CHAPTER 20
How to design a study that everyone will believe: Emergency department operations and systems

Vaishal Tolia

Department of Emergency Medicine, University of California, San Diego Health System, San Diego, CA, USA

Studying operations and flow in an acute care setting

So how do you study the processes of an acute care setting, such as the emergency department (ED)? Well, as has been mentioned previously in other chapters (e.g., Chapter 3), the first step in carrying out research is asking a thoughtful and important question. This continues to hold true even if the focus of research is based on operations and systems of an emergency department. In industry, as well as in medicine, operations-based research helps in solving problems that need timely decisions. Data-driven analytic techniques help manage the ability to make challenging decisions in dynamic environments, and there are, indeed, none more dynamic or complicated than the emergency department.

Medical directors often have to find unique solutions to each problem specific to their institution and geography, as well as keep a very diverse faculty and resident group happy and feeling that they have the tools to effectively and efficiently do their jobs. If that is not hard enough, managing relationships with nursing and hospital administration is another major operational area. All of these problems and innovative solutions require lots and lots of evaluation (in other words, research!), and there are multiple questions that can be asked. The most common problems (er, opportunities for research) that you will encounter in your emergency department include:

Common goals of medical directors and department operation leaders include improved patient care, enhanced patient experience, fewer patients who leave the ED without being seen by a provider (LWBS), decreased length of stay (LOS) for all patients, fewer “diversion hours” where ambulances go to other emergency departments, and meeting state and national benchmarks for excellence [1]. Outside of the interesting research questions involved with all of these, measuring success and failure is important for medical directors, so that they can both appropriately recognize high performers as well as help lesser performers improve areas that are below par. This is important to realize, because unlike data collected purely for research purposes in other types of studies (Chapter 11), data collected for operations research in the emergency department is collected solely to help the department improve. Thus, the type (and even the quality) of the data is based solely on the needs of particular department. This, of course, may make it difficult to generalize a study in this area to other areas of the country (or even in the same hospital!). Other areas of the hospital, such as the intensive care unit, operating room, and so on have their own set of challenges and metrics that require focused research. Many of the same concepts are applicable here as well.

Benchmarking

If every department collected data differently, of course, it would be difficult, if not impossible, to carry out impactful research in this area. However, there are areas of notable agreement. Welch et al. described a consensus on operational metrics in 2010 at the second summit of convened experts in emergency department operations. These experts concluded that in order to improve performance in all areas of ED care and processes, terminology and metrics needed to be standardized, so as to more easily compare between departments [2]. There has been an increased emphasis on standardization due to the regulatory burdens that every emergency department are facing and because outside organizations are analyzing flow in performance in comparing it to national “standards” in order to set goals for accreditation [3, 4].

c20-fig-0001

Figure 20.1 Timeline of ED timestamps and intervals.

Emergency departments (naturally!) try to reduce many of these times to optimize flow. Medical directors often look at these metrics on a monthly basis (if not more frequently) to evaluate trends and focus on areas of improvement. One notable change in ED flow patterns at several institutions that was an outcome of this data set was the creation of “pods.” These are care-specific areas in the ED with the staff divided into smaller teams with improved communication and reduction in commonly cited delays. Several institutions (including our own) have shown a reduction in many of these benchmarked metrics, such as overall length of stay for discharge patients.

This benchmarking allows easy comparison between emergency departments and allows any ED to target particular areas where performance is lacking. Obtaining this data is also useful for operation leaders and medical directors to create improvement, since systems and processes impact these metric measurements. In many situations, internal funding, external funding, and reimbursement are directly tied to performance. Thus, having a standardized set of data to present to both hospital administration and other regulatory bodies is essential to optimal functioning and support. Again, all of this data means that you can ask questions about how a particular factor impacts the day-to-day operations of your ED.

Informatics

If a question about benchmarking does not interest you, there are other interesting questions about how all of this benchmarking data are stored and utilized. The field of informatics is concerned with exactly that. Informatics in medicine generally is an important topic, since not only is the meaningful use of electronic health records a major issue, but the ability to perform high quality research is based upon the ease of access to accurate and interpretable patient data. In addition, mobile technologies, web-based technologies, and simulation are important tools. Most emergency department charting systems are being replaced by electronic health records, and most modern physicians rely on electronic sources of information instead of upon printed textbooks. In other industries, such as the airlines, simulation-based education and training have been mainstays, but are now becoming ever more popular in emergency medicine.

The conversion from print to electronic methods of storing and delivering information raises a whole host of questions for researchers in this area: How and what information should be stored? What is the optimum way to provide needed information so that clinicians do not have to “hunt” through various parts of the chart? How can patient information be utilized in order to make a simulation more realistic? From a regulatory standpoint, how can patient privacy and maintenance of records be protected, yet easily available to providers who need it? Informatics research has spawned fellowships and many research opportunities that involve a wide variety of topics, some of which include: ease of health record access, meaningful use of the EMR and data it produces, and identification of at risk situations in patient care (among many others).

Improving operations-based research

As noted previously, studies in this area often cannot be generalized to other departments (i.e., lack external validity). Still, you have probably had the experience (like many people) of returning from a great conference where you discussed innovative solutions with amazing physicians and administrators from other departments. If it works well somewhere else, why can you not replicate it immediately at your home institution with similar or better results?

Wise administrators know that the most important questions in a busy department do not necessarily involve the specific intervention of interest, but rather how the intervention is going to affect the current environment and processes already in place. For example, one of our emergency departments suffers from a common problem, that of holding inpatient admissions for an extended period (often days), which may pull valuable resources such as nursing away from other patients. This is a unique challenge to this particular clinical site and comes with it its own set of solutions. This clinical site is also the only one that has a dedicated emergency department base observation unit, which has had a positive impact on this boarding problem by creating alterative care pathways that are more efficient. Focusing on solutions often involves the medical director further highlighting the problem to hospital leadership, so that specific areas (operating room scheduling, inpatient discharge, nurse staffing, etc.) can be addressed, as they all impact (and can help disimpact!) the flow. Our other clinical site, on the other hand, is our level I trauma centre and primary teaching facility and we deal with the challenges of a very vulnerable population as well as a high proportion of psychiatric patient’s that require significant time and resources.

One of the process improvement measures that we employed was to send patients who are stable directly to the CT scan and have them unloaded onto the scanner directly from the paramedic gurney. We had a large volume of data prior to this singular intervention, which we employed at both of our clinical sites, and we were able to show a significant reduction in time to lytic therapy, which improved the care for the patient as well as our measured metrics related to the care of the acute stroke patient. Having a strong study design is fundamental to provide both internal and external validity, particularly when the impact of the study may require resources, finances, and oversight from those outside of the department of emergency medicine. Having a dedicated research network and staff that can focus on operational issues is important for not only study design but also implementation of findings as well as multicenter application of the study and ultimately the expansion of better process methods. Randomization is also easier with multiple sites, as department-specific measured confounding variables are less likely to have an impact. Though it goes without saying, patient-based research and outcome measures are generally the focus of emergency department operations studies. Developing partnerships with funding agencies, specialty colleges in emergency medicine, and industry, especially with a focus on improving healthcare delivery efficiency and cost, is an important first step to dealing with emergency department-based operational challenges.

Asplin and Yealy, in their February 2011 article in Annals of Emergency Medicine, describe other key requirements for operations-based research in the emergency department [5]. The following are necessary to conduct operations-based research: proper infrastructure; ability to perform multicenter analysis; study design focusing on patient centered outcomes, especially those that go along with benchmarked metrics (mentioned previously); and projects that have the opportunity for impact and propagation. In addition, the institution has to have a focus and an interest in operation-based research, as well as the ability to act upon findings that improve health care delivery and patient outcomes.

Conclusion

Although many physicians may not like the fact that data about their performance with patients is being collected and compared in multiple ways, collecting and analyzing data about the processes and flows of a typical emergency department are essential for efficiency. Research on this data, or operations-based research, allows for the comparison of this data across different systems. Like it or not, modern emergency departments operate in the current era of benchmarking data, evolving informatics, increasing regulatory burdens, and demands from society to become more patient-oriented. Medical directors will continue to need high quality data to inform these decisions. The area of operations-based research has tremendous need for skilled researchers, especially ones who can use rigorous methodology and high-quality study designs.

References

  1. 1 Rathlev, N.K., Chessare, J., Olshaker, J., et al. (2007) Time series analysis of variables associated with daily mean emergency department length of stay. Ann Emerg Med, 49:265–271.
  2. 2 Welch, S.J., Stone-Griffith, S., Asplin, B., et al. (2011) Emergency department operations dictionary: results of the second performance measures and benchmarking summit. Acad Emerg Med, 18:539–544.
  3. 3 The Joint Commission (2005) A Comprehensive Review of Development and Testing for National Implementation of Hospital Core Measures. http://www.jointcommission.org/assets/1/18/A_Comprehensive_Review_of_Development_for_Core_Measures.pdf (last accessed 10 May 2015).
  4. 4 Glickman, S.W., Schulman, K.A., Peterson, E.D., et al. (2008) Evidence-based perspectives on pay for performance and quality of patient care and outcomes in emergency medicine. Ann Emerg Med, 51:622–631.
  5. 5 Asplin, B.R. and Yealy, D.M. (2011) Key requirements for a new era of emergency department operations research. Ann Emerg Med, 57(2):101–103.