Case Study: Imaging and Clearcanvas PACS at Zinga Children’s Hospital Tanzania
There are several options when deploying open source PACS systems. In this case, we installed a proven server and viewer, which is Windows-based, i.e., ClearCanvas. This software is very well suited to be deployed in a small clinic, such as this one, with 3 modalities doing a couple of procedures/day. The system was funded by a Rotary International Global grant of a little over $100k, which provided a stationary Fuji X-ray and CR, Ultrasound, and dental X-ray with sensors. The X-ray was sourced and installed by a local dealer, and together with the Ultrasound, located in a newly constructed building, an Ethernet cable connected these modalities to the central facility, which is a few hundred feet away. The images can be previewed on the Fuji monitor and QA’d while diagnosis is done on a medical 2MP display. The PACS software runs on a locally sourced desktop server, which has a NAS as a backup.
Lessons Learned:
We installed this a few years back and at that time selected to use a CR (removable plate system) instead of a DR (direct digital plate) for the X-ray to save money. However, the difference between CR and DR is currently narrowing. If we were to do this again today, we might choose DR technology as the incremental cost might be justified by the increase in reliability.
I was surprised to see how many deployments of ClearCanvas there are in Africa. It is impossible to find out where and how many of these installs there are, as anyone can download the software, and there is no record except for the 50k+ downloads, but it has proven to be a workhorse. The fact that it is Windows-based makes it easy accessible and supportable. The open source version lacks web functionality, which is why Orthanc would have been a better choice if it had been done today. (A case study of that will be covered in a next newsletter)
ClearCanvas has an extensive built-in Help feature, and the local staff can typically learn how to use it rather quickly. However, some level of coaching is still essential. For example, at another site in Kenia that uses this software, the radiologist told me that some of his function keys had suddenly disappeared. I had to teach him how to re-arrange them and get them back, which shows that a user group and/or online resource is critical.
Using the viewing software is not hard to learn, and its functionality is easy to grasp. Workflow, routing, archiving, and overall PACS system functionality require some level of additional training, especially with regard to troubleshooting (what if I can’t find the images I just sent?)
We used refurbished dental detectors and a portable X-ray unit, which saved quite a bit of money and does not require additional shielding. The Fuji X-ray room was appropriately shielded and was inspected by the local governmental agency responsible for radiation protection.
Deployments are very challenging as local suppliers and dealers, as well as government approvals, follow their own, flexible schedule, which is hard to match with resources from the US with a relatively fixed, preplanned schedule. Combined with challenges to import equipment and getting it through customs, it makes it hard.
The Ultrasound was imported directly from China based on recommendations of another local hospital, which used this model successfully, which meant we could buy it for half the price it would have cost in the US.
To justify these systems for future grant applications, it is critical to collect data regarding usage, the procedures performed, and their impact (outcome measurements), which is often challenging.
A significant gap in the availability of open-source software is at the front-end of the workflow, i.e., having a simple scheduling and ordering system (a mini-RIS). This also provides unique patient and procedure identification. There is OpenMRS and a few more, but there is still a lot to be done in this field.
Radiation safety is a big issue and requires training. The radiation safety officer from the government found leaks through the door of a restroom connected to the X-ray room, as well as a non-lead glass panel to the control room.
Medical-grade monitors are a must, and in my opinion, using a low-grade OTS monitor is not acceptable as it simply does not represent all greyscale levels and is not calibrated. The additional cost of a 2MP monitor, which allows for zooming up to full resolution, is only a small incremental cost of the overall project. Awareness of this issue is critical in these countries.
Physical security is a significant issue in these resource-limited countries. One hospital installed a security camera at every workstation. In this particular site, they have Masai guards who literally live on the campus and walk around with their machetes to keep strangers out.
In conclusion, a relatively inexpensive (US$100k) package with an X-ray, Ultrasound, and dental system that includes an open source PACS system is very supportable and feasible for a small clinic. It can make a significant impact on the lives of these children.
