In the quest to mine and analyze meaningful, reliable, and useful data from the burgeoning plethora of electronic and online sources, healthcare organizations can allow the big picture to overshadow many underlying and valuable components contributing to patient care improvement.
The clinical data and diagnostic images in radiology information systems (RIS) and picture archiving and communication systems (PACS) remain two examples. For clinical imaging and radiology executives, these visual clues and cues are necessary for effective, efficient decision support.
Certainly a growing number of manufacturers and information technology companies recognize this – even if many healthcare providers have not yet reached the point where they can tackle the necessary underlying infrastructure beyond the planning and strategic stages. As a result, they’re offering providers a light at the end of the tunnel.
Streaming expanded horizons
“The latest generation of reporting capabilities can help improve the utilization of imaging data for diagnostic decision making,” says Cristine Kao, Global Marketing Director for Healthcare Information Solutions, Carestream.
An NIH study concluded that oncologists and radiologists prefer quantitative reports that include measurements as well as hyperlinks to annotated images – with tumor measurements, for example. A report by Emory and ACR shows eight out of 10 physicians will send more referrals to facilities that can offer interactive multimedia reporting – citing the ability to better collaborate with radiologists.”
Connecting all of the technology and tools remains important, too, for a visually rich information view, according to Todd Winey, Senior Advisor, Strategic Markets, InterSystems.
“Images need to be blended into the comprehensive patient record, along with genomic information, claims data, wearables, and other disparate healthcare data sets,” Winey tells Health Management Technology. “Picture archiving and communications system images represent a significant percentage of the worldwide medical data volume. However, these images have remained locked in proprietary PACS systems that have had limited connectivity to systems outside of radiology. All that is changing. Rich standards to support image exchange are becoming more widely available. Vendors are promoting vendor-neutral archives (VNAs) to unlock images across disparate PACS systems, and health information exchanges are leveraging images as a component of the comprehensive patient record for all care team members.”
For the clinical and diagnostic data to play a more valuable role in patient care improvement, these trends need to be accelerated, Winey insists, which isn’t without challenges. “VNAs remain only marginally deployed,” he laments. “Many of the advances in radiology information systems and PACS have been focused on productivity improvements for radiologists and are not yet fully supporting advanced interoperability.”
Kao agrees with the foundational importance of a VNA but adds that it shouldn’t stop there.
“A vendor-neutral archive that links radiology, endoscopy, pathology, cardiology, lab results, and other systems is a good foundation,” she says. “But having a repository is just the first step – providers also need to provide the clinical context for unstructured data within those systems to enable better decision support. Carestream’s Clinical Collaboration Platform offers optional services to a VNA – or other backbone – by using the latest interoperability standards to identify and aggregate data from disparate systems to create a holistic view of the patient. Once you have a comprehensive view of patient data it can be maintained and distributed to authorized users for different uses, such as clinical diagnostics, analytics, etc.”
Depending on an organization’s capabilities, imaging data must be accessible to more than just one clinical segment to be included as part of the decision support process, according to Winey.
“For imaging data to truly drive patient care improvement, imaging data must become part of the unified health record, available in real time to the entire care team, not the exclusive domain of radiologists,” he says. “Bringing RIS/PACS data into the era of accountable care requires a focus on sharing available imaging studies to avoid duplicative diagnostic tests, extensive use of decision support guidelines to minimize unneeded imaging studies, broader access to imaging data during patient encounters, and the ability for patients to own their own records to avoid duplication. These have been tenets of the RSNA Imaging 3.0 education efforts and are critical steps towards a patient-centered imaging process.”
Kao says she fully anticipates future reporting functions may include “more intuitive searching capabilities that will link pertinent patient information for a specific condition or disease, even if previous reports did not include the specific word involved in the search command. The goal for enhancing the entire diagnostic process is to provide clinically relevant information when and where it’s needed. New advanced reporting techniques provide information that can lead to improved decision support and diagnostic outcomes.”
Imaging can contribute to the outcomes of the healthcare system by providing meaningful information to drive decision making, and by contributing to the patient’s visual medical history, according to Sham Sokka, Ph.D., Head of Radiology Solutions, Philips Healthcare.
“The future of imaging depends on the integration of data from imaging modalities and clinical informatics, allowing for optimization and improvement in outcomes, both economic and clinical,” Sokka says. “[At] Philips, we are developing integrated imaging to drive meaningful change towards achieving the Triple Aim. This integration of imaging toolsets – from acquisition to clinical informatics to care coordination to outcomes management – enables future innovations to achieve better outcomes and overall health of populations.”
Healthcare organizations continue to struggle with the demands of improving outcomes while reducing per-capita costs, Sokka says, adding that the amount of available data complicates matters.
“Data can help direct improvements, yet most hospital systems have a plethora of unaligned data sources, such as EMR, RIS, PACS, business intelligence/workflow engines, HL7, and DICOM,” he says. “In addition to people and process transformations, improving outcomes while reducing costs requires aggregation of the multiple silos of data in healthcare and converting that data into information and intelligence for data-driven decision making.”
Sokka says that practice management is the solution that ties all of this information together to drive continuous improvement.
“To transform your practice, you need an analysis and intelligence framework to monitor and to improve operational, financial, and clinical aspects,” he continues. “The practice management platform could act as a command center that could be customized to the needs of different users like director of imaging services, radiology administrator, chief of radiology, etc. The solution also provides a toolset and services for healthcare provider leadership to understand the impact of imaging at the system level, and to drive collaboration across clinical service lines to deliver improved performance.”
The “longitudinal picture or visual medical record” plays a key role, Sokka says.
“Your medical record is the written version of your history, but there’s a lot of meaningful information in images, videos, and pictures collected along the care pathway,” he says. “Radiology has typically been the imaging information source within the hospital network, and can continue to contribute to this but also drive the broader aggregation of the full visual history of the patient.”
Susan Niemeier, R.N., BSN, MHA, Chief Nursing Officer, CapsuleTech, stresses the foundational nature of data analytics within the clinical decision support system.
“Data analytics is the engine of a clinical decision support system that can help describe what has occurred in the past, why something is happening, and what is predicted to happen in the future,” Niemeier says. “Data analytics is built on evidenced-based criteria and unique insight into the association of data to certain events. Data analytics coupled with near real-time, comprehensive, and accurate data is the foundation for a clinical decision support system and the delivery of its full promise of increasing quality of care, enhancing health outcomes, minimizing errors and adverse events, improving efficiency, reducing costs, and improving staff and patient satisfaction.”
Yet in today’s alert-addled industry, clinicians need help navigating through the noise, according to Kathleen Aller, Director of Business Development for HealthShare, InterSystems.
“Analytics should inform us of baseline states, help to target where clinical decision support can be applied to greatest benefit, and then quantify the impact, positive or negative,” Aller says. “Given the risks inherent in over-alerting, and the disruptive impact of inappropriate interruptions to care processes, knowing what is and is not moving the dial in the right direction is a critical factor in successfully deploying decision support interventions.”
Holistic partnerships
Data-driven decisions may lead to improved decision-making so long as the clinical and operational perspectives factor in the patient.
“Data analytics is the process of compiling raw data from various sources and turning that data into information, that information into knowledge, and that knowledge into action,” says Thomas Van Gilder, M.D., Chief Medical Officer and Vice President, Informatics and Analytics, Transcend Insights. “Successful analytics should be shared with physicians, caretakers, and patients in a way that is relevant, easy to understand, and results in action. It should also fit well into the care team’s workflow.”
Those decisions ideally should be made in a more objective and timely manner, according to Jason Williams, Vice President, Business Analytics, Financial Solutions, McKesson RelayHealth Financial.
“Lots of decisions are made in a vacuum of objective data or deferred for lack of facts,” Williams says. “We’ve seen a number of examples in interactions between clinicians and operations teams. When the data is made available, it elevates the discussion above anecdotes and perceptions, creates imperative for change, and delivers facts. Moreover, change in actions can be prescribed prior to reports of poor results, such as high documentation-based claim denials.”
Yet Larry Schor, Senior Vice President, Medecision, cautions against a “just-the-facts” philosophy, particularly for more acute and chronic patients as they are drawn into the care process.
“Analytics will be increasingly important for patients with complex care needs where alternative treatment options require analysis and consideration,” Schor says. “And weighing options will mean not just facts about the patient objective history information, but also self-reported preferences and personal expectations that will need to be considered. Decision support and analytics will no longer be the sole province of the practitioner, but include the patients in important treatment decisions that consider how alternatives rank in terms of outcomes, side effects, and lifestyle.”
Sarah Corley, M.D., FACP, Chief Medical Officer, NextGen Healthcare, recommends embedding analytics and decision support into practice procedures from the start to capitalize on the value of data already collected.
“There are many areas of medicine of which we do not have solid evidence as to what is the best treatment option in terms of outcomes, complications, and costs,” Corley says. “Data analytics can play a vital role in unlocking the answers from data already existing once it is aggregated. It can assist in prioritizing care when time or resources are limited.”
“Moreover, one can derive value out of the data by putting analytics into the workflow in order to improve clinical and economic outcomes, thereby increasing patient satisfaction,” she says.
Population to personal
Manu Varma, Vice President, Strategy, Philips Hospital to Home, sees the precision dynamics within data analytics for individualized care.
“Analytics can help spot trends inside large amounts of data being generated in healthcare IT systems,” he says. “These trends can then be codified into decision support tools that can allow superior provider and patient experience.”
As an example, Varma cites his company’s “Discharge Readiness Score” created for patients in intensive care units. “We used data from [more than] 1 million ICU stays of patients to develop a scoring mechanism to spot patients ready to be discharged from the ICU,” he says. “What used to depend entirely on physician review of large amounts of data is now supported through a simple score. Tools like this have helped telehealth programs show as much as 20 percent reduction in length of stay.”
Rather than distinguish data analytics from decision support, Chris Hobson, M.D., Chief Medical Officer, Orion Health, classifies the former merely as a form of the latter to shape population health initiatives. Analytics shows “stratified populations of patients with gaps in care, plus the ability to go from a list of patients with gaps in care directly to each individual patient record, so the provider can take action on that patient’s care gap,” he says.
“Data analytics are typically applied at a whole population level,” Hobson says. “We translate that population view into an individual patient view that shows care gaps and tasks, and their relative priorities. Imagine analytics that stratify your patient population and auto-enroll them onto condition-specific pathways with action items, deadlines, and owners. Now imagine that these owners are automatically alerted as to the next step in the patient’s care in whatever way suits their workflow.
For example, a care coordinator owner may see an action item for a particular patient on a COPD pathway within the care coordination system or module he uses each day. A physician owner may be alerted via a text message or right within her EMR, and a patient may be alerted via email that there is a task awaiting his attention in the patient portal. This personalization of action allows all stakeholders to work in the manner most efficient to them but with the added value derived from population level and individual patient data analysis.”
Donald Voltz, M.D., Department of Anesthesiology and Medical Director of the Main Operating Room at Aultman Hospital in Canton, OH, recognizes the inherent value in data analytics with and within decision support.
“One of the greatest potentials in healthcare comes from how patients and healthcare professionals use the data and information that is generated,” he says. “There is a great deal of success on using population data to assess the effectiveness of treatments and interventions. There is no question this will continue to be utilized to guide physician decisions and help patients become more involved with these decisions.”
But Voltz sees the communication connections within analytics as holding potential for understanding how the system functions.
“One of the problems that remains in healthcare is communication between health professionals as well as between patients and their care providers,” he says. “Data analytics on what information is required, or of most interest, to making my decision goes a long way to ensure that a high quality of care is being delivered. Without such an understanding, and the use of decision support to filter the expanding amounts of patient health data, healthcare professionals and patients will be overwhelmed with the data deluge.”
In fact, Voltz envisions technology enhancements to current systems to drive more efficiency and generate more informed decisions to improve patient care.
“Imagine a smart system, one that sits on top of one or more EHRs, that recognizes all the members of the care team,” he says. “When an abnormal result, a missed finding to support a diagnosis, or a previously ordered test is entered into a different system, decision support aids in addressing the needs by the most appropriate member of the team. In addition, tracking on how this is accomplished brings about the ability to redesign systems, visual displays of information, or processes to improve the flow of information in the future. Predictive presentation or smart displays of relevant information bring consistency and less risk of errors from missed data or duplication of services.”
