Questions of Trust

Everyone agrees that any system of networked medical records must be trustworthy. The medical information in such a system must be complete, free from error and secure, and the confidentiality of the material must be maintained. But how free from error will the conversion of current paper-based medical records to an electronic format be? One study reported a 10.2 percent error rate in transcribing just 2,098 children’s vaccine records from paper to electronic records. Other studies indicate that a 2 percent to 5 percent transcription error rate is not unrealistic. Who will be responsible for finding and fixing these types of errors? The doctor? The patient? If someone makes a typo, who would know the difference between Flomax or Volmax?

The need for maintaining privacy is another requirement of the medical network that everyone agrees must be built in from the start. However, even the Department of Defense has a hard time keeping its networks safe from hacking. Even Guidance Software, a leading company that sells forensic software to detect hackers, recently got hacked. Given that 80 percent of U.S. doctors are in practices of eight or fewer physicians, can we expect them to install and keep updated all the security controls required to keep EHRs private?

Psychiatrists especially are alarmed that their patients may not want their records put into EHRs because of the risk of disclosure. Unfortunately, mental illness and even temporary behavioral ailments still carry a societal stigma, and can hinder employment and some types of insurance coverage. If psychiatric patient records are included in EHRs, patients may decide to stop their treatments. Other patients may decide to opt out as well because of the potentially embarrassing information being captured. Are we going to end up with a medical digital divide as a result?

Even assuming a secure network, the availability of EHRs may encourage not-so-subtle medical redlining by employers and insurance companies. For instance, some insurance companies and employers are promising not to use DNA results in determining eligibility for their health care or benefits plans. The availability of cheap, fast and widespread DNA testing in less than a decade makes such promises suspect. And keep in mind that with access to EHRs and data mining, it is possible to create a genetic profile of a person without ever resorting to DNA.

Who’s in Charge?

Another risk that gives me pause is the number of players involved in creating the national EHR network. Not just the 650,000 doctors, 5,800 hospitals, and countless other health-care facilities, but all the players who wish to be involved, from standard bodies to medical associations to EHR suppliers. As of this time, there are 18 bills related to EHRs being contemplated in Congress, each with slightly (or greatly) different consequences for a national system. It has gotten so bad that the HHS has had to politely ask Congress not to be so "helpful."