Staffing models. Kirk Jensen, Dan Kirkpatrick, and Thom Mayer

Transcription

1 Date:15/4/14 Time:10:51:40 Page Number: 212 Section 3 Chapter 20 Operational principles Staffing models Kirk Jensen, Dan Kirkpatrick, and Thom Mayer Key learning points Define the key variables in staffing an emergency department. Identify the key concepts that define strategies for meeting staffing needs. Consider how to build staffing models based on different priorities. Describe the subtleties of staffing academic emergency departments. Examine case studies illustrating fundamental building blocks of a staffing model and how to optimize flow and resources. Introduction Focusing on staffing in your emergency department (ED) is necessary, and even crucial, in order to optimize what amounts to 75% or more of professional staff expenditures in delivering ED services. In this chapter, we will analyze how best to determine staffing needs, examine how to optimize and leverage your current staff, and, finally, review alternative staffing models. What are your staffing needs? Let us look briefly at the key factors that drive your staffing need, keeping in mind that they are often interrelated. We will also define key terms that will be used throughout this chapter. Patient volume Certain efficiencies are achieved once the ED delivers more than visits annually (as well as , , , etc.) As these volume marks are achieved, EDs find they can better utilize physicians and physician extenders (physician assistants, medical assistants, and nurse practitioners, generally speaking, are healthcare professionals licensed to practice medicine under the supervision of a licensed physician) and can become even more efficient through segmenting different patient flow streams (through acuity or through various triage and intake servers based on volume at different presentation times). We will assume patient volume data come from registration logs. Patient acuity Higher-acuity patients require additional staffing resources for evaluation, management, treatment, and disposition. 1 Patient length of stay (LOS) Longer LOS requires more staffing attention, although not necessarily clinical staff; longer LOS also reduces the available beds. This will reduce capacity to treat higher volumes. It is crucial that we fully understand our LOS and available bed capacity in order to understand whether we have a capacity problem in treating incoming patients. We calculate LOS by measuring the time from registration until ED departure (treat and release LOS) or departure from ED to an inpatient unit (treat and admit LOS). Boarded patients If we are responsible for boarding patients (those awaiting admission to an inpatient unit but who are still located in the ED), our staffing resources will be reallocated in order to monitor these patients. Also, each boarded patient reduces our ability to better utilize bed capacity. We calculate boarding time as the number of minutes beyond 120 after a physician has documented a decision to admit. 212 Emergency Department Leadership and Management, ed. Stephanie Kayden, et al. Published by Cambridge University Press. Cambridge University Press 2014.

2 Date:15/4/14 Time:10:51:41 Page Number: 213 Chapter 20: Staffing models Capabilities of clinicians (physicians and physician extenders) More experienced physicians will be in a better position to treat patients in a more efficient manner. Also, physicians more experienced in dealing with the local medical staff and understanding transfer issues, and more familiar with the current hospital, will be able to more consistently achieve efficiencies in treating patients. When using rotating physicians, or residents, establishing standard operating procedures (SOP) can help reduce variability and improve reliability in practice. Role of non-physician staff How the physician group utilizes physician extenders both for less acute patients and as supplements for more acute patients will have a significant impact on staffing needs. 2 Physician extenders are not an option in many countries. Development of SOPs and collaboration with nursing staff in developing and adopting advanced treatment protocols (e.g., clinical pathways for diagnostics ordering) are two excellent approaches for improving physician effectiveness. Hospital s expectations The hospital s expectations for physician involvement with all patients, door-to-doctor time, and the physician s involvement with all aspects of decision making and communications with attending staff, transfer orders, transfer organizations, etc. can markedly change the staffing requirements. Nursing expectations and nursing skills The nursing culture of involvement by the clinical staff will have a large impact on staffing needs. The scope of nursing care will impact clinical staffing. Many nurses and physicians come from a variety of backgrounds and training experiences, and thus the use of routine evaluation protocols, SOPs, and advanced treatment protocols can create greater consistency for the ED team. The greater the consistency amongst the nurses and physicians, the greater the effectiveness of medical assistants. 3 Into practice: creating your schedule As you better understand your tactical drivers, the next step is to determine how, in practice, staffing decisions actually get made. Most hospitals establish staffing goals through an annual budgeting process. Many times the budgeting process is based on historic patterns and/or algorithms taking into account historic plus blended service hours per patient visit. For the professional staff, changes in staffing patterns should result from careful analysis of patient demand. Patient arrivals You must thoroughly understand your ED s patient arrivals. From a macro perspective, review annual arrivals over the past five years in order to understand trended historic growth and anticipate future growth. Review average daily visit volume for each of the most recent 24 months to determine seasonal fluctuations. Review patient arrivals by hour of the day (HOD) and by day of the week (DOW). While we know that Sundays, Mondays, and the day following a holiday are generally heavier-volume days, you will want to compare average volumes and variation from the average for each day. We recommend identifying heavy (greater than average) and light (less than average) days. Creating different staff schedules for these days is prudent resource use. By reviewing your patient arrival curve by HOD, you can schedule staff to stay ahead of the patient arrival curve. Benchmarking Establish goals for how many patients per hour your physicians will treat by benchmarking externally and internally. The following groups are recommended for external benchmarking: Medical Group Management Association ( Emergency Nurses Association ( The International Conference on emergency Medicine ( org). You should also do your own benchmarking. This may be done by discussing staffing patterns and visiting local colleagues who direct EDs. This can also be expanded outside of your immediate market area to colleagues within the region. As you compare your ED staffing needs, be sure to understand similarities and dissimilarities with hospitals with which you are benchmarking, e.g. admission percentage, LOS, etc. Performance metrics and determinants of staffing needs As we are assessing our staffing needs, some reasonable performance metrics, which will allow us both to 213

3 Date:15/4/14 Time:10:51:41 Page Number: 214 Section 3: Operational principles establish goals and to determine whether we are effectively meeting them, include the following: bed placement to clinician exam test results available to clinician review total ED length of stay percentage of admitted ED patients treat-and-release ED length of stay left without being seen (LWBS) rate overall customer satisfaction with the ED Each of these metrics helps fully understand successful patient flow. Understanding the hospital s goals surrounding patient flow will be crucial in order to create goals for the professional group staffing. In summary, as we look at what is determining our staffing needs, we need to capture historic data and goals or expectations on the following: Patient complexity (we recommend a comparative metric of average RVUs per E/M code for the USA; we recommend admission rate for other EDs). 4 Note: RVU (relative value unit) refers to a methodology of normalizing work done by medical providers; in the United States it has been widely adopted in reimbursement policy. E/M codes (evaluation and management) refer to the current procedural terminology codes that describe physician patient encounters. Customer service (understand your patient satisfaction survey tool and metric relating specifically to clinician performance). Skilled workforce shortages (fully understand your supply of experienced nurses, new nurses, nursing technicians, paramedics, unit secretaries, medical assistants, transportation technicians, etc.). ED crowding or boarders awaiting admission occupying scarce beds. Current practices that may represent high risk factors. The group s preference on using physician extenders independently or only under direct physician supervision. 5 The viability of using scribes. The current medical record and hospital information management system. Determining your staffing strategy Ultimately, each ED should allocate staff by anticipating patient demand, and ensuring that a reasonable asset velocity (patients evaluated per hour) is employed for the clinician treating the arriving patients. Patient arrivals should be broken down by HOD and preferably by Emergency Severity Index (ESI) or some indexing algorithm to understand acuity. Additionally, as you look at allocating staff, you need to take into consideration 24-hour/7-days-a-week coverage and whether you have clinicians who want to work only nights versus those that will rotate nights, evenings, days, weekends, etc. Ease of recruiting and your group s historic staff retention are crucial drivers to understanding your flexibility in creating staffing allocations. Certain EDs are easier to staff than others. Staffing in the heart of a major city with several emergency medicine training programs and plenty of physicians and nurses is vastly different than staffing and scheduling an ED in a rural area with no training programs and fewer amenities. In planning and managing your ED team, you will be more successful if you develop an overall strategy and if you leverage tactics on your staffing and scheduling goals prior to establishing the best staffing numbers for you and your department. Questions you need to ask include: Do you want to staff safely despite a shortage of physicians? Do you need to minimize the number of physicians in your department? Do you want to maximize patient satisfaction? Do you want to have a waiting room full of patients ready for the clinician? Do you want to have a LWBS rate approaching zero? Agree upon an asset velocity (how many patients per hour treated) and then start to back in to duration of shifts and how many hours annually you expect your clinicians to work. Once you have assembled key strategies for these questions, you can begin both generating a schedule and clearly identifying needs for your department. Adding coverage There are several different ways of looking at both assembling core coverage and identifying trigger points for adding extra coverage. The most fundamental way is looking at patients seen per hour. If your asset velocity (number of patients treated per hour) is two patients per hour and you begin routinely exceeding two patients an hour, you should look towards adding 214

4 Date:15/4/14 Time:10:51:41 Page Number: 215 Chapter 20: Staffing models an additional provider. Normally, single coverage (24 physician hours per day) will be able to handle an ED volume of visits annually. You should also budget or forecast monthly volume against monthly clinician hours so that, as volume is growing, you can identify thresholds indicating trigger points for adding staff. Other key drivers that are indicative of the need to consider additional staff include: (1) Turnaround times are becoming progressively higher. (2) LWBS rate is unacceptably high. (3) Your clinicians are concerned that the shifts are too long. (4) Patient satisfaction survey results are unacceptably low. (5) There are frequent concerns or complaints about clinician behavior in a stressful environment. (6) Heavier-volume days (usually Sunday, Monday, Tuesday) become days that both clinicians and nursing staff do not want to work, and key performance metrics are unacceptable on these heavier volume days. It is important to differentiate routine variation in patient volume from trended or progressive increases in volume. While both of these result in additional demand and complexity for the ED clinician, the solutions will be different. In the case of routine variation, it is essential that patient arrivals be analyzed by both HOD and DOW in order to forecast and adequately project additional staffing in anticipation of patient surges. The entire ED clinician team should be in agreement on a surge capacity plan (on call, ability to open additional beds, ability to re-deploy both clinicians and nursing staff to handle a newly created patient stream). Different staffing models offer different solutions for different EDs, and for the same ED at different times in its staff s development and evolution, as well as when transitioning to different models. Patient flow is clearly predictable. Within 85% accuracy, you can predict when patients are coming, what the diagnoses are, and what sort of lab and x-ray needs they are going to have. Again, this is the reason we collect and analyze arrival data, plotting patient arrivals by HOD, DOW, acuity, diagnoses, and tests required. Staffing with other clinicians You should always seek to employ the least expensive resource to accomplish the mission. In the case of an ED, 25 35% of the cases can be adequately and successfully seen independently by physician extenders. In countries where physician extenders are not a viable alternative, SOPs and advanced treatment protocols developed and implemented with nursing participation should be explored. Family practitioners or internists can see 75% or more of the cases that emergency physicians see in the ED. Normally, the use of residents in the ED is only a net gain when you are using senior-level residents (final year). In general, new residents only add complexity and slowness to the EM clinician s day. See below for more discussion on academic training considerations. Single coverage with 12-hour shifts This approach can accommodate anywhere from to visits per year. This will require the ED to treat patients per hour. Groups will try to push their asset velocity up to three patients per hour, allowing single coverage to accommodate up to patients per year. Unfortunately, while many physicians think they can see three patients an hour, rarely can they sustain seeing three patients an hour for an entire 12-hour shift, let alone 12-hour shifts for an entire year. One of the significant challenges in an ED is to deal with the patient arrival curve and the necessary asset velocity during different times of the day. In fact, 64% of the daily ED volume arrives between 10 a.m. and 10 p.m. Consequently, on the typical ED, with annual visits, patients are being processed at 2.63 patients per hour during this peak presentation period. During the remainder of the day (10 p.m. to 10 a.m.), patients are seen at less than two per hour. Most EDs cannot function at 2.63 patients per hour. And the resulting dysfunction includes increased wait time, patient dissatisfaction, poor patient satisfaction, and dissatisfaction amongst the team. Workable strategies to accommodate this increased demand during the 10 a.m. to 10 p.m. shift include productivity-based compensation, templatebased charting, ED efficiency initiatives, scribes or personal productivity assistants, rapid medical evaluation, on-call clinician backup, and transition to eight-hour flex length shifts (shifts that can be two or more hours shorter or longer depending on patient demand). Scheduling methodologies center on an understanding of five key questions: (1) What is the volume coming in? (2) What is the workload? (3) What is the acuity? 215

5 Date:15/4/14 Time:10:51:41 Page Number: 216 Section 3: Operational principles (4) What are your performance measures? (5) What are your priorities for meeting performance measures? Once you have an understanding of these questions, and have developed strategies depending on their answers, you can begin establishing approaches to scheduling. The approaches to scheduling include a review of historical staffing patterns and an understanding of advantages and disadvantages. This allows establishment of the best fit by trial and error. Rule-based computer programming allows efficient generation of draft schedules that follow prescribed rules and allow better fairness as it relates to weekends, nights, holidays, etc. Clearly, clinicians operate more effectively and efficiently when performance and compensation are more closely aligned. The caveat to remember here is that the lowest-cost staffing resource should always be maximized first. Team-based processes should be employed and characterized by front-loading your care, having a physician or medical assistant in the front seeing patients, ordering diagnostic and treatment protocols, and then handing them off to the fast-track or fasttrack-like environment. This approach leverages the sensitivity and specificity of the diagnostic skills of the physician, while also leveraging the expertise of the physician extender in procedural-type services. 2 Case study Demand capacity: determining patient arrivals and aligning physician staffing This study walks the reader through the process of gathering sufficient data to plot a patient arrival curve (demand) and align physician staffing (capacity). Step 1. Gather patient arrival data by hour of the day (HOD) and by day of week (DOW). You should be able to collect these data from the patient registration system; if the data are not available from a registration system, use a spreadsheet to tally one month s patient arrivals by documenting date and time of arrival for each visit to the ED. Calculate averages for each hour of the day and each day of the week. These data points lead you to the next step. Step 2. Plot HOD and DOW patient arrival curves (Figs. 20.1, 20.2) Visit / Hr Av Visits Figure 20.1 Average visits by hour of the day (HOD). Step 3. Initial analysis. The HOD curve reveals the consistent patient arrival pattern replicated by EDs throughout the world. The DOW graph reveals heavy days (visits > average) as Saturday, Sunday, and Monday and lighter days (visits < average) as Tuesday, Wednesday, Thursday, and Friday. This pattern, too, is consistent with most EDs. We should note that our patient arrival ramp up begins at 8 a.m. and arrivals quickly increase to six per hour, which is sustained until 9 p.m. until receding to a level of one visit per hour from 2 a.m. to 7 a.m. In this example, 75% of patient arrivals occur from 10 a.m. to 10 p. m.; this fact alone has huge ramifications for staffing. Step 4. Decide on staffing model. We will assume that our physicians can treat two patients per hour. We will also assume, for the purposes of this example, that we have adequate beds and nursing staff. So, 90 patients/day divided by 2 patients/hour ¼ 45 hours of daily clinical coverage. 216

6 Date:15/4/14 Time:10:51:41 Page Number: 217 Chapter 20: Staffing models Daily Arrival Average Sun Mon Tue Wed Thu Fri Sat Average Variance 7 6 (3) (4) (5) (5) 4 Figure 20.2 Average visits by day of week (DOW) Figure 20.3 Staffing model. Av Visits Staff The model shown in Figure 20.3 was developed by using two 10-hour, two 9-hour, and one 7-hour shifts. Our challenge is the < 2 patient/hour arrivals during the early morning hours, which does not allow our night clinician to achieve our goal. Consequently, we are not able to accommodate the ramp up (10 and 11 a.m.), nor are we able to better staff the evening hours (9 p.m. through 12 a.m). A better variation of this model could be achieved by scheduling five 8-hour shifts and one 5-hour shift. You and your physician group will need to decide on staffing issues which will include: length of shifts, how many patients/hour physicians are expected to treat, whether physicians will have to rotate shifts or wheher you have night coverage that allows others to not have to work nights, and what type of on-call backup system to implement. Backup systems The best physician backup systems are formalized and focused on expediting bed placement to physician exam. A backup system should be resisted unless the hospital provides its members of the backup team to support the ED when the ED is overwhelmed. In other words, having physician on-call when there will not be additional nursing staff to support the additional physician is both inefficient and unsuccessful at accomplishing the goal of heightened treatment services. Backup systems are most effective when they have predefined thresholds, triggers, and next-actions that have been trialed and agreed upon before the crisis ever happens. 217

7 Date:15/4/14 Time:10:51:41 Page Number: 218 Section 3: Operational principles Coordination between clinician staffing and nursing staffing While the clinicians cannot control nurse staffing, there is a fundamental management paradox incorporated herein: you need to know what your staffing levels are, you need to know what their staffing levels are, you need to know what benchmark data they are using, and then you need to know how many nursing shifts are going unfilled. Even though the clinicians do not have control over it, nursing staffing severely impacts what clinicians can do. Emergency physicians may be the scarcest resource in the ED, but they are certainly not the most valuable. That bears repeating: Emergency physicians may be the scarcest resource in the ED, but they are not the most valuable. In many EDs, nurses certainly run the department, and it is nurses who keep things flowing. If nurse staffing levels and experience are not where they need to be, then no amount of physician coverage can compensate for it. According to the 2001 Emergency Nurses Association Benchmark Guide, the average ED patient requires 1.57 hours of direct ED nursing care. 3 Scribes and personal productivity assistants According to one study, scribes can improve patient velocity or patients per hour from 2.2 to On average, a good scribe program will add half a patient per hour. In other words, you need to see two extra patients a shift to pay for a scribe. What will the scribe do for you? The scribe will complete the chart, order x- rays and labs, and keep you on task. They allow more complete charting, they prompt you for elements that will result in optimizing your coding, and they assist you in promptly getting test results, particularly when they relate to multiple patients. Additionally, the scribe can assist in real-time problem solving by being an extender for the physician, can improve coding, and can improve overall asset velocity. The scribe can act as an assistant to perform patient rounding assistance for comfort and follow-up with patients and assist nursing and medical-assistant team members in improving overall patient flow. The following data from a hospital in Virginia certainly defends the case for using scribes % increased charge capture (reduction in downcodes; downcodes occur when record documentation fails to substantiate care rendered). Improved asset velocity (before scribes 1.9; after scribes 2.3). Improved RVU per hour production of 15 20%. Improved lab documentation (before scribes 55%; after scribes 89%). Improved ratio of compliments to complaints per 1000 visits (before scribes 5 : 1; after scribes 9 : 1). Staffing the academic emergency department Providing care to ED patients in the academic setting produces unique challenges, which require unique solutions. Like all areas of academic medicine, EDs combine elements of patient care, teaching, and research, all of which must be effectively combined for the good of the patients and for those who care for those patients. As the poet Seneca said, If you do not know where you are going, no wind is the right wind. With that wisdom in mind, each academic ED should take the time to ensure that those charged with providing care not only understand the sometimes delicate balance between clinical care, education, and research, but also have a clear statement of the vision (why we exist) and the mission (what we are trying to do). Because many non-north American EDs rely on physicians and residents from varying specialties, this constancy of purpose is even more important. Without it, the physicians and residents cannot possibly know what is expected of them, nor can the nurses and support staff have a clear conception of what to expect of the physicians in any of the three areas. The leadership of the academic ED must also have data on the types and numbers of patients who will be cared for, what their acuity is, when they will arrive, and what they will need in order to receive appropriate care. The type, number, and time of arrival of patients seen should not be a surprise to the ED staff. Once those data are known, ED managers should assess the resources that will be required, from both a clinical and a support-staff standpoint, to meet these needs (Fig. 20.3) The educational mission is a related but distinct part of that analysis, since providing education to residents, medical students, and nursing students requires additional time and resources, over and above those needed for effective patient care. Evidence-based medicine (EBM) approaches to emergency patients are increasingly a core part of the provision of effective ED care, but they are even more 218

8 Date:15/4/14 Time:10:51:42 Page Number: 219 Chapter 20: Staffing models essential in EDs with educational missions, for several reasons. First, many of the earliest and most widely adopted EBM approaches arose from critically ill or injured patients, including those with major trauma (advanced trauma life support), myocardial infarctions (advanced cardiac life support and time to open artery protocols), stroke (thrombolytic and interventional therapies), and pediatrics (advanced pediatric life support and pediatric advanced life support). Each of these are examples whereby a collegially developed EBM protocol was married to a widely accepted educational program, which ensures that patients receive consistent care with consistent results, but also rnsures that those who provide that care have the confidence, derived from effective training, to provide the care. Second, rotating residents and students should be provided with training in these EBM approaches, which effectively ensures that both consistent clinical care and education are met proactively. Third, an EBM approach ensures that the nurses and support staff know what to expect when faced with certain clinical situations, which makes both clinical care and education more effective. Fourth, in those settings in which attending physicians are not available in the department during night hours, EBM approaches are even more important, since those providing the care better understand both what is expected of them and how to provide care efficiently and effectively. The downstream effect that such an EBM approach can have on the education of residents, medical students, and nursing students is extremely important. Regardless of the specialty or the setting in which physicians and nurses eventually practice, they will have received an invaluable education in how to evaluate, approach, and treat patients with the most acute illnesses and injuries. Finally, the academic setting is an excellent place in which non-clinical aspects of the art of medicine can and should be taught. Whether they are called service or communication skills, students and residents should be mentored in building the skills and techniques required to address patients psychological and spiritual needs, as well as their clinical ones. As the great physician, musician, and theologian Albert Schweitzer said, One can save one s life as a human being, as well as one s professional existence, if one seizes every opportunity, however unassuming, to act humanly towards another human being. The future of the world depends on it. Perhaps there is no better place to teach these skills than in the ED. Case study Patient flow: flow symptoms and staffing analysis Hospital A is a community-based private organization in northern Virginia, USA. The ED treats patients annually with an admission rate of 12%. Boarding admitted patients in the ED, a left without being seen (LWBS) rate of > 8%, patient satisfaction survey results < 50th percentile, and fast-track treat and release LOS > 170 minutes (Table 20.1) were characteristics precipitating a thorough patient flow analysis and construction of a demand capacity staffing model. Table 20.1 Baseline ED metrics Metrics Trend: before Goal Your ED LWBS 8.5% < 3% Fast-track LOS 171 min < 100 min Admitted LOS 394 min > 252 min Overall physician satisfaction percentile 11th percentile > 75th percentile Monthly boarding hours > 500 hours < 100 hours Patient arrival analysis Data were captured in order to fully analyze patient arrivals by hour of day (HOD) and day of week (DOW) (Fig. 20.4). Additionally, patient arrivals were segmented by severity, or acuity. We identified sufficient daily variation to categorize heavy and light days to drive our staffing allocation. Next we built our staffing model by plotting our patient demand (arrivals-based) against the current clinical staffing. This ED staffs three areas: the main room (24 hours/day 7 days/week), the fast track (19 hours/day 7 days/week), and a pediatric ED (10 hours/day 7 days/week) (Fig. 20.5). 219

9 Date:15/4/14 Time:10:51:42 Page Number: 220 Section 3: Operational principles Sun 163 Mon 75 th Percentile Tues Wed Thurs Fri 154 Sat Figure 20.4 Patients seen per day. Figure 20.5 Demand versus capacity in various units. 220

10 Date:15/4/14 Time:10:51:42 Page Number: 221 Chapter 20: Staffing models Findings The patient arrival and staffing (demand vs. capacity) graphs above highlight the following mismatches: Main missing the patient arrival ramp-up (begins at 10 a.m.) and overstaffing twice later in the day (2 p.m. and 10 p.m.). Fast track understaffing from 10 a.m. to 4 p.m. Pediatrics seems to be a fixed server that can treat approximately 2 patients per hour; as such, it is inelastic to surges and cannot readily adjust to assist either the main or the fast track. The intervention Currently, fast track and pediatrics demonstrate little capacity to flex to accommodate volume surge. Pediatrics treats roughly 20 patients per day and fast track treats roughly 35 patients per day. We reorganized these two services into one (Table 20.2). Lower-acuity patients have testing started in the post-triage area ( results waiting area) and are staged awaiting test results. Patients are not being kept in a bed for purposes other than being evaluated or while having a procedure performed, e.g., suturing. Fast-track beds are the first to be used, and when additional beds are needed, pediatric beds are used. Pediatric patients requiring procedures can be treated in pediatric or fast-track rooms, as appropriate. Table 20.2 Revised staffing model Main ED Current Heavy Light Physician 7 a.m. 5 p.m. 6 a.m. 4 p.m. 6 a.m. 3 p.m. Physician 2 p.m. 12 a.m. 9 a.m. 8 p.m. 9 a.m. 8 p.m. Physician 4 p.m. 2 a.m. 3 p.m. 1 a.m. 3 p.m. 12 a.m. Physician 9 p.m. 7 a.m. 8 p.m. 6 a.m. 8 p.m. 6 a.m. Physician extender 7 a.m. 5 p.m. 6 a.m. 3 p.m. 6 a.m. 3 p.m. Physician extender 9 p.m. 7 a.m. 9 p.m. 6 a.m. 10 p.m. 6 a.m. Fast-track/pediatric Current Revised Physician 2 p.m. 12 a.m. 8 a.m. 6 p.m. Physician extender 8 a.m. 6 p.m. 11 a.m. 10 p.m. Physician extender 4 p.m. 2 a.m. 6 p.m. 2 a.m. Outcomes These recommendations were executed in July We better aligned staffing (capacity) with demand (patient arrivals and clinician treatment requirements), which has resulted in initial reductions of LWBS to < 5%. Treat-and-release LOS in the fast track is < 120 minutes and door-to-doctor time has decreased by 25% (Table 20.3). 221

11 Date:15/4/14 Time:10:51:43 Page Number: 222 Section 3: Operational principles Table 20.3 ED metrics two months after staffing changes were implemented Metrics Trend after Goal LWBS 4.6% < 3% Fast-track LOS 111 min < 100 min Admitted LOS 286 min < 252 min Overall physician satisfaction percentile 51st percentile > 75th percentile Monthly boarding hours < 90 hours < 100 hours References 1. Jenkins PF, Barton LL, McNeill GB. Contrasts in acute medicine: a comparison of the British and Australian systems for managing emergency medical patients. Medical Journal of Australia 2010; 193: Patrick VC, Lazarus J. A study of the workforce in emergency medicine: 2007 research summary. Journal of Emergency Nursing 2010; 36 (6): e Twigg D, Duffield C, Bremner A, Rapley P, Finn J. The impact of the nursing hours per patient day (NHPPD) staffing method on patient outcomes: A retrospective analysis of patient and staffing data. International Journal of Nursing Studies 2010; 48: Albrecht R, Jacoby J, Heller M, Stolzfus J, Melanson S. Do emergency physicians admit more or fewer patients on busy days? Journal of Emergency Medicine 2011; 41: Unterman S, Kessler C, Pitzele HZ. Staffing of the ED by non-emergency medicinetrained personnel: the VA experience. American Journal of Emergency Medicine 2010; 28: Arya R, Salovich DM, Ohman- Strickland P, Merlin MA. Impact of scribes on performance indicators in the emergency department. Academic Emergency Medicine 2010; 17: