Addressing Information Overload in the ICU

2016 - 1 February - Technological Game Changers
Brian R. Jacobs, MD; Ying Ling Lin, BEng, MEng; Patricia Trbovich, PhD; Anne-Marie Guerguerian, MD, PhD, FRCPC, FAAP
Experts address what innovative methods and technologies critical care practitioners can leverage to best approach information overload in the intensive care unit. See what approaches and technologies they identified as efficacious.
 
Health Information Technology in the Intensive Care Unit
Brian R.Jacobs, MD
 
Considering the critical state of intensive care unit (ICU) patients, health information technology has become a key resource required to improve care delivery, quality and research. During the past decade, Children’s National Health System in the Washington, DC, USA area has realized the benefits of digitizing its pediatric, cardiac and neonatal ICUs. The organization has implemented multiple electronic solutions in these areas, including computerized provider order entry, physician and nursing electronic documentation, smart infusion pumps, integrated communication tools, aggregated patient summary boards and biomedical device interoperability to the electronic health record (EHR).

These solutions have certainly improved the accuracy, completeness and legibility of the essential health data required for optimal care. In addition, care delivery efficiency, accessibility to the EHR anywhere and anytime, and the overall safety of our care delivery systems have benefited greatly from these changes. Furthermore, digitizing the healthcare environment in the ICU setting has permitted us to use large datasets for quality improvement efforts, educational opportunities, enhanced communication, clinician decision support and clinical research. 
 
Associated Challenges

These advances in health information technology have also introduced new challenges to our ICU enterprise. Voluminous additional data have contributed further to information overload in an already overwhelming environment. It is not unusual for EHR-based decision support to be introduced with low positive predictive value, which can result in busy clinicians overriding valuable information about their patients. Constant device alarming for trivial issues or false alarms can result in alert fatigue and difficulty separating important messages from overall noise levels. A recent study at Children’s National noted false-positive cardiopulmonary monitor alarm rates of 85%. Despite the many advances, safety events such as seeing the wrong patient or prescribing look-alike medications can still occur as they did in the paper-based environment. Information clutter and copying forward documentation can result in bloated daily progress notes and misinformation. Finally, clinicians have become so dependent on high-performing electronic systems for care delivery that there is now little tolerance when performance deteriorates or systems go down, situations that lead to potential safety and quality concerns. 

Addressing Challenges

During the past decade, we have learned much about health information technology. While the benefits undoubtedly outweigh the risks, we should not lose sight of the latter. The safe introduction and use of health information technology requires adherence to several important principles. Clinicians must be engaged in the design, testing and implementation of new platforms. The technology must be designed to meet the needs of the population being served (i.e., pediatric weight-based dosing, developmental considerations, staffing and work flow). Niche systems must be interoperable with the EHR to ensure that essential health data is easily accessible. In order to avoid alert fatigue, decision support should be integrated into the work flow, targeted to specific needs, actionable and effective. New technology implementations must be well supported both during and after implementation, with problems identified early and remediated quickly. After implementation, clinicians require ongoing support, and systems require continuous development and optimization. Finally, patient and clinician work flow and technology changes must be matched to ensure that they are compatible. An organization should never automate dysfunctional work flow.  

The benefits of health information technology can, and have been, achieved by many organizations across the globe through adherence to these principles and avoidance of well-documented historical misadventures.
 
 
A Case Study of the Adoption of an Innovative Monitoring Software in a Pediatric Critical Care Setting Facilitated by Human Factors Engineering
Ying Ling Lin, BEng, MEng, Patricia Trbovich, PhD, and Anne-Marie Guerguerian, MD, PhD, FRCPC, FAAP
 
In the field of medicine, the critically ill patient with multiple underlying diseases and conditions arguably stands to benefit the most from big data through customized, evidence-based decision-making. In the past, close approximation using epidemiologic data rarely existed for these patients(1) or was unavailable at the bedside where critical decisions are made. Advances in computer hardware and software make it easier to collect continuous monitoring data, instantly visualize large data sets and calculate derivative indicators in real time to predict future patient states and their associated interventions. These combined features can powerfully inform decision-making by providing contextual and relevant clinical information.

Currently, noncontinuous clinical information sources, including laboratory results, clinician notes and personal patient information, add to the sense of information overload.(2) Studies have indicated that poorly designed technologies in high-risk environments have led to incidents and accidents,(3) and software-related recalls of medical devices are on the rise.(4) In critical care, this could mean poor outcomes for patients and, in the worst case, mortality. The addition of vast arrays of continuous data from monitoring technologies to the clinical information environment risks aggravating the sense of information overload if the data is not managed appropriately. This may be partially addressed through well-designed software and the use of human factors engineering techniques used to adapt technologies to end users. Practically, software can be adapted to front-line critical care staff through careful study of the different clinical work flows contributing to team care.

To this end, three groups have formed a close collaboration to derive meaningful use from big data in pediatric critical care: the Hospital for Sick Children in Toronto, Ontario, Canada; the HumanEra Group at the Centre for Global eHealth Innovation; and Etiometry, Inc. The research team at the Hospital for Sick Children, led by intensive care physicians Peter Laussen and Anne-Marie Guerguerian, aims to develop, adapt and integrate monitoring technologies through different engineering solutions. The HumanEra group, led by Patricia Trbovich, has extensive knowledge and experience with translating findings from human factors studies to design recommendations for medical devices used in the Canadian public health system. Etiometry, Inc. is the up-and-coming developer of the T3 (tracking, trajectory and trigger) monitoring platform, which displays and derives meaning from multimodal monitoring signals through novel population-based and etiology-based algorithms.

Multiple perspectives of team care are collected and analyzed, and capabilities and constraints of the existing information technology infrastructure are gathered from database analysts, human factors engineers specializing in complex sociotechnical healthcare systems and product developers open to rapid adoption of design recommendations based on clinician needs. This collaboration takes the form of a multiphase project aiming to safely and effectively adapt the T3 monitoring platform to actual clinical work flow through the application of human factors engineering methods. Each phase of the project provides design recommendations for the T3 monitoring interface that are determined through the application of accepted design principles or safe testing in simulation laboratories with clinicians. As part of the iterative design cycle, if such revisions to the software are deployed in the critical care unit, a potentially practice-changing technology can be integrated in the safest manner possible.

Undeniably, critical care is technology driven and continually advances through cutting-edge monitoring discoveries. As these monitoring technologies find relevance in critical care, they increase the complexity of data management and push the human limits of clinical decision-making through the sheer plurality of data. These efforts to transition from an increasingly unmanageable data-driven environment to one in which technology aids in deriving meaning from data lead to two main benefits: first, providing real-time, evidence-driven critical care and, second, minimizing clinician information overload, thus leading to improved workflow. Collaborations between clinician end users and technology developers can be efficiently facilitated by neutral human factors groups to ultimately develop technologies for critical care that are proven to be safe and effective with minimal risk to patients.

 References:

1. Wheeler DS, Wong HR, Shanley TP, eds. Pediatric Critical Care Medicine: Care of the Critically Ill or Injured Child. London, UK: Springer-Verlag; 2014.
2. Manor-Shulman O, Bevene J, Frndova H, Parshuram CS. Quantifying the volume of documented clinical information in critical illness. J Crit Care. 2008;23(2):245-50.
3. Thimbleby HP, Oladimeji P, Cairns P. Unreliable numbers: error and harm induced by bad design can be reduced by better design. J R Soc Interface. 2015 Sep 6;12(110): 0685.
4. Simone LK. Software-related recalls: an analysis of records. Biomed Instrum Technol. 2013 Nov-Dec;47(6):514-522.