Why Health Care Mergers Can Be Good for Patients

Hospitals and physician groups are merging into large health systems at unprecedented rates, fueled in part by the Affordable Care Act. Provider consolidation takes many forms, but the general trend can make it easier to share electronic records systems, coordinate care of patients, and eliminate redundant costs. Another potential benefit of integrated health systems is in addressing the persistent problem of variation in health care practice and outcomes, particularly in surgery.

Decades of scientific research confirm the obvious: Patients who undergo complicated operations fare worse when their hospitals or surgeons rarely perform them. (In the medical community, we classify these surgeons and hospitals as “low-volume.”) The problem has long been known, and the purchaser coalition called Leapfrog Group has tried to curb it. However, recent analyses confirm that thousands of Americans whose surgeries are performed by low-volume surgeons and hospitals die unnecessarily every year.

The Dartmouth-Hitchcock Health System (where I do surgery), Johns Hopkins, and the University of Michigan have recently taken a “volume pledge” — the first public commitment to manage the low-volume surgery problem. The three institutions have identified 10 complex operations (mainly cancer, cardiovascular, and orthopedic procedures) for which scientific evidence shows that surgical volume matters, and they have set minimum standards for all surgeons and hospitals in their systems. The volume bars, established by consensus among surgeons in each specialty, are set high enough to reduce risk meaningfully for patients, but not so high as to be impractical or impede access to care. Reasonable exceptions are made for surgeons just out of training, surgeons coming back from sabbatical or medical leave, and emergency clinical situations.

Dartmouth-Hitchcock, as part of its One DH effort, is also working to implement standardized care protocols across its affiliate network for common conditions and procedures — based on evidence where possible, clinician consensus where not. For example, for total knee replacement, a patient will soon be able to expect not only a high-volume surgeon but also a consistent clinical experience, regardless of which DH practice or affiliated hospital provides the care. That experience will include shared decision making, to ensure that both scientific evidence and patient preferences influence decisions about surgery. Standardized perioperative care pathways will also guide consistent, evidence-based prophylaxis against infection, blood clots, and other complications. Orthopedic surgeons across the system will even agree to use the same types of joint prostheses — a big stride in collaboration within this specialty.

Geisenger’s ProvenCare model for cardiac surgery is another example of how standardization in an integrated health system can improve quality and reduce costs within hospitals. Regional health systems that comprise many affiliated hospitals have an opportunity to apply such a model on a much larger scale and in heterogeneous care environments, where limiting variation in quality is most important.

How to move forward

Even within most committed, mission-driven health systems, standardizing care and reducing variation in quality is difficult. Given our early experience, I believe that two cultural shifts are key:

1. Physician leaders must be willing to push through unpopular changes. When we announced the surgical volume pledge at Dartmouth-Hitchcock, most physicians, including surgeons, were strongly supportive. But a small but vocal subgroup accused us of “Big Brother” tactics, asking questions like “Who are you to tell me what procedures I can and can’t do?” We offered to help low-volume surgeons better their skills in doing specific procedures, differentiate their practices, and clear the new volume bars. We would not bend the policy itself, however, and told a couple of recalcitrant surgeons, “Dartmouth-Hitchcock may not be the best fit for you.” That kind of firmness is essential.
2. Organizational structure must also be changed. Like many regional health systems, Dartmouth-Hitchcock has grown incrementally, one community group practice and one affiliated hospital at a time. Until recently, clinical practice issues were left entirely to local leaders and governance. Our move this year to a specialty-specific “service line” model is greatly accelerating our efforts to reduce variation in care. In that structure, one physician in each specialty has authority and accountability for managing care across all practice sites in the system. For the musculoskeletal disease service line, for example, one orthopedist is responsible for “clinical programming” (which procedures are done where and by whom), for optimizing the size of the clinical workforce to match population demands, and for aligning physician compensation with clinical performance. That same service-line leader is also charged with implementing consistent standards and care pathways for the most common orthopedic conditions and procedures.

Clearing the hurdles

Despite early progress in reducing variation, practical obstacles remain. Although Dartmouth-Hitchcock can establish clinical policies centrally as a health system, a lot of the “teeth” in implementation and enforcement resides within individual hospitals. For practical reasons, we have elected not to undertake the laborious process of changing credentialing standards and bylaws at each hospital. Instead, we are implementing the volume requirements as a clinical safety standard, just as we require checklists to be completed before any surgery. Any breach in the surgery-volume standard would be addressed by detailed review and remediation, in line with expectations for staff accountability established by the Joint Commission.

Standardizing care also often requires persuasion and peer pressure rather than brute enforcement. For example, although we have little direct leverage over the subset of surgeons at our affiliate hospitals who are not employed by Dartmouth-Hitchcock, we can discourage our primary care physicians from referring patients to any surgeon who does not adhere to safe practices, including our volume requirements.

Rapidly expanding health systems, like ours, are learning as they go and will inevitably take different paths toward reducing variation in practice. Regardless of how they get there, patients will benefit.