Regional and Age Group Differences in Sickle Cell Anemia Hospitalization Rates among African-American Adults: 1990-2002.

Background: In 1995, investigators from the Multicenter Study of Hydroxyurea in Sickle Cell Anemia (MSH) published data demonstrating that patients on hydroxyurea (HU) experienced fewer hospitalizations, fewer incidences of acute chest syndrome, and required fewer blood transfusions than patients not receiving HU. The objective of this study was to look at national trends in sickle cell anemia (SCA) admissions since the publication of the MSH results.

Methods: We examined data from the National Hospital Discharge Survey (NHDS) for the period 1990–2002. The NHDS is a nationally representative survey of non-federal short-stay hospitals in the U.S, which allows for the estimation of inpatient utilization data. Inclusion criteria for our analysis were African American race, age 18+ years, and the following ICD9 codes: 28261 and 28262 (sickle cell anemia with or without crisis). To control for population growth, denominators for the calculation of hospitalization rates were estimated using U.S. Census population estimates included in the NHDS dataset, and a SCA prevalence of 1 in 600 AA. Negative binomial regression models were used to test the statistical significance of trends in SCA hospitalization rates overall, and stratified by age group (18 to 23, 24 to 29, 30 to 35, 36 to 41, and 42+) and region (Northeast, Midwest, South, and West). A linear spline term was created for the year 1996 (one year after MSH publication) to compare annual changes in mean hospitalization rates for the periods pre (1990–1996) and post (1997–2002) MSH publication.

Results: In the overall analyses, the hospitalization rate increased by 1.5% annually from 1990 to 2002 (p=0.062). The pre/post analysis for all patients found no statistical differences in the change in hospitalization rates pre (0.9% increase, p = 0.626) or post (1.2% increase, p = 0.729) MSH publication. In the stratified age group analyses, we found that 24 to 29 year olds experienced an overall mean annual increase in hospitalization rates of 7.1% (p < 0.001). However, when viewed pre/post MSH, we found that 24 to 29 year olds experienced a 14% mean annual increase in rates pre-MSH (p < 0.001), and a 12% mean annual decrease in rates post-MSH (p = 0.003). The reverse trend was true for 36 to 41 year olds, who experienced a mean annual decrease in rates of 17% pre-MSH (p < 0.001), and mean annual increase of 36% post-MSH (p < 0.001). No statistical differences in annual rates of change were found for the other age groups. Regionally, we found that the South experienced a 7% mean annual increase in its relative hospitalization rates (p < 0.001) for the entire time period. The South experienced a 2.5% mean annual increase in rates pre-MSH (p = 0.156), and a 9% annual increase in rates post-MSH (p = 0.006). The West saw a 7% decrease in rates for the overall time period (p = 0.001). The Midwest experienced a 6% mean annual increase in rates pre-MSH (p = 0.058), and an 11% decrease in mean annual rates post-MSH (p = 0.034).

Conclusions: Estimates of national SCA hospitalization rates for black adults, after MSH publication suggests that there has not been an overall decrease in hospitalization rates when patients and the country are examined as a whole. However, there are important age group and regional differences that need to be explored further to better understand the impact HU has had on this population.