Trends in Emergency Department Visits Among Medicaid Patients

To the Editor: Dr Tang and colleagues¹ concluded that emergency department (ED) visit rates have been increasing most among Medicaid patients. We believe this conclusion may be unwarranted. The analysis does not appear to have incorporated changes that occurred over the years in the coding of the variable “primary expected source of payment” in the data source used by the authors, the National Hospital Ambulatory Medical Care Survey (NHAMCS) conducted by the National Center for Health Statistics (NCHS). The only information on payment sources that NHAMCS collected from 1997 through 2004 was the primary expected source of payment.² Starting in 2005, hospital respondents could indicate multiple expected sources of payment per visit and did not designate a primary source. From 2005 through 2007, the data files included separate variables for each expected payment source plus a variable that assigned the primary expected source using a hierarchy with Medicaid first, followed by Medicare, private insurance, worker's compensation, self-payment, and no charge.

We suspect that many hospital respondents from 1997 through 2004 reported Medicare as the primary expected payment source for patients dually eligible for Medicare and Medicaid. It appears that for 2007, Tang et al used NHAMCS’ hierarchical payment variable that classified dual eligibles as having Medicaid. This approach led to a report of ED visit rates between 1997 and 2007 that increased by 36.5% for adult patients with Medicaid and decreased by 2.5% for Medicare (Table in the article). We recomputed the number of visits by coding the primary payment source for dually eligible patients as Medicare instead of Medicaid for 2007. Using information in the Table¹ to generate denominators for the visit rate, the estimated number of ED visits for Medicaid patients would be 14.2 million and the visit rate 759.9 per 1000; for Medicare, the number of visits would be 19.9 million and the visit rate 485.9 per 1000. Therefore, the visit rate for adults from 1997 through 2007 would increase 9.5% for Medicaid and 17.6% for Medicare. These latter estimates may be more realistic than those presented by Tang et al.

The NCHS has added further explanatory notes to the NHAMCS documentation for 2005 through 2007. Beginning with 2008 data, NCHS has also modified the hierarchical payment variable so that dual eligibles are classified as having Medicare as the primary payment source, which better reflects the source of payment amounts to hospitals from these patients. We hope these changes better support future analyses.