Picture this: a hospital administrator answers an urgent call one morning. The film room, where all the patient imaging studies printed on acetate film are stored, has been emptied overnight. Only empty shelves and some scattered folders strewn about the floor remain. On the security camera videos, the administrator sees people with covered faces stuffing films into 55-gallon drums and moving them out.
Criminal gangs are not the only threat to healthcare images. Analog films can be hard to share, lost, or destroyed by heat and humidity. In the analog world, physical constraints mean that clinicians and patients are sometimes unable to benefit from the full potential of imaging. Decades’ worth of prior exams and invaluable clinical data could disappear overnight, compromising patient care and causing reportable HIPAA/HITECH Act Protected Health Information (PHI) breaches[1] for targeted organizations, resulting in fines and negative publicity.
The challenges of storing imaging data
Starting in the 1980s[2] computed radiography, followed by fully digital modalities, the development of the DICOM (Digital Imaging and Communications in Medicine) standard, and the adoption of PACS (Picture Archiving and Communication Systems) allowed radiology to embrace digital transformation. Breaking free from the limitations of film and moving to a fully digital workflow with benefits like reduced time from image acquisition to report availability and diagnosis, improved patient care, and increased throughput and utilization of imaging modalities. Cardiology soon followed suit as did most other specialties when digital photography became commonplace, and especially when handheld devices such as smartphones and tablets started including high-resolution cameras. This evolution gave rise to the field of Enterprise Imaging, encompassing all imaging used in healthcare beyond the original specialties of radiology and cardiology, including dermatology, gastroenterology, emergency medicine, surgery, pathology, orthopedics, and many other specialties.
Repurposing film rooms and other physical remnants of the analog era was a welcome side effect of the digitization of radiology. All those images, however, had to be stored somewhere. Datacenters and computer rooms were sprouting up in hospitals, medical office buildings, imaging centers, and all sorts of other healthcare locations. The racks full of servers, storage arrays, and networking equipment that filled them were in large part dedicated to processing and storing image data. Unstructured data, including imaging, represents nearly 80%[3] of all data in healthcare. Being outside of databases, unstructured data presents different challenges for performance, availability, and protection. The file systems included in most operating systems do not offer enough scale, performance, or reliability to hold unstructured clinical data in general, and imaging data in particular. Those requirements led to this data being stored in purpose-built systems like Network Attached Storage (NAS) or, more recently, object storage.
Even though most jurisdictions have regulations controlling the retention requirements for health data, it is uncommon for a healthcare provider organization to delete any clinical data including images. There can be diverging interests between different stakeholders such as clinicians, who may be interested in keeping as much data as possible in case it can help with patient care, attorneys, who may be interested in reducing liability and following the letter of the law by deleting data as soon as it is no longer legally required to keep it, and IT professionals, who may be focused on information security, infrastructure costs, and overall system stability and manageability. More recently, the significant value of these labeled datasets (images, metadata, and radiology or other specialty reports) to train AI models for detecting and classifying abnormalities, segmenting organs, and quantifying clinical measurements to name but a few use cases, has made it even less likely that image data would be deleted even after the legal retention requirements have been met.
Cloud comes to the rescue if it has the right pricing model for Enterprise Imaging
Enterprise Imaging went full circle, from requiring film rooms and warehouses to hold physical film and paper copies of reports to requiring datacenters full of compute, storage, and network infrastructure to hold digital versions of images, metadata, and reports. This data, being more valuable now than ever, is unlikely ever to be deleted leading to the need for ever expanding IT resources to hold it safely and in such a way that it can be available to clinicians through the systems of record like electronic health records (EHR) and PACS. At the same time that this trend is intensifying, healthcare organizations are facing increased pressure to focus on their core competencies in an era of ever-decreasing margins, staff turnover, clinician burnout, and increased regulatory pressures. A common consequence is the realization that running datacenters is not a core competency for most healthcare providers, leading to the common refrain of “we need to get out of the datacenter business”.
Public cloud services offer a perfect solution to this conundrum, allowing for a healthcare organization to decide how much or how little they want to shift from on-premises to the cloud. Oracle Cloud Infrastructure can help you map out a cloud journey that can start with archiving old images, like those older than ten years, to the cloud to free up space in your datacenters to limit previously unchecked infrastructure growth. From there moving the secondary imaging environments, disaster recovery or business continuance solutions, is a logical progression. Then implementing full cloud-based PACS or vendor neutral archives (VNAs), as many healthcare providers are doing.
Not all public cloud services are equal when it comes to Enterprise Imaging. In the public cloud space, there is an unfortunate tendency to build “Hotel California” where you can “check out anytime you want but you can never leave”. Ease of use, elasticity, and abundant services entice you to move your data and applications in, but egress charges mean that any time that you want to use your own data, not to mention move it out, you get hit with significant charges. In Enterprise Imaging, where hundreds of terabytes are standard even for a medium-sized hospital, these egress charges can be an unwelcome surprise that upends all pricing projections and makes you liable for exorbitant bills. At Oracle, we believe that your data should not be held hostage. In many cases there is no charge to access or move your data, and when there is, it is a fraction (think 1/20th) of the industry standard.
Oracle Cloud Infrastructure is the second-generation cloud built for industries such as healthcare. Let us show you how we can help you start your Enterprise Imaging cloud journey with no spiraling bills in a way that will make clinicians and IT happy.
To learn more about how you can use Oracle Cloud Infrastructure for healthcare, contact one of our representatives.
[1] https://www.hipaajournal.com/what-are-the-penalties-for-hipaa-violations-7096/
[2] https://pubmed.ncbi.nlm.nih.gov/34744119/
[3] https://blog.hyland.com/hyland-healthcare/unstructured-data-in-healthcare-the-missing-link-to-interoperability/