| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Review |
1 Department of Radiology, Massachusetts General Hospital and Harvard Medical School, 55 Fruit St., White 270, Boston, MA 02114.
Received September 25, 2007;
accepted after revision January 25, 2008.
Address correspondence to G. W. L. Boland
(gboland{at}partners.org).
Abstract
 Top Abstract IntroductionDiscussionCharting the Workflow...Increasing CT Capacity Through...Financial ImplicationsSummaryReferences |
|---|
CONCLUSION. Many organizations are challenged to meet stakeholder demands of providing additional CT capacity and reduction of patient waiting lists. However, much can be achieved through workflow redesign, the addition of key personnel, and implementation of information system platforms and databases.
Keywords: CT • efficiency • information systems • operations management • productivity • reimbursement • stakeholders • technologists • workflow
|
TopAbstractIntroductionDiscussionCharting the Workflow...Increasing CT Capacity Through...Financial ImplicationsSummaryReferences |
|---|
Meanwhile, now that outpatient CT has become highly profitable for hospitals, administrators and many radiologists are recognizing the importance of retaining their customer base [11, 12, 14]. Business interests aside, most care providers would also argue that it is important to scan patients within network, enabling the radiology report and images to be available on a common information technology (IT) platform and for on-site radiologists and referring physicians to readily discuss the findings, if necessary [11, 15]. On the inpatient side, hospitals are trying to cut costs due to pressure from third-party payers and are requiring inpatients to be scanned as quickly as possible to expedite patient care [11].
The overall impact of these pressures is that key radiology stakeholders (patients, referring physicians and administrators, and many radiologists) are placing increasing demands and expectations on radiology departments and practices to reduce both inpatient and outpatient waiting lists and increase capacity [11]. This has become particularly critical for CT services, which have now become central to the investigation and diagnosis of many disease processes [11, 16].
One solution to adding more capacity (and therefore improving patient access) would be to add more equipment. This approach, however, not only is expensive but also may be unnecessary because increased imaging capacity can often be obtained through workflow redesign with only a marginal addition al cost in resources [17, 18]. In the hope of improving service and financial state ments, radiology departments have theref ore attempted various strategies and tactics to improve CT productivity. Most of these strategies do not require the purchase of additional CT equipment, particularly if the organization already has an MDCT-installed base [11, 12]. This article will therefore identify those strategies and tactics that can be used to enhance both inpatient and outpatient CT productivity. It will also discuss the potential benefits to radiology stakeholders (patients, referring physicians and administrators, and radiologists) including those who must monitor the financial impact to the organization.
|
TopAbstractIntroductionDiscussionCharting the Workflow...Increasing CT Capacity Through...Financial ImplicationsSummaryReferences |
|---|
Assuming that no additional equipment is purchased, CT capacity can be increased by either scanning more patients per hour, scanning for longer hours, or both. Although the concept of these metrics is simple, the methods that can be used to increase capacity are more complex. Ideally tactics that address both these metrics are likely to have the biggest impact on patient throughput in CT, but for the purposes of this article, they will be discussed separately. In any event, improving performance on either metric requires change to either the number or type of personnel, the way these people work, or the technology and information systems they are using. Each will be discussed in turn. Before any changes can be made, however, a comprehensive analysis of the existing workflow should be performed.
Charting the Workflow (Flowcharting)
|
TopAbstractIntroductionDiscussionCharting the Workflow...Increasing CT Capacity Through...Financial ImplicationsSummaryReferences |
|---|
The initial goal of this process is to identify each unique step in the workflow process, which is best portrayed as a flowchart (also known as a workflow diagram or process map) because it visualizes an overview of the operation. Each box within the flowchart represents a unique step or process within the operation, sometimes minor, sometimes critical, but each step is deemed necessary to process a patient through CT. Some of these steps may subsequently be considered redundant (an outcome of the operations management improvement process), but typically each organization has a workflow that has evolved idiosyncratically over time depending on the equipment, staffing levels, technologic infrastructure, and needs of the organization. Usually no two workflows from different organizations are alike. Part of the process of operations improvement is to identify best practices in the industry and incorporate them into the workflow. For this reason, some organizations prefer to outsource the flowcharting process to consultant groups, given that such groups can benchmark the host organization against other organizational flowcharts. Furthermore, the development of a comprehensive flowchart can be complex, time consuming, and involve multiple personnel from the organization, who may have little experience of the process.
The workflow diagram should outline the work process from beginning to end by interviewing every individual who interfaces with that workflow: patients, referring physicians, schedulers, reception staff, technologists, IT personnel, and others [19]. Once the diagram is constructed, a CT operations management team, composed of key individuals (technologists, radiologists, IT personnel, radiology schedulers), performs a gap analysis to identify failure points that provide opportunities for improvement. Further analysis may be required within each "box" (i.e., the unique step that has been identified). Such analysis may include chronologic data (the time it takes a transporter to deliver a patient to the CT suite from the inpatient floor), the number and type of personnel (to determine if the staffing ratio to meet demand is sufficient), the technologic infrastructure (do the radiology information system [RIS] work lists automatically download onto the CT console workstations?), and financial data (are the employees cost-effective?). Changes made to the CT operations all refer back to this flowchart, as discussed next.
Once the workflow has been outlined in a flowchart, the operations management team can evaluate areas of opportunity to increase CT throughput. There are, however, three critical tactics necessary to increase inpatient and outpatient CT productivity: addition of technologists; redefining the workflow of these technologists (and other essential personnel in the CT operation); and enhancement and integration of existing information systems (RIS and PACS) as well as the use of simple databases to manage the CT workflow more productively. Implementation of these three tactics is likely to have the greatest impact on improving CT throughput and productivity for most organizations. Each will be discussed in turn.
Addition of Technologists
Technologists must perform a lengthy series of tasks to process a patient
through the CT scanner, starting from the moment the patient has been
delivered to them, either by the reception staff (for outpatients) or the
transporters (for inpatients). For instance, it has been shown that a
technologist must perform 34 unique tasks to process a patient through the CT
scanner [17]. A single
technologist or individual has to perform the majority of these 34 tasks one
after another (in series) rather than at the same time (in parallel). This can
have a profound negative effect on patient throughput. For instance, it has
been shown that it can take a single technologist an average of 27 minutes to
perform all of these tasks successfully before being able to move onto the
next patient [17]. Therefore,
a single technologist can generally only scan an average of approximately two
to three patients per hour.
On the other hand, adding personnel considerably expedited the time it took for a patient to be processed through the same scanner because many of the 34 technologist tasks could be performed in parallel [17]. It has therefore been suggested that CT scanners should be operated with either two or three technologists, depending on the time of day. When both inpatient and outpatient scanners are at their busiest, usually between 10:00 am and 4:00 pm, three technologists are used to accommodate the frequent "add-on" patients sent from referring physician's offices or from the hospital wards. Otherwise, the workflow is well managed with two technologists.
It could be argued that a radiology nurse should be responsible for placing the IV catheter and that technologist helpers assist in the workflow. However, it has been recommended that technologists are more suitable, given they can also perform the multiple other tasks on this work list. This concept of technologist flexibility is essential to improving productivity because any technologist can perform any of the tasks, and the workflow then becomes free of its previous dependence on any single individual. In this manner, the technologists can seamlessly switch to any of the tasks at any time, which minimizes time delays between the different tasks. However, given that different organizations have different needs (IV cannulation is rarely performed by CT technologists in Europe, for instance), it is less important that technologists are used exclusively. However, it is essential that multiple personnel manage the workflow rather than the single-technologist model because the multiple-personnel model provides the opportunity to scan up to 20,000 additional patients annually [17].
Reframing the Workflow Process
As a result of the flowcharting procedure, it should be possible to
identify multiple opportunities for process improvement, which will differ
from institution to institution and will depend on whether the patient is an
inpatient or outpatient. For instance, one study showed that 13 areas for
improvement on the outpatient side of the operation and eight areas on the
inpatient side were necessary as part of the improvement process
[20]. A synopsis of a
flowchart for two CT technologists is shown in
Figure 1. Many of these
initiatives are common sense, but they may not become evident until an
analysis of the flowchart has been performed
[21]. Specific examples
relevant to most organizations are included below.
|
Decision support indications also correspond to the correct International Classification of Diseases, 9th rev. [24], which is necessary to bill appropriately for the requested examination. This minimizes, if not eliminates, the proportion of requested examinations that are considered inappropriate and for which payment would usually be denied by third-party payers [23].
Relocating preparatory procedures—A rel atively simple but highly effective change is to insert the IV catheters into patients in a holding area adjacent to the CT suite rather than waiting until the patient is on the CT table (Fig. 1). Once the patient has been delivered to the technologist and is ready for the examination, IV placement can take the greatest proportion of time. It is therefore far more productive to scan a different patient while the IV is being placed into another patient outside the room. In addition, on the in patient side of the operation, oral contrast media should be made available on the patient floor (and administered to the patients in advance of the scan) so patients can be scanned immediately on arrival to the CT suite.
Patient and radiologist reminders—Reception staff should telephone outpatients the day before their scans, not only to confirm their appointments but also to remind them to arrive on time. At imaging center facilities, the reception staff should also telephone the covering radiologist to confirm timely arrival for the following day's schedule. A CT workflow is significantly disrupted if the covering radiologist fails to arrive on time because no IV contrast studies can be performed.
Timely patient transport—As part of operations improvement, a dedicated radiology transporter can be hired who is trained to understand the importance of timely patient arrivals to the CT suite. In turn, the staff on the patient floors can be educated to ensure that the patient is ready for when the dedicated radiology transporter arrives. A telephone call from radiology to the inpatient floor 30 minutes before the scheduled arrival of the patient transporter helps to ensure that the patient is ready on time.
Restocking materials—Material should be restocked after hours rather than during the day, which otherwise takes up valuable time for a technologist and reduces the time the room is available for scanning patients.
Professional development—Additional procedural changes include further training for technologists on how to use information systems more efficiently [25–28]. Technologists can also be trained by radiologists to identify a limited number of easily recognizable, relatively common, CT findings that might require additional imaging for lesion characterization (i.e., liver lesions requiring delayed CT). This enables subsequent imaging to be performed expeditiously, rather than having to find a radiologist to request further imaging. The radiologist's advice should, in general, only be requested for unusual or more difficult cases. The goal is to remove radiologists from the scanning process itself as much as possible so as not to slow down patient throughput. On the other hand, radiologists should be relatively close at hand, either to quickly answer specific technologist or patient questions or for rapid response in the event of a patient reaction to IV contrast material.
Furthermore, it is also important that radiologists devise and implement department-wide standardized and succinct disease-specific CT protocols (including those that require reformations) that are adhered to by all radiologists and technologists in the organization. Otherwise, patient throughput is inevitably delayed when technologists attempt to verify whether a particular protocol for a given clinical indication is acceptable to the covering radiologist who might not approve another colleague's idiosyncratic protocol.
Although some of the aforementioned procedural changes might be seen as simplistic, it is important for each department to critically evaluate each failure point in the CT workflow process and develop solutions appropriate to the institution. These changes will expedite patient throughput. Solutions with the greatest impact on CT throughput will likely include those that address planning ahead and improving background documentation and professional development. Not all the recommended solutions, however, will have equal impact on improving CT throughput and productivity, but it is still important that even those with less immediate and observable impact be incorporated into the new workflow.
Integrated RIS and PACS
Although the technical details of the advantages and disadvantages of
different in formation systems relevant to radiology workflow productivity are
beyond the scope of this discussion, an RIS is critical in evaluating,
implementing, and measuring productivity. Many, if not most, organizations
have already implemented an RIS but may not be using the technology to full
advantage. Even before analyzing the productivity of a CT operation, it is
important to obtain data pertinent to the operation, usually available on the
RIS [27,
29]. Such data will include
examination length and type and hourly, daily, weekly, monthly, and annual
throughput. If they are to be efficient, technologists need to have accurate
and seamless electronic access to the CT schedule to appropriately plan
studies (ideally the day before), and they need access to the patient
demographic data. Ideally the RIS-scheduled work list should be integrated
electronically with the CT console workstation, without the need for further
patient demographic data entry.
With these systems in place, technologists can then choose the correct patient from this work list when the patient arrives at the CT scanner. They can then input additional data relevant to the examination, including any adverse events. The goal is to remove paper as the primary driver of the workflow process, because paper is inefficient, sometimes inaccurate, and not easily available to all relevant personnel [29–31]. Furthermore, without an effective RIS, patients cannot be efficiently scheduled, and the daily work lists will be difficult to retrieve for reception staff and technologists alike. Ideally the RIS should be integrated into the hospital information system (HIS) for further ease and accuracy of scheduling.
It has also been shown that technologist productivity is enhanced through the use of a PACS [27, 31–34]. Technologists no longer need to print and separate the film into separate jackets for different subspecialty radiologists. In some institutions, it has been the responsibility of the technologist to deliver the films to the film library or radiologist and sometimes even to hang films on alternators. Therefore, PACS offers technologists, not just radiologists, clear operational advantages in managing workflow productively.
Increasing CT Capacity Through Changing the CT Schedule: Use of Simple Databases
|
TopAbstractIntroductionDiscussionCharting the Workflow...Increasing CT Capacity Through...Financial ImplicationsSummaryReferences |
|---|
A simple computerized spreadsheet can help managers determine the best combination of scheduled slot time and hours of operation [35]. A sensible approach to schedule change should reflect the existing nonurgent patient waiting list, or backlog, and the time in days or weeks before the next available appointment. If it is known how long it is until the next available nonurgent appointment, the spreadsheet can be used to determine the number of days it would take to reduce or minimize this backlog by reducing the time of the examination slots or increasing the hours of operation.
Consider the following scenario: a CT schedule operates for 11 hours per day with 20-minute scheduled time slots for 5 days per week. This schedule permits a total of 33 slots per day or 165 per week (Table 1). However, the delay until the next available, nonurgent, appointment for this department is 4 weeks or 20 operating days because the department does not operate on the weekend or evenings. This equates to a backlog of 660 waiting patients (165 patients x 20 days) (Table 2). From the spreadsheet, it can be seen that using this existing scheduling template, the waiting list will never be reduced, particularly as demand continues and new patients are added to the waiting list. Under circumstances such as this one, referring physicians naturally become frustrated and may begin to refer outpatients outside of the network.
|
|
However, if the examination length was reduced by 5 minutes to a total time of 15 minutes, it can be seen from the spreadsheet that an additional 11 slots per day would be generated, resulting in 55 additional examination slots per week (Table 3). From the spreadsheet, it can be calculated that it would now take 60 operating days (or 12 weeks) to bring the time to the next available, nonurgent, appointment down to zero days. Alternatively, if one were to increase the number of operating hours in the evening by 3 hours and open for 10 hours on Saturday (still using a 20-minute examination slot), then an additional 75 examination slots per week become available (Table 4). Finally, combining the new 15-minute examination slot time with the weekly additional 25 hours of operating time will yield 280 additional examination slots per week (Table 5). The waiting list would then take just over 3 weeks to clear. If the examination time was further reduced to 10 minutes, then 480 additional slots per week become available, reducing the time to the next available appointment to about 1.5 weeks. The opportunity for increased patient throughput is even higher if the schedule is open into the late evening hours (say 10:00 pm) and on both weekend days.
|
|
|
The goal is to create enough flexibility in the CT schedule so that referring physicians know that an urgent add-on patient (both for inpatients and outpatients) can generally be accepted (or that routine scans are obtained in a timely manner). Otherwise, for outpatients, referring physicians may decide to send their patients to a competing institution. Once a referring physician believes that patients are better served elsewhere, then it may be very hard to convince that physician to refer patients back within the network.
Although the examples described in the previous paragraph are simplistic, they do permit a radiology manager to understand the dynamics between the size of the waiting list and the opportunity for reduction through manipulation of the CT schedule. Fundamentally, it is through reduction in the length of CT scheduled slots combined with longer operational hours that organizations are able to reduce both inpatient and outpatient waiting lists and meet many of the stakeholders' expectations. If these adjustments are made, it is highly likely that referring physicians would consider that the radiology department offers improved value for their inpatients and outpatients and would likely continue to refer both urgent and nonurgent patients to that facility, given that they can be accommodated with relative ease.
Naturally, the productivity gains will depend on whether the scanner is used for outpatient purposes only, for mixed inpatient and outpatient services, or uniquely for inpatient services. Indeed, inpatients are continuously disruptive to outpatient imaging if both are scanned using the same equipment [36]. Usually, inpatients and outpatients must compete with each other for the limited number of available scheduled appointments, with outpatient appointments often delayed because of the unpredictable nature of the inpatient workflow. It has therefore been suggested that inpatient and outpatient CT services are effectively different businesses and should be operated as such [36]. This may require the allocation or purchase of a dedicated outpatient scanner (ideally located outside of the hospital) to take full advantage of the potential productivity gains possible on dedicated outpatient scanners. With such a strategy, a hospital can then redirect most outpatients away from inpatient scanners, which will now have the additional capacity to scan inpatients expeditiously, a key stakeholder expectation [36].
However, although the opportunity for CT productivity gain through reengineering the workflow is likely to be greatest for outpatient rather than inpatient scanners, much can still be achieved on inpatient (or mixed inpatient and outpatient) scanners. Indeed, most strategies recommended in this article are applicable to both inpatient and outpatient facilities. The most important strategies affecting both operations will include staffing the CT scanner to optimal levels, reducing the scheduled slot times, extending the hours of operation, and fully integrating information systems into the workflow.
|
TopAbstractIntroductionDiscussionCharting the Workflow...Increasing CT Capacity Through...Financial ImplicationsSummaryReferences |
|---|
Given this potential for scheduled slot elasticity, the financial implications now become more evident and the potential additional revenue can run into millions of dollars. For instance, it has been shown that a typical scanner can scan approximately 12,000 patients per year using a single-technologist model, compared with 20,000 or 30,000 patients when using two or three technologists, respectively [17]. Assuming an average technical reimbursement of $300 per CT examination (the average technical Medicare reimbursement rate for a CT examination in many U.S. states), a single-technologist model can generate, at maximal productivity, approximately $3.6 million per year whereas the two- and three-technologist models can generate up to $6 and $9 million, respectively (which could also be considered as the cost of several new MDCT scanners) [17].
The operating costs to generate these differences are marginal in comparison. Although the cost of the equipment and room space are fixed and independent of the operating hours or scanner productivity, the remaining costs are variable: technologist staffing, reception and physician coverage, contrast media, lighting, heating and film (if there is no PACS), among others. The major variable cost incurred when moving to a multitechnologist model is staff salaries, but this is still relatively minor compared with the potential revenue that can now be generated. Depending on the number of additional technologists employed, the overall increase in annual technologist salaries (or other personnel) per scanner is between $120,000 and 280,000 [17]. The overall variable cost per scan is then ap proximately $50–60 (inclusive of technologist salaries), so the overall costs for performing the additional scans with the two- and three-technologist model are approximately $400,000 and $900,000, respectively, far less than the $2.4 and $5.4 million in incremental revenue these respective models can generate. Given that many patients receive scans of more than one body part during their CT visit, the margin will be even higher if the revenue is based on examination rather than patient numbers. Despite the recent enactment of the Deficit Reduction Act (whereby the technical payment at outpatient facilities for the second body part scanned at the same time will only be paid at a reduced rate compared with the first body part), the costs to operate a scanner more efficiently are marginal compared with the revenue [38]. It is therefore imprudent to consider it too costly to operate a CT scanner more efficiently. Furthermore, it does not make sense to purchase additional equipment as a means of increasing CT capacity unless one has already maximized the potential of existing CT assets.
|
TopAbstractIntroductionDiscussionCharting the Workflow...Increasing CT Capacity Through...Financial ImplicationsSummaryReferences |
|---|
|
TopAbstractIntroductionDiscussionCharting the Workflow...Increasing CT Capacity Through...Financial ImplicationsSummaryReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
