ICU Management & Practice, ICU Volume 14 - Issue 1 - Spring 2014

Creating the ICU of the Future: A Day of Innovation

share Share

Peter J. Pronovost, MD, PhD, FCCM1-3

Clifford S. Deutschman, MS, MD, FCCM4,9

Kathryn E. Roberts, MSN, RN, CNS5

Alan D. Ravitz, PE6

Bradford D. Winters, PhD, MD, FCCM1,2

Jo M. Leslie, MBA1

Douglas S. Solomon, PhD, MPH7

George W. Bo-Linn, MD, MPH, FACP8

Adam Sapirstein, MD1,2

for the Johns Hopkins Armstrong Institute ICU Innovation Consortium

Armstrong Institute for Patient Safety and Quality, Johns Hopkins Medicine, Baltimore, USA.1; School of Medicine, The Johns Hopkins University, Baltimore, USA.2; Bloomberg School of Public Health, The Johns Hopkins University, Baltimore, USA3; The Society of Critical Care Medicine, Prospect, U.S.A.4; American Association of Critical-Care Nurses, Aliso Viejo, USA5; Johns Hopkins Applied Physics Laboratory, Laurel, USA6; IDEO, Palo Alto, USA7; The Gordon and Betty Moore Foundation, Palo Alto, USA8; Perelman School of Medicine, University of Pennsylvania, Philadelphia, USA9

[email protected]


In this manuscript, we describe the process and results of a workshop to apply a no pre-conceived notions, outside-the-box approach to design an ideal ICU that we can begin implementing now. This design effort largely focused on technology, but is mindful that the technology must be used by clinicians and will be embedded in work processes and culture.



More than a decade has passed since the Institute of Medicine issued its report, Crossing the Quality Chasm: A New Health System for the 21st Century (Institute of Medicine 2001). In the interim, facets of the United States healthcare system have changed. A concerted effort to deliver quality care has decreased central line-associated bloodstream infections (CLABSI) in intensive care units (ICU) nationwide (Pronovost et al. 2011; AHRQ 2012). Overall, hospitalised patients continue to experience preventable harm, such as hospital-acquired infections, and receive therapies they neither wanted nor needed. Empirical evidence suggests that we can do better.


Hospitalised patients are exposed to harm in a myriad of distinct ways. A subtle but equally devastating harm is disrespectful or undignified care, in which patients and families are excluded from the decision-making process or from information essential to understand the course of care (Institute of Medicine 2001). The historic approach to managing harm has largely relied on the clinician’s heroism, requiring them to recognise and immediately intervene when harm is imminent. This approach is inefficient, unreliable, and inferior to a safe systems approach that prevents many potentially harmful situations from arising. In the ICU, a patient receives almost 200 interventions daily (Donchin et al. 1995), most directed at avoiding the dozen or more consequential harms they are at risk of suffering (Unpublished data; Armstrong Institute workshop, December 7, 2012). Missing any one intervention could have catastrophic results. Yet, our healthcare system relies on the memory and vigilance of busy clinicians, rather than technology, to ensure that all of these therapies are delivered.


Technology has not provided healthcare with the safety or productivity benefits that other highly technological industries have experienced. In part, this shortcoming reflects a limited use of systems engineers to design care systems that support clinicians and protect patients. It also reflects a failure to integrate a composite of innovative technologies into an intelligent information ecosystem, in which the electronic medical record (EMR) is one part of the system. For example, an aeroplane cockpit is dramatically safer today. Thirty years ago a dizzying array of instruments bombarded pilots with vast quantities of unfiltered information that was difficult to assimilate. Today pilots rely on automation, and only receive context-sensitive information that provides exactly what he or she needs to make wise decisions.


Today an ICU is far more complex and potentially dangerous than 30 years ago. Each room is filled with more monitors, wires, and devices that typically do not communicate with each other or with clinical information systems. Clinicians are the safety net for patients. Clinicians are piloting a 30-year-old cockpit, bombarded with more data and left to prioritise or make sense of the information. Few tasks are automated. Despite significant investments, information technology does not help clinicians identify or predict patients at risk for harm. The systems do not describe what therapies are appropriate nor monitor the administration of needed treatments. Most tellingly, information technology lacks intelligence to learn or offer guidance that would improve global outcomes. Finally, patients and their families have not been actively engaged in the hospital setting, leaving them feeling isolated with inadequate information and unmet needs. The desired goal of truly personalised medicine, beyond genetic testing (Behrens et al. 2008), is not imminent.


A systems approach to improve ICU care needs to address culture, people, processes, and technology. Substantial efforts have explored interventions, such as the Comprehensive Unit-based Safety Program, to improve ICU culture and teamwork. Minimal work has explored how to ensure that technology serves clinicians, rather than making clinicians support technology.



Researchers from the Armstrong Institute for Patient Safety and Quality (AI), an entity of Johns Hopkins Medicine, hosted a one-day workshop on December 7, 2012. A diverse group of 69 people representing patients and families, clinicians, patient safety researchers, hospital administrators, professional society leaders, and device manufacturing leaders participated in the workshop (See Appendix 1 [on website,] for a list of participants and their affiliations). The workshop format followed an innovation process. The overarching goal was to design an ideal ICU environment that eliminated preventable harms, optimised patient outcomes and the patient and family experience, and reduced waste in healthcare delivery. We identified four tasks or objectives based on this goal:

1.Better understand the needs of patients, family members, and caregivers in the ICU.

2.Work in multidisciplinary teams to brainstorm innovative solutions and build a system that met patient, family, and caregiver needs.

3.Include in that system a mechanism to identify new or unanticipated needs.

4.Ensure that the system is flexible enough to address new or unanticipated needs.


Strategy – The Innovation Process

We used an innovation process called design thinking that was created by IDEO (Ravitz et al. 2013), a global consultancy that takes a humancentred, design-based approach to help organisations innovate and solve problems ( The workshop had three main phases: inspiration, ideation, and implementation (Figure 1).



The work was grounded in a presentation highlighting how current ICU medicine and healthcare in general are wasteful of human life, labour and money. The presenter (PJP) recounted real stories of young lives lost in ICUs because systems were not integrated. These stories anchored participants’ perspectives in the experiences of patients, families, and clinicians. The presentation included examples of how care can be unintentionally disrespectful to patients, and also to clinicians, who struggle with poorly designed systems and ever-increasing workloads, prompting some to eventually leave the ICU workforce.


The presenter described ways current ICU systems cannot handle the complexities inherent in caring for critically ill patients, and how this overburdens clinicians. Workshop participants were challenged to understand the demands and the often obscure interactions that connect seemingly independent systems in an ICU and to develop a new ICU that met the needs of patients, family members, and clinicians (Stokols et al. 2008; Pronovost and Bo-Linn 2012).


The presentation was followed by a visit to seven ICUs (adult: medical, coronary care, general surgery, cardiac surgical, trauma, neurosurgical, surgical oncology, and paediatrics) at The Johns Hopkins Hospital. Workshop participants also visited several patients who were recently discharged from an ICU to an intermediate care unit. Appropriate permission was obtained and Health Insurance Portability and Accountability Act compliance maintained. The participants were briefed for 30 minutes on observation methods, given an observation guide (Appendices 2a , 2b2c, available on the website), and asked to write down their observations. Careful observation while immersed in a real world environment plus context driven inquiry of individuals functioning within the system is a common method to deconstruct complex systems and then reconstruct innovative alternatives (Gurses et al. 2009).


To obtain a broad range of observations in a short period of time, workshop participants were divided into three groups. Each group was assigned a specific area of observation:

• People (e.g., communication, human interactions, human-machine interaction), including interviews with patients and families;

• Technology and physical space (e.g., available space, ICU environment, monitors, alarm systems, other technologies, EMR);

• Processes and workflows (e.g., tasks, work processes, sequencing of tasks, interdependency of tasks).


Observers worked in pairs, with each member from a different specialty or area of interest. Assignments forced participants to observe and evaluate unfamiliar environments, activities or situations. For example, we asked ICU physicians and ICU nurses to interview patients recently transferred to a step-down inpatient unit, and we asked technologists and engineers to observe people rather than devices. All participants reported that this approach brought another perspective and provided novel insight.


After visiting the ICUs, participants were organised into seven teams of 8 to 10 people. We constructed teams that represented each field, each ICU visited, and each observation assignment. Findings were shared with other team members, summarised, and grouped based on similar themes. The teams then named each group of observations based on common characteristics. This exercise helped participants organise their observations of people (patients, health workers, and families), technology, space, and processes around themes that impressed them as key determinants of current paradigms in the ICUs.



In this next phase, teams chose two themes or areas for improvement in the ICU that they felt provided ample opportunities for creativity and impact. One facilitator was assigned to each team to help develop these themes into broad discussion topics that began with a “How might we…?” method. This question format prompted teams to ask optimistic questions that offered maximal opportunities for new thinking. For example, the question, “How might we create an ICU that ensures dignity and respect for patients and their family members?” casts a wide net to pull in novel ideas.


The facilitator provided a short training session on brainstorming, a technique that emphasises presentation of ideas no matter how far-fetched they may seem to an open-minded audience. The advantage of brainstorming is generating a wide range of innovative ideas. Each team completed two 20-minute rounds of brainstorming to address their “How might we…?” questions. Answers were grouped by similar ideas. The teams then chose three ideas that inspired them and were, in their estimation, the most relevant and promising for designing a more ideal ICU.



In this last phase, teams constructed prototypes of their ideal ICU. The exercise involved synthesising concepts and using standard ICU equipment and some basic materials to create a product. This exercise was designed to energise and challenge the teams’ creativity, enabling them to truly step outside the box and create what an ICU of the future could look like and how it could operate. Each team then presented key elements of their novel designs and prototypes and a facilitator synthesised the comments.



Teams generated novel, futuristic, and challenging ideas. Some could be immediately implemented, while others required advances in technology to implement. The proposals primarily focused on changes in the physical ICU structure, workflow, technology support requirements, patient and family experiences, and operations. The goals and main concepts created by each team are briefly described below (see Table 1 for details).


Team 1
sought to use technology to deliver evidence- based care and to design patient and family- centred care processes. Their prototype included wireless technology and equipment (e.g., implantable patient monitors and treatment microsystems), and voice-activated information and communication systems to access and integrate history, physical findings, and diagnostic data.


Team 2 sought to improve situational awareness and coordination of care between providers and families. Their design centred on a wallmounted computer display to continuously orient clinicians and family members to current and future plans for care. The system provided computerised facial identification of providers and their specific roles in care delivery, displayed reminders of required care and virtual huddles for handoffs, and had electronic location monitors to eliminate delays from tracking down staff, patients, or family members.


Team 3 focused on converting data into knowledge and action. They invented an automated head-mounted technology controlled by voice or eye blinks. Caregivers used it to access, document, and share patient data, record rounds, and complete hospital requirements (e.g., populating the patient record, clinical credentialling). This technology had a patient dashboard of key clinical indicators. Hospital administrators and supervisors could also use the technology to train providers, and, through biometrics, monitor clinician stress levels to prevent burnout, excessive fatigue, and medical errors related to these impairments.


Team 4 considered how an artificial intelligence agent might reduce caregiver anxiety. They posited that the major causes of anxiety were a sense that heroic action was frequently needed and a fear that making a mistake would harm a patient. The team conceived a voice-prompted, tabletbased, virtual personal assistant (avatar) to completely support caregiver needs (e.g., calculate medication dosages, remind caregiver about a needed intervention).


Team 5 sought to more effectively involve families in the care of their loved ones. They envisioned using tablet-based technology to improve communications between patients and their family members (offsite and in-hospital), and between patients, families and the care team. The tablet would let families and patients access their EMRs, provide lay translations of medical terminology, translate English to other languages and vice-versa, share care team information, deliver educational materials, and recommend accommodations and facilities close to the hospital.


Team 6 sought to improve the patient and family experience and foster collaboration among clinicians, family members, and patients. They modified the physical environment, creating a family area in each room with amenities that enabled families to stay for longer periods (e.g., comfortable chairs, shelves, a refrigerator, and noise-cancelling technology). They invented an instant hand sanitiser and wall displays or tablets to enhance communication, retrieve information, and control room functions (e.g., lighting). They included a voice-activated system for clinical data entry, and order entry for automated activation of devices, such as ventilators or infusion pumps.


Team 7 sought to enable collaboration in care between caregivers and family members by improving the communication of all relevant information. They envisioned a wall-mounted interactive monitor in each patient room that communicated information to the care team, patient, and family. This system contained patient and family-based applications (e.g., care journal, photo wall, understandable explanations of medical procedures) and caregiver-based applications (e.g., clinical diagnostics, daily goals, a patient dashboard of medical status). This technology allowed caregivers to remotely access a patient’s medical information from any mobile device.



The workshop was undertaken to step outside current ICU medicine and invent an ICU that optimised care. We undertook this project because drastic changes are needed in current ICU systems to eliminate preventable harm, improve patient outcomes and experience, provide clinicians with better decision support tools, and reduce waste of healthcare resources. Intensive care was established to improve the health of critically ill patients, but care is more complex than 30 years ago. Studies in the ICU found an estimated 38.8 sentinel events per 100 patient ICU days, with 31% of patients experiencing multiple events (Valentin et al. 2006), and multiple complications in a single patient following abdominal aortic surgery (Pronovost et al. 2001). Multiple events in a single patient (588 events in 423 unique patient admissions) were also discovered in a retrospective study of patient admissions (Landrigan et al. 2010).


All seven teams noted how poorly the current ICU information technology meets clinicians’ and patients’ needs. They commented how the technology in their personal lives was far superior to the technology in their professional lives. Simply put, current approaches are inadequate to achieve our overarching goal of optimising ICU care.


The participants in this conference described seven prototypes that, if implemented, promise to substantially improve care delivery. We found common attributes among these different designs that fell into several overlapping imperatives.


1. Create intelligent clinical systems

• Use computer-based technology to automate tasks (e.g., voice dictation), improve access to all data from any device (e.g., smartphone, tablet), and establish safeguards and reminders to assist clinicians in medical care and decisions.

• Such a system would eliminate duplicate documentation and data input, improve clinical diagnostics, decrease clinician’s reliance on memory, and reduce the potential for medical error.


2. Create an integrated information ecosystem

• Build an electronic platform to connect devices (e.g., ventilators, medication pumps), monitors, intelligent clinical systems (described in point 1), and the patient’s EMR. The ecosystem would use wireless technology, have accurate sensors, and communicate information in real time to clinicians and patients. For example, the ecosystem would couple smart alarms to the patient’s room equipment and EMR and communicate.


3. Engage the patient and family in the medical care experience

• Design a patient room with smart screens that display patient information in a timely and understandable manner, offer multilingual translations, and play instructional videos tailored to the patient’s medical situation.

• Reduce clutter (e.g., wireless connection of equipment, described in point 2) and provide an area of comfort for family members. Involve the family in the patient’s care (e.g., provide oral care, help with mobility). This last point will increase engagement and reduce the sense of hopelessness felt when a loved one is critically ill (Hibbard and Greene 2013). Such involvement will also reduce the burden on caregivers.


4. Use technology to improve communications

• One key recommendation was the use of technology to allow everyone involved in care – patient, family, and care team – to communicate with each other as easily and often as possible.


Future Directions

Workshop participants compiled a list of key elements or actions to start designing a more ideal ICU, including always partnering with patients and families to co-create the design. Table 2 outlines future directions and provides a list of design principles for the next generation ICU. The main principles that should govern future ICU designers are to meet the needs of patients and clinicians, actively engage family members in care, employ technology to reduce workload and improve quality of care, provide continuous feedback to patients, families, and providers, and be humble and respectful by learning and improving care delivery. The group agreed to meet again to conduct a similar exercise to review the prototypes they developed, review progress, and brainstorm additional ideas.


We will use the systems lifecycle framework shown in Figure 2 as our work continues to design the ideal ICU. Throughout this work, collaboration and transparency are essential features of this framework. This manuscript informs the conceptual development and user’s needs. Future work with the system lifecycle and design the ideal ICU.


In summary, our innovation process helped workshop participants identify ways in which an ICU can be redesigned to eliminate patient harm, optimise the patient and family’s experience and outcomes, and reduce wastefulness. This process, which is deeply rooted in the needs of all stakeholders, could be widely applied in healthcare to move from a health system where clinician heroism is the safety net to one in which the net relies on safe, innovative design.

Appendices - please see bottom of page
Appendix 1.  Workshop Participants and Organisations 
Appendix 2a. ICU Observations. People. 
Appendix 2b. ICU Observations. Technology and Space.
Appendix 2c. ICU Observations. Work Process/ Work Flow.

The Johns Hopkins Armstrong Institute ICU Innovation Consortium consists of, in alphabetical order by last name, Hanan J. Aboumatar, MD, MPH; Christine Bechtel, MA; Michael Becker, PhD, RN; Melania Bembea MD, MPH; John Benson, MS; Michael Blomquist ; Howard T. Carolan MPH, MBA; Jon Coleman MBA; Maria Cvach RN, MS, CCRN; Danielle DaSilva; Richard Dean, PhD; Shirley Dugdale, MA; Karen A. Earsing, RN, MS; Nancy Edwards, MSB; Ayse P. Gurses, PhD; John R. Deopuria, MS; Jennifer Di Mattina, MS; Todd Dorman, MD, FCCM; Pete Doyle, PhD; Jordon Duval-Arnould MPH, CPH; James Fackler, MD; J. Christopher Farmer, MD; Ayse P. Gurses, PhD; Deborah B. Hobson, RN, BSN; Charles W. Hogue, MD; Lucas Huang ; Elizabeth A. Hunt, MD, MPH, PhD; Robert C. Hyzy, MD; Corry J. Kucik, MD, MA, DMCC, FCCP; Soo-Hoon Lee; Mary Logan JD, CAE; Leah C. Lough, MBA; Sally W. MacConnell, MBA; Pedro A. Mendez-Tellez, MD; Michael P. McLoughlin, MSEE; Vinay M. Nadkarni, MD, MS; Edward Ovando, BS; Michael O’Reilly, MD, MS; Stephanie Peditto, MHS; Phillip Phan, PhD; Julius Pham, MD, PhD; Dan Rice; Fred Richards, BS; Mark Romig, MD; Michael A. Rosen, PhD; Judy A. Schroeder, RN, BSN, MS; Maureen Seckel, MSN; Brian Strawn, MDes, MArch, Scott A. Swetz, MSCE; Tom Ulseth; Aaron Webb, BS; Laura Winner, MBA, RN; Jason Wong, MSPH; and Rhonda Malone Wyskiel, BSN, RN. The authors thank Christine G. Holzmueller, BLA for helping organise, write and edit the manuscript content. 

Source of Funding and Conflicts of Interest:  The Gordon and Betty Moore Foundation funded this research, and will provide support to prototype many of the ideas described in the Future Directions section of this manuscript.  Dr. Pronovost has received grant support from the Agency for Healthcare Research and Quality, the National Institutes of Health and The Commonwealth Fund (past), support for board membership from the Cantel Medical Group, consultant fees from the Association of Professionals in Infection Control and Epidemiology, speaking honoraria from the Leigh Bureau and various hospitals, and book royalties from the Penguin Group. Dr. Deutschman receives royalties from Elsevier for a textbook on Evidence-based Practice of Critical Care, has pending grant applications with the NIH and the Harrington Foundation, and previously received support for board membership from the Society of Critical Care Medicine and consultancy fees from the Centers for Disease Control & Prevention.  Dr. Bo-Linn is employed by the Gordon and Betty Moore Foundation. Dr. Solomon receives consulting fees from the Armstrong Institute for Patient Safety and Quality and from various other health and technology organisations as an innovation consultant, speaking honoraria from Health 2.0, California Hospital Association and other organisations, and a fellow at IDEO.  Dr. Sapirstein has no conflicts.

«« ISICEM 2014: iMDsoft Presents Unique ICU Data Management Capabilities

New Atrial Fibrillation Treatment Device »»


AHRQ (2012) Eliminating CLABSI, a national patient safety imperative. Final report on the National On the CUSP: Stop BSI Project. Rockville, MD: Agency for Healthcare Research and Quality. 

Behrens KM, Marburger JH, Kvamme EF and PCAST. (2008) Priorities for personalized medicine. Washington, DC: Executive Office of the President of the United States, President's Council of Advisors on Science and Technology.

Donchin Y, Gopher D, Olin M (1995) A look into the nature and causes of human errors in the intensive care unit. Crit Care Med, 23(2): 294-300. 

Gurses AP, Murphy DJ, Martinez EA, Berenholtz SM, Pronovost PJ (2009) A practical tool to identify and eliminate barriers to compliance with evidence-based guidelines. Jt Comm J Qual Patient Saf, 35(10): 526-32. 

Hibbard JH, Greene J. (2013) What the evidence shows about patient activation: better health outcomes and care experiences; fewer data on costs. Health Affairs, 32 (2): 207-214. 

Institute of Medicine. (2001) Crossing the quality chasm: a new health system for the 21st century. Executive summary. Washington, DC: National Academy Press. 

Landrigan CP, Parry GJ, Bones CB et al. (2010) Temporal trends in rates of patient harm resulting from medical care. N Engl J Med, 363(22): 2124-34. 

Pronovost PJ, Bo-Linn GW (2012) Preventing patient harms through systems of care.  JAMA, 308(8): 769-70. 

Pronovost PJ, Dang D, Dorman T et al. (2001) Intensive care unit nurse staffing and the risk for complications after abdominal aortic surgery, Eff Clin Pract, 4(5): 199-206. 

Pronovost PJ, Marsteller JA, Goeschel CA (2011) Preventing bloodstream infections: a measurable national success story in quality improvement, Health Affairs, 30(4): 628-34. 

Ravitz AD, Sapirstein A, Pham JC, Doyle PA (2013) Systems approach and systems engineering applied to health care: improving patient safety and health care delivery. Johns Hopkins APL Technical Digest, 31 (4): 354-65. 

Stokols D, Hall KL, Taylor BK et al. (2008) The science of team science: overview of the field and introduction to the supplement, Am J Prev Med, 35(2 Suppl): S77-89. 

Valentin A, Capuzzo M, Guidet B et al. (2006) Patient safety in intensive care: results from the multinational Sentinel Events Evaluation (SEE) study, Intensive Care Med, 32(10): 1591-8. 

Related IssueArticles

Author Ross Freebairn, FCICM Consultant, Intensive Care Services & Clinical Director,... Read more

Is there such a thing as a perfect ICU? Organisation and design of the ICU has evolved exponentially since the first units delivering... Read more

An Interview with Dr. Jerry Nolan Dr. Jerry Nolan was co-editor of the European Resuscitation Council Guidelines for... Read more

Authors Peter J. Pronovost, MD, PhD, FCCM1-3 Clifford S. Deutschman, MS, MD, FCCM4,9 Kathryn E. Roberts, MSN, RN, CNS5 Alan D. Ravitz, PE6 Brad...

No comment

Please login to leave a comment...

Highlighted Products