[Skip to Content]
Access to paid content on this site is currently suspended due to excessive activity being detected from your IP address 54.163.92.62. Please contact the publisher to request reinstatement.
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
Download PDF
Figure. Preventable and Nonpreventable Inpatient Costs Among High-Cost Patients, Medicare 2010
Figure. Preventable and Nonpreventable Inpatient Costs Among High-Cost Patients, Medicare 2010

Dark gray bars indicate conditions that are considered preventable; light gray bars indicate those that are considered nonpreventable. GI indicates gastrointestinal; TIA, transient ischemic attack; COPD, chronic obstructive pulmonary disease.

Table 1. Patient Characteristicsa
Table 1. Patient Characteristicsa
Table 2. Preventable and Nonpreventable Emergency Department Costs, Medicare, 2010
Table 2. Preventable and Nonpreventable Emergency Department Costs, Medicare, 2010
Table 3. Preventable and Nonpreventable Inpatient Costs, Medicare, 2010
Table 3. Preventable and Nonpreventable Inpatient Costs, Medicare, 2010
Table 4. Inpatient Hospitalizations Among Persistently High-Cost Patients
Table 4. Inpatient Hospitalizations Among Persistently High-Cost Patients
Table 5. Regional Variation in Preventable Spending Per Capita for High-Cost Patientsa
Table 5. Regional Variation in Preventable Spending Per Capita for High-Cost Patientsa
1.
Niefeld MR, Braunstein JB, Wu AW, Saudek CD, Weller WE, Anderson GF. Preventable hospitalization among elderly Medicare beneficiaries with type 2 diabetes.  Diabetes Care. 2003;26(5):1344-1349PubMedArticle
2.
Riley GF. Long-term trends in the concentration of Medicare spending.  Health Aff (Millwood). 2007;26(3):808-816PubMedArticle
3.
Smulowitz PB, Lipton R, Wharam JF,  et al.  Emergency department utilization after the implementation of Massachusetts health reform.  Ann Emerg Med. 2011;58(3):225-234PubMedArticle
4.
Ballard DW, Price M, Fung V,  et al.  Validation of an algorithm for categorizing the severity of hospital emergency department visits.  Med Care. 2010;48(1):58-63PubMedArticle
5.
Billings J, Anderson GM, Newman LS. Recent findings on preventable hospitalizations.  Health Aff (Millwood). 1996;15(3):239-249PubMedArticle
6.
Oster A, Bindman AB. Emergency department visits for ambulatory care sensitive conditions: insights into preventable hospitalizations.  Med Care. 2003;41(2):198-207PubMed
7.
Braunstein JB, Anderson GF, Gerstenblith G,  et al.  Noncardiac comorbidity increases preventable hospitalizations and mortality among Medicare beneficiaries with chronic heart failure.  J Am Coll Cardiol. 2003;42(7):1226-1233PubMedArticle
8.
Congressional Budget Office.  Lessons from Medicare's Demonstration Projects on Disease Management, Care Coordination, and Value-Based Payment. Washington, DC: Congressional Budget Office; 2012
9.
Dartmouth Institute for Health Policy and Clinical Practice.  The Dartmouth Atlas of Health Care. 2012. http://www.dartmouthatlas.org/. Accessed July 22, 2012
10.
Garber AM, MaCurdy TE, McClellan MB. Persistence of Medicare expenditures among elderly beneficiaries. In: Garber AM, ed. Frontiers in Health Policy Research. Vol 1. Cambridge: National Bureau of Economic Research, Massachusetts Institute of Technology; 1998:153-180
11.
Bhattacharya J, Garber AM, MaCurdy TE. The narrowing dispersion of Medicare expenditures 1997 to 2005. In: Wise DA, ed. Research Findings in the Economics of Aging. Chicago, IL: National Bureau of Economic Research, University of Chicago Press; 2010:387-407
12.
NYU Wagner.  NYU ED algorithm: background.  2000. http://wagner.nyu.edu/faculty/billings/nyued-background.php. Accessed March 13, 2013
13.
Wharam JF, Landon BE, Galbraith AA, Kleinman KP, Soumerai SB, Ross-Degnan D. Emergency department use and subsequent hospitalizations among members of a high-deductible health plan.  JAMA. 2007;297(10):1093-1102PubMedArticle
14.
Agency for Healthcare Research and Quality.  Prevention Quality Indicators overview. 2011. http://www.qualityindicators.ahrq.gov/modules/pqi_overview.aspx. Accessed January 15, 2013
15.
Basu J, Friedman B, Burstin H. Primary care, HMO enrollment, and hospitalization for ambulatory care sensitive conditions: a new approach.  Med Care. 2002;40(12):1260-1269PubMedArticle
16.
Bindman AB, Grumbach K, Osmond D,  et al.  Preventable hospitalizations and access to health care.  JAMA. 1995;274(4):305-311PubMedArticle
17.
Jiang HJ, Russo CA, Barrett ML. Nationwide Frequency and Costs of Potentially Preventable Hospitalizations, 2006. Rockville, MD: Agency for Healthcare Research & Quality; April 2009. HCUP Statistical Brief 72
18.
Brown RS, Peikes D, Peterson G, Schore J, Razafindrakoto CM. Six features of Medicare coordinated care demonstration programs that cut hospital admissions of high-risk patients.  Health Aff (Millwood). 2012;31(6):1156-1166PubMedArticle
19.
Cebul RD, Rebitzer JB, Taylor LJ, Votruba ME. Organizational fragmentation and care quality in the US healthcare system.  J Econ Perspect. 2008;22(4):93-113PubMedArticle
20.
Miller DC, Gust C, Dimick JB, Birkmeyer N, Skinner J, Birkmeyer JD. Large variations in Medicare payments for surgery highlight savings potential from bundled payment programs.  Health Aff (Millwood). 2011;30(11):2107-2115PubMedArticle
21.
Congressional Budget Office.  High-Cost Medicare Beneficiaries. Washington, DC: Congressional Budget Office; 2005
22.
Baicker K, Chandra A. Medicare spending, the physician workforce, and beneficiaries' quality of care.  Health Aff (Millwood). 2004;(suppl web exclusives)  W4-184-W4-197PubMed
Original Contribution
ONLINE FIRST
June 26, 2013

Contribution of Preventable Acute Care Spending to Total Spending for High-Cost Medicare Patients

Author Affiliations

Author Affiliations: Departments of Health Policy and Management (Drs Joynt, Gawande, and Jha) and Biostatistics (Dr Orav), Harvard School of Public Health, Cardiovascular Division (Dr Joynt) and Division of General Internal Medicine (Drs Orav and Jha), Department of Medicine, and Department of Surgery (Dr Gawande), Brigham and Women's Hospital, and VA Boston Healthcare System (Drs Joynt and Jha), Boston, Massachusetts.

JAMA. 2013;309(24):2572-2578. doi:10.1001/jama.2013.7103
Abstract

Importance A small proportion of patients account for the majority of US health care spending, and understanding patterns of spending among this cohort is critical to reducing health care costs. The degree to which preventable acute care services account for spending among these patients is largely unknown.

Objective To quantify preventable acute care services among high-cost Medicare patients.

Design, Setting, and Participants We summed standardized costs for each inpatient and outpatient service contained in standard 5% Medicare files from 2009 and 2010 across the year for each patient in our sample, and defined those in the top decile of spending in 2010 as high-cost patients and those in the top decile in both 2009 and 2010 as persistently high-cost patients. We used standard algorithms to identify potentially preventable emergency department (ED) visits and acute care inpatient hospitalizations. A total of 1 114 469 Medicare fee-for-service beneficiaries aged 65 years or older were included.

Main Outcomes and Measures Proportion of acute care hospital and ED costs deemed preventable among high-cost patients.

Results The 10% of Medicare patients in the high-cost group were older, more often male, more often black, and had more comorbid illnesses than non–high-cost patients. In 2010, 32.9% (95% CI, 32.9%-32.9%) of total ED costs were incurred by high-cost patients. Based on validated algorithms, 41.0% (95% CI, 40.9%-41.0%) of these costs among high-cost patients were potentially preventable compared with 42.6% (95% CI, 42.6%-42.6%) among non–high-cost patients. High-cost patients accounted for 79.0% (95% CI, 79.0%-79.0%) of inpatient costs, 9.6% (95% CI, 9.6%-9.6%) of which were due to preventable hospitalizations; 16.8% (95% CI, 16.8%-16.8%) of costs within the non–high-cost group were due to preventable hospitalizations. Comparable proportions of ED spending (43.3%; 95% CI, 43.3%-43.3%) and inpatient spending (13.5%; 95% CI, 13.5%-13.5%) were preventable among persistently high-cost patients. Regions with high primary care physician supply had higher preventable spending for high-cost patients.

Conclusions and Relevance Among a sample of patients in the top decile of Medicare spending in 2010, only a small percentage of costs appeared to be related to preventable ED visits and hospitalizations. The ability to lower costs for these patients through better outpatient care may be limited.

High and increasing health care costs are arguably the single biggest threat to the long-term fiscal solvency of federal and state governments in the United States. One compelling strategy for cost containment is focusing on the small proportion of patients in the Medicare programs who account for the vast majority of health care spending. We know from prior work that Medicare spending is highly concentrated: 10% of the Medicare population accounts for more than half of the costs to the program.1

By far the biggest sources of spending among high-cost beneficiaries are those related to acute care: emergency department (ED) visits and inpatient hospitalizations, which make up more than 55% of costs for this population.2 As a result, many interventions targeting high-cost patients have focused on case management and care coordination, aiming to prevent ED visits and hospitalizations for conditions thought amenable to improvement through high-quality outpatient management programs. The premise behind these and related interventions is that high-quality outpatient care should reduce unnecessary hospitalizations for high-cost patients.

However, there are few data on the proportion of inpatient hospitalizations among high-cost patients that are potentially preventable. Prior studies have shown that among patients with select chronic diseases, a substantial proportion of ED visits3,4 and hospitalizations may be preventable.1,57 However, the degree to which these findings apply to high-cost patients more generally is unknown. Furthermore, little is known about the supply-side factors (such as the number of primary care physicians in a community) that affect spending on preventable hospitalizations within this population. A recent Congressional Budget Office report suggested that the various Medicare Coordinated Care Demonstration programs focused on high-cost patients have generally failed to save money, at least in part because little is known about the drivers of costs for this group.8

Therefore, in this study, we sought to quantify the preventability of high-cost patients' acute care spending, which is critical to both designing appropriate interventions and predicting their potential clinical and economic benefits. Specifically, the 3 goals of this study were to: (1) determine the proportion of acute care episodes and spending that are attributable to the high-cost patients in the Medicare population; (2) determine the proportion and amount of this spending that is likely preventable using standard criteria; and (3) determine whether specific supply-side variables, including the number of primary care physicians and specialists, are associated with preventable acute care spending.

METHODS
Patients

We used MedPAR as well as standard 5% Medicare outpatient and carrier files from 2009 and 2010. Patients younger than 65 years, those not continuously enrolled during the study period, and those with any Medicare Advantage enrollment were excluded. We excluded patients who died during 2009 or 2010 because assessing preventability of end-of-life costs was beyond the scope of this analysis and might bias us toward overestimating preventable spending. Patient race was categorized in the Medicare data according to self-report. We used the Centers for Medicare & Medicaid Services Hierarchical Condition Categories coding to assign comorbidities to each patient in our database based on their inpatient and outpatient diagnoses. International Classification of Diseases, Ninth Revision (ICD-9) codes that were used to identify each major comorbidity are provided in eTable 1.

To assess the effect of supply-side variables on our outcome of interest, we used data from the Dartmouth Atlas of Health Care9 at the level of the hospital referral region (HRR) to obtain the supply of primary care physicians, specialist physicians, and ED physicians per 100 000 residents and the supply of hospital beds per 1000 residents.

This study was approved by the Harvard School of Public Health Office of Human Research Administration; the requirement of informed consent was waived because of the deidentified nature of the data.

Identifying High-Cost Patients

We created standardized costs for inpatient care using the MedPAR files. We began with the amount paid by Medicare for each hospitalization, subtracted out duplicate costs from the carrier file, and adjusted for Medicare Wage Index, graduate medical education, and disproportionate-share payments. We used published Medicare fee schedules to assign standardized Medicare costs to each outpatient and carrier file service, regardless of the actual amount Medicare paid for each service (see eAppendix for detailed methods on standardized costs). The use of standardized costs allows us to identify patients who use a comparable amount of medical care across areas of the country in which the actual spending may vary significantly. Costs were summed across the year and across settings for each patient in our sample. We defined patients in the top decile of total cost in 2010 as high-cost patients and those in the top decile in both 2009 and 2010 as persistently high-cost patients based on prior work showing only modest persistence of high expenditures in the Medicare population.10,11

Identifying Preventable Emergency Department Visits

To identify preventable ED visits, we used an algorithm created by Billings et al.12 This algorithm, which has been validated4 and used in prior published work,3,13 uses diagnosis codes to separate ED visits into 4 categories: nonemergent; emergent but primary care treatable; emergent, ED care needed, but preventable; and emergent, ED care needed, and not preventable. We defined nonemergent, emergent/primary care treatable, and emergent/ED care needed/preventable or avoidable visits as preventable ED visits. Because Medicare data combine ED costs with inpatient costs if a patient is admitted to the hospital, we limited our sample of independent ED visits to visits not leading to an admission.

Identifying Preventable Hospitalizations

We used the Agency for Healthcare Research and Quality Prevention Quality Indicators software to identify potentially preventable hospitalizations.14 This algorithm defines potentially preventable hospitalizations as those related to conditions, such as heart failure, diabetes, hypertension, and asthma, for which good outpatient care can likely prevent the need for hospitalization, and it has been validated and used in prior work on the Medicare population.1517 The software was altered to include patients admitted from nursing homes, who were excluded from the Prevention Quality Indicators algorithm in the 2009 version. eTable 2 provides a list of the preventable hospitalization diagnoses and their associated ICD-9 codes. For purposes of comparison, we then grouped nonpreventable diagnoses into clinically similar groups (eg, anterior myocardial infarction, subendocardial infarction, and coronary atherosclerosis were all grouped into ischemic heart disease).

Statistical Analysis

We first compared patient characteristics between high-cost and non–high-cost patients. We then compared characteristics for the hospitalizations within each group. We calculated the proportion of total 2010 ED visits and inpatient hospitalizations that were for high-cost patients vs non–high-cost patients, as well as the associated costs. Next, we calculated the proportion of ED visit costs and the proportion of short-stay acute care hospital costs that were potentially preventable within each group, using the algorithms described above. For the high-cost cohort, we created a histogram at the patient level of the proportion of acute care costs that were potentially preventable. We repeated these analyses for the persistently high-cost cohort.

We also classified each high-cost individual in our database into 1 of the 306 HRRs in the country and calculated the average per capita preventable acute care costs for high-cost patients for each HRR. We calculated summary statistics for these costs at the HRR level. We then created HRR-level linear regression models in which the average per capita preventable costs, calculated at the HRR level, were our outcome of interest, and supply-side variables, including primary care physician, specialist physician, and emergency department physician supply and hospital bed supply, were our primary predictors. We included HRR-level age, sex, race, and comorbidity burden as covariates. A 2-tailed P <.05 was considered statistically significant.

All analyses were performed using SAS software, version 9.3 (SAS Institute Inc).

RESULTS
Patient Characteristics

There were 1 114 469 patients in our 5% Medicare sample, of which 113 341 constituted the high-cost cohort. High-cost patients were older (median age, 78 vs 77 years), more often male (44.5% vs 41.7%), and more often black (8.5% vs 7.1%) than non–high-cost patients, and were more often Medicaid eligible. High-cost patients had a higher burden of comorbid illnesses, including heart failure, diabetes, and cancer, as well as higher rates of mental illness and substance abuse (Table 1). Hospitalizations for high-cost and non–high-cost patients were distributed similarly across hospital types, though a higher proportion of hospitalizations for high-cost patients were in hospitals in the South (38% vs 45%), hospitals located in urban areas (85% vs 78%), and major teaching hospitals (21% vs 13%) and safety-net hospitals (23% vs 17%). Twenty-six percent of hospitalizations for high-cost patients and 30% for non–high-cost patients originated in the ED.

Proportion of Acute Care Services and Costs From the High-Cost Group

The high-cost patient cohort, which included 10% of the patients in our sample, was responsible for 30.8% (95% CI, 30.7%-31.0%) of ED visits not resulting in an admission and 32.9% (95% CI, 32.9%-32.9%) of ED costs. High-cost patients accounted for 56.7% (95% CI, 56.5%-57.0%) of admissions and 79.0% (95% CI, 79.0%-79.0%) of inpatient costs. In total, 73.0% (95% CI, 73.0%-73.0%) of acute care spending in 2010 was attributable to the 10% of patients in our high-cost group.

Preventable Emergency Department Visits

Within the high-cost cohort, 42.6% of ED visits were deemed to be preventable (95% CI, 42.4%-42.9%) according to our algorithm. These visits were associated with 41.0% of the ED costs within this group (95% CI, 40.9%-41.0%, Table 2). Patterns were similar for the non–high-cost cohort, with 44.2% of visits (95% CI, 44.0%-44.4%) and 42.6% of costs (95% CI, 42.6%-42.6%) deemed preventable.

Preventable Hospitalizations

The most common reasons for preventable hospitalization in high-cost patients were congestive heart failure, bacterial pneumonia, and chronic obstructive pulmonary disease (Table 3), and the most common reasons for nonpreventable hospitalization were orthopedic conditions, ischemic heart disease, and cancer and chemotherapy (eTable 4); many of the diagnoses that were associated with the highest overall costs in this cohort were considered nonpreventable (Figure).

Within the high-cost group, 15.8% (95% CI, 15.8%-15.8%) of the admissions were attributable to preventable causes; 9.6% (95% CI, 9.6%-9.6%) of hospital costs were attributable to preventable hospitalization, while the remaining 90.4% (95% CI, 90.4%-90.4%) were attributable to other causes of hospitalization (Table 3). Within the non–high-cost group, though overall spending was significantly lower, a higher proportion of inpatient costs were potentially preventable (16.8%; 95% CI, 16.8%-16.8%).

Combining the ED and inpatient settings, 10.0% (95% CI, 10.0%-10.0%) of high-cost patients' costs were considered potentially preventable and 19.1% (95% CI, 19.1%-19.1%) of non–high-cost patients' costs were considered potentially preventable. Only 10% of the high-cost cohort had acute care costs that were considered preventable (eFigure 1).

Persistently High-Cost Cohort

The 31 263 patients in the persistently high-cost cohort were older, more often black, and more often Medicaid eligible and had a higher burden of medical comorbidities than patients who were not persistently high-cost (eTable 5). In this group, 44.7% (95% CI, 44.2%-45.1%) of ED visits were deemed preventable according to our algorithm. These visits were associated with 43.3% (95% CI, 43.3%-43.3%) of the ED costs within this group (eTable 6). Within the persistently high-cost group, 20.3% (95% CI, 20.1%-20.5%) of the admissions and 13.5% (95% CI, 13.5%-13.5%) of the costs were attributable to preventable hospitalization, while the remaining 86.5% (95% CI, 86.5%-86.5%) of the costs were attributable to other causes of hospitalization (Table 4).

Regional Variability in Per Capita Preventable Costs

We found significant variability in preventable acute care costs for high-cost patients across HRRs in 2010, ranging from $681 to $5217 per capita across the total population of each HRR (eFigure 2). After adjusting for age, sex, race, and comorbidities of the patient population in each HRR, our supply-side variables were positively associated with these costs: HRRs with the lowest primary care physician supply had $1954 (95% CI, $1837-$2071) in preventable costs per capita vs $2186 (95% CI, $2065-$2307; P = .009) for HRRs with the highest primary care physician supply (Table 5). Higher specialist physician supply was also associated with higher per capita preventable costs, as was hospital bed supply.

DISCUSSION

We found that more than 70% of the roughly $91.7 billion in acute care costs in the Medicare population in 2010 were for the 10% of patients that comprise the high-cost cohort. Approximately 10% of these costs were for what were deemed potentially preventable causes as calculated using standard algorithms; the percentage was slightly higher for the persistently high-cost cohort. Hospital referral regions with a higher primary care or specialist physician supply had higher annual preventable costs per capita.

The biggest drivers of inpatient spending for high-cost patients were catastrophic events such as sepsis, stroke, and myocardial infarction, as well as cancer and expensive orthopedic procedures such as spine surgery and hip replacement. These findings suggest that strategies focused on enhanced outpatient management of chronic disease, while critically important, may not be focused on the biggest and most expensive problems plaguing Medicare's high-cost patients. Indeed, while a proportion of these very expensive inpatient episodes may be potentially preventable (such as acute myocardial infarction or degenerative joint disease leading to orthopedic procedures), their prevention would likely require a long time horizon and substantial investments in population wellness. Such investments are critically important for ensuring the health of the population, but the time frame needed to see cost savings is likely years, not weeks or months.

These findings may shed light on why many recent efforts to control costs for these very medically complex, high-utilizing patients, including the Medicare Coordinated Care Demonstration programs, have failed to do so,8 even in cases in which there was a small decrease in hospital admissions.18 The majority of these programs have focused on providing enhanced outpatient services, such as frequent telephone and in-person contact, patient education, enhanced medication management services, and assistance with transitional care following a hospitalization.8,18,19 These types of services are targeted toward reducing ambulatory care–sensitive hospitalizations, and investing further in disease management programs may lead to reductions in avoidable ED visits and hospitalizations. Although these visits are still very expensive in aggregate, our findings suggest that they make up a small proportion of the total acute care spending among the costliest of patients. As a result, while disease management may yield cost savings, even a substantial reduction in these preventable hospitalizations is unlikely to have a large effect on overall spending levels within this cohort.

Our findings suggest that a complementary approach to saving money on acute care services for high-cost patients may be to additionally focus on reducing per-episode costs for high-cost disease entities through clinical innovation and care delivery redesign. It is feasible that current policy interventions, such as bundled payments,20 might spur such innovation in inpatient cost control. Furthermore, as shared-savings programs such as accountable care organizations become more common, clinical leaders may need to prioritize both reducing preventable admissions and reducing the cost of hospitalizations for catastrophic or acute disease to reap meaningful savings.

We found that HRRs with high primary care physician supply or high specialist physician supply had higher preventable spending for their high-cost patients; it is unclear whether this is due to supply-induced demand for ED visits and hospitalizations or simply reflective of a more complex underlying patient population. However, it suggests that simply increasing access to primary care services may not in itself lead to lower health care spending for high-cost patients.

Our study adds to prior literature examining acute care spending for high-cost patients. For example, the Congressional Budget Office examined Medicare data from 2001 and showed that high-cost patients were more likely than non–high-cost patients to use short-term hospital care as well as ED care and that per-episode costs in these categories were higher in the high-cost group as well.21 Riley2 studied spending trends over time on high-cost Medicare patients and found an increasing proportion of their costs coming from services such as skilled nursing and home health care between 1975 and 2004, though inpatient spending remained the dominant source of costs throughout. However, to our knowledge, this is the first study to examine preventability, especially of acute care services, within this cohort. Our findings regarding the effect of primary care and specialty care on per-capita preventable costs are somewhat in contrast to prior work by Baicker and Chandra22 demonstrating that states with high primary care supply had lower costs and higher quality for their Medicare beneficiaries, although these authors did not focus on high-cost patients, and the relationship in this group, who may have a higher need for services, may be different.

There are limitations to this study. Although we used well-established algorithms to define preventable acute care episodes, it is likely that these represent a spectrum of preventability and that admissions for perforated appendicitis or lower extremity amputation, for example, may actually have not been preventable through better outpatient care. However, this would have biased us toward finding a greater proportion of spending due to preventable causes, and thus our data are likely conservative. Our findings that regions with more primary care and specialist physicians have higher per-capita preventable acute care spending may be due to factors such as a generally sicker or more complex patient population that uses more health care services and, thus, may be demand driven rather than supply driven.

CONCLUSION

Among a sample of patients in the top decile of Medicare spending in 2009, only a small percentage of costs appeared to be related to preventable ED visits and hospitalizations. The ability to lower costs for these patients through better outpatient care may be limited. Also needed may be strategies, perhaps focused on high-cost disease entities, that make hospital care more efficient so that each episode of inpatient care is less expensive regardless of its cause.

Back to top
Article Information

Corresponding Author: Karen E. Joynt, MD, MPH, Brigham and Women's Hospital, 75 Francis St, Boston, MA 02115 (kjoynt@partners.org).

Published Online: June 24, 2013. doi:10.1001/jama.2013.7103

Author Contributions: Dr Joynt had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Joynt, Jha.

Acquisition of data: Jha.

Analysis and interpretation of data: All authors.

Drafting of the manuscript: Joynt.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Orav.

Obtained funding: Joynt, Gawande, Jha.

Administrative, technical, or material support: Jha.

Study supervision: Jha.

Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Gawande reported receiving honoraria for lectures and teaching about improvement of quality and safety in health care from clinical organizations and associations that are financially affected by the design of the health system and receiving royalties from multiple publishers for books, writing, and a documentary on health care systems and performance. No other disclosures were reported.

Funding/Support:This study was funded by the Rx Foundation and the West Wireless Foundation.

Role of the Sponsors: The funders of this study had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, and approval of the manuscript; or decision to submit the manuscript for publication.

REFERENCES
1.
Niefeld MR, Braunstein JB, Wu AW, Saudek CD, Weller WE, Anderson GF. Preventable hospitalization among elderly Medicare beneficiaries with type 2 diabetes.  Diabetes Care. 2003;26(5):1344-1349PubMedArticle
2.
Riley GF. Long-term trends in the concentration of Medicare spending.  Health Aff (Millwood). 2007;26(3):808-816PubMedArticle
3.
Smulowitz PB, Lipton R, Wharam JF,  et al.  Emergency department utilization after the implementation of Massachusetts health reform.  Ann Emerg Med. 2011;58(3):225-234PubMedArticle
4.
Ballard DW, Price M, Fung V,  et al.  Validation of an algorithm for categorizing the severity of hospital emergency department visits.  Med Care. 2010;48(1):58-63PubMedArticle
5.
Billings J, Anderson GM, Newman LS. Recent findings on preventable hospitalizations.  Health Aff (Millwood). 1996;15(3):239-249PubMedArticle
6.
Oster A, Bindman AB. Emergency department visits for ambulatory care sensitive conditions: insights into preventable hospitalizations.  Med Care. 2003;41(2):198-207PubMed
7.
Braunstein JB, Anderson GF, Gerstenblith G,  et al.  Noncardiac comorbidity increases preventable hospitalizations and mortality among Medicare beneficiaries with chronic heart failure.  J Am Coll Cardiol. 2003;42(7):1226-1233PubMedArticle
8.
Congressional Budget Office.  Lessons from Medicare's Demonstration Projects on Disease Management, Care Coordination, and Value-Based Payment. Washington, DC: Congressional Budget Office; 2012
9.
Dartmouth Institute for Health Policy and Clinical Practice.  The Dartmouth Atlas of Health Care. 2012. http://www.dartmouthatlas.org/. Accessed July 22, 2012
10.
Garber AM, MaCurdy TE, McClellan MB. Persistence of Medicare expenditures among elderly beneficiaries. In: Garber AM, ed. Frontiers in Health Policy Research. Vol 1. Cambridge: National Bureau of Economic Research, Massachusetts Institute of Technology; 1998:153-180
11.
Bhattacharya J, Garber AM, MaCurdy TE. The narrowing dispersion of Medicare expenditures 1997 to 2005. In: Wise DA, ed. Research Findings in the Economics of Aging. Chicago, IL: National Bureau of Economic Research, University of Chicago Press; 2010:387-407
12.
NYU Wagner.  NYU ED algorithm: background.  2000. http://wagner.nyu.edu/faculty/billings/nyued-background.php. Accessed March 13, 2013
13.
Wharam JF, Landon BE, Galbraith AA, Kleinman KP, Soumerai SB, Ross-Degnan D. Emergency department use and subsequent hospitalizations among members of a high-deductible health plan.  JAMA. 2007;297(10):1093-1102PubMedArticle
14.
Agency for Healthcare Research and Quality.  Prevention Quality Indicators overview. 2011. http://www.qualityindicators.ahrq.gov/modules/pqi_overview.aspx. Accessed January 15, 2013
15.
Basu J, Friedman B, Burstin H. Primary care, HMO enrollment, and hospitalization for ambulatory care sensitive conditions: a new approach.  Med Care. 2002;40(12):1260-1269PubMedArticle
16.
Bindman AB, Grumbach K, Osmond D,  et al.  Preventable hospitalizations and access to health care.  JAMA. 1995;274(4):305-311PubMedArticle
17.
Jiang HJ, Russo CA, Barrett ML. Nationwide Frequency and Costs of Potentially Preventable Hospitalizations, 2006. Rockville, MD: Agency for Healthcare Research & Quality; April 2009. HCUP Statistical Brief 72
18.
Brown RS, Peikes D, Peterson G, Schore J, Razafindrakoto CM. Six features of Medicare coordinated care demonstration programs that cut hospital admissions of high-risk patients.  Health Aff (Millwood). 2012;31(6):1156-1166PubMedArticle
19.
Cebul RD, Rebitzer JB, Taylor LJ, Votruba ME. Organizational fragmentation and care quality in the US healthcare system.  J Econ Perspect. 2008;22(4):93-113PubMedArticle
20.
Miller DC, Gust C, Dimick JB, Birkmeyer N, Skinner J, Birkmeyer JD. Large variations in Medicare payments for surgery highlight savings potential from bundled payment programs.  Health Aff (Millwood). 2011;30(11):2107-2115PubMedArticle
21.
Congressional Budget Office.  High-Cost Medicare Beneficiaries. Washington, DC: Congressional Budget Office; 2005
22.
Baicker K, Chandra A. Medicare spending, the physician workforce, and beneficiaries' quality of care.  Health Aff (Millwood). 2004;(suppl web exclusives)  W4-184-W4-197PubMed
×