Governance Advice for Hospital Boards: Population Health

"For 60 or 90 days of post-discharge care?"

Writing in the April 26 issue of JAMA, Michael Jillinek of Lahey Health has some important advice for healthcare organization boards of directors.

As income shifts from fee-for-service to global payments, the insurance risk transfers that underlie much of "population health" are an important threat to these enterprises' viability.

After a compact and well-written summary of the growth of population health, he offers six suggestions for these boards:

1. Plan on having "forthright discussions" about the difficult tradeoffs between still-remunerative fee-for-service activities (such as high-dollar imaging, lucrative surgical services) and having to invest in the Triple Aim (care coordination personnel, improving quality measures for persons with chronic illness).

The Population Health Blog suspects most boards will ask why they can't have both the FFS cake and the global payment icing. If that's the case, these boards need to plan on having forthright and very lengthy discussions. It's organizationally difficult to have one mission on the 4th floor of the hospital and another in the emergency room.

2. If the organization's employees are enrolled in a "self-insured" health plan, bring them into a population health program sooner rather than later.

Not only is this an important opportunity for a board to understand the revenue versus savings versus expenses involved in driving the clinical and care experience outcomes of population health, its only right to take this for a personal test drive before subjecting your patients to it.

3. Look for common ground between old fee for service and new global payment arrangements.  The author suggests reducing readmissions is a good start.

The PHB suggests boards ask their management teams to also pursue the care coordination "chronic care management" payments offered by CMS.

4. Start demanding population health metrics from your management team, "such as details of total medical expenditures."

More details on the work of measurement can be found here.  The PHB has also humbly suggests here that health organizations should be prepared to invest significant resources - and discipline - into the process.

5. Invest in primary care, care coordination teamwork and pursue "population health pilot programs."

Since the PHB believes well-intentioned CMS' programs are star-crossed (see here and here), it suggests working with local commercial insurers for starters.  As it reviews resources like this, they seem to have a better track record. 

6.  Ask your management team to be open population health contracting.

Hear hear, says the PHB.  But it also cautions that the board needs to have individuals with the kind of industry knowledge necessary to provide oversight of these contracts.  

Image from Wikipedia

A Primer on Population Health

If you need to help a colleague who is clueless about cost-saving potential  of population health, the Population Health Blog recommends you refer him or her to this efficiently written primer from The Health Care Transformation Task Force's High Cost Patient Work Group.

Basically, if the small number of patients who are destined to be high cost can be prospectively a) identified and b) helped, a lot of money could be saved.  By focusing resources on a small fraction of insured people, the costs for the entire risk pool can be decreased.

In the meantime the majority-remainder of the population can be aided with other lower-cost resources that include, but are not limited to wellness and prevention.

Who are the small numbers of people?

1) patients with a) advanced illness who are nearing end of life who may benefit from b) hospice.

2) patients with a) high spending patterns who may benefit from b) coordination of care by a dedicated health provider (such as a nurse or a lay community care worker).

This population does not include:

3) high cost patients with a) any illness who b) are destined to get better all by themselves or c) unlikely to derive any benefit.

Algorithms to spot that small number of patients use diagnosis codes, treatment codes and medication utilization data from the electronic record or insurance claims databases.  Other useful insights can be gained from patient surveys (and a number are available), using public data to ascertain socioeconomic status (zip codes are destiny), asking physicians to refer patients who are at risk and the design of insurance benefit (patients may not be aware that certain services are covered).

The science behind the use of these inputs is imperfect but getting better, and the more inputs, the better.  Don't let the perfect be the enemy of the good, however, because simple algorithms based on readily available data will get you started.

One of the advantages of this approach is that the cost of coordination of care is variable.  It can start small and be flexed up as expertise grows and opportunities arise

The Population Health Blog couldn't have said it better.