which of the 6 nurse coordinators to interview formally and which ones to shadow (3 were shadowed while 5 were formally interviewed). I would have liked to see the justification for that in the paper.

    Moreover, why was this particular hospital chosen over others? How is the use of clinician-designed information tools at this hospital representative of other trauma hospitals elsewhere?

  3. With respect to the research study, I am not quite sure if the research question is appropriate. The ethnography is about learning the culture of a group or society through mostly participant observation and informal interviews.

    1. Based on the definition of ethnography, the research question is not quite clear; did the authors try to understand the characteristics of the tool, or the culture of the nurse coordinators with respect to using the tool?

    2. The details of the sampling were not clear. Can we call this a sampling since there were only 6 nurse coordinators entirely and all participated in the study?

    3. Saturation is an important element of qualitative methodologies. The authors could not perform saturation due to the limited number of informants.

    4. Why the authors did used grounded theory approach to analyze the clipboard?

  4. I found that this paper, although with interesting points, was not very novel. In fact, I would argue that all of the information contained within the study is old news.

    One point of interest for me was the sheer number of references cited in the paper. Thus, it makes me wonder if the results were actually deduced from the study itself, or by doing a review of the literature.

    Marjan asked most of the questions that I had. I was specifically interested in why the authors used such a small sample size. In addition, why were some nurse coordinators left out of the formal interviews (1/6) or why was the content validity of the instrument assessed by conducting interviews with only 3 of the 6 nurses.

    One last note about Table 2. What was the unit of time: mean or median. This is not clear to me.

  5. I was intersted in the intervention that was studied however I am unsure if their findings support their conclusions. Do their findings really support the conclusion that a paradigm shift is needed towards user designed HIT. Doesn't that presuppose that the end user has a firm grasp of various IT tools and how they can be applied to their work environment?

    I would think that their findings actually support the notion of a closer/better examination of the users work flow and needs. Which as James has noted is not so new.

  6. It is refreshing to see an actual research that examines the workarounds after an IT application (despite well designed workflow reviews and systems) with a view to improving HIT design. Research of this nature will be beneficial to health informatics.

    What is missing from the study, is a comparative study of nurses using the existing HIT tools to accomplish the same task and a comparative analysis done to determine the extent to which the workaround is due to user design vs workflow management given the lengthy time it takes to prepare the clipboard. Also, the training should be examined to see if this might also be a factor.

    An aside with regards to privacy, one observation is the possible violation of privacy given that photographs were taken of clipboard while it was being assembled before handoffs vs nurse coordinators removed all personally identifiable information from the sample clipboards before sharing them. Was there identifiable information in the photographs see fig 1 & 2?

  7. I would be interested in learning whether the authors had a rationale behind using the "Grounded Theory Approach" for this study. Given that the research was a fairly recent one (2006), it would have been more appropriate to use the "Framework Analysis" approach to analyze the qualitative data that uses a priori as well as emergent concepts. Moreover Framework Analysis seems to have gained popularity in recent years as mentioned in a report from the NHS UK which also provides a good source of reference for qualitative data analysis (Available from http://www.rds-eastmidlands.org.uk/resources/cat_view/13-resource-packs.html - click on Qualitative Data Analysis that opens in a PDF). I am somewhat advocating for the latter approach since the authors in this case explored a particular setting (trauma centre) to address information tools designed by a specific group of users (clinicians).

  8. In response to James and Marijan, I believe for ethnographic studies, it's been shown that only 6 members are needed before information saturation occurs and no new information can be gleaned. I don't think this is the type of information system or even an article that can be useful for designers. The clipboard acts as little more than an integrated data repository. This would be mediated with interoperability being addressed. But unfortunately, we all knwo this is not the case. Overall, I think this study lacked a clear way to improve the way systems are designed. As mentioned previously, (and I agree) it did not offer a novel approach to the concept of HIT design.

  9. I found the discussion about particular methods used in this study to be quite entertaining. There may be issues with them, but I thought that the authors made an honest attempt to explain much of what they have done.

    This study relates back to the design of an information system - who do you study for gathering the requirements? Evidently, in this research the authors used one individual hospital with 6 nurses. I'm not sure why the authors chose not to explore other hospitals for this, because I would not have been surprised to see a very different process for composing the clipboard in other hospitals. Also, do note that along with the information needs, there is also the need of good 'usability' of the layout of the clipboards.

    One question that I would like to raise is, did the authors make an attempt to identify any information system that the hospital has that DOES NOT meet the need of the nursing coordinators? The reason is, this may very well be a gap in the information need they have, which MAY be satisfied by designing a system with the design ideas gathered from the nurse coordinators, rather than classifying it as the limitation of the systems that may not have been implemented with this very idea in mind.

  10. How did the three authors independently conduct content analysis. And why did they also use grounded theory?

  11. Forgot to also mention that the abstract noted six participants but page 669 stated that they conducted interviews with five nurse coordinators. So was the sample size five or six?