Estimated cost savings to hospitals from the use of electronic health records are quite large: For example, a RAND study published in 2005 calculated that the inpatient savings from efficiency alone could total $57 billion annually within 15 years, assuming a 90 percent adoption rate. But those savings included only $2.4 billion of annual savings in medical record handling, which one might assume would be greatly reduced after the introduction of EHRs.
One reason why paper record costs might not decline much with EHR adoption is that state laws require hospital records to be archived for long periods of time. For example, Massachusetts mandates that all of a patient's records be stored for 25 years after the last time they were used. While state requirements vary quite a bit--Michigan's is only seven years, for instance, for adult records--all of this archiving is quite costly. Iron Mountain, a data storage and protection company that claims to work with about a third of the nation's hospitals, estimates that the 6,000 hospitals in the U.S. spend about $10 billion a year to store and manage 500 million patient records.
Ed Santangelo, senior vice president of healthcare at Iron Mountain, says that most hospitals do not scan in paper charts when they implement an EHR, as many physician practices do. "The cost to convert historical files would be outrageously expensive," he says, citing a $1 billion estimated price tag for Johns Hopkins Hospital alone.
Iron Mountain has two suggestions for resolving the dilemma: First, the federal government, as part of its effort to stimulate the adoption of EHRs, should set a uniform national standard of 10 years for the retention of adult hospital records and 10 years after a patient's age of majority for pediatric records. The firm maintains this would cut the amount of archived records by 35 percent. Second--and this would obviously benefit the company--Iron Mountain proposes that hospitals centralize their archived records to gain better control over document destruction and reduce the labor required to retrieve old charts.
Requests for archival information fall off dramatically over time, Santangelo notes. In an adult acute-care hospital, for instance, 85 percent of the requests for radiology records occur in the first two years after they were created. Few requests for other medical records will be made after five years. In pediatrics, most of the requests occur within five years for radiology and 12 years for general medical records.
The longest intervals for information requests typically come from anesthesiologists, he adds. Because of their high likelihood of being sued, anesthesiologists are very risk-averse. Even if a patient insists they have never had an adverse reaction to anesthesia drugs, the sedating physicians want to look back through the old records to cover themselves.
Santangelo claims that with the 10-year limit and best-practice storage policies, hospitals could generate much of the money they need to buy EHRs. Whether or not that's true, it does seem like a waste of the nation's health-care resources to archive records far longer than anyone needs them.