contents

business
 
CDS Tools Set to Change Way Medicine Is Practiced

Clinical decision support (CDS) tools, technologies that provide information to aid the diagnosis and treatment of patients, are set to fundamentally change the way medicine is practiced. This is according to the latest report from independent market analyst Datamonitor. The report, "Clinical Decision Support in Healthcare: One Step Closer to the Omniscient Clinician", expects clinical intelligence (CI) solutions to be the next major trend in CDS support tools, followed by patient-centric and diagnosis-related CDS. However, the report points out that medical culture will be the major obstacle to overcome in CDS adoption.

CDS technologies range from online reference materials to guidelines to alerts built into electronic prescribing (eRx) and computerized physician order entry (CPOE) to data mining to artificial intelligence. While most CDS tools today are targeted to providers (hospitals, physicians, nurses, physician assistants, pharmacists, physical therapists, etc.) and payers, governments and patients will also use CDS to a greater extent in the future.

The culture of medicine is resistant to CDS

CDS solutions improve patient care, potentially decrease healthcare costs over time and make it easier for providers to take part in pay-for-performance (P4P) initiatives. However, the fundamental cultural changes that need to take place in medicine, lack of technology adoption and steep investment that must be made to fully implement CDS, are high barriers to adoption.

According to Datamonitor, medical culture will be the most difficult barrier to CDS adoption. The idea that a computer could be more accurate than a physician is difficult for providers to accept despite numerous studies which have shown that algorithms and computers do outperform most doctors on some tasks.

CDS tools of the future will be patient-centric and focus on diagnosis
As the healthcare system becomes more patient-centric, CDS will as well. This focus on individual patients will be evident in a number of ways. For example, alerts and reminders will be personalized to each patient and genetic information will be included in patient records. Patients will even use CDS tools themselves to help aid in their own diagnosis and treatment.

Additionally, today's CDS solutions focus on aiding providers with their treatment plans for patients. While this is necessary, CDS should also be used by providers to help diagnose patients. Misdiagnoses occur often in medicine; furthermore, the correct diagnosis is sometimes not reached until multiple incorrect diagnoses have been tested. With the technology and information available today, providers should not be complacent with the current misdiagnosis rate. Chang states, "if Facebook is able to predict who an individual might be friends with based on who he/she is already friends with, why shouldn't CDS be able to determine what diagnosis patients may have based on their health information?" Despite the need to improve in this area, the use of CDS tools for diagnosis and computer-assisted diagnosis (CAD) technologies will be slow until a fundamental change in medical culture occurs.

The amount of clinical data available for research will grow exponentially with the greater adoption of EHRs, but the full value of the information collected will not be reached unless healthcare practitioners have the tools to analyze it.

Implementing most CDS tools is no easy feat. Even the low hanging fruit is difficult to pick, particularly if the technology is not user friendly. CPOE with CDS, for example, is now already widely accepted, but rarely used appropriately. The number one complaint physicians have regarding CPOE with CDS is that too many inappropriate alerts pop up on the computer screen. Providers begin to ignore the alerts, even the correct ones, negating the reason why the alerts were set up in the first place. Alerts and reminders need to be accurate, relevant to the patient, unobtrusive to the provider's workflow and quick to use. Tracking how alerts are used and which are over-ridden may be the most valuable information for early adopters to share with their peers.

Similarly, CI solutions face their own set of implementation problems. CI may be great at analyzing data, but if the data the analysis was built upon is incorrect, then the validity of the research is compromised. Thus, the quality of the data being entered into EHRs must be verified. Finally, without interoperability, the capabilities of CDS will be limited as CI and alerts will not have the information they need to work well. The biggest barrier to interoperability is not, unfortunately, technology, but convincing all the stakeholders to participate, which is a much more difficult task to overcome.



write your comments about the article :: © 2008 Computing News :: home page