Coordination of Care for Complex Pediatric Patients: Perspectives from Providers and Parents

Abstract

Coordinators help patients requiring complex chronic care manage frequent ambulatory visits and services received at home or from community-based agencies. EHRs directly support only a few of the required tasks as they do not allow access to all parties involved in care. Our goal was to examine how technology was used to coordinate efforts and to describe common barriers and facilitators. Insights may inform the design of tools that would effectively support identified goals. We conducted five hours of interviews with sixteen parents and six clinicians and characterized emergent themes from transcripts. Situational awareness, care and visit planning, document aggregation, abstraction and interpretation were tasks essential to coordination yet generally poorly supported by EHRs. Providers communicated primarily by email, telephone and by exchanging paper and scanned documents. A preliminary model of coordination that could be used in the planning and testing stages of a User Centered Design process is described.

Introduction

Inadequate coordination has been identified as a significant impediment to effective care,^1,2 contributing to increases in costs and negatively affecting quality.^3–5 As specialist care is increasing and accounts for more than half of all ambulatory visits, teams consisting of several providers located at different institutions are common.⁶ Patients with multiple chronic conditions may visit up to 16 physicians in a single year.⁷ Each provider often needs to review results, findings, relevant patient history and other information gathered by the others in order to make care decisions. Exchange of information across technological and institutional boundaries, however, has been a time-consuming, complicated process fraught with care delays, duplication of effort and a source of frustration.

Most currently available electronic records systems (EHR) do not support coordination and collaboration among caregivers adequately.⁸ A survey of clinicians using vendor EHRs found that the technology provided only limited support for many common coordination tasks.⁹ For example, an accurate, updated list of care team members is not yet widely available.¹⁰ The low prevalence of this resource in current systems is reflected in the modest requirement by the Office of National Coordinator (ONC) to maintain such lists for at least 10% of patients in order to receive Meaningful Use Stage 2 certification.¹¹ The referral process for specialty care is often described as a frustrating series of disjointed phone calls, faxes and mail that may result in delay, lapses and duplication of effort.¹²

Care for patients with chronic conditions, especially congenital and childhood-onset diseases, is highly complex, involves non-medical professionals and requires significant involvement of patients and their family members. Coordination, characterized as a flow of information between providers to ensure that they all act toward a common goal, encompasses health services such as ambulatory, hospital, post-acute and home care, and social support services from state, community and private organizations.^13–16 Vendors have so far not adequately addressed the limitations of EHRs in supporting coordination beyond institutional boundaries.¹⁷ Professionals providing patient care and services therefore need to rely on other technologies to communicate¹⁸ and often turn to web-based applications to improve the process.

We conducted a qualitative study to gather and analyze insights from clinicians, coordinators and parents of children with multiple chronic conditions to formulate a preliminary model of care coordination intended to inform the design of electronic support tools. It models information needs and flow between participants, describes common technical and organizational barriers and outlines the complexities of information exchange that advanced software tools will need to support. Studies of similar scale and purpose are usually performed in the initial phase of a User-Centered Design (UCD) process¹⁹ as necessary precursors to functional and design specifications that are sensitive to appropriate fit to task and clinical context. UCD is established in software development as a robust, reliable process across many industries and has recently been required for HIT design under the 2014 Meaningful Use Act.^20,21

Methods

We interviewed twenty-two individuals at two institutions affiliated with a large urban healthcare delivery network in five sessions lasting an hour each. Two primary care physicians and one nurse practitioner were interviewed individually and a primary care physician, a social worker and a coordinator as a group. Sixteen parents of children with multiple chronic problems formed a focus group during their scheduled monthly meeting. The interviewer asked open-ended questions intended to elicit descriptive narratives about their routine tasks and activities in coordinating care for their children. Participants were encouraged to describe facilitators and barriers to effective care management, their most frequent modes of communication and to give examples of the types of documents they gather, receive, store or exchange with other parties, including patients and their families. The interview structure was gradually adapted over time to focus on salient themes emerging from preceding sessions and to follow up on specific topics. Interviewees were selected as a sample of convenience but attention was given to including providers from both hospital and community settings in order cover a broader set of personal perspectives.

All sessions were recorded, transcribed and analyzed using NVivo 10 software.²² Codes were developed and refined in three iterations by one investigator in which statements representing discrete meaningful concepts were either aggregated into broader categories based on similarity or differentiated into more descriptive units (axial and selective coding).²³ The final structure was reviewed and refined in two consecutive sessions in collaboration with a second investigator.

We analyzed all transcripts by using a constant comparative method²⁴ that involves identifying ideas in discrete statements (units of observation) and aggregating them into sets. This method produces categories derived from the participants’ expertise and language and those the researcher identifies as significant to the focus of inquiry (e.g., coordination, communication). The goal is to develop theoretical insights into processes that characterize the flow of information and construction of goals in a socio-technical system (STS).²⁵ The sociotechnical framework was used to interpret and analyze the findings as components of goal-oriented activity that encompasses humans, technology and artifacts.²⁶

Interview analysis

We identified and coded statements about roles, communication and information technology and characterized actors, technological agents and artifacts. For example, we analyzed accounts of compiling, abstracting, and interpreting information from visit notes and other sources routinely done by physicians and administrative staff in order to update patients on the course of treatment and to plan future tests and visits. Analysis of the parent focus group followed themes that were unique to their experience and those that provided their perspective on processes and issues identified in interviews with providers.

Results

Descriptive and analytical results from the interviews are described below. We identified main participants in communication and information exchange (parents, families, care and assistance providers), technology (EHR, email, mail, phone, fax) and goals and activities characterizing coordination (aggregation, abstraction, interpretation, visit planning, situational awareness, care planning, medications and problems). We also describe how these findings would inform the design of software tools for coordination.

Patients and families

Families with children who had three or more chronic problems or disabilities requiring frequent treatment and continuous care were assisted by coordinators whose task was to manage most communications with the care team, interpret and clarify medical findings provided by specialists and advise parents on the progress of care and next steps. For example, one coordinator described the treatment complexity of most patients as “having four to five subspecialists who are seen regularly every few weeks to every few months; almost all having physical, occupational or speech therapy.” Parents were regarded as the main source of information about previous care and as agents who maintained contacts to community-based assistance providers and mediated the forwarding and transferring documents between offices and agencies without direct formal affiliation. However, primary care providers and coordinators actively maintained communications and activities for those not able to carry out the tasks themselves.

Parents described similar levels of care complexity as did clinicians and coordinators although individual experiences varied. Several parents reported receiving care regularly from about ten providers, “six of them out of the system,” and worked extensively to maintain communication and timely updates. Their frequency of visits was typically “between six times a month to about five times every two months, with physical therapy sessions occurring several times per week.” Collectively, they expressed the feeling of “being intermediaries, especially for episodic care like ED visits,” or “being the clearinghouse for information,” closely echoing the perceptions of clinicians.

Care and assistance providers

Medical care providers were either co-located in one physical setting (e.g., a hospital or a clinical center) or had private offices in other locations, sometimes in a patient’s home community. A list of professional roles of care givers participating in coordination activities, as described in the interviews, is in Table 1. Those affiliated with the same care network shared an EHR system although some records such as emergency department visits, discharge summaries or specialty visit notes may have been stored on non-integrated systems and therefore available only to some clinicians. Primary care physicians (PCP) provided routine care and either communicated with a coordinator or assumed the responsibilities of a medical home. Nurses provided care at clinics, patient homes (visiting services) or at schools, therapy centers or other facilities and organizations. Providers not directly affiliated with the network generally maintained their own notes and gave occasional updates to a PCP. There were several coordinators of clinical care and of social services who worked at state agencies (e.g., Departments of Social Work, Early Intervention, Families and Children, etc.) or at community support centers, schools and other institutions. Their contact with PCPs and hospital-based care coordinators and social workers was primarily through email and telephone. Several parents noted that their PCP was not central to coordination and mostly “was there for your typical colds and well visits.” However, this arrangement required them to do more information forwarding and updating as “all specialists refer you back to your PCP.”

Table 1.

Roles of collaborating professionals

Role	Description
Primary Care Physician	Located in large care centers or in patient’s home community
Specialist Physician	Affiliated with hospital network or in independent clinics and offices
Nurse	Clinic, community, school, home care workers, visiting nurses
Therapists	Physical, occupational, speech, and specialized forms of therapy
Administrator	Supports PCPs and other clinicians in care management.
Coordinator	Services and care coordination at clinics, state and private agencies
Social Worker	Services and management at hospital centers and state agencies

Communication and information technology

Coordinators and providers relied primarily on email, phone and mail to reach others across institutional boundaries. These forms of communication were often inadequate, laborious and time consuming, requiring extra and duplicative effort to manually update electronic records and private notes. For example, information circulated by mail was not always received or reviewed in time to effectively support face-to-face conversations and maintain situational awareness. As one coordinator explained, “if I have a question, I’ll call the PCP and he says that we didn’t receive a letter from you although I sent it the day after I saw the patient. They may be too busy to follow up with us.” Specialists who did not have access to a shared EHR had difficulties forwarding information such as visit notes and records of prior care. They often scanned documents to a PDF and emailed them or used a fax or the mail.

Delays and other complications resulting from slow communication sometimes affected care decisions. Several parents whose children received immunizations, sick and other routine care from PCPs in their community and more extensive chronic and specialty care at hospital network centers decided to transfer all their care to the centers even if it entailed more travel and time. However, coordination and communication between the providers became less problematic as a result. A coordinator noted that “this may not be the case if we were able to communicate better.”

All interviewed professionals expressed a strong preference for a common electronic platform that would accommodate automated updates of such basic information as contact records and also clinical and other personal data for the entire team, including patients and their families. According to one social worker, “just knowing who is on the team is a huge challenge for parents and also for the vast army of involved people, the folks that help the children at home, physical therapists, early intervention people, the school nurse, probably the teachers and others.” This frustrating situation was described by a PCP who noted that “reaching out to the community side, agencies responsible for approving equipment the patient needs, finding out what letters need to go with that and recording it is enormously laborious.” A list of the most frequently used communication technology and its limits for documentation exchange, as described in interviews, is in Table 2.

Table 2.

Communication technology

Technology	Limitations
EHR System	Limited to affiliated providers only. Duplicate sets of data need to be maintained separately by other parties.
Email	Bridges institutional boundaries but is often a poor workflow fit. Group conversations are difficult to maintain and security and privacy is a concern.
Mail	Substitutes for lack of interoperability. Extra work is needed to abstract data for coded entries into an EHR.
Telephone	Poor support for group decision making. Synchronous conversations do not often fit into workflows.
Fax, PDF	Documents are mostly not searchable and need to be abstracted for EHR use.

Direct access to selected information by patients and family members was considered essential for reducing their dependence on mediation by clinicians, coordinators and social workers. For minors, elderly patients and those without computer skills, “care planning still has to be done the old fashioned way, either in person or on the phone,” according to one PCP. However, many others would benefit from direct access and could update or correct information and be more engaged in their own care.

Care coordination

Activities directly associated with care coordination that were discussed in the interviews were, a) aggregation of documents and information, b) abstraction of key events from charts for EHR entry, c) interpretation of medical findings and advice for patients, d) scheduling and tracking of referrals and visits, e) maintaining long-term situational awareness, f) development and updates of care plans and g) maintenance of medication, allergy and problem lists. The associated tasks and activities are summarized in Table 3 and described in detail in the sections below.

Table 3.

Activities associated with care coordination

Activity	Description
Aggregation	Compile records generated by clinicians, professionals and family across all settings.
Abstraction	Identify and summarize key points from extensive records of prior care.
Interpretation	Collate and interpret visit notes for patients, clarify results, requirements and next steps in care.
Visit planning	Initiate and track appointments, test, consultations and referrals.
Situational Awareness	Maintain awareness of new events, care progress, upcoming and missed visits.
Care Plan	Individual course of interventions, monitored for situational awareness.
Medications, Problems	Reconciliation of lists at hospital and community locations.

Aggregation

Before the first visit of a new or a transfer patient, administrators, nurses and coordinators collected all available electronic, paper and scanned documents they were able to obtain from prior and current providers. Patients or their families were central to this process as they often provided names, contact information and key dates although missing information still had to be identified and added. Some PCPs were able to “off-load much of this information gathering” to staff but for very complex patients or those with cognitive impairments clinicians had to “figure out who in their life are key people who could help with information.” However, this process was time consuming and incurred considerable delays as information arrived through the mail as paper documents or printouts of electronic notes from systems that were not interoperable. As one coordinator noted, “you may have seen the child a couple of times before the records arrive and that is incredibly frustrating for parents.”

Abstraction

Physicians often had to abstract, compare and summarize from multiple records core health problems, patient history, prior treatment and procedures and to identify information that is missing or requires updates and more detailed description. This work had to be completed by clinicians (often the PCP) rather than support staff. As a coordinating physician noted, “it means that I have to hunt through hundreds of pages and find what’s important although summaries may exist for some patients in a one or two page document that has all the key information in one spot.” Direct phone conversation with identified referring physicians can sometimes clarify “what the key issues, hospitalizations and problems are” if the record is too extensive to review.

Patients receiving complex care had a large volume of records of which only a fraction – those generated by providers from the same care network – were electronic. Very frequently, documents were forwarded across settings in scanned PDF files that could be “hundreds of pages long” and from which “immunization record, operative, advanced directive and other things I will need to search for” had to be extracted and the information entered in a coded form into the EHR. PDF documents were mostly not searchable automatically. Paper charts, forms, documents and even printouts from systems that were not interoperable were routinely scanned and entered into EHRs as attachments to clinical note entries.

Abstraction and comparison of documents from multiple providers was also done to reduce redundant or conflicting recommendations. A physician noted that “I want to know what the social worker has done so that I’m not asking the same battery of questions and recommending the same services already put in place.” As the notes were coming from many different sources “it took a long time to do our homework – there is no one place to read all of this.”

Interpretation

Clinicians and coordinators compiled and interpreted for patients and their families important findings and instructions from numerous specialist and procedures that were not well understood during the visit, either for their complexity or the sheer volume of information. Coordinators typically talked to a patient after three or more medical visits have taken place (often on a single day), “collating who they saw and what that person wanted to happen next” and to “look at the big picture of who recommended what, and to synthesize information.” This overview often generated further queries to the providers to confirm particular requirements or to get more detailed instructions. For example, a coordinator noted that “I can address any mismatches or ask questions – if the cardiologist said this patient needs to have a catheterization I can make sure that the patient understood it.” Patients or parents often asked coordinators to follow up with questions they did not raise during visits for lack of opportunity or understanding. A coordinator noted that “it took a lot of emailing behind the scenes to get a sense. For example, if a kid was started on new epileptics, are they supposed to be seizure-free or are we hoping for a 50% decrease? What’s the time period that you would expect this, what are the side effects and can he continue to go to school?”

Visit planning and referrals

Patients typically needed to schedule and keep regular appointments with four or more sub-specialists, therapists, social workers and other assistance providers. Coordinators helped patients navigate new environments such as large hospital and clinical centers so that they did not get “overwhelmed by the volume of information and moving parts” and called and emailed multiple providers to arrange visits so they to co-occurred on the same day in one location. In some cases, complex scheduling was arranged by a dedicated service center available at the hospital. When patients did not have access to a coordinator, their PCPs had fewer opportunities to call and schedule multiple visits, so often “the parents were pretty much on their own for making those appointments.” Knowing which appointments needed to take place, what procedures or tests will need to be done and coordinating planned care was essential for avoiding redundant visits or missing some due to conflicts.

Tracking visits and their outcomes was difficult, required careful planning and active outreach to providers to find out whether an appointment did take place and what were the next follow-up steps. For example, a coordinator noted that “a large proportion of all referrals never actually get completed and without proactively keeping track I can’t assume that they were done.” The EHR was considered to be only marginally useful in displaying scheduled and missed appointments as many specialists did not share the same system. A physician described the task as “actively, manually tracking what are the specialties and clinics, who needs to follow up with them and when.”

Reasons for missing appointments varied widely but coordinators felt that they could effectively intervene to prevent some of those if they were sufficiently informed of the situation. For example, a social worker noted that families needing transportation services from a state agency may “miss six appointments in a day if the required paperwork did not go through.” A more effective, integrated communication system may allow coordinators to notice such gaps and contingencies and act in time to make sure that medical care can be provided or rescheduled.

Situational awareness

Awareness of activity and communication taking place in a socio-technical system is essential for understanding how events and actions of all agents – human and technological – affect current personal and common objectives and future plans. The design of most EHRs generally does not allow clinicians to maintain this level of activity awareness, especially across different institutions.²⁷

A physician described updating others on the same care team about planned actions: “After a patient visit I usually talk to their various other providers to make sure we’re all on the same page.” There was no central place to “see the care team and when they want to see the patients next and what the active issues are.” The physician had to recreate the sense of the current state of care from “looking at notes,” a time-consuming task of aggregating and reading through electronic and paper records and emails. Coordination of tasks for very complex patients required in-person meetings and appointments with patients to “go through all of the specialty visits, answer questions, think ahead and troubleshoot any care deficits.”

Workarounds were devised by coordinators to keep them informed about unscheduled events, underscoring the inadequate support for this task by their EHRs. For example, the IT department developed by request an email service to automatically generate an alert message when patients on a tracking list were admitted, discharged or seen at the emergency department or same-day surgical units. Events and visits taking place outside of the hospital network, however, could be tracked only if patients reported them. Another repurposing of email to remind about future events, as reported by a physician, was to set up messages to arrive on days of important follow-up appointments and to “call afterwards to make sure that they went.” However, this strategy was not suitable for tracking a large number or frequent visits. The physician in such cases required patients to “come back and figure out if they’ve missed things – but I can’t be calling.” Sub-specialists would sometimes notify PCPs by email about missed care and then “somebody would reach out to the family.”

Poor situation awareness and event tracking may contribute to adverse events resulting from missing or delayed care. A physician reported that “one of my patients didn’t go to catheterization because they didn’t understand the conversation with the cardiologist and I didn’t know about it, then the patient got hospitalized.” Tracking abnormal radiographs and tests was also seen as a recurring problem with many opportunities for omission.

The lack of reliable, integrated information sharing between medical and non-medical care providers confounded situation awareness for the entire team. Most communications took place through email, over the phone or by postal service, requiring secondary manual documentation and updates. Non-medical events with the potential to negatively affect care were difficult to discover in time to intervene. For example, a physician may learn about a change in a patient’s social services or about serious incidents “only if the family informs me or if someone from DPH tries to contact me, but that hasn’t really happened.” In cases where children are “moved to a different home or foster parents” their local records may not be transferred with them. As one PCP noted, “when they come back in 2 or 3 years, I don’t know what happened, did they get their vaccines, etc. It takes digging into records and talking to the social workers and getting all that information.”

Coordinators or primary care physicians usually received information from social and community workers only “on demand” and in case of serious problems when “they will reach out to us to come up with a plan together.” Several clinicians thought that “sometimes folks in the community feel very disconnected from teams at the hospital” and that better access to shared records would benefit all involved and allow them to make better decisions. Some questions from patients could be directly answered by the most appropriate person rather than being mediated by physicians. For example, families now approach their primary care providers for “community related things like they need to know how to get a wheelchair, ramp, accessible housing, transportation, a better day program and a host of other non-medical things. Finding out which case worker needs to know that a request for equipment was filed is very time consuming.”

Communicating with patients is often done through secure email and then documented by “copying all the emails to the chart.” Some physicians felt that it is currently difficult to extend or route these conversations to those who can provide the best answers such as “administrators, nurses, social workers and the other people on the team.”

Parents noted that they often need to actively monitor their own care and recognize that certain events should take place which could be difficult for very complex treatment plans. An automated system of reminders about “appointments or blood draws that they have to do through email or something outside of their own scrap paper” was regarded as highly desirable.

Care plan

As the number of problems, providers, and sites increase, overall coordination of different provider-specific treatment plans, discipline-specific plans of care, and the production of a master care plan are needed.²⁸ Physicians providing primary and specialty care as well as therapists develop their own respective plans that need to be monitored for progress and completion of goals and adjusted if necessary. Core information from plans such as due dates, missed and achieved goals and new or unplanned events are important components of situational awareness. Coordinators needed to review and monitor this data periodically in order to take appropriate action. This function was approximated by reviewing notes to “figure out if they haven’t seen some person, are due for something, or there are upcoming things involving patient care.” The process had almost no support by information technology. A physician reported that “I can put one concrete piece of plan in a clinical message to my nurse and the nurse can call but if my plan is complex and evolves through a discussion then I need to do that myself.” Email and phone communication, however, was ineffective for team discussions and real-time updates as “there are so many points where any little thing can slip up that will erode that communication and it can be disastrous sometimes.”

Increased situational awareness may improve appointment scheduling, event prioritization and better responsiveness to emergent care needs. Such system, however, would need to bridge the existing electronic boundaries delineated by limited access to shared electronic information and replace the often poor workflow fit of email and phone communication with more appropriate forms. A parent noted that they felt as if long-term planning was not prioritized or well maintained by anyone. For example, ”I don’t feel that there is a management system for the long term things; critical care is well managed and followed up, but it’s up to me to figure out what needs to be taken care of next and send emails so that it doesn’t fall off the radar screen.”

Medication and problem lists

Lists of currently active and prior medications and problems were usually maintained by the primary care physician on an EHR. Care providers in the community such as visiting and school nurses administered the medications and kept their own records. Although dosing changes and substitutions need to be approved by a physician, central updates and reconciliation of lists was usually done at intervals coinciding with visits that sometimes substantially lagged behind actual events. As reported, many reconciliations took up patient visit time that would be otherwise focused on care issues. For example, patients may have a medication list from visiting and community nurses and the primary care physician “hopes that when they come in, it matches mine, but most of the time it doesn’t so the majority of the visit is spent reconciling medications.” Reconciling over the phone may “take 20 minutes or so” and fax communication may never close the confirmation loop if not returned in time.

Parents acknowledged that from their experience, reconciliation was a laborious and often imprecise process. For example, one parent reported that “it’s painful; they ask you to verify it and I see the same things that we told them the last time to take off the list. Something was prescribed two years ago by a specialist and then it just sits there on the list of your standard medications.”

Differences in separately maintained lists could be substantial if a primary care physician sees a patient once every 2–3 months. Updates accumulated over time between home, school and hospital discharge lists may take several hours to complete, especially with patients who may be taking 40 or more medications. A more efficient way to update lists electronically and to propagate changes to all when they occur was a common request by all interviewed coordinators and physicians.

Care coordination model

A web-based tool would need to support, at minimum, the following tasks and activities: increased situation awareness by allowing plan, visit and event tracking, allowing medication and problem list updates via secure email and embedded links, provide a two-way patient-facing portal for care instructions and requests and integrating updated information such as medication changes with the primary EHR after a reconciliation and confirmation process. We hypothesized that while it was unreasonable to expect that all participants would be willing to change their preferred workflows to navigate to a common website in order to communicate with others or update shared lists, they would likely respond to email prompts with one-click links to give and receive information updates that would be valuable for their own work.

A model of possible integration of a coordination tool with EHRs at separate institutions and with direct access from patients is in Figure 1. EHRs at hospitals and clinics can implement basic interoperability of medication, problem and contact updates through email links, state agencies may exchange electronic forms, community and visiting nurses can update medications through email automation and patients can review care instructions and educational materials online. Participants may copy emails to others on a shared server and update their contact information. Electronic forms can support semi-automated entry of coded information through text processing and confirmation.

Figure 1.

Model of possible integration with existing technology

Discussion

The core themes about barriers to effective care coordination that emerged from the interviews showed that lack of integration and system interoperability within and primarily across institutions and professions adds considerable effort to the work of most clinicians and providers of services. Real-time situation awareness that is indispensable for the coordination of actions and reduction of duplicative effort is difficult to derive from existing fragmented and mixed use of electronic and paper documentation. Compiling, tracking and following up on events across the medical-social services boundary was the most challenging part of coordination as the EHR – virtually the only shared electronic resource – does not extend outside of a network of affiliated clinics and offices.

Clinicians and assistance services managers disproportionately relied on paper or scanned documents when information had to be sent to others on the care team. However, it seems that many transactions (e.g., requests for standard social services or medical equipment approvals) used paper forms and collected fairly structured information that could be easily replicated in electronic form. We also noted that some tasks described as having no electronic support could in fact be accomplished with existing systems although some clinicians were not aware of the possibility.

Our results correspond with previous reports about inadequate EHR support for team-based chronic care management and that monitoring, plan feedback and assurance that plan components were completed are core requirements for effective coordination.²⁹ Our model, however, outlines the necessity to integrate non-medical professionals with the care team more closely. We also propose that improvement of EHR design alone can benefit only those clinicians who have access and that bridging communication gaps to care and service providers outside the network may require innovative approaches that integrate several technologies such as web and email-based forms and include interoperable connections between different EHR systems.

We formulated design advice and proposed a model primarily to support the work of care coordinators serving as liaisons between patients requiring complex care and the care team. Primary care physicians with responsibilities to maintain a medical home may also benefit from a similar system that is tightly integrated (e.g., as a module) with their EHR. They often take care of young adults who have transitioned from specialized pediatric care centers but may require as much coordination as before without adequate staff support.

Limitations

This study was conducted on a scale that may not be sufficiently large for generalization or validation of the coordination model as a comprehensive and predictive construct. It was intended, however, to give adequate understanding of the needs of patients, clinical and other professionals involved in the process of chronic care as a pre-requisite for the design of an advanced electronic coordination tool. Studies of similar size are commonly done in the first, planning stage of a user-centered design process and give informaticians and designers insights into unmet needs, communication and information flow complexities and requirements that the software will need to support and accommodate. Our methods and results should serve as valid reference and a comparison points to investigators and developers engaged in the design of similar electronic tools and looking for guidance.