Today’s critical care units are both busy and complex, where nurses, doctors and other health professionals use their knowledge, skills and technology to provide patient care. In fact, this complexity makes errors a common occurrence; one large international study in 205 Intensive Care Units (ICU) showed that 39 serious adverse events occurred per 100 patient days.¹ The Institute of Medicine (IOM) in the USA defines quality of health care as ‘the degree to which health services for individuals and populations increase the likelihood of desired health outcomes and are consistent with current professional knowledge’.² Critical care nurses are well known for their skills in patient assessment. In fact, this ongoing surveillance of patient condition means that nurses are ideally positioned to prevent, discover and correct medical errors.³ Thus, nurses play a key role in improving quality and safety in health care. This chapter provides a review of quality and safety in critical care. First, an overview of evidence-based nursing and clinical practice guidelines is given to provide a foundation to consider quality and safety. Next, quality and quality monitoring is considered. Included in this section are the topics of care bundles, checklists, rapid response systems and information and communication technologies. Finally patient safety, including safety culture is described. In Chapter 2 we addressed risk management, clinical governance and the role of clinical leaders and managers in delivering critical care services; this information is complementary to what will be discussed in Chapter 3.

Evidence-Based Nursing

Evidence-based nursing (EBN) is the ‘Application of valid, relevant, research-based information in nurse decision-making.’⁴ Research evidence, however, is only one of four considerations in making a clinical decision. Three other considerations include: (1) knowledge of patients’ conditions (i.e. preferences and symptoms); (2) the nurses’ clinical expertise and judgment; and (3) the context in which the decision is taking place (i.e. setting, resources). Figure 3.1 provides a schematic representation of EBN, using an example of a decision about weaning a patient from a mechanical ventilator.

FIGURE 3.1 Schematic representation of evidence based nursing including an example of weaning from mechanical ventilation.

EBN has emerged as a way to improve nursing practice by considering the care that nurses give to patients, and whether this care results in the best possible outcomes for patients. It has been viewed as both an attitude and a process. As an attitude, it is a way of approaching practice that is critical and questioning. As a process, a number of steps in EBN have been described. Figure 3.2 identifies these steps, with more details about each step being provided below.

FIGURE 3.2 Steps in the evidence based nursing process.

Translate a Clinical Query into a Structured Question

In situations where nurses have to make clinical decisions, it is important for them to carefully consider the issue or problem they are facing as it influences what research evidence should be used to make decisions. Thus, the first step in the EBN process is translating a clinical query into a well-defined, answerable, structured question. A well recognised approach is the Population, Intervention, Comparison, Outcome format, more often referred to as PICO. The Population reflects the patient group or clinical scenario of concern. The Intervention is one option for the particular nursing practice. The Comparison is the current practice, or the second option for practice. Finally, the Outcome is the effect that the nurse is hoping to achieve, which should reflect a patient outcome. Table 3.1 provides three examples of PICO questions relevant to critical care nursing.

TABLE 3.1 Examples of clinical questions using the PICO format

Locate the Best Evidence

After well-defined, answerable, structured questions have been developed, nurses can turn to reviewing the literature to find the answers. First, the evidence has to be located, which involves searching library databases. Some of the databases generally searched include Ovid CINAHL, Medline and Cochrane. Articles that relate to the question then have to be retrieved. These articles may be reports about primary research (i.e. written by the person conducting the research); systematic reviews of existing research; or clinical practice guidelines that have been developed from primary research and systematic reviews.

Critically Appraise the Evidence

Once the various sources of evidence have been retrieved, they are then assessed for their quality and relevance to the clinical question. In Australia, the National Health and Medical Research Council (NHMRC)⁵ has described strategies to assess research evidence on the effectiveness of interventions. It provides a useful framework to consider research evidence for improving nursing interventions, and identifies three questions to ask regarding potential interventions:

1. Is there a real effect?

2. Is the size of the effect clinically important?

3. Is the evidence relevant to practice?

This first question regarding the real effect relates to the strength of the research that has been conducted. The strength of the research has three dimensions: level of evidence, quality of the individual studies, and their statistical precision (denoted by P values or confidence intervals). Although there are a number of different evidence hierarchies (e.g. see Jennings & Loan, 2001),⁶ the framework used by the NHMRC⁵ is displayed in Table 3.2.

TABLE 3.2 NHMRC’s level of evidence ⁵ designation for levels of evidence in the studies of effectiveness

Level of evidence	Study design
I	Evidence obtained from a systematic review of all relevant randomised controlled studies
II	Evidence obtained from at least one properly designed randomised controlled trial
III-1	Evidence obtained from well-designed pseudorandomised controlled trials (alternative allocation or some other method)
III-2	Evidence obtained from comparative studies (including systematic reviews of such studies), with concurrent controls and allocation not randomised, cohort studies, case-control studies, or interrupted time series with a control group
III-3	Evidence obtained from comparative studies with historical controls, two or more single-arm studies, or interrupted time series without a parallel control group
IV	Evidence obtained from case series, either post-test or pre-test/post-test

The second question focuses on whether meaningful improvements to patient care and outcome will result if the research findings are applied in practice. It also considers how the intervention compares with current practices in terms of patient care and outcomes.

The third question conveys the notion that potential benefits or outcomes of the intervention must be both important to the patient, and be able to be replicated in other settings. The NHMRC ⁵ identifies three types of outcome: surrogate, clinical, and patient-relevant (which are not mutually exclusive) (see Table 3.3). Surrogate outcomes are often used in critical care where measurement of the actual physiological change (e.g. oxygen-carrying capacity of the blood) is replaced by a more accessible, and equally acceptable, parameter (e.g. oxygen saturation). Clinical outcomes are those of direct relevance to clinical practice, and patient-relevant outcomes are those likely to be articulated as significant by the patient/carer. When assessing research evidence, the type of outcome used in the research should be considered. Assessing the evidence results in an understanding of its quality of evidence for a particular nursing practice.

TABLE 3.3 Types of outcome

Outcome	Definition	ICU example
Surrogate	Some physical sign or measurement substituted for a clinically meaningful outcome	• Oxygen saturation • Vital capacity

Clinical Outcome defined on the basis of the problem

• Ventilator days

• Survival

Patient-relevant Outcomes that are important to the patient

• Functional ability

• Quality of life

Integrate the Evidence into Practice

When good quality evidence for a particular practice is identified, it is important to then consider this evidence alongside nurses’ expertise, patient preferences and available resources. In essence, evidence may suggest that a particular practice achieves the best patient outcomes, but if the nurse does not have the skills needed to implement the practice, if the resources are not available, if the patient either does not want the intervention, or their situation is such that the intervention may not be appropriate for them, then this practice should not be implemented. However, in many situations the practice will be applicable to the patient and nurses will have the skills and the resources to implement the practice. At times, implementing this new practice may take the form of developing a clinical practice guideline or protocol for a particular nursing activity. Clinical practice guidelines are described in the next section.

Evaluate Clinical Performance

Once a new practice has been implemented, it is important for nurses to assess whether it is having the desired effect. At the individual patient level, this often involves assessing the patient, whereas at the unit level, it may involve either a practice audit or research. Practice audits often involve reviewing patient charts to determine both the extent to which the new practice has been implemented and its outcome on the patient. Research may seek to understand similar things, but generally takes a more formal approach, addressing issues such as appropriate study designs, ethics approvals, etc.

Clinical Practice Guidelines

The development and use of clinical practice guidelines (CPGs) is one strategy to implement EBN. CPGs are statements about appropriate health care for specific clinical circumstances that assist practitioners in their day-to-day practice.⁷ They are systematically developed to assist clinicians, consumers and policy makers in healthcare decisions and provide critical summaries of available evidence on a particular topic.⁸ Other terms that are often used synonymously with CPGs include protocols and algorithms.

There are a number of benefits of using CPGs. They are seen to be central to quality patient care because, in essence, they standardise care.⁹ They can guide decisions and can be used to both justify and legitimise activities and practices.⁹ However, limitations have also been identified. Poorly developed guidelines may not improve care and may actually result in substandard care⁹. In the critical care area, the Intensive Care Coordination and Monitoring Unit of New South Wales Department of Health has led the development of CPGs associated with six common nursing interventions: (1) Eye care; (2) Oral care; (3) Suctioning a tracheal tube; (4) Endotracheal tube stabilisation; (5) Central line care; and (6) Arterial line care.¹⁰ Clinical audits are often used to establish the need to develop new protocols at the local unit level. Clinical audits generally involve chart reviews, but may also use direct observation or surveys of practice. Clinical audits often establish variation in practice without adequate justification.

Developing, Implementing and Evaluating Clinical Practice Guidelines

A number of steps are undertaken when developing clinical practice guidelines. Table 3.4 provides an overview of these steps, which has been adapted from Miller and Kearney’s work.⁹

TABLE 3.4 Steps in developing clinical practice guidelines ⁹

Step	Description
Find the evidence	After deciding on what is considered evidence, databases such as CINAHL and Medline must be searched to find relevant studies and expert opinions.
Evaluate the evidence	Relevant studies and expert opinion papers must be critically appraised for their strengths and weaknesses. This may or may not incorporate a systematic review.
Synthesise the evidence	General summary statements about the state of knowledge on a particular topic are developed.
Design the guidelines	Written summaries, algorithms and/or summary sheets will be developed that include statements about appropriate healthcare practices and their rationale.
Appraise the guidelines	Validity, reliability, clinical applicability, flexibility and clarity are some criteria that can be used to assess the guidelines.
Disseminate and implement the guidelines	Specific strategies such as seminars and patient chart reminders must be developed to increase awareness, acceptance and implementation of the guidelines.
Review and reassess the guidelines	Clinical audits and research may be used to regularly evaluate the impact the guidelines have had on patient care and outcomes.

While research, systematic review and expert opinion form the foundation for CPGs, the quality of evidence must be assessed and overall summaries of the knowledge to date are essential. These summaries are then used to develop the guidelines, which generally include a series of statements about the care to be provided and a rationale for this care.

Once the guidelines are developed, a group of experts and users should assess the guidelines for accuracy, clinical utility and comprehension. Recently, international experts developed a 23-item appraisal instrument, termed the Appraisal of Guidelines for Research and Evaluation (AGREE), that assesses five domains: (1) scope and purpose of the CPG (3 items); (2) stakeholder involvement in CPG development (4 items); (3) rigour of development (7 items); (4) clarity and presentation (4 items); and (4) applicability (5 items). Instruments such as AGREE can be used to assess the quality of CPG.

Based on the assessment of the CPG, revisions may be required. Next, strategies for disseminating and implementing the guidelines should be developed. Importantly, simply publishing and circulating CPGs will have a limited impact on clinical practice, so specific activities must be undertaken to promote CPG adherence. The following seven strategies have been shown to be moderately effective in promoting guideline adherence: (1) interactive small group sessions; (2) educational outreach visits; (3) reminders; (4) computerised decision support; (5) introduction of computers in practice; (6) mass media campaigns; and (7) combined interventions.⁷ Finally, a process for regularly evaluating and updating the guidelines must be developed, which may involve quality improvement activities or clinical research. In summary, by developing, using and evaluating clinical practice guidelines, nurses may improve patient care and outcomes. Additionally, use of CPGs should ensure that nursing practice is based on the best available evidence.

Quality and Safety Monitoring

This section discusses unit-level measures used to evaluate the quality and safety of care for critically ill patients. Quality and safety in healthcare is commonly described in terms of Donabedian’s approach ¹¹ with three major domains:

1. Patient outcomes – the results of care in terms of recovery, restoration of function and/or survival (e.g. mortality, health-related quality of life).

2. Process – the practices involved in the delivery of care (e.g. pressure ulcer prevention strategies).

3. Structure – the way the healthcare setting and/or system is organised to deliver care (e.g. staffing, beds, equipment).

More recently, a fourth domain of culture or context has been suggested specifically for patient safety models to evaluate the context in which care is delivered.¹² The contemporary model for healthcare improvement recognises that the resources (structure) and activities carried out (processes) must be addressed within a given context (culture) to improve the quality of care (outcome). The overall aim of quality improvement (QI) is to provide safe, effective, patient-centred, timely, efficient and equitable health care.¹³ QI activities identify and address gaps between knowledge and practice. Importantly, these activities need to reflect the most recent and robust clinical evidence to improve patient care and reduce harm. The most common approach used for rapid improvement in healthcare is the plan–do–study–act (PDSA)¹⁴ method where four essential steps are carried out in a continuous fashion to ensure processes are continually improved:

1. Plan – identify a goal, specify aims and objectives to improving an area of clinical practice, and how that might be achieved (i.e. how to test the intervention).

2. Do – implement the plan of action, collect relevant information that will inform whether the intervention was successful and in what way, taking note of problems and unexpected observations that arise.

3. Study – the results of the intervention, particularly its impact on practice improvement, noting any strengths and limitations of the intervention.

4. Act – determine whether the intervention should be adopted, abandoned or adapted for further rapid cycle testing recommencing at the Plan phase.

A variety of specific activities have been used in the ICU setting to translate findings from the literature to improve clinical practice.¹⁵ Quality monitoring includes measurement of, and response to, the incidence and patterns of adverse events (AEs). Adverse events occur in up to 17% of all hospital admissions,¹³ and cost the Australian healthcare system an estimated $2 billion per year.¹⁶ About 18,000 hospital deaths per year are associated with AEs¹³ and generally occur as a result of system errors. Half the AEs were deemed preventable with such strategies as improved protocols, better-quality monitoring, enhanced training and opportunities to consult with specialists or peers on clinical decisions.¹⁷ Studies have identified specific contributing factors for adverse events related to patient airway¹⁸ and intra-hospital transports.¹⁹

A number of methods for reporting AE such as direct observation chart audit and self or facilitated reporting can be used; each has its strengths and limitations. Trained observers report more unintended events but this method is expensive, labour intensive and vulnerable to the Hawthorne effect.²⁰ Both chart audits and incident reporting only reflect what is charted or reported, but even when chart audit, incident reporting, general practitioner reporting and external sources, such as coronial review, are used together, some adverse events will be missed.²¹ Importantly, self or facilitated reporting, such as the Australian Incident Monitoring Study (AIMS)^22,²³ are routinely used surveillance methods in many countries.

Medication administration is the most common intervention in health care, but the medication management process in the acute hospital setting is complex, and creates risk for patients. As a result, medication-related events are the commonest AE for hospitalised patients.²⁴ Adverse drug events (ADEs) are common in Australian hospitals, with preventable, high-impact events involving anticoagulants, anti-inflammatories and cardiovascular drugs (over 50% of ADEs), as well as antineoplastics, opioids, steroids and antibiotics (commonly used in critical care units).¹⁷ Events are clinically significant in 20% of cases.¹⁷ A number of strategies have been instituted in Australia under the auspices of the National Medicines Policy, including the quality use of medicines (QUM) framework. There, however, remains a lack of consensus on how to measure medication safety²⁵ – either by error or adverse event – where:

• error is a failure in clinical management, resulting in potential harm to the patient

• adverse events relate to actual patient harm (injury).¹⁷

The actual incidence of both measures is higher than what is reported.^17,²⁶ Fortunately, most healthcare errors do not result in patient harm because of safety-net processes.²⁴ Despite this, it has been estimated that one potentially serious intravenous drug error occurs every day in a 400-bed hospital.²⁷ Approximately 5% of medication errors relate to infusion pumps. These pumps are used to administer high-impact medications, such as inotropes, heparin or antineoplastics.²⁸ It is therefore important to evaluate interventions that can reduce the incidence and impact of adverse intravenous drug events, particularly in critical care settings.^29,³⁰ Recent evidence suggests that nurses who are interrupted whilst administering medications may have an increased risk of making medication errors,³¹ prompting calls for all healthcare workers to make concerted efforts to reduce interruptions to clinical tasks.³² Other activities examining quality of care include the analysis of incident reports such as the Australian Incident Monitoring Study (AIMS),^22,²³ Quality in Australian Health Care Study (QAHCS)¹⁷ and the Australian Council on Healthcare Standards (ACHS) indicators.³³ Current ACHS indicators for intensive care include:

Similar activities are evident internationally, where concepts of ‘safety science’ (error reduction and recovery) are being applied to critical care practice.^29,³⁴^–³⁶ Process indicators of quality care have been developed, including care related to the prevention of ventilator-associated pneumonia (VAP) and central venous catheter management. Table 3.5 outlines process indicators with good clinical evidence and/or strong recommendations for use by professional bodies, such as the Agency for Healthcare Research and Quality (AHRQ) in the USA.

TABLE 3.5 Evidence based process indicators

Process	Process indicator
1. Central venous catheter management	Maximum sterile barriers Real-time ultrasound guidance during insertion Antibiotic-impregnated catheter
2. Prevention of ventilator-associated pneumonia	Elevated head of bed Continuous aspiration of subglottic secretions Stress ulcer prophylaxis
3. Reducing mechanical ventilation	Low tidal volumes for acute respiratory distress syndrome Weaning protocols Sedation protocols Appropriate use of analgesia and sedation
4. Pressure ulcer prevention	Use of pressure-relieving materials

A range of clinical support tools have been developed and are used to measure compliance with these best practice clinical standards. Daily goals forms, for example, have been used to aid communication between clinicians during and after multidisciplinary ward rounds and ensure that all staff are aware of what care the patient should be receiving and what the clinical plan is.^37,³⁸ A popular mnemonic developed for use by ICU clinicians during patient assessment is the ‘FASTHUG’ which stands for Feeding, Analgesia, Sedation, Thromboembolism prophylaxis, Head-of-bed elevation, stress ulcer prevention and Glucose management.³⁹ Along with care bundles and checklists (detailed below) these tools facilitate standardised care and improve communication between clinicians.³⁸

Care Bundles

An evolving QI approach to the optimal use of best practice guidelines at the bedside is the development of ‘care bundles’. A care bundle is a set of evidence-based interventions or processes of care, applied to selected patients. A number of bundles have been developed for critical care by the Institute for Healthcare Improvement (IHI) in the USA (see Table 3.6).

TABLE 3.6 Institute for Healthcare improvement care bundles

Bundle name	Aim	Bundle components
Central line	Prevent central-line associated bacteraemia	• hand hygiene • maximal barrier precautions upon insertion • chlorhexidine skin antisepsis • optimal catheter site selection with avoidance of the femoral vein for central venous access in adult patients • daily review of line necessity with prompt removal of unnecessary lines

Ventilator care Prevent ventilator-associated pneumonia

• elevating the head of the patient’s bed to 30–45 degrees

• daily ‘sedation vacations’ or gradually lightening sedative use each day

• daily assessment of the patient’s readiness to extubate or wean from the ventilator

• delivering both peptic ulcer disease and deep vein thrombosis prophylaxis

Sepsis resuscitation Reduce mortality due to severe sepsis

• serum lactate measured

• blood cultures obtained prior to antibiotic administration

• improve time to broad-spectrum antibiotics

• treat hypotension and/or elevated lactate with fluids

• apply vasopressors for ongoing hypotension

• maintain adequate central venous pressure

• maintain adequate central venous oxygen saturation

Sepsis management Reduce mortality due to severe sepsis

• administer low-dose steroids by a standard policy

• administer Drotrecogin Alfa (Activated) by a standard policy

• maintain adequate glycaemic control

• prevent excessive inspiratory plateau pressures

Table 3.7 outlines studies examining the process of care delivery in critical care units, including those where care bundles were implemented and evaluated. Increased bundle compliance was associated with decreased ICU length of stay (LOS), reduced ventilator days and increased ICU patient throughput,⁴⁰ and decreased rates of ventilator-associated pneumonia.⁴¹ Other quality improvement studies targeted similar processes of care without taking the bundled approach. A range of measures demonstrated improved outcomes:

• decreased VAP,^42,⁴³ catheter-related bloodstream infection (CR-BSI) rates and LOS⁴³

• increased days between CR-BSIs ⁴⁴

• decreased hospital mortality as the number of process interventions increased ⁴⁵

• reduction in severity-adjusted total hospital costs related to improvements in process measures of care, including glucose control, use of enteral feeding and appropriate sedation.⁴⁶

TABLE 3.7 Studies describing process of care delivery in critical care units⁵⁷

Although studies revealed improvements in both processes and outcomes, variation in levels of compliance with process measures were also reported (see Table 3.7 for detail). One study⁴⁷ revealed the more unwell the patient was, the less likely they were to have received practices they were eligible for.

Checklists

Checklists have the potential to prevent omissions in care by serving as reminders to healthcare providers for the delivery of appropriate quality care for every patient, every time, in complex clinical environments. A checklist typically contains a list of action items or criteria arranged in a systematic way, allowing the person completing it to record the presence or absence of individual items to ascertain that all are considered or completed.⁴⁸

In critical care settings, checklists have been used to facilitate staff training, detect errors, check compliance with safety standards and evidence-based processes of care (such as those outlined previously), increase knowledge of patient-centred goals and prompt clinicians to review certain practices on morning rounds in the ICU. Findings from studies noted that checklists:

Three studies suggested that checklists also contributed to improved outcomes: (1) reduced LOS, ventilator days, unit mortality;⁴⁹ (2) reduced catheter-related bloodstream infections;⁴⁴ and (3) reduced mean monthly rates of VAP.⁵⁴ However, the lack of methodological rigour in these studies prevents inferring causal links between checklist use and improved outcomes.⁵⁷

Information and Communication Technologies

Health departments continue to develop systems and processes that will result in a complete electronic medical record. In Australia, a national e-health strategy has been established,⁵⁸ with the National E-Health Transition Authority (NEHTA), a company established by the Australian, State and Territory governments, assigned responsibility for establishing the foundations (including the development of standards) for e-health across the Australian health sector.⁵⁹ The ultimate goal is an individual electronic health record system designed to provide ‘a consolidated and summarized record of an individual’s health information for consumers to access and for use as a mechanism for improving care coordination between care provider teams’.⁵⁸ ^(p.13)

In combination with these initiatives, information and communications technologies (ICT) are also expanding into clinical practice.⁶⁰ Critical care in particular is at the forefront of these developments, with bedside clinical information systems, order-entry strategies, decision support, handheld technologies and telehealth initiatives continuing to evolve and influence practice. This section examines the current and future impact that these technologies will have on patient care and safety, and on clinician workflows and practices, as clinical information fully assimilates with evidence-based practice and clinical decision support systems.

Clinical Information Systems

A clinical information system (CIS) enables improved data collection, storage, retrieval and reporting of patient-based information, and can facilitate unit-based outcomes research and quality improvement activities.⁶¹ Computerisation of monitoring and therapeutic activities for critically ill patients began in the 1960s, and has now evolved to encompass all aspects of patient care such as cardiorespiratory monitoring, mechanical ventilation, fluid and medication delivery, imaging and results of diagnostic testing.^62,⁶³ Patient-based bedside CIS offers increasingly sophisticated functionality and device interfaces,⁶⁴ enabling real-time data capture, trending and reporting,⁶² and linkage to relational databases.^65,⁶⁶

The introduction of intravenous ‘smart pump’ technology is one application aimed at reducing adverse drug events and improving patient care by supporting evidence-based guidelines for medication management.⁶⁷ The operator-error prevention software is based on a device-based drug library with institution-established concentrations/dosage limits incorporated in the function of the pump. Resulting software functions include clinician alerts (for keystroke errors)²⁸ and transaction log data (post-incident analysis).⁶⁸ Medication errors and adverse drug events can be detected by this software, but further technological and nursing behavioural factors must be addressed before a measurable impact on serious adverse drug errors can be achieved.⁶⁹

The proportion of ICUs in Australia and New Zealand using a CIS is not known, while the estimate for units using electronic charting in North America is 10–15%.^64,⁷⁰ Early generation systems held promise of improved efficiencies but did not demonstrate actual decreases in nursing workload or activity patterns, including in one Australian site.⁷¹ Current third-generation systems (Windows NT operating system [or equivalent] with relational databases and enhanced graphic displays and user interfaces)⁶³ have reduced documentation time (52 minutes per 8-hour shift) and increased the proportion of time on direct care activities.⁷² Despite these positive findings, it is noted that a CIS would not enable a reduction in nursing staff; on the contrary, at least a half-time nursing position is required to administer the system.⁷² An Australian study demonstrated significant reductions in medication and intravenous fluid errors and the incidence of pressure areas, and improved variance between ventilator orders and settings, after implementation of a CIS.⁷³ A sample of nursing staff perceived that the CIS also increased time on patient care and decreased documentation time, while staffing recruitment and retention rates improved.⁷³ Findings that critical care nurses are accepting of new technologies were previously noted.⁷⁴

Other issues also need consideration. Accuracy of data (correctness and completeness of the data set) from both manual and automated inputs to the information system requires evaluation. While automated entry eliminates transcription errors from other data sources,⁷⁵ the use of ‘carry-over’ data to new fields, sampling frequency, and clinician acceptance of monitor-generated data can erroneously affect data accuracy (e.g. damped pulmonary artery waveform not checked, with erroneously low readings documented).⁶¹ In addition to errors related to entering and retrieving information, errors can also arise if systems are not designed to enhance communication between healthcare workers and facilitate coordination of work processes.⁷⁶ Further, ‘clinical alert’ functions can lack the specificity for detecting clinically important events⁷⁰ and may compromise patient safety when used excessively in clinical settings with one study demonstrating 49–96% of drug safety alerts were overridden by clinicians.⁷⁷

To tackle these and other limitations, future systems will provide wireless capabilities, remote access, ‘smart’ alerts, handwriting recognition, clinician-configured forms, flowcharts and reports using standardised data structures and terminology. This level of functionality will enable decision support with online evidence-based clinical practice guidelines, real-time clinical alerts, and online patient historical information via a complete electronic medical record.⁷⁸

Computerised Order Entry and Decision Support

Computerised physician (or provider) order entry (CPOE) is viewed as an important innovation in reducing medical errors, through minimising transcribing errors,²⁵ triggering alerts for adverse drug interactions and facilitating the adoption of evidence-based clinical guidelines.^70,^73,^79,⁸⁰

Computerised order entry is used for medication and intravenous fluid prescribing, diagnostic test ordering and results management, and mechanical ventilation or other treatment orders.^79,⁸¹ Implementation of CPOE and related clinical decision support systems (CDSS) have demonstrated significant reductions in medication errors⁷⁹ and redundant or unnecessary order requests,⁸²^–⁸⁴ and improved compliance with practice guidelines.⁸⁵^–⁸⁶ Clinical decision support systems interface with hospital databases to retrieve patient-specific and other relevant clinical data and to generate recommended actions.⁸⁷ Importantly, clinical decision making at the bedside can be enhanced by providing clinicians with a readily available tool that incorporates relevant clinical information and evidence-based medicine.⁸⁸ Clinician alerts (e.g. allergies or interaction effects) or prompts (e.g. to check coagulation when prescribing warfarin) can be generated. A number of studies have demonstrated improved delivery of patient care after the introduction of such reminders.⁸⁹^–⁹¹ As with CIS implementation, examination of clinician workflow and care delivery patterns⁸¹ and detailed planning is required for successful implementation of a CPOE process.⁹² In particular, order decryption, prioritisation and translation steps within the medication or treatment order process require review to minimise potential errors.⁹² Additional developments involving wireless communication, personal digital assistants and closed-loop delivery systems will improve the efficiency, effectiveness and adoption of this innovation in clinical practice.⁷⁹ Closed-loop delivery adjusts drug or fluid delivery based on active feedback from the target parameter (e.g. inotropic dosages adjusted to a range for mean arterial pressure).

Handheld Technologies

Wireless applications enable both clinical access and portability and mobility within a critical care environment at the point of care. Clinical uses for personal digital assistant (PDA) and Smartphone technologies continue to evolve at a rapid pace.⁹³ These handheld computers use operating systems and pen-like styluses that enable touch-screen functionality, handwriting recognition, and synchronisation with other hospital-based computer systems. An increasing array of clinical applications and content are available for downloading to PDAs, including drug reference information (e.g. MIMS on PDA), clinical guidelines, medical calculators and internet-based literature searches.⁹³^–⁹⁵ PDA use has been reported as a helpful nursing education tool,^96,⁹⁷ with nursing students reporting specific benefits of PDA use such as having access to readily available data, validation of thinking processes and facilitation of care plan re-evaluation.⁹⁷ In critical care, PDAs have been used to document clinical activities, such as logging critical care procedures, which was demonstrated as feasible and useful, although adoption and user acceptance was not uniform.⁹⁸ They have also been used to deliver point-of-care decision support to improve antibiotic selection⁹⁹ and prescribing,⁸⁴ and an interactive weaning protocol that assisted care providers wean patients from mechanical ventilation more efficiently when compared with the use of a paper-based weaning protocol.¹⁰⁰

The benefits of this mobile computing also create concerns, particularly regarding confidentiality of patient information. Health services therefore need policies for managing handheld devices, including password protection, data encryption, authenticated synchronisation and physical security.⁹⁵ In particular, wireless applications require appropriate standards for data security (e.g. wireless-fidelity protected access 2 [WPA2] compliance).⁷⁸ As these issues are addressed, these technologies will form an integral component of routine clinical practice in critical care.

Telehealth Initiatives

Remote critical care management (eICU) using telemedicine/telehealth technologies is expanding as the necessary high bandwidths for transmitting large amounts of data and digital imagery become available between partner units or hospitals. Videoconferencing functions enable direct visualisation and communication of patients and on-site staff with the ‘virtual’ critical care clinician or team. Review of real-time physiological data, patient flowcharts and other documents (e.g. electrocardiograms, laboratory results) or images (e.g. radiographs) provide a comprehensive data set for patient assessment and management.¹⁰¹

This technology-enabled remote care initiative is of particular value for critical care units where no or limited on-site intensivist resources are available. Despite various methodological limitations,¹⁰² several studies using ‘before and after’ comparisons have indicated improved outcomes such as decreases in severity-adjusted hospital mortality, incidence of ICU complications, ICU length of stay, and ICU costs.^101,^103,¹⁰⁴ One study demonstrated improved outcomes for neurological ICU patients through the use of a robotic tele-ICU system that made rounds in response to nurse paging.¹⁰⁴ More recent studies, however, have not found improvement in patient outcomes as a result of telemedicine technology,¹⁰⁵^–¹⁰⁷ highlighting the complex nature of these initiatives and the difficulties evaluating them. One local study instead observed improvements in patient management (i.e. increased discharges and decreased transfers) for moderate trauma patients upon implementation of a virtual critical care unit that linked a district hospital ED with a metropolitan tertiary hospital ED. Further studies that include detailed descriptions of system implementation are required in determining the most effective elements of this technology in critical care settings.^102,¹⁰⁸

In addition to remote patient assessment and management, telecommunications have been used to deliver continuing education to rural healthcare professionals for many years via audio, video and computer.¹⁰⁹ More recently, distance education has been delivered via web-based courses accessed over the internet.¹¹⁰ For example, a web-based educational tool was used to provide information about the classification of pressure ulcers and the differentiation between pressure ulcers and moisture lesions to both student and qualified nurses.¹¹¹ The potential use of web logs or ‘blogs’,¹¹² online communities and virtual preceptorships¹¹⁰ in nursing education has also been discussed. Continuing professional development (CPD) opportunities are also provided on-line, for example AusmedOnline contains a range of resources and learning activities that count towards CPD for registration requirements. However, more work is required to determine how successful these technological advances are on educational outcomes.¹¹³

Patient Safety

The signing of the Declaration of Vienna in 2009 ¹¹⁴ (Appendix A4) committed critical care organisations around the world, including the World Federation of Critical Care Nurses, to patient safety.¹¹⁴ Patient safety is viewed as a crucial component of quality.¹¹⁵ Over the years, numerous definitions of patient safety have emerged in the literature. The Institute Of Medicine¹¹⁶ described it as the prevention of harm, however, more recently, the European Agency, Safety Improvement for Patients in Europe,¹¹⁷ asserted it was about identifying, analysing and minimising patient risk. This latter description is appealing as it leads us to consider the degree of risk situations pose for patient harm and targeting those that are either high risk or frequent in occurrence.

Three techniques used to understand patient risk are analysing reports of adverse events, root cause analyses and failure mode and effect analysis. Recent research on adverse events in critical care has helped to both better understand patient risks and target improvement activities. For example, medications, indwelling lines and equipment failure were the three most frequent types of adverse events in a study of 205 Intensive Care Units world-wide.¹ Focusing on analysing the narratives written about adverse events is viewed as an important way to learn from errors. Root cause analyses is a structured process generally used to analyse catastrophic or sentinel events.^118,¹¹⁹ Learning from both incident reporting such as AIMS and root cause analyses is based on the premise that the information they contain is of sufficient quality to allow accurate analysis, interpretation and detection of the root causes of problems, and even more importantly, the formulation and implementation of corrective actions. Failure mode and effect analysis identifies potential failures and their effects, calculating their risk and prioritising potential failure modes based on risk.¹²⁰ In addition to examining patient risk, another strategy has focused on understanding the safety culture of a unit or organisation, with subsequent activities aimed at improving components of this culture.

Safety Culture

Measurement of the baseline safety culture facilitates an action plan for improvement. Safety culture has been defined as ‘the product of individual and group values, attitudes, perceptions, competencies, and patterns of behaviour that determine the commitment to, and the style and proficiency of, an organisation’s health and safety management.’¹²¹ It is commonly referred to as ‘the way we do things around here.’¹²² A widely used instrument to measure safety culture, the Safety Attitudes Questionnaire (SAQ), focuses on six domains: teamwork climate, safety climate, job satisfaction, perceptions of management, working conditions and stress recognition.¹²³ Interestingly, two studies in the USA that used the SAQ showed that nurses and doctors differed in their perceptions of safety culture.^124,¹²⁵

One strategy to improve the safety culture has involved identifying factors that make organisations safe, which in turn allows initiatives to be developed that target areas of specific need. For example, five characteristics of organisations that have been able to achieve high reliability include:¹²⁶

1. safety viewed as a priority by leaders

2. flattened hierarchy that promotes speaking up about concerns

3. regular team training

4. use of effective methods of communicating

5. standardisation.

Many of these five factors fall under the category of ‘non-technical’ skills.¹²⁷ Other non-technical skills include situational awareness and decision making. Importantly these skills can be learned. For example, the Anaesthetists’ Non-Technical Skills (ANTS) is a training program developed in Scotland, that focuses on task management, team working, situational awareness and decision making.¹²⁷ A second training program, Team Strategies and Tools to Enhance Performance and Patient Safety (TeamSTEPPS), developed in the US, is designed to develop competency in four areas: team leadership, situational monitoring, mutual support (or back-up behaviours) and communication.¹²⁸ Thus, training programs can be used to help develop various non-technical skills, ultimately promoting a culture of safety within the critical care environment.

Rapid Response Systems

Rapid Response Systems (RRS) are systems that have developed to first recognise, and then to provide emergency response to, patients who experience acute deterioration.^129,¹³⁰ The Australian Commission on Safety and Quality in Health Care have identified eight essential elements in a RRS (Table 3.8).¹²⁹

TABLE 3.8 Essential elements of a rapid response system ¹²⁹

Domain	Element	Description
Clinical processes	Measurement and documentation	Vital signs, oxygen saturation and level of consciousness should be undertaken regularly on all acute care patients
	Escalation of care	A protocol for the organisation’s response in dealing with abnormal physiological measures and observations including appropriate modifications to nursing care, increased monitoring, medical review and calling for assistance.
	Rapid response systems	When severe deterioration occurs, medical emergency teams, outreach teams or liaison nurses are available to respond.
	Clinical communication	Structured communication protocols are used to hand over information about the patient.
Organisational prerequisites	Organisational supports	Executive and clinical leadership support and a formal policy framework for recognition and response systems should exist.
	Education	Education should cover clinical observation, identification of deterioration, escalation protocols, communication strategies, and skills in initiating early interventions.
	Evaluation, audit and feedback	Ongoing monitoring and evaluation are required to track changes in outcomes over time and to check that the RRS is operating as planned.
	Technological systems and solutions	As relevant technologies are developed, they should be incorporated into service delivery, after considering evidence of their efficacy and cost as well as potential unintended consequences.

Recently, the recognition aspect of RRS has been referred to as the afferent limb, whereas the response aspect has been called the efferent limb.¹³¹ The afferent limb involves the use of various track and trigger systems to identify patients at risk of deterioration. The efferent limb is comprised of teams of specialists who provide treatment and care to the deteriorating patient. Each of these components is briefly described.

Afferent Limb

Recognising the deteriorating patient, the afferent limb has focused on measuring clinical signs including vital signs, level of consciousness and oxygenation as well as acting on abnormalities in these measurements.^129,¹³⁰ A variety of scoring systems to identify ward patients with clinical deterioration have evolved as part of the development of critical care outreach,¹³² including the MET,¹³³ early warning scoring (EWS),¹³⁴ and patient-at-risk (PAR) criteria^135,¹³⁶ (see Table 3.9).

TABLE 3.9 Commonly used risk assessment scores

Other modified criteria are also in use.^131,¹³⁷^–¹³⁹ All systems identify abnormalities in commonly measured parameters (e.g. respiratory rate, heart rate, blood pressure, neurological status). Other parameters used in patient assessment are oxygen saturation, temperature in PAR, urine output in PAR and EWS. The EWS/PAR systems include an ordinal scoring approach used as calling criteria for contacting the admitting medical team, ICU staff, the critical care outreach team or the MET, depending on the severity of the patient’s clinical deterioration and the resources available in the local clinical environment.

Efferent Limb

The efferent limb involves the response to clinical deterioration. Two types of services have emerged to respond to deteriorating ward patients: Rapid Response Team (RRT) and ICU Liaison Nurses (LN). RRT is an umbrella term that refers to the teams responding to deteriorating patients.¹³¹ In Australia and New Zealand these teams are known as Medical Emergency Teams (MET), while in the United Kingdom they are referred to as Critical Care Outreach Teams (CCOT) and in North America the umbrella term of RRT is used. Irrespective of the title used, RRT generally comprise an experienced nurse and a doctor, and in the case of North America, may include a respiratory therapist. RRT have replaced the traditional ‘cardiac arrest’ team in many hospitals, although the evidence base on the effectiveness of the system is not clear. RRT assess deteriorating patients and then initiate emergency treatments to stabilise patients. Table 3.10 summarises some of the recent research on RRT. To note, most of the research has been undertaken in Australia, where MET were first developed.

TABLE 3.10 Examples of recent rapid response system research

The second type of efferent limb service is the ICU LN. LN services are a proactive strategy aimed at ward patients who have complex care needs that may overextend the skills of ward staff.¹⁴⁰^–¹⁴² In some hospitals LNs are part of the RRT. The role of the ICU LN is described as one of education of both staff and patients, supervision, follow-up of patients discharged from ICU, liaison and coordination with ward staff, assessment, assistance in development and coordination of the discharge plan, preparation of written documentation and referral.¹⁴³ Despite the broadly defined nature of this role, one of its primary aspects involves supporting other staff, both within and beyond the ICU, in providing continuity of care for ICU patients.¹⁴³ The scope of practice, qualifications and job titles of LNs have yet to be standardised, although a LN Special Interest Group now exists under the auspices of the Australian College of Critical Care Nurses. This group is working to develop a standard role description and core competencies for the LN role. Interestingly, while the literature about LNs is predominantly from Australia and New Zealand, the service is now emerging in countries such as Canada and Scotland.