Professional Documents
Culture Documents
Aldridge M. Migth Cost 2015
Aldridge M. Migth Cost 2015
Aldridge M. Migth Cost 2015
December 2015, Vol 105, No. 12 | American Journal of Public Health Aldridge and Kelley | Peer Reviewed | End-of-Life Care | 2411
COMMENTARY
2412 | End-of-Life Care | Peer Reviewed | Aldridge and Kelley American Journal of Public Health | December 2015, Vol 105, No. 12
COMMENTARY
fact that longitudinal data repre- for our aging population are For example, a recent com- a specic intervention. For exam-
senting all payers in our health potentially profound.28 mentary2 suggested that an ple, using data regarding the pop-
care system do not exist in a single estimated 22% of health care ulation with chronic conditions
population-based source, and thus POPULATION-LEVEL expenditures are related to poten- and functional limitations30 and
we synthesized data and estimates INTERVENTIONS tially avoidable complications, our estimates with respect to the
from multiple sources for this such as hospital admissions for population at the end of life, we
analysis. First, note that MEPS A substantial proportion of the patients with diabetes with ketoa- can imagine a hypothetical inter-
data are limited to the civilian high-cost group has persistently cidosis or amputation of gangre- vention and 3 potential target
noninstitutionalized US popula- high costs,17,25,26 underscoring the nous limbs or for patients with populations: individuals with
tion. To overcome this omission potential scal impact of targeting congestive heart failure and chronic conditions and functional
of the LTC population, we relied this group with high-value shortness of breath due to uid limitations, older adults with
on other sources of health care (i.e., high-quality and lower-cost) overload.2,31 Furthermore, the chronic conditions and functional
expenditure data for nursing interventions. The challenge re- disproportionally higher costs for limitations, and individuals at the
home residents.1,20 sides in appropriate prospective this group may reect a lack of end of life. If we assume, for in-
Second, our estimates of end- identication of this group. A adequate community-based care stance, that the percentage of the
of-life costs are based on the fee- recent analysis of US health care and supportive services, leaving target population that will be af-
for-service Medicare beneciaries spending revealed that chronic patients with no alternative but to fected by the intervention is 50%
included in the Health and Re- illnesses account for 84% of total access the acute care hospital sys- and that the potential reduction in
tirement Study. We hypothesize health care costs, but the group tem by calling 911 or visiting the costs is 10%, we can compare
that this may have biased our with chronic illnesses is extremely emergency department. between-intervention cost savings,
end-of-life cost estimates toward large.29 Some evidence suggests as shown in Table 1.
higher results, a conservative that the presence of functional ESTIMATED EFFECTS An intervention that targets all
approach for our study and one limitations in conjunction with OF TARGETED individuals with chronic conditions
that may have overestimated the chronic conditions is a better pre- INTERVENTIONS and functional limitations (an esti-
proportion of decedents among dictor of high health care costs than mated 45 million people), affects
the most expensive 5% of the number of chronic conditions In our opinion, our analysis half of that population, and reduces
patients. alone.30 The combination of suggests that identication of the costs by 10% will theoretically
Finally, our cost analyses did chronic conditions and functional appropriate target population for achieve nearly double the reduc-
not include estimates for costs limitations may be associated with high-quality, cost-saving interven- tion in health care costs ($45
such as informal caregiving and higher health care costs owing to tions is critical given the substan- billion vs $27 billion) of an identi-
lost wages. Consideration of these the complexity of care coordination tial variation in the size of different cal intervention that targets only
costs must be included in the across multiple providers and set- target populations, the costs gen- adults aged 65 years and older
context of any new or reformed tings, increased use of specialists erated by different populations, with chronic conditions and func-
health service design, as the eco- and procedures, or an increased and the proportion of the target tional limitations (22 million peo-
nomic implications of these costs likelihood of hospitalization. population likely to be affected by ple). If the intervention targets all
Source. The percent distributions of populations and costs by age and chronic condition/functional limitation categories were obtained from The Lewin Group ; total population and health care
30
costs were obtained from 2011 Medical Expenditure Panel Survey data21 adjusted to include the nursing home population.1,20,24
December 2015, Vol 105, No. 12 | American Journal of Public Health Aldridge and Kelley | Peer Reviewed | End-of-Life Care | 2413
COMMENTARY
individuals at the end of life (2.5 in the highest cost group are in Medicine Committee on Approaching 13. Lubitz JD, Riley GF. Trends in
Death. Medicare payments in the last year of life.
million), affects half of that popu- their last year of life. Efforts to
N Engl J Med. 1993;328(15):1092---
lation, and reduces costs by 10%, it improve the quality of care for this 1096.
will theoretically achieve less than group are clearly warranted25; References
1. Centers for Medicare & Medicaid 14. Riley GF, Lubitz JD. Long-term
a quarter of the reduction in health however, expecting such interven- Services. National health expenditure trends in Medicare payments in the last
year of life. Health Serv Res. 2010;45
care costs ($10 billion vs $45 tions, if limited to those at the end data. Available at: http://www.cms.gov/
Research-Statistics-Data-and-Systems/ (2):565---576.
billion) of an identical intervention of life, to have a meaningful impact
Statistics-Trends-and-Reports/ 15. Riley GF. Long-term trends in the
that targets individuals with on overall health care costs is mis- NationalHealthExpendData/index.html. concentration of Medicare spending.
chronic conditions and functional guided. Not only is this group small, Accessed September 12, 2015. Health Aff (Millwood). 2007;26(3):
limitations (45 million people). but the window of time for a sig- 808---816.
2. Emanuel EJ. Where are the health care
This projection highlights the nicant impact on costs is limited cost savings? JAMA. 2012;307(1):39---40. 16. Reschovsky JD, Hadley J, Saiontz
Martinez CB, Boukus ER. Following
fact that interventions aimed at by the patients life expectancy. 3. Congressional Budget Ofce. The
the money: factors associated with
long-term outlook for health care spend-
individuals in their last year of life Furthermore, our ndings con- ing. Available at: http://www.cbo.gov/
the cost of treating highcost Medicare
will generate smaller reductions in rm the need to focus on those beneciaries. Health Serv Res. 2011;46
ftpdocs/87xx/doc8758/MainText3.1.
(4):997---1021.
cost savings than interventions with chronic serious illnesses, shtml. Accessed September 12, 2015.
17. Cohen S, Yu W. The concentration
that target those with chronic functional debility, and persistently 4. Davis K, Buttorff C, Leff B, et al.
and persistence in the level of health
Innovative care models for high-cost Medi-
conditions and functional limita- high costs. Although many pro- expenditures over time: estimates for
care beneciaries: delivery system and
tions given the signicantly grams have begun to target the U.S. population, 2008---2009.
payment reform to accelerate adoption. Am
Available at: http://meps.ahrq.gov/
smaller size of the end-of-life pop- segments of this population or J Manag Care. 2015;21(5):e349---e356.
mepsweb/data_les/publications/
ulation and the limited time frame evaluate health system and reim- 5. Brown RS, Peikes D, Peterson G, st354/stat354.shtml. Accessed September
Schore J, Razandrakoto CM. Six features 12, 2015.
for cost reduction. In addition, bursement models that may incen-
of Medicare coordinated care demon-
given the complexity of identifying tivize higher-value care,4--10 outcome stration programs that cut hospital ad-
18. Shugarman LR, Campbell DE, Bird
CE, Gabel J, Louis TA, Lynn J. Differences
individuals in their last year of life data and cost-effectiveness ana- missions of high-risk patients. Health Aff
in Medicare expenditures during the last
relative to identifying individuals lyses are still forthcoming. Cer- (Millwood). 2012;31(6):1156---1166.
3 years of life. J Gen Intern Med. 2004;
with chronic conditions and func- tainly, the greatest strides in 6. Peikes D, Chen A, Schore J, Brown R. 19(2):127---135.
Effects of care coordination on hospitali-
tional limitations, an end-of-life improving the quality and contain- 19. Fassbender K, Fainsinger RL, Carson
zation, quality of care, and health care
M, Finegan BA. Cost trajectories at the
intervention may have an even ing the costs of health care for the expenditures among Medicare benecia- end of life: the Canadian experience. J
smaller impact on costs than we highest cost population will be ries: 15 randomized trials. JAMA. Pain Symptom Manage. 2009;38(1):
2009;301(6):603---618. 75---80.
have described because it would achieved by focusing research and
7. Smith SM, Soubhi H, Fortin M, 20. Sing M, Banthin JS, Selden TM,
likely affect less than 50% of the clinical interventions on this
Hudon C, ODowd T. Managing patients Cowan CA, Keehan SP. Reconciling med-
terminal population. vulnerable and complex group. j with multimorbidity: systematic review of ical expenditure estimates from the MEPS
interventions in primary care and com- and NHEA, 2002. Health Care Financ
munity settings. BMJ. 2012;345:e5205.
CONCLUSIONS About the Authors Rev. 2006;28(1):25---40.
8. Fisher ES, Wennberg DE, Stukel TA, 21. Agency for Healthcare Research and
The authors are with the Department of
Gottlieb DJ, Lucas FL, Pinder EL. The Quality. Medical Expenditure Panel
Maximizing value (i.e., increasing Geriatrics and Palliative Medicine, Ichan
implications of regional variations in Survey. Available at: http://meps.ahrq.
School of Medicine at Mount Sinai, New
quality while reducing costs) in the York, NY, and the James J. Peters VA
Medicare spending. Part 1: the content, gov/mepsweb. Accessed September 12,
care of high-cost, seriously ill in- Medical Center, Bronx, NY. quality, and accessibility of care. Ann 2015.
Correspondence should be sent to Melissa Intern Med. 2003;138(4):273---287.
dividuals is a major public health 22. Aldridge MD, Kelley AS. Epidemi-
D. Aldridge, PhD, MBA, Department of 9. Fisher ES, Wennberg DE, Stukel TA, ology of serious illness and high utiliza-
challenge facing the nations health Geriatrics and Palliative Medicine, Icahn Gottlieb DJ, Lucas FL, Pinder EL. The tion of health care. Available at: http://
care system and economy. Many School of Medicine at Mount Sinai, One implications of regional variations in www.nap.edu/read/18748/chapter/14.
proposals to reduce health care Gustave L. Levy Place, Box 1070, New York, Medicare spending. Part 2: health out- Accessed September 12, 2015.
NY 10029 (e-mail: melissa.aldridge@mssm. comes and satisfaction with care. Ann
costs in the United States target the edu). Reprints can be ordered at http://www. Intern Med. 2003;138(4):288---298.
23. Kelley AS, Ettner SL, Morrison RS,
Du Q, Wenger NS, Sarkisian CA. Deter-
high cost of end-of-life care, yet at ajph.org by clicking the Reprints link.
10. Frank RG. Using shared savings to minants of medical expenditures in the
the population level the cost of This article was accepted August 25,
foster coordinated care for dual eligibles. last 6 months of life. Ann Intern Med.
2015.
caring for individuals in their last N Engl J Med. 2013;368(5):404---405. 2011;154(4):235---242.
year of life accounts for only 13% 11. Hogan C, Lunney J, Gabel J, Lynn J. 24. National Center for Health Statistics.
Contributors Medicare beneciaries costs of care in the Health, United States, 2012: Table 109.
of total annual health care spend- The authors contributed equally to the last year of life. Health Aff (Millwood). Nursing homes, beds, residents, and oc-
ing. That is, although the majority study concept and design, analysis and 2001;20(4):188---195. cupancy rates by state: United States,
of decedents are in the highest cost interpretation of data, and the writing of selected years 1995---2011. Available at:
12. Hoover DR, Crystal S, Kumar R,
the article. http://www.cdc.gov/nchs/data/hus/
group, the majority of individuals Sambamoorthi U, Cantor JC. Medical ex-
penditures during the last year of life: hus12.pdf#109. Accessed September 12,
in that group are not in their last 2015.
Acknowledgments ndings from the 1992---1996 Medicare
year of life. Specically, we esti- This study was based on a commissioned Current Beneciary Survey. Health Serv 25. Institute of Medicine. Dying in
mate that only 11% of individuals paper prepared for the Institute of Res. 2002;37(6):1625---1642. America: Improving Quality and Honoring
2414 | End-of-Life Care | Peer Reviewed | Aldridge and Kelley American Journal of Public Health | December 2015, Vol 105, No. 12
COMMENTARY
December 2015, Vol 105, No. 12 | American Journal of Public Health Aldridge and Kelley | Peer Reviewed | End-of-Life Care | 2415