Evidence from medical home (PCMH) pilots suggests that reductions are on the way for specialty and hospital services, and that integrated health systems should look to other avenues to make PCMHs financially sustainable.
Bottom line up front: Volume reductions likely
We received several questions recently from health system leaders asking whether there might be a volume upside for specialist and hospital services affiliated with PCMHs. This is a good question because certain aspects of the PCMH -- for example, changed referral patterns or increased patient screening activity -- might seem to translate into increased downstream volume.
But despite a couple of documented cases to the contrary, most evidence from pilot sites points toward reduced downstream utilization (which makes sense under the theory of accountable care).
Future declines would be driven by two main factors:
- Services being rendered unnecessary as patients are kept healthier through compliance with primary care EBM
- Some traditional specialist visit types shifting to PCPs
One silver lining from a specialty/hospital business perspective would be a higher-complexity case mix for specialists, since the less complex visits would be the ones shifting to PCPs. Also, to a certain extent, PCMHs could be leveraged to capture incremental market share that would backfill new cases into the system.
Anecdotal evidence of PCMH-linked upticks in downstream utilization
Group Health's PCMH reported sites about 5% more specialist encounters than control sites, but the differences between sites appeared to decline over time (Source: Reid, R. "The Group Health Medical Home At Year Two: Cost Savings, Higher Patient Satisfaction, And Less Burnout For Providers", Health Affairs, May 2010)
In the BCBS of ND/Meritcare medical home pilot:
"There was a statistically insignificant increase in the frequency if IP admissions ... driven by elective admissions for hip and knee replacements, which demonstrated an increase of 40% from pre-study period to (15%) the study period (21%) for all IP admissions (p=.582)..... We postulate that this occurred because of improved chronic disease management strategies for patients with multiple chronic conditions, such that the chronic conditions were no longer "quality of life limiting" and that physicians and patients could now manage other "quality of life" medical problems such as severe degenerative disease of the hip and knee by referring patients for elective arthroplasties" (Source: Hanekom, D. "Blue Cross Blue Shield of ND: The Advanced Medical Home: Results & Lessons Learned A Four-Year Longitudinal Study," April, 2009)
Evidence that PCMHs tend to reduce downstream utilization overall
Overall, evidence from PCMH pilots points to success in reducing hospitalizations, specialty services, and total population spending.
Anecdotal examples of PCMH-linked downstream volume declines include:
- BCBS of MI PCMH program: In PCMHs vs non-PCMH practices, 2% lower rate of adult radiology usage; 1.4% lower rate of adult ER visits; 2.6% lower rate of adult IP admissions; 2.2% lower rate of pediatric ER visits (Newsroom, "Michigan Blues' Patient-Centered Medical Home model succeeding," Newsroom Blues Blogs Multimedia. Jun. 4, 2010)
- BCBS of SC and Palmetto Primary Care Physicians: Compared to previous year: 10.4% fewer IP hospital days than the previous year; 12.4% fewer ER visits than the previous year. Compared to diabetes patients treated by non-PCMH local practices: 10.7% fewer hospital admissions among medical home patients; 36.6% fewer IP hospitals days; 32.3% fewer ER visits (Arvantes, James, "South Carolina Insurer Embraces PCMH Model of Care," AAFP News Now. Nov. 9, 2010.
- Group Health Cooperative's medical home pilot: The PCMH prototype clinic experienced 6% fewer hospitalizations (in comparison to the 19 other area Group Health clinics) after 2 years (Reid, Robert J. et al, "The Group Health Medical Home at Year Two: Cost Savings,
Higher Patient Satisfaction, And Less Burnout For Providers," Health Affairs 29, No. 5 (2010)
Intermountain Healthcare's Management Plus PCMH model experienced a 10% reduction in total hospitalizations (Grumbach, Kevin, "The Outcomes of Implementing Patient-Centered Medical Home Interventions: A Review of the Evidence on Quality, Access and Costs from Recent Prospective Evaluation Studies," PCPCC. Aug. 2009)
- AtlantiCare's pilot medical homes focusing on high utilization/high cost patients: surgical procedures reduced 25% , hospital admissions and ED visits reduced "by more than 40%" (Gawande, A. "The Hot Spotters: Can we lower medical costs by giving the neediest patients better care?", The New Yorker, January 24, 2011)
Very little granular detail is available about exactly which downstream services are most likely to register volume changes, or to what degree. For hospitalizations, a reasonable starting point would be those DRGs that AHRQ has identified as "ambulatory-sensitive", i.e., preventable if primary care is delivered to evidence-based standards, such as admissions for complications of diabetes, CHF, etc. (See the AHRQ overview of quality indicators).
The Advisory Board has a customized-assessment tool that forecasts magnitude of downstream ambulatory-sensitive volume changes for a specific hospital or health system. The Preventable Admissions Identification Assessment is part of the Health Care Advisory Board's suite of tools--i.e., freely accessible to all participants in the Medical Home Project.
Impact of PCMHs on Specialty Services
Just as transformation to a PCMH model will have a difficult-to-forecast effect on site-level primary care volumes (with some visit types increasing, some decreasing as non-face-to-face modes are introduced), specialists who work with PCMHs will likely see a change in volumes, patient interaction types and/or case mix.
Greater coordination between PCMH practices and specialists, including more phone consults and greater information exchanged, could even replace certain types of specialist visits. Ongoing management for certain conditions could migrate out of specialty offices altogether as medical homes become more focused on treating chronic conditions. This would mean PCMH-affiliated specialists should experience reduced utilization, but potentially more complex case mix, because the lower-complexity care management activities would be the ones making the shift:
"Specialists [would be] freed up from routine care of chronically ill patients (specifically chronic obstructive pulmonary disease/asthma, low back pain, diabetes mellitus, coronary artery disease/congestive heart failure, chronic kidney disease, and depression). Assuming that half the direct/indirect care for these patients were offloaded to PCPs, relevant specialists would free up a considerable amount of time for each specialist practice, potentially resulting in higher revenue per remaining patient/unit of service" (Hollingsworth, J et al. "Specialty Care and the Patient-Centered Medical Home," Medical Care, November, 2010)
To keep these specialist practices running at capacity, and backfill with additional, higher-complexity cases, new patients would have to be brought into the system.
Why wouldn't increased screening activity result in greater downstream advanced testing/procedural volumes?
Despite tremendous, long-standing industry interest in any data linking increased screening to gains in downstream volume for advanced testing, procedures, or admissions, our experts' consensus is that there is no such reliable analysis. The lack of data stems in part from the difficulty in tracking patterns of usage across the continuum of care (though it might be possible to conduct an analysis like this within an integrated delivery system). More importantly, we think that the reason no such data exists is because, assuming a constant patient base, greater primary screening does not reliably translate into greater downstream procedural volumes.
The reasoning here is that greater screening of existing patients does not create new cases as much as it catches cases earlier. "Catching cases earlier" has a basically impossible-to-quantify total impact on downstream testing/procedural services per patient, which will differ by service line and condition and will include not just volume changes, but service mix changes as well.
Therefore, the only clear-cut way to capture incremental volume from screening activity at the primary care level is to ensure that some of these screens are being performed on patients that are new to the system.
How would PCMHs increase market share and bring new patients into the system?
From a system business perspective, since the PCMH model "shrinks the pie," the main opportunity for volume increase would lie in capturing greater market share. A system affiliated with PCMHs might capture new cases in some or all of these ways:
- Taking market share from alternative screening sites: If PCMHs increase their screening activity, this may mean screening some patients who otherwise would have been screened at alternative sites not affiliated with the system.
Improved recruitment/retention of PCPs interested in the medical home model
- Larger panel sizes, including patients new to the system), supported by changes in practice workflow and/or new patients steered by payers to PCMHs/narrow network high performers
- Change in specialist referral protocols: Referral protocols may be more formalized and/or more closely integrated with the system in PCMHs than in typical primary care settings, due to increased focus on care coordination and information-sharing, leading to PCP-specialist service agreements, use of shared IT platforms, etc.
A closer look: specialists could capture incremental business from PCMHs
Extra volume associated with changed referral protocols would hinge on the ability of those specialists to win patient volume by working with PCMHs cooperatively. At a minimum, those practices would need to meet access and communication protocols. As explained in a recent Advisory Board interview with a chief medical officer at a medical home pilot in the northeast:
"The medical home physician in the [PCMH pilot] is doing it "by the book". In other words, he makes phone calls to specialists when referring patients and expects letters in return--which we need for pay-for-performance. If the letter doesn't come, we call the specialist. And, if the specialist is not doing it, we don't use that specialist anymore." (The Advisory Board, "Transforming Primary Care: Building a Sustainable Network for Comprehensive Care Delivery", 2011)
Beyond improving coordination with PCMHs, recent interviews with health plans highlight a trend toward emerging payment designs that would further reinforce connections between PCMHs and "preferred" specialty providers on the basis of demonstrated specialist performance against cost and quality standards.
For example, mid-Atlantic insurer CareFirst plans to share specialist cost-effectiveness data (such as which specialists are designated "high performance" providers) with PCP participants in its PCMH program. PCPs would receive incentives based, in part, on reductions in total PMPM spending that the PCP could help inflect by selecting these preferred providers. In addition to steering patients to this narrow network, the CareFirst program envisions PCPs forging tighter bonds with these specialists on the basis of shared agreements about how to treat patients--all of which together would lay the groundwork for bundled payments and/or accountable care contracting (CareFirst. "Patient Centered Primary Care Medical Home Program: Program Description and Guidelines", www.carefirst.com/providers/pcmh/ -- see page 50)
Summary -- and contracting implications for health systems
All told, PCMHs are unlikely to generate substantial downstream volume for specialist or hospital services. In fact, the reverse is likelier. The main PCMH-related avenue for capturing incremental volumes would be to utilize the PCMH model's capacity to grow system market share.
The strong probability that PCMHs will reduce downstream volumes reinforces our general take on PCMH contracting in a system context: The system must participate in joint contracts that include gainsharing/shared savings components in order to protect total-system financial viability in the long term.