Computers in Radiology
Internet-Based Radiology Order-Entry, Reporting, and Workflow Management System for Coordinating Urgent Study Requests During Off-Hours
Audio Available | 
Share
OBJECTIVE. Our aim was to develop a simple, low-cost, Internet-based application for radiology order-entry, reporting, and workflow management during off-hours.
CONCLUSION. The system was quickly accepted by users both within and outside the radiology department, and it required very modest resources to develop, deploy, and support. In a busy on-call setting at a high-volume academic institution, the system described was effective in obtaining more thorough patient histories from referring physicians, reducing the number of telephone calls required, and documenting more rigorously the communication between radiologists and clinical services. These benefits allow the generation of more informative and timely radiology reports.
| Introduction |   |
|---|
The on-call radiologist in a busy academic environment is charged with the difficult task of coordinating and reporting multitechnique radiology studies while optimizing limited radiology resources. This job includes taking requests from clinicians; obtaining relevant clinical history and study-specific data; communicating that information, including special instructions, to technologists; and finally distributing study results to all clinical services in charge of a given patient. Even after a study is interpreted and reported to the ordering physician, other clinical services may call to review the findings directly with the radiologist. This process can be time-consuming, inefficient, and frustrating for all parties involved.
Clinical histories provided either on paper requisitions or even on computer order-entry systems are often inadequate. For a variety of reasons, histories frequently contain one or two words, such as “abdominal pain,” even though a more thorough medical and surgical history may help the radiologist provide a more meaningful report. Requisitions that are misplaced or lost add further delay to the reporting process.
During off-hours, preliminary reports are often given verbally and poorly documented. This practice has risk-management implications because clinical decisions are often made on the basis of the preliminary report alone. Attending radiologists who review a case in the morning need a clear understanding of what information was communicated to referring clinicians overnight.
We describe an Internet-based application called “3RADS,” named after the radiology on-call pager number at our institution, that serves as an order-entry, reporting, and workflow management system. Advantages include custom tailoring to workflow, ease of development and support, low cost, and hospitalwide availability.
| Materials and Methods | |
|---|
Minimum hardware requirements are an Intel-compatible computer with a 90-MHz processor, 32 megabytes of random access memory, and a static Internet address. Data storage and Web-serving capabilities were provided by a FileMaker Pro 6 (FileMaker, Inc.) database package costing approximately $1,100. A relational database was created with multiple tables that store study information, preliminary reports, and the names and pager numbers of radiologists on call for any given technique. Data entry and retrieval were programmed using FileMaker-specific CDML instructions embedded in the HTML code controlling Web layout. The system could be accessed using any Web browser.
User authentication is handled through a Java applet embedded in the main Web page that checks a user name and password against a list of valid users in the institution's electronic patient record. If there is a match, access to 3RADS is granted. Separate login screens and passwords are provided for radiologists and technologists to distinguish their roles from the clinicians.
This system was developed in-house by a radiology resident with a moderate programming background but no previous experience with FileMaker. Hospitalwide radiology workflow was comprehensively analyzed during system development.
Three different parties use the system, all of whom direct their Internet browser to the same Web site. After a clinician is identified as a valid user, he is offered three options (Fig. 1). The first option searches the database and provides a list of the names and pager numbers for the radiologists on call for any technique. The second option allows the clinician to make a request for a radiology study and eliminates the need for a paper requisition. Ordering guidelines are provided on the basis of the type of study requested (Fig. 2). For a CT study to evaluate suspected pulmonary embolism, for example, the clinician is made aware that central lines cannot be used for contrast injection and that peripheral venous access in the upper extremity is necessary. The system then prompts the ordering clinician to enter unique information based on the study type, such as the patient's serum creatinine value for a CT scan with IV contrast agent. After the clinician submits the required information, an alphanumeric page is sent to the radiologist on call for that technique. The last of the three options provided to the clinician allows him to track the progress of any study and to obtain a preliminary report when one is available. Study tracking can be performed only with knowledge of the patient's medical record number. For American Health Insurance Portability and Accountability Act of 1996 compliance, 3RADS logs the clinician's name and the date every time a preliminary report is viewed. With this data, we calculated the frequency with which clinicians requested studies and viewed reports on 3RADS.
 View larger version (60K) | Fig. 1. —Screen shot of main 3RADS screen provided to clinician after login authentication. |
 View larger version (60K) | Fig. 2. —Screen shot of ordering guidelines for CT. |
The radiologist who accesses the system can view studies according to status and technique (Fig. 3). After review of the information provided and any necessary discussion with the clinician, the radiologist may change the status of a study from “Pending Approval” to an appropriate alternative, including “Approved,” “Preliminary Report Complete,” “Awaiting Physician Call,” or “Cancelled.” She can mark a study as urgent and provide special instructions on study protocol to the technologist. Finally, the radiologist can enter a preliminary report that can be accessed by clinicians and by the attending radiologist who is reviewing the cases in the morning.
 View larger version (94K) | Fig. 3. —Screen shot of sample radiologist work list, currently set to view all studies that have been approved but do not yet have preliminary reports. |
The technologist accessing 3RADS has a technique-specific work list of studies to perform and can prioritize studies on the basis of urgency as directed by the radiologist. When the technologist marks the study complete, it appears highlighted on the radiologist's work list, indicating that it is ready for interpretation.
| Results | |
|---|
3RADS was launched hospitalwide on May 9, 2003, for sonography requests only. Requests were accepted during the night radiology resident's shift, between 5:00 pm and 7:30 am. Preliminary reports could be viewed at any time of day. On July 4, 2003, CT requests were made available. Although requests for studies of other techniques (e.g., MRI or nuclear medicine) could not be made on the system, the person to contact for any technique was posted on 3RADS and diligently updated. Initially, clinicians still could obtain studies by calling the resident personally. About 3 weeks into use, any clinician calling to request CT or sonography was asked to use 3RADS.
Figure 4 shows the average number of CT and sonographic studies requested per day during on-call hours. Figure 5 shows the number of times preliminary reports were viewed per study. Instead of requiring multiple telephone calls between clinician, radiologist, and technologist, a typical study performed overnight that has insignificant findings now requires no telephone calls. Significant findings affecting patient treatment are still communicated verbally to the requesting clinician, and this communication is documented in the online preliminary report.
 View larger version (27K) | Fig. 4. —Graph shows average number of CT and sonographic studies requested per day from 5:00 pm to 7:30 am between May 9, 2003, and January 31, 2004. 3RADS system was not available for CT requests before July 4, 2003. |
 View larger version (30K) | Fig. 5. —Graph shows number of times preliminary reports were viewed per study between May 9, 2003, and January 31, 2004. 3RADS system was not available for CT requests before July 4, 2003. |
Both radiologists and technologists report increased satisfaction with the history provided with each study. A median of 14 words per request was provided in the “Symptoms,” “Medical/Surgical History,” and “Suspected Diagnosis” fields from May 2003 to January 2004. For these three fields, ordering physicians were given specific online examples of the kind of information expected. Furthermore, unlike paper requisitions that allocate only a small space for clinicians to include this type of history, the space provided in 3RADS allows much more information to be communicated to the radiologist. A sample history provided by a clinician is shown in Figure 6.
 View larger version (16K) | Fig. 6. —Screen shot shows sample clinical history provided by referring physician in required fields. |
| Discussion | |
|---|
3RADS has been an effective tool for communication between the on-call radiologists, clinicians, and technologists. Before implementation, one inefficiency commonly cited by radiologists was having to review the same case with multiple clinicians at different times. In January 2004 (after implementation), the radiologist received an average of 35 study requests per night shift, and clinicians viewed preliminary reports an average of 63 times per day. If only one-half of the viewed preliminary reports averted a telephone call in that month, this conservative estimate suggests that the night radiologist had 66 (35 + 63/2) fewer pages to answer or calls to make to discuss findings. This estimate does not include saved calls to technologists. If the time spent in returning a page and discussing a case averaged between 1–2 min, the overall time saving was 1–2 hr per day.
Additional benefits include tight communication and control. Technique-specific information is provided to the clinician before a study request. This is a form of medical-decision support cited in similar projects [1]. This feature has reduced the number of patients who are returned to the floor from the radiology department because they are unprepared for the examination. An alphanumeric page sent to the on-call radiologist after submission of a study request reduces delays in approval of the study. The ability of the radiologist to provide special instructions to the technologist ensures that the proper study is completed in a timely manner and that the appropriate questions are answered. This enhanced communication gives the radiologist more time to interpret studies with fewer interruptions, improving overall clinical service.
The concepts of physician order entry, study tracking, and electronic reporting are not new. Similar projects have shown improved accuracy of information, heightened efficiency of workflow, and higher overall user satisfaction [2]. Although these other projects have included advanced features such as links with the radiology information system (RIS) and PACS that 3RADS does not provide, we have achieved similar end points at a fraction of the cost of other systems. Although our institution's RIS does offer basic order-entry features to technologists, it is not easily adaptable to provide an environment for physician order entry. Our system has been tailored to the unique needs and workflow structure of the on-call radiology service through close collaboration with all parties using 3RADS.
Understanding users' needs contributed to the success of 3RADS, as in similar projects [3]. It was originally believed that the number of required fields would make the system unpopular among clinicians. In fact, most clinicians found that these fields provide a valuable avenue of communication with the radiologist on-call, to the extent of providing a median of 14 words of clinical history per request, significantly higher than the brief histories provided on paper requisitions or some clerical systems. The extent of the history provided can be attributed to the ease of the user interface, the provision of examples of what information is expected in each field, the visibility of the history to all users, and the knowledge by the ordering physician that an adequate history would eliminate the need for further information exchange by telephone with the radiologist before the study. Furthermore, clinicians themselves, and not clerks who may lack sufficient medical training, are entering the data, thus improving accuracy.
Documentation of preliminary reports in 3RADS formalizes a previously informal process of communication. Unlike systems in which preliminary reports are entered directly into the hospital electronic patient record and later overwritten by the attending radiologist, 3RADS stores the preliminary report separately and permanently. Clinicians find that this documentation protects the medical management decisions they make on the basis of preliminary reports, and attending radiologists have a clear knowledge of the findings conveyed overnight when deciding whether any corrective actions should be taken. This form of documentation also serves as an educational tool for residents who can now more easily and comprehensively compare their reports to those of the attending radiologists.
A key technologic factor that contributed to the success of this project is ease of development, using a FileMaker database on a Windows (Microsoft) platform, and widespread availability on any networked hospital computer with an Internet browser. Another feature is the Java applet that authenticates users against a preexisting institution database of valid users. This obviates creating and managing a separate list of valid users, significantly reducing the effort required to support 3RADS.
The 3RADS project has provided meaningful insights into the needs of different users and the unique workflow environment of on-call radiology. These insights will form an important element in planning for a physician order-entry system currently being implemented at our institution. Features of the system that go beyond order entry, however, will ensure that 3RADS will continue to play an important role in our institution's radiology workflow. Therefore, plans are underway to expand the system to provide full-time ordering of all imaging studies in our department. We anticipate that the expansion will be accompanied by further analysis of and subsequent improvements in clinical operations, both within and outside the radiology department.
Address correspondence to K. Juluru ([email protected],).
| References | |
|---|
