Graduate medical education (GME) programs face increasing pressure to train physicians who deliver high-value care, which encompasses delivery of timely and effective care, cost-effectiveness, patient experience, and outcomes.1 Groups such as the Medicare Payment Advisory Commission have recommended increasing GME training in value, as well as reallocating funds to incentivize programs to train residents who deliver high-value care.2,3 Differences in how health systems function and available resources may affect trainees’ exposure to a high-value care culture and environment.
Culture is defined as a system of shared assumptions, values, beliefs, and norms existing within an environment.4 A health system’s culture is a powerful force that shapes health care provider practice patterns and may correlate with institutional value outcomes.4–7 Measuring culture by residents’ perceptions can help GME programs identify areas where local culture may be improved.7 However, it is likely that culture varies across different types of health systems with different patient populations and resources.
Therefore, we aimed to describe and assess factors associated with high-value care culture among a diverse group of GME training programs (safety net based, community based, and university based). We assumed that residents are shaped by and contribute to the learning environments within their institutions, each of which varies in value-based outcomes. Therefore, we assessed whether high-value care culture reported by internal medicine (IM) residents correlated with institutional value-based care delivery as measured by the Centers for Medicare and Medicaid Services (CMS) Value-Based Purchasing (VBP) program. The VBP program is an initiative that rewards roughly 3,000 acute care hospitals with incentive payments for the quality of care they provide to Medicare beneficiaries.1
Residents at 12 IM GME programs across California completed a cross-sectional online survey between January and June 2016 assessing their perceptions of high-value care culture in their training programs. We asked first- through third-year residents to respond to questions based on their experiences in a specific preidentified center within their respective institutions where they practiced for the greatest number of weeks. We approached all 750 residents across the 12 sites to complete the survey. Potential study participants were contacted by e-mail invitation to participate in the study and had the option to participate in a drawing to receive an iPad if randomly selected. E-mail reminders were sent to nonresponders, and time was provided in the workday for survey completion. We excluded residents who were of preliminary or transitional status during the study period due to their temporary participation and residents who had practiced for less than six months within their residency programs.
We measured high-value care culture by the High-Value Care Culture Survey (HVCCS), which includes four subdomains: leadership and health system messaging, data access and transparency, comfort with cost conversations, and blame-free environment.7,8 The 24-item survey uses a Likert agreement scale (where 0 = strongly disagree, 25 = disagree, 50 = neutral, 75 = agree, and 100 = strongly agree) where a lower score indicates poorer high-value care culture defined by the specific survey item. Survey items are averaged to obtain subdomain scores, and the four subdomains are averaged to obtain the overall HVCCS score. In the HVCCS, value is defined as the quality of care provided to patients in relation to the costs required to deliver that care, and high-value care is care that is designed to maximize quality while minimizing costs. Quality is defined as the degree to which health services increase the likelihood of desired health outcomes that are safe, effective, patient centered, timely, and equitable and are consistent with current professional knowledge. Cost is defined as the negative financial, physical, and emotional effects to patients and the health system.
The HVCCS was developed using a national modified Delphi process, and internal structure validity evidence was collected from two academic institutions.7 Each of the survey’s four domains had Cronbach alphas in excess of 0.7, and, in confirmatory factor analysis, this four-factor model fit the data well (Bentler–Bonett normed fit index 0.98, root mean squared residual 0.06). Higher HVCCS scores equate to having a more robust high-value culture. We linked institutional mean resident HVCCS scores to hospital-level data from the CMS Impact, VBP, and Hospital Compare files from fiscal year 2015.
Medical center type and institutional value-based outcomes
We chose 12 medical centers (see Supplemental Digital Appendix 1, available at http://links.lww.com/ACADMED/A635): 4 medical centers with IM GME training programs from each of 3 types of medical centers (safety net, community, university) in California. We categorized responses to medical centers based on the greatest average number of weeks that residents practiced within specific medical centers. For example, if IM residents spend more than 35% of their time at a university-based medical center, then their results were classified as representing a university-based medical center. Safety-net medical centers were differentiated from community-based medical centers by being in the top quartile (greater than 0.5) of their disproportionate share index (DSH), which measures Medicaid patient load.9,10 Safety-net medical centers are traditionally defined by their mission to serve the nearly 1 in 10 low-income, medically, and socially vulnerable patients in the United States who reside in underserved areas.9–12 We then shared medical center categorization with local study leads from all sites for face validity, and no further adjustments were made. Each of the three types of medical centers represented institutions with low, middle, and high VBP performance (split in tertiles) as reported by the CMS VBP program in fiscal year 2015. We selected at least one program from each tertile.
The VBP program data include chart-abstracted process-of-care measures and assess hospitals’ compliance with evidence-based guidelines, with 97% of all hospitals in the United States reporting their performance, including all of our study sites.13 CMS calculates a VBP total performance score as a composite of four domains: clinical processes of care (20% of total performance score), patient satisfaction (30%), patient outcomes including mortality and complications (30%), and cost defined by Medicare payment per beneficiary (20%).14 Established quality measures are based on data reported by participating hospitals during a performance period (2011–2014 claims), some of which are risk adjusted. The Medicare Spending per Beneficiary measure assesses Medicare Part A and Part B payments provided during episodes of care. The other three components of the VBP are measured at the patient level and rolled into the hospital-level VBP total performance score.
We determined frequencies, means, range, and standard deviations (SDs) of resident demographics and training experiences, as well as institutional overall mean high-value care culture and subdomain scores measured by the HVCCS and hospital characteristics. The HVCCS assesses providers’ perceptions of leadership and health system messaging, data transparency and access, comfort with cost conversations, and blame-free environment within their departments or training programs.7 Descriptive statistics were stratified by type of medical center (safety net, community, university). We assessed the relationship between hospital type and individually reported high-value care culture scores using bivariate and multilevel regression. We adjusted for physician characteristics including age, gender, track in training, and number of months training within a safety-net medical center, as well as for medical center size and type. The smaller sample size limited the ability to conduct a cluster analysis. To evaluate our findings further, we conducted multilevel linear regression specifically among the safety-net medical centers. Statistical significance was determined at a P value < .05.
Finally, we estimated Spearman rank correlations of institutional mean HVCCS scores with the CMS institutional VBP scores among university and community sites separately from safety-net sites because of the CMS metric adjustment methodology. The Spearman rank correlation measures the strength and direction of association between two measures compared in rank order. VBP scores are not adjusted for characteristics of medical center patient populations such as socioeconomic status and race.14 Whereas other clinical measures of care intensity have been used as proxies of value-based care,15,16 we used this publicly reported measure of value because it is widely used and affected reimbursements for 80% of hospitals in the CMS VBP program in 2015.15 We obtained institution-level data from the CMS VBP program and Hospital Compare files.
This study was approved by institutional review boards at all 12 sites. We conducted all analyses using Stata statistical software, version 13.0 (StataCorp, College Station, Texas).
Correlates of high-value care culture among GME training programs
Across all sites, 750 residents were e-mailed surveys, and 517 (68.9%) residents completed the survey. Two hundred eight (54.2%) were female. Residents were on average 29 years old, in the categorical track (i.e., not primary care) (439; 85.4%), and had trained 9.4 months (SD 10.1) in a safety-net medical center. After stratifying by the three hospital types, safety-net medical centers had a larger number of residents in the categorical track (P < .001) (Table 1). Participating residents were of similar age and race/ethnicity compared with national resident physician demographics, though they included fewer men and categorical residents.17 There were no differences in HVCCS scores based on resident training level.
Medical center characteristics.
Of the 12 study sites, there were 4 each from safety-net-based, community-based, and university-based sites, and 4 representing each value-based purchasing performance tertile (high, middle, and low). Eleven (91.7%) were located in urban areas with an average DSH index of 0.42 (SD 0.25), case mix index (CMI, which describes the medical complexity of patients) of 1.97 (SD 0.29), and bed size of 435.5 (SD 146.0).18 Compared with other medical centers training IM residents in California, safety-net medical centers had higher DSH indices but no significant difference in CMI or bed size.10,17 Across all sites, the mean HVCCS score was 51.2 (SD 11.8) on a 0–100 scale. Mean subdomain scores were 63.7 (SD 14.6) for leadership and health system messaging, 38.1 (SD 23.6) for data transparency and access, 53.4 (SD 18.8) for comfort with cost conversations, and 49.4 (SD 21.9) for blame-free environment. After stratifying by hospital type, there was a significant difference between groups in overall HVCCS scores (P < .001) and the leadership and health system messaging subdomain (P < 0.001) (Table 1).
Safety-net medical centers and high-value care culture.
In unadjusted data and multilevel regression modeling, residents from safety-net-based training programs had a lower mean HVCCS score (β = −4.4; 95% confidence interval [CI]: −8.2, −0.6). A four-point difference creates a meaningful separation because it would move a medical center from the top quartile to a lower quartile of high-value care culture among the medical centers. There is no separate association with the number of months that residents spent training within a safety-net medical center (Table 2). In multilevel regression modeling, male gender was also associated with lower HVCCS scores (β = −2.7; 95% CI: −4.7, −0.6) (Table 2).
In our sample, most residents at safety-net medical centers believe that their leadership aimed to promote health care value (78; 60.9%), were interested in participating in efforts to improve value (82; 64.1%), and believed that clinicians encouraged consideration of costs in clinical decision making (78; 60.9%). There was variation in resident responses to having open communication, role modeling of high-value care, and leadership actions that provided support to improve systems change, promoted value, and showed that value is a top priority (Table 3). Residents reported that previous efforts to promote value had not been met with success. They also reported poor access to both quality data and cost data, concern for blame in the work environment, and variation in agreement about clinicians having comfort with cost conversations (Table 3).
Relationship between HVCCS scores and institutional value-based outcomes
Excluding the four safety-net medical centers, we found that the mean institutional HVCCS scores among the four community-based and four university-based sites positively correlated with VBP scores (Spearman r = 0.71; P < .047). Among the four safety-net medical centers, HVCCS scores and VBP scores had a negative correlation (Spearman r = −1.0; P < .001). We expected this finding because the CMS VBP scoring does not adjust for some patient-specific factors, including socioeconomic status and race, and testing for a correlation may be limited for safety-net medical centers.1
In this cross-sectional study, residents across all study sites reported low average HVCCS scores across the domains of data transparency and access, comfort with cost conversations, and blame-free environment. Residents from safety-net-based training programs reported lower overall mean HVCCS scores, with significantly lower performance in the leadership and health system messaging domain. The HVCCS mean scores among the university and community sites also positively correlated with institutional value-based outcomes, which suggests an important association between having a higher-value care culture and value-based outcomes. Residency programs strive to create learning environments that support delivery of high-value care. Although previously published editorials have outlined the importance of high-value care culture shaping these environments,19–21 to our knowledge ours is the first study to describe and assess factors associated with high-value care culture among a diverse group of GME training programs.
GME and health system leaders have an opportunity to invest in their local culture and clinical training environments to drive high-value practice, as residents are the pipeline of the future health care workforce.5,7 The HVCCS was designed to identify targeted interventions that can support a high-value care culture for trainees or faculty. Individual sites can focus on areas identified by the HVCCS to improve the high-value care learning environment by increasing trainee exposure and engagement with health system leaders, supporting infrastructure for improvement efforts and training to residents and faculty role models, and openly discussing the quality and costs of care delivered.
Residents across all programs reported lower average scores in data transparency and access compared with the other three domain average scores. To improve those scores, all sites could consider initiatives to increase cost and utilization transparency among trainees. Although data transparency alone has not shown improvements in reducing expenditures, when combined with strategies to engage executive leadership, change culture, and educate clinicians, institutions such as the University of Utah have reduced costs.22,23
Residents across sites also reported lack of a blame-free environment. Training programs could focus on creating an environment that promotes open communication of poor outcomes and errors to avoid “defensive medicine.” Programs could also train residents to weight the risks and benefits of defensive medicine and rely more heavily on evidence-based practices.15
Only 25% to 30% of residents from all types of medical centers reported comfort with cost conversations, with little variation by medical center type. Perhaps this is an emerging needed skill set, as a prior study has suggested that two-thirds of patients report that they want to talk about costs with their physicians.24 Training programs could adapt commonly used teaching techniques for challenging patient–provider conversations to discuss the costs of care.25,26 As well, programs could consider using interactive modules or standardized patients.
Safety-net medical centers and high-value care culture
Nationally, numerous IM GME programs train residents at safety-net medical centers. Trainees gain unique insight into working with vulnerable patients and improving health equity, and residents who train within safety-net centers are more likely to return to practice in the same type of setting.27 We initially hypothesized that residents based out of safety-net medical centers may have reported greater high-value care culture because they would have greater exposure to the financial constraints faced by their patients, limited resources (i.e., potentially reducing their ability to overuse diagnostics and treatments), and non-fee-for-service-based physician payment arrangements (i.e., salaried).28–32 In this study, however, safety-net centers performed worse on measures of high-value care culture across all subdomains as reported by participating residents.
We hypothesize that these performance shortfalls of safety-net medical centers may reflect their well-documented longer delays in care and lower patient experience scores compared with their counterparts.31–36 Although we know that the safety-net centers in our study have low publicly reported spending compared with their university and community counterparts,37 value is a more complex concept than cost alone, and decision making incorporates a nuanced balance among quality, experience, and cost.
Immersion in the value culture of safety-net centers involves both daily patient–provider interactions as well as health system supports for value promotion. Safety-net centers traditionally have fewer resources to initiate a value-promoting infrastructure, which has been exacerbated by recent national policies that place safety-net centers at a greater financial disadvantage.14,33,38–41 Safety-net medical centers, therefore, may face greater challenges to fully streamline care coordination, which can increase inappropriate health care utilization, repeat diagnostic evaluations, and delays in obtaining appointments. Additionally, the patients seen at safety-net medical centers manifest greater socioeconomic challenges that affect health care utilization, including poor housing and food insecurity, and have limited access to technological advances that could be used to promote value-based care.42 In initial HVCCS development we found that infrastructural support was an important component of the largest domain, leadership and health system messaging. We also found that residents among safety-net-based programs reported lower scores in this domain, which may reflect, from their vantage points, the limited visibility of infrastructural support within resource-limited institutions.
The financial structures and drivers of medical center total costs of care and patient out-of-pocket costs are fundamentally different among the safety-net medical centers compared with the other study sites. Three of four sites deploy partial capitated payment models in their billing, which are payment arrangements to physician groups based on the number and complexity of patients assigned to them rather than the volume of services they deliver. These sites do not follow traditional billing practices. Additionally, unlike uninsured patients, indigent patients often do not have co-pays or deductibles, so clinical decisions affecting out-of-pocket patient costs may differ. It is unclear how these payment arrangements may affect institutional infrastructure and messaging around value. Further qualitative work is needed to characterize safety-net residents’ and faculty members’ understanding of value, including how financial structures and the combination of total costs of care, patient affordability, quality, and patient experience affect perceptions of culture.
Residents trained within these centers may benefit from having mixed clinical exposures, such as rotations within medical centers that have met value benchmarks.19 In parallel, safety-net medical centers can begin to target areas for improvement identified by residents through the HVCCS. For example, residents from our study reported interest in participating in efforts to improve value within their programs and indicated belief that leadership engagement feeds value promotion. These motivations can be leveraged through interventions targeting faculty and resident training to role model discussions about costs of care with patients; hold open discussions with trainees about health system value improvement goals; and provide resources to access quality performance and relative costs of care for alternative diagnostics, therapies, and care settings.7 Faculty can also begin to align resident value improvement projects with broader health system goals.42
Relationship between HVCCS scores and institutional value-based outcomes
Culture measured by other surveys, such as the Patient Safety Culture Survey, has been correlated with high-quality care outcomes including patient morbidity and mortality.43–45 High-value care culture as measured by the HVCCS also demonstrated a significant correlation between institutional high-value care culture and value-based outcomes at university and community medical centers. However, further study is needed to explore the best measures of institutional value-based care delivery that are valid across multiple care settings and to evaluate this relationship among a large number of medical centers with different patient populations.
Male gender also was associated with lower HVCCS scores. Further qualitative study will be needed to further explore gender differences in this area.
Our sample size creates a limitation in assessing the correlation between the HVCCS and institutional VBP scores. The magnitude of the complete correlation (r = −1.0) between safety-net centers and HVCCS scores is likely in part related to the small sample size. To further understand the correlation’s negative direction and residents’ understanding of value, next steps should include qualitative studies among trainees at safety-net medical centers.
There is also no clear gold standard metric to assess high-value care in the field. The VBP measure is currently the best standard available across our study sites—providing recent value data for over half of hospitals representing a broad array of hospital types—because it can be followed over time and it presents true policy implications. The VBP measure focuses on inpatient measures, some of which are out of the resident purview of influence; however, the majority of required service weeks are in the inpatient setting. The VBP also is limited because it does not adjust for patient factors.
While potential selection bias was limited in our study given the high completion rates among the residents sampled, our study also could be limited by generalizability because it evaluated high-value care culture among IM residents across 12 residency training centers in California alone. Although the study sites we chose represent various medical center types, institutional VBP performance, and size, a larger multisite study will be required to evaluate the range of variation across all regions of the country, ambulatory care settings, and other specialties in medicine. This study is also cross-sectional and may benefit from further evaluation of organizational culture over time.
As GME programs widely reflect on how their own training environments can support value improvement, the HVCCS can help to identify clear target areas such as improving data transparency and creating blame-free environments across GME training programs. Safety-net medical centers have a vital role in GME training to serve community needs, but trainees at those study sites reported lower scores in leadership and health system messaging of high-value care culture.46–49 All medical centers may benefit from increasing faculty training to role model value-promoting clinical decision making, use of transparent quality and cost data, and open communication about the value structure within their institutions.
The authors wish to thank Michael Lazarus, MD, Sara-Megumi Naylor, MD, and Cody Dashiell-Earp, MD, from the University of California, Los Angeles; Robert Wachter, MD, James Harrison, PhD, and Victoria Valencia, MPH, from the University of California, San Francisco; Mithu Molla, MD, from the University of California, Davis; Gregory Seymann, MD, from the University of California, San Diego; Bindu Swaroop, MD, and Alpesh Amin, MD, from the University of California, Irvine; Jessica Murphy, DO, and Danny Sam, MD, from Kaiser Permanente Santa Clara; Thomas Baudendistel, MD, and Rajeeva Ranga, MD, from Kaiser Permanente Oakland; Yile Ding, MD, from California Pacific Medical Center; Anshu Abhat, MD, MPH, from the LA BioMed Institute at Los Angeles County/Harbor-UCLA Medical Center; Steve Tringali, MD, from Community Regional Medical Center Fresno; and Dan Dworsky, MD, from Scripps Green Hospital, for their site leadership and participation with the study. They did not receive any funding or payment for their participation.
1. Centers for Medicare and Medicaid Services. The Hospital Value-Based Purchasing (VBP) Program. https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Value-Based-Programs/HVBP/Hospital-Value-Based-Purchasing.html
. Revised August 2, 201. Accessed January 10, 2019.
2. Chaudhry SI, Lien C, Ehrlich J, et al. Curricular content of internal medicine residency programs: A nationwide report. Am J Med. 2014;127:1247–1254.
3. Hackbarth G, Boccuti C. Transforming graduate medical education to improve health care value. N Engl J Med. 2011;364:693–695.
4. Ravasi D, Schultz M. Responding to organizational identity threats: Exploring the role of organizational culture. Acad Manage J. 2006;49:433–458.
5. Chen C, Petterson S, Phillips R, Bazemore A, Mullan F. Spending patterns in region of residency training and subsequent expenditures for care provided by practicing physicians for Medicare beneficiaries. JAMA. 2014;312:2385–2393.
6. Dzeng E, Colaianni A, Roland M, et al. Influence of institutional culture and policies on do-not-resuscitate decision making at the end of life. JAMA Intern Med. 2015;175:812–819.
7. Gupta R, Moriates C, Harrison JD, et al. Development of a high-value care culture survey: A modified Delphi process and psychometric evaluation. BMJ Qual Saf. 2017;26:475–483.
8. High Value Care Culture Survey. Identifying targets to improve high-value care culture. www.highvaluecareculturesurvey.com
. Revised March 30, 2018. Accessed January 10, 2019.
9. Chatterjee P, Joynt KE, Orav EJ, Jha AK. Patient experience in safety-net hospitals: Implications for improving care and value-based purchasing. Arch Intern Med. 2012;172:1204–1210.
10. Centers for Medicare and Medicaid Services. Disproportionate Share Hospital (DSH). https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/AcuteInpatientPPS/dsh.html
. Revised October 4, 2018.Accessed January 10, 2019.
11. Doximity. Doximity Residency Navigator 2018–2019. https://residency.doximity.com
. Revised August 2018. Accessed January 10, 2019.
12. Access Granted: The Primary Care Payoff. 2007.Bethesda, MD: National Association of Community Health Centers, The Robert Graham Center, Capital Link.
14. Favini N, Hockenberry JM, Gilman M, et al. Comparative trends in payment adjustments between safety-net and other hospitals since the introduction of the hospital readmission reduction program and value-based purchasing. JAMA. 2017;317:1578–1580.
15. Centers for Medicare and Medicaid Services. Medicare Program; Hospital Inpatient Value-Based Purchasing Program. Final rule. Fed Regist. 2011;76:26490–26547. https://www.federalregister.gov/documents/2011/05/06/2011–10568/medicare-program-hospital-inpatient-value-based-purchasing-program
. Accessed January 10, 2019.
16. Sirovich BE, Lipner RS, Johnston M, Holmboe ES. The association between residency training and internists’ ability to practice conservatively. JAMA Intern Med. 2014;174:1640–1648.
17. Accreditation Council for Graduate Medical Education. Data Resource Book: Academic Year 2016–2017. https://www.acgme.org/About-Us/Publications-and-Resources/Graduate-Medical-Education-Data-Resource-Book
. Accessed January 10, 2018.
19. Gupta R, Arora VM. Merging the health system and education silos to better educate future physicians. JAMA. 2015;314:2349–2350.
20. Sklar DP, Hemmer PA, Durning SJ. Medical education and health care delivery: A call to better align goals and purposes. Acad Med. 2018;93:384–390.
21. Gupta R, Sehgal N, Arora VM. Aligning delivery system and training missions in academic medical centers to promote high-value care. Acad Med. 2019;94:1289–1292.
22. Sedrak MS, Myers JS, Small DS, et al. Effect of a price transparency intervention in the electronic health record on clinician ordering of inpatient laboratory tests: The PRICE randomized clinical trial. JAMA Intern Med. 2017;177:939–945.
23. Lee VS, Kawamoto K, Hess R, et al. Implementation of a value-driven outcomes program to identify high variability in clinical costs and outcomes and association with reduced cost and improved quality. JAMA. 2016;316:1061–1072.
24. Alexander GC, Casalino LP, Meltzer DO. Patient–physician communication about out-of-pocket costs. JAMA. 2003;290:953–958.
25. Edinger W, Robertson J, Skeel J, Schoonmaker J. Using standardized patients to teach clinical ethics. Med Educ. 1999;4:4306.
26. Heron SL, Hassani DM, Houry D, Quest T, Ander DS. Standardized patients to teach medical students about intimate partner violence. West J Emerg Med. 2010;11:500–505.
27. Phillips RL, Petterson S, Bazemore A. Do residents who train in safety net settings return for practice? Acad Med. 2013;88:1934–1940.
28. Gusmano MK, Fairbrother G, Park H. Exploring the limits of the safety net: Community health centers and care for the uninsured. Health Aff (Millwood). 2002;21:188–194.
29. Felland LE, Felt-Lisk S, McHugh M. Health care access for low-income people: Significant safety net gaps remain. Issue Brief Cent Stud Health Syst Change. 2004;84:1–4.
30. Chen AH, Kushel MB, Grumbach K, Yee HF Jr.. Practice profile. A safety-net system gains efficiencies through “eReferrals” to specialists. Health Aff (Millwood). 2010;29:969–971.
31. Knudtson ML, Beanlands R, Brophy JM, Higginson L, Munt B, Rottger J; Canadian Cardiovascular Society Access to Care Working Group. Treating the right patient at the right time: Access to specialist consultation and non-invasive testing. Can J Cardiol. 2006;22:819–824.
32. Levin A. Consequences of late referral on patient outcomes. Nephrol Dial Transplant. 2000;15(suppl 3):8–13.
33. Andrulis DP, Siddiqui NJ. Health reform holds both risks and rewards for safety-net providers and racially and ethnically diverse patients. Health Aff (Millwood). 2011;30:1830–1836.
34. Neuhausen K, Katz MH. Patient satisfaction and safety-net hospitals: Carrots, not sticks, are a better approach. Arch Intern Med. 2012;172:1202–1203.
35. Gilman M, Adams EK, Hockenberry JM, Milstein AS, Wilson IB, Becker ER. Safety-net hospitals more likely than other hospitals to fare poorly under Medicare’s value-based purchasing. Health Aff (Millwood). 2015;34:398–405.
36. Jiang HJ, Reiter KL, Wang J. Measuring mortality performance: How did safety-net hospitals compare with other hospitals? Med Care. 2016;54:648–656.
: Hospital Compare. https://www.medicare.gov/hospitalcompare/search.html
. Revised August 2018. Accessed January 3, 2019.
38. Rosenbaum L. The whole ball game—Overcoming the blind spots in health care reform. N Engl J Med. 2013;368:959–962.
39. Neuhausen K, Davis AC, Needleman J, Brook RH, Zingmond D, Roby DH. Disproportionate-share hospital payment reductions may threaten the financial stability of safety-net hospitals. Health Aff (Millwood). 2014;33:988–996.
40. Taylor R, Bower A, Girosi F, Bigelow J, Fonkych K, Hillestad R. Promoting health information technology: Is there a case for more-aggressive government action? Health Aff (Millwood). 2005;24:1234–1245.
41. Zaman O, Cummings L, Laycox S. America’s Safety Net Hospitals and Health Systems, 2010: Results of the Annual NAPH Hospital Characteristics Survey. 2012.Washington, DC: National Association of Public Hospitals.
42. Lewis VA, Larson BK, McClurg AB, Boswell RG, Fisher ES. The promise and peril of accountable care for vulnerable populations: A framework for overcoming obstacles. Health Aff (Millwood). 2012;31:1777–1785.
43. Nieva VF, Sorra J. Safety culture assessment: A tool for improving patient safety in healthcare organizations. BMJ Qual Saf Health Care. 2016;12:ii17–ii23.
44. Mardon RE, Khanna K, Sorra J, Dyer N, Famolaro T. Exploring relationships between hospital patient safety culture and adverse events. J Patient Saf. 2010;6:226–232.
45. Berry JC, Davis JT, Bartman T, et al. Improved safety culture and teamwork climate are associated with decreases in patient harm and hospital mortality across a hospital system [published online ahead of print January 7, 2016]. J Patient Saf. doi:10.1097/PTS.0000000000000251
46. Schroeder SA, Zones JS, Showstack JA. Academic medicine as a public trust. JAMA. 1989;262:803–812.
47. White KL, Connelly JE. The medical school’s mission and the population’s health. Ann Intern Med. 1991;115:968–972.
48. Woollard FR. Caring for a common future: Medical schools’ social accountability. Med Educ. 2006;40:301–313.
49. Frenk J, Chen L, Bhutta ZA, et al. Health professionals for a new century: Transforming education to strengthen health systems in an interdependent world. Lancet. 2010;376:1923–1958.