The Value of a Cross-Disciplinary Approach to Human and System Performance Research in Obstetrics and Neonatology: AHRQ’s Patient Safety Learning Laboratory

Abstract

Objective:

In creating an Agency for Healthcare Research and Quality (AHRQ) Patient Safety Learning Laboratory (PSLL), our objective has been to establish a multidisciplinary research environment focused on the safe care of pregnant women and newborns. This manuscript describes work performed under grants P30 HS023506 (obstetric focus) and R18 HS029123 (neonatal focus).

Methods:

We follow AHRQ’s 5-step approach to systems engineering in health care: problem analysis, design, development, implementation, and evaluation. Within this 5-step approach, methods used include interviews, focus groups, direct observation, teamwork scales, flow disruption analysis, the Systems Engineering Initiative for Patient Safety model, design thinking, and simulation-based testing of processes and prototypes.

Results:

Grant P30 HS023506 is completed. The physical characteristics of 10 labor and delivery units were examined, finding significant heterogeneity in size, design, and organization. Task analysis revealed multiple obstacles to optimal team performance. We designed and tested a delayed cord clamping cart to address inherent ergonomic challenges. Finally, we identified common lapses in verbal communication during obstetric emergencies. Grant R18 HS029123 is ongoing. Eighteen Need Statements serve as the basis for exploratory work in mitigating threats to neonates during resuscitation, including a task analysis to determine points of intervention. We are developing (a) novel resuscitation platforms, (b) improved methods of equipment/supply organization, (c) new means of acquiring, displaying, and processing multiple data streams, and (d) innovative techniques and devices for neonatal intubation.

Conclusions:

The approach to systems engineering in health care supported by AHRQ’s PSLL funding mechanism fosters critical thinking about safety issues by facilitating the integration of investigators with diverse, complementary expertise. By encouraging such collaboration, AHRQ’s 5-step process enables important questions to be answered. The PSLL mechanism is a valuable resource for the patient safety community.

The Agency for Healthcare Research and Quality (AHRQ) established its Patient Safety Learning Laboratory (PSLL) grant program in 2014. Since its inception, AHRQ has funded 47 multiyear grants directed at investigations across the clinical spectrum. Per the AHRQ website, “Patient safety learning laboratories (PSLLs) take a systems engineering approach to allow researchers and practitioners to evaluate clinical processes and enhance work and information flow to improve patient safety. The learning laboratories use cross-disciplinary teams to address the patient safety-related challenges providers face. This approach can involve evaluating the physical (built) environment, technological factors such as health information technology (IT), and clinical workflow processes relevant to the patient’s condition. Emphasis is placed on the system-level confluence of these multiple factors in producing better patient safety.”¹ By supporting a systems engineering approach, AHRQ’s PSLL program represents an important source of support for investigators desiring to approach important patient safety problems by focusing on key issues in human and system performance. This mechanism has proved critical to 2 projects at our institution focused on maternal and neonatal patient safety:

• P30 HS023506: optimizing safety of mother and neonate in a mixed methods learning laboratory (2014-2019).

• R18 HS029123: applying human factors science, design thinking and systems engineering to mitigate threats to neonates undergoing resuscitation and stabilization (2022-projected 2026).

This manuscript will describe how the application of systems engineering principles has enabled our teams to identify key safety questions and study objectives, develop and implement effective methods (such as highly realistic simulation), generate objective evidence leading to promising patient safety improvements, and reach important conclusions.

Our focus on maternal and neonatal care at the time of delivery is driven by several factors. In the obstetric domain, the spectrum of patient acuity ranges from healthy women undergoing a natural though dynamic process, to critically ill women whose physiologic parameters are dramatically affected by the changes associated with pregnancy, labor, and the immediate postpartum period. Obstetric patient acuity is rising, with increased rates of advanced maternal age; chronic medical co-morbidities such as obesity, diabetes, and hypertension; and mounting rates of cesarean delivery.^2,3 After peaking during the COVID pandemic at >30 deaths per 100,000 live births, the U.S. maternal mortality rate has fallen to approximately 20 deaths per 100,000 live births in 2024; yet this rate remains one of the highest of developed countries.⁴ Although most pregnancies involve the delivery of healthy full-term (39-40-wk gestation) newborns, not all are healthy nor delivered at term. Every year in the US, ~1 in 10 of all newborns (400,000 neonates) requires resuscitation, a time-pressured activity requiring teams of health care professionals (HCPs) to carry out invasive procedures in a specific sequence of steps in a relatively constrained volume of space (Fig. 1).⁵ This need for resuscitation and stabilization results from problems such as prematurity, infection, perinatal depression, the presence of congenital anomalies, or other issues that prevent a normal transition to extrauterine life. Even the most uncomplicated delivery occurring in a low-acuity, low-volume hospital or birth center can evolve within minutes to become a life-threatening emergency for the mother and/or newborn. Although many variables affect maternal and neonatal morbidity and mortality, and not all are related to the time period immediately surrounding birth, it can be easily appreciated that the perinatal period is nevertheless fraught with danger to both mother and newborn. As such, it is an important area for study and intervention.

FIGURE 1.

Neonatal resuscitation in the delivery room: space-constrained environment.

METHODS

AHRQ requires that grantees applying for the PSLL systems engineering funding mechanism take a 5-step approach: problem analysis, design, development, implementation, and evaluation (Table 1). Within this 5-step approach, our team has applied a number of different methods drawn from diverse yet interrelated fields of research. A description of these methods as applied in both of our PSLL grants follows.

TABLE 1.

Patient Safety Learning Laboratory Methods

Step	General approach
Problem analysis	Observation of actual clinical practice. Interviews and focus groups with clinicians and/or families. Study of key problem areas in the simulated clinical setting.
Design	From the problem analysis, identify key design features for modification or innovation.
Development	Create or modify equipment, physical layout, policy / procedure, or behavioral intervention. Follow an iterative process of study, re-design, and development.
Implementation	Conduct a randomized controlled trial to determine if the innovation can improve safety. For innovations ready to test in the clinical setting, implement in the appropriate unit of the hospital or clinic.
Evaluation	Analyze results of the controlled trial for both objective and subjective evidence of clinical efficacy and improved human and system performance. Follow clinical outcomes and process measures for the implementation of the innovation in the clinical setting.

Problem Analysis

Identification of problem areas to explore is accomplished through: (1) interviews and focus groups with clinicians and family members (where appropriate); and (2) direct observation of actual or simulated clinical practice in real-time or by review of audiovisual recordings or transcripts of individual and team performance. Field notes are taken during observations, full descriptions written as soon as possible after the observation is concluded, and interpretations checked with clinicians in real-time when possible.^6–9 This allows us to build a theoretical model describing the components or subsystems of the existing patient care system, analyze the function of these subsystems and identify threats to optimal system functionality, and finally generate hypotheses regarding how to modify subsystems to improve the safety, efficiency, and effectiveness of care.

Specific methods used to assess subsystem performance include: (1) teamwork scales to assess behavioral performance (e.g., NOTECHS);^10,11 (2) Flow Disruption analysis (FD) of deviations from the expected or optimal state,^12–14 and (c) the Systems Engineering Initiative for Patient Safety (SEIPS) model that describes how people, tasks, tools, technology, working environment, organization and components of the external environment (e.g., federal and state regulation) interact to affect performance.^15–17 Elements of SEIPS 1.0, 2.0, and 3.0 are used to describe the work system from the perspectives of individual HCPs, health care organizations and patients and their families. We also use grounded theory, a systematic method for studying human social interaction, to guide qualitative data collection and analysis. Grounded theory facilitates analysis across individual, group, unit, and organizational levels, consistent with the SEIPS 2.0 model.^18–20

Design

On the basis of the problem analysis, the design phase identifies key procedural or physical features whose modification or innovation are hypothesized to remediate threats to patient care. We use characteristics of design thinking to develop our design objectives (Table 2). This includes not only asking clinicians, patients, and family members their perception of what is needed to address overt problems, but also discerning latent needs that may not be readily apparent to these groups. The focus at this stage is on subsystems.

TABLE 2.

Characteristics of Design Thinking to Inform Design Objectives

Design thinking characteristic	Explanation
UNDERSTAND (empathy)	Understanding is the first phase of the design thinking process, in which investigators will be immersed in learning. They conduct interviews, focus groups and qualitative studies in several settings. The goal is to develop background knowledge through these experiences. Developing understanding is a springboard for beginning to address design challenges.
OBSERVE (again and again and again and again…)	Investigators need to become keen people watchers in the observation phase of the design process. They watch how people behave and interact and observe physical spaces and places in several settings. They talk to people about what they are doing, ask questions, and reflect on what they see. The understanding and observation phase creates empathy.
DEFINE (develop design objectives based on identified needs)	In this phase of design thinking, the focus is on becoming aware of peoples’ needs and developing insights. The phrase “How might we….” is often used to define a point of view, which is a statement of the: user + need + insight. This statement ends with a suggestion about how to make changes that will have an impact on peoples’ experiences.
IDEATE (brainstorm)	Ideating is a critical component of design thinking. Investigators brainstorm a myriad of ideas and suspend judgment. Ideating is all about creativity and innovation. In the ideation phase, quantity is encouraged. A hundred ideas may be generated in a single session. Investigators become savvy risk takers, wishful thinkers, and dreamers of what might be possible.
PROTOTYPE (make)	Prototyping is a rough and rapid portion of the design process. A prototype can be a sketch, model, or a cardboard box. It is a way to convey an idea quickly. Lesson: it’s better to fail early and often through prototypes. This is a key element of CAPE’s philosophy.
TEST (on real users)	Testing is part of an iterative process that provides designers with feedback. The purpose of testing is to learn what works and what doesn’t, and then iterate. This means going back to the prototype and modifying it based on feedback. Testing ensures that investigators learn what works and what doesn’t work for users. This is also a key element of CAPE’s philosophy and is integrated into each project’s goals.

Development

Subsystem development is optimized through repeated testing and modification of original design ideas. Simulation facilitates safe testing of prototypes and rapid iterative modification of novel procedures and technologies in real-time. By minimizing confounding variables across scenarios and subjects, intense focus is placed on specific design features.²¹

Implementation

Once individual subsystems are designed and developed, they must then be integrated to determine how they function when used together within the system as a whole. Again, this is readily accomplished using highly realistic simulated clinical environments that are configured in different ways to study the effects of variables such as room layout, visibility, noise, and other performance-shaping factors. Realistic simulated development environments include the inanimate objects intrinsic to real health care environments; realistic patient simulators; and the human beings who respond in an authentic manner to the events that occur in those environments.²² Complex visual, auditory, and tactile cues are recreated in high fidelity to engender individual and team behaviors and subsystem/system performance consistent with an operational context. Events and discussion occurring during simulated implementation are recorded with multiple cameras and microphones, enabling post hoc review and quantitative and qualitative data collection and analysis. Objective debriefings are conducted after comprehensive simulations; our debriefing methods are described in detail elsewhere.²³ In this way, simulation facilitates understanding “work as done” as opposed to the idealized concept of “work as imagined.”

Evaluation

Comprehensive evaluation of the novel procedures and technologies is performed during highly realistic simulated clinical scenarios and, where possible, in the real clinical environment. Once again, tools such as NOTECHS, FD, and SEIPS are used at this stage.

During all 5 phases, any data entered on paper is quickly transcribed to digital files and paper documents are shredded. These digital files and all devices and media used to record and store digital data are encrypted per Stanford University policies.

RESULTS

The work conducted under grant P30 HS023506 focused on the obstetric patient is completed and will be summarized first.

Our multidisciplinary team initially assessed the physical characteristics of a convenience sample of ten labor and delivery (L&D) units in California.^24,25 Data were gathered through interviews, observation, and physical measurements. These hospitals represented a range of centers with annual delivery volumes ranging from <1000 to >5000. Similarly, cesarean section rates ranged from 19.6% to 39.7%. Together, these hospitals accounted for 34,161 deliveries based on California Maternal Quality Care Collaborative (CMQCC) data. Significant heterogeneity in labor room (LR) and operating room (OR) size, count, and number of configurations was found to exist not only among the units of the ten different hospitals but also within units in the same hospital. There was no standard found for either LR or OR design or overall size. Specific aspects of current design that were consistently identified as problematic included (a) lack of rapid access to blood to treat maternal and fetal/neonatal hemorrhage, (b) inadequate operating space for neonatal resuscitation teams, and (c) deficient methods for restocking and organization of equipment and supplies. This lack of standardization and optimization creates human factors challenges for teams operating in those spaces and represents an opportunity to improve human and system performance when designing future L&D units.

To better understand how the physical OR impacts human performance during emergency cesarean section, a task and equipment analysis was conducted through focus group interviews and surveys that were administered to a convenience sample of 34 multidisciplinary obstetric team members.²⁶ Data were coded and mapped by specialty to identify areas of congestion and the time spent in each area. Complex interdependencies between specialties were identified. Thirty task groupings requiring over 20 pieces of equipment were identified. Perceived areas of congestion and areas of time spent in the OR varied by clinical specialty. The main challenges encountered during an emergency cesarean delivery included (a) limitations imposed by the size, layout, and orientation of the OR, (b) the volume and location of the equipment within the OR, and (c) issues with patient transport into the OR. This represents yet another opportunity to improve human and system performance when designing L&D units.

Delayed (as opposed to immediate) clamping of the umbilical cord (delayed cord clamping, DCC) after birth in otherwise healthy newborns for at least one minute is standard of care. Investigation of the practicality and safety of DCC in neonates requiring resuscitation is ongoing; this requires a platform, or delayed cord clamping cart (DCCC), capable of supporting such an intervention. Our team designed and built a prototype of such a DCCC and conducted simulation-based testing of the device.²⁷ We then evaluated the feasibility of its use in 20 low-risk neonates delivered through nonemergency cesarean section.²⁸ Data were collected through direct observation in real-time, recorded post hoc debriefings, and surveys of HCPs and patients. Forty-four HCPs responded to written surveys: 16 (36%) were very satisfied, 12 (27%) satisfied, 13 (30%) neutral, and 3 (7%) were somewhat dissatisfied with use of the DCCC in the OR. Twenty patients provided subjective feedback: 18 (90%) reported that they felt safe with the device in use. Our team has subsequently applied for a patent on this device and additional testing during use in real deliveries is ongoing.

Our team also investigated communication in the OR.²⁹ We first identified common lapses in verbal communication during simulated obstetric emergency scenarios. All scenarios were recorded and reviewed to identify questions that were repeated within and across scenarios. Certain questions were commonly repeated both within and across 27 simulated scenarios. The median number of questions asked was 27 per simulated scenario. Commonly repeated questions focused on 3 general topics: (a) historical data/information such as estimated gestational age, (b) maternal clinical status such as estimated blood loss, and (c) presence of specific personnel. Verbal transfer of data/information was found to be inefficient at best, inadequate at worst during simulated obstetric emergencies. Other more reliable and effective means of communicating key static historical information and specific dynamic clinical data are necessary to facilitate optimal human performance.

As an extension of our work on communication, we designed a visual display to improve the flow of crucial information between HCPs during emergencies in L&D.³⁰ This central monitor presented critical information and cognitive aids (such as checklists) to guide treatment during specific emergency situations. Relevant information was obtained through the following:

• Focus group interviews of HCPs in obstetrics, obstetric anesthesia, and obstetric nursing.

• Review of simulated obstetric emergencies, specifically to identify information that was repeatedly requested by team members.

• Consultation with experts in human factors and ergonomics.

Appropriate clinical checklists were also embedded in the display. Usability testing was performed on a prototype during simulated clinical events and subjective feedback from the subjects testing it was generally favorable. However, a significant limitation to use of the display was the need to manually input real-time data such as maternal blood loss that is not automatically uploaded into the electronic medical record (EMR). This is a significant issue with current EMRs and will need to be solved before the EMR can be used for real-time decision support.

The work conducted under grant R18 HS029123 is focused on the neonatal patient and is ongoing. To date our multidisciplinary team has produced the following preliminary results:

Multiple focus groups were conducted with HCPs who actively participate in caring for neonates in need of resuscitation in the delivery room. These focus groups identified a series of clinical problems that have been translated into a list of 18 formal Need Statements modeled after the approach used by the Stanford Mussallem Center for Biodesign program.³¹ Each Need Statement lists a discrete need (requirement) followed by a set of criteria to be met to adequately meet the need. In addition, potential thoughts as to how to address the need, generated by members of the focus groups and augmented by our investigator team, are listed as a starting point for design and testing. These need statements serve as the basis for our initial exploratory work in mitigating identified threats to neonates during resuscitation. An example follows:

Need Statement

A neonatal resuscitation environment that enables optimal human performance by HCPs with differing anthropometric characteristics.

Criteria

• Flexible features/components

• Easily and quickly adjustable components

Initial Thoughts

• Define the range of anthropometric measurements in a population of HCPs.

• Define features of resuscitation environments/devices that can be varied to accommodate needs (e.g., bed height).

Midway through the current grant, we have generated preliminary results while exploring the following identified needs:³²

Need: A way to improve adherence to Neonatal Resuscitation Program (NRP) guidelines during newborn resuscitation to reduce errors, neonatal morbidity/mortality, and legal liability.

The NRP sets the evidence-based standard in the US for resuscitation of newborns. To better understand potential methods for improving human and system performance during this task, our team is taking a human factors approach to creating a detailed task analysis of the NRP’s algorithm. By reviewing recorded simulated neonatal resuscitations and recorded actual resuscitations and directly observing simulation-based resuscitation training and debriefings of those events at CAPE, we are documenting the complexity and time-pressured nature of the neonatal resuscitation environment. This serves as the basis for determining points of intervention when addressing the other 17 needs statements.

Need: A way to create more physical space in the room for neonatal resuscitation team members.

As a starting point in experimenting how to address this need, we are conducting a literature search for references to guidelines, codes and user experiences regarding the design and construction of labor and delivery rooms, cesarean section rooms, and neonatal resuscitation and stabilization suites used to resuscitate newborns immediately after delivery. Key development criteria include the ability of any redesigned space to accommodate larger numbers of HCPs on resuscitation teams (as well as the equipment and supplies needed for resuscitation) while not compromising the space available to the delivering patient and the obstetric team and avoiding any requirement for new construction or extensive renovation. One of the first results of this work is a novel resuscitation bed design (discussed in the following paragraphs) and taking a 3-dimensional approach to the use of space around the bedside for the placement of people, equipment, and supplies.

Need: A way to acquire, process, and optimally display multiple data streams before and during neonatal resuscitation.

Neonatal resuscitation teams benefit from advance knowledge of both chronic and acute fetal conditions that predict the need for resuscitation at the time of birth. We have found critical fetal data are located in multiple different electronic and physical locations within our hospital. We have also found that accessibility to this data, especially in time-pressured situations, is limited and, in some circumstances, nonexistent. We are actively identifying all sources of this important data, logging its location, and determining the optimal means of making it available to resuscitation teams. Because data must be translated into actionable information to positively impact patient care, we have begun to explore data translation issues including but not limited to location at the bedside, font size/color/brightness, text versus icons, numbers versus graphs, types of alerts (visual, auditory, tactile) and other variables associated with data display. Although this will initially involve working with traditional electronic displays, the need for experimentation with novel methods of data display, as used in other high-risk industries such as the military and aerospace, has become clear to our team.

Need: A way to minimize handling/transfers of neonatal patients.

For neonates requiring resuscitation, care in the delivery room includes DCC (carried out in close proximity to the maternal bed), transport of the neonate to a radiant warmer for resuscitation and stabilization, and then transport to the appropriate nursery for ongoing care. This requires multiple movements onto different surfaces, each with the potential for trauma, cold stress, dislodgement of or disconnection from monitoring and life support equipment, and other sequelae. We are currently simulating the use of different resuscitation bed sizes and orientations in a series of increasingly complex resuscitation scenarios involving the need for chest compressions, umbilical vein access and drug/volume delivery, intubation, and chest tube placement. We have determined an optimal bed design and have begun to outfit that with additional design features including but not limited to focused lighting and recording technologies.

Need: A way to improve rates of successful neonatal intubation.

Intubation is both a lifesaving and a life-threatening technical intervention in the neonate. We have developed a new concept in laryngoscope design to facilitate this task. Using 3-D printing, we have modified the blade and handle and have begun preliminary testing of this early prototype on preterm and full-term patient simulators. We plan to use a variety of assessment technologies (4-angle video, including intraoral views; measurement of force and torque; motion-tracking; vision-tracking) to compare this new design to the industry standard.

DISCUSSION

Because this manuscript is written to illustrate the value of the multidisciplinary approach that is fostered by the AHRQ’s PSLL grant mechanism, it focuses on the collective outcomes of 2 separate grants, one that has been completed and one that is in progress, with a goal of outlining a template for conducting this type of research. The PSLL mechanism emphasizes a systems engineering approach to problems in health care. This first requires defining the system of interest and then generating appropriate hypotheses. As an example, for our second PSLL we first defined the extensive system that directly and indirectly influences the care of newborns in the delivery room (Fig. 2).

FIGURE 2.

Neonatal resuscitation in the delivery room: the subsystems comprising the comprehensive system of care.

As described previously, SEIPS is one of the most commonly used models in the science of patient safety. SEIPS considers clinical outcomes (such as safety, efficiency, wellbeing) as arising from a complex and dynamic mix of people, tasks, technologies, working environment and organization—the subsystems comprising the overall system of care. Studying these subsystems, not just in isolation but also as they interact with each other, allows the identification of weaknesses that place patients at risk.³³ Addressing these weaknesses by reconfiguring subsystem function and studying the effects, it is possible to design better systems that will help the humans working within them to deliver optimal patient care.

Another important early step in the process involves assembling the investigative team. A systems engineering approach requires the collective efforts of a broad range of professionals who do not historically interact on a regular basis to identify and solve problems in patient safety and human performance. In our training and research center, we use a small team that possesses certain types of expertise (e.g., clinical, simulation, design of human and system performance research, etc.). This core group is complemented by other professionals with additional expertise who are recruited and funded through grant mechanisms, depending on the specific needs of the type of research being undertaken and the funding sources involved. Depending on their role, different professionals are involved at different stages of the research process. Our teams have included the following professionals:

• Physicians

• Nurses

• Respiratory therapists

• Human factors scientists

• Engineers

• Industrial designers

• Health care simulation operations specialists

• Statisticians

• Patients

• Parents

These team members bring together a broad base of experience that includes expertise in the following areas:

• Clinical medicine

• Patient safety

• Human and system performance

• Simulation-based research

• Simulation operations

• Audiovisual technology

• Quality improvement

• Epidemiology

• Human factors and ergonomics

• Implementation science

• Biostatistics

• Biodesign

• Industrial design

• Systems engineering

• Mechanical engineering

• Biomedical engineering

• Computer science

• Information systems

• Human-computer interaction

• Teaching/learning (multiple platforms)

• Mentoring/coaching

The AHRQ PSLL funding mechanism, by emphasizing a systems engineering approach, not only encourages such diverse teams but also facilitates them.

As stated previously, the general approach advocated by AHRQ includes 5 steps: analysis, design, development, implementation, and evaluation. Initial efforts at analysis are accomplished through focus groups and generation of needs statements, although in reality, analysis continues throughout the investigative process. The information obtained is used to first generate hypotheses regarding how to change subsystem components in an effort to improve safety. We next design prototypes of technologies and modifications of procedures to remediate identified weaknesses in those subsystems. Solutions to threats to subsystem performance are developed and tested in a simulated clinical environment. Once refined in isolation, the procedures and technologies that have been developed are integrated into the overall system of care. Function of the overall system is then evaluated during highly realistic simulated clinical scenarios and, where possible, in the real clinical environment.

Recruiting busy HCPs as human subjects to participate in focus groups and simulated clinical scenarios can be challenging. The team at CAPE has established a reputation among clinical colleagues for examining novel procedures and devices in a highly realistic simulated clinical environment. Subjects know that, when they provide consent for participation in studies conducted at CAPE, they will be challenged, their time used efficiently and modestly compensated, and their input valued and used to enhance the care of patients. It is through the relationships and reputation that have been built over time, coupled with modest financial compensation to cover expenses associated with their participation as human subjects (such as fuel, parking, childcare, etc.), that these busy people can be recruited to participate. Modest financial incentives, in the form of gift cards, are typically built into grant requests.

At appropriate times in this process, provisional patents are applied for promising new technologies. Results are then reported in presentations at meetings across multiple domains and manuscripts submitted for publication in the peer-reviewed literature. This process facilitates not only the exchange of information within academic medicine but also helps to translate the discoveries generated by the multidisciplinary research teams into innovative technologies and procedures used in the care of real patients. It also generates new questions to be investigated, in the same manner that our initial PSLL focused on maternal care has led to multiple questions regarding neonatal care, the focus of our second PSLL.

CONCLUSIONS/LESSONS LEARNED

The multidisciplinary approach to systems engineering in health care supported by AHRQ’s PSSL funding mechanism expands critical thinking about health care safety issues and other clinical problems by encouraging and facilitating the integration of investigators with expertise in clinical medicine, human factors and ergonomics, engineering, design, and other relevant professions. AHRQ’s approach involving analysis, design, development, implementation, and evaluation allows teams to answer important questions in an iterative manner, refining the answers and generating new questions along the way. This truly exploratory approach to solving clinical problems is a unique and valuable resource for all who are willing to look beyond the confines of their own domain to solve the many challenging issues in healthcare today.