Emergency Department: Investigating the hidden
effect of treatment rooms on increased patient
waiting times
John Bos
University of Groningen
Faculty of Economics and Business
Master thesis: Supply Chain Management
February 28, 2014
Supervisors:
Dr. J.T. van der Vaart
2
Emergency Department: Investigating the hidden
effect of treatment rooms on increased patient
waiting times
Abstract
The aim of this research is to improve the throughput times of patients visiting the emergency department. This thesis aims to make recommendations to design the emergency department in such a way that the waiting times of patients waiting inside treatments rooms are minimized. The results of the research show that changes in the room allocation policy could potentially reduce the time patients spent waiting inside a treatment room. More specifically, assigning patients to treatment rooms in a controlled way with the use of triage codes as a guideline when there are treatment rooms available reduces the utilization of the treatment rooms, which improves the overall quality of the emergency department.
Keywords: emergency department, treatment rooms, patient waiting times, emergency
3
Table of contents
Abstract ... 2
1. Introduction ... 4
2. Theoretical background ... 5
2.1 Emergency department overcrowding ... 6
2.2 Emergency department triage... 7
2.3 Emergency department capacity ... 8
2.4 Conceptual model and research question ... 9
3. Methodology ... 10 3.1 Case description ... 10 3.2 Research method ... 11 3.3 Data collection ... 11 4. Results ... 13 4.1 Introduction ... 13
4.2 Room allocation policy ... 13
4.2.1 New patient arrivals ... 13
4.2.2 Patient arrivals from other departments ... 13
4.2.3 Patient arrivals via ambulance ... 13
4.2.4 Time until triage ... 14
4.3 Dataset analysis ... 15
4.3.1 General findings ... 15
4.3.2 Arrival and departure pattern ... 16
4.3.3 Throughput diagram analysis ... 18
4.4 Questionnaire ... 20
5. Conclusions ... 22
6. Recommendations ... 23
6.1 Recommendations for the emergency department of the MCL ... 23
6.2 Limitations ... 23
6.3 Suggestions for further research ... 23
References ... 24
Appendix A – Output of frequencies analysis time until triage... 27
Appendix B – Table arrival and departure pattern ... 28
Appendix C – Throughput diagrams days with long throughput times ... 29
4
1. Introduction
Nowadays, in many western health services, the capacity of the emergency departments (EDs) is one of the most pressing concerns for both policymakers and health department managers (Fitzgerald, Sloan, Simoff, Samaranayake, & Johnston, 2008). A high variability in the arrival process of patients in emergency care units of hospitals can lead to massive waiting times. The capacity of EDs can be separated to staff capacity and room capacity. Staff capacity refers to the nursing- and medical staff working within the ED and room capacity refers to the standard and specialized treatment rooms within an ED. Not only should the physical capacity of EDs be sufficient to cope with the patient demand, it should also be designed in a way that excess waiting times within treatment rooms are avoided and overcapacity is minimized. The main goal of this thesis is to make recommendations to design the emergency department in such a way that the waiting times of patients within treatment rooms are minimized. The research is performed at the Medical Center Leeuwarden (MCL), a large teaching hospital in the Northern part of the Netherlands.
5
From a scientific point of view, it would be interesting to examine how the minimization of patient waiting times within treatment rooms affects the overall quality of the ED. It could be, for example, that by minimizing the waiting times within treatment rooms, the ED can become more responsive to fluctuations in patient demand. Moving the waiting time from the treatment rooms to the waiting area would imply that the utilization of the treatment rooms decreases. This decrease in room utilization implies that the ED could react better to sudden increases in patient demand. This thesis aims to expand the current body of literature on the subject of improving throughput times in emergency departments.
From a practical perspective, insights on the impact of the minimization of patient waiting times within treatment rooms on the quality of the ED could be very beneficial. Health department managers, for example, may reconsider the budget that is spent on improving the quality of the ED, in case the quality of the ED is improved when patient waiting times within treatment rooms are minimized. Another benefit is that, should the minimization of waiting times within treatment rooms lead to an improvement in patient throughput times, ED overcrowding will occur less often. Also, the ED will be able to handle more patient admissions, which will improve the overall quality of the ED.
This thesis aims to provide an answer to the following research question: “How can the emergency department be designed in such a way that waiting times within treatment rooms are minimized?”. In order to give an answer to this research question, a case study is carried out at the Medical Center Leeuwarden (MCL) .
The next chapter will discuss the theoretical background, the hypotheses and the conceptual model. Chapter three elaborates on the applied methodology. Chapter four will discuss the results and chapter five discusses the conclusions based on the research. Finally, chapter six will state recommendations for the specific ED investigated in this case study, along with limitations of the research and suggestions for further research.
2. Theoretical background
6
the overall throughput times. After this section, the chapter continues with a review on the issues regarding emergency department capacity. The chapter ends with a conceptual model.
2.1 Emergency department overcrowding
Emergency departments have to be available to receive patients in urgent need of treatment at any time (Van der Vaart, Vastag, & Wijngaard, 2011). Patients can arrive at the ED at their own initiative, through a referral of their general practitioner, by ambulance or helicopter, or by a transfer from another hospital or hospital department. The timing and amount of patients arriving at the ED is very uncertain, as well as the urgency of the needed care. The increased volume in ED visits over the past years is one of the main causes of overcrowding (Asaro et al., 2007; Derlet, 2002; Solberg, Asplin, Weinick, & Magid, 2003). Overcrowding has become increasingly widespread and continues to be a major factor affecting EDs worldwide (Paoloni & Fowler, 2008).
Emergency department overcrowding is a subject that is well grounded in the current body of literature. Some of its causes are beyond the scope of individual hospital control, such as increased complexity and acuity of patients presenting to the ED, managed care problems, and increases in elderly patients, often requiring a higher level of care (Derlet, 2002; Liu, Hobgood, & Brice, 2003). Also, the effect of non-urgent visits or frequent flyers (patients who have used the ED three or more times in six months (Michelen, Martinez, Lee, & Wheeler, 2006)), on ED overcrowding has been studied in literature (Hoot & Aronsky, 2008), and a clear consensus of this relationship is missing, as one analysis concluded that visits by patients with non-urgent complaints were not associated with the most severe crowding at large hospitals (Sprivulis, Grainger, & Nagree, 2005).
7
departure time, exceeding eight hours (Forero et al., 2004). Access block occurs when a patient remains in an ED for more than eight hours (after the patient is first seen by a doctor) consequent to the limited availability of hospital beds (Paoloni & Fowler, 2008). Access block has several negative consequences. It increases the risk to the patients’ health (Thompson, 1999) and it affects the allocation of limited public resources and the management of healthcare services (Fitzgerald, Dadich, & Sloan, 2010). Access block is associated with decreased efficiency in the ED (Derlet, 2002) and increased inpatient stays (Richardson, 2002). Access block is likely to hinder patient flow throughout a hospital (Fitzgerald et al., 2010). Also, access block has been demonstrated to decrease the triage benchmark performance of non-urgent and lower-acuity patients (Paoloni & Fowler, 2008).
Because of the highly stochastic nature of emergency care, rules were developed for situations where demand exceeds available resources (Van der Vaart et al., 2011). The need for these decision rules or ‘triage’ is enhanced by the growing imbalance between needs and resources resulting from the twin challenges of access block and growing demand (FitzGerald, Jelinek, Scott, & Gerdtz, 2010). The section below discusses the Manchester Triage System (MTS), a triage system that has been implemented in several European countries (Roukema et al., 2006).
2.2 Emergency department triage
Triage is an essential element of modern medical care as it is necessary to assign relatively scarce resources to unlimited medical needs. Such assignments become necessary where there is a mismatch in quantum, time or location between the medical needs of patients and available resources (FitzGerald et al., 2010). In the emergency department, triage refers to the methods used to assess patients’ severity of injury or illness within a short time after their arrival, assign priorities, and transfer each patient to the appropriate place for treatment (Van Veen et al., 2008).
8
flowcharts and consists of 52 flowchart diagrams (e.g. head injury or abdominal pain) that are specific for the patient’s presenting problem (Christ et al., 2010). The patient’s principal presenting complaints are allocated to one of these diagrams. The triage nurse selects the most suitable flow chart for each presenting problem and uses general and specific discriminators (life threat, pain, hemorrhage, acuteness of onset, level of consciousness, and temperature) to identify the patient’s acuity (Van Veen et al., 2008). Every patient is classified in one of the five urgency categories: red (needs to see a doctor immediately), orange (can wait ten minutes), yellow (can wait one hour), green (can wait two hours), and blue (can wait four hours) (Van der Wulp, Van Baar, & Schrijvers, 2008). If the nurse does not agree with the assigned urgency category, the system can be overruled (Van Veen et al., 2008).
There are several articles that evaluate the reliability and the validity of the MTS. Van der Wulp et al. (2008) assessed the reliability and validity of the MTS in a general ED in the Netherlands. They found a inter-rater reliability of “moderate” to “substantial” and a high test-retest reliability, meaning that the MTS has a high reliability. The validity was assessed by calculating sensitivity, specificity, and the percentages of overtriage (patient vignettes receive a more urgent triage category rating from the triage nurse than from the expert) and undertriage (the opposite of overtriage). The authors concluded that undertriage mainly occurs in the triage categories orange and yellow, and that the MTS is more sensitive for children requiring immediate or urgent care than for other patients in the ED.
The introduction of structured triage by specially trained nursing staff in the ED helps to accurately identify patients whose lives are endangered, especially at times of insufficient treatment capacity (Christ et al., 2010). An important determinant of patient flows in EDs is the capacity of the EDs. The next section will discuss two types of capacity; staff capacity and room capacity.
2.3 Emergency department capacity
9
overcrowding. Staff capacity refers to the nursing- and medical staff working within the ED and room capacity refers to the standard and specialized treatment rooms within an ED.
The challenge that many EDs nowadays face is to ensure sufficient capacity to cope with the patient demand, in particular during peak periods when there is a sudden increase in demand from new patients, and during the busiest time of the week. The exact arrival times of new patients are to a large extend unpredictable, but the daily pattern of ED visits can be predicted, based on historic patient-arrival times (Asplin, Flottemesch, & Gordon, 2006). Also, the busiest time of the week is predictable, as historical data shows that Monday has the highest peak in patient demand (Asplin et al., 2006). The capacity of EDs should not only be sufficient to cope with patient demand, there is always a fundamental trade-off between the efficiency of a process as measured by its utilization and its responsiveness as measured by its waiting time (Terwiesch et al., 2011).
2.4 Conceptual model and research question
The conceptual model shown in Figure 1 depicts the variables and the relations between these variables relevant for this research.
Figure 1 - Conceptual model
10
of patients that are inside treatment rooms are also a determent of how long patients spent waiting inside treatment rooms, as a crowded ED requires the ED personnel to divide their attention among multiple patients, which decreases their availability for individual patients.
3. Methodology
In order to provide an answer to the research question, a single case study will be conducted. Case research has consistently been one of the most powerful research methods in operations management, particularly in the development of new theory (Karlsson, 2008). This chapter will discuss the case description, the research method, and the collection of data.
3.1 Case description
This research is performed at the emergency department of the Medical Center Leeuwarden (MCL), a large teaching hospital in the province of Friesland, which is located in the Northern part of the Netherlands. In 2012, the MCL had a capacity of 623 beds and employed 204 medical specialists. The hospital had over 153.000 outpatient visits and handled about 29.500 admissions. Since 2004, the ED applies the Manchester Triage System (MTS) (Van der Vaart et al., 2011), as well as many other EDs across Europe. Figure 2 shows the layout of the ED of the MCL.
11
When patients arrive at the reception desk, they are registered in the computer system. This registration process takes a couple of minutes. Patients requiring immediate care are directly allocated to an emergency room and their data are entered at a later point in time. After arriving at the reception, patients proceed to the waiting area where they wait for the triage process. During the triage, a nurse applies the MTS to assess the urgency level of the patient and assigns them with one of the five triage colors. Depending on the assigned triage color and the occupancy rate of the ED, the patient is either sent back to the waiting room or transferred to an emergency room for treatment. After the treatment, the patient will be either sent home because further treatment is not required, or the patient will be transferred to another department in the hospital for further treatment.
3.2 Research method
The first step is to identify and analyze the way that the patients are currently allocated to the treatment rooms and nurses of the emergency department. This current room allocation policy will be mapped by observing the work floor of the ED. The implications of the room allocation policy and its performance will be discussed. Next, quantitative data will be analyzed in order to further elaborate on the implications of the current room allocation policy. The average daily arrival and departure pattern of patients will be made and discussed. Furthermore, throughput diagrams will be made, highlighting the performance of the room allocation policy. Finally, a questionnaire is distributed among the nurses of the ED, in order to map their opinions about the occurrence of patients that are still waiting for treatment, once they are placed in a treatment room. The questionnaire tries to find solutions of this phenomenon, by asking questions about the way nurses prioritize the order in which they treat patients, the allocation of patients to nurses by the care coordinator, the workload of the ED nurses, and the coordination within the ED.
3.3 Data collection
In order to execute this research, both qualitative and quantitative data is required.
12
treatment, the department of the destination, the diagnosis, the starting data, the time of arrival, the date and time the triage was executed, the date and time of the first encounter with the ED nurse, the date and time of the first encounter with the specialist, the date and time the treatment is finished, the date and time of departure, the name of the nurse involved, the name of the specialist involved, the date and time of the start and finish of the lab requests. Most data was present. However, a lot of data considering the date and time of the first encounter with the specialist and the lab data were mostly missing.
13
4. Results
4.1 Introduction
This chapter presents the results of the case study. The first section of this chapter elaborates on the way the patients are currently allocated to the treatment rooms. The second section presents the dataset analysis, consisting of some general findings, the average daily arrival and departure pattern, and a throughput diagram analysis. The third section presents the results of a questionnaire that was distributed along the nurses of the MCL, in order to gain additional insights in the findings presented in the second section.
4.2 Room allocation policy
This section presents the current room allocation policy of the emergency department of the Medical Center Leeuwarden. This current policy is mapped by observing the work floor of the ED.
4.2.1 New patient arrivals
The allocation of patients to treatment rooms is an important role of the care coordinator. Patients that arrive at the ED have to undergo the triage process before they are treated, as described in section 3.1. After this process, the care coordinator assigns the patient to a treatment room, depending on the acuity of the complaint. Patients that have received an acute triage code are assigned by the care coordinator to a treatment room that lies close to the central post of the ED (see Figure 2) and less-acute patients are assigned to a treatment room that lies further away from the central post. The more acute patients are assigned to treatment rooms close to the central post because these patients usually require intensive care and the care coordinator likes to have a clear view over these patients.
4.2.2 Patient arrivals from other departments
Patients that are referred to the ED by other departments of the hospital are directly transferred to a treatment room, depending on the acuity of the complaint and the number of patients in the waiting area. Acute patients that arrive from another department are directly placed in a treatment room, meaning that the triage process is skipped. The triage code of these patients is often registered at a later point in time. When the ED is crowded, it occurs that stable, less-acute patients that are being treated are temporarily moved from the treatment room to the waiting area, in order to treat the newly arrived acute patient.
4.2.3 Patient arrivals via ambulance
14
waiting in the waiting area. If all treatment rooms are occupied upon arrival, the patient temporarily waits in the hallway until a treatment room becomes available. Patients that arrive at the ED via ambulance skip the triage process. Their triage code is often registered at a later point in time, just like the patient arrivals via other departments. The exact arrival time of patients arriving by ambulance is unknown. Therefore, when there are patient expected to arrive via ambulance, one or two treatment rooms are reserved for ambulance arrivals when the ED is crowded.
4.2.4 Time until triage
The nurse that performs the triage on the new patients also starts up treatments with another nurse when there are no new patients in the waiting area. This implies that during the time the triage nurse is helping with the treatment of a patient, he or she is unavailable to perform triage on new patients. The standard is that the triage should be executed within ten minutes after a patient arrives at the ED, on quiet as well as busy periods. The dataset of a fourteen-month period, from October 2012 until November 2013, was analyzed to gain insights in the time it takes before the triage process takes place. The patient arrivals via ambulance, helicopter, and other departments were filtered out from the analysis, as well as missing values. As a result, 13.696 out of the 26.114 available data points were used in this analysis. The results of the analysis are presented below.
Time until triage after patient arrival Cumulative percentage
0:00:00 24,0%
0:05:00 55,1%
0:10:00 74,3%
0:15:00 83,5%
0:20:00 89,3%
Table 1 - Time until triage after patient arrival
15
4.3 Dataset analysis
This section presents the analysis of the dataset. First, some general findings are presented. After that, the average daily arrival and departure pattern of patients visiting the ED is analyzed. Finally, a throughput diagram analysis is presented.
4.3.1 General findings
As mentioned above, the dataset consist of a fourteen-month period, from October 1, 2012 until November 30, 2013. During this period, a total of 26.114 patients have visited the emergency department. Figure 3 shows the age distribution of the patients.
Figure 3 - Age distribution of the patients
As can be seen in this figure, there is a peak in the amount of patients around the age of 20 and the age of 65. The age of the patients varies between 0 and 113 ages and the average age of the patients is 51,5 years. The most frequently visited specialisms by the patients are surgery (35,8% of the total patients), internal medicine (16,6%), cardiological (15,5%), lungs (7,4%), and neurological (7,1%). The average throughput time (the time between the arrival and departure time of the patient) is two hours, four minutes, and forty-seven seconds. Table 2 shows the amount of patients that visit the ED per day.
16
Day Amount of patients Average per day
Monday 4.122 68 Tuesday 3.772 62 Wednesday 3.683 60 Thursday 3.871 63 Friday 4.181 69 Saturday 3.326 55 Sunday 3.159 53
Table 2 - Amount of patients per day
This table shows that Friday is the most busy day of the week, as this day has the most patient visits. This finding contradicts the findings of Asplin et al. (2006), stating that Monday has the highest peak in patient demand, although the difference in patient visits between these days is only one patient on average. Ekelund et al. (2011) studied six EDs and they found that Monday was the busiest day and Saturday the least busy day. Table 2 shows that Sunday is the least busy day in this ED with 53 patients on average, although the differences in patient visits between Saturday and Sunday is again minimal, with two patients on average. Overall, the average amount of patient visits per day is 61,3 patients.
4.3.2 Arrival and departure pattern
17 Figure 4 - Average daily arrival and departure pattern
As can be seen from this figure, the amount patient arrivals starts to increase around 8:00 hours and reaches its peak around noon until around 17:00 hours, with over 4,5 patients arriving at the ED per hour. The amount of patient departures starts to increase around 9:00 hours. Around noon, there is a nod in the departure curve. After this nod, the departure curve is increasing at a slower pace than it was before, which points to the fact that the ED has trouble handling the increase in patient arrivals earlier in the day. The horizontal distance between the arrival and departure curve increases during the day, which means that the average throughput time is increased during this period.
It is interesting to note the sudden drop in the patient departure curve around 16:00. Between 14:31 and 15:30 hours, the average departure rate is 5,2 patients. This amount drops to 4,5 patients between 15:31 and 16:30 hours, and increases to 5,2 patients again between 16:31 and 17:30 hours. More specifically, from 15:31 until 16:00, the average departure rate is 2,0 patients and 2,5 patients in the period from 16:01 until 16:30. Table 3 displays the schedules of the nurses in the ED.
0,0 1,0 2,0 3,0 4,0 5,0 6,0 A m o u n t o f p atien ts Time
Average daily arrival and departure pattern
18
Type of shift Time Amount of nurses
A shift 7:30 – 16:00 3
C shift 9:30 – 18:00 1
D shift 15:30 – 23:30 4
Night shift 23:15 – 7:45 2
F shift 13:00 – 21:00 1
Table 3 - Schedules of the nurses (Table adapted from Nanninga (2005))
This table shows that, between 15:30 and 16:00 hours, there is half-an hour overlap between the A and D shifts. This overlap is planned to discuss and hand over patients (Van der Vaart et al., 2011). Observing the work floor shows that around 15:30 hours, the nurses of the A shift come together to discuss and evaluate the day. This planned evaluation takes about 10 minutes. During this evaluation the nurses of the A shift are not available to discuss the patients with the nurses that just have started the D shift. This could explain the sudden drop in the patient departure curve in Figure 4.
4.3.3 Throughput diagram analysis
Previous research executed at the MCL showed that on 74% of the days of a twelve-month period (November, 2011 until October, 2012), there is a moment in time where all ED rooms are occupied, and on around 10% of the days the treatment rooms are operating at maximum capacity for two hours, implying that the physical capacity of the ED is a bottleneck during certain periods of time (Verbree, 2013). This section analyses the throughput times of a few days with a long average throughput time, in order to investigate whether the ‘hiding’ effect of treatment rooms, meaning that patients are actually waiting inside treatment rooms instead of being treated, occurs. Days were chosen where the age of the patients and number of patient visits did not deviate strongly from the average (51,5 years and 61,3 patients) and where the required data is as complete as possible. Table 4 shows the selected days.
Date Average age Number of patients Average throughput time
February 1, 2013 57,7 71 2:39:47
March 14, 2013 51,3 69 2:35:49
November 21, 2013 53,9 60 2:33:54
December 30, 2012 56,0 51 2:32:44
Table 4 - Days selected for the throughput diagram analysis
19
between the cumulative input and cumulative output curve depicts the cycle time, in this case the throughput time of the patients. The vertical distance between the cumulative input and cumulative output curve depicts the work in process, in this case the patients that are being treated. Figure 5 shows the throughput diagram of February 1, 2013. The throughput times of the other selected days shown in Table 4 can be found in Appendix C.
Figure 5 - Throughput diagram February 1, 2013
The blue line represents the cumulative arrival times of the patients. The red line depicts the cumulative amount of triages performed throughout the day. The green line depicts the moments in time where a new patient is under treatment, meaning that from that moment on, the patient is inside a treatment room. The purple line represents the cumulative departure times. Figure 5 shows that after 9:00 hours the throughput times are increasing, as the horizontal distance between the arrival and the departure line is increasing. Also, the work in process increases sharply around 11:00 hours, as the vertical distance between the arrival and departure line increases.
There is a sharp increase in new patient arrivals around 15:00 hours. From that point on until 19:00 hours, there is a notable horizontal distance between the arrival and the start treatment line, implying that these patients are waiting inside the waiting area. Before 15:00 hours, the amount of time the patients spent waiting in the waiting area are minimal, as the horizontal
20
distance between the arrival and start treatment line is relatively small. This highlights the current room allocation policy of the ED. Patients are allocated to a treatment room if this room is unoccupied or not reserved for ambulance arrivals (refer to the detailed analysis of the current room allocation policy to section 4.2). This implies that the utilization of the treatment rooms is not considered when new patients are assigned to treatment rooms, and that the triage codes assigned to the patients are only effectively used when the physical capacity of the ED is insufficient to treat a new patient.
Around 12:30 hours, the amount of patients that is inside a treatment room is thirteen (the vertical distance between the start treatment curve and the departure curve). At this point in time, there are four nurses working at the ED (see Table 3 for the schedules of the nurses). This implies that each nurse is assigned to at least three patients (assuming the care coordinator has taken the workload of the nurses into account when assigning new patients). It is not unthinkable that one of these patients has to wait for treatment while the assigned nurse is treating the other patients. In order to investigate the occurrence of treatment rooms acting as a ‘second’ waiting room, a questionnaire is distributed among the nurses of the ED.
4.4 Questionnaire
The goal of the questionnaire is to map the opinions of the nurses of the ED about the occurrence of patients that are still waiting for treatment once they are placed in a treatment room. The remainder of the questionnaire focuses on solutions of this phenomenon by asking questions about the way nurses prioritize the order in which they treat patients, the allocation of patients to nurses by the care coordinator, the workload of the ED nurses, and the coordination within the ED. The full questionnaire can be found in Appendix D.
Four out of the seven respondents agreed with the following statement: “It sometimes occurs that patients are placed in a treatment room, but that it takes a while before the actual treatment of these patients starts.”. This finding confirms the presumption of section 4.3.3, stating that treatment rooms might act as a ‘second’ waiting room where patients are actually waiting for treatment instead of being treated.
21
One nurse reported that once a patient enters a treatment room, nurses often unconsciously “re-triage” the patient and that the result of that triage determines the order of patient treatment. The assigned triage codes are based on a snapshot in time, which could be different once the patient enters a treatment room.
The respondents were asked to respond to the following statement: “In my opinion, the care coordinator takes the workload of the nurses into consideration when he or she assigns patients to the nurses.” Six out of the seven nurses agreed with this statement, implying that this statement is correct.
The nurses were also asked to give their opinion about the following statement: “In my opinion, the number of nurses per shift is sufficient at all times to meet the patient flow.” Five out of the seven nurses disagreed with this statement. One nurse reported that the ED sees the most patients between 11:00 and 21:00 hours, and that there are not enough nurses present to meet the patient demand during this period. Another nurse likes to see an extra nurse present at the ED during the period between 12:00 and 21:00 hours. Another ED nurse gives roughly the same explanation, stating that there are not enough nurses between 11:00 and 21:00 hours. Furthermore, four nurses agreed and three nurses disagreed with the statement “I work harder when there are many patients present in the ED.”, implying that the perceived workload affects some ED nurses. Also, two nurses disagreed with the following statement: “When I have to treat four patients at the same time, I do my job just as well as when I have to treat only one patient.” This implies that the number of patients that are being treated in parallel has a negative influence on the performance of the treatment by some ED nurses.
22
5. Conclusions
The main focus of this research is to make recommendations to design the emergency department in such a way that the waiting times of patients within treatment rooms are minimized. In this chapter, the most important findings are presented and discussed in order to answer the research question.
“How can the emergency department be designed in such a way that waiting times within treatment rooms are minimized?”
When evaluating the current room allocation policy and analyzing the throughput diagrams, it quickly became clear that the waiting area of the ED is only used for its purpose when the treatment rooms are occupied or partially reserved. The utilization of the treatment rooms is not considered when new patients are assigned to treatment rooms, and the triage codes assigned to the patients, which determine how long a patient can wait before he or she needs to be seen by a doctor, are only effectively used when the physical capacity of the ED is insufficient to treat a new patient.
23
6. Recommendations
The last chapter contains several recommendations for the emergency department of the MCL to improve the throughput times and the overall quality of the ED. Furthermore, this chapter pays attention to the limitations of the research and points out suggestions for further research.
6.1 Recommendations for the emergency department of the MCL
Along with the change in the room allocation policy mentioned in the previous chapter, this section contains some additional recommendations. The findings suggest that the triage nurse should solely focus on the task of executing the triage on new patients, as more than a quarter of the patients were triaged too late. Additionally, the triage nurse can start the blood tests in the triage room, as it can take up to an hour before the results from the laboratory have arrived. Currently, the blood test is only started in the triage room when the treatment rooms are occupied.
As explained in section 4.3.2, the sudden drop in the output curve between 15:31 and 16:00 hours could be caused by the planned discussion and evaluation of the day by the A shift nurses, at 15:30 hours. An idea is to move this meeting to 15:50, implying that the nurses that start the D shift at 15:30 have more time to discuss patients with the A shift nurses.
Another recommendation is to schedule in an additional nurse. When the triage nurse solely focuses on executing the triage and the blood tests, he or she is unavailable to start up new treatments. Additionally, the majority of the respondents of the questionnaire say that the number of nurses is insufficient to meet the patient flow at all times. They argue that they are understaffed during peak hours.
6.2 Limitations
A big limitation during this research was the received dataset. More specifically, there was barely any data available about what happens during the process of treatment. The start time of the treatment and the departure time of the patients were mostly correct. But the steps between the moment the treatment starts and ends were mostly missing. Therefore, the amount of time patients spent waiting while they are under treatment could not be documented.
6.3 Suggestions for further research
24
References
Asaro, P. V, Lewis, L. M., & Boxerman, S. B. 2007. The impact of input and output factors on Emergency Department throughput. Academic Emergency Medicine, 14(3): 235–42. Asplin, B. R., Flottemesch, T. J., & Gordon, B. D. 2006. Developing models for patient flow and
daily surge capacity research. Academic emergency medicine : official journal of the Society for Academic Emergency Medicine, 13(11): 1109–13.
Bazzoli, G. J., Brewster, L. R., May, J. H., & Kuo, S. 2006. The transition from excess capacity to strained capacity in U.S. hospitals. The Milbank quarterly, 84(2): 273–304.
Bellow, A. A., Flottemesch, T. J., & Gillespie, G. L. 2012. Application of the emergency department census model. Advanced emergency nursing journal, 34(1): 55–64.
Christ, M., Grossmann, F., Winter, D., Bingisser, R., & Platz, E. 2010. Modern triage in the emergency department. Deutsches Ärzteblatt international, 107(50): 892–898.
Derlet, R. W. 2002. Overcrowding in emergency departments: Increased demand and decreased capacity. Annals of Emergency Medicine, 39(4): 430–432.
Derlet, R. W., & Richards, J. R. 2000. Overcrowding in the nation’s emergency departments: complex causes and disturbing effects. Annals of emergency medicine, 35(1): 63–68. Ekelund, U., Kurland, L., Eklund, F., Torkki, P., Letterstål, A., Lindmarker, P., et al. 2011. Patient
throughput times and inflow patterns in Swedish emergency departments. A basis for ANSWER, A National Swedish Emergency Registry. Scandinavian journal of trauma, resuscitation and emergency medicine, 19(1): 37.
Fitzgerald, A., Dadich, A., & Sloan, T. 2010. Doing more with less: ways to improve patient flow in hospital settings. Asia Pacific Journal of Health Management, 5(2): 36–46.
Fitzgerald, A., Sloan, T., Simoff, S., Samaranayake, P., & Johnston, M. 2008. Visual workflow and process optimisation: A method of analysis for patient flow in the hospital emergency department. ANZAM (Australian and New Zealand Academy of Management) Conference.
FitzGerald, G., Jelinek, G. A., Scott, D., & Gerdtz, M. F. 2010. Emergency department triage revisited. Emergency medicine journal, 27(2): 86–92.
FitzGerald, G., Jelinek, G. A., Scott, D., & Gerdtz, M. F. 2011. Working with capacity limitations: operations management in critical care. Critical care (London, England), 15(4): 308. FitzGerald, Gerard, Jelinek, G. a, Scott, Deborah, & Gerdtz, Marie Frances. 2010. Emergency
department triage revisited. Emergency medicine journal : EMJ, 27(2): 86–92.
Forero, R., Mohsin, M., Bauman, A. E., Ieraci, S., Young, L., Phung, H. N., et al. 2004. Access block in NSW hospitals, 1999-2001: does the definition matter? The Medical journal of Australia, 180(2): 67–70.
25
Howell, M. D. 2011. Managing ICU throughput and understanding ICU census. Current opinion in critical care, 17(6): 626–33.
Hwang, U., McCarthy, M. L., Aronsky, D., Asplin, B., Crane, P. W., Craven, C. K., et al. 2011. Measures of crowding in the emergency department: a systematic review. Academic Emergency Medicine, 18(5): 527–38.
Karlsson, C. 2008. Researching operations management. Routledge.
Liu, S., Hobgood, C., & Brice, J. H. 2003. Impact of critical bed status on emergency department patient flow and overcrowding. Academic Emergency Medicine, 10(4): 382–5.
Michelen, W., Martinez, J., Lee, A., & Wheeler, D. P. 2006. Reducing frequent flyer emergency department visits. Journal of health care for the poor and underserved, 17(1 Suppl): 59– 69.
Nanninga, S. 2005. Wachttijden voor spoedpatiënten: Een onderzoek naar doorlooptijdverkorting van spoedpatiënten. Rijksuniversiteit Groningen.
Paoloni, R., & Fowler, D. 2008. Total access block time: a comprehensive and intuitive way to measure the total effect of access block on the emergency department. Emergency Medicine Australasia, 20(1): 16–22.
Richardson, D. B. 2002. The access-block effect: Relationship between delay to reaching an inpatient bed and inpatient length of stay. Medical Journal of Australia, 177: 492–495. Roukema, J., Steyerberg, E. W., Van Meurs, A., Ruige, M., Van der Lei, J., & Moll, H. a. 2006. Validity
of the Manchester Triage System in paediatric emergency care. Emergency Medicine Journal, 23(12): 906–10.
Solberg, L. I., Asplin, B. R., Weinick, R. M., & Magid, D. J. 2003. Emergency department crowding: consensus development of potential measures. Annals of emergency medicine, 42(6): 824–834.
Sprivulis, P., Grainger, S., & Nagree, Y. 2005. Ambulance diversion is not associated with low acuity patients attending Perth metropolitan emergency departments. Emerg Med Australas, 17: 11–15.
Thompson, J. 1999. Coroners and lack of emergency department resources. Journal of Emergency Medicine, 17(3): 541–542.
Van der Vaart, T., Vastag, G., & Wijngaard, J. 2011. Facets of operational performance in an emergency room (ER). International Journal of Production Economics, 133(1): 201–211. Van der Wulp, I., Schrijvers, a J. P., & Van Stel, H. F. 2009. Predicting admission and mortality
with the Emergency Severity Index and the Manchester Triage System: a retrospective observational study. Emergency Medicine Journal, 26(7): 506–9.
26
Van Veen, M., Steyerberg, E. W., Ruige, M., Van Meurs, A. H. J., Roukema, J., Van der Lei, J., et al. 2008. Manchester triage system in paediatric emergency care: prospective observational study. British Medical Journal, 337.
27
28
Appendix B – Table arrival and departure pattern
29
Appendix C – Throughput diagrams days with long throughput times
32
Appendix D – Questionnaire
Deze vragenlijst is bedoeld voor de verpleegkundigen van de spoedeisende hulp afdeling van het Medisch Centrum Leeuwarden.
Geachte deelnemer,
De voor u liggende vragenlijst is opgesteld in kader van mijn afstudeeronderzoek naar het verbeteren van de patiëntdoorlooptijden op de spoedeisende hulp. Uit het onderzoek dat vorig jaar is uitgevoerd binnen het MCL door mijn collega-student Jan Marten Verbree is onder andere naar voren gekomen dat een toename van nieuwe patiënten die zich in de ochtend aanmelden kan leiden tot capaciteitsproblemen later in de dag. Het doel van deze vragenlijst is om inzichtelijk te krijgen wat uw ervaringen zijn met het werken op de spoedeisende hulp.
Het invullen van deze vragenlijst duurt ongeveer vijf minuten en uiteraard is de anonimiteit gewaarborgd (het invullen van persoonlijke gegevens is niet nodig). De resultaten van deze vragenlijst zullen alleen gebruikt worden voor het lopende onderzoek, waarna ze vernietigd zullen worden. Deze resultaten komen dus in geen geval terecht bij derden.
Ik dank u bij voorbaat voor het invullen van deze vragenlijst.
John Bos,
Student Supply Chain Management aan de Rijksuniversiteit Groningen
33
Hieronder volgen een aantal stellingen en vragen. Geef na elke stelling aan in welke mate u het er mee eens bent. Gelieve elke vraag zo compleet mogelijk in te vullen.
1. “Soms heb ik het idee dat een drukke SEH met veel patiënten voorkomen had kunnen worden door eerder op de dag beter te reageren op nieuwe patiënten.”
2. “Het komt wel eens voor dat patiënten worden geplaatst in een behandelkamer, maar dat het een tijdje duurt voordat deze patiënt in behandeling wordt genomen.”
3. Als u meerdere patiënten tegelijk dient te behandelen, hoe bepaalt u dan welke patiënt u als eerste behandelt? Prioriteert u bijvoorbeeld op basis van de triage-code of behandelt u eerst de patiënt die de kortste behandeltijd vereist?
4. “Als ik meerdere patiënten tegelijk dien te behandelen, dan ben ik volledig vrij in de keuze welke patiënt ik als eerste behandel.”
5. Is de door de triage-verpleegkundige toegewezen triage-code nog leidend wanneer patiënten zich in behandelkamers bevinden en behandeld dienen te worden? Met andere woorden; wordt er altijd eerst naar een code geel patiënt gekeken en daarna pas aan een code groen patiënt?
34
7. “Ik vind dat het aantal verpleegkundigen dat per dagdeel volgens het dienstrooster aanwezig is te allen tijde voldoende is om aan de patiëntenstroom te voldoen.”
* Indien u het oneens bent met deze stelling, hoe zou u het aantal aanwezige verpleegkundigen wijzigen?
8. “Als er veel patiënten aanwezig zijn op de SEH, dan werk ik harder.”
9. “Als ik vier patiënten tegelijk moet behandelen, dan doe ik mijn werk net zo goed als wanneer ik één patiënt moet behandelen.”
* Indien u het oneens bent met deze stelling, kunt u dan uitleggen waardoor dit komt?
10. “Over het algemeen werk ik ’s middags na 12 uur harder dan in de ochtend.”
* Indien u het eens bent met deze stelling, kunt u dan uitleggen waardoor dit komt?
11. “Als alle behandelkamers bezet zijn, dan ben ik nog steeds prima in staat om elke patiënt te behandelen die aan mij is toegewezen.”
12. “Als een patiënt klaar is voor vertrek en verplaatst dient te worden naar een andere afdeling, dan wordt deze binnen een redelijke termijn opgehaald.”
Als u verder nog opmerkingen of aanvullende informatie heeft dan kunt u die hieronder kwijt.
