When recognizably modern hospitals appeared in the late nineteenth century, the physician hierarchy was detached from the administrative leadership structure.14 Although this separation still exists, increasing integration is gradually obscuring the distinction between clinical leadership and organizational management.

Our findings show that certain features of healthcare organizational structures may carry predictive potential when applied to the US News and World Report rankings. Physician CEO was positively associated with US News and World Report rank. Furthermore, univariable analysis suggested a possible association between number of physician executive team members and US News ranking, a finding that raises important and complex questions about the importance of physician leadership and its relationship to healthcare quality as measured by US News and World Report. Previous studies investigating the frequently observed link between physician leadership and hospital performance have offered several explanations. In “Why the Best Hospitals Are Managed by Doctors,” Stoller et al. propose that physician-leaders who earn credibility in the clinical domain understand the needs and challenges of clinicians working underneath them and bring a patient-focused culture that non-clinician leaders are less adept to replicate.15 A different study showed that physician business leaders frequently bring a diverse collection of professional goals and compensation practices to their respective institutions that prioritize the patient experience, improve healthcare provider satisfaction, enhance institutional loyalty among workers, and strengthen employee trust in leadership.16 Another study found that among the U.S. News and World Report top 100 hospitals, there was a 25% difference in overall quality between physician-led hospitals and non-physician-led hospitals.17 These findings suggest that technical expertise at the executive level not only increases worker happiness and productivity but is likely among the most influential variables contributing to worker satisfaction.18

The results of the current analysis suggest an association between the flagship bed size and US News and World Report rank, an observation at least partially attributable to the well-established notion that providing highly specialized care is only financially viable within institutions of sufficient scale. Niche specialties such as cardiac oncology and transplant infectious disease can only exist at hospitals large enough to offset the costs associated with providing such specialized care.19 Although the odds ratio per unit change was smaller for bed count compared to other continuous variables, this could be attributed to the fact that unlike other continuous variables in the multivariable model that are less numerous, bed count has much wider range often exceeding 1000. In addition, our findings suggest that this observation may not be exclusively attributable to the influence of combining hospitals into larger health systems that inevitably have more beds. Furthermore, although consolidation may provide short-term financial gain by decreasing competition and centralizing resource allocation, no study has demonstrated that the financial costs of salaried, non-clinical administrative staffing are offset by superior patient care or efficiency.20

Our findings suggest that institutions utilizing group practice models achieve greater incentive alignment than their non-group practice competitors, a valuable element of the institutional architecture that fosters cooperation and collaboration between specialists. Although previous research in the social sciences has repeatedly demonstrated that cooperation and competition are not mutually exclusive, group practice models are more often associated with a collaborative work environment while overly competitive incentives can discourage mutually beneficial practices.21

Throughout the preceding decades, corporate influence and financial stakeholder involvement in United States healthcare has accelerated.22 Advocates of the for-profit model emphasize the positive impact of market pressures on innovation and the value of competition for maximizing the efficiency of healthcare delivery and reducing costs.23,24 Critics argue that greater involvement of profit-seeking financial institutions and corporate actors negatively impacts hospitals in a variety of essential ways. Quality improvement studies have repeatedly demonstrated a strong association between private equity ownership and the erosion of the patient-physician relationship and increased likelihood of physician burnout and moral injury.25,26 This reduces the relationship to transactional rather than focusing on cultivating trusting and transparent relationships with patients.27 Our finding that for-profit status was associated with lower US News and World Report ranking is aligned with these reports, and although for-profit institutions comprise a minority of all health systems, they represent a sizeable portion with respect to hospital number and bed count and should not be overlooked.

Our study contains several limitations that are important to acknowledge. First, our analysis was limited by the number of available predictor variables. It is possible that our findings could have revealed different associations, perhaps even supporting alternate conclusions, had additional predictor variables been available and included in the analysis. Our approach was to analyze all available variables simultaneously in a multivariable model, and variables found to be significantly associated with the outcome in the multivariable model were analyzed individually in univariable analyses. Adherence to this approach limited our capability to identify relationships within data-scarce territories. For example, our study could not account for heterogeneity between for-profit and not-for-profit sub-organizations of a non-profit parent because the data volume and granularity required to perform such an analysis is currently lacking. Furthermore, we sought to maintain an accurate and consistent definition of “group practice” without conflating the relationship between physicians practicing within a single group practice and other relationships that exist between administrative and clinical personnel. In pursuit of this aim, our study defined “group practice” using the AHRQ definition as a group of physicians that offer services under a shared entity, and this definition would have excluded non-integrated groups of physicians who provide patient care as a collective entity which could impact our results and conclusions. Lastly, demographic leadership variables such as age, gender, and medical specialty were not amenable to systematic collection and were not pursued in this analysis.

Our investigation focused on a single ranking system, increasing the study susceptibility to inaccuracies within the US News and World Report institutional practices and vulnerability to the influence of biases within their methodological processes. The limitations of US News and World Report hospital ranking metrics are well-documented, and although institution rank has never been demonstrated to correlate with patient outcomes, concluding that because hospital ranking is correlated with volume which has repeatedly been linked to patient outcomes suggests the presence of a connection between patient outcomes and US News and World Report ranking could be misguided. Furthermore, reputational score was part of the US News and World Report ranking and, as such, a subjective metric made its way into a numeric score. Therefore, we recognize that US News and World Report hospital rankings represent imperfect surrogates for health system performance and the overall ranking system remains a controversial measure of institution quality which potentially lessens the value of our conclusions. Such observations underscore the importance of exercising caution when interpreting and extrapolating data.

US News and World Report relies on data provided by institutions and third-party entities, so our study would be negatively impacted by any shortcomings or substantial omissions in data reporting, as well as the possibility that the strength of association between hospital bed count and US News and World Report rank is impacted by their methodologic processes. Lastly, because there is no simple binary classifier for a coarse-grained description of an institution’s academic status, exploring the impact of academic affiliation on health outcomes quality was determined to be beyond the scope of analysis for this study.

In summary, our study highlights an array of factors that are associated with superior institutional performance as measured by the annual US News and World Report rankings. It is likely that physician leadership, incentive practices, flagship hospital size, and not-for-profit status all contribute to organizational performance in complex ways, of which some are better understood than others. Nevertheless, we hope that our findings may be used to inform policymaking at the state or national governmental level while acknowledging that if we strive to deliver healthcare that is effective, affordable, and equitable, our leadership models and governance practices must be evidence-based, and our healthcare institutions and decision-makers within them must reflect the values and priorities of the patients they serve.