Medicare help line needs improvement, GAO finds

Audit shows that fewer than two-thirds of callers got complete and accurate answers.

Less than two-thirds of the calls to Medicare's toll-free hot line, 1-800-MEDICARE, last summer resulted in complete and accurate answers to questions about the program, according to a Government Accountability Office report published Wednesday.

GAO conducted an audit of the 24-hour help line in July. Researchers selected six common questions to ask customer service representatives and posed each question 70 times. Representatives working in call centers provided complete, accurate answers to 256 out of 420 calls, or 61 percent.

Representatives gave inaccurate or misleading information in response to 29 percent of the inquiries. Researchers were not given answers on 10 percent of the calls. Calls were disconnected, transferred to offices that were not open or routed to wrong numbers. Other attempts to ask questions failed because a Web-based tool upon which representatives rely was temporarily unavailable.

Mark McClellan, administrator for the Centers for Medicare and Medicaid Services, pointed out in a response to the report (GAO-05-130) that the audit took place during a period of adjustment. The 2003 Medicare Prescription Drug, Improvement, and Modernization Act caused the number of calls to skyrocket because of questions about prescription discount cards, which went live in June, and a Medicare drug benefit, which takes effect next month.

CMS had expected to receive about 7 million calls in fiscal 2004, but nearly 4 million calls came in during May 2004 alone. "We believe we responded as well as we reasonably could, given the unique and demanding circumstances," McClellan wrote.

In response to the volume of calls, CMS added more than 800 representatives in the first six months of 2004. They received two weeks of classroom training and took written and practical exams. An understanding of Medicare programs is not a requirement for hiring. On the job, representatives can access nearly 700 scripts that provide program information language that is supposed to be easy to understand.

McClellan said that the high number of calls is evidence of seniors' great interest in drug discount cards. It is also proof that seniors have many questions about the cards. There are more than 70 options to choose from, and each offers different deals on different drugs at different pharmacies. Enrollment in the program has lagged far behind expectations, even though the cards provide low-income seniors with a $600 yearly drug benefit.

Despite the complexity, Medicare representatives scored well on questions about how to choose the discount cards. In 76 percent of the calls, they provided the name of at least one card with the lowest total price for the drugs the caller listed. The representatives also were able to answer 81 percent of questions about whether seniors who have Medicare coverage and a private Medicare supplement plan are eligible for the drug discount cards.

On a more complicated question, in which representatives had to assess whether an individual's income would qualify for the $600 credit associated with the drug cards, representatives gave the correct answer on only 20 percent of the calls. They needed to read one script to determine the maximum income amount allowed, and then read a second script to determine which kinds of income are counted. Most failed to look at the second script, and therefore didn't realize that the income described does not disqualify an individual for the credit.

Problems using the scripts accounted for most of the inaccurate responses, GAO concluded. Some scripts contain confusing language. In answering a question about coverage for power wheelchairs, one representative misunderstood the term "trunk strength," which was meant to indicate upper-body strength, for the trunk capacity in a car. Other representatives didn't realize that a script was available to help them answer certain questions, failed to choose the most useful script, or didn't read enough information from the script.

GAO recommended better training and evaluation for representatives, testing of scripts, and monitoring of questions that frequently cause problems. McClellan basically agreed with the recommendations and said that CMS has resolved some of the issues and taken steps to address others. A spokesperson for Pearson, the Arlington, Va., contractor that manages the majority of Medicare's call centers, said the company already has addressed all of the report's recommendations.