Formative Evaluation of Care Nexus: a Tool for the Visualization and Management of Care Teams of Complex Pediatric Patients

Abstract

Complex and chronic conditions in pediatric patients with special needs often result in large and diverse patient care teams. Having a comprehensive view of the care teams is crucial to achieving effective and efficient care coordination for these vulnerable patients. In this study, we iteratively design and develop two alternative user interfaces (graphical and tabular) of a prototype of a tool for visualizing and managing care teams and conduct a formative assessment of the usability, usefulness, and efficiency of the tool. The median time to task completion for the 21 study participants was less than 7 seconds for 19 out of the 22 usability tasks. While both the prototype formats were well-liked in terms of usability and usefulness, the tabular format was rated higher for usefulness (p=0.02). Inclusion of CareNexus-like tools in electronic and personal health records has the potential to facilitate care coordination in complex pediatric patients.

Introduction

Children and youth with special health care needs (CYSHCN) generally have or are at a risk of developing chronic medical and mental health conditions, resulting in an increased need for specialized medical, therapeutic, equipment, family support, and other services¹. 15.6% (approximately 11 million) of the pediatric population has special needs and accounts for about a third of the total healthcare spending associated with children². Coordinating care for these medically complex patients is crucial for enabling efficient use of resources, reducing costs, enhancing communication between patient/family and provider, and improving patient/family and provider satisfaction^3–5.

Caring for CYSHCN often involves a large number of participants from disparate settings, working independently, and serving in various roles that may wax and wane in importance or need over the continuum of patient care^6,7. The clinical care teams often include the primary care physician and a number of specialists, care coordinators, therapists, and social workers. However, the care teams of CYSHCN also extend far beyond the hospital/clinic walls and may include schools; community resources such as support groups and family advocacy groups; cultural and charitable organizations; multiple payers and funding agencies; and family members, neighbors, and friends⁷. In a previous study, we identified and described information needs of physicians, care coordinators, and parents to support care coordination of CYSHCN and categorized them into information goal types⁷. One of the goal types we identified was care networking and we defined it as “building a patient’s care team or network, knowing team member identities and roles, and sharing pertinent information to enable activities/actions as a team”. Keeping track of who is involved in a patient’s care at any given time, what their roles are, their goals and feedback, and preferred contact information; and finding new care team members to fit patient and family needs, was described as extremely challenging by the interviewees. Because they lacked appropriate tools, physicians, care coordinators, and parents of CYSHCN resorted to workarounds such as spreadsheets, hand-written sticky notes, refrigerator magnets, and memorization to store and track care team-related critical information.

In spite of the great potential for electronic health records (EHRs) to help providers coordinate care, current EHR systems do not adequately support the needs of care coordination^8–10. Vawdrey et al. note that we need better tools to support care team-related information in commercial EHR systems¹¹. Usability of EHRs also falls short in supporting the unique needs of representing information about complex patients^12–14. A recent systematic review reported that lack of appropriate software functionality and poor user interfaces were linked to patient safety concerns¹⁵. Electronic personal health records (PHRs) have been proposed as a strategy to support care coordination¹⁶. However, few PHRs have been developed for the domain of pediatrics due to lack of standards for pediatric content and customizations needed for chronic conditions¹⁷. With lack of PHRs to support care networking, parents of CYSHCN bear the burden of maintaining information about their child’s care team and repeating their “story” while coordinating care among a large and diverse care network.

In this study, we partner with care team members in key and complementary roles: physicians, parents, and care coordinators of CYSHCN, to gain insights into designing an application for clinicians and patients/families for the purpose of care networking. We iteratively design two alternative user interfaces to view, understand, share, and manage patient care team information. We then implement the designs and conduct a formative evaluation of the usability, usefulness, and efficiency of the user interfaces.

Methods

This study used a within-subject design comparing the interactions of physicians, parents, and care coordinators of CYSHCN with two user interface designs (graphical and tabular) of a prototype software we named “Care Nexus” to accomplish goals and tasks related to creating and managing patients’ complex care networks. The study involved both, granular tasks designed to assess usability, and high-level tasks focused on solving a care coordination problem described in vignettes. The study addressed the following research questions: 1) to what degree are features offered by Care Nexus to create, understand, and manage care networks of CYSHCN easy to use and efficient?; 2) how useful are the features offered by Care Nexus?; 3) how do the graphical and tabular displays of the care networks compare in terms of usability and usefulness? The study was approved by the University of Utah Institutional Review Board under protocol #IRB_00096357.

Care Nexus tool design

The design of Care Nexus was guided by Information Foraging theory¹⁸, Shneiderman’s principles for information visualizations¹⁹, and Jakob Nielsen’s heuristics for user interface design²⁰. The Information Foraging theory draws an analogy between a bird foraging for food and humans foraging for information. The optimal foraging effort seeks maximum “benefit” from minimal “cost” of information seeking by identifying rich information patches. We enabled users to optimize their information seeking effort by providing information patch enrichment (i.e. providing ways to get to the relevant content quickly and easily). We also applied Shneiderman’s visualization principles by offering an overview of information at the first level, then implementing zoom-in/zoom-out functions for the information, and finally providing information details on demand. Additionally, we have incorporated Nielsen’s principles for user interaction design to the Care Nexus prototype.

The design of Care Nexus followed an iterative design methodology based on rapid prototyping, analyzing, and refining cycles guided by feedback from representative users from each of the target user roles: physician, parent, and care coordinator. Tabular representation of medical data in the form of charts are common in current EHRs (e.g. Cerner™ uses tables to display patient care teams). This made the tabular design an obvious choice. A recent systematic review on innovative visualization of EHR data reported that color, lines, shapes, and visual diagrams have been effectively used to render patient data²¹. Thus, we opted to design a graphical interface that depicts the care team as a visual diagram as an alternative to the tabular format. We started off with “low-fidelity” prototypes in the form of whiteboard diagrams and software mockups. As the design matured, we transitioned to web-based “high-fidelity” prototypes using the Angular JS™ framework, Java^TM RESTful Web Services, and My SQL™ database. We further incorporated the critique from human factors and usability experts into the mature designs.

The functional specification for Care Nexus is derived from the results of our previous work focused on eliciting information needs and associated goals that are raised by physicians, care coordinators, and family members while coordinating the care for CYSHCN⁷. Care Nexus is designed with the objective of supporting the information goal of care networking, specifically creating, understanding, and managing patient care networks or care teams. Care Nexus does this by supporting previously identified goal sub-types of care networking: 1) care team building, 2) care team member identities, 3) contact information, 4) shared team knowledge, and 5) shared team action.

Participants and setting

Participants in the iterative design phase and the study were recruited from primary care sites that participate in pediatric Patient-Centered Medical Home (PCMH) Demonstration projects in Utah²². Each practice has a designated care coordinator and has one or more “family partners” who are actively-engaged parents of CYSHCN. The iterative user design phase included one of the co-authors (CN), one parent, and one care coordinator. For the formative evaluation, we recruited a purposive sample of 21 subjects (7 each of physicians, parents, and care coordinators) with the following criteria: 1) a minimum of 2 years of current experience caring for CYSHCN; 2) experience across a wide range of clinical and patient conditions; and 3) no previous exposure to the Care Nexus tool. The participants were invited by email to join the study by the project director of the PCMH Demonstration and co-author CN. We determined the sample size by following recommendations from the literature^23,24.

Case vignettes

The two case vignettes used in the study were adapted from the “Essential Information for Children with Special Healthcare Needs” project headed by the HL7 Child Health work group²⁵. The primary author contributed to this project by suggesting use cases and writing story boards for the selected use cases. The case vignettes were representative of the clinical conditions, information needs, and challenges that often face this cohort of patients and are comparable in complexity. Each case vignette consisted of a narrative about the patient’s clinical and social context and a care coordination problem related to a current event or episode in her/his life that needed to be resolved using Care Nexus. The two case vignettes were further customized to the role of the study participant (physician, parent, or care coordinator). The case vignettes and the associated data were synthetic and were approved by the users in each of the three roles who participated in the user interface design phase.

Procedure

The study was conducted either in an office setting at the work sites of the participants or at their homes. The two case vignettes and two interface designs resulted in four possible case vignette/display format combinations: 1) case vignette 1 + graphical format; 2) case vignette 1 + tabular format; 3) case vignette 2 + graphical format; and 4) case vignette 2 + tabular format. Each of the participants interacted with two of the four combinations in random order such that all participants interacted with both the displays and both case vignettes.

The study session began with a brief introduction of the study. In study part 1, the participants were asked to complete the following steps for their first case vignette/CareNexus display format combination: 1) usability tasks: perform 22 tasks (Table 1) distributed over the 5 goal subtypes of care networking⁷; and 2) problem-solving: identify care team members to communicate with using CareNexus to resolve a problem related to care coordination for a current episode of care as posed by the case vignette. This was followed by a questionnaire that assessed the usability and usefulness of one display format of CareNexus. These steps were repeated for their second case vignette/CareNexus display format in study part 2. Finally, the study subjects were asked to rate the usefulness of a set of CareNexus features and provide open-ended comments and suggestions. The participants were not provided with a tutorial of CareNexus. The goal was to assess the intuitiveness, usability, usefulness, and efficiency of CareNexus without any prior exposure to the user interface of CareNexus. The user sessions were recorded using Hypercam, a screen capture software.

Table 1.

Usability tasks and time to completion (average, median, range; in seconds).

Usability task	Average time (s)	Median time [min-max]
Identify the “inactive members” of the patient’s medical care team.	1 ± 0.2	1 [1-1]
Identify the “less active” member(s) of the patient’s family network.	1 ± 0.3	1 [1-1]
Identify the “less active” member(s) in the patient’s medical network	1 ± 0.3	1 [1-1]
Identify the patient’s event timeline.	1.1 ± 0.3	1 [1-2]
Find the patient’s conditions.	1.2 ± 0.5	1 [1-3]
How many total members are in the patient’s family network?	2 ± 0.2	2 [2-2]
Identify the “active” member(s) of the patient’s family network.	2 ± 0.2	2 [2-2]
Who are the medical specialists the patient is actively/currently seeing?	2 ± 0.2	2 [2-2]
What is the preferred contact number for the patient’s main contact?	2 ± 0.2	2 [2-2]
Find the patient’s name on the screen.	2.1 ± 0.4	2 [2-3]
Identify the care team action(s) of the currently logged in user.	2.4 ± 0.8	2 [2-5]
How many of the care team actions are “Done”?	2.8 ± 1.0	2 [2-5]
Find the patient’s main contact person (by name or role).	2.8 ± 1.2	2 [2-6]
What is the contact information for an “active” care team member of the patient’s non-medical care team?	3 ± 0.2	3 [3-3]
Zoom in and zoom out on the event timeline.	4.2 ± 1.0	4 [3-6]
Identify the events related to the current patient episode on the timeline.	5.5 ± 1.6	5 [3-10]
Delete the care team member you added.	5.5 ± 0.8	6 [5-8]
What are the goals and/or feedback of any one of the active specialists in the patient’s care team?	6.6 ± 2.7	6 [4-17]
How many care team actions are currently displayed?	6.7 ± 1.7	6 [5-10]
Search for a new care team member.	13.4 ± 2.3	14 [8-17]
Identify events on the timeline ± 6 months from today.	14.9 ± 5.1	15 [8-29]
Manually add a new care team member.	32.8 ± 5.2	31 [26-43]

Data analysis

Using the video recordings from Hypercam, each of the usability tasks were coded for: 1) ability to carry out the usability task to successful completion, and 2) time to completion. Given that the data represented repeated measurements of ratings, where the user rated two different interface designs, a paired sample data analysis was used. Comparison between the ratings of the two interface designs was performed using mixed-effects linear regression, with repeated measurements nested within user, controlling for the covariates of perceived vignette complexity, experience of the study subject with the patient conditions in the vignette, and the sequence in which the interface designs were evaluated. We developed a 14-item questionnaire with Likert-scale response options (1=strongly disagree; 5=strongly agree) to assess the usability and usefulness of CareNexus. The questionnaire included five questions from the System Usability Scale²⁶ and nine questions that measured self-perceived ability to understand the gist, create, and manage care networks and related information. The individual questions were aggregated into two composite scales: usability (questions 1, 3, 8, 12, and 13) and usefulness (questions 2, 4, 5, 6, 7, 9, 10, 11, 14) to maximize reliability and generalizability. Reliability analysis was performed using Cronbach’s alpha by aggregating ratings for the two user interface designs for the composite scales.

Results

CareNexus user experience

The design of the user interface of CareNexus required ten iterations that were performed before the formative evaluation. The resulting user interface of CareNexus with the graphical care network format for case vignette 1 is shown in Figure 1. Figure 2 shows the tabular care network display for case vignette 2.

Figure 1.

CareNexus user interface with a graphical view of the care network for case vignette 1.

Figure 2.

CareNexus user interface with a tabular view of the care network for case vignette 2.

The patient banner (section #1, Figures 1 & 2) gives a quick overview of the patient and indicates additional needs if applicable (e.g. need for a language interpreter). Section #2 below shows the event timeline with boxes intuitively labeled to indicate clinical events (e.g. outpatient, inpatient, and emergency room visits) and other significant events in the patient’s life providing information-patch enrichment. The start and end of timeline defaults to six months before and three months after the current date respectively. The user can view a wider or narrower timeline range by using the navigation menu buttons or the mouse wheel. Clicking on individual events, displays detailed information about the primary care team member involved in the event in the bottom left-hand side of the screen (section #4). Shared team knowledge of the patient demographics, clinical conditions, a quick reference to the contact information of the person most involved in the patient’s care, and a quick summary of the patient’s encounters aims to provide patient context efficiently and accurately.

Users can visualize the patient’s care network in section #3 (Figures 1 & 2). Knowing who is currently involved in the patient’s care, in what role, and how to contact them is vital to care networking. The care network is displayed either in a graphical (Figure 1) or tabular format (Figure 2). In the graphical format, the patient is shown in the center of the network and is surrounded by either the family (yellow), medical (green), or non-medical (blue) care network member nodes. The tabular format displays the care team members in a table sorted by status (“active” at the top, followed by “less active”, and finally “inactive”). The family network includes immediate and extended family members involved in the patient’s care. Care team members belonging to a clinic setting are categorized as medical network members (e.g. primary care physician, care coordinator, neurologist, and social worker). The non-medical team members include the extended care network such as the school, community support groups, and durable medical equipment providers. The color gradient of the nodes of the care network and the thickness and style of the connecting lines indicate closeness of the care team to the patient’s care. There are three levels of closeness or importance: 1) active (color: darkest, connecting line: bold and solid); 2) less active (color: lighter, connecting line: medium and solid), and inactive (color: grey, connecting line: medium and dashed). The closeness or importance of a care team member to the patient’s care network can be manually assigned or can be inferred based on whether the team member has been involved in the patient’s care within a certain time frame: 1) “active” indicates activity within the last 3 months; 2) “less active” indicates activity within last 6 months; and 3) “inactive” indicates no activity for over 6 months. The label across the top gives a quick summary of the number of active, less active, and inactive care team members. Our design goal was to provide optimal cues to users to help them understand the gist of the patient’s care network and the ability to get more information on demand, per Shneiderman’s visualization principles.

The bottom left of the screen (section #4) shows details of the care team member with the team member identity and contact-information of the selected care team entity (primary care team member involved in a timeline event or a member from the care network). This information display resembles a “business card” following Nielsen’s design principle of matching real world and software system representations. Next to it we display the status, recent appointments, patient care goals, and feedback (if applicable and available) of the care team member. Building and managing a patient’s care team (section #5) can be accomplished by using the search, add, update, and delete functionality conveniently co-located in bottom center of the screen. Users can update and/or delete only those care team members who have been added by them. Finally, shared care team actions are supported by providing information about who is responsible for which task, due date (if applicable), and the status of the task. The users can view all team members’ tasks and add/delete/update their own task list. Following Nielsen’s design principles, we have maintained simplicity and color/font consistency throughout the design of CareNexus and minimized the need for user recall by providing convenient tool tips.

Ease of use, efficiency and usefulness ratings

The study participants were able to successfully complete all of the 22 usability tasks (Table 1). The median time to completion was less than 7 seconds for all but 3 tasks (searching for and manually adding care team members, and identifying events took > 7 seconds). All users also successfully identified all the care team members needed to resolve the care coordination problems posed by the case vignettes. Users highly rated all CareNexus features with highest ratings for the timeline, team member business/contact cards, and most recent and next appointments (Table 2).

Table 2.

Usefulness ratings of CareNexus features.

Feature description	Rating (1=not at all useful; 5=very useful)
	Mean	Std. Dev.	Min	Max
Patient’s primary contact on the patient banner	4.85	0.35	4	5
Timeline of events	4.95	0.21	4	5
Zoom-in/Zoom-out for event timeline	4.33	1.01	2	5
Display of three separate care networks	4.81	0.40	4	5
Display of number of care team members per network	4.67	0.58	3	5
Color-scheme supported display of “active”, “less active”, and “inactive” care team members	4.81	0.51	3	5
“Business cards” for the selected care team member	4.90	0.30	4	5
Most recent and next appointments for the selected care team member	4.90	0.30	4	5
Goals for the selected care team member	4.86	0.36	4	5
Feedback from the selected care team member	4.71	0.46	4	5
Search for new care team members	4.62	0.59	3	5
Add to (search and manual), delete, and update the care network	4.86	0.36	4	5
Care team actions	4.71	0.56	3	5

Comparison between CareNexus user interface designs

Cronbach’s alpha for the composite variables of usability and usefulness are given in Table 3. There was a significant lower mean usefulness rating for the graphical interface compared to the tabular interface, after controlling for complexity, experience, and the sequence in which the interfaces were evaluated (adjusted mean difference=-0.12; 95% CI: -0.22,-0.01; p=0.02). Although statistically significant, the differences are very small and may not indicate clinical significance. There was a non-significant lower mean usability rating for the graphical interface compared to the tabular interface, after controlling for the same criteria (adjusted mean difference=-0.12; 95% CI: -0.26,0.03; p=0.12). Given our sample size of 21 we had 80% power using a two-sided alpha 0.05 comparison to detect a paired sample standardized mean difference of 0.64, which represents a moderate to large effect size by Cohen’s criteria²⁷. Table 4 gives the adjusted mean ratings for the individual questions and the composite variables assessing usability and usefulness of CareNexus.

Table 3.

Cronbach’s alpha for the composite variables.

User interface design format	Usability	Usefulness
Graphical	0.79	0.87
Tabular	0.83	0.94

Table 4.

Ratings of the CareNexus interface designs.

Measurement	Tabular design		Graphical design		P-value
	Adjusted Mean*	Sth. Err.	Adjusted Mean*	Sth. Err.
Q1. I thought the system was easy to use.	4.72	0.10	4.55	0.10	0.21
Q2. I was able to grasp the gist of the patient's care network.	4.81	0.11	4.51	0.11	0.06
Q3. I found the various functions in this system were well integrated.	4.62	0.11	4.56	0.11	0.50
Q4. I was able to find the care team members relevant to the case vignette.	4.82	0.09	4.65	0.09	0.124
Q5. I was able to find the pieces of information I needed to accomplish the tasks in the case vignette.	4.85	0.08	4.76	0.08	0.18
Q6. It was easy to understand the meaning of the information presented.	4.86	0.10	4.61	0.10	0.05
Q7. I was able to find the contact information for the care team member(s) I need to communicate with.	4.85	0.08	4.77	0.08	0.20
Q8. I would imagine that most people would learn to use this system very quickly.	4.82	0.08	4.69	0.08	0.15
Q9. It was easy to search for newcare team members.	4.82	0.10	4.69	0.10	0.21
Q10. I was able to find goals of the specialists working with the patient.	4.86	0.08	4.80	0.08	0.24
Q11. I was able to find feedback of specialists working with the patient.	4.86	0.08	4.75	0.08	0.14
Q12. I think that I would like to use this system frequently.	4.82	0.08	4.79	0.08	0.64
Q13. I found the system very cumbersome to use. (reversed criteria)	1.08	0.09	1.33	0.09	0.04
Q14. Compared to the tools/workflow I currently usefor care networking, I thought that CareNexus made it easier to accomplish care networking.	4.86	0.08	4.79	0.08	0.46
Usability (composite scale)	4.77	0.07	4.66	0.07	0.12
Usefulness (composite scale)	4.84	0.06	4.72	0.06	0.02

^*adjusted for perceived vignette complexity, experience with patient conditions in the case vignette, and sequence in which the interface designs were evaluated.

Open-ended comments

The study participants echoed their appreciation for CareNexus in their comments and offered suggestions for improvements (Tables 5 and 6).

Table 5.

Open-ended comments by participant roles.

Role of participant	Comments
Physician	“The care network was easy to use. It was extremely helpful to have the timeline to assess where the patient has been and where she is heading with her case.”
	“This was quite easy to navigate.”
	“Excellent tool.”
	“Nicely divided into family, medical, and non-medical.”
	“Timeline is fantastic.”
	“It is actually helpful to know the missed appointments. I have to go to two different screens to see the missed ones in my current EMR.”
Care coordinator	“The visual timeline is a great way to help patients with appointments.”
	“I like that it is all on one screen and there aren’t a lot of tabs to navigate through.”
	“I like the color coding.”
Care coordinator	“I like that it is all in one place.”
	“Nice interface compared to the current EMR.”
	“I found the graphical visually more over stimulating or busy.”
	“Well developed and user friendly. This app would replace our Excel registry. I can’t say enough positive about the app –love it.”
Parent	“I could use this on a regular basis finding what I need.”
	“I currently do not have a tool for care networking other than a notebook.”
	“Looks to be very exciting and useful.”
	“CareNexus is very user friendly.”
	“I love the display of goals and feedback.”

Table 6.

Suggestions by participant roles.

Role of participant	Suggestions
Physician	“It would be nice to connect the timeline events to notes.”
	“Let the user choose the interface format: tabular vs. graphical.”
Care coordinator	None.
Parent	“It would be nice to have this tool available in languages besides English.”
	“Being able to scan in documents, prescriptions, IEP documents would be helpful.”
	“Reminders to schedule specialist appointments would be great.”

Discussion

Pediatric patients with special needs have large care teams with members in various roles corresponding to different aspects of patients’ lives, such as treatment and management of health conditions, developmental challenges, educational needs, and financial support. Previous research in the domain of patient care teams has addressed availability of care team-related information in the inpatient setting^11,28, supporting team work within the same care setting²⁹, and tailored applications for the needs of patients with certain conditions³⁰. However, less has been done for designing clinician- and patient-facing applications to address the needs of medically complex patients that cross the boundaries of specific settings and conditions. The goal of our research is to address this gap by designing, developing, and evaluating two alternative user interface designs for a prototype of CareNexus, a tool to visualize, understand, share, and manage care team related information for complex pediatric patients.

The study participants highly rated the feature set and the overall user interface of CareNexus in terms of usability and usefulness. These findings are important based on the technology acceptance model (TAM) which stipulates that perceived usefulness and ease-of-use are predictors of actual use³¹. Several factors may have contributed to these findings, including deriving the requirements from a systematic information needs analysis⁷, the early involvement of representative users, an iterative design approach based on the information foraging theory¹⁸, Shneiderman’s visualization principles¹⁹, and Nielsen’s usability principles²⁰. All users completed 100% of the usability tasks (time to completion for 19 out of the 22 tasks was less than 7 seconds). The problem-solving segment of our evaluation approach encouraged the users to quickly grasp the gist of the care network, identify the care team member(s) relevant to the problem, and access their contact information. The patient event timeline was the highest ranked component of the application. Users found it to be a very intuitive and quick way to gain shared team knowledge about recent visits, issues, and current status of the patient’s care. The tabular format of the care network was preferred over the graphical format, however the differences in the ratings were small. Participants liked both the formats and users should be allowed to choose between the two formats, per individual preference, as suggested by one of our physician participants. Further research is needed to investigate if our findings regarding tabular versus graphical displays generalize to other applications as well as the different factors that may influence display format preferences. Display of three separate networks and the associated color coding made it easier for the users to identify care team members. Having the contact information along with the preferred contact readily available can be very beneficial, especially in emergent situations which was described as a need for this cohort of patients⁷. Sharing the goals and feedback provides a way to create a shared sense of common ground³² between the team members. Finally, care team actions enable processes that require shared responsibility between care team members ensuring that members of the care team (including the parents) are “on the same page”. Viewing the care team actions and their status is valuable for all team members and may reduce the need for time-consuming and sometimes unreliable person-to-person communication.

The results of our formative evaluation are promising and warrant future work on: 1) analyzing the information sources for the care team-related information displayed in CareNexus, 2) developing algorithms to automatically populate applications like CareNexus, and 3) integrating CareNexus into EHR and PHR workflows. Future studies should also focus on extending the findings in this study to other patient populations needing chronic care management.

Limitations

The case vignettes used in this study were adapted from the use cases identified by the HL7 Child Health work group. Although they are representative of clinical and patient conditions of children with special needs, it is possible that different conditions may require other design features. The formative evaluation assumes availability of accurate and up-to-date care team information. High-quality care team data may not be readily available in real world systems and that may influence users’ perception of usefulness. Also, CareNexus is designed to be used in tandem with EHR and PHR systems. Further studies should investigate the usability, usefulness, and efficiency of CareNexus integrated in the user workflow.

Conclusion

We describe the design and formative evaluation of two alternative user interfaces of CareNexus, a prototype of a tool to view, understand, share, and manage patient care team information. We followed an iterative design approach guided by the information foraging theory, information visualization principles, and user interface design heuristics. In addition, feedback from representative users was incorporated early into the design. Twenty-one users participated in the formative evaluation of the resulting graphical and tabular user interfaces. Users highly rated the usability, usefulness, and feature set of CareNexus, and were able to complete the usability tasks in a short amount of time. The tabular format was rated higher for usefulness but the difference was small indicating that the users liked both the formats. Tools that enable understanding the gist of a patient’s care network across organizational boundaries, the temporal nature of care team relationships, details of contact information, goals and feedback of those involved in the patient’s care, and ability to find providers to match patient/family needs have the potential to facilitate care coordination and team collaboration.