Growth of hospital affiliations in California: impact on hospital charges and quality

Introduction

Since the 1970s, hospitals have increasingly joined systems through mergers and acquisitions, leading to significant consolidation in the healthcare industry (1-3). Beyond full mergers, hospitals have been strategically aligning by forming clinically integrated networks, joint ventures, and group purchasing organizations (4-6). These strategic alignments can enhance bargaining power and lead to higher prices without necessarily improving quality (7-9).

In addition, hospitals are establishing affiliation networks, which are defined as strategic partnerships between hospitals that maintain separate ownership but share resources and branding (10). Unlike full mergers, affiliations allow hospitals to collaborate while potentially retaining their independence, affecting competitive dynamics in healthcare markets (11). The decision for a sponsor and another hospital to affiliate is significant, as it can impact financial stability, quality of care, patient referral patterns, brand recognition, and hospital management (10,12). The goals of these affiliations are multi-pronged, including enhancing the quality of care provided by local hospitals and clinicians, and allowing patients to receive top-level care locally through collaborative networks, as opposed to merging, which requires greater strategic investment and might draw scrutiny from antitrust regulators (13,14). Sponsors may benefit financially and reputationally from these affiliations through referrals and increased brand awareness in affiliates’ markets (8,15).

However, sponsors and affiliates usually keep the details of these affiliation agreements private, including financial terms, resources available for affiliates, co-branding and marketing arrangements, patient referral obligations and incentives, and changes to hospital boards and management (12). In addition to broad affiliations, some affiliations are specific to high-risk conditions in the United States, such as cancer, which are among the most expensive to treat (16). While hospital affiliations in California have grown from 8 in 2012 to 33 in 2021 (Figure 1), little is known about the impact on charge and quality at affiliate hospitals.

Figure 1 Trends in hospital affiliations (by year). Authors analysis of data on affiliations collected using web search.

Research on hospital affiliations has been limited by several challenges, including difficulty in identifying sponsors and affiliates, the intensity of data collection required due to the lack of established datasets, variation in affiliation forms and offerings, and a previously small number of affiliations. However, a few studies exist. A pair of studies examined 203 hospitals that had, by 2015, affiliated with the Mayo Clinic, Cleveland Clinic, MD Anderson Cancer Center, or Memorial Sloan Kettering Cancer Center, all of which are nationally recognized health systems. Hospitals were more likely to join one of these affiliation networks if they had higher patient acuity, were located in areas of higher utilization intensity, were in urban areas, or were nonprofit (10). The second study used a causal-inference research design to compare financial outcomes and quality measures before and after the affiliation to a comparison group of hospitals that did not affiliate, which is, to our knowledge, one of the few studies to examine hospital-to-hospital affiliations that do not involve common ownership (12). They found that compared to the comparison group, affiliated hospitals experienced an increase in net income and operating margin but did not experience a difference in clinical quality measures, including mortality and readmissions. However, there was an improvement in patient experience scores, suggesting that affiliations may positively influence patient-perceived quality, possibly through shared resources, even if they do not significantly affect clinical outcomes.

Because of the lack of systematic data on affiliations, case studies and surveys have been employed to gain insights into the development of affiliations, including resource requirements and challenges. The University of Kentucky (UK) operates UK HealthCare, an academic medical center that has become a regional referral center through a network of affiliated hospitals and clinics (16,17). UK HealthCare’s strategy is to invest in its internal facilities and specialists, particularly for care involving transplants; hence, it was important to be a regional referral center to ensure sufficient patient volume for those needing complex care (16). Furthermore, UK HealthCare operates the Markey Cancer Center Affiliate Network. Its administrators and clinicians reported that it had successfully enhanced cancer care quality in Kentucky community hospitals through affiliations, emphasizing that network membership helped maintain cancer care accreditation, as well as the role of specialized services access, performance feedback, and a culture of quality improvement (17).

In summary, hospital affiliations—short of a merger—are becoming more common, but their impact on healthcare charges and quality has not been well studied. Prior research on hospital consolidation suggests that increased market power can lead to higher charges without corresponding improvements in quality. This study aims to fill that gap by focusing on affiliations in California, where there has been a significant increase in the number of affiliations since 2012. The study first identifies the predictors of whether a hospital chooses to affiliate and then uses a causal-inference model to estimate the impact of affiliating on hospital charges and quality. This research not only contributes to the sparse literature on hospital affiliations but also healthcare policy, because it is important to understand whether these strategic alliances are benefiting patients—either through higher-quality or less-costly care, or both—and whether they warrant more transparency and scrutiny from antitrust regulators.

Methods

Identifying hospital affiliations

We identified hospital affiliations in California through manual web searches from January 2012 to December 2021. Our strategy involved searching each hospital’s name in combination with the term “affiliation” across multiple sources, including hospital websites, news articles, and historical timelines. Web pages we found with information about the hospital’s affiliation included the websites of the hospital, news articles, and history timelines of the hospital. For each affiliation, we determined the start dates using a web search to verify or update announcement dates. Then we categorized each hospital as a sponsor or an affiliate. For example, to discover the affiliation for El Centro Regional Medical Center, we conducted a search using the term ‘El Centro Regional Medical Center affiliations’. As a result, we successfully found that it was affiliated with University of California (UC) San Diego. Moreover, with our awareness of the affiliation between El Centro Regional Medical Center and UC San Diego, we input the query “El Centro Regional Medical Center and UC San Diego affiliation” into the search bar. This led to the emergence of numerous internet sites and web documents, expanding our access to a wider array of information on the subject. On the UC San Diego Today news site, we successfully discovered precise details regarding the commencement date of the affiliation between El Centro Regional Medical Center and UC San Diego, as well as the comprehensive contents of the affiliation agreement. When encountering difficulties in obtaining information about the affiliations, we opted to investigate the hospital’s history timeline. By consulting the timeline, we aimed to uncover precise dates and gather additional details about any previous affiliations the hospital may have had. We rely on publicly available information to identify affiliation agreements, which can introduce limitations.

Study population

Our charge and quality analyses include non-profit, non-teaching, general acute care hospitals in California from 2012 through 2021, including 22 hospitals that were affiliated with a health system during the study period. Study hospitals met two criteria. First, the hospitals had to be viable candidates for affiliation membership given the specialties and directives of the sponsor hospitals. Thus, for-profit hospitals, federally owned facilities, psychiatric facilities, long-term care facilities, hospitals that treat alcoholism, hospitals that went through a merger or acquisition, hospitals focused on intellectual disabilities, and sponsor hospitals were excluded. Kaiser facilities were excluded as well. Second, the hospitals had outcome data available for all ten years of our study period. This led to one of the 22 affiliates dropping out of our charge analysis and two of the affiliates dropping out of our quality analyses. We restricted our analysis to non-profit, non-teaching, general acute care hospitals because these institutions are more likely to be viable candidates for affiliation based on their service profiles and strategic priorities. This selection helps to minimize heterogeneity in organizational structures and market strategies that could confound our results. However, we acknowledge that limiting our study to this subset may affect the generalizability of our findings to for-profit, teaching, or specialty hospitals.

Predicting affiliations

Following Jin and Nembhard (10) we consider case-mix index (patient acuity), patient care (utilization intensity), market competition, and net income as the major motivators for a hospital to affiliate. To determine predictors of affiliating with a sponsor, we estimated a logit model where an indicator variable for affiliation status served as the dependent variable. The following variables that are shown in Table 1 were included as independent variables in the model.

Case-mix index/patient acuity

Patient acuity was determined based on a hospital’s case mix index (CMI) using data from the Department of Healthcare Access and Information (HCAI) Annual Financial Report. The CMI of a hospital is calculated as the average diagnosis-related group (DRG) relative weight for that hospital. DRGs indicate the type of care a patient receives and correspond to payment categories. The CMI is computed by summing the DRG weights for all Medicare discharges and dividing it by the number of discharges (10,18). Therefore, it provides an indication of the diversity, clinical complexity, and resource needs of all patients in a hospital. CMI is a widely accepted indicator for hospital disease severity and is available for most hospitals in California.

Utilization intensity/Medicare cost

Building upon previous research (10), we assessed the propensity for higher or lower healthcare service utilization in a hospital’s region by utilizing standardized, risk-adjusted Medicare costs per capita data at the county level. This approach aims to account for regional variations in healthcare costs. Standardizing Medicare costs involves removing geographic differences in costs, while risk-adjusting accounts for variations in the health status of beneficiaries across different regions. These adjustments help to better capture physician practice patterns and beneficiaries’ willingness and ability to seek healthcare services. The data on utilization intensity was obtained from the Centers for Medicare and Medicaid Services (CMS) geographic variation file, which provides information on regional variations in healthcare utilization across the United States.

Market competition/HHI

Market competition was measured by calculating the market concentration in each hospital’s county using the Herfindahl-Hirschman Index (HHI), a commonly used metric to estimate market concentration (19). The HHI is computed by summing the squared market shares of each hospital in the county, based on patient admissions reported in the American Hospital Association (AHA) survey data. By squaring the market shares and summing them, the HHI places greater weight on larger hospitals, reflecting their relative dominance in the market. Because the market shares are measured as percentages, the resulting HHI values range from just above 0, indicating a perfectly competitive market, to 10,000, indicating a monopoly. We divide the HHI by 100 in our analysis for ease of interpretation.

Financial security/net income

One of the most significant factors that can influence a hospital’s decision to join an affiliation with a health system is financial considerations (10). Hospitals may consider affiliating with a health system to gain access to greater financial resources, including capital for investments in technology and infrastructure, as well as funding for research and development. Financial security was assessed by examining a hospital’s net income, which is defined as the surplus of revenues over expenses. Net income serves as a crucial metric for evaluating organizational profitability. In our analysis, we scaled the net income variable to enhance interpretability, such that each unit represents $1 million. To obtain data on financial security, we utilized the HCAI Annual Financial Data, which provides comprehensive financial information for hospitals in California.

Hospital and county characteristics

Based on past studies, we included five covariates in our analyses that are potential confounders of the relationship between study variables and participation in an affiliation network: type of ownership [for-profit, non-profit, government (city/district)], teaching status, urban/rural location, and the number of total hospital beds. Nonprofit status, type of ownership, teaching status, and rural location were binary variables coded as 1 when the characteristic is present. These data are compiled from HCAI. Additionally, we included county-level characteristics in our model—population, utilization, unemployment rate, median household income, and number of physicians per 100,000 population. We calculated descriptive statistics (means, standard deviations, and frequencies) to summarize hospital characteristics and outcome variables. We tested for the difference in means for treatment and comparison groups using a two-sample t-test and report the results in Table 2.

Estimating the impact of affiliations on charge and quality

We estimated difference-in-differences and event study models to measure the effect of affiliations on the charge and quality of affiliates. We begin this section by describing our charge and quality dependent variables and then proceed to the details of our econometric models.

Hospital charge

Our charge analysis is conducted at the discharge level, with each discharge serving as our measure of hospital charge. The amounts insurers and patients actually pay (i.e., negotiated rates) can differ substantially from charges, but a recent study showed “commercial negotiated rates were typically calculated consistently in increments of a 5 percent discount from chargemaster prices” so we expect affiliations that increase hospital charges would also increase negotiated rates (20).

Hospital quality

Using publicly available data from CMS Hospital Compare, we assessed hospital performance on quality. The data included three measures: all-cause rate of readmission within 30 days after discharge, the rate of death within 30 days after admission, and patient experience from 2012–2021. The 30-day readmission metric quantifies the proportion of patients who are rehospitalized (unplanned) within a 30-day period following their initial discharge, irrespective of the hospital to which they return. This indicator spans a wide array of healthcare services, including medical, surgical, gynecological, neurological, cardiovascular, and cardiorespiratory, monitored from 2012 through 2021. An elevated readmission rate may reflect on the quality of care provided by the hospital, with higher rates potentially indicating inferior care quality. Meanwhile, the 30-day mortality rate is concerned with the mortality rates among surgical patients who encounter serious yet manageable complications post-surgery. This statistic has been consistently tracked for six specific conditions: acute myocardial infarction (AMI), coronary artery bypass grafting (CABG) surgery, chronic obstructive pulmonary disease (COPD), heart failure (HF), stroke (STK), and pneumonia (PN) from 2012 to 2021. A higher mortality rate may suggest deficiencies in post-operative care or emergency response, thus implying lower quality of hospital services. Furthermore, the Patient Experience metric is derived from the CMS Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) Survey, which collects opinions from a random sample of adult patients discharged between 48 hours and six weeks prior. This survey evaluates several aspects of patient care, including the quality of communication with nurses and doctors, the responsiveness of hospital staff, the clarity of discharge instructions, and the overall patient rating of the hospital. The scores are on a scale from 0 to 100, with higher scores denoting more favorable patient experiences and, correspondingly, higher hospital quality.

We implemented a difference-in-differences (DIDs) model to estimate the impact of affiliations on charges and quality. The treatment group includes non-profit, non-teaching hospitals affiliated with a sponsor hospital between 2013 and 2021. Hospital affiliations in 2012 or earlier are not included in the model as they would be treated for the entire study period. The comparison group includes all non-profit, non-teaching hospitals that were not affiliates, sponsors, or targets in hospital mergers during the study period. In the DID model, before affiliating, affiliates are considered to be unaffiliated, but in the year the affiliation occurs, the model considers them to be affiliated from that year forward. Furthermore, our analysis considers only the non-sponsor hospitals as affiliates; sponsor hospitals are excluded from the quality and charge analyses. This approach ensures that the observed effects can be attributed specifically to the strategic partnership of the affiliate hospitals rather than the sponsor institutes. Our estimation equation for our charge analysis is

Yihct =ϕAff0jt+βAff2jt+ζZit+ΩWht+γXct+αj+τt+εihct[1]

where i indexes discharges, j indexes hospitals, c indexes counties, and t indexes years; Aff0 is an indicator variable equal to one for affiliates in the year of affiliation and the first full year post-affiliation, Aff2 is an indicator variable equal to one for affiliates in years two or more post-affiliation, Z is a vector of discharge-level characteristics—length of stay, DRG fixed effects, and the age and sex of the patient; X is a vector of county-level characteristics—population, median household income, unemployment rate; and ε_i,h,c,t is the error term. We included hospital (α) and year (τ) fixed effects in our model, CMI as a covariate, and clustered standard errors at the hospital. All hypothesis tests were two-tailed with a significance level set at 0.05. The separation of the affiliation variable into Aff0 and Aff2 is meant to capture the fact that hospital reimbursement contracts are generally renegotiated every 2 to 4 years, so it is unlikely that any measurable charge effect would be detected immediately upon affiliation. As such, Aff2 is our variable of interest and thus an estimate of β is the average treated on the treated (ATT) coefficient we focus on for the remainder of the paper. Our quality estimating equation is similar to Eq. [1] with the differences attributable to the quality dataset being at the hospital-level rather than the discharge-level (i.e., no discharge-level characteristics included as control variables). We also utilized an event study framework to validate the parallel trends assumption. All analyses were performed using Stata 16.

The above equation estimates the change in hospital charges following an affiliation. For a given post-transaction year, the differential change represents the difference between the observed charge for affiliated hospitals and their expected charge/quality if the pre-transaction difference had remained unchanged in the post-transaction period (i.e., the estimated effect of affiliation). Equivalently, the model compares the average difference between affiliated hospitals and comparison hospitals during the pre-transaction period with the difference in each post-transaction year. We also estimate event study versions of Eq. [1] in order to detect any parallel trends violations and estimate dynamic effects.

Results

The following sections provide empirical evidence on the trends in affiliations, predictors of affiliations, and whether the hoped-for improvements in quality and charges were realized.

Affiliation trends and details

Figure 1 shows the number of hospital affiliations by year in California from 2012 to 2021. In 2012, there were 8 affiliations. The number of affiliations increased steadily at a rate of 1–5 per year, culminating in 33 affiliations in 2021—a four-fold increase reflecting a significant shift towards strategic collaboration in healthcare. In general, these affiliations are pursued to enhance access to specialized care, improve administrative efficiency, and expand service offerings. The table (available online: https://cdn.amegroups.cn/static/public/jhmhp-24-134-1.docx) lists the sponsor hospitals, their affiliated hospitals, and the primary objectives stated in each affiliation agreement. These affiliations, marked by a diverse range of focuses including administration, general care, specialized medical services, research, and teaching, indicate a significant trend towards collaborative efforts in enhancing healthcare delivery and governance across different regions. Starting with the affiliations between Adventist Health and multiple hospitals (Rideout, Lodi Memorial, Mendocino Coast, and Dameron Hospital), these partnerships initiated between 2018 and 2020 highlight a strategic move towards managed healthcare services. Adventist Health’s role in these affiliations extends to managing hospital services, with notable governance changes such as the sponsor becoming the president of Lodi Memorial and obtaining 13 board seats in Mendocino Coast. These affiliations signify a push towards expanding healthcare options and ensuring financial stability through backing from the affiliate sponsor. Other notable affiliations include those with University of California San Francisco (UCSF), University of California San Diego (UCSD), University of California Irvine (UCI), which focus on expanding access to specialized care and integrating innovative healthcare models. For instance, the affiliation between Dominican Hospital and UCSF in 2017 emphasizes broadening cancer services, establishing new primary care clinics, and expanding mental health services, without altering the hospital’s governance structure. Similarly, affiliations with UCSD, like El Centro Regional Medical Center and Tri-City Medical Center, aim to enhance strategic planning and healthcare service management, with significant governance changes like UCSD being given five board seats in the Tri-City Medical Center affiliation. The partnerships also reveal a focus on specific healthcare areas like behavioral health, rehabilitation, and specialized surgery. For example, the affiliation between Antelope Valley Hospital and Kindred Health in 2020 aims to operate and build a behavioral health and rehabilitation hospital, showcasing a targeted approach to addressing specific healthcare needs without altering governance. Pediatric care, research, and teaching emerge as focal points in several affiliations, such as the partnership between Valley Children’s Hospital and UHS in 2019, which centers on enhancing behavioral health services and establishing a new psychiatry residency program. These affiliations often do not involve changes in governance but focus on expanding healthcare services and access. Significant investments and strategic initiatives mark some affiliations, like the partnership between San Ramon Regional Medical Center and John Muir Health in 2013, which includes a $100 million investment and the development of outpatient services. This indicates a financial and strategic commitment to healthcare improvement and expansion. The ending of Hoag Memorial Hospital Presbyterian and Hoag Orthopedic Institute from Providence in 2022 is a unique case, indicating the dynamic nature of healthcare affiliations and the potential for reevaluation and restructuring over time.

In summary, these affiliations collectively demonstrate a strategic shift in healthcare towards strategic collaborative models that aim to enhance healthcare quality, access, and specialization. The table reflects a broad spectrum of affiliations, from financial investments and management changes to expansions in specialized care and research initiatives. These partnerships are pivotal in shaping the future of healthcare delivery, indicating a trend towards integrated healthcare systems that leverage the strengths of both hospitals and healthcare sponsors to meet the evolving needs of communities. Through these affiliations, hospitals not only gain access to expanded resources and specialized services but also engage in governance restructuring to align more closely with the strategic goals of their healthcare sponsors.

Baseline differences

Table 2 provides an overview of the baseline (year 2012) differences between affiliate and non-affiliate hospitals in terms of the hospital and county characteristics used in our affiliation prediction model (case-mix index, utilization, net income, HHI, bed count, urban/rural location), other county characteristics and our outcome measures—charge, readmission, mortality, and patient experience. There were no significant differences in case-mix index, utilization, net income, market competition, or bed count between the two groups. There were statistically significant differences between affiliate and non-affiliate hospitals in location intensity (90% affiliates were located in urban regions). Affiliate hospitals were also located in counties with higher physician count per 100,000 population, as compared to non-affiliates (304 vs. 227 for non-affiliates). There were no significant differences in outcomes at the baseline—hospital charge, readmission rate, mortality rate, and patient experience between affiliates and non-affiliates at the start of our sample period.

Predictors of affiliations

Table 1 presents the results of our analysis examining the factors associated with hospitals joining affiliation networks. It suggests that hospitals with a higher case-mix index/more complex patients are more likely to become affiliates [odds ratio (OR) =2.36, 95% confidence interval (CI): 1.15–4.83]. Hospitals located in areas of higher utilization intensity, as evidenced by higher Medicare costs, are less likely to affiliate than those in areas with less utilization intensity (OR =0.96, 95% CI: 0.94–0.99). Urban hospitals are more likely to affiliate than rural hospitals (OR =4.78, 95% CI: 1.97–11.57). Hospitals with higher net income are less likely to affiliate (OR =0.59, 95% CI: 0.36–0.98). Market competition does not predict the likelihood of hospitals affiliating with health systems.

Impact on charge and quality

Table 3 shows the impact of hospital affiliation on charge, readmission, mortality, and patient experience. Charges at affiliate hospitals increased by 8.1% (P=0.04, 95% CI: 0.6% to 16.2%) relative to control hospitals. There were no statistically significant differences in readmission or mortality rates following affiliations. Patient experience was statistically significant (P value =0.01) and increased by 2 percentage points for affiliates. Figure 2 shows the event study results, suggesting a delayed increase in patient experience, starting four years after affiliation. These findings suggest limited improvement in affiliate hospital quality with evidence of significant charge increases.

Figure 2 Event study results for charge and quality. Data on hospital characteristics and charges from the Department of Health Care Access and Information (2012–2021).

Discussion

As of 2021, 33 hospitals in California had affiliated with sponsor hospitals, usually academic medical centers and health systems. The four-fold increase in hospital affiliations from 8 in 2012, as detailed in our descriptive analysis, provides important context about market dynamics. As more hospitals enter into affiliations, market dynamics and care coordination practices are reshaped, ultimately affecting charges and quality. We found that hospitals in urban areas and those with more complicated case mixes were more likely to affiliate with a sponsor, likely so that patients with complicated conditions could be referred to the sponsor. Methodologically, our study employs a DID framework that incorporates hospital and year fixed effects to control for unobserved, time-invariant factors and common temporal shocks. The pre-trends analysis supports the validity of the parallel trends assumption, indicating that affiliated and non-affiliated hospitals followed a similar trajectory before the affiliation. As compared with non-affiliated hospitals, the affiliates experienced an 8.1% charge increase and were 3.3% higher on patient experience measures, but no statistically significant impact on readmissions and mortality was found. While this suggests that affiliations may enhance a hospital’s bargaining power and improve aspects of care coordination, potentially reflected in better patient experience, the absence of significant changes in clinical outcomes raises important questions. Specifically, it remains unclear whether the higher charges incurred by affiliated hospitals are justified by the improvements in care quality.

Our findings are consistent with two studies by Jin and Nembhard, who examined affiliations that did not involve an ownership relationship, the same type of affiliation that we examined (10,12). We found that affiliates experienced higher charges, which might be due to an enhanced negotiating position with payers because of the reputational effect of being affiliated with a well-known sponsor hospital. We also did not find an effect on mortality and readmission rates. These null findings could be due to several factors, such as the lack of necessary resources and expertise being transferred from the sponsors to their affiliate hospitals, the measure being too broad to capture an affiliation that focuses on a narrow clinical scope, or the quality improvements occurring at the sponsor hospital, such as when high-acuity patients are transferred from an affiliate to a sponsor. But in contrast to Jin and Nembhard, we found a positive effect on patient experience measures, perhaps because we used a broader set of measures, including patient communication measures, which may have improved from better coordination of care between the affiliate and sponsor hospital. This coordination was more likely to occur in our study because the sponsors generally had only a few affiliate hospitals, whereas the four sponsors in Jin and Nembhard had an average of 51 affiliated hospitals. Although our primary quality measures include readmission and mortality rates, these indicators may not fully capture the broader effects of hospital affiliations. Alternative measures such as outpatient visit rates, care coordination metrics, referral patterns, and comprehensive patient satisfaction surveys could potentially offer a more complete assessment of the value generated by these affiliations.

Affiliations are one type of strategic alliance among hospitals in which the affiliate gains resources, training, and expertise from the sponsor, with the aim of improving quality. Because the sponsors tend to be academic medical centers and health systems, they are better equipped to treat complex patients, with the affiliation enabling the affiliate hospitals to routinely refer these patients to such hospitals. Although sponsor-affiliate agreements are not considered to be a full merger or acquisition, such transactions fall under the purview of federal and state antitrust agencies. For example, the federal Merger Guidelines highlight such transactions, particularly those that involve competitors or give the sponsor rights in the affiliate, such as the right to appoint board members, which was a part of several affiliation agreements in California (see guideline 11) (19). If warranted, the antitrust agencies will assess the pro- versus anti-competitive effects of an affiliation to determine, on net, whether it improves competition by examining prices, output, quality, and innovation. However, the sponsor and affiliate hospitals in many of these transactions are not required to file a pre-affiliate notification because the value of the transaction is less than the threshold that requires a filing, currently $126.4 million at the federal level, making it more difficult for antitrust agencies to review these transactions, particularly the terms of affiliation.

Our findings provide valuable insights for policymakers and hospital administrators by highlighting the factors influencing whether to affiliate and the effect on charges coupled with a limited impact on quality. However, it is important to consider the limitations of our study, and further research is needed to expand upon these findings. First, a limitation of our data collection method is its reliance on publicly available information. Consequently, we may have underestimated the number of affiliations because many are not reported and there are no reporting requirements for hospital affiliations, which could bias our results toward the null. We expect such instances to be rare and to have a minimal effect on our results. Second, our results represent average effects of affiliations, potentially obscuring the benefits or harms of specific affiliations. Third, obtaining detailed information about affiliation agreements was challenging. The lack of specific information within affiliation agreements limits the dataset’s granularity. Fourth, while we examined a broad set of quality measures, we were unable to capture all dimensions of quality. Fifth, our difference-in-differences analyses, in which exposure to treatment is voluntary, are susceptible to potential selection bias. Finally, the restriction to non-profit, non-teaching hospitals may limit the external validity of our findings. Hospitals with different ownership or teaching statuses might experience affiliations differently, and future research should examine these groups to provide a more comprehensive picture. Researchers and stakeholders should consider the potential gaps in historical coverage, the challenges associated with accessing and obtaining detailed information from affiliation agreements, the discrepancies in reporting practices, the dynamic nature of affiliations, incentives such as referrals and branding, and the regional focus when analyzing similar data.

Conclusions

The number of hospital affiliations in California has significantly increased since 2012, having reached 33 affiliations in 2021. Affiliations represent a growing type of strategic alliance that does not involve a merger. Hospitals, regulators, and policymakers should continue to monitor the significant and growing trend of affiliations in the healthcare industry to ensure they do not increase charges without a commensurate improvement in quality. This also suggests that affiliations may enhance a hospital’s bargaining power with payers, enabling higher list charges, while concurrently fostering better care coordination that improves patient experience. Ongoing evaluation and appropriate regulatory oversight are essential to ensure that hospital affiliations serve the public interest by promoting high-quality, affordable healthcare.

Acknowledgments

The authors thank Natalie Zazula and Jordan Wolf for their research assistance.

Footnote

Peer Review File: Available at https://jhmhp.amegroups.com/article/view/10.21037/jhmhp-24-134/prf

Funding: This study was funded by Arnold Ventures (No. 20-05350) to the Regents of the University of California, Berkeley campus.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jhmhp.amegroups.com/article/view/10.21037/jhmhp-24-134/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Balan DJ, Brand K. Simulating hospital merger effects in differentiated product markets with quality competition. J Ind Econ 2020;68:260-96.
2. Cutler DM, Scott Morton F. Hospitals, market share, and consolidation. JAMA 2013;310:1964-70. [Crossref] [PubMed]
3. Dranove D, Lindrooth R. Hospital consolidation and costs: another look at the evidence. J Health Econ 2003;22:983-97. [Crossref] [PubMed]
4. Burns LR, Briggs AD. Hospital purchasing alliances: Ten years after. Health Care Manage Rev 2020;45:186-95. [Crossref] [PubMed]
5. Ridgely MS, Buttorff C, Wolf LJ, et al. The importance of understanding and measuring health system structural, functional, and clinical integration. Health Serv Res 2020;55:1049-61. [Crossref] [PubMed]
6. Hague EL, Brown TT, Brewster A, et al. Hospital Characteristics Associated With Clinically Integrated Network Participation. Med Care 2023;61:521-7. [Crossref] [PubMed]
7. Melnick G, Keeler E. The effects of multi-hospital systems on hospital prices. J Health Econ 2007;26:400-13. [Crossref] [PubMed]
8. Cuellar AE, Gertler PJ. Strategic integration of hospitals and physicians. J Health Econ 2006;25:1-28. [Crossref] [PubMed]
9. American Hospital Association. (2024). Fast facts on U.S. hospitals – 2024. Chicago, IL. Available online: https://www.aha.org/system/files/media/file/2024/01/fast-facts-on-us-hospitals-2024-20240112.pdf
10. Jin B, Nembhard IM. Who Joins the Franchise, Affiliation Model of Hospital Networks? An Analysis of Hospital and Market Characteristics of Members. Med Care Res Rev 2021;78:660-71. [Crossref] [PubMed]
11. Vogt WB, Town R, Williams CH. How has hospital consolidation affected the price and quality of hospital care? Synth Proj Res Synth Rep 2006;15231.
12. Jin B, Nembhard IM. Effects of affiliation network membership on hospital quality and financial performance. Health Serv Res 2022;57:248-58. [Crossref] [PubMed]
13. Tsai TC, Jha AK. Hospital consolidation, competition, and quality: is bigger necessarily better? JAMA 2014;312:29-30. [Crossref] [PubMed]
14. Haas-Wilson D, Garmon C. Hospital mergers and competitive effects: two retrospective analyses. Int J Econ Bus 2011;18:17-35.
15. Krishnan R. Market restructuring and pricing in the hospital industry. J Health Econ 2001;20:213-37. [Crossref] [PubMed]
16. Gao X, Schroeder MC, Lizarraga IM, et al. Improving cancer care locally: Study of a hospital affiliate network model. J Rural Health 2022;38:827-37. [Crossref] [PubMed]
17. Edwards RL, Lofgren RP, Birdwhistell MD, et al. Challenges of becoming a regional referral system: the University of Kentucky as a case study. Acad Med 2014;89:224-9. [Crossref] [PubMed]
18. Mendez CM, Harrington DW, Christenson P, et al. Impact of hospital variables on case mix index as a marker of disease severity. Popul Health Manag 2014;17:28-34. [Crossref] [PubMed]
19. U.S. Department of Justice & Federal Trade Commission. (2023). Merger Guidelines. Washington, DC. Available online: https://www.ftc.gov/system/files/ftc_gov/pdf/P234000-NEW-MERGER-GUIDELINES.pdf
20. Wang Y, Meiselbach MK, Cox JS, et al. The Relationships Among Cash Prices, Negotiated Rates, And Chargemaster Prices For Shoppable Hospital Services. Health Aff (Millwood) 2023;42:516-25. [Crossref] [PubMed]