Best Hospitals for Common Care
U.S. News & World Report Health

Post

http://health.usnews.com/best-hospitals/rankings

History

Best Hospitals for Common Care first published in May 2015. The purpose of report is to rate overall quality of inpatient care for procedures and conditions that many hospitals commonly encounter. Currently, U.S. News has identified 19 procedures and conditions as candidates for evaluating hospital performance. First five sets of ratings in isolated heart bypass surgery, hip replacement, knee replacement, heart failure (HF) and Chronic Obstructive Pulmonary Disease (COPD) has been published; work on the remaining 14 and on other procedures and conditions is proceeding. Unless otherwise noted, the metrics discussed refer only to the five ratings in the initial release.

Please note that definitions for these conditions and procedures were created for this report and are not the same as those used by CMS for its performance indicators.

One of the most significant changes in 12 specialties was that the weight of patient safety was doubled, to 10% of each hospital’s overall score. The weight given to hospital reputation dropped from 32.5% to 27.5%.

Scope of Report

Report’s Intended End Users

Consumers can research and select hospitals and physicians per specialty and per region, state or city.

Understanding the Score

Ranking

U.S News ranks the top hospitals in 16 different specialties from Cancer to Urology. For 12 of the 16 specialties, each hospital receives an overall score from 0 to 100, based on four elements: reputation, patient survival, patient safety and care-related factors. Hospitals were evaluated in at least one of the five procedures or conditions. A hospital may be rated as “High Performing”, “Average” or “Below Average”. The categories for hospital recognition are:

  • Honor Roll: Hospitals with very high scores in at least six specialties
  • High-Performing: Hospitals that are not nationally ranked in a specialty but have scores in the top 25% of the analyzed centers.
  • The Best Regional Hospitals: A hospital that is ranked or high-performing in at least one specialty is further recognized as Top 25% within its state and metro area or similar region.
  • Top 50: The hospitals with the top 50 scores in each specialty.

Rankings in the four remaining specialties, Ophthalmology, Psychiatry, Rehabilitation and Rheumatology, do not depend on hard data rather is based on hospital reputation, determined by a physician survey. Many structural and outcomes measures are not applicable to these specialties because procedures are performed largely on an outpatient basis and pose a very small risk of death. These specialties are referred to as Reputation-Only Rankings.

Note: the highest-ranking hospitals in a specialty are not necessarily the best choices for care. Each specialty ranking evaluates hospitals according to their performance across a wide range of conditions and procedures. One hospital might rank lower than another overall but do much better in treating a particular condition, like chronic obstructive pulmonary disease.

Reporting

The Best Hospital rankings are updated annually in July. Results are posted on US News website. US News is developing “The Best Hospitals Dashboard”, which gives hospitals an opportunity to communicate directly with US News making it easier for hospitals to provide updates and ask questions.

Inclusions

  • All nonfederal community hospitals of any size included in the American Hospital association (AHA) 2012 database.
  • Except as noted, ratings reflect care received only by inpatients age 65 and above who are covered by traditional fee-for-service Medicare. Many hospital patients enrolled in commercial Medicare Advantage plans are also included.
  • A hospital qualified for consideration in the data-driven specialties if it met any of four possible criteria:
    • Member, Council of Teaching Hospitals (COTH)
    • Medical school affiliation (AMA or American Osteopathic Association)
    • At least 200 hospital beds set up and staffed
    • At least 100 beds set up and staffed plus availability of at least four of eight important key technologies
  • Volume of at least 25 patients in the Inpatient LDS SAF data for particular procedure or condition. A hospital with below-minimum volume was considered eligible for a specialty if it had a reputation score of 1% or greater.
  • Specialties are as follow: Cancer, Cardiology & Heart Surgery, Diabetes & Endocrinology, Ear, Nose & Throat, Gastroenterology, Geriatrics, Gynecology, Kidney Disorders, Neurology & Neurosurgery, Ophthalmology, Orthopedics, Psychiatry, Pulmonology, Rehabilitation, Rheumatology, Urology.

Exclusions

  • Federal facilities.
  • VA and military hospitals.
  • Psychiatric, rehabilitative,  substance abuse treatment, and Long Term Care facilities.
  • Results were not reported on groups of fewer than 11 patients to avoid release of information that might result in patient identification.

Data Source

The data used in the analysis derive from government and proprietary sources including:

  • The Centers for Medicare & Medicaid Services (CMS) administrative claims data from the Hospital Compare web site. Inpatient Limited Data Set Standard Analytical Files (Inpatient LDS SAF)
  • The Centers of Disease Control & Prevention (CDC); National Healthcare Safety Network (NHSN)
  • Society of Thoracic Surgeons (STS) clinical registry data
  • American Hospital Association Annual Survey; all U.S. hospitals annually (including AHA non-members)
  • Hospital Consumer Assessment of Healthcare Providers and Systems Survey (HCAHPS). The federal government releases quarterly rolling 12-month assessment of inpatients’ opinions about their stay in various respects. HCAHPS data are aggregated across each hospital. Patients’ opinions about their care in specific departments cannot be determined
  • Physician Survey: Four specialties, Ophthalmology, Psychiatry, Rehabilitation and Rheumatology, do not depend on hard data since most care is delivered on an outpatient basis.  Therefore rate for these specialties is based on hospital reputation, determined by a physician survey.

Methodology

Methodology was developed by Dr. Foster Limited, a London-based health data analytics organization. Hierarchical Logistic Regression model is used to calculate expected number of incidents. To indirectly standardize for differences in patient case mix, the model was adjusted for factors such as: Age at admission, Sex, Year of hospital admission, Elixhauser comorbidities and dementia, Medicare Status code, Subgroup where procedure group was known to be heterogeneous,  Method of admission (via emergency room or transferred to the hospital to hospital), and Socioeconomic Status (SES).

After transforming data into a Z-Score, the degree of over-dispersion in the data was estimated and adjusted. The Bayesian hierarchical model was implemented using the statistical program WinBugs. Using posterior distribution of the rank, a 13- Tier Banding system was created, with 13 being the best and 1 being the worst.

To aggregate indicators to overall rating, an “Unobserved Components Model” was employed. The model weighs indicators according to two principles; correlation with other indicators and degree of variation. Because of weight driven from this model weight of each indicator is different for each condition and procedure. For details please refer to “U.S. News & World Report Best Hospitals for Common Care Methodology” © 2015. The overall hospital score was banded according to the degree of confidence of where a hospital ranks relative to the mean score of zero. Hospitals were placed into one of five performance bands and then collapsed to three- tier “High Performing”, “Average” or “Below Average” for the purpose of public reporting.

Reputation-Only Rankings is by reputation alone and based on the latest three annual physician surveys. Those receiving nominations from an average of at least 5% of the responding specialists and the last three surveys are nationally ranked Best Hospitals. Those nominated by at least 3% but less than 5% of physicians are High-Performing. A hospital that fell short could still make it through if nominated by at least 1% of the physicians in a specialty who responded to the 2012, 2013 and 2014 reputational surveys.

Measures

Improvement Tools

Improvement tools and resources can be found at http://www.hqinstitute.org/tools-resources