The CIGNA-Healthspring deal: local share key to top-line synergies but still missing from the equation

By

Summary

  1. Provider discounts are a key priority for national accounts – which puts CIGNA (CI) and Aetna at a disadvantage; CI responding in part by trying to get closer to providers
  2. A provider collaboration strategy requires a critical mass of patients and provider mindshare. CI does not have it; nor will the Healthspring (HS) acquisition provide it given the limited geographic overlap between the two companies
  3. CI must therefore grow share in key markets to capture the deal’s potential provider collaboration synergies (though other synergies are certainly accessible)
  4. If CI relies on organic growth going forward, it may have to wait 5-7 years before matching today’s Medicare platforms of Wellpoint or United. Hard to imagine CI’s national provider collaboration strategy can wait that long
  5. Yet more acquisitions will be required to get the requisite local share; and bigger the faster: One more reason to consider a CI-Aetna combination.
* * *

In 2011, the weak economy appears to have increasingly driven many national employers to focus on short-term savings from superior provider discounts vs. longer-term servings of sophisticated payer-orchestrated health management. As a result, Wellpoint and United have gained, while smaller Aetna and CI have been losing at the national account level (Note: CI continues making progress with ASO for smaller groups leading to a net gain in ASO so far). It is a frequent theme in the quarterly earnings calls with the leaders celebrating their strategic deep share position, and the second-tier talking about retooling and retrenching. Given the continued body blows to the economy, “deep provider discounts as table stakes” is likely to be a long-lasting feature of the market place.


Among CI’s responses is a shift towards working much more closely and collaboratively with providers to reduce medical costs (embedded in its “Go Deep” strategy). This means working with a subset of physicians (typically primary care providers or PCPs) to improve how their members are managed (and therefore how much they cost) and changing how providers are reimbursed to incentive the effort. Broadly, the approach should displace a zero sum arms race of payer vs. provider negotiating leverage with an upstream care vs. downstream care dynamic (better care delivered upstream in the primary care setting results in less cost downstream in the specialist setting). More provider friendly (as least to some providers!)

CI calls its version of the strategy “Collaborative Accountable Care” — adding “collaborative” because the insurer plays a key role in making the model work:

  • CI embeds a nurse care coordinator into the participating providers (who uses CI data streams and tools such as predictive modeling, engages in population management activities such as outreach and arranging for gaps in care to be addressed, and recruits patients into CI health management programs)
  • CI also provides its health management programs for patients (presumably at no charge to the provider even if the provider may benefit from their impact through shared savings)

I think CI’s CAC strategy is fundamentally constrained by the same issue which limits is negotiating leverage vs. providers: lack of adequate share in local markets. To systematically change how providers deliver care requires mindshare. CI is doubly handicapped: its lives are spread across the country and they are commercial lives with lower volumes and less complex care than Medicare eligibles (less complex = less potential impact for provider-driven coordination = less potential savings to be shared with PCPs). Perhaps for this reason, CI’s CAC model appears to be structured to be pretty easy for providers to implement (CI does a lot of the work through the care coordinator and its telephonic population health capabilities).

The results from CI’s CAC pilots so far seem unspectacular. 8 CAC pilots have been launched since June 2008 (CI also has 6 multi-payer patient centered medical home pilots which – being multi-payer – are less competitively interesting) and cost results for only 2 have been reported: one achieved a 2% lower costs than market (within the statistical margin of error of an ACO per CMS); the other reported 7% lower costs than the market (mediocre vs. the 13% lower costs BSC reported for its ACO on a CalPERS population — though BSC did not specify the basis of comparison; it may have been versus the previous year, therefore benefited from recession-driven utilization declines).

Significantly in my view, this best reported result was with the Cigna Medical Group which CI owns and where it had the largest share of the PCP panel among its pilots (an estimated 10% vs. somewhere between 5% and 9% for the others). (Note: CI wasn’t giving up on the strategy despite the meager results: A March press release said they wanted to have 30 programs implemented or being planned by the end of the year. Since then, they have announced 3 so things haven’t been proceeding quite so quickly). Implication: To make real progress, CI needs a critical mass of patients with the targeted providers. One way to get there quickly is by acquiring Medicare lives – thus, in part, a logic for acquiring HS.

With its 300K Medicare Advantage lives largely concentrated in five states, HS is in a much better mindshare with its providers. A quick analysis suggests that HS has the same or better provider mind share across its entire network in key states as CI has within its targeted pilots – no doubt the actual effective concentration with high volume provides is much higher (like Florida). A few notes on the analysis:

  • I assume an average panel of 2.3K patients per PCP but also multiply the HS lives by an adjustment factor of 3 to reflect their Medicare volume and complexity of care.
  • Some of the geographies still reflect the broader legacy provider networks from the Bravo acquisition (e.g. Pennsylvania)
  • HS networks need to comply with CMS access rules, so I expect the number of PCPs in network is actually inflated well above the actual PCPs treating most of the patients.
  • The regions where HS talks about having the most success are in Florida, Alabama, middle of Tennessee and in the Houston area (so either in areas where the network is quite narrow or in portions of states where the patient concentration can be leveraged).


HS also seems to operate a more disruptive model of working with providers – organizing IPAs, driving towards more accountability and full capitation, etc. These two factors are likely key to HS’s success in engaging providers to manage medical costs (leading to a lot of gushing from analysts).

Ideally, HS and CI would have major overlaps in geography so that the HS model can be implemented in its network. Geographic overlap is also important for some of the other synergies CI cited for the merger (for example, transitioning Medicare age-ins in CI groups to HS will produce volume in states where CI has lives but only be differentiated in states where HS has the geographic depth to push its model).

However, the geographic overlaps are relatively small:

  • In FL, CI has ~9% of the lives. But HS’s care is very concentrated in the Leon Medical Centers (LMC) and there may not be either many CI lives which go there or HS lives which see providers other than LMC
  • In TN, CI has about 7+% of the lives, so there may be some opportunity to push the HS model out from the middle TN to the east and West. For example, CI’s CAC pilot with Holston Medical Group located in eastern part of the state may be a good platform for expansion.
  • Otherwise, the overlaps are minimal (perhaps some in TX).

If CI is going to really roll the HS model out across its service areas, it will need to increase its share in more key local markets. And given that these kinds of provider arrangements take time to negotiate, time to pilot and demonstrate effectiveness before they can be built into and fully priced into insurance products, it will need to get that critical mass quickly.

It is hard to imagine that organic growth can deliver the critical mass to CI’s Medicare business fast enough. Consider this back of the envelope:

  • Assume HS can keep a 17%+ CAGR in membership which it achieved between 2008/9 (before the Bravo acquisition boosted 2010 membership)
  • Assume about 2.5% of CI’s 11M commercial lives age into Medicare each year and that CI can convert 50% of them to a HS Medicare Advantage plan, yielding ~140K more lives each year. 50% is probably generous but HS’ depth in Florida and CI’s presence in Arizona through its Cigna Medical Group will certainly help with snowbirds.

Under these assumptions, it will take CI 5 years to grow the Medicare business to the current levels of Wellpoint (1.4M lives) and just under 8 years to reach current levels of United (2.2M lives).

A much faster route would be a tie up with Aetna. Aetna has 400K Medicare Advantage lives to contribute, along with 16M+ commercial lives whose age-ins might find the option of Medicare plan powered by HS provider collaboration incrementally more enticing. Adding Aetna’ current 400K lives and share of age-ins to the growth formula would allow CI’s Medicare business to match Wellpoint in less than 2 years and match United’s membership in 3 years. There has been speculation about a CI-Aetna tie up for some time. The HS acquisition (and the implied strategy) to my mind strengthens this logic, does not weaken it.

Given at least 2 of the five synergies of the HS acquisition identified by CI require greater share in key regions (and 2 of the 3 top-line growth synergies – the other being about retail capability in preparation for post 2014 Exchange environment), it is hard to believe that CI will leave this project unfinished or on the slow track. Major inorganic moves must be on the agenda!

  • This CI analysis is spot on in many aspects. There’s no question that CI is very late in the game of building/growing market-share presence in retiree space. According to Citi, CI paid about $500 per customer for HS PDP members, and about $10k per member for its 340K MA members ,a very big premium given recent years acquisitions which have normally ranged from $3k -5k per member!

    When you look at longer term membership scenarios and HCR implications the commercial business is heading in wrong direction quickly especially “down-market” which is where the majority of their book of business resides. In addition, the private exchange marketplace will be heavily influenced by unit cost position which to your findings, CI is disadvantaged (by significant % in many markets) in half the markets they compete in. There’s no conceivable strategy in closing these gaps with care management and/or provider model innovations.

    Another significant challenge that remains a heavy burden is that CI SG&A expense is highest among the national health plan vendors. They have a big gap across the functional areas with claim and call being highest compared to competitors, last I knew this was somewhere around 6-7% higher!

    Bottom line is that neither CI nor Aetna will survive post-reform, they’re doomed to continued shrinkage with higher fixed costs per covered member because you can’t cut enough costs nor quick enough in order to maintain margins in a zero-sum game. Merger or acquisition is only valid option with the two 800 pound gorillas sitting next to them!

  • nice post.