Size matters

A study that appeared last week in the Journal of the American Medical Association documented a decade-long rise in 30-day mortality rates at critical-access hospitals, most of which are rural with fewer than 25 beds. It no doubt set off alarm bells in the C-suites at those institutions. The fear is that the government will soon start financially penalizing poorly performing CAHs in the same way it now penalizes larger hospitals, which showed slightly declining mortality rates in the same study.

The study was disturbing because the differences appear to be mounting. A decade ago, CAHs were roughly comparable in outcomes to larger facilities when treating heart attack, congestive heart failure and pneumonia patients. By 2010, the combined 30-day mortality rates at CAHs were nearly two percentage points higher: 13.3% versus 11.4%.

Policy analysts have long been aware of the relationship between patient volume and outcomes. It is at work in the wider world, too, not just at CAHs.

If one looks only at hospitals in the prospective payment system (CAHs are reimbursed based on costs), there are similar gaps between larger and smaller hospitals. A June 2012 special report by the Medicare Payment Advisory Commission, which reviewed heart failure and pneumonia mortality rates at hospitals of different sizes, found those with more than 8,000 discharges a year had significantly lower mortality rates compared to those with 1,000 to 2,000 discharges a year, whether they were in urban or rural settings.

That same report revealed the size dynamic was at work even within the CAH universe—at least for those sending data to the CMS (not all do because they're not required to). Hospitals with more than 10 admitting physicians had risk-adjusted mortality rates that were one to two percentage points lower than those with less than five admitting physicians.

The MedPAC report offered what appear to be obvious reasons for the difference: Physicians practicing at CAHs have fewer colleagues to talk to; less experience because they see fewer patients; and less practice on difficult cases.

For at least two decades, some experts have called for merging competing CAHs or those that serve adjacent territories into single facilities. But policymakers have repeatedly rejected that approach, arguing making people drive 20 miles or more to reach a hospital could jeopardize lives.

More to the point, it's politically a nonstarter. No legislator willingly votes for hospital closings. Indeed, Washington representatives and senators from rural states have had remarkable success in funneling special subsidies to CAHs, which now total $4 billion a year.

The quick-fix solution would be to impose incentive models that push CAHs to adhere to best practices, such as the value-based purchasing program now being used by the CMS with larger hospitals. However, the same MedPAC report found that frontier hospitals (those most likely to be CAHs) showed “significantly worse performance” on some process measures, but better or comparable on others when compared to larger hospitals. It's not clear what works in improving outcomes.

The National Rural Health Association, which represents small hospitals and other rural healthcare providers, has developed a set of process measures with what it calls “rural relevance.” How well do hospitals conform to emergency transfer protocols for complex patients, for instance. Other core measures, such as patient satisfaction surveys, should be modified to take into account the smaller number of patients seen in rural hospitals, according to the NRHA plan.

The Patient Protection and Affordable Care Act took note of the special circumstances facing CAHs by authorizing a demonstration project aimed at determining which processes led to better outcomes. Unfortunately, the agency missed the deadline for issuing the regulations, which were due in March 2012.

In any case, a demonstration program would be only a start because it wouldn't be large enough to truly determine what works in improving outcomes in small hospitals. That point was made in another article in JAMA by Dr. John P.A. Ioannidis of Stanford University School of Medicine, who has a well-deserved reputation for being one of the more insightful critics of the medical literature. “Trying to impose quality data collection and reporting in such hospitals that have already strained resources may actually do more harm than good,” he warned.

Ioannidis called for a yearlong randomized trial of 20 CAHs that would test a range of quality improvement initiatives and changed payment practices. That way, the CMS could scientifically determine which, if any, actually improved outcomes. It's a good idea. It would be nice to have some evidence before the agency imposed financial penalties on CAHs.

Merrill Goozner, Editor