Hospital networks: Configurations on the exchanges and their impact on premiums

Articles

Our third intelligence brief on ACA exchange dynamics sets forth five observations based on analysis of new network configurations across 20 urban rating areas.

The configuration of hospital networks is changing on the new public health exchanges and, in many cases, having a direct impact on premium levels. We are observing a phenomenon not dissimilar to that noted in the emerging private defined contribution exchanges: specifically, the “segmentation and valuing” of choice in the form of a proliferation of products with varying breadths of hospital networks. Across the markets we analyzed, there is a greater breadth of network options available on the individual exchanges compared to the 2013 individual market, with nearly every rating area offering consumers products with networks spanning from very narrow to broad. This trend is consistent with what we see in most well-functioning consumer markets ranging from cell phone plans to automobiles – a variety of choices comprising different value propositions at different price points.

The growing prevalence of products utilizing narrow and “ultra”-narrow hospital networks is evident,1 and the motivation is bi-lateral. Hospitals in some cases have removed themselves from network consideration, believing the value of their services exceeds the amount carriers are willing to offer in terms of reimbursement. Carriers are adapting to affordability imperatives by actively excluding some higher cost hospitals while collaborating more closely with those willing to accept lower reimbursement rates. This latter group of hospitals is perhaps motivated by the potential to increase volume in what has been a flat to modest growth environment. These are both legitimate strategic postures and only time will resolve the prudence of either or both.

As defined by the strict network adequacy requirements set out initially by state insurance regulators, and subsequently by the Affordable Care Act (ACA), these narrower networks appear to be compliant in terms of both the number and types of providers, and the distance to the nearest provider. As well, we found no meaningful difference between participating and non-participating hospitals in these exchange networks when we examined the Centers for Medicare and Medicaid Services’ (CMS) performance figures comprising a composite score of 20 quality and patient satisfaction metrics (the value-based purchasing score), and separate figures covering rates of readmissions.

Our analyses show, however, that products comprising narrower hospital networks correlate with a lower premium. Indeed, across the markets we analyzed, the median increase in the premium for the same product type (e.g. HMO, PPO), offered by the same carrier, in the same metal tier, but utilizing a broad versus narrow hospital network is 26 percent. Thus, the trade-off between price and choice of hospital breadth in a network in the exchanges has been established for the consumer. The consumer will now need to evaluate this price/breadth trade-off and choose a product accordingly.

Previous McKinsey research from exchange simulations2 conducted within the last 12 months has revealed consumer demand at many price/network breadth combinations, including some consumers willing to select a narrower network product in return for lower monthly premiums or lower out-of-pocket costs. Empirical evidence of how consumers value this trade-off will emerge shortly, yet it may well take several years, after consumers have had experience with different types of networks and have made subsequent renewal decisions, until we understand any enduring impact.

To inform our observations, the McKinsey Center for U.S. Health System Reform compiled hospital network data from 120 unique 2014 individual exchange market products in the silver tier offered by 80 carriers, and a subset of corresponding hospital network data from 2013 individual market products. Our analyses span 20 urban rating areas across a broad geographic range, and these rating areas include close to one-fourth of the U.S. non-elderly uninsured population. In our analyses, we categorize each network based on its degree of narrowing – broad, narrow, and ultra-narrow – defined by percent participation of the 20 largest local hospitals by bed size in that rating area.

We identified five key observations from our analyses:

  • Narrow and ultra-narrow hospital networks are more prevalent (70 percent of all networks), increasing the variety of network configurations available to consumers
  • Products with broad hospital networks reveal higher premiums, with a median premium increase of 26 percent between broad and narrower networks of the same carrier, product type (e.g., HMO, PPO), metal tier, and rating area. Also, the majority (84 percent) of lowest-price silver products utilizes narrow or ultra-narrow networks
  • Network breadth and product type (e.g., HMO, PPO) are correlated: the majority (76 percent) of ultra-narrow networks is coupled with HMO designs, and the majority of HMOs (58 percent) is coupled with ultra-narrow networks
  • Frequency of narrow networks differs notably by carrier type, as does the carrier’s ability to translate narrower networks into competitively-priced products
  • Academic medical centers are participating predominantly in broader, higher-priced exchange offerings (10 percent higher premium on average)

 

  1. Networks in our sample are categorized as follows: a) broad networks have less than 30 percent of 20 largest hospitals by bed size in the “relevant area” (area within 50 miles of rating area’s most populous zip code) not participating, b) narrow networks have 30-69 percent of 20 largest hospitals not participating, c) ultra-narrow networks have at least 70 percent of 20 largest hospitals not participating; largest 20 hospitals represent ~80-85% of beds on average across rating areas.
  2. Jenny Cordina et al, “Winning on the individual exchanges: Driving high consumer participation,” McKinsey white paper, June 2012.
  3. For each rating area, we identified all on-exchange silver tier networks. When the same carrier offered multiple products based on the same network, the lowest-price product was used. Carrier count represents unique carriers at a state level. See methodology in appendix for further detail.
  4. Cities include: Atlanta, Bridgeport, Chicago, Dallas, Denver, Houston, Indianapolis, Los Angeles, Louisville, Miami, Nashville, Philadelphia, Pittsburgh, Portland (ME), St. Louis, Salt Lake City, San Jose, Seattle, Tampa, Washington DC.
  5. Non-elderly uninsured defined as those over 100 percent FPL in non-Medicaid expansion states and over 138 percent FPL in Medicaid expansion states.