Being a safe, non-invasive and highly accurate means of diagnostic radiology, Magnetic Resonance Imaging (MRI) is considered now as an indispensible element of the diagnostic process for a myriad of pathologies. However, the growing popularity of MRI imposes rather complex challenges for hospital administrators and healthcare practitioners. Two major constraints on MRI units – namely scanner time and human capital – require new thinking on the means by which MRI services are organized and provided.
Current literature on the performance of MRI units stresses the importance of adopting new patterns of management, reorganization of working processes and changes in the allocation of responsibilities. This paper analyzes and maps the current flow of service in the MRI unit at the Kuwaiti Amiri Hospital, while giving special emphasis on patient journey and the efficiency of current processes. Moreover, in order to improve patient satisfaction and organizational performance, this paper proposes a novel design for workflows in the MRI unit, which takes into account the current constraints of the system.
Overview: The MRI Department at Amiri Hospital
The Amiri Hospital is a major general hospital in Kuwait City, serving roughly 400,000 patients per year. Al- Amiri’s Department of Clinical Radiology has two superconducting MRI scanners with highfield capabilities (>0.5 T). In terms of equipment (but not human capital, as described later), the department is equipped well enough to perform merely all types of diagnostic assignments. It has two different types of scanners (one 0.7 T open-bore system and one 1.5 T closed-bore system), as well as organ-specific coils for all common examinations. When used optimally, this flexibility allows for more efficient scanning processes, shorter scanning time and reduction of human movement artifacts compared to more rigid settings.
Until several years ago, Amiri patients enjoyed high availability of MRI resources, as the number of MRI referrals was rather small. Today, however, the average waiting time for an MRI appointment has reached three months and continues to increase as the number of patients grows rapidly. Moreover, patients referred to the MRI unit must endure outdates work processes and clunky bureaucratic procedures.
The Amiri hospital is considered to be one of the major hospitals receiving the majority of patients who are living in Kuwait city. This huge health organization is structured and formed by multiple processes in different medical and administrative fields. The MRI appointment is considered to be a combination of both fields. The need to change the process becomes an urgent matter to the department that must be analyzed and improved. The cooperation needed from the clinicians, IT technologist and the patient themselves was a huge challenge to the department to successfully go through the transformation phase and to insure the appropriate diagnosis and treatment for each patient.
Mapping Patient Journey at the MRI Unit
Process mapping is a workflow diagram to bring forth a clearer understanding of a process or series of parallel processes. Patient journey mapping is a highly effective patient-focused analysis, which gives planners and administrators understanding of the interaction between their facilities and customers. It may also assess gaps in service, unforeseen delays and inefficient use of resources, thereby helping to enhance the quality of service by indicating specific flews in the provision of care (NHS Institute for Innovation and Improvement, 2008).
MRI appointments involve a long journey that starts from the day the patient sees a physician until taking the examination results. Although it sounds easy and smooth, Figure 1 below shows that taking an MRI appointment at Amiri hospital is not simple.
The major steps involved in the process of a typical patient at the Amiri MRI unit are as follows: 1. The patient enters the radiology department with an MRI request. Depending on the nature of the request (i.e. the organs and the type of examination/potential pathology involved), the receptionist will ask the patient to provide one or more blood tests before having the appointment sheet. 2. The patient will go to perform the blood test, the results of which may take several days to obtain from the busy laboratory. 3. Upon receiving the blood results, the patient goes back to Department of Radiology to submit his or her results. This involves waiting in a long queue for several hours, where there are hundreds of patients are waiting to submit their different requests to the appointment desk. The receptionist will then collect patient information and prepare the request for a review by the radiologist in charge. 4. The patient waits for the radiologist decision. The latter can either approve the request and order a specific scan, or reject the request due to clinical reasons. It is also possible that the radiologist will ask for more information from the patient’s GP and/or additional tests to be submitted for further review. 5. Upon the radiologist’s consent to perform the examination, the patient must wait again for the receptionists to schedule an appointment. As mentioned earlier, non-acute cases would not receive MRI appointments within less than two months. The usual waiting time is 14 weeks. The receptionist prints out an appointment form, which entails instructions as for preparations to be made prior the test (e.g. nutritional instructions and not to bring metallic objects), as well as a notification that late arrival will result in canceling appointment. 6. In the day of the appointment the patient presents in the MRI unit and meets the responsible nurse to prepare him for the procedure. After completing the preparations and the scan, the patient is requested to return to the reception desk, where a paper with the date of examination is provided. 7. Examination results are usually provided within two weeks from the day of the scan. The obvious result of this rather lengthy process is deterioration in patient satisfaction and possible exposure to risks. The main underpinning of low patient satisfaction is the long waiting time, during which patients are left without definitive diagnosis and may suffer pain and/or symptoms of an undiagnosed condition. Furthermore, patient safety may be a stake here because late diagnosis may lead to complications and lack of timely treatment, particularly in the case of tumors and nervous system pathologies (Mackenzie et al. 211).
Defining Specific Needs for Change
The work of the MRI Unit can and should be divided into two major categories, each one with different settings, stakeholders, strengths and weaknesses. These are (1) the service process, and (2) the manufacturing of clinical imaging, namely the production and interpretation of clinically valuable findings in the patients’ bodies. Nevertheless, the segmentation into two distinct processes does not imply that two are not heavily interconnected, nor do they are constrained by two distinct sets of conditions. In fact, external constrains such as the extent of demand for service and limited budget affect both aspects of the process and should be accounted for when designing work processes.
The MRI Unit’s service operations can be defined as the set of interactions between the unit and its stakeholders, which are basically the patients (either inpatients of outpatients) and their families, the hospital staff, practitioners outside of the hospital (such as GPs and specialists) and the Kuwaiti healthcare system. All of these are obviously involved to some extent in the unit’s production processes (albeit in different aspects of it), but their main interest is the ‘bottom line,’ namely receiving cost-effective, timely and accurate results, while maintaining smooth and clear interactions with the unit’s staff.
Based on this broad definition, it is possible to describe the transformation process of the service operations in terms of using non-medical resources (most notably administrative staff and overheads) to allow for a smooth flow of patients from the initial referral (or more accurately from the first contact with the unit) to the delivery of results. Main improvements in efficiency thus imply either (1) shorter interaction time with patients (since handing patients is the main time- and resource-consuming aspect of the unit), or (2) increased number of patients handled per resource per unit of time. Ideally, both of these two goals must be improved significantly, as they are very strongly interdependent.
When coming to assess the specific elements in the process that are to be improved, this paper takes three factors as given and generally not malleable. First, the scanner time is determined, so more scanning hours are not possible (specific recommendations to better utilization of scanner time are discussed later in this paper). Second, there cannot be any increase in human and/or capital resources; that is, the size of the personnel must remain constant, there cannot be any material increase in equipment, and there is no possibility to send overflows to external service providers (namely other MRI units in Kuwait City). Third, demand for the service is expected to grow, and it is not unlikely that the clinical requirements before the examination (e.g. blood tests) will become more detailed, lengthy and complex.
Quantitative measurement of the dynamics of service operations reveals that the most variable (and therefore unforeseen) element in the delivery process is the presentation of new patients in the reception. These are the patients that have been recently referred by their GPs or specialists (i.e. outpatients) and urgent inpatient cases from other departments of the hospital. The reception staff’s inability to control the flows of patients in the various stages that precede the scan causes several major flaws in the service transformation process:
In addition, the interaction with clinical stakeholders is a very serious constraint on the effectiveness of the administrative staff. Mounting evidence suggest that information flows within MRI units and between units and other suppliers of care (such as ambulatory clinics and other departments at the hospital are the weakest and most prone to variably process (Cavagna et al. 205). Al-Amiri’s MRI Unit is yet another classical example for this flew.
Several sub-processes here require substantial improvement:
The final product of the MRI Unit, namely an MR-based assessment of pathologies, is the result of a sequence of activities taken place in the MR suite. These activities involve (counterclockwise from bottom left):
These tasks are conducted by the radiology staff, which includes radiologists, nurses and MR technicians. Moreover, the staff is involved in several other tasks, the most problematic of which, as described earlier, is the paper-based review and signature of a radiologist on the so-called ‘MRI approval form.’ This is a simple form that includes (1)the type of examination to be contacted (0.7 T or 1.5 T), (2) body part as checked by receptionist, and whether there are (3)previous examinations, (4)normal creatinine level, (5)history of allergy, or of (6)bronchial asthma. Discussions with the radiology stuff clearly indicates that this form, which may take several valuable minutes to fill out (depends on the availability of all relevant data in the patient’s file) is a strictly unnecessary and time-consuming tasks for the radiologist in duty.
There are more than a few reasons for that. Some common arguments include:
In addition to the redundant process described above, quantitative and qualitative review of the scanning schedules reveals inefficient allocation of scanner resources. The current practice in the MRI Unit is to divide each day into slots according to patient specifications. With the exception of Tuesday, in which only pediatric examinations are performed, each day is constructed as follows (for every scanner):
8am-9am: MR breast
It should be noted, however, that exam (scanning) times are highly variable. There exist many factors that may influence scan time, including the type of scanner and software used, the organ under investigation, the type of examination (e.g. T1 vs. T2-weighted imaging), etc. There seems to be no organized way to account for this variability, which often results in delays and/or idle time, as scans may take longer or shorter time than the assigned appointments. The average number of scans per day is 8 (with rather large SD of 1.8) for both scanners.
Principles for Redesigning and Implementing Work Flows
After completing the three initial steps in the six sigma process – namely defining, measuring and analyzing – the rest of this paper is going to deal with improving the work flows in the unit and controlling results. As mentioned earlier, this paper takes capital, physical and clinical constraints as given and tries to tackle the same cost environment by improving specific work processes. Based on the discussion so far, the proposed changes are divided into the two core operations of the unit (namely administration/service and radiology/manufacturing).
Proposed Changes in the Service Operations
In order to enhance both the quality of care and the efficiency of the service, it is vital to reduce the time spent on each patient prior to the examination. As Figure 1 reveals, patients must present at least three times in the unit prior to the day of their appointment. The first time a patient arrives to the unit seems rather redundant and can be eliminated by obliging referring physicians to perform the standard blood tests prior to referring the patient. Furthermore, the referral with the blood results can be emailed or faxed to the unit directly from the physician office, thereby eliminating the second presentation to the unit (t=4). Similar allocation of responsibilities was tried and successfully implemented in CT settings, resulting in increased throughput and decreased outpatient-scheduling backlog (Loree et al. 49).
This basically means that the radiologist approval can be taken without the actual presence of the patients in the unit, avoiding the long queues. Finally, patients will be contacted by phone to set the appointment and may choose to come to the unit to collect their information sheet or to receive it electronically. Less patient handling time will also allow the administrative staff to have more time for transcribing reports, shortening the delay between scans and results.
Proposed Changes in the Manufacturing Operations
The two most scant resources are scanner time and radiologists. Eliminating the time spent on reviewing requests can easily save radiologists’ time. A radiologist can review special cases, but evidence shows that these cases are rather infrequent. More time can be therefore allocated for interpretation of imaging data, thereby allowing higher capacities and potentially reduce diagnostic errors (Halsted et al. 331).
Finally, more strict separation between outpatients and inpatients, as well as between the open-bore and closed-bore systems, may help to enhance the flow of patients and reduce buffer times. This is mainly because the types of patients and examinations have smaller variability within-groups than between-groups.
The proposed redesigned schedule is as follows:
Close-bore scanner: Sunday-Thursday (except Tuesday): 8am-10am: Outpatients 10am-11am: MR breast 11am-1pm: Inpatients Tuesday: Pediatric day
Open-bore scanner: Sunday to Thursday: 8am-12am: Outpatients 12am-1pm: Inpatients
This may allow for a significant increase in cases per day of about 4 cases (50% increase). Recommended cases may take the empty slots, especially in the variably scheduled MR breast slots. Overtime may be also possible, albeit in a limited array of examinations due to lack of technicians.