Can a Ris/Pacs System Really Improve Execution and Reporting Times? ()
A study of the process timing and organization has been done in a Radiology Department where Radiology Information System (RIS) and Picture Archiving and Communication System (PACS) are working. The workflow times have been directly read from RIS database. In particular, Execution and Reporting steps were considered. Afterwards, data have been analyzed through a statistical elaboration and results were compared with data, coming from studies performed on similar process, collected with a stopwatch method. Possible “bottlenecks” can be evidenced in the process structure. Moreover, comparing hospitals with the same RIS/PACS system, different internal organization is emphasized.
1. Background
Many studies have showed the positive impact of the introduction of Radiology Information System (RIS) and Picture Archiving and Communication System (PACS) on the work organization of a Diagnostic Imaging Department. Kotter and Langer present results about the introduction of a RIS/PACS in a German Hospital [1]. Filmless and paperless operations are typical of a RIS/ PACS system. In fact, manually intensive tasks are reduced or eliminated. Reiner, Siegel et al., discuss the general situation and possible strategies for management optimization [2,3]. New working practices, following the introduction of RIS/PACS in an hospital, are discusssed by Shaw [4].
Workflow studies evidenced a significant reduction of the number of steps, needing to complete the process, after PACS utilization [2].
We collected data from Radiology Department of Macerata Hospital, equipped with a RIS/PACS system. RIS database was interrogated to measure process times for Execution and Reporting steps, two main parts of the whole image acquisition procedure, useful to analyze the workflow.
Results were compared with similar time data, manu ally measured and presented in a previous study [5]. In this work, Mariani, Tronchi, et al., cut the whole process in a number of steps with their elementary operations. Afterwards, they measured using a stopwatch method, the interval time needing for each elementary operation. The comparison of the data collected with two different methods can support a bottleneck analysis. Finally, we analyze differences in the organization of two other hospitals, but working with the same RIS/PACS digital system installed in Macerata Hospital.
2. Methods
A RIS/PACS system is working in the Radiology Department of Macerata Hospital since January 2004. This system offers several advantages, linked to the implementation of a speech recognition reporting technology, too.
RIS was interrogated to get a file containing data of about 33.000 examinations, performed from January 2005 to December 2005. The ethics committee of ASUR-Marche Z.T. 9 examined, before commencement of the study, the research proposal and the actual activity plan.
We considered only examinations on external patients and three exams: X-Rays (XR), Computed Tomography (CT) and Magnetic Resonance Imaging (MRI). Data were checked for anomalies and wrong values.
RIS records three times:
1) end of patient registration at Admittance Front Office;
2) end of the examination;
3) signature of the report.
The time interval from the end of the registration to the end of examination has been defined as T1’, Figure 1.
In this same figure, we define T2’ as time, running from the end of the examination to when film and report were available for delivery (Report Turnaround Time). A statistical elaboration was performed for collected times, computing their mean values with standard errors.
3. Results
Time frequency distribution of raw T1’ data for XR, is presented in Figure 2.
As one can see, about 45 % of the examinations is executed in 20 m or less, while about only 7 % takes a time ranging from 60 to 120 m. Similar histograms can be represented for CT and MRI: about 56 % of examinations is performed within 40 m for CT, as one can see in Figure 3, while the value decreases to about 42 % for MRI. All CT examinations are performed at maximum in 200 m, while the time increases to 240 m for MRI.
Figure 4 presents the time frequency distribution of raw T2’ data referring to XR: as one can see, about 49 % of the examinations is reported in 20 m or less and only about 10% takes a Reporting time longer than 140 m.