Vertical integration in oncology: what does it mean for patients with cancer?

The health-care industry in the United States has witnessed market consolidation since the 1990s (1,2). As mentioned by Alpert and colleagues (2), consolidation started out as horizonal integration (eg, merging with similar health-care entities such as hospitals or physician practices) and progressed to vertical integration (eg, integrating physician practices to hospitals) such that it has become the dominant form of consolidation in the health sector. Vertical integration of physicians and hospitals has introduced a major shift in care delivery in the United States. Economic studies suggest that the advantages of vertical integration include mitigation of quality deficiencies through better care coordination and the potential to curb health-care costs from more streamlined process in care (eg, avoiding duplicate services) and administration (3). Economic studies also voiced concerns that higher market concentration as a result of vertical integration would increase hospitals’ bargaining power over private health insurers, which could lead to higher prices without noticeable gains in the quality of care (3-5).

Research has found that physician–hospital integration is more common for specialties providing lucrative care to hospitals, such as cardiology and oncology (6). A sharp increase in vertical integration was observed in the oncology market, growing from 30% before 2009 to 60% in 2015 (2). Will vertical integration in oncology improve the welfare of patients with cancer? On the one hand, better care coordination in an integrated system could reduce treatment delays and improve access to multidisciplinary teams, which has been shown to be associated with better outcomes (7). On the other hand, patients, especially those with private insurance, may find their out-of-pocket (OOP) payment growing faster as the market becomes more concentrated. Moreover, patients with less generous insurance coverage could face a shrinking number of in-network providers and may have to choose between travelling longer distances to stay within network or incurring higher OOP payments associated with receiving care from out-of-network providers of oncology care. Given the trend toward widespread vertical integration in the oncology market, it becomes increasingly important to understand how oncologist–hospital vertical integration affects patients with cancer. Hu and colleagues (1) offered one of the few empirical studies in oncology to address this important topic.

Using 2008-2017 Surveillance, Epidemiology, and End Results–Medicare linked data, Hu and colleagues (1) investigated the impact of oncologist–hospital vertical integration on health services utilization, Medicare payments, and quality of care among patients with metastatic castration-resistant prostate cancer (mCRPC). They creatively linked the Provider Practice and Specialty File to Medicare claims to quantify vertical integration through billing patterns, which enabled them to categorize oncologists into 3 groups: nonintegrated, always integrated, and those who switched from nonintegrated to integrated (ie, the “change” group). They then took advantage of the longitudinal data structure and applied the difference-in-differences method to estimate the effect of oncologist–hospital integration by comparing between the “change” and “unchanged” groups using this elegant study design. They concluded that vertical integration increased the use of bone-modifying agents but did not statistically significantly change the duration of systemic therapy, survival, and end-of-life resource utilization. Interestingly, although the decomposition analysis found that the magnitude of increase in outpatient payment ($5190) was larger than the decrease in professional service payment (−$4757) when oncologists switched billings from a physician office to hospital outpatient (ie, the change group), there were no statistically significant changes in the total per-patient Medicare payment in the first 3 months following the initiation of systemic therapy.

The study by Hu and colleagues (1) contributes to the literature of the harm–benefit assessment of vertical integration in the US health-care system. Findings from this study represent longitudinal, population-based data from 21 regions covering 34.6% of the US population. These findings help policy makers understand the unintended consequences of vertical integration in the oncology market as well as potential trade-off between efforts in cost containment and quality improvement for older patients with cancer. Although economic theory suggested that vertical integration could potentially reduce medical costs, the empirical evidence from this study does not point in that direction. Notably, empirical evidence of the impact of vertical integration in other diseases has been mixed (8). A systematic review examining vertical integration across disease areas reported improvement in care quality for diabetes through better care coordination but found limited evidence supporting efficiency gains quantified by avoidable utilization and lower health-care spending (8). Of greater concern is the lack of improvement in health outcomes among patients with mCRPC documented by Hu and colleagues (1). Indeed, this pattern echoed the concerns expressed by policy makers that vertical integration could aggravate an already concentrated health-care market and cultivate anticompetitive behaviors of providers without benefiting patients or the health-care system (8).

In addition to the outcomes examined in the study by Hu and colleagues (1), it is important to further explore access to care, disparity implications, patient satisfaction, and financial hardship related to cancer treatment to better understand the impact of oncologist-hospital integration on patients’ welfare. Findings that a higher proportion of patients in the “change” group resided in non-metro areas and areas with lower socioeconomic status and were treated by oncologists with lower service volume warrant more discussions. This suggests that vertical integration may be a business survival strategy for community oncologists practicing in rural and less affluent communities. For patients residing in these neighborhoods, oncologist–hospital integration could provide better access to facilities with newer technologies and a team of more experienced experts—this will be a welfare improvement. However, if oncologist–hospital integration requires travelling long distances to receive treatment and paying higher OOP payments for chemotherapy administered in a hospital outpatient setting, the integration would likely expand the access to care for high-income well-insured patients while creating financial barriers for lower-income patients with less generous insurance coverage, thus widening the disparities between these 2 groups. Although Hu and colleagues (1) found that “on average” there was no statistically significant difference in time on systemic therapy or survival, it is worth exploring whether the same conclusion holds in subgroup analyses stratified by rurality or neighborhood socioeconomic status. In addition, their study cohort consisted of patients with mCRPC who initiated systemic therapy; we are curious to know whether vertical integration may affect the likelihood of treatment initiation or time from diagnosis to treatment. The impact of vertical integration on patient satisfaction is another topic that deserves more attention in future research. As recognized by the systematic review article mentioned earlier (8), few studies have incorporated patient-centered outcomes, such as patient satisfaction, in their examination of the effect of vertical integration.

In terms of the implication on patient financial hardship, although it is reassuring that the study by Hu and colleagues (1) reported no statistically significant changes in total per-patient Medicare spending, we cautioned unlike private insurers that negotiate the amount of payments with providers for services rendered to patients with private insurance, payments to providers for Medicare beneficiaries were set by Medicare program. Therefore, Medicare spending is likely less sensitive to vertical integration. Although prior research has examined the impact of vertical integration on the private insurance market (4), it is critically important for future research to explore how oncologist–hospital integration affects cancer care costs because 40%-50% of patients with cancer reported experiencing financial hardship (or financial toxicity) (9), which disproportionally inflicts younger, working-age patients (10). Furthermore, the proportion of workers enrolled in high-deductible plans increased from 4% in 2006 to 30% in 2019, placing greater financial burden on patients with employer-sponsored insurance (11). Indeed, a recent analysis of claims data from private insurance estimated an 15% increase (after inflation adjustment) in OOP payments from 2009 to 2016 for patients diagnosed with breast, colorectal, lung, and prostate cancer, with deductibles accounting for 30% of OOP payment in 2009 that rose to 40% in 2016 (12).

More research like the study by Hu and colleagues (1) is needed to generate high-quality empirical evidence to guide more targeted policy interventions that maximize the potential benefits of vertical integration while minimizing its anticompetitive effects. To better understand oncologists’ decision to integrate with hospitals and the extent of integration, it is imperative to collect information on physicians’ training and vocational experience and their attitude and preferences regrading financial compensation and autonomy. In addition, characteristics of hospitals (eg, “flagship” hospital, hospital size, ownership) can influence the integration decision, shape the trend of vertical integration in the local market, and ultimately affect patient welfare. It is also critical to examine the role of primary care physicians in the midst of oncologist–hospital integration because they can improve the overall health outcomes of patients with cancer through better management of concurrent chronic conditions and care coordination.

The trend of vertical integration is likely to continue in the US health-care system, making it ever more important to understand how it affects the welfare of patients with cancer. The study by Hu and colleagues (1) enhanced our knowledge on the impact of oncologist–hospital vertical integration among patients with mCRPC. Their study lays the groundwork for future oncology health services research to assess the potential benefits and harms of vertical integration for patients who are diagnosed with other cancers and/or received other forms of cancer treatment (eg, surgery or radiation therapy).

Funding

Dr Shih acknowledges funding from the National Cancer Institute (R01CA207216, R01CA225647).

Notes

Role of the funder: The funder had no role in the writing of this editorial or the decision to submit it for publication.

Disclosures: HJT has no disclosures. Y-CTS received consulting fees, travel, and accommodations for serving on a grants review panel for Pfizer Inc and an advisory board for AstraZeneca in 2019. Y-CTS, JNCI Associate Editor and co-author on this editorial, did not play a part in the editorial review of this editorial or decision to publish it.

Author contributions: Hyo Jung Tak, PhD (Writing—original draft, Writing—reviewing and editing); Ya-Chen Tina Shih, PhD (Writing—original draft, Writing—reviewing and editing).

Data availability

No new data were generated or analyzed for this editorial.