West AN (White River Jct Research Enhancement Award Program (REAP)), West PA
(White River Jct REAP), Weeks WB
(White River Jct REAP)
To determine whether VA enrollees living beyond urban areas rely more on non-VA hospitalizations paid by Medicare than urban enrollees do, as well as whether their relative reliance differs depending on age group (younger or older than 65) or on major diagnostic category treated.
Data included all hospitalizations in VA’s Patient Treatment Files and in the VA-Medicare discharge files for the years 1997 through 2004. Each hospitalization was classified according to major diagnostic category and age at admission; the patient’s ZIP code of residence also was classified as urban or one of two rural levels using Rural-Urban Commuting Area codes. For each age group separately, PTF and VA-Medicare admission counts for urban and rural veterans, overall and within each of 23 major diagnostic categories (excluding Pregnancy and Newborn MDCs for low N), were compared using chi square analysis.
VA enrollees 65 or older had 4.5 times as many Medicare as VA hospitalizations; enrollees living in areas of greater or less “ruralness” both had over 5 times as many Medicare as VA admissions, a significantly higher ratio than for urban residents (4.2; p < .0001). Chi-squares were highly significant for 21 MDCs, with 20 showing greater reliance on Medicare among rural enrollees; for Neoplasms, however, reliance on Medicare was greater for urban veterans (p < .0001). Among VA enrollees younger than 65, urban residents had 2.2 times as many VA as Medicare admissions, while for residents of less rural areas the ratio was 1.8 and for more rural veterans it was 1.6 (p < .0001). Twenty-two MDCs yielded significant differences, with 20 showing greater reliance on Medicare among rural enrollees; for Mental Disorders (p < .0001) or HIV (p < .001), however, reliance on Medicare was greater for urban veterans.
For most major diagnostic categories, VA enrollees with Medicare are likely to use non-VA hospitals, particularly if they live beyond urban areas. Rural veterans, however, may have limited access to non-VA hospitals for inpatient cancer, psychiatric, or HIV care.
Planning for the hospitalization needs of VA enrollees should include analyses of all their non-VA utilization, broken down by where veterans live.