The most mature implementation of critical care clinical information systems consists of tools which meet the needs of the hands-on caregivers. Historically, massive amounts of data documenting an ICU patient’s clinical status and treatment have been recorded on large double-sided paper flow sheets, which are plagued with problems of legibility, inaccurate calculations, and use restricted to a single person at a time. By replacing this document with computer screens, each customized to a specific purpose, these problems have been essentially solved.

Going beyond simple replacement of paper documents provides an opportunity to present information such that patterns are more easily recognized. For example, correlation of measures of physiologic status, clinical status (such as urinary output and body weight), and administration of medications can facilitate clinical analysis by juxtaposing interdependent variables. Less obviously, trends over longer periods of time can be easily displayed while these could only be laboriously drawn by hand.

Optimal use of clinical information systems should also guide the hands-on caregivers to provide care using evidence-based protocols. Simply creating a place to document the position of head of the bed underscores that this is important issue to be managed in prevention of ventilator-associated pneumonia (VAP). Explanatory information can also be provided on a just-in-time basis to encourage protocol compliance. Computer provider order entry prompts and order sets can also facilitate standardization of care.

Simply collecting and displaying information electronically, while an advance over a paper record, vastly underutilizes the capability of the computer system. The data are being gathered in a computable form and consequently are subject to continuous analysis, enabling detection of patterns that could signify clinical decompensation. Vastly larger datasets than can be retained and analyzed in the human brain can be evaluated and, furthermore, it can be done continuously on all monitored patients, simultaneously. Such an early warning system could trigger evaluation that might otherwise be delayed.

Finally, computable information that describes in detail both the patient’s status and treatment can be used to analyze compliance with protocols for optimal care, resource utilization, and outcomes. Monitoring on a near real-time basis provides timely feedback and is an opportunity to intervene to improve ongoing care.

Once all of these capabilities are available and used by the hands-on caregiving team, their individual capabilities can be optimized. Nonetheless, the realities of the critical care environment are such that patients may be critically ill and yet not be in a setting where their care needs can be expeditiously met. For example, a patient might be in a distant hospital where intensivist coverage is not available. Or, even in a sophisticated medical center, patients may decompensate outside the ICU and, indeed, even in the ICU after hours, an intensivist might not be physically available to respond.

Two technological approaches to dealing with this problem have developed, each dependent on a suitably trained intensivist sitting at a remotely located computer that is equipped with a microphone and speaker and a high-bandwidth Internet connection. Each approach also has one or more high-resolution cameras which can be controlled by a remote physician and means to communicate with caregivers and patients and family who are in the patient’s room. Medical devices such as stethoscopes can sometimes be connected as well.

Connectivity to additional clinical information systems varies according to institutional capabilities. For example, some systems have as many as eight monitors arrayed such that the remote physician can see the real-time electrocardiogram tracing, access the institution’s image archiving and communication systems, and review all elements of a comprehensive clinical information system, simultaneous with viewing and talking with the patient. Without question, availability of this full suite of technological capabilities allows a comprehensive evaluation of the patient that far surpasses the limited verbal interaction between the bedside caregiver and a physician connecting by telephone. It is now well documented that such interactions can provide for more timely and therapeutically appropriate interventions [3]. Nonetheless, such evaluations are still limited in that hands-on physician diagnostic and therapeutic maneuvers are not available when the physician is remote. It has been documented that remote proctoring of a procedure being performed by a house officer who is in the hospital is a practical alternative when immediate interventions are required. Furthermore, even in the case where the physician or patient will need to travel to the point of care, useful temporizing measures may be deployed.

A form of technology that is particularly well suited for the interaction with an individual patient is a mobile robot, offering what is referred to as “robotic telepresence.” One form of this device can actually be driven remotely by the physician from its storage location to the patient’s location in the appropriately equipped facility. Using wireless connectivity, the robot establishes a similar connection to that which exists in rooms that have been specifically hardwired for these capabilities. Because of the costs of connectivity, institutions frequently limit fixed installations to ICUs. Nonetheless, it is clear that patients in other patient care locations can decompensate and care may be needed elsewhere. Such robots provide a lower cost means to provide similar capabilities and could be used to augment the expertise of rapid response teams.

Interactions may be initiated by the caregiving staff from any care location. In such circumstances acceptance has been generally very favorable. Nurses feel that there are trained physicians who are awake and available in the middle of the night and can be provided with all of the information needed to provide care. As a consequence, nurses may be more confident that the patients are receiving quality care. A limitation is that the remote physician may have less of an appreciation for the patient’s clinical course than a physician who has seen the patient daily. However, in some ICUs, the physician on call at night at home and using the robot may be the same person who rounded on the patient that day. Interaction between remote and primary treating physicians remains an essential element of care.

In institutions where multiple ICUs have been equipped with cameras in each room and connectivity to clinical information systems and other clinical data sources has been established, a
team is established at a central monitoring location which may be distant from the ICUs and hospital(s), frequently off campus in less-expensive commercial office space. Analysis of signals from bedside monitors and other devices as well as the results of laboratory tests alert off-site providers to perform patient assessments. Alternatively, bedside providers can request evaluation and off-site management. Interventions, including the ordering of diagnostic tests, medications or consultations, or the manipulation of life support devices can be done by off-site providers or by on-site providers. Thus, a single patient interaction may be initiated by the remote physician as well as by hands-on caregivers. Like any team endeavor, effectiveness is determined in part by communication timeliness and dynamics of trust and responsibility among the bedside and off-site team members.

The primary responsibilities of the remote monitoring team are identification of unfavorable trends and to intervene to enhance best practice adherence, perform care plan reviews for patients admitted after day time hours and provide ICU pharmacist [4] review of after hours provider medications orders which provides an additional safety net for patients in the ICU [3,4]. Bedside caregivers have the potential to be overwhelmed by the need to care for multiple patients, and the requirement to deal with the mechanics of providing care may interfere with always maintaining perspective on the patient’s course.

Information systems that power the central monitoring station have been equipped with series of rules that evaluate clinical information as it is being gathered at the bedside and returned from the laboratory. By correlating this data, alerts can be fired to draw the attention of the remote monitoring team. The team then has the clinical information available to judge whether this is a new or serious development which then prompts interactions with the bedside caregiving team. Such tools may also be available to the bedside caregiving team; however, their many clinical duties can often result in a delayed response. Furthermore, many bedside clinical information systems are much less sophisticated in this area than are the systems designed for use in monitoring a population of patients.

An important capability is the opportunity to perform virtual rounds on the sickest patients. The acuity status is used to identify which patients might most benefit from closer observation. In this way, the remote physician can perform virtual rounds at intervals to judge the effect of medications which may have been administered to determine if physiologic responses are improving or deteriorating. This surveillance can be an important complement to bedside care.

An essential element for the success of remote monitoring of critically ill patients is the effective collaboration between the hands-on caregivers in the central monitoring team. The bedside critical care multidisciplinary team that is responsible for the patient and sees the patient and family on an ongoing basis is best positioned to establish the daily plan of care for each patient. The role of the off-site team members is to keep the patient on the intended trajectory and to communicate with the bedside providers when the patent’s course has deviated from that path. In ICUs where full-time 24-hour day coverage is not available, which is the vast majority of ICUs, physician interaction that may be necessary to ensure that the goals of care are achieved may be sporadic and untimely. The remote team serves as a surrogate for the bedside team at times when they are not able to attend to the patient. In recent years, evidence has accumulated that ongoing availability of intensivist is associated with improved outcomes. If there are an insufficient number of trained intensivists to cover the ICUs that exist, such remote monitoring is being used to increase the availability of trained staff.

It has also been established that implementing certain protocols for care of critically ill patients is associated with better outcomes in the management of sepsis and the avoidance of VAP. Nonetheless, it has proven challenging not only to achieve initial compliance with such protocols, but even more difficult to maintain compliance at a high level. An additional role played by a central monitoring team is to identify when patients who are eligible for a protocol are not receiving such care.

To the extent that the remote monitoring team functions completely independently from the on-site caregivers, there is opportunity for miscommunication and compromise of trust. Indeed, bedside caregivers have been reported to feel threatened by the sense of someone looking over their shoulders all the time and the primary treating physicians could resent intrusions that alter the plan of care set out by them [5]. A substantial investment in relationship building and acceptance by all members of the on-site and remote teams of the importance of minimizing medical errors is thought to be associated with larger improvements in outcomes.

In the fall of 2009, a new technological sea change is that the Blackberry and Apple iPhone are beginning to not only take over the previous place of the medical pager, but, due to their ubiquitous access to high-speed Internet, provide the means to deliver high-resolution bedside monitoring device (BMDI) data, as well as complete access to the electronic medical record from any location at any time. Although telemedicine has enabled new healthcare structures, as mentioned earlier, these new technologies delivered to the individual practitioner are likely to transform medicine just as has happened in the business world [6].

Clinical decision support (CDS) has been defined as a system that uses two or more items of patient data to generate case-specific advice [7]. In practical terms, CDS includes a wide range of functions, including predefined rules, alerts, reminders, workflow, and collaboration tools—and associated content—for improved medical decision making. CDS is often intended to facilitate the introduction of and conformance to evolving evidence-based medical protocols and standards of care while enabling appropriate individual physician discretion (such as during order entry). Rules are, at their core, built on IF/THEN logic statements that allow a tremendous amount of flexibility and power to be added to systems within critical care and across the hospital or integrated health system.

Over the past decade, the business end (e.g., the user experience) of CDS has been the alert box. A growing number of studies are beginning to reveal the critical limitation of alerts that, by design, interrupt the clinician’s workflow, in particular, during order entry [8,9]. The primary reason for this limitation lies in CDS systems designed mainly to alert post hoc after the clinician has requested a particular item (e.g., drug dosage, test).

CDS has the potential to provide special value in settings like the ICU due to the density of data assailing the busy critical care physician and the ability of computers to combine, synthesize, and correlate these data and then create more complex rules and information interpretation displays [10]. Studies have demonstrated that critical care rounds may challenge the physician with 20 times more data elements than the human brain can simultaneously process [11]. In the past, we have relied on the team approach to cope with this onslaught. In the current practice reality of competing priorities of intensivist time, numerous handoffs among providers, the need for IT to take more of a facilitation role for the ICU physician and
nurse is substantial. The next emerging developments in CCIS are likely to be in both the areas of visual design and complex rules and algorithms to predict and inform clinicians about patient circumstances by multiple means. This is discussed, along with emerging techniques for ICU performance management and related metrics, in a later section of this chapter.

The following steps enable the physician, in combination with other stakeholders such as nurses and pharmacists, to evaluate, select, and obtain maximal benefits from CCIS systems, with the assistance of a professionally certified and experienced project manager (typically from the IT department). It is the project manager who coordinates overall project planning, ensures that the required technical resources will be available on time, and monitors tasks and milestones among the project team. Technical needs such as interfaces to other IT systems and to medical devices, hardware, power, physical space, network access, and system security are necessary parts of this coordinated planning in addition to software delivery and configuration.