In 2004 President Bush put forth a Health Information Technology Plan calling for widely available electronic health records within 10 years, arguing that patients would greatly benefit from having their full health record easily available to all of their physicians.¹ If one examines the current state of health records, this is quite an ambitious initiative.

A typical medical record today is a hybrid including paper physicians' or nursing notes, various reports, digitized demographics, insurance data, and images such as CT scans, x-rays or MRIs.² Records for a single patient may be widely scattered between various offices and hospitals. To compound the problem, each entity is often unable to access the records of another in a timely manner.

Why is it so difficult to combine all of a person's medical history into one accessible yet private digital package, and why is there hesitance in implementing technologies toward this goal?

The answer is complex. The first and most obvious barrier is the cost of the system, which must be entirely absorbed by the health system or individual office. This up-front cost includes a large time investment for training both the physician and the ancillary staff, and there is often a significant change in work flow.

There is also a looming concern for patient privacy. Web-based information is notoriously insecure, even with the latest encryption and firewall technologies. Isolated and locally maintained databases don't provide the necessary availability to make them viable alternatives. The currently available software is cumbersome and often requires extensive customization for a particular clinical setting.

Because physicians' offices vary so widely in their specializations and protocols, packaged software may have unnecessary features or may lack something basic to the practice. Perhaps the most difficult is the challenge of bridging the gaps between multiple data sources, often including ambiguous and non-standardized handwritten physician notes.^{3, 4}

While it may seem like the typical patient chart is an antiquated mess, computerization is certainly not new to healthcare. Computers have been used for many years to collect demographic information and generate bills to insurance companies. This early effort toward digitization was driven largely by economics; an electronically submitted claim yields a much speedier payment than a paper claim.

Laboratories use their own computer systems to record and analyze results, as do radiology offices. Pharmacists use computers to help track potentially dangerous combinations of medications. Where the system has yet to succeed is in pulling all of the relevant information together, and in collecting the clinical information from a doctor-patient encounter in a meaningful way.

Will this really be the decade where we overcome these obstacles? As a computer-savvy generation of health professionals comes into practice, they will likely demand better access to better records.

In addition, these professionals have been educated to accept more standardization and structure in clinical thinking, a trait necessary to for creating and maintaining a truly useful electronic record.⁵ As health care organizations assess the technologies supporting a transformation to electronic patient records, clinicians must advocate preserving the practice of medicine in the best interests of the patient. Anything less is an unacceptable compromise.

• CIN: Computers, informatics, nursing (available electronically through OVID or in print in the Scott Library)
• Healthcare informatics (also in print at Scott Library)
• Journal of the American Medical Informatics Association (available in print at Scott Library)

Send us your comments.