Dr. Vivek Kumar, K19617, Dr. Parth Rana, Dr. Kashyap Patel, Dr. Shreya Shah
Abstract:
Purpose: To determine the impact of a Digital Imaging and Communications in Medicine (DICOM) workflow on the linkage of demographic information to ophthalmic data.
Design: Evaluation of technology.
Participants: 2414 visual field, OCT and Fundus imaging performed by 2 technicians.
Methods: At 3 months before and 12 months after implementation of a DICOM workflow, technicians recorded the work required to enter, confirm, or edit patient demographics in each visual field, OCT and fundus imaging devices. We also determined the proportion of imaging tests sent to an error queue for manual reconciliation because of incorrect demographic information before and 28 months after the DICOM workflow was established.
Main Outcome Measures: The proportion of testing encounters for which staff had to enter, edit, or merge patient demographics and the proportion of misfiled images.
Results: Staff entered, edited, or merged data for 19% of patients before implementation (n= 497). This decreased to 2.2% within 12 of implementing the DICOM archive (n=2414). Staff could locate a patient in a DICOM work list for 99% at 12 months. Before implementation, 18.59% of the images required additional intervention to be associated with the correct patient (n=497). This decreased by 2.2% over 12 months (n=2414; P < 0.05). There was a reduction in the percentage of misfiled images between pre implementation and post implementation 12 from 2.2% to 0.95 p<0.05).
Conclusions: Implementation of a DICOM-compatible workflow in an ophthalmology clinic reduced the need to enter or edit patient demographic information into imaging or testing devices and reduced the need to manage misfiled images. In a clinical environment that demands both efficiency and patient safety, the DICOM workflow is an important update to current practice.
Key words: Workflow; DICOM; OCT; HFA; Electronic Health Record
Ophthalmic imaging is very important component of ophthalmic investigation.
The importance of ophthalmic imaging was recognized as early as 1886, when an article in the Philadelphia Photographer described a technique by which to photograph the “retina of the living eye.” [1, 2] Today, the modern ophthalmologist can count on dozens of digital imaging and testing methods to glean high-quality information about ocular disease, to monitor disease course, and to document treatment effects. [3] However, the rapid expansion of digital imaging is not without its problems. Quickly and accurately collecting, transmitting, storing, and retrieving results for all of the patients seen in a clinic undergoing a different combination of testing represents a true informatics challenge. [3]
Currently, All data is being collected in databases of standalone ophthalmic devices, when review is required during sub sequent visit clinician either may use hard copy or should have access to devices like fundus camera, HFA or OCT and most ophthalmic testing methods store data in proprietary databases or formats that cannot be readily imported into an electronic medical record or a image archive. Often, this necessitates either printing images for review by the ophthalmologist or implementing proprietary interfaces to extract printed reports in digital format more over clinical data may not readily available. Consequently, ophthalmologists cannot take full advantage of the digital paradigm, either in streamlining data management or in using software to analyze the raw test data. [2, 3, 4]
Images are important for documentation, monitoring of diseases.
During decision making process if a clinician have access to images from all three devices along with other clinical data it may be very useful for work flow as multiple visits may be visible.
A single ophthalmologist may involve medical records personnel, front desk staff, technicians, photographers, and multiple physicians. This is very complex method of workflow, there is great chance for losing or misfiling paper based patient data. The field of radiology came to realize that physical film workflows resulted in limited ability to make longitudinal comparisons between studies, unnecessary repeat examinations, delayed clinical decision making because of missing examinations, or diagnoses applied to incorrect patients., [5, 6] in response to growing evidence about the role that misidentification plays in iatrogenic errors, the efforts for improving the accuracy of patient identification essential. [1, 7] Thus, a key issue for the modern ophthalmologist is developing a system to access easily all the right information, for the right patient, every time.
In response to their data management challenge, radiologists developed the Digital Imaging and Communication in Medicine (DICOM) standard, which has since been extended for use in ophthalmology. [3, 4] First developed in 1985, DICOM is a mature, universal, and non-proprietary standard that “provides all the necessary tools for diagnostically accurate representation and processing of medical imaging data.” 7 Because DICOM was designed with digital data transfer, storage, and processing in mind, the standard confers several advantages over printed images or, in fact, non-DICOM digital images. The DICOM image files contain a header section that includes data about the image itself acquisition parameters, filters, image dimensions das well as more than 2000 demographic and medical attributes including patient’s name, date of birth, diagnosis, and provider. Images and their context are linked and are far less prone to identification errors and are more amenable to further processing and analysis. [8, 9]
The usefulness of this standard in ophthalmology was heretofore limited by slow adoption of electronic medical records and a lack of ophthalmology specific DICOM standards. Recently rapid increase in popularity of electronic medical records 2 and the growing trend toward DICOM-compatible ophthalmic imaging, we set out to explore the impact of the DICOM workflow on staff efficiency and the rate of misfiled images in an academic ophthalmology practice.
Acquired images then are sent to and stored in a central server known as an archive and can be viewed via networked workstations. This combination of imaging methods, a DICOM archive, and review stations is known as a Picture Archiving and Communication System (PACS). [ 9]
METHOD:-
This study was determined to be nonhuman subject research by the Drashti Netralaya Institutional Review Board and therefore was exempt from review. Before the implementation of a DICOM-based workflow for in office testing, ophthalmic technicians manually entered patient demographic data into each machine. So, if a patient required a visual field test, fundus imaging and optical coherence tomography
(OCT), the technician entered the patient’s name, birth date, medical record number, or a combination thereof separately into each machine. After testing, each machine then generated separate reports in both paper and digital formats. Digital images then were sent to an image management system (Lekhisoft; Digital Healthcare, Pithampur, India) that double-checked final demographic data associated with the image against the patient registration system before finally serving the images to the physician on request. On demand of clinician selected images made available.
In late September 2013, the Drashti Netralaya deployed a DICOM-based archive (Forum version 3.2; Carl Zeiss Meditec, Dublin, CA) to accommodate data from several testing devices distributed across both subspecialty clinics at our primary location at Drashti Netralaya and our geographically distributed satellite clinics. The server software was installed on a single central server managed jointly by existing institutional information technology staff. No additional staffs were added to accommodate this project. It was also necessary to install new DICOM licenses on some of the devices that were integrated with the archive. Licenses for existing devices cost approximately €2000 per device, and the new devices added during the evaluation were specified to include DICOM capability at the time of delivery. Licensing of the archive software itself was included in an overall agreement between Carl Zeiss Meditec and Drashti Netralaya, so the incremental cost of the new software is not available. Also at the time of implementation, all existing visual field data were uploaded from each of our Humphrey field analyzers (HFAs; Carl Zeiss Meditec) to the DICOM archive so that prior studies could be available for progression analysis.
At the time of the initial deployment, 1 HFA, 1 Cirrus OCT devices, Fundus camera FF 450+and 1 IOL master 500 (Carl Zeiss Meditec) were integrated with the DICOM archive.
The DICOM archive was, in turn, linked to our central patient registration system via a Lekhisoft interface. The goal of this approach was to have demographics flow from the registration system to the DICOM archive, then from the archive to the individual devices as a DICOM method work list. Technicians then selected the appropriate patient from the list (rather than manually entering demographics). After testing was complete, the testing data then flowed back to the DICOM archive for storage and forwarding. As with the pre-DICOM workflow, test data were sent to the image management software (Lekhisoft) and then were available for review by clinicians.
urvey of Ophthalmic Technicians:-
To determine the impact of the DICOM-based workflow on ancillary clinic staff, we assessed the work performed by 2 technicians in the Drashti Netralaya during 2 time periods: 3 months before the implementation of the new workflow (June, September 2013), and 12 months immediately after implementation (September 2014). At each of the evaluation periods, we asked each technician to complete a paper form indicating the work required during each encounter.
Misfiled Images:-
Another source of inefficiency in the pre-DICOM workflow was the requirement that information technology staff had to work actively through a queue in our digital image management software containing those images and reports that could not be linked to a real patient based on the demographic data. This situation arose each time a technician mistyped a patient’s name, birth date, or medical record number.
To measure this phenomenon, we analysed the log files of our image management system (Lekhisoft) to count the number of tests that ended up in this queue. This was performed concurrently with the surveys of ophthalmic technicians 3 months before (June September 2013) and 12 months after (September 2014) after implementation of the DICOM workflow. Accordingly, the log files reflect data from the additional devices that were networked to the DICOM archive in March 2012.
Statistical Analysis:-
Differences between individual categorical variables were assessed using one test ANOVA tests, and differences in overall work distribution were assessed. Statistical analyses were performed using SPSS statistical software package (version 22). Differences were considered statistically significant if the P value was less than 0.05.
Results:
Survey of Ophthalmic Technicians
Before implementation of the DICOM workflow, 19% of testing encounters required some level of intervention to add or edit demographic information. At 12 months, only 2.2% of testing encounters required merging or entry of patient data. A greater proportion of encounters trended toward having correct demographics available to the technician, although this was not statistically significant (81% vs. 91%; P<0.22). After 12-month evaluations, the vast majority of encounters did not require any data entry by technicians. Relative to the 12-month post-DICOM period, technicians were using the work listless (7.04% vs. 0.91%; P < 0.05) and instead were using the DICOM archive more to select specific patients and to push their records down to the devices (HFA, OCT and Fundus imaging) (4.8% vs. 0.689%; P < 0.05). (Table.1)
Misfiled Images
Over the 24 months of evaluation, the misfiled image rate decreased significantly (2.2% to 0.91%; P < 0.01; Table 2).
This assessment offers a thorough evaluation using DICOM workflows to address the growing challenge of managing electronically stored ophthalmic testing data. Use of a workflow enabled by DICOM allowed our clinic to reduce by 19% the number of patients requiring technicians to manipulate their demographics, while also enabling those technicians to access the correct patient demographic data at the devices(HFA,OCT and Fundus imaging) more than 97.8% of the time. Consequently, the rate of misfiled images decreased from 2.2 to 0.9%. These findings have implications both for clinic efficiency and for patient safety.
Any new technology, of course, has unexpected effects, and the DICOM archive was no exception. Because of the pre population of the archive with existing device (HFA, OCT and Fundus imaging) data, technicians were required to merge existing, incorrect demographics with those received from the clinic registration system. If each duplicate patient record was merged appropriately, we would have expected the proportion of duplicate records to decrease over time, further reducing the time required by technicians at the archive interface.
In the context of the system we evaluated, there were 3 ways to associate a patient with a particular test: (1) select the demographics from a work list in the device, (2) push a patient record from the DICOM archive to the device, or (3) enter demographics de novo. At 12 months, technicians chose the first 2 options more than 97% of the time. However, we note that there was a statistically significant decrease in the number of patient records selected on the HFA work list over time and a statistically significant increase in patient encounters pushed from the DICOM archive. One possible explanation is that technicians were overwhelmed with the size of the work list presented at the devices (HFA, OCT and Fundus imaging); it listed not only speciality (glaucoma, retina) patients, but also every patient registered across the department that day. Based on conversations with staff, we found that it became a nuisance to scroll through hundreds of names to find the particular patient they were testing.
Technicians were now working at 2 interfaces the DICOM archive and the testing device instead of 1, and an argument could be posited that this complicated their workflow. It was our experience that by installing the DICOM archive software on the technicians’ computers right next to the devices (HFA, OCT and Fundus imaging) and by virtue of only having to click names instead of typing out full demographic data, the workflow was greatly simplified. Alternatively, there are potential risks to making things too simple. That is, if now all that is required is clicking on a name, someone could quickly and accidentally click on the wrong patient record.
Although the rate of misfiled images decreased by more than 75% over 12 months, despite a near 5-fold increase in image volume (497 to 2414), the rate did not fall to 0. After consulting with our information technology people, we discovered several possible explanations. By design, patient demographic data from our satellite clinics also flow into our central clinic registration system and subsequently into the image management software. We also noted that our medical records department also unable to find specific patients because of demographic errors occasionally issued new medical record numbers to existing patients, resulting in duplicate patient information.
Ophthalmology is a data driven and increasingly technology driven specialty. The key to improving efficiency and enhancing patient outcomes relies on an ophthalmologist’s ability to organize and use that data. From a clinical perspective, DICOM integration promotes a patient centric versus method centric approach to ophthalmic imaging. Accordingly, such a workflow offers ophthalmologists the opportunity to integrate and review more easily all of the relevant data for a given patient. Moreover, access to raw testing data, with the ability to zoom, apply filters, perform 3-dimensional transformations, or view as video, may extract enhanced clinical value from the same test. From a patient safety perspective, clinical images and their context are now inseparable, reducing the risk of misidentification. We describe here a solution by which we can use the DICOM workflow an open and proven technology to achieve these critical goals.
Acknowledgments: The authors thank Drashti Netralaya Biostatistics Consulting Centre for assistance in statistical analysis.
References
- Ravi R. Pandit, Michael V. Boland. Impact of Digital Imaging and Communications in Medicine Workflow on the Integration of Patient Demographics and Ophthalmic Test Data Ophthalmology 2015;122:227-232
- Bennett TJ, Barry CJ. Ophthalmic imaging today: an ophthalmic photographer’s viewpoint a review. Clin Experimental Ophthalmol 2009; 37:2–1
- Boland MV, Schuman JS, Mattox CG. Patient Care Committee of the American Glaucoma Society. American Glaucoma for clinical practice. J Glaucoma 2013; 22:174–
- Chiang MF, Boland MV, Brewer A, et al. American Academy of Ophthalmology Medical Information Technology Com Special requirements for electronic health record systems in ophthalmology. Ophthalmology 2011; 118:1681–7.
- Morishita J, Watanabe H, Katsuragawa S, et al. Investigation of misfiled cases in the PACS environment and a solution to prevent filing errors for chest radiographs. Acad Radiol 2005; 12:97–10
- Schulze OC, Ackerman C, Greyling J, et al. Talking PACS: Part 2-why should we change to PACS? S Afr J Radiol 2007; 11: 86–90. Available at: https://www.sajr.org.za/index.php/sajr/article/ view PDF Interstitial/19/3 Accessed May 28, 2014.
- The Joint Commission. National Patient Safety Goals Effective January 1, 2014. 2013. Available at: https://www. org/assets/1/6/HAP_NPSG_Chapter_2014. pdf. Accessed March 2, 2014.
- Pianykh OS. Digital Imaging and Communications in Medicine (DICOM): A Practical Introduction and Survival Guide. 2nd ed. Berlin: Springer; 2012:4–1
- Graham RN, Perriss RW, Scarsbrook AF. DICOM demystified : a review of digital file formats and their use in radiological practice. Clin Radiol 2005; 60:1133–4
Table 1. Work Required by Technicians at the HFA, OCT and Fundus imaging Before and 12 months after the Deployment of the Digital Imaging and Communications in Medicine Workflow
| 3 months Before(n=497) | After 12 month(n=2414) | P value | |
| Correct Entry | 403(81.0) | 2361(97.8) | <0.22 |
| Number of repeat entry | 35(7.04) | 22(0.91) | <0.05 |
| Number of merge data | 24(4.82) | 15(0.62) | <0.05 |
| Number of editing of data | 25(5.03) | 6(0.24) | <0.05 |
| Number of spelling variation | 10(2.01) | 10(0.41) | <0.05 |
| Total | 497(100) | 2414(100) |
Table 2. Misfiled Images in Image Management Software by Stage of Implementation
| Misfield | P value | |
| 3 months before implementation of DICOM | 14(2.21) | D |
| 12 months after implementation of DICOM | 23(0.95) | P<0.05 |
Leave a Comment