Letters
3. Rural Assistance Center. Funding by type: loan repayment programs. http: //www.raconline.org/funding/type_details.php?type=Loan Repayment Programs. Accessed August 19, 2013. 4. Pathman DE, Konrad TR. Growth and changes in the National Health Service Corps (NHSC) workforce with the American Recovery and Reinvestment Act. J Am Board Fam Med. 2012;25(5):723-733. 5. US Department of Health and Human Services; Health Resources and Services Administration. FY 2012 online performance appendix. http://www.hrsa.gov/about/budget/performanceappendix2012.pdf. Accessed October 11, 2013.
Conflict of Interest Disclosures: The author has completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and reported being a Centers for Medicare & Medicaid Services innovation advisor, for which his department received funds to support his travel and participation in the program. 1. Joynt KE, Gawande AA, Orav EJ, Jha AK. Contribution of preventable acute care spending to total spending for high-cost Medicare patients. JAMA. 2013;309(24):2572-2578.
6. US Department of Health and Human Services Data Warehouse. Designated health professional shortage areas statistics: third quarter of fiscal year 2013 designated HPSA quarterly summary. http://datawarehouse.hrsa.gov /quickaccessreports.aspx. Accessed August 26, 2013.
2. Winter WG. Nonoperative treatment of proximal femoral fractures in the demented, nonambulatory patient. Clin Orthop Relat Res. 1987;(218):97-103.
COMMENT & RESPONSE
4. Baicker K, Chandra A. Medicare spending, the physician workforce, and beneficiaries’ quality of care. Health Aff (Millwood). 2004;(suppl web exclusives):W4-184-97.
Preventable Acute Care Spending for Medicare Patients To the Editor The study on preventable acute care spending by Dr Joynt and colleagues1 had a number of limitations. The authors did not look specifically at end-of-life costs and the savings generated by a physician who keeps a patient out of the hospital through the appropriate provision of palliative or hospice services would not be counted. For example, if a nonambulatory patient with dementia and living in a nursing home rolled out of bed and broke her hip but did not have surgery, this event would not count as a preventable ED admission according to the methods used in the study, even though nonoperative management may be the best approach.2 The authors concluded that 41.0% of ED costs for high-cost patients were potentially preventable. A recent study estimated the downstream costs of ED care may be as high as 10% of the total US health care costs,3 which implies there are more total cost savings than they calculated. In addition, in the Discussion section, the authors wrote: “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 Chandra demonstrating that states with high primary care supply had lower costs and higher quality for their Medicare beneficiaries … .” The study by Baicker and Chandra4 found a correlation between higher quality and lower costs and family physician supply, not all primary care physician supply. There is no evidence that internist or pediatrician supply is associated with higher quality and lower costs. In fact, higher internist supply is associated higher ED use, more hospital admissions, and higher costs; the opposite is true for family physicians.5 There are differences among primary care physicians that result in system effects that the authors did not capture or analyze, which is an important limitation of their study. Richard Young, MD Author Affiliation: Department of Family Medicine, JPS Hospital, Ft Worth, Texas. 1984
Corresponding Author: Richard Young, MD, JPS Hospital, 1500 S Main, Ft Worth, TX 76104 ([email protected]).
3. Lee MH, Schuur JD, Zink BJ. Owning the cost of emergency medicine: beyond 2% [published online April 24, 2013]. Ann Emerg Med. doi:10.1016/j.annemergmed.2013.03.029.
5. Petterson S, Xierali I, Phillips R. Impact of primary care provider supply in urban and rural areas in the United States. http://www.stfm.org/fmsup/napcrg /fmconferencesupplement.cfm?confid=137. Accessibility verified October 8, 2013.
To the Editor Dr Joynt and colleagues1 demonstrated limitations in cost savings achievable by reducing emergency department (ED) visits and hospitalizations for some of the most common preventable diseases faced by Medicare patients in a high-cost group. However, the study used an algorithm created by Billings et al2 to define preventable visits as those ED diagnosis codes that were determined to be nonemergent; emergent but primary care treatable; and emergent, ED care needed, but preventable. Even though the algorithm by Billings et al2 remains a convenient way for researchers to measure preventable ED visits, its convenience is outweighed by its fallibility. Determining preventability using a retrospective application of ED visit diagnostic codes can result in bias. Joynt and colleagues1 reported that 41.0% of ED visits in high-cost patients were potentially preventable, which is a strikingly high proportion. For example, a middle-aged man who comes to the ED with chest pain and has a normal electrocardiogram and a negative troponin level may be discharged with a preventable diagnosis of esophageal reflux. Yet if the testing had indicated that he had acute coronary syndrome, the determination of preventability would have been different. Recent work by Raven et al3 highlighted the major flaws of this algorithm and the limitations of using ED diagnostic codes to determine the urgency or preventability of a visit. We support abandoning use of the algorithm by Billings et al,2 except perhaps for its original intent of tracking the status of the health care safety net. Despite these limitations, the conclusions of the study by Joynt et al1 may underemphasize the absolute amount of cost savings available by reducing hospitalizations that originate in the ED. The total amount of savings attributable to reducing 10% of admissions from the ED may be 2.5% of federal health expenditures, 4 accounting for $7 billion annually, which is no small sum with respect to bending the cost curve.
JAMA November 13, 2013 Volume 310, Number 18
Copyright 2013 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ by a Ndsu Library Periodicals User on 05/19/2015
jama.com
Letters
Scott G. Weiner, MD, MPH Peter B. Smulowitz, MD, MPH Author Affiliations: Department of Emergency Medicine, Tufts Medical Center, Boston, Massachusetts (Weiner); Department of Emergency Medicine, Beth Israel Deaconess Medical Center, Boston, Massachusetts (Smulowitz). Corresponding Author: Scott G. Weiner, MD, MPH, Tufts Medical Center, 800 Washington St, PO Box 311, Boston, MA 02111 ([email protected]). Conflict of Interest Disclosures: The authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported. 1. Joynt KE, Gawande AA, Orav EJ, Jha AK. Contribution of preventable acute care spending to total spending for high-cost Medicare patients. JAMA. 2013;309(24):2572-2578. 2. Billings JC, Parikh N, Mijanovich T. Emergency Room Use: The New York Story. New York, NY: Commonwealth Fund; 2000. 3. Raven MC, Lowe RA, Maselli J, Hsia RY. Comparison of presenting complaint vs discharge diagnosis for identifying “nonemergency” emergency department visits. JAMA. 2013;309(11):1145-1153. 4. Smulowitz PB, Honigman L, Landon BE. A novel approach to identifying targets for cost reduction in the emergency department. Ann Emerg Med. 2013;61(3):293-300.
In Reply Dr Young points out that we did not examine endof-life care and suggests that we may have underestimated preventable spending. We agree that high-quality hospice or palliative care may reduce admissions at the end of life, including for conditions that are considered nonpreventable. The frequency with which these costly inpatient services, such as hip replacement or major cardiac surgery, are occurring in people with advanced terminal illness is unknown. On the other hand, patients at the end of life may be sufficiently sick that their ED admission for typically preventable conditions, such as urinary tract infections, may actually be more difficult to avoid. Because of the difficulty in assessing these complex issues in administrative data, we chose instead to exclude these patients from this study but agree with Young that better understanding of the preventability of spending at the end of life is important. Drs Weiner and Smulowitz express concerns that the algorithm created by Billings et al may overstate the preventability of ED visits. The difficulty with this algorithm is primarily that visits deemed preventable based on the final diagnosis (eg, costochondritis) might not be preventable based on the presenting complaint (eg, chest pain).1 We attempted to account for this by conducting sensitivity analyses in which preventability was assessed based on the diagnosis code for each service (eg, if a stress test was ordered with chest pain as a diagnosis), but agree that this is likely an incomplete fix. On the other hand, both Weiner and Smulowitz and Young suggest that the potential savings from preventable ED visits may have been larger if we had included the downstream effect on hospitalizations. The costs of any ED visit leading to a hospitalization are rolled into the inpatient costs for that hospitalization with Medicare; thus, these
downstream costs were largely captured by our analysis of preventable inpatient care. Furthermore, as Smulowitz et al 2 pointed out, even if we prevent 10% to 25% of hospitalizations that come from “intermediate/complex” ED visits, those savings will account for a small proportion of total health care spending. Young also points out that there are differences among primary care physicians in terms of costs and quality. Although a more detailed breakdown of costs for patients of family practitioners vs internists was not available for this analysis, we concur that this may be an important area of study in the future. In addition, we agree that, even though we found that only 10% of spending was preventable using existing algorithms, investing in outpatient care is still an extremely important strategy for reducing health care costs, especially over the longer run. Furthermore, our findings suggest that for the other 90% of spending, achieving cost savings safely would likely require new strategies, including care redesign. Improving care for high-cost patients represents an important opportunity to reduce waste in the US health care system. Given the complexity of the task, our data suggest that we need a multifaceted approach. Karen E. Joynt, MD, MPH Atul A. Gawande, MD, MPH Ashish K. Jha, MD, MPH Author Affiliations: Department of Health Policy and Management, Harvard School of Public Health, Boston, Massachusetts. Corresponding Author: Karen E. Joynt, MD, MPH, Brigham and Women’s Hospital, 75 Francis St, Boston, MA 02115 ([email protected]). Conflict of Interest Disclosures: The 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. 1. Raven MC, Lowe RA, Maselli J, Hsia RY. Comparison of presenting complaint vs discharge diagnosis for identifying “nonemergency” emergency department visits. JAMA. 2013;309(11):1145-1153. 2. Smulowitz PB, Honigman L, Landon BE. A novel approach to identifying targets for cost reduction in the emergency department. Ann Emerg Med. 2013;61(3):293-300.
New Definition of Term Pregnancy To the Editor The Viewpoint on defining term pregnancy1 presented an oversimplified construct for stratifying term pregnancy. Dr Spong1 framed her discussion around identification of “… the optimal timing of delivery for a healthy pregnancy.” However, a woman with a healthy pregnancy may have common risk factors that may lower her optimal time to deliver.2,3 Spong1 proposed “39 weeks 0 days through 40 weeks 6 days’ gestation” as the optimal time for delivery of healthy pregnancies, but in many studies the interval associated with
jama.com
JAMA November 13, 2013 Volume 310, Number 18
Copyright 2013 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ by a Ndsu Library Periodicals User on 05/19/2015
1985
3. Rural Assistance Center. Funding by type: loan repayment programs. http: //www.raconline.org/funding/type_details.php?type=Loan Repayment Programs. Accessed August 19, 2013. 4. Pathman DE, Konrad TR. Growth and changes in the National Health Service Corps (NHSC) workforce with the American Recovery and Reinvestment Act. J Am Board Fam Med. 2012;25(5):723-733. 5. US Department of Health and Human Services; Health Resources and Services Administration. FY 2012 online performance appendix. http://www.hrsa.gov/about/budget/performanceappendix2012.pdf. Accessed October 11, 2013.
Conflict of Interest Disclosures: The author has completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and reported being a Centers for Medicare & Medicaid Services innovation advisor, for which his department received funds to support his travel and participation in the program. 1. Joynt KE, Gawande AA, Orav EJ, Jha AK. Contribution of preventable acute care spending to total spending for high-cost Medicare patients. JAMA. 2013;309(24):2572-2578.
6. US Department of Health and Human Services Data Warehouse. Designated health professional shortage areas statistics: third quarter of fiscal year 2013 designated HPSA quarterly summary. http://datawarehouse.hrsa.gov /quickaccessreports.aspx. Accessed August 26, 2013.
2. Winter WG. Nonoperative treatment of proximal femoral fractures in the demented, nonambulatory patient. Clin Orthop Relat Res. 1987;(218):97-103.
COMMENT & RESPONSE
4. Baicker K, Chandra A. Medicare spending, the physician workforce, and beneficiaries’ quality of care. Health Aff (Millwood). 2004;(suppl web exclusives):W4-184-97.
Preventable Acute Care Spending for Medicare Patients To the Editor The study on preventable acute care spending by Dr Joynt and colleagues1 had a number of limitations. The authors did not look specifically at end-of-life costs and the savings generated by a physician who keeps a patient out of the hospital through the appropriate provision of palliative or hospice services would not be counted. For example, if a nonambulatory patient with dementia and living in a nursing home rolled out of bed and broke her hip but did not have surgery, this event would not count as a preventable ED admission according to the methods used in the study, even though nonoperative management may be the best approach.2 The authors concluded that 41.0% of ED costs for high-cost patients were potentially preventable. A recent study estimated the downstream costs of ED care may be as high as 10% of the total US health care costs,3 which implies there are more total cost savings than they calculated. In addition, in the Discussion section, the authors wrote: “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 Chandra demonstrating that states with high primary care supply had lower costs and higher quality for their Medicare beneficiaries … .” The study by Baicker and Chandra4 found a correlation between higher quality and lower costs and family physician supply, not all primary care physician supply. There is no evidence that internist or pediatrician supply is associated with higher quality and lower costs. In fact, higher internist supply is associated higher ED use, more hospital admissions, and higher costs; the opposite is true for family physicians.5 There are differences among primary care physicians that result in system effects that the authors did not capture or analyze, which is an important limitation of their study. Richard Young, MD Author Affiliation: Department of Family Medicine, JPS Hospital, Ft Worth, Texas. 1984
Corresponding Author: Richard Young, MD, JPS Hospital, 1500 S Main, Ft Worth, TX 76104 ([email protected]).
3. Lee MH, Schuur JD, Zink BJ. Owning the cost of emergency medicine: beyond 2% [published online April 24, 2013]. Ann Emerg Med. doi:10.1016/j.annemergmed.2013.03.029.
5. Petterson S, Xierali I, Phillips R. Impact of primary care provider supply in urban and rural areas in the United States. http://www.stfm.org/fmsup/napcrg /fmconferencesupplement.cfm?confid=137. Accessibility verified October 8, 2013.
To the Editor Dr Joynt and colleagues1 demonstrated limitations in cost savings achievable by reducing emergency department (ED) visits and hospitalizations for some of the most common preventable diseases faced by Medicare patients in a high-cost group. However, the study used an algorithm created by Billings et al2 to define preventable visits as those ED diagnosis codes that were determined to be nonemergent; emergent but primary care treatable; and emergent, ED care needed, but preventable. Even though the algorithm by Billings et al2 remains a convenient way for researchers to measure preventable ED visits, its convenience is outweighed by its fallibility. Determining preventability using a retrospective application of ED visit diagnostic codes can result in bias. Joynt and colleagues1 reported that 41.0% of ED visits in high-cost patients were potentially preventable, which is a strikingly high proportion. For example, a middle-aged man who comes to the ED with chest pain and has a normal electrocardiogram and a negative troponin level may be discharged with a preventable diagnosis of esophageal reflux. Yet if the testing had indicated that he had acute coronary syndrome, the determination of preventability would have been different. Recent work by Raven et al3 highlighted the major flaws of this algorithm and the limitations of using ED diagnostic codes to determine the urgency or preventability of a visit. We support abandoning use of the algorithm by Billings et al,2 except perhaps for its original intent of tracking the status of the health care safety net. Despite these limitations, the conclusions of the study by Joynt et al1 may underemphasize the absolute amount of cost savings available by reducing hospitalizations that originate in the ED. The total amount of savings attributable to reducing 10% of admissions from the ED may be 2.5% of federal health expenditures, 4 accounting for $7 billion annually, which is no small sum with respect to bending the cost curve.
JAMA November 13, 2013 Volume 310, Number 18
Copyright 2013 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ by a Ndsu Library Periodicals User on 05/19/2015
jama.com
Letters
Scott G. Weiner, MD, MPH Peter B. Smulowitz, MD, MPH Author Affiliations: Department of Emergency Medicine, Tufts Medical Center, Boston, Massachusetts (Weiner); Department of Emergency Medicine, Beth Israel Deaconess Medical Center, Boston, Massachusetts (Smulowitz). Corresponding Author: Scott G. Weiner, MD, MPH, Tufts Medical Center, 800 Washington St, PO Box 311, Boston, MA 02111 ([email protected]). Conflict of Interest Disclosures: The authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported. 1. Joynt KE, Gawande AA, Orav EJ, Jha AK. Contribution of preventable acute care spending to total spending for high-cost Medicare patients. JAMA. 2013;309(24):2572-2578. 2. Billings JC, Parikh N, Mijanovich T. Emergency Room Use: The New York Story. New York, NY: Commonwealth Fund; 2000. 3. Raven MC, Lowe RA, Maselli J, Hsia RY. Comparison of presenting complaint vs discharge diagnosis for identifying “nonemergency” emergency department visits. JAMA. 2013;309(11):1145-1153. 4. Smulowitz PB, Honigman L, Landon BE. A novel approach to identifying targets for cost reduction in the emergency department. Ann Emerg Med. 2013;61(3):293-300.
In Reply Dr Young points out that we did not examine endof-life care and suggests that we may have underestimated preventable spending. We agree that high-quality hospice or palliative care may reduce admissions at the end of life, including for conditions that are considered nonpreventable. The frequency with which these costly inpatient services, such as hip replacement or major cardiac surgery, are occurring in people with advanced terminal illness is unknown. On the other hand, patients at the end of life may be sufficiently sick that their ED admission for typically preventable conditions, such as urinary tract infections, may actually be more difficult to avoid. Because of the difficulty in assessing these complex issues in administrative data, we chose instead to exclude these patients from this study but agree with Young that better understanding of the preventability of spending at the end of life is important. Drs Weiner and Smulowitz express concerns that the algorithm created by Billings et al may overstate the preventability of ED visits. The difficulty with this algorithm is primarily that visits deemed preventable based on the final diagnosis (eg, costochondritis) might not be preventable based on the presenting complaint (eg, chest pain).1 We attempted to account for this by conducting sensitivity analyses in which preventability was assessed based on the diagnosis code for each service (eg, if a stress test was ordered with chest pain as a diagnosis), but agree that this is likely an incomplete fix. On the other hand, both Weiner and Smulowitz and Young suggest that the potential savings from preventable ED visits may have been larger if we had included the downstream effect on hospitalizations. The costs of any ED visit leading to a hospitalization are rolled into the inpatient costs for that hospitalization with Medicare; thus, these
downstream costs were largely captured by our analysis of preventable inpatient care. Furthermore, as Smulowitz et al 2 pointed out, even if we prevent 10% to 25% of hospitalizations that come from “intermediate/complex” ED visits, those savings will account for a small proportion of total health care spending. Young also points out that there are differences among primary care physicians in terms of costs and quality. Although a more detailed breakdown of costs for patients of family practitioners vs internists was not available for this analysis, we concur that this may be an important area of study in the future. In addition, we agree that, even though we found that only 10% of spending was preventable using existing algorithms, investing in outpatient care is still an extremely important strategy for reducing health care costs, especially over the longer run. Furthermore, our findings suggest that for the other 90% of spending, achieving cost savings safely would likely require new strategies, including care redesign. Improving care for high-cost patients represents an important opportunity to reduce waste in the US health care system. Given the complexity of the task, our data suggest that we need a multifaceted approach. Karen E. Joynt, MD, MPH Atul A. Gawande, MD, MPH Ashish K. Jha, MD, MPH Author Affiliations: Department of Health Policy and Management, Harvard School of Public Health, Boston, Massachusetts. Corresponding Author: Karen E. Joynt, MD, MPH, Brigham and Women’s Hospital, 75 Francis St, Boston, MA 02115 ([email protected]). Conflict of Interest Disclosures: The 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. 1. Raven MC, Lowe RA, Maselli J, Hsia RY. Comparison of presenting complaint vs discharge diagnosis for identifying “nonemergency” emergency department visits. JAMA. 2013;309(11):1145-1153. 2. Smulowitz PB, Honigman L, Landon BE. A novel approach to identifying targets for cost reduction in the emergency department. Ann Emerg Med. 2013;61(3):293-300.
New Definition of Term Pregnancy To the Editor The Viewpoint on defining term pregnancy1 presented an oversimplified construct for stratifying term pregnancy. Dr Spong1 framed her discussion around identification of “… the optimal timing of delivery for a healthy pregnancy.” However, a woman with a healthy pregnancy may have common risk factors that may lower her optimal time to deliver.2,3 Spong1 proposed “39 weeks 0 days through 40 weeks 6 days’ gestation” as the optimal time for delivery of healthy pregnancies, but in many studies the interval associated with
jama.com
JAMA November 13, 2013 Volume 310, Number 18
Copyright 2013 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ by a Ndsu Library Periodicals User on 05/19/2015
1985
