Concentration of health spending in a given period of time typically is found among a small portion of the population, for example 10% of Medicare patients account for half of spending in that program. Commercial spending is even more concentrated. Identifying where these patients are might be helpful in designing interventions to lower spending, so research carried in the American Journal of Managed Care attempts to ascertain whether these patients tend to be clustered in certain hospitals or geographic markets. (AJMC Article) The researchers used Medicare fee-for-service data from 2011 and 2012. Adjusting costs so that they would equal across regions, the researchers then identified high-spend beneficiaries, stratifying them by decile. Hospital relationships with high-cost patients were examined because hospital use tends to be the largest category of care and spending for these high-cost beneficiaries. The percentage of claims from high-cost (top decile) patients was calculated for each hospital in the study. The researchers also calculated the percentage of high-cost beneficiaries in each hospital referral region.
Hospitals and hospital referral regions were considered high concentration of high-cost patients if their percentage of these patients was in the top 10% of all hospitals or HRRs, respectively. Since many high-cost patients turned out to be end-of-life cases, the authors ran their analyses with and without these patients. In their analysis, the researchers found that the top 10% of beneficiaries accounted for 55% of all spending and their average annual cost was ten times higher than the average cost for the remaining 90% of beneficiaries. These patients were somewhat more likely to be African-American, twice as likely to be Medicaid-eligible and had higher rates of chronic disease. These patients obviously accounted for a large number of Medicare claims at every hospital. The highest decile of hospitals had 69% of their claims from high-cost beneficiaries, compared to only 33% in the lowest decile. But overall concentration is low; the highest decile hospitals in total accounted for only 14% of all Medicare hospital spending, modestly above the 10% that would be expected if the distribution was even across all hospitals. Similar modest concentration was shown in the geographic analysis by HRR. In the top decile of HRRs by concentration of high-cost patients, only 9.7% of Medicare spending occurred, slightly below the 10% that would result from an even distribution.
Hospitals in the top decile for high-cost patients were more likely to be for-profit and to be teaching hospitals. They were more likely to get disproportionate share payments and to have lower patient/nurse staffing ratios. They had higher 30-day readmission rates but lower 30-day mortality rates. The HRRs with the greatest concentration of high-cost patients had a higher percent of African-Americans in the population, more poverty and were more likely to be urban. They also had more specialists and a lower-supply of hospital beds. These high HRRs had both high costs of care and high utilization. Their end-of-living spending was over 20% higher than for HRRs in the other deciles. Overall, the results suggest a limited concentration of high-cost patients by hospital or HRR, but it may be enough to warrant some additional focus.