Behind the Best Children's Hospitals Rankings

A 65-page hospital survey and the opinions of thousands of specialists went into the scoring.


Just as the objective of the Best Hospitals rankings is to help adults with uniquely challenging medical needs, kids with rare or life-threatening illnesses such as cystic fibrosis or cancer or severe heart defects are the focus of America's Best Children's Hospitals. Case volume, research fellowships, reputation for managing complex illnesses, and the other factors reflected in the children's hospitals rankings usually aren't the kinds of things that loom large in the course of routine pediatric care. To a young patient who is at serious risk, however, they matter. The bar for a child in need of the best care should be set high.

The facilities most likely to provide exceptional pediatric care are freestanding children's hospitals and large multispecialty pediatric services. With the help of the National Association of Children's Hospitals and Related Institutions (NACHRI), we identified a universe of 160 such centers nationwide and ranked the top 30 in 10 specialties: cancer, digestive disorders, heart and heart surgery, neonatal care, neurology and neurosurgery, and respiratory disorders, plus four—diabetes and endocrine disorders, kidney disorders, orthopedics, and urology—that are new this year. We have also introduced an Honor Roll this year, to recognize the few hospitals that earned a spot in all 10 specialty rankings.

[See America's Best Children's Hospitals and search for one near you.]

The data used to compare institutions were gathered in a 65-page survey completed by the hospitals, covering care-related information from surgical death rates and compliance with known weapons against infection to involvement in research activities and the availability of key subspecialists such as pediatric anesthesiologists. It had to be collected directly from the hospitals because there are no large independent sources of data to draw upon akin to the enormous Medicare files we use to assess adult care.

To help us refine the standards for judging hospital performance, we sought advice from nearly 70 medical directors, department chairs, infection specialists, and other experts, grouped into specialty panels. Their recommendations, developed over four months, were reviewed by RTI International and incorporated into the survey. (RTI, a large research and consulting firm, also oversees the Best Hospitals project.) Of the 160 hospitals contacted, 98 responded. The remaining 62 effectively removed themselves from contention.

Hospitals were ranked on scores made up of three major elements: reputation, medical outcomes, and care-related indicators such as patient volume, nurse staffing, and availability of specialized programs. Because of space constraints, only the most important measures are displayed. A full listing and descriptions of all of the measures, as well as an explanation of how they were chosen and the methodology constructed, are available in an extensive methodology report. Here are the basics of the three fundamental elements behind the approach:

Reputation. Pediatric specialists selected at random from a comprehensive national database in a way that weighs all parts of the country equally were surveyed and asked to name up to five hospitals they believe can provide the best care for patients with serious or difficult problems in their specialty. The responses from 1,500 specialists surveyed in 2009, plus responses from 900 others surveyed last year in the original six specialties, counted as 50 percent of the final score in nine specialties and 60 percent of the final score in cancer (see below). Nearly 60 percent of the physicians surveyed responded, a rate considered extraordinarily high for such a survey.

Outcome. Surgical death rate, transplant survival, and bloodstream infections were among the outcomes measures that made up 10 percent of the final score in all specialties other than cancer. Surgery and procedures leading to bloodstream infections are relatively uncommon in cancer patients, and other credible indicators were absent.