Development of a Quick Scan for assessing
stroke pathway performance
A case study on Groningen stroke pathway Master Thesis, MSc, Supply Chain Management University of Groningen, Faculty of Economics and Business
August 14, 2013 Shangyi Liu Student number: 2158779 E-mail: mailto:s.y.liu@student.rug.nl Supervisor Dr. ir. D.J. van der Zee
Co-assessor
Abstract
Table of content
Abstract ... 1
1. Introduction ... 1
2. Literature review ... 3
2.1 Background of stroke care ... 3
2.1.1 Organizational models of stroke pathway ... 3
2.1.2 System description of stroke pathway... 3
2.1.3 Delays in stroke care pathway ... 7
2.1.3.1 Pre-hospital delays ... 7
2.1.3.2 In-hospital delays ... 10
2.1.4 Medical standards and protocols ... 12
2.2 The Quick Scan methodology ... 13
2.2.1 Diagnostic methodologies for supply chain practice ... 13
2.2.2 The Quick Scan methodology ... 15
3. Development of Quick Scan for stroke pathway ... 20
3.1 Overview of Quick Scan for stroke pathway ... 20
3.2 The Quick Scan for stroke pathway ... 22
3.2.1 Data collection ... 24
3.2.2 Performance assessment ... 24
3.2.3 Solution findings ... 25
4. Methodology ... 27
5. Applying the Quick Scan for stroke pathway ... 30
5.2 Performance assessment ... 33
5.2.1 Performance analysis ... 33
5.2.1.1 Benchmark analysis ... 33
5.2.1.2 Spread analysis of EMS Performance ... 34
5.2.1.3 Compare study ... 35
5.2.2 Cause & effect analysis ... 37
5.2.3 Score system ... 38
5.3 Solution findings ... 40
5.3.1 Solutions & suggestions ... 40
5.3.2 Summarize findings ... 41
6. Discussion & Conclusion ... 42
Reference ... 44
1. Introduction
Stroke is the second most common cause of death and a major cause of disability in adults who are over 65 years old (Bouchaert et al, 2009). Intravenous tissue-type plasminogen activator (tpA) is the unique pharmacological treatment for acute ischemic stroke, and the therapy is only effective within 4, 5 hours after stroke onset (Mazighi et al, 2010; Bouchaert et al, 2009). Even within the narrow time window, for every minute of delay in treatment, 1, 9 million neurons are lost permanently (Chew et al, 2008; Bouchaert et al, 2009). Hence, the efficient timely access to medical treatment is crucial in curing acute stroke (Meretoja et al, 2012; Evenson et al; 2009). However, delays in stroke care delivery are the main barrier causing under-treatment (Kwan et al, 2004). Due to delays, many stroke patients do not arrive at the hospital within 4, 5 hours after the stroke onset. And therefore these patients lose the chance of receiving tPA treatment (Lahr et al, 2012). In UK, only 25% to 31% patients arrive at the hospital within 3 hours of stroke onset, and the delays vary largely (Kwan et al, 2004). Thus, a quick and comprehensive diagnostic tool is required to assess delays in stroke pathway. Although many researchers have reviewed factors causing delay, a systematic diagnosis approach for stroke pathway, which acts as a step-by-step guideline to assess delays in a short period of time, currently lacks.
Within the business context such diagnosis approach already exists. Given this context, Quick Scan is a diagnostic approach, which examines the supply chain operations of a business organization, and aims to understand and document the business processes within logistics and production (Basnet & Childerhouse, 2003; Naim et al, 2002). It enables a health check to identify and rank areas that yield most value with minimal resources of a supply chain (Berry et al, 1999). This Quick Scan approach attempts to identify long-term strategic actions which simplify and optimize the processes in the supply chain (Towill, 1997; Basnet & Childerhouse, 2003).
2. Literature review
To answer the research question and develop a Quick Scan for stroke pathway, certain knowledge is required to build the model. In this section, firstly, an introduction is given of the background of stroke care, including organizational models of stroke pathway, process description of stroke care system and factor analysis of delay in stroke chain, as well as an introduction of widely used medical standards and protocols on stroke pathway. Then it elaborates upon ‘the Quick Scan methodology’.
2.1 Background of stroke care
2.1.1 Organizational models of stroke pathway
There are four organizational models of acute stroke delivery system, which are primary stroke centers (PSC), comprehensive stroke centers (CSC), telemedicine and mobile stroke unit (MSU) (National Stroke Association, 2007).Among the identified organizational models in stroke care, PSC, CSC, and telemedicine focus on in-hospital care, while MSU works on pre-hospital phase of bringing the treatment to patients (Lahr et al, 2012). In this paper, we develop the Quick Scan for primary stroke centers and comprehensive stroke centers (PSC & CSC).
2.1.2 System description of stroke pathway
This section provides a detailed process description from the perspectives of key participants in the stroke care system. The process of stroke care delivery is illustrated in figure 1.
First respondents
Face-Arm-Speech-Test. Alternatively, the doctor could decide to visit the patients. If the doctor confirms a stroke onset, they would call 112 on behalf of the patients. When the patients call at night, or outside office hours, the patients’ call would be connected to on-duty nurses from the GP system. In that case nurses can diagnose stroke. If the patients have stroke symptoms, the nurse would call 112 to request for an ambulance.
Role of A&E service
If the patients choose A&E (Ambulance & Emergency) service, they will get treatment conform the standardized emergency rescue procedure. The emergency services staff would then activate the ‘Stroke Alert Call’ through the accident and emergency department, and dispatch an ambulance according to the level of urgency (A1 = blue-lighted, most urgent, ambulance must arrive within 15 minutes; A2 = less urgent, ambulance can arrive within 30 minutes) and distance to the nearest hospital (Mazighi et al, 2010). For EMS service (Emergency Medical Service), time is strictly recorded to measure the performance of delivery of care. The clock starts to kick after the dispatchers confirm the stroke symptoms and get in touch with the ambulance center. Time is recorded in three main periods, called ‘response time’, ‘time on scene’, and ‘transport time’. ‘Response time’ refers to the time that passes from the moment the ambulance department receives the call until they drive to patient’s location. Within the time periods, EMS staffs are called to the ambulance and drive to patient’s location. Subsequently, ‘Time on scene’ refers to the time from arriving at the patients till loading the patient into the ambulance. The main activities at scene include diagnosing the patient, obtaining a blood sample and then load them in the ambulance. After loading the patient in the ambulance, the patients are brought to hospital with the corresponding level or urgency. Before the patients arrive at the hospital, EMS staff has to inform the ER about the condition of the patient. Besides that, ER staff request for a neurologist before arrival of the patient.
Role of ER
Figure 1.Flow chart of stroke pathway (Lahr et al, 2013. Unpublished article) Emergency departmen t Nero imaging examination Treatment decision tpA mixin g Neurological examination Hospitalization Laboratory evaluation Stroke onset Call for help 911 call Ambulance
2.1.3 Delays in stroke care pathway
To quickly identify delay, it is necessary to understand the factors that might influence delay in the stroke pathway. This section illustrates delay factors on pre-hospital phase and in-hospital phase. Thereafter, several approaches will be discussed which might reduce those delays.
2.1.3.1 Pre-hospital delays
Delay in seeking medical attention
Low recognition of stroke symptom
Another reason causing delay in pre-hospital phase is low recognition of stroke symptom. Paramedics and medical staff diagnose stroke during the patients’ call and triage patients based on their diagnosis. Previously studies have shown that only 30-50% of strokes are correctly identified by emergency medical services dispatchers (Mazighi et al, 2010). In addition Mazighi et al (2010) found, low diagnosis accuracy causes big barrier for patients in getting proper medical treatment. The low recognition of stroke may also lie in the fact that the symptoms are not obviously or they are the mimic stroke. In fact, about 20% of presumed stroke symptoms are caused by other diseases (Fassbender et al, 2013). Study shows that if the symptoms of stroke included of facial droop, speech problem or the patients had a previously history of stroke, the likely of recognizing the stroke symptom is higher (Bouckaert et al, 2009).
Suggestions for reducing pre-hospital delays
Public awareness
Pre-notification
Furthermore, hospital pre-notification of patient arrival is helpful to shorten in-hospital delay (Ian et al, 2007). Studies have shown positive effects on implementing notification protocols in improving quality indicators and the rate of tPA treatment (Mazighi et al, 2010). It is also proven that the introduction of pre-hospital notification of the stroke team has big impact on reducing door-to-needle time, which leads to increased access to thrombolytic treatment and improved rate for tPA treatment (Mazighi et al, 2010). Thus, the American Heart Association and the European Stroke Organization recommend that the EMS personnel should pre-notify the receiving hospital that a stroke patient is on the way. In addition, EMS dispatchers are suggested to provide initial patient data, such as the time of onset of symptoms or performed treatment to patients, so that the hospital could prepare and mobilize resources before patients arrive at the hospital (Fassbender et al, 2013).
Telemedicine
2.1.3.2 In-hospital delays
Delay for neurology evaluation & CT scan
For in-hospital phase, delays are mainly from waiting time for medical assessment and CT scan. To be specific, delays appear as waiting for neurologist, transferring patient to CT scan, and getting CT test (Kwan et al, 2004; Evenson et al, 2009).
Suggestions for reducing in-hospital delays
Delays in stroke care pathway
Factors causing delay Pre-hospital
Causes Recommendations
1. Patient delay
Delay in seeking medical attention
1. Unawareness of stroke symptoms 2. Choices of medical services 3. Wake-up stroke
4. Personal reasons
1. Educational campaign
2. Public education by mass media
2. Clinical delay
Low recognition of stroke symptom
1. Unawareness of stroke symptoms 2. Difficulty of recognizing stroke 3. Mimics of stroke
4. Misdiagnosis of stroke to other diseases 1. Educational program 2. Medical training 3. Telemedicine 4. Pre-notification 5. Interventions 6.Standardized instruments 7. Medical protocols
Factors causing delay In-hospital
Causes Recommendations
1. Neurology evaluation 1. No-shows for neurologist 1. Pre-notification
2. CT scan 1. waiting in line 1. Moving CT to ER
2. Medical training
2.1.4 Medical standards and protocols
In this section, medical standards and protocols that are used on stroke pathway are elaborated upon. These guidelines are seen as benchmarks for examining the performance of the stroke pathway.
Guidelines and protocols
The generalized use of existing stroke management guidelines are ‘the American Heart Association’ and ‘the European Stroke Organization’. Protocols of telephone interview include ‘Medical Priority Dispatcher Systems Stroke protocol’ and ‘the Cincinnati pre-hospital stroke scale’ (Porteous et al, 1999; Ramanujam et al, 2008; Govindarajan et al, 2011).
FAST test
The well-known FAST test is a useful tool to assess the stroke symptom. It includes three key elements of facial weakness, arm weakness and speech disturbance. The symptoms are measured as facial droop: one side of the face does not move as well as the other; Arm drift: one arm does not move or drifts downward when held extended; Speech: patient slurs words, uses the wrong word, or cannot speak at all; Time: get help immediately (Fassbender et al, 2013; Massey et al, 1999).
Benchmarks
The NINDS published a benchmark for acute stroke pathway in 1996 (shown in table 2). It recommends several measurements for time delivery (Evenson et al, 2009). In addition, for EMS service, it is required that at least 95% ambulance must response within 15 minutes if the patient is targeted with blue-lighted (A1) and 30 minutes if targeted with A2 level.
Benchmark
Door-to-CT 25min
Door-to-needle 60min
Onset-to-needle 2hours
2.2 The Quick Scan methodology
2.2.1 Diagnostic methodologies for supply chain practice
A number of diagnostic methods exist to identify problems through implementation and operational processes. A summary of diagnostic methods is presented in table 3. Typical improvement tools such as SMED (Single Minute Exchange of Dies) are identified to reduce and simplify the setup time during changeovers (Moreira & Garcez, 2013). In addition, Profit Pool Mapping is developed to identify where the margin is generated from a value chain (Gadiesh & Gilbert, 1998). Furthermore, the Ernst and Young Navigator is a tool box as a best-practiced database, presenting example work sheets and implementation methodologies (Towill, 1999). In addition, the Eindhoven Technical University developed a so-called Quick Scan to assess performance and identify bottlenecks for business processes from customer to supplier (Scheaffer, 1992). The Quick Scan is developed as an originally stand-alone, systematic approach to diagnose supply chain performance (Naim et al, 2002). This method consists of interviews, data collection, and data analysis including brainstorming and cause-and-effect analysis and a feedback stage (Scheaffer, 1992; Naim & Childerhouse, 2002).
Accordingly, as an extension of the Quick Scan from the Eindhoven Technical University, a Quick Scan Audit Methodology (QSAM) performs learning activities on site, which includes statistical analysis and identification of best practices (Banomyong et al, 2005). QSAM is used as a key performance metric, and aims to measure the uncertainty factor. It is developed as a generic approach covering a wide range of supply chain issues and providing best means for understanding a supply chain’s current practices. It also identifies changes in resource-consuming which could be improved (Banomyong et al, 2005).
like a focused operation management approach, whereas, the QS methodology is a stand-alone, systematic approach to diagnose supply chain performance. Therefore, the QS method is most likely to be used as a generic approach in diagnosing delays on stroke pathway.
Diagnostic methodology Description
Master class sessions Industry experts look at the technology and manufacturing processes adopted by company. Improvement techniques include SMED and process control (Childerhouse & Basnet, 2003). Profit pool mapping Technique to identify where the margin in a value
chain is generated (Gadiesh et al, 1998).
Ernst and Young Navigator A tool box approach with a best practice database, sample work sheets and an implementation
methodology (Towill, 1999). Eindhoven University
Quick Scan
‘Quick Scan’ approach which focuses on business process from customer to supplier and
concentrates on gathering indicators of performance and identifying bottlenecks (Scheaffer, 1992).
Quick Scan Audit Methodology
Extends of the Eindhoven approach. QSAM is developed as a generic approach including statistical analysis and identification of and evidence for best practice (Childerhouse et al, 2005).
2.2.2 The Quick Scan methodology
This section will illustrate the Quick Scan methodology (from Eindhoven University) according to the research of building a supply chain diagnostic methodology published in 1999 by Childerhouse et al. The Quick Scan methodology comprises of 7 procedures and is designed to be completed within two week’s time window (shown in figure 2). After the two first stages of identifying a suitable supply chain and obtaining buy-in from business champion, three on-site days are required. The Quick Scan team normally consists of a core of four researchers and a business champion, together with a few additional internal members. The first stage of identifying a suitable supply chain is mostly straightforward and relies on a perceived problem that is identified by organizations and business units. This perceived problem indicates the need to conduct a Quick Scan. The perceived problem is usually defined by a business champion so that at the second stage, ‘buy-in from the business champion’ is to outline the method, explain the resource implications and determine the scope of the Quick Scan (Childerhouse et al, 2002).
Preliminary presentation
This section elaborates on the Quick Scan methodology and highlights specific objectives for their business processes. Representative products are then selected for in-depth analysis. Interview plans are developed to gather information from personal involved in the business processes.
Conduct the Quick Scan via four data collection techniques
cross-Uncertainty Circle model (Mason-Jones & Towill, 1998) which is utilized to identify the disturbances that may by encountered by a supply chain. The disturbances include the demand, supply, control or process areas that require the mostly re-engineering focus. Archival data is to pay attention to understand and document the four areas of uncertainty.
The first day of on-site involves these four formats of data collection and culminates through a brainstorming session designed to evaluate further avenues of investigation. This leads to a more focused approach on the second day to validate and further investigate key issues.
Analyzing the findings
Triangulation of the different ways the information is collected enhances a balanced perspective. A large number of analytical tools are utilized at this stage, such as Cause and Effect Analysis, Pareto Analysis and financial performance rankings. The first step at this stage is to agree on an outline of the supply chain and major business processes. This usually enables a development of a supply chain map agreed by a Quick Scan team, where everyone is given a chance to contribute, so as to avoid the potential bias from any individual players.
An initial brainstorming session is conducted to identify and document as many different first impressions as possible. It is important not to question the validity of any suggestions put forward at this stage, but to keep the impressions, whether good or bad. The next stage is to quantify and justify these first impressions. And in order to validate these first impressions it may be required to collect additional data on specific issues.
improvement activity. Finally, each improvement opportunity is ranked based on the evaluation of cost and time it takes to process the improvement and gain the associated benefit. The findings of the analysis stage are then summarized in a short list of validated and prioritized improvement opportunities for implementation.
Feedback presentation
Identify a suitable supply chain business process
Get buy in from the business champion
Preliminary presentation (half a day, on site) -Identify product / issue questionnaires
-Identify personnel for interview / develop interview plan -Quick tour / agree dates for feedback
-Explain purpose / issue requests for data requirements
Conduct the quick scan via four data collection techniques (2 Days, on site) -Completing and collecting questionnaires
-Process mapping / structured interviews
-Data collection –supply / demand / process & control uncertainties
-Brainstorm initial findings and conduct a more detailed investigation via the 4 techniques Analyzing the findings (3 Days)
-agree outline of the supply chain -good / bad first impressions
-quantify / justify first impressions and debate -identify additional data requirements and collect -identify the key business cost drivers
-identify the major pain from bad points
-create a ‘cause and effect’ diagram around the pain -identify the root causes
-utilize the best practice database to overcome root causes -brainstorm the remaining problems
-identify improvement opportunities
Feedback presentation (half a day, on site)
-present findings to management & business champion -initiate a round table discussion of findings
-develop an agreed action plan Write up the report (3 days)
3. Development of Quick Scan for stroke pathway
3.1 Overview of Quick Scan for stroke pathway
The development of quick scan for stroke pathway is based on the Quick Scan methodology built by Childerhouse et al in 1999. Childerhouse’s model is composed of four main phases, the preliminary presentation, data collection, data analysis and a feedback presentation. When building the quick scan for stroke pathway, we rephrase the model and re-organize the structure more dedicated to apply it on stroke pathway.
The first stage of Childerhouse’s model is preliminary presentation. The main activities in this stage include make questionnaires, develop interview plans, agree dates for feedback and explain purpose. Sequentially, data collection is the second stage. Stage two conducts the quick scan via four data collection techniques of collecting questionnaires, process mapping, semi-structured interviews and archival information. When develop the Quick Scan for stroke pathway, we combined the first two stages in one stage of ‘data collection’. Because the main objective of the first two stages is to get prepared for quick scan and collect data. We set the goal of QS as reducing delays in stroke pathway and collect data according to the four techniques. Through these methods, we describe all key activities in the stroke pathway and present first impression of performance.
develop the model not to take much consideration of cost and profit, but to identify delays and optimize process of stroke pathway. To present performance and give first impressions, we use data analysis approaches, such as benchmark analysis, statistics analysis and compare analysis. Secondly, we find root causes. We followed steps g) & h) to utilize analytical methods and tools to identify root causes. They utilize a large number of analytical tools to find root causes, including cause & effect analysis, Pareto Analysis and financial performance rankings. We only create cause & effect diagram to assess the stroke pathway, because as mentioned, we don’t pay much attention to cost analysis in assessing stroke pathway. We do not follow step i) because Best Practice Database aims for business context. Step j) is a suggestion of a brainstorming to discuss the performance presented from data. In our model, we will also seek wiser advice from expertise to interpret gathered data and also use literature to help defining factors influencing delay. Lastly, we follow step k), l) and m) to identify improvement opportunities. We will create a score system to score delays by ranking them to the level of severity and the efforts needed to improve. The scored delays are seen as improvement list for further implementation.
The last stage for Childerhouse’s model is feedback presentation. This stage aim to discuss improvement opportunities and identify proposed solutions. Activities include a) present findings to client; b) discussion of findings; and c) develop an agreed action plan. We rephrase this phase into ‘solution finding’ by providing solutions and evaluating the scan process. Because compare to the original scan model, the developed model aims for not only diagnosing delay but find reasons causing delay and provide solutions reducing delay. We will give solutions and suggestions depending on the improvement list and provide solutions according to the summary from literature (see table 1) and useful advises from expertise. We also consider the efforts needed to initiate the actions and discover the necessity of further investigation. At the end, we conduct a summary finding of the quick scan.
3.2 The Quick Scan for stroke pathway
The model of ‘Quick Scan for stroke pathway’ is displayed in figure 3. Each phase will be detailed described in the following sections.
The Quick Scanfor stroke pathway
Data collection Objective:
Process description
Performance presenting
Activities:
Develop questionnaires (example shown in appendix 1);
Describe key activities on stroke care pathway, visualize process by graphed displays, use flow chart to present the process mapping,
Develop semi-structured interview with key participants (including neurologist, GP doctor, EMS expertise, etc.)
Collect performance data from profile (Data gathering includes time recorded per key activity, diagnosis accuracy, and tPA treatment rate)
Performance Assessment Objective:
Give first impression by triangulating of gathered information Utilize analytical methods and tools to identify root causes Discover improvement opportunities
Activities:
Data analysis (Use medical standardize or protocol for benchmark mentioned from table 2; Use statistics tool / excel for data analysis; Compare performance under different conditions)
Use analytical tools analyze reasons causing delays (Use delay factor analysis from table 1; Draw Cause & Effect diagram)
Figure 3. The Quick Scan for stroke pathway Solution Finding
Objective:
Solutions & suggestions
Summary of findings
Further investigation & additional research Activities:
Provide solutions and suggestions according to the score system (Find solution from table 1; Inquire feedbacks / advices from key participants / experts) sum main findings of the quick scan
3.2.1 Data collection
This stage aims to describe key activities and obtain an overview of the performance of the stroke pathway. By doing so, we firstly develop interview plans to get overview of the process of stroke care delivery. Categories of experts to be includes are (1) GP doctors; (2) EMS expertise; (3) Medical staff. The involvement of these experts is inquired to obtain detailed process descriptions for their department, and point out problems and issues they consider to be necessary to improve. Based on the gathered information, the process of stroke care system is mapped into a flow chart. The flow chart or process mapping is a graphic depiction of process showing inputs, outputs, and steps within the process. It gives clear picture of what activities are carried out and how activities are performed on stroke pathway (Mclaughlin & Olson, 2008). Process mapping is an important purpose of this stage.
Next, performance is presented through more collected data. Data are collected in three formats: (1) time spent per activity, (2) diagnosis accuracy; and (3) tPA treatment rate. Many time slots are recorded to show the performance of stroke pathway. The time slots include door-to-examination time, door-to-CT time, door-to-lab time, door-to-needle time (DNT), onset-to-needle time (ONT), and patient travelling time. We track the patient’s movement and record time spent per each activity from the registration by EMS system and hospital system. We gathered diagnosis accuracy, as well as tPA treatment rate from the interviews. Once collected, these data is displayed in tables and diagrams to assess the performance of stroke system.
3.2.2 Performance assessment
each type of situation. For instance, when EMS is analyzed by spread, certain trends can be discovered. To analyze trends in the data, causes are found to explain the variance found. In this way, spread analysis helps to analyze and discover relationships. The third approach is to compare the performance under different conditions. For example, a comparative study could be made to compare the performance of patients treated with tPA to those who are not treated with tPA. In case of remarkable differences, arguments are required to explain the cause of these differences. And by explaining the reasoning, more findings can be generated.
Next, we further identify reasons causing delay. We have two large supports to find reasons, ‘defined delay factors from literature’ and ‘pointed out reasons from expertise’. We develop a ‘cause & effect diagram’ to explore the factors which lead to delay. This diagram is used to investigate and eliminate a problem as clearly as possible. The diagram does so exploring the root causes through identification of the causes until every possible cause has been identified (Mclaughlin & Olson, 2008).
Thirdly, we score delay in terms of its severity and its possibility to reduce. The score system gives indication of biggest delays and takes consideration of the possibility for improve. Delays are scored on the scale of 0 to 10 (10 being the highest). Longer delay gets a higher score. In addition, delay is weighed considering the level of improvement. Delays are marked as plus (+) or negative (-). Easier reduced delays get more plus. In the score system, longer and easier-reduced delays rank higher. Reducing the highest ranked delay would get the higher return with the least effort. Thus, the rank from the score system is considered to be an improvement list which seeks to gather the biggest benefits while minimizing efforts.
3.2.3 Solution findings
4. Methodology
Researcher Robert K. Yin defines case study research method as an empirical inquiry that investigates a contemporary phenomenon within its real-life context; when the boundaries between phenomenon and context are not clearly evident; and in which multiple sources of evidence are used (Yin, 1984, p. 23). Case study method is utilized to build upon theory, to produce new theory, to dispute or challenge theory, to explain a situation, to provide a basis to apply solutions to situations, to explore, or to describe an object or phenomenon (Karlsson, 2009). This research aims to produce a new model based on existing methodology. Thus, this study uses case study method to develop the new model.
Step 1. Determine and Define the Research Questions
The aim of this research is to develop a scan tool to quickly and comprehensively diagnose delay in stroke pathway. It begins with a review of literature to determine what prior studies have done about this topic and we use literature to give overview of existing diagnostic methodologies, and provide knowledge background of stroke pathway. Due to the conducted literature review, we are certain that no previous study has been performed to investigate the suitability of the Quick Scan in healthcare. We then determine the research question to develop a scan system to quickly and comprehensively diagnose delays in stroke care pathway.
Step 2. Select the Cases and Determine Data Gathering and Analysis Techniques
We consider multiple sources of data for this study. We use document examination and open-ended interviews with key participants of each organization. We record facts, opinions, and insights during each interview. We use tools to collect data including surveys, interviews, documentation review, observation, and even the collection of physical artifacts (Yin, 1984). We develop rounds of semi-constructed interviews with key stakeholders in stroke pathway. Relevant stakeholders include EMS dispatchers and medical staff such as stroke physicians and nurses, and other relevant participants such as family physicians. We collect three formats of data: (1) time spent per activity, (2) diagnosis accuracy; and (3) tPA treatment rate. Collected time slots include door-to-examination time, door-to-CT time, door-to-lab time, door-to-needle time (DNT), onset-to-needle time (ONT), and patient travelling time. In order to clarify the data collection, several definitions are used. The time of stroke onset is defined as the time when the stroke symptoms started (Vemmos et al, 2000). For patients who noticed symptoms after awakening, stroke onset was defined as the time when the patient went to sleep or was last seen awake without symptoms (Panagiotis et al, 2010). Pre-hospital was defined as the time from the stroke onset to arrival at the ER; delay in the CT scan was recorded as the time from presentation at the ER to the completion of the CT scan, and total delay refers to the time that passes from symptoms till the patient gets treatment (Panagiotis et al, 2010).
Throughout the research design, we make sure that the study is well constructed to ensure construct validity, internal validity, external validity, and reliability. According to Karlsson (2009), within-case examination along with literature review enhances external validity. We will pay attention to the stability, accuracy, and precision of measurement to remain reliable. Also, we will examine the case study design to ensure the process is well-documented so that the model can be developed with same process over again (Karlsson, 2009).
Step 3. Prepare to Collect the Data
We prepare to collect data by first contacting each expertise to be interviewed to gain their cooperation, explain the purpose of the study, and request for relevant organizational documents.
Step 4. Collect Data in the Field
study and analyzing the literature found, we conducted another round of interviews with more specific questions. Notes were taken during the interview and transcribed when the interview is completed. We interpret the interviewed data in support with literature, and narrow down the research domain to specify the exact focus of the research.
Step 5. Evaluate and Analyze the Data
The analysis of data is separated into two parts, quantitative data with real data sets and qualitative data of transcript of interview conversations. The actual data sets, such as lead time, will be observed and analyzed using certain software, such as SPSS or Excel, to find out the current bottlenecks in the system. The material of conversations is categorized and highlighted to analyze and support the findings. In the process of analyzing this information, the transcript from different stakeholders is categorized and organized based on the content of the conversation. Within each category, arguments were be coded with key opinions that help to answer the research question.
Step 6. Prepare the Report
5. Applying the Quick Scan for stroke pathway
5.1 Data collection
The application of the quick scan on Groningen stroke pathway starts from this chapter. This phase provides a detailed process description of the Groningen stroke pathway and indicates the overall performance of the stroke care system.
5.1.1 Process description
5.1.2 Performance presenting
This part presents an indication of the overall performance of Groningen stroke pathway. The performance is presented from three aspects: tPA treatment rate, time spent per key activities and diagnosis accuracy.
tPA treatment rate
This section illustrates tPA treatment rate on Groningen stroke pathway (shown in figure 4). In total, 280 patients are diagnosed with ischemic stroke. Among them, 159 of the ischemic patients are delivered over 4, 5 h, which are not suitable for thrombolytic therapy. Among the 121 patients that are delivered within 4, 5 h, 62 patients are treated with tPA. Thus, of all 280 ischemic stroke patients, 22.1 % of the patients get the TPA treatment. The relative low tPA treatment rate is partly because not all stroke patients are suitable for tPA treatment. For example patients with mild stroke, wake-up stroke, or brain bleeding are not suitable for tPA treatment.
4 with Ischemic stroke patient
N = 280, (100%)
62 (51.24%) received tPA Ischemic stroke patients
< 4, 5 h: N=121 (43.2%)
59 (48.76%) not treated with tPA
Delivery time
The overall performance of Groningen stroke care pathway is shown in table 5. Of the 58 patients with complete data records, the average onset-to-needle time is about 2 hours (125, 98 minutes), within this time period, it takes about 50 minutes (47, 31min) before patients arrive at the hospital (ambulance response time + time on scene + transport time), and 40 minutes (39, 41min) for treatment after they arrive the stroke center. In the pre-hospital phase, on average, it takes about 8 minutes to arrive to patient’s house and around 20 minutes to pick them up to the ambulance and 16 minutes to transfer the patients to hospital.
Table 5. Descriptive statistics of delivery time.
Diagnose accuracy
Apart from transferring the patients, another important decision in pre-hospital phase is diagnosis accuracy. Nearly 70% of the stroke patients are correctly diagnosed by EMS dispatchers, whereas 30% stroke patients are not diagnosed as stroke during the call, but are examined as stroke after they arrive at the hospital.
Descriptive Statistics N Minimum (minutes) Maximum (minutes) Mean (minutes) Std. Deviation Variance (minutes)
Ambulance response time 58 3 20 8,33 3,909 15,277
Ambulance time on scene 58 9 42 22,17 7,651 58,531
Ambulance transport time 58 4 33 16,81 7,608 57,876
Door-to-neurological 58 0 8 0,76 1,867 3,485
Door-to-CT scan 58 2 40 10,64 7,622 58,095
Door-to-laboratory 27 12 50 30,56 7,827 61,256
Door-to-needle 58 14 120 39,41 21,149 447,264
Onset-to needle 58 52 260 125,98 42,593 1814,193
5.2 Performance assessment
This phase assesses the performance and diagnoses delays on Groningen pathway. It contains three main procedures, performance analysis, cause and effect analysis and discovery of improvement opportunities.
5.2.1 Performance analysis
In this section, performance is analyzed upon three approaches, benchmark analysis, data analysis, and variable analysis.
5.2.1.1 Benchmark analysis
Benchmark performance of ER
Compared to the benchmark NINDS published in 1996, Groningen pathway performs rather excellent. Door-to-CT scan, door-to-needle, and onset-to-needle are three important indicators for performance evaluation. As shown in table 6, Groningen stroke pathway over-achieved the published standards.
Delivery time Performance of Groningen stroke pathway
NINDS benchmarks
Door-to-CT 10min 25min
Door-to-needle 40min 60min
Onset-to-needle 3hours 2hours
5.2.1.2 Spread analysis of EMS Performance
Figure 5. Histogram for time spent of EMS service delivery.
5.2.1.3 Compare study
Patients treated with tPA versus patients not treated with tPA
To evaluate the performance of Groningen stroke pathway, we compare the outcome between patients treated with tPA and those who are not treated with tPA. As shown in table 7, all
0 50 100 0 5 10 15 20 25 30 35 40 More Fr e q u e n cy Bin
Ambulance response time
Frequency 0 20 40 60 0 5 10 15 20 25 30 35 40 More Fr e q u e n cy Bin
Time on scence
Frequency 0 20 40 0 5 10 15 20 25 30 35 40 More Fr e q u e n cy BinAmbulance transport time
targeted level of urgency, etc. We firstly find out that, on average, no-tPA-treated patients spent more than one and half hours than tPA-treated patients to get treated. To explain the reason, we consider patients’ choice of first responder. We found out that except for patients who have stroke in hospital, the rest tPA-received-patients are all delivered to hospital by EMS transport and are blue-lighted (A1). In comparison, for patients who did not receive tPA, 71% of the patients are delivered by EMS service and among which, 79% of the patients are targeted with A1. Thus, we can explain that patients who are treated with tPA are much likely to receive faster response in receiving medical help and thus get quicker access to treatment.
Compare study Part one Activity durations in minutes
(all patients) All patients (n=280) Patients treated with tPA (n=62) Patients not treated with tPA
(n=218) Stroke onset to call for help 90 (0 – 490) 25 (0 – 127) 135 (0 – 583) First responder (911/ GP) 12 (0 - 55) 11 (0 – 63) 13 (0 – 53) EMS
Response time 10 (3 - 29) 8 (9 – 15) 11 (4 – 51)
Time spent on scene 21 (8 – 40) 22 (9 – 41) 20 (7 – 21)
Transport time 17 (1 - 44) 17 (4 – 32) 17 (1 – 48)
Time to neurological examination 2 (0 - 15) 1 (0 – 9) 2 (0 – 20)
Time to CT examination 12 (2 - 45) 12 (2 – 42) 12 (2 – 50)
Time to laboratory examination 32 (13 - 63) 32 (12 – 62) 33 (24 – 64) Time to treatment decision 9 (0 – 16) 9 (0 – 16)
Time to mix tPA 5 (2 – 8) 5 (2 – 8)
Part two
Diagnostics Percentage Percentage Percentage
Table 7. Comparative study between tPA-treated patients versus no-tPA-treated patients (Lahr et al, 2013, unpublished articles)
5.2.2 Cause & effect analysis
This section investigates the causes which lead to delays. As shown in figure 6, cause and effect are analyzed from four perspectives, patient, EMS, GP, and hospital (ER).
Patient delay among others results from a low recognition of stroke symptoms and asuboptimal preference of the first care-provider. Delay occurs when patients fail to recognize stroke symptoms and request their GP to visit them, at the patients’ location. Patient delay and delay
3. Intra-hospital 3 3 1
Choice of first responder 1. 911 call 2. GP consult by phone 3. GP consult by visit 30 19 27 50 16 29 24 26 21 EMS transport, level of urgency
distance to be covered. In comparison, a relatively longer waiting time (over 20 min) for ‘time on scene’ can be caused because many activities were taken in this period, including examining the patient, symptom inquiring, getting a blood sample, and transferring them to the ambulance. Also, in some cases, it requires more time to reach and load the patient to the ambulance. Once arrived at the hospital, the biggest delay is the waiting time for the lab results and for neurologist examination.
Figure 6. Cause & effect diagram
5.2.3 Score system
Delays are scored to the severity of delay and the improvement opportunity. For Groningen stroke pathway, the biggest delay is ‘time on scene’ and the ‘time-for-waiting-lab-results’. On-scene activities play an important role of first assessment. But time spent on On-scene is largely depending on the condition of the patient, thus it is hard to be reduced. Delay of ‘time on scene’ is then graded as 7 with ‘+/-’ in the score system.
The activity of checking lab results, on the other hand, is a routine procedure in hospital and can be improved by advanced technologies. A new device, called point of care (POC), is going to be introduced to Groningen stroke pathway. POC takes only 2-3 min for test results, which shows
Waiting time
Patient GP
EMS Hospital
Not recognize
stroke symptom Wait for family
physician Pay visit to patient home
big advantage towards regular tests and saves more than 15 min of waiting time. After applying POC, the possibility of reducing the waiting time for lab results can be largely improved. The door-to-needle time is expected to be largely improved from currently 45 min to 30 min. Thus, delay of waiting lab results is graded 9 and a ‘++’ in the score system.
Next, the ambulance transportation time is another important delay factor. However, the travelling time is largely relying on the distance between patients’ home to the stroke center. It is therefore hard to improve and is graded as 6 with ‘+/-’ in the score system.
Furthermore, GP plays an important role in possibly reducing delay in stroke pathway. Although the time spent per visit to patient is not measured in time slots, and the percentages of GP doctors paying visit to patients’ home is unknown. It is confirmed that family visits still cause big delay in pre-hospital phase. If the GP doctors could transfer the patients to EMS service at once, more potential stroke patients could get treated before it is too late. Thus, GP delay is graded as 8 with ‘+/-’.
In addition, patient’s delay is a seriously delay factor in the whole stroke pathway, and it is easier to be improved by education programs. Thus it is ranked as 9 with ‘+’.
A last factor mentioned is the impact of the availability of neurologist on the efficacy of ER. In most of cases, the neurologist is informed before the patients arrive at the hospital, and is standby in ER before patient arrives. Nevertheless, there are some exceptional cases where the neurologist is not available at the moment and patients need to wait long time for them to show. Thus delay for neurologist no-show is graded as 6 with ‘+’.
Table 8. Scoring system
5.3 Solution findings
5.3.1 Solutions & suggestions
In this section, we give suggestions according to the improvement list (ranks in score system) to reduce delay. Firstly, as checking lab result takes longer, we suggest introducing the new technique of POC by shortening the DNT from 40 to around 30 minutes. Secondly, to reduce patient delay, an educational campaign is suggested to improve thepatients’ knowledge of stroke symptoms. This campaign can be launched through public mass media. In addition, care brochures can be send to elderly people, gathering volunteers to give information during face-to-face meetings or presenting audio and slide shows at public events (Bouckaert et al, 2009). Thirdly, to avoid delay from GP’s, the GP can be stimulated to transfer patients at once to the A&E service in case they found the patient might suffer from a stroke. Next, to reduce delay of ‘time on scene’, EMS dispatchers can be informed to transfer the patient to the hospital as quickly as possible. In addition, to avoid the situation of no-shows of the neurologist, pre-notification is considered to be the best suggestion. A pre-notified message send by the EMS dispatcher to the medical staff of the ER ensures the neurologist is aware a patient will arrive. Lastly, no suggestions can be given with regard to the ambulance transport time, because this largely depends on the distance to be covered
Scoring system
Activities Score Mark Phase
Checking lab results 9 ++ In-hospital
Patient delay 9 + Pre-hospital
GP family visit 8 +/- Pre-hospital
Time on scene 7 +/- Pre-hospital
Neurologist 6 + In-hospital
5.3.2 Summarize findings
Main findings of applying the Quick Scan on Groningen stroke pathway is summed in table 9.
Table 9. Main findings for application of Quick Scan in Groningen pathway
Activities causing delay Solutions and suggestions reducing delay
1. Checking lab results Introduce technique of POC to shorten time for waiting lab result
2. Patient delay Develop education campaign to increase awareness of stroke
symptom
3. GP family visit Avoid family visit and transfer patient to EMS as quickly as
possible
4. Time on scene Instead of spending long time examining patient at patients’
home, but quickly transfer patients to hospital
5. Neurologist no-shows Encourage pre-notification, enable more corporation
between EMS and ER member
6. Discussion & Conclusion
The main proposition of this research is to develop a diagnosis tool to quickly and comprehensively diagnose delays in the stroke pathway. To achieve this objective, a Quick Scan is developed basing on the QS methodology as already available within the supply chain context. The developed Quick Scan is used to identify unnecessary process-steps and optimize the stroke care process to reduce delays. The improved stroke pathway, which results from the usage of the Quick Scan tool, enables more stroke patients to get a proper treatment in time, and thereby reduce damage to their health. Given the enhanced care for patients, the practical relevance of this study is obvious.
To enable a quick diagnosis, a Quick Scan is designed as a well-structured and procedure-driven diagnosis tool. The Quick Scan is developed into three phases: ‘data collection’, ‘performance assessment’, and ‘solution findings’. The scan provides step-by-step instructions to identify delays and gives suggestions to reduce the identified delays.
To enable a comprehensive diagnosis of the performance of the Groningen stroke pathway, the Quick Scan method examines all process steps taken in the stroke chain. In developing the method, several key participants and experts in the stroke pathway are interviewed. Based on the information gathered from these expertise, we present a comprehensively system description and cover a wide range of identified bottleneck. In addition, recommendations are made on how to tackle these bottlenecks.
Therefore, the Quick Scan for stroke pathway has particular strengths. Firstly, it is a relatively quick and efficient process. Secondly, it undertakes a deeper diagnostic within a short period by triangulating four formats of data sources. Thirdly, it gives a holistic overview of the stroke pathway by seeking advice from key participants in the chain.
Reference
Atte M., Daneil S., Satu M., Turgut T., Perttu J.L., Markku K.. 2012. Reducing in-hospital Delays to 20 Minutes in Stroke Thrombolytic. American Academy of Neurology, 79 (4): 306-313.
Banomyong, R., Basnet, C., Childerhouse, P., Deakins, E., Disney, S. M., Naim, M. M., & Towill, D. R. 2005, July. Internationalising the quick scan audit methodology. In 18th International Conference on Production Research Proceedings.
Basnet, C., & Childerhouse, P. 2003. Application of a supply chain diagnostic to a New Zealand manufacturer–a case study. In 38th Annual Conference of the Operational Research Society of New Zealand, Hamilton: 253-262.
Berry D., Evans G.N., Mason-Jones R., and Towill D.R. 1999. The BRP Scope Concept in Leveraging Supply Chain Performance. Business Process Management Journal 5(3): 254-274. Bouckaert, M., Lemmens, R., & Thijs, V. 2009. Reducing prehospital delay in acute stroke. Nature Reviews Neurology, 5 (9): 477-483.
Chew, S., Kausar, S. A., & Sturman, S. 2008. Thrombolysis in acute ischaemic stroke pathway. International Journal Care Pathways, 12(1):10-22.
Childerhouse P., Disney S.M. and Naim M.M. 1999. “A Quick Scan Method for Supply Chain Diagnostics.” Proceedings of the 4th International Symposium on Logistics, Florence, July: 755-760.
Daniel B. M., Julie M. H.2008. Healthcare Operations Management. Chicago: Health Administration Press.
Eisenhardt, K. M., & Graebner, M. E. 2007. Theory building from cases: Opportunities and challenges. Academy of management journal, 50 (1): 25-32.
Ford, A. L., Williams, J. A., Spencer, M., McCammon, C., Khoury, N., Sampson, T. R., & Lee, J. M. (2012). Reducing door-to-needle times using Toyota’s lean manufacturing principles and value stream analysis.Stroke, 43(12): 3395-3398.
Gadiesh, O., & Gilbert, J. L. 1998. Profit pools: a fresh look at strategy. Harvard Business Review, 76(3): 139.
Golinska, P., Fertsch, M., & Pawlewski, P. 2011. Production flow control in the automotive industry–quick scan approach. International Journal of Production Research, 49(14): 4335-4351.
Govindarajan, P., Ghilarducci, D., McCulloch, C., Pierog, J., Bloom, E., & Johnston, C. 2011. Comparative evaluation of stroke triage algorithms for emergency medical dispatchers (MeDS): prospective cohort study protocol. BMC neurology,11(1): 14.
Harbison, J., Massey, A., Barnett, L., Hodge, D., & Ford, G. A. 1999. Rapid ambulance protocol for acute stroke.The Lancet,353(9168): 1935.
Harbison, J., Hossain, O., Jenkinson, D., Davis, J., Louw, S. J., & Ford, G. A. 2003. Diagnostic accuracy of stroke referrals from primary care, emergency room physicians, and ambulance staff using the face arm speech test.Stroke, 34(1): 71-76.
Harraf, F., Sharma, A. K., Brown, M. M., Lees, K. R., Vass, R. I., & Kalra, L. 2002. A multicentre observational study of presentation and early assessment of acute stroke.BMJ: British Medical Journal,325(7354): 17.
Hayward, Leanne M. 2012. How applicable is lean in mental health? A critical appraisal. International Journal of Clinical Leadership. 17 (3): 166-173.
Kuo M.H., Borycki E., & Kushniruk A., 2011. A Healthcare Lean Six Sigma System for Postanesthesia Care Unit. Manage Health Care. 20 (1): 4-14.
Lahr, M. M., Luijckx, G. J., Vroomen, P. C., van der Zee, D. J., & Buskens, E. 2012. Proportion of patients treated with thrombolysis in a centralized versus a decentralized acute stroke care setting. Stroke, 43 (5):1336-1340.
Lahr, M. M. H., Luijckx, G. J., Vroomen, P. C. A. J., Van Der Zee, D. J., & Buskens, E. 2013. A simulation based approach for improving utilization of thrombolysis in acute brain infarction. Working paper. University Medical Centre Groningen, Groningen, The Netherlands.
LaMonte, M. P., Xiao, Y., Hu, P. F., Gagliano, D. M., Bahouth, M. N., Gunawardane, R. D., ... & Cullen, J. 2004. Shortening time to stroke treatment using ambulance telemedicine: TeleBAT.Journal of Stroke and Cerebrovascular Diseases,13(4): 148-154.
Lasserson, D. S., Chandratheva, A., Giles, M. F., Mant, D., & Rothwell, P. M. 2008. Influence of general practice opening hours on delay in seeking medical attention after transient ischaemic attack (TIA) and minor stroke: prospective population based study.BMJ: British Medical Journal,337.
Levine, S. R., & Gorman, M. 1999. “Telestroke” The Application of Telemedicine for Stroke.Stroke,30(2):464-469.
Mazighi, M., Derex, L., & Amarenco, P. 2010. Prehospital stroke care: potential, pitfalls, and future. Current opinion in neurology, 23 (1): 31.
Meredith, J. 1998. Building operations management theory through case and field research. Journal of operations management, 16(4): 441-454.
Moreira. A.C., & Torres. G.P.M.. 2013. Implementation of the Single Minute Exchange of Die (SMED) Methodology in Small to Medium-sized Enterprises: A Portuguese Case Study. International Journal of Management. 30 (1): 66-87.
Naim M. M., Childerhouse P., Disney S. M., Towill D. R., 2002. A Supply Chain Diagnostic Methodology ~Determining the Vector of Change, Computers and Industrial Engineering, 43: 135-157.
National Stroke Association (NSA). Building the Case for a Primary Stroke Center, A Resource Guide. -2007. Tex.: National Stroke Association; 2007.
Porteous, G. H., Corry, M. D., & Smith, W. S. 1999. Emergency medical services dispatcher identification of stroke and transient ischemic attack. Prehospital Emergency Care,3(3): 211-216.
Ramanujam, P., Guluma, K. Z., Castillo, E. M., Chacon, M., Jensen, M. B., Patel, E., ... & Dunford, J. V. 2008. Accuracy of Stroke Recognition by Emergency Medical Dispatchers
andParamedics-San Diego Experience. Prehospital Emergency Care,12(3): 307-313.
Rosamond, W. D., Gorton, R. A., Hinn, A. R., Hohenhaus, S. M., & Morris, D. L. 1998. Rapid response to stroke symptoms: the Delay in Accessing Stroke Healthcare (DASH)
study.Academic Emergency Medicine,5(1): 45-51.
Saver, J. L., Smith, E. E., Fonarow, G. C., Reeves, M. J., Zhao, X., Olson, D. M., & Schwamm, L. H. 2010. The “Golden Hour” and Acute Brain Ischemia Presenting Features and Lytic Therapy in> 30 000 Patients Arriving Within 60 Minutes of Stroke Onset.Stroke,41(7): 1431-1439.
Schaeffer, G. 1992. Exchange of quick scan know how in EUNiL. Internal Document, Eindhoven University.
Siggelkow, N. 2007. PERSUASION WITH CASE STUDIES. Academy of Management Journal, 50(1):20-24.
Towill, D.R. 1999. Management theory: Is It of Any Practical Use? Engineering Management Journal. June: 111-121.
Voss, C., Tsikriktsis, N., & Frohlich, M. 2002. Case research in operations management. International journal of operations & production management, 22 (2): 195-219.
Waring, J. J., & Bishop, S. 2010. Lean healthcare: Rhetoric, ritual and resistance. Social science & medicine, 71 (7): 1332-1340.
Wester, P., Rådberg, J., Lundgren, B., & Peltonen, M. 1999. Factors Associated With Delayed Admission to Hospital and In-Hospital Delays> in Acute Stroke and TIA A Prospective, Multicenter Study.Stroke,30(1): 40-48.
Scott, W. M., & Walton, D. A. 2010. Maximizing Case Efficiency: Lessons Learned From Lean-Process Management Philosophy Utilized In Automotive Manufacturing.[Article]. FDCC Quarterly, 61(1): 2-16.
Yin, R. K. (1984). Case study research: Design and methods. Newbury Park, CA: Sage.
Appendix 1
Example open questions for interview Part one: General questions for Data Collection
What is the role of for your department in the stroke pathway? what do you think your main duty in providing care?
What are the specific procedures in your department? Do you follow some rules/guidelines/ protocol during the process of care delivery?
What are the barriers causing delays in your department? What issues you think need to be improved in your department?
Do you think the collaboration of departments in the stroke chain need to be improved? If so, in what way?
Part two: Dedicated questions for Data Analysis
To GP doctors:
What is the role of GP doctors in treating acute disease like stroke? If you confirm the stroke symptom during patients’ call, would you pay visits to patients’ home or transfer them directly to 911?
To EMS expertise:
As shown from the spread, what reasons cause the big variance on ‘time on scene’? What main activities are taken on scene? Is it possible to conduct the on-scene procedures in the ambulance instead of in patients’ house?
To Neurologist:
