Pediatric emergency medicine trisk 23

CHAPTER 3 ■ CLINICAL DECISION SUPPORT
JOSEPH J. ZORC, NAVEEN MUTHU

Clinical decision support (CDS), broadly considered, encompasses any
system, device, or document that is intended to support decision making by
any individual involved in providing or receiving health care. CDS has
existed as long as clinical decisions have been made, and covers a wide
variety of tools in the acute pediatric setting ranging from algorithms such
as resuscitation guidelines or clinical pathways, devices such as
measurement tapes for dose calculation, or complex smartphone
applications to make diagnoses or apply clinical decision rules at the
bedside. More recently, CDS has become a focus within the field of clinical
informatics, where it has been defined by Osheroff as content that “provides
clinicians, staff, patients, or other individuals with knowledge and personspecific information, intelligently filtered or presented at appropriate times,
to enhance health and health care.” Academic research and quality
improvement work has often focused on CDS within the electronic health
record (EHR), given its widespread recent adoption and central role within
systems of clinical care. However, the EHR is only one of many channels
available to provide CDS.

PEOPLE, PROCESS, THEN TECHNOLOGY
When considering an opportunity to use CDS to improve clinical care, the
first step is to develop a team and identify the appropriate individuals who
can best lead, inform, implement and sustain the project. An adage within
clinical informatics is the importance of focusing on people and process
analysis before launching a proposed technology solution. Much as erecting
a building relies on a solid foundation, good CDS rests on an informed
understanding of the individuals who will interact with the potential system,
including their knowledge, attitudes, beliefs and limitations. These team
members can then map out the current state of the clinical process to
provide a baseline for proposed interventions. In parallel, understanding the

goals of stakeholders and leadership will create champions and investment
in resources that will be critical to CDS implementation and sustenance.


One of these stakeholders is the local information services team and
understanding the availability of knowledge and resources within this group
may provide insight into important technical constraints that affect which
CDS intervention is chosen.
A systematic comprehensive approach to designing CDS in the acute
pediatric setting has been well described in Sheehan et al., a project
implementing EHR-based CDS to support a decision rule to reduce
unnecessary imaging for children with head injury. This project conducted a
formal “sociotechnical analysis” including focus groups, workflow
observation, and interviews to map out themes related to how staff interact
with the EHR while providing care for these children. This “gold standard”
example from researchers within the Pediatric Emergency Care Applied
Research Network (PECARN) provides an excellent reference for themes
that are common to many acute care clinical processes. However, when less
resources are available, a well-led team can often do an effective job in a
few meetings by bringing the right team members together to map the
process on a blackboard. Often, producing a “fishbone diagram” identifying
themes or a workflow map listing the steps in the process is an effective
approach in a group meeting. A “driver diagram” of key primary and
secondary themes can identify key issues that may need to be addressed
within or outside of technology systems.

CDS OPPORTUNITIES ACROSS A HEALTHCARE
ENCOUNTER
Determining the right approach within in a clinical process to provide CDS
is another key decision. Often CDS designers may fall back on familiar

tools such as pop-up alerts that may be the first option to come to mind
given their ubiquitous presence in EHRs. However, in isolation, these popup alerts may be one of least effective methods of CDS. Prior to selecting
an intervention, the team should consider the full breadth of options for
CDS within and outside of the EHR. Osheroff and others have developed
conceptual models for potential CDS opportunities across the span of a
healthcare encounter, from before a patient arrives to after discharge.
Fortunately there are examples of each of these opportunities applied to
acute pediatric care available in the literature (see Table 3.1 ).


Pre-Visit
Prior to a visit, information can be provided actively through referral from
another provider or gathered passively from systems such as the EHR or
emergency medical services (EMS) systems. As an example, Dandoy
describes how one pediatric ED group developed a process for oncology
providers to provide recommendations in advance of an ED visit for a child
referred for fever. These recommendations were implemented at the time of
referral using an order set. The lead time provided by early ordering greatly
enhanced timeliness so that antibiotics could be administered rapidly after
arrival.

Assessment and Plan Formulation
During the initial assessment phase multiple opportunities exist for CDS in
the workflows of various team members. For the PECARN head injury
decision rule project described above, these opportunities were leveraged to
gather information about risk factors through as many avenues as possible,
including nursing triage and physician data entry. A key theme identified in
the project was the importance of delivering the CDS as early as possible,
ideally prior to the provider’s evaluation of the patient, as they may have
already discussed imaging with the family and made the decision by the

time they used the EHR for order entry. When multiple team members enter
data to trigger CDS, this introduces the issue of validity and potential
disagreement between those sources, and the need to update them based on
additional information, which was also addressed in this project.
Documentation tools such as note templates that may be used by providers
during history-taking can provide a useful noninterruptive and easily
implemented technology method to guide data collection and provide links
to CDS resources. However, clinicians may often wait to document until
after clinical care is delivered, limiting its timeliness.


TABLE 3.1
CLINICAL DECISION SUPPORT APPLIED ACROSS THE
PHASES OF PEDIATRIC ACUTE CARE