Institute of Medicine Reports: The Case for Improving Healthcare Safety and Quality

In 1999, the Institute of Medicine (IOM) report To Err is Human: Building a Safer Health System estimated that 100,000 people die every year in the United States as the result of preventable errors made by healthcare professionals.⁵¹ The statistics presented in this study ranked medical errors as the eighth leading cause of death and demonstrated that more people die because of medical errors than from motor vehicle crashes, breast cancer, or AIDS.^7,18,51 The report attributed 7000 of the deaths to medication errors and estimated the annual cost of errors at $17 billion to $29 billion.⁵¹

The To Err is Human report exposed the inadequacies of healthcare systems, focused attention on patient safety, and fostered programs dedicated to measuring and improving performance. Table 22-1 outlines the recommendations for healthcare improvement that were presented in the To Err is Human report and endorsed by the IOM’s Quality of Health Care in America Committee. In a follow-up to the To Err is Human report, in 2001 the IOM published Crossing the Quality Chasm: A New Health System for the 21st Century, to more fully address quality problems related to patient-family satisfaction, treatment disparities, and resource accessibility.^21,64,65 Crossing the Quality Chasm defines the ideal twenty-first century healthcare system as one that is safe, effective, patient-centered, timely, efficient, and equitable. Table 22-2 defines the six aims set forth by the IOM and highlights the recommendations of the Crossing the Quality Chasm report for achieving the ideal healthcare system.

Table 22-1 The Institute of Medicine Quality of Health Care in America Committee’s Recommendations for Health Care Improvement

Tier	Goal	Action
Leadership and knowledge	Establish national leadership, research, tools, and protocols to enhance patient safety knowledge	Development of a Center for Patient Safety within the Agency for Healthcare Research and Quality
Identifying and learning from errors	Identify and learn through errors via mandatory and voluntary reporting efforts with an emphasis on using information to make systems safer	Development of reporting standards by the National Forum for Health Care Quality Measurement and Reporting and analysis of the data by the Center for Patient Safety
		Requirement for state departments of health to report adverse events
		Designation of federal funding for the development of reporting systems
		Creation of legislation to extend peer review protections to data related to patient safety and quality improvement
Setting performance standards and expectations	Raise the standard and expectations for the improvement of safety through the actions of oversight organizations, professional organizations, and group purchasers	Inclusion of patient safety indicators in performance standards
		Implementation of patient safety programs by regulators and accreditors, public and private healthcare purchasers and health professional licensing, certifying and credentialing agencies
		Development of patient safety committees within professional societies
		Increased post-marketing monitoring of drugs by the U.S. Food and Drug Administration
Implementing safety systems in healthcare organizations	Create safe systems within healthcare organizations by implementing safe practices at the delivery level	Incorporation of patient safety programs, with executive sponsorship, into all healthcare institutions
		Adoption of safe medication practices by all healthcare institutions

From Kohn KT, Corrigan JM, Donaldson MS: To err is human: building a safer healthcare system, Washington, DC, 1999, National Academy Press; National Academy of Sciences: To err is human: building a safer healthcare system executive summary. Available at http://www.nap.edu. Accessed October 8, 2011.

Table 22-2 Institute of Medicine Six Aims for Improvement of Health Care Systems

Aim	Definition	Recommendations for Achieving the Six Quality Aims
Safe	Avoid injuries to patients from care that was intended to help them.	• Designate federal funding to healthcare quality improvement initiatives. • Reduce illogical variability in treatment between individual clinicians and institutions and promote the use of evidence-based practices. • Increase transparency of information regarding the performance of healthcare institutions. • Allow patients to access to their own medical records and clinical knowledge. • Implement mechanisms for active communication and exchange of information among healthcare providers. • Create an information technology infrastructure that eliminates the need for handwritten documentation of care. • Align payment policies with quality improvement. • Prepare the workforce to understand and implement quality improvement aims.

Effective Provide services that are based on scientific knowledge to all who could benefit and refrain from providing services to those who are not likely to benefit. Patient-centered Provide care that is respectful of and responsive to individual patient preferences, needs and values and ensures that patient values guide all clinical decisions. Timely Reduce wait time and harmful delays for those who both receive and give care. Efficient Avoid waste such as wasteful use of supplies, ideas and energy. Equitable Provide care that does not vary in quality because of personal characteristics such as gender, ethnicity, geographic location and socio-economic status.

From Committee on Quality Health Care in America and Institute of Medicine: Crossing the quality chasm: a new health system for the 21st century. Washington, DC, 2001, National Academy Press; National Academy of Sciences: Crossing the quality chasm: a new health system for the 21st century executive summary, 2003. Available at http://www.nap.edu. Accessed October 8, 2011.

Through the implementation of the IOM recommendations and the work of organizations that have evolved to improve the safety and quality of healthcare (see the table in the Chapter 22 Supplement on the Evolve Website for a list of organizations and their focuses and websites), individual institutions have demonstrated incremental improvement in care. As late as 2005, however, national statistics had not improved substantially.^2,56 Furthermore, the data compiled in 1999 may have underestimated deficiencies in patient safety.⁵⁶ For example, the Centers for Disease Control and Prevention (CDC) data subsequent to the To Err is Human report suggest that healthcare-acquired blood stream infections may account for as many as 90,000 deaths every year.^19,56 In light of these projections, it is clear that an essential component of contemporary healthcare delivery is improving the safety and quality of healthcare for patients of all ages.

To be successful, improvement efforts must be evidence based. To build an evidence-based foundation, the healthcare improvement movement uses clinical outcome research to identify practices that promote positive patient outcomes. To effectively implement best practices, the discipline adopts improvement theories and methods from businesses such as the automotive, nuclear energy, and aviation industries. Refer to the Chapter 22 Supplement on the Evolve Website for a history of quality improvement.

Human Error

Refer to the Chapter 22 Supplement on the Evolve Website for a history of the industry-based theories of human error. Error theorists recommend a systems approach to improving workplace performance. Extrapolated from successful programs in the nuclear power and aviation industries, a systems approach to preventing human error starts with the notion that people are intrinsically fallible. In addition, it encourages error-proofing the systems in which people work. In a systems approach to error-proofing, layers of defensive interventions that respond to failure-prone human behaviors are used to prevent or trap errors before they cause harm. Often referred to as the Swiss Cheese Model (Fig. 22-1), Reason’s system theory⁷⁹ illustrates that despite layers of defense, it is possible for holes in the safeguards to align in such a way that an error passes through the system. The model proposes that to be maximally effective, multiple defensive layers are necessary to reduce the likelihood that the holes will align and allow the system to fail. In environments such as critical care units, layers of defense are often technologic, such as redundant monitor alarms that sound at the bedside and at a central monitoring station and may also alert staff via a visual cue, such as a flashing light.⁷⁹

Reason notes that errors in systems are caused by either “active failures” or “latent conditions.”⁷⁹ In healthcare, an active failure is an unsafe act performed by a professional who is in direct contact with a patient. A latent condition is a flaw in the design of the care delivery system and its layers of defense. In keeping with the Swiss Cheese Model, patient safety is promoted by barriers to both the active failures of professionals and the latent conditions of a system. Successful error prevention only adds defensive layers and minimizes the number of potential failures that exist in the system. A common approach to prospectively analyzing a system for potential failures is failure mode effect analysis (FMEA). Root cause analysis (RCA) is a technique for assigning a cause to errors that have actually occurred. The methodologies for FMEA and RCA are described later in this chapter. Such preoccupation with errors, their cause, and their prevention is a key characteristic of safe and high reliability industries.⁷⁹

High Reliability Organizations

Reliability is the rate at which a system produces a desired effect without failure.⁸⁰ High reliability organizations perform well in hazardous situations that depend on technology and human interactions.¹¹⁴ (Refer to the Chapter 22 Supplement on the Evolve Website for information about the study of nuclear power and aviation accidents.) By understanding the interplay between the technical and human causes of accidents, the nuclear power and aviation industries have significantly improved the safety of their practices and are now considered high reliability organizations. Regardless of the industry, all high reliability organizations share the following characteristics:

Box 22-1 lists the characteristics of high reliability organizations. In addition to a high rate of positive patient outcomes and a low rate of error and complications, evidence of reliability in a healthcare organization includes continual learning from the analysis of potential system flaws and actual adverse events, the ability of staff at any level to raise concerns and make decisions that are within the scope of their expertise, standardization of processes, and leadership support for safety practices.^16,60,77

Culture of Safety

The ability of a healthcare system to be highly reliable hinges on its underlying organizational culture and leadership support for a culture of safety. To fully support a culture of safety, leadership must prioritize and build consensus on the importance of safety and quality improvement goals and must remove barriers and conflicts of interest to achieving these goals. Leadership can also promote a culture of safety by using nonpunitive approaches to investigating errors and by promoting transparency for the purpose of learning from adverse events.^37,92 Evidence that a culture of safety exists in an organization includes sufficient staff and resources, a flat organizational hierarchy, open dialogue about problems without fear of repercussion, clear lines of communication and effective intradisciplinary and interdisciplinary teamwork.^63,72 The Agency for Healthcare Research and Quality provides a survey that can be used to assess the culture of safety in a hospital³ (Box 22-2).

Crew Resource Management

Effective teamwork and communication are key features of the safety culture in high reliability organizations. Initially developed by the National Aeronautics and Space Administration Aerospace Human Factors Research Division to train flight crews, crew resource management (CRM), is essential for the effective functioning of multidisciplinary healthcare teams (Box 22-3). CRM includes team training via simulations and interactive group briefings that focus on the development of behaviors critical to decision making and performance in high-risk and high-stress environments that depend on technology, process, and interpersonal communication.⁷² The ultimate goal of CRM is a “shared mental model” that aligns all team members to the goals and the strategies for any team activity.⁹⁶ Behaviors taught to team members via CRM include:

In addition to teaching these core behaviors, CRM provides team members with the opportunity to develop technical proficiency and to practice avoiding and trapping errors and mitigating their consequences, and it helps participants appreciate the effects of stress, fatigue, and work overload on performance. Serial performance appraisal of team members is also an important component of CRM programs.^72,96

Healthcare team training programs have gained acceptance over the past 10   years, and performance of labor and delivery, surgical, and anesthesia teams have demonstrated the positive effects of CRM.^30,59,69 The Joint Commission concluded that communication failures contribute to more than half of all errors.⁴⁰ In response, the Joint Commission and other patient safety interest groups have mandated the use of the structured inquiries and briefings that are central to CRM. Examples of such mandated CRM methods include the performance of “universal protocol” or “time-out” to confirm the correct patient, procedure, and site before invasive interventions, and the use of standardized templates or scripts for hand-off communication between caregivers.^{12,40,41,70,84,95}

An appreciation for the importance of CRM has also expanded the use of simulation in healthcare. In critical care settings, the tradition of simulating resuscitations has taken new direction and is now used to teach delegation and communication skills and to reinforce clinical decision-making algorithms and procedural techniques.^29,69 Simulation, which was once a tool reserved for advanced life support and crisis resource management training (such as learning to manage a difficult airway), can be used to develop virtually any aspect of critical care team performance, including noncrisis activities such as daily rounds.^40,69

Model for Improvement

Theories of quality improvement and human error provide a foundation for programs in patient safety and quality improvement, but healthcare teams also need practical methods for orchestrating improvement initiatives. Developed by Associates in Process Improvement, the Model for Improvement (Fig. 22-2) is widely considered an appropriate quality improvement method for healthcare.¹⁰⁴ The Model for Improvement is a two-step process consisting of (1) answering questions to define the improvement opportunity and (2) conducting plan-do-study-act (PDSA) cycles to test interventions that are hoped to improve performance.^24,44,54,104

To successfully improve a process, it is necessary to define the problems associated with the current state, describe the ideal future state, and assign interventions (called improvement actions or changes) to produce the ideal future state. To identify these elements of an improvement effort, the first step of the Model for Improvement requires answering the following questions:

Answering these three fundamental questions leads to the development of an aim for the initiative. The aim of an improvement initiative must be time specific and measurable, and it must articulate the specific population and scope that will be affected by the effort. An example of a quality improvement aim is: central line-related blood stream infections in patients receiving care in the pediatric critical care unit will be reduced to one infection per 1000 line days (or fewer) within 6   months.

Although the need for improvement may come from external sources such as hospital administration or a regulatory agency, a team of healthcare providers who work within the targeted process or system should be assembled to create the aim, identify improvement actions, and then perform the improvement effort. To comprehensively represent all stakeholders, a typical quality improvement team is composed of the following:

When the initiative includes direct patient care or operations that support patient care, the team consists of members of the clinical microsystem. The clinical microsystem is the group of people who work together on a regular basis to provide care to a discrete population of patients. Members of the clinical microsystem are best able to provide subject matter expertise and encourage the adoption of change throughout an organization.⁶⁸

Once the team determines an aim, improvement actions intended to achieve the aim can be tested using PDSA cycles, also called tests of change. These tests can be conducted repeatedly to refine an improvement action. A single PDSA cycle consists of:

When an improvement effort requires multiple improvement interventions, each intervention can be tested via a separate PDSA cycle. All the related PDSA cycles are then linked to create a package of interventions that have been proven to foster improvement (Fig. 22-3). Effective improvement actions can then be spread beyond the initial test setting and population. The use of PDSA cycles to test and refine improvement actions can help to overcome an organization’s natural resistance to change.^24,44,54,105

Performance Measurement

Measures must be developed to identify opportunities for improvement, to test changes, and to continually monitor an organization’s performance. A measure is a quantitative indicator of performance. There are three types of improvement measures:

The goal of reducing the incidence of central line-associated blood stream infections can be used to illustrate the three types of performance measures. The following are sample measures for a central line-associated blood stream infection improvement project:

To analyze performance patterns represented by data, it is helpful to graphically display and apply statistics to the data. Run charts and control charts, which plot data over time, are two essential improvement tools. Refer to Evolve Figs. 22-1 and 22-2 in the Chapter 22 Supplement on the Evolve Website for examples of such charts.

Failure Mode and Effects Analysis

Before the Model for Improvement can be used to manage performance, opportunities for improvement must be identified. As mentioned earlier, high reliability organizations are preoccupied with failure and proactively analyze processes and systems for potential failures, even before failures occur. An FMEA is a tool for prospective analysis of healthcare system vulnerabilities, called failure modes. Steps to a healthcare FMEA:

Table 22-3 provides a sample documentation grid and RPN calculation tool for conducting an FMEA. An FMEA can be used to identify new improvement projects or as a tool to further understand special cause variation demonstrated in the data generated by an existing improvement project. An improvement team that includes all key stakeholders, including members of the clinical microsystem, should be used to conduct FMEAs.⁴³

The Institute for Safe Medication Practices has strongly encouraged the use of FMEA, and it has been shown to improve medication safety.^81,91,115 In response to the overwhelming evidence that the technique contributes to patient safety, the Joint Commission mandates that healthcare organizations perform at least one FMEA per year.³³

Incident Reporting

Healthcare organizations should use tools, such as FMEA, to anticipate and prevent failures before they occur. However when a failure takes place, incident reporting becomes central to the safety and quality improvement effort. Colonel John C. Flanagan, the Director of the Division of Aviation Psychology, developed the critical incident technique for military training in 1954. The critical incident technique was a structured incident reporting process that included the use of personnel directly involved in a failed process to create, at the time of discovery, a qualitative narrative of the event. Flanagan’s work also established common terminology for incident reporting that has been adapted for use in healthcare (Table 22-4).^28,51,109 Healthcare incident reporting is the process by which a clinician documents adverse events, medical errors, and near-miss events, including deviations from standard procedures that could have resulted in an adverse event or error, in real time and at the point of care. Incident reports may also require clinicians to assign a perceived cause and severity grade to the event.^31,42,55,109

Table 22-4 Healthcare Incident Definitions

Term	Definition
Adverse event	An injury that was caused by medical management rather than by a patient’s underlying disease
Medical error	The failure of a planned action to be completed as intended or to achieve its intended goal or the use of the wrong action to achieve a goal
Serious error	An error that causes permanent injury or transient but potentially life-threatening harm
Minor error	An error that causes harm that is neither permanent nor potentially life threatening
Near miss	An error that could have caused harm, but did not because of timely intervention or chance
Sentinel event	An unexpected occurrence involving death or serious physical or psychological injury, or the risk thereof

Adapted from Garbutt J, et al: Reporting and disclosing medical errors: pediatricians’ attitudes and behaviors. Arch Pediatr Adolesc Med 161:179-185, 2007; Joint Commission: Sentinel events. Available at http://www.jointcommission.org/SentinelEvents. Accessed April 25, 2008; Kohn KT, Corrigan JM, Donaldson MS: To err is human: building a safer healthcare system. Washington, DC, 1999, National Academy Press.

Since 1975, the U.S. Food and Drug Administration has mandated the reporting of blood transfusion reactions, and in 1995 the Joint Commission mandated reporting of all other adverse events, errors, and near misses as part of the Sentinel Event Policy. The hallmark IOM report To Err is Human called for the enhancement of voluntary and mandatory incident reporting systems in individual organizations and for public disclosure. Furthermore, and as noted previously, documenting incidents is an important characteristic of High Reliability Organizations.^55,109 Despite a keen focus on incident reporting, however, there are many barriers to the use of incident reports by healthcare providers.^31,97 In a 2004 survey by Taylor et al,⁹⁷ 34% of pediatricians and pediatric nurses surveyed indicated that they reported fewer than 20% of their own medical errors, and approximately 33% of those surveyed indicated that they reported fewer than 40% of the errors committed by their colleagues. Other studies have indicated incident reporting rates as low as 10%.⁵¹ Barriers to incident reporting can be inadequacies of reporting systems, organizational culture, or the attitudes or knowledge deficits of healthcare professionals (see Barriers to Incident Reporting in the Chapter 22 Supplement on the Evolve Website for examples).

Despite the barriers to incident reporting, pediatric healthcare providers indicate a willingness to report if and when appropriate and an interest in understanding incidents to improve care.³¹ Facilitating the use of incident reports is a key improvement effort of all hospitals. Effective strategies to improve incident reporting rates include:

All hospitals should implement a system that provides timely information to front-line clinical leadership about the occurrence of and trends in incidents. Electronic incident reporting systems help to accomplish this goal.^98,100 Continual feedback to clinicians about the improvement actions taken in response to incident reports is also essential to increasing reporting rates and to building a culture of transparency and learning. Often a reported event will undergo an extensive, systematic evaluation, such as an RCA, that will generate a new improvement initiative for the clinical microsystem.

Root Cause Analysis

Performing an RCA of actual adverse patient events, like FMEA and incident reporting, is a characteristic of a high reliability organization and is mandated by the Joint Commission, as well as by many state regulated incident reporting programs. Reporting programs, which are often public, typically require an RCA of a “sentinel event” or an “unexpected occurrence involving death or serious physical or psychological injury, or the risk thereof.”⁴⁶ The goal of an RCA is to identify the factors that contributed to an adverse patient event in order to safeguard the system from the same event occurring again. Although an RCA may identify multiple system vulnerabilities, the emphasis of the exercise is to assign and understand the causes of the failure that resulted in death or serious morbidity or the potential for such.²⁶ Steps for conducting an RCA include:

To be effective, the approach to conducting an RCA should be consistent from one event to the next. RCA should be conducted by representatives of all stakeholders in the system in which an event occurred, including members of the clinical microsystem. To provide an objective analysis of the event, one or more members of the organization who sit outside the system under examination may also be included. In addition, a neutral facilitator should assist with the organization of the RCA and moderate related discussions and debates.^9,15,107

Environment of Care

Pediatric critical care is highly vulnerable to errors and adverse patient events.^61,85 The needs of critically ill children are complex and heterogeneous, the systems for providing care to critically ill children are tightly coupled, and the work of critical care providers is psychologically stressful and physically exhausting. Accidents related to technology, including dislodgement of invasive catheters and airways and ineffective use of monitoring alarms by staff, are prevalent in critical care, and gaps in care continuity and cognitive biases are particularly problematic when fast-paced decision making is needed.^61,102,103 Elements of the critical care physical environment, such as open layouts and excessive noise, produce interruptions and distractions during care delivery. These elements are also known to impede patient safety in pediatric critical care units.⁶¹ In a study that applied an aviation “sterile cockpit” approach to critical operations, medication errors were reduced when interruptions experienced by the nurse during medication administration were minimized.^61,74 Furthermore, the general effects of fatigue on clinician performance are applicable, if not amplified, in pediatric critical care, where on-call physicians have little time to sleep and nurses frequently work extended shifts. The documented effects of sleep deprivation include a 36% increase in medication and diagnostic errors and a 30% increase in surgical procedure times.^53,61 Similarly, research has shown that nurses are twofold to threefold more likely to make an error after working more than 12.5 consecutive hours.^61,82,87 A 2006 study conducted by the American Association of Critical-Care Nurses reported that 44% of critical care nurses are routinely scheduled to work shifts longer than 12.5   hours and that during a 1-month period, only 13% of nurses left work at the end of their assigned shift and 61% worked 10 or more overtime shifts.^61,87 The ideal hospital environment culture of safety includes leadership support, teamwork, learning from mistakes, work hour restrictions, work load distribution, appropriate physical space, and decision support aids that are readily available at the point of care.^37,73

Improving the hospital work environment is clearly pertinent to nursing. Leape et al⁵⁶ reported that as many as 86% of harmful errors are intercepted by nurses before they reach patients. Several studies have shown that when patient-to-nurse ratios increase and when nurses spend less time with patients, there is an increase in complications such as infection, gastrointestinal bleeding, and cardiac arrest, as well as extended length of hospital stay.^52,67,78 Aiken et al^4,5 have published hallmark research that correlates low nursing education level and high patient-to-nurse ratio with increased patient mortality rates and increased failure to rescue patients who are experiencing physiologic instability resulting from an unexpected complication. Aiken et al estimated that for every patient receiving care beyond an assignment of four patients there is a 7% increase in the risk of a patient dying within 30   days after discharge.⁵ These statistics are particularly daunting in light of the estimated 29% to 40% nursing shortages predicted by the year 2020.³⁷

In response to evidence that nursing care plays a pivotal role in patients’ outcomes, the IOM Committee for the Work Environment of Nurses and Patient Safety published a manual titled Keeping Patients Safe: Transforming the Work Environments of Nurses. The manual, intended to be a companion to the IOM report To Err is Human, provides a blueprint for the ideal work environment for nurses.⁷³ The blueprint builds on research that demonstrates the effects of nursing education, management, staffing, work hours and processes, and the ergonomics of the work space. In addition, the blueprint links the development of a safe work environment with the evolution of a learning organization—that is, an organization that is willing to adapt over time to best meet the needs of its clients.^37,38,73 The requirements set forth by the IOM for a safe and effective work environment for nurses include an organizational commitment to patient safety, effective and evidence-based management and leadership, interdisciplinary collaboration, adequate staffing and work hour restrictions, workplace design that facilitates safe care, decision support aids at the point of care, and opportunities for ongoing learning.

Failure to Rescue

The ultimate outcome of a healthy work environment is the prevention of failure-to-rescue events. Failure to rescue is any event during which a clinician is unable, because of a complication of care or a condition that was not present on admission, to save a hospitalized patient’s life.⁴ To recognize and intervene early when a life-threatening complication occurs, nurses use surveillance such as clinical assessment for signs and symptoms of physiologic instability, and environmental safety checks, such as verifying the settings of monitor alarms.⁴ A workforce and work environment that support surveillance, recognition, and early intervention by nurses are the backbone for rapid response teams (RRTs). RRTs are hospital-based teams of critical care providers who are available to assess, provide consultation, and assist with the care of a patient whose status is deteriorating prior to the need for resuscitation for cardiopulmonary arrest.¹¹⁶ RRTs have been successfully implemented in pediatric hospitals and have been shown to reduce the rate of cardiopulmonary arrests that occur outside of critical care units and reduce patient mortality. The use of RRTs is strongly encouraged by the Joint Commission and other patient safety interest groups.^48,88,90,116

Role for Information Technology

There is great promise that evolving information technology will enable safe environments of care. Electronic health record (EHR) and computerized provider order entry (CPOE) systems offer benefits to clinicians, including expanded access to patients’ health information, particularly during and after transitions of care, and decision support such as drug-allergy interaction checking and dose calculation.^99,112 By offering identification and tracking applications, electronic information systems can also enhance patient identification and patient flow.⁷⁶ Furthermore, electronic capture of clinical information facilitates data mining for clinical research and measurement of performance. The IOM, the Joint Commission, the Center for Medicare and Medicaid Services, the Institute for Safe Medication Practices (ISMP), public interest groups, and many private insurers include electronic information systems in their healthcare improvement requirements and pay-for-performance plans.^6,34

It is important to note that there are risks associated with EHR and CPOE systems. Because EHR and CPOE systems require changes in routine clinical operations and processes, it will be important to verify that clinical workflow is safe and effective after implementation. Other risks include lack of information security, alert fatigue, downtime, inefficiency, and the evolution of new types of medical errors, such as keystroke mistakes.^11,110,112

In addition to the generic risks of clinical information systems, there is national concern that commercially available EHR and CPOE systems do not accommodate the unique needs of pediatric healthcare.^{6,89,93,110,112} The Joint Commission mandated that information system vendors develop functionality to help prevent pediatric medication errors,^46,47 and the American Academy of Pediatrics has developed position statements on pediatric requirements for clinical information technology.^6,93 Table 22-5 outlines the pediatric-specific requirements for maximal safety and effectiveness of EHR and CPOE systems.

Table 22-5 Requirements for Safe and Effective Pediatric Electronic Information Systems

Requirement	Electronic Health Record	Computerized Provider Order Entry and Prescribing
Immunizations	Ability to record immunization data and link to state immunization registries	Immunization decision support with reminders regarding immunization schedules and electronic entry of immunization orders
	Ability to print an immunization record for families
Growth documentation and tracking	Automated growth charts with alerts when children fall outside of growth parameters for age, including gestational and adjusted age for premature infants	Weight displays on all order screens
	Ability to document gestational age	Reminders to document a patient’s weight
	Accommodation of normative values that change specific to the child’s age (e.g., vital sign norms)	Stopping functions that prevent providers from ordering medications without a weight entry in the system
Medication management	Presence of medication lists and the ability to view medication history and a medication administration record	Dosing by body weight or body mass index
	Electronic medication administration record with decision support prompts for nurses	Dose range checking
		Rounding to safe and convenient doses
		Age-based (including gestational age) dosing decision support
		Indication-based dosing decision support
		Drug allergy checking, drug-drug interaction checking, drug-food interaction checking, and duplicate medication alerts
Patient identification	Availability of two patient identifiers on all screens and printed reports	Availability of two patient identifiers on all order screens and printed copies of orders
	Mechanism for newborn infant identification before a name or Social Security number has been assigned
	Ability to store prenatal data for an identified fetus and connect this data with the patient after birth, without relying on maternal identifiers
	Ability to change an infant’s name
	Ability to assign unknown or ambiguous sex
Privacy	Ability to flag and limit access to protected information, including adolescent health issues as dictated by state laws	Ability to flag and limit access to orders reflecting protected information, including adolescent health issues as dictated by state laws (e.g., pregnancy test)
	Ability to accurately and confidentially identify foster and custodial care

Adapted from American Academy of Pediatrics Council on Clinical Information Technology, Gerstle RS: Electronic prescribing systems in pediatrics: the rationale and functionality requirements. Pediatrics 119:1229-1231, 2007; Joint Commission: Preventing pediatric medication errors. Sentinel Event Alert, 39, April 2008. Available at http://www.jointcommission.org/SentinelEvents/SentinelEventAlert/sea_39.htm. Accessed April 25, 2008; Spooner SA: Council on Clinical Information Technology, American Academy of Pediatrics. Special requirements of electronic health record systems in pediatrics. Pediatrics 119:631-637, 2007.

Additional technologies, such as bar code identification systems and programmable infusion pumps, also have the potential to improve patient safety in pediatric critical care units. Despite somewhat conflicting evidence in the literature, it appears that bar code technology reduces medication administration and lab labeling errors^20,35,62 and that programmable or “smart” infusion pumps reduce duplicate infusion and infusion rate errors.⁷¹ Additional research is needed to determine the most appropriate uses and the true effects of such technologies.^20,58

Common Pediatric Critical Care Unit Improvement Opportunities

There are unique impediments to improving patient safety in pediatric critical care. Many of the safety and quality improvement efforts launched by regulatory and public interest groups primarily focus on adult populations and may not match the safety risks specific to critically ill children.⁸⁵ In light of the high risk for iatrogenic complications and the special needs of sick children, pediatric nurses and physicians must identify the key safety and quality indicators for pediatric critical care.⁸⁵ A number of professional organizations, such as the National Initiative for Children’s Health Care Quality (NICHQ), the Child Health Corporation of America (CHCA), and related work groups have been formed to define and prioritize pediatric specific safety and quality initiatives and measures. Box 22-5 lists targeted pediatric safety and quality issues; it can be used by pediatric critical care providers to develop their own unit-specific improvement goals. Medication safety and healthcare-acquired infections are the two most consistent themes of all proposed pediatric safety and quality improvement programs.