Difference between revisions of "Dashboard"

From Clinfowiki
Jump to: navigation, search
(Design Considerations)
 
(13 intermediate revisions by one user not shown)
Line 1: Line 1:
A '''clinical dashboard''' summarizes information from an associated electronic health record (EHR) relating the real-time condition of a patient cohort in a visually concise and usable interface. Dashboards provide feedback to clinicians and managers '''''at the point of care''''' for clinical decision support.
+
A '''Clinical Dashboard''' summarizes information from an associated electronic health record (EHR) relating the real-time condition of a patient cohort in a visually concise and usable interface. Dashboards provide feedback to clinicians and managers '''''at the point of care''''' for clinical decision support.
  
'''Quality Dashboards''' or '''Health System Dashboards''' provide feedback on entire departments or enterprises. They can display individual patient, provider, or aggregate metrics (such as performance or quality indicators) for the purposes of improving healthcare quality, efficiency, safety, or revenues.
+
'''Quality Dashboards''' or '''Health System Dashboards''' provide feedback on entire departments or enterprises. They can display individual patient, provider, or aggregate metrics (such as performance or quality indicators) for the purposes of improving healthcare quality, efficiency, safety, or revenues. These operational dashboards provide an at-a-glance view to help managers quickly uncover problems and make immediate decisions.
  
 
== History and Terminology ==
 
== History and Terminology ==
Dashboards were developed in the business sector as a way to "summarize and integrate key performance information across an organization into a visual display as a way of informing operational decision making". <ref name="Dowding 2015">Dowding D, Randell R, Gardner P, et al. Dashboards for improving patient care: review of the literature. Int J Med Inf. 2015;84(2):87-100.</ref>  
+
Dashboards were developed in the business sector as a way to "summarize and integrate key performance information across an organization into a visual display as a way of informing operational decision making". <ref name="Dowding 2015">Dowding D, Randell R, Gardner P, et al. Dashboards for improving patient care: review of the literature. Int J Med Inf. 2015;84(2):87-100.</ref>
  
The terms "registry" and "dashboard" are often used interchangeably in the literature. However, the term "clinical data registry" is usually in reference to viewing a retrospective patient cohort for the purposes of research or quality improvement. The data collected by a registry can have multiple sources, providing information about a patient's longitudinal care. The information is not necessarily displayed with an interface intending on real-time decision making.  
+
The terms "registry" and "dashboard" are often used interchangeably in the literature. However, the term "clinical data registry" is usually in reference to a retrospective patient cohort comprising a database for the purposes of research or quality improvement. The data collected by a registry can have multiple sources, providing information about a patient's longitudinal care. The information is not necessarily displayed with an interface intending on real-time decision making.  
  
''Clinical dashboards'' provide point-of-care feedback to administrators, physicians, nurses, pharmacists, and other ancillary staff.
+
''Clinical dashboards'', on the other hand, provide operational/point-of-care feedback to physicians, nurses, pharmacists, administrators, and other ancillary staff.
  
 
== End-Users of Dashboards ==
 
== End-Users of Dashboards ==
Line 15: Line 15:
 
* Hospital administrators for patient placement and flow.
 
* Hospital administrators for patient placement and flow.
 
* Physicians for monitoring their patient panels, whether in a hospital unit or outpatient clinic.
 
* Physicians for monitoring their patient panels, whether in a hospital unit or outpatient clinic.
* Nursing to monitor their patient panels and nursing-specific workflows
+
* Nursing to monitor their patient panels and nursing-specific workflows.
 
* Pharmacists to assess medication use.
 
* Pharmacists to assess medication use.
 
* Care managers for coordinating discharge and follow up.
 
* Care managers for coordinating discharge and follow up.
Line 21: Line 21:
  
  
== Use Cases and Examples ==
+
== Potential Use Cases and Examples ==
 
=== Outpatient Clinic ===
 
=== Outpatient Clinic ===
*'''Primary care physician''' who wants to identify high-risk or high-cost patients from their patient panel that may benefit from close follow-up or care coordination.
+
*'''Primary care physicians''' can identify patients from their panel that are high-risk, that have a recent ED visit, or that have recent admission.
*'''Specialty clinician''' who wants view the HbA1c and medication refill status of his diabetic patients from a single screen.
+
*'''Specialty clinicians''' can view the HbA1c and medication refill status of his diabetic patients from a single screen.
*'''Medical assistant & schedulers''' can manage the check-in, rooming, and check-out status of clinic visits
+
*'''Medical assistants & Clinic Managers''' can manage the check-in, rooming, and throughput status of clinic visits.
  
=== Hospital ===
+
=== In-Hospital ===
Acute Care Physician: Monitoring assigned patient list, rounding support
+
*'''Acute care physicians''' can track quality measures and identify patients at high risk for AKI or line infections.
Intensive care unit physician: Monitoring for compliance with ventilator and line saefty bundles
+
*'''Intensivists''' can monitor their unit’s compliance with safety bundles, such as for ventilators and lines.
Emergency Department physician:
+
*'''Emergency Departments''' can display a patient-level table of triage information, bed status, provider assignments, resulting labs, and more. An aggregate dashboard can also display throughput metrics and overcrowding information.
Surgeon:
+
*'''Surgical staff''' can similarly use dashboards to display scheduling and throughput information
Nurse:
+
*'''Nursing''' can see a unit-wide visual representation of bed status, provider contacts, and equipment availability.
Health Unit Coordinator (HUC):
+
*'''Hospital command centers''' can monitor throughput and incorporate predictive modeling to optimize census levels while preventing overcrowding.
 +
*'''Pharmacists''' can monitor for potential adverse drug events, high-risk medication scenarios, supply issues, and to safeguard against over-use as with antibiotic stewardship programs.
  
Pharmacist: Potential adverse drug event dashboard to highlight high-risk medication scenarios. for potential errors, supply issues, and to safeguard against over-use as with antibiotic stewardship programs.
+
=== Population Health ===
 +
*'''Coordinated Care Organizations (CCOs)''' can compare provider performance. Risk stratification with predictive modeling can help to identify high-cost patients that may benefit from care coordination. The multi-disciplinary care team can use the interface to coordinate their actions and communicate in one place.<ref name="Sharp 2018">Sharp J. Registries in Accountable Care. Agency for Healthcare Research and Quality (AHRQ); 2018.</ref>
  
Hospital Administration: Command center
+
=== In-depth: Emergency Department throughput dashboard at The University of Texas Health Science Center at Houston ===
Physician leadership: High-risk patient dashboards displaying those with elevated creatinine, positive line cultures. Quality
+
Franklin et al. describe the design, implementation, and evaluation of a real-time throughput dashboard for improving efficiency within their emergency departments. In pre-planning, they considered supporting situational awareness and optimal time visualization as top priorities. They sought to “improve throughput in-the-moment by focusing on breaking down bottlenecks within stages of care and improving the turnaround for the individual patient.” Throughput was measured as the “duration of time within an interval or stage of care”.  Temporal events (timestamps) in the EHR were mapped to meaningful stages of care. Each stage was given its own throughput time goal.
  
=== Population Health ===
+
In considering the dashboard visualizations, they wanted to highlight actionable information, including identification of bottlenecks and individuals who are approaching or are over time thresholds for a stage of care. To integrate the dashboard into workflows, providers were shadowed/observed to develop a work domain ontology (WDO). The dashboards themselves were developed in Tableau. Patients and their throughput status are represented by colored symbols. In addition to throughput time data, the displays also support providers with census info, laboratory and imaging order status, and individual physician workloads.
Coordinated Care Organizations:
+
 
Public health departments:
+
The prototype dashboard was refined during presentations and focus groups at multiple EDs. During implementation, multiple surveys and sample observations were conducted.<ref name="Franklin 2017">Franklin A, Gantela S, Shifarraw S, et al. Dashboard visualizations: Supporting real-time throughput decision-making. J Biomed Inform. 2017;71:211-221.</ref>
  
 
== Effects on Workflows ==
 
== Effects on Workflows ==
 
=== Benefits of clinical dashboard utilization ===
 
=== Benefits of clinical dashboard utilization ===
* Reduced information overload, simplification of clinical data
+
* Reduced information overload.
* Improved ability to find information effectively
+
* Improved ability to find information effectively.
 
* Provides clinical decision support, guiding clinicians on the priority of care and on the priorities of the organization.
 
* Provides clinical decision support, guiding clinicians on the priority of care and on the priorities of the organization.
* Facilitates communication between providers caring for common patients
+
* Facilitates communication among multiple providers and multidisciplinary team members caring for common patients.
  
 
=== Unintended consequences of clinical dashboard utilization ===
 
=== Unintended consequences of clinical dashboard utilization ===
* Tunnel vision: Focusing on the metrics that are displayed, while ignoring other unmeasured indicators
+
* Tunnel vision: Focusing on the metrics that are displayed, while ignoring other unmeasured indicators.
 
* Measurement fixation: Focusing on improving the quantitative representation of performance and quality (meeting the target number), rather than on the overarching purpose of care.<ref name="Dowding 2015" />
 
* Measurement fixation: Focusing on improving the quantitative representation of performance and quality (meeting the target number), rather than on the overarching purpose of care.<ref name="Dowding 2015" />
 
  
 
== Design Considerations ==
 
== Design Considerations ==
The presentation of information affects the decisions of end-users.
+
The presentation of information affects the decisions of end-users.<ref name="Dowding 2015" />
  
 
==== Visualization ====  
 
==== Visualization ====  
Additional information about the urgency of data can be imparted using color coding, or with symbols such as warning signs or traffic lights. Recent trends can be simplified on a dashboard display using spark-lines.  
+
Additional information about the urgency of data can be imparted using color coding, or with symbols such as warning signs or traffic lights. Recent trends can be simplified on a dashboard display using spark-lines.<ref name="Martinez 2018">Martinez DA, Kane EM, Jalalpour M, et al. An Electronic Dashboard to Monitor Patient Flow at the Johns Hopkins Hospital: Communication of Key Performance Indicators Using the Donabedian Model. J Med Syst. 2018;42(8):133.</ref>
  
 
==== Accessibility ====   
 
==== Accessibility ====   
An emergency department may have a large screen permanently displaying the ED census and important metrics, such as abnormal vitals, and bed assignments, and notes from staff. A dashboard can be a screensaver or the first page when logging into an EHR. Studies of dashboard implementation have shown that difficult or inconvenient accessing is a major impediment to their usage. <ref name="Dowding 2015" />  
+
An emergency department may have a large physical screen permanently displaying the ED census and important metrics, such as triage info and bed assignments. A dashboard can be a screensaver or the first page when logging into an EHR. Studies of dashboard implementation have shown that difficult or inconvenient accessing is a major impediment to their usage. <ref name="Dowding 2015" />  
  
 
==== Interaction ====
 
==== Interaction ====
The dashboard should allow for convenient linking to more detailed information, by allowing the user to hover for expanded data or click to access that element in the patient's chart.
+
The dashboard should allow for convenient linking to more detailed information, by allowing the user to hover for expanded data, clicking to drill-down, or clicking to access an element in the patient's chart.<ref name="Franklin 2017" />
 
+
==== Risk Scoring ====
+
 
+
  
 +
==== Predictive Modeling ====
 +
Risk scoring algorithms or risk stratification can make use of multiple patient metrics to provide a visual or numerical representation of a patient’s risk for poor outcomes, including mortality, care utilization, or re-admission.<ref name="Sharp 2018" />
  
 
== References ==
 
== References ==

Latest revision as of 11:47, 21 October 2019

A Clinical Dashboard summarizes information from an associated electronic health record (EHR) relating the real-time condition of a patient cohort in a visually concise and usable interface. Dashboards provide feedback to clinicians and managers at the point of care for clinical decision support.

Quality Dashboards or Health System Dashboards provide feedback on entire departments or enterprises. They can display individual patient, provider, or aggregate metrics (such as performance or quality indicators) for the purposes of improving healthcare quality, efficiency, safety, or revenues. These operational dashboards provide an at-a-glance view to help managers quickly uncover problems and make immediate decisions.

History and Terminology

Dashboards were developed in the business sector as a way to "summarize and integrate key performance information across an organization into a visual display as a way of informing operational decision making". [1]

The terms "registry" and "dashboard" are often used interchangeably in the literature. However, the term "clinical data registry" is usually in reference to a retrospective patient cohort comprising a database for the purposes of research or quality improvement. The data collected by a registry can have multiple sources, providing information about a patient's longitudinal care. The information is not necessarily displayed with an interface intending on real-time decision making.

Clinical dashboards, on the other hand, provide operational/point-of-care feedback to physicians, nurses, pharmacists, administrators, and other ancillary staff.

End-Users of Dashboards

  • Healthcare organizations (ie hospitals, clinics, and government bodies) to monitor performance and quality indicators.
  • Hospital administrators for patient placement and flow.
  • Physicians for monitoring their patient panels, whether in a hospital unit or outpatient clinic.
  • Nursing to monitor their patient panels and nursing-specific workflows.
  • Pharmacists to assess medication use.
  • Care managers for coordinating discharge and follow up.
  • Ancillary staff for scheduling, transport, and more.


Potential Use Cases and Examples

Outpatient Clinic

  • Primary care physicians can identify patients from their panel that are high-risk, that have a recent ED visit, or that have recent admission.
  • Specialty clinicians can view the HbA1c and medication refill status of his diabetic patients from a single screen.
  • Medical assistants & Clinic Managers can manage the check-in, rooming, and throughput status of clinic visits.

In-Hospital

  • Acute care physicians can track quality measures and identify patients at high risk for AKI or line infections.
  • Intensivists can monitor their unit’s compliance with safety bundles, such as for ventilators and lines.
  • Emergency Departments can display a patient-level table of triage information, bed status, provider assignments, resulting labs, and more. An aggregate dashboard can also display throughput metrics and overcrowding information.
  • Surgical staff can similarly use dashboards to display scheduling and throughput information
  • Nursing can see a unit-wide visual representation of bed status, provider contacts, and equipment availability.
  • Hospital command centers can monitor throughput and incorporate predictive modeling to optimize census levels while preventing overcrowding.
  • Pharmacists can monitor for potential adverse drug events, high-risk medication scenarios, supply issues, and to safeguard against over-use as with antibiotic stewardship programs.

Population Health

  • Coordinated Care Organizations (CCOs) can compare provider performance. Risk stratification with predictive modeling can help to identify high-cost patients that may benefit from care coordination. The multi-disciplinary care team can use the interface to coordinate their actions and communicate in one place.[2]

In-depth: Emergency Department throughput dashboard at The University of Texas Health Science Center at Houston

Franklin et al. describe the design, implementation, and evaluation of a real-time throughput dashboard for improving efficiency within their emergency departments. In pre-planning, they considered supporting situational awareness and optimal time visualization as top priorities. They sought to “improve throughput in-the-moment by focusing on breaking down bottlenecks within stages of care and improving the turnaround for the individual patient.” Throughput was measured as the “duration of time within an interval or stage of care”. Temporal events (timestamps) in the EHR were mapped to meaningful stages of care. Each stage was given its own throughput time goal.

In considering the dashboard visualizations, they wanted to highlight actionable information, including identification of bottlenecks and individuals who are approaching or are over time thresholds for a stage of care. To integrate the dashboard into workflows, providers were shadowed/observed to develop a work domain ontology (WDO). The dashboards themselves were developed in Tableau. Patients and their throughput status are represented by colored symbols. In addition to throughput time data, the displays also support providers with census info, laboratory and imaging order status, and individual physician workloads.

The prototype dashboard was refined during presentations and focus groups at multiple EDs. During implementation, multiple surveys and sample observations were conducted.[3]

Effects on Workflows

Benefits of clinical dashboard utilization

  • Reduced information overload.
  • Improved ability to find information effectively.
  • Provides clinical decision support, guiding clinicians on the priority of care and on the priorities of the organization.
  • Facilitates communication among multiple providers and multidisciplinary team members caring for common patients.

Unintended consequences of clinical dashboard utilization

  • Tunnel vision: Focusing on the metrics that are displayed, while ignoring other unmeasured indicators.
  • Measurement fixation: Focusing on improving the quantitative representation of performance and quality (meeting the target number), rather than on the overarching purpose of care.[1]

Design Considerations

The presentation of information affects the decisions of end-users.[1]

Visualization

Additional information about the urgency of data can be imparted using color coding, or with symbols such as warning signs or traffic lights. Recent trends can be simplified on a dashboard display using spark-lines.[4]

Accessibility

An emergency department may have a large physical screen permanently displaying the ED census and important metrics, such as triage info and bed assignments. A dashboard can be a screensaver or the first page when logging into an EHR. Studies of dashboard implementation have shown that difficult or inconvenient accessing is a major impediment to their usage. [1]

Interaction

The dashboard should allow for convenient linking to more detailed information, by allowing the user to hover for expanded data, clicking to drill-down, or clicking to access an element in the patient's chart.[3]

Predictive Modeling

Risk scoring algorithms or risk stratification can make use of multiple patient metrics to provide a visual or numerical representation of a patient’s risk for poor outcomes, including mortality, care utilization, or re-admission.[2]

References

  1. 1.0 1.1 1.2 1.3 Dowding D, Randell R, Gardner P, et al. Dashboards for improving patient care: review of the literature. Int J Med Inf. 2015;84(2):87-100.
  2. 2.0 2.1 Sharp J. Registries in Accountable Care. Agency for Healthcare Research and Quality (AHRQ); 2018.
  3. 3.0 3.1 Franklin A, Gantela S, Shifarraw S, et al. Dashboard visualizations: Supporting real-time throughput decision-making. J Biomed Inform. 2017;71:211-221.
  4. Martinez DA, Kane EM, Jalalpour M, et al. An Electronic Dashboard to Monitor Patient Flow at the Johns Hopkins Hospital: Communication of Key Performance Indicators Using the Donabedian Model. J Med Syst. 2018;42(8):133.


Submitted by Jonathan Sachs