Imaging Equipment Productivity; Benchmarking Device Efficiency
Using DICOM
 
Authors:
Mengqi Hu, Arizona State University; William Pavlicek, PhD; Teresa Wu, PhD; Muhong Zang, PhD; Steven G. Langer, PhD; Vicki Place, RT(R); Rafael Miranda
 
Background:
Acute awareness of the costs associated with medical imaging is nearly ever present with the current debate regarding “sky-rocketing health” care costs. In truth, expensive imaging equipment and costs associated with providing staff support have always been monitored – especially in light of the reduction in reimbursement that has been successfully implemented in recent years. The tools used by medical imaging professionals for monitoring and benchmarking efficiency continue to lag. Logbooks of time-based actions are either hand written or printed copies of patient lists which are normal documents used by administrative staff in assessing their facility. Data envelopment analysis (DEA) is a sophisticated tool for efficiency and is useful, but is quite limited in the details of understanding the changes needed for performance improvements. Observational studies are excellent in extracting valuable information, but can be costly and time consuming. Comparing efficiency evaluations are sensitive to variability in data gathering methods. Benchmarking, as used in many areas involving multiple sets of complex activities, is yet to be agreed upon or established.

We have developed a new tool for automatic and continuous monitoring of exam duration, inter-exam delays, and inter-series delays of all imaging equipments that are using DICOM for storage. The “Imaging Time Monitor” is comprised of a DICOM Parser that receives copies of all performed examinations as they are moved into PACs. This device automatically opens the exam folder, inspects each image folder, deletes the image content, and forwards the DICOM image header information to an Oracle relational database. The schema for the database is designed so equipment is fully identified from the DICOM tags, as are their physical locations. Since every image has been recorded, the exam start time is defined as the time that the first image is acquired, while the exam end time is the time of acquisition of the last image. Inter-scan delays are simply the time interval between last image of one patient and the first image of the next patient. For devices that have multiple series, it is possible to use similar methods to obtain inter-series delay information. The equipment efficiency in a specified period is calculated as the summed total of exam times on this equipment, divided by the total time of scheduled working hours in this period.
 
Evaluation:
Our initial use of the “Imaging Time Monitor” was an evaluation of four CT units: two evaluations were at an outpatient facility while another two were hospital based. The scanner operation is provided with a send function that permits all examinations selected to be forwarded from the patient list. Upon receipt these are parsed in the “Imaging Time Monitor” and the DICOM headers are recorded in the database. DICOM tags are inspected for those four scanners and vendor compliance with the DICOM standard permits the use of approximately 25 tags, including Patient ID, Study Description, Series Description, Equipment (Modality, Manufacturer, Model, and Software version), and Image Acquisition Times. Exam times, inter-scan delays, and inter-series delays were computed as described above. Reporting tools were developed that permitted ad hoc retrieval of scanner use and efficiency for administrative use. Below are sample graphical presentations for Exam Times, Inter Scan Delay Times, and Inter Series Delay Times. “CT54496” and “mcsct2vct” are at an outpatient facility and “CT54426” and “mchgehdct1” are at hospital. The “Imaging Time Monitor” tool groups the Exam times, the Inter Scan Delay times, and the Inter Series Delay times by minutes and shows each group separately using a colored bar.

Figure 1

Figure 2

Figure 3

Between August 26 and 27, 2009, the total working time is 960 minutes, and the summed total of exam times for “CT54496,” “mcsct2vct,” “CT54426,” and “mchgehdct1” is 45.1 minutes, 40.4 minutes, 32.7 minutes, and 93.1 minutes, respectively. An efficiency metric can be computed for any equipment. The efficiency for these four equipments in this example using limited (developmental) test data is 45.1/960=4.7%, 40.4/960=4.2%, 32.7/960=3.4%, and 93.1 /960=9.7%, respectively.
 
Discussion:
The “Imaging Time Monitor” has several important advantages as a measure of efficiency in image acquisition.
  • It automatically records accurate time stamps of exactly when acquisition actually took place.
  • It standardizes the method of exam time duration and inter-patient time for benchmarking.
  • It couples this activity with the patient and the Exam Description so different examinations can be compared for duration. For example, since the CT protocol used is automatically captured, a comparison can be made of the impact of changing protocols, or types of scanners. Vendors’ claims regarding software features which may improve efficiency can be documented. Inter scan delay times are accurately recorded, and disruptions in the serial acquisitions of multiple series examinations are identified.
  • It is inexpensive to collect this data. All data is directed to PACs, collecting a copy is easily implemented.
  • It makes data continuously available. A highly regarded attribute of the “Imaging Time Monitor” is that it generates detailed reports, essentially upon demand, via the intranet by the administrative staff.

This tool provides a non-arbitrary metric of efficiency that is fully standardized across facilities. Its use may facilitate evaluations using operations research and provide insights into behavior changes that can bring about cost savings and system productivity.

 
Conclusion:
A monitoring tool for the measurement of examination duration and inter-examination delay times was developed using DICOM tag information. In this example, four CT systems were monitored for their efficiency. A relational database provides full access to queries and generates reports ad hoc for administrative use.
 
References:
1. Andriole KP. “Productivity and Cost Assessment of Computed Radiography, Digital Radiography, and Screen-Film for Outpatient Chest Examinations.” Journal of Digital Imaging. September 2002;Vol.15(3):161-169.

2. Reiner BI, Siegel EL, Hooper FJ, et.al. Multi-institutional analysis of computed and direct radiography: Part I. Technologist productivity. Radiology. August. 236(2):413-9.

3. Reiner BI, Siegel EL, Hooper FJ, et.al. Multi-institutional analysis of computed and direct radiography: Part II. Economic analysis. Radiology. August. 236(2):420-6.

4. Law MY, Liu B. “Informatics in Radiology: DICOM-RT and Its Utilization in Radiation Therapy.” Radiographics. 2009;29(3):655-667.

5. “Realize the full potential of your equipment and staff”, available at http://www.healthcare.philips.com/main/products/mri/options_upgrades/utilization_services/index.wpd.

6. Geis JR. “Medical Imaging Informatics: How It Improves Radiology Practice Today.” J Digit Imaging. June 2007;20(2):99-104.

7. Chan WP, Chiu WT, Chen WM, et al. “Applying Six Sigma methodology to maximize magnetic resonance imaging capacity in a hospital.” International Journal of Healthcare Technology and Management. 2005;Vol.6(3):321-330.

8. Gale B, Safriel Y, Lukban A, Kalowitz J, Fleischer J, Gordon D. “Radiology report production times: voice recognition vs. transcription.” Radiol Manage. 2001;23(2):18–22.