Are Nursing Home Preferred Networks Good for Patients' Outcomes? Evidence From the Veterans Health Administration

In the Veterans’ Administration (VA), medical centers contract with community nursing homes to provide care to Veterans. As a purchaser, the VA could pursue a strategy of selecting a high-quality network; alternatively, it could focus resources on oversight by its nursing-home coordinators. The question of whether narrow networks are good for Veterans’ outcomes, conditional on quality, therefore, needs empirical investigation. We examined the effect of network concentration on hospital admissions, conditional on Veterans’ clinical acuity. We operationalized network concentration as the number of Veterans already in residence at the time of admission, and controlled for publicly reported quality measure (star rating). We identified 93,805 VA-paid admissions to nursing homes between 2013 to 2016. To address selectin bias, we estimated effects using a distance- based instrumental variable (IV) for each measure, with the log of distance to the nearest nursing home with a specified number of Veterans at the facility in the previous month (1-4, 5-9, and 10-13, and 14+ Veterans). Going to a facility with 10-13 or 14+ Veterans had a higher hospitalization probability (6.2 and 3.3 percentage points higher, respectively), than going to a facility with 1-4 Veterans. If quality rating improves outcomes, then broader networks are beneficial if consumers (Veterans) choose based on quality, given a broader choice set. Conditional on quality, concentrated networks do not seem to lead to fewer hospital admissions. Our results suggest that the VA could do more in its oversight role to work with these nursing homes to decrease hospital admissions.

Policy Strategies for the “New Normal” in Healthcare to Ensure Access to High-Quality Cancer Care

In recent years, oncology has seen a rapid increase in the introduction of high-cost innovative therapies while scrutiny around drug pricing has simultaneously amplified. Significant policy shifts impacting health coverage and benefit design are also being implemented, including narrow network health plans, uncertainty around the Affordable Care Act insurance exchanges, and threats to preexisting condition protections. Shifting health coverage policy combined with high drug prices and outdated reimbursement systems may create barriers to patient access to innovation and high-quality cancer care. To understand how trends in health policy are impacting the oncology ecosystem, NCCN convened the NCCN Policy Summit: Policy Strategies for the “New Normal” in Healthcare to Ensure Access to High-Quality Cancer Care on June 25, 2018. The summit included discussion of how innovation is changing cancer treatment, care delivery, and ways health systems are responding; the impact of narrow networks on access to academic cancer centers; and how the evolving health policy landscape is affecting access to high-quality cancer care for patients.

Relation Between Narrow Networks and Providers of Cancer Care

Purpose Health insurers offer plans covering a narrow subset of providers in an attempt to lower premiums and compete for consumers. However, narrow networks may limit access to high-quality providers, particularly those caring for patients with cancer. Methods We examined provider networks offered on the 2014 individual health insurance exchanges, assessing oncologist supply and network participation in areas that do and do not contain one of 69 National Cancer Institute (NCI)–Designated Cancer Centers. We characterized a network’s inclusion of oncologists affiliated with NCI-Designated Cancer Centers relative to oncologists excluded from the network within the same region and assessed the relationship between this relative inclusion and each network’s breadth. We repeated these analyses among networks offered in the same regions as the subset of 27 NCI-Designated Cancer Centers identified as National Comprehensive Cancer Network (NCCN) Cancer Centers. Results In regions containing NCI-Designated Cancer Centers, there were 13.7 oncologists per 100,000 residents and 4.9 (standard deviation [SD], 2.8) networks covering a mean of 39.4% (SD, 26.2%) of those oncologists, compared with 8.8 oncologists per 100,000 residents and 3.2 (SD, 2.1) networks covering on average 49.9% (SD, 26.8%) of the area’s oncologists ( P < .001 for all comparisons). There was a strongly significant correlation ( r = 0.4; P < .001) between a network’s breadth and its relative inclusion of oncologists associated with NCI-Designated Cancer Centers; this relationship held when considering only affiliation with NCCN Cancer Centers. Conclusion Narrower provider networks are more likely to exclude oncologists affiliated with NCI-Designated or NCCN Cancer Centers. Health insurers, state regulators, and federal lawmakers should offer ways for consumers to learn whether providers of cancer care with particular affiliations are in or out of narrow provider networks.