About 440,000 people die in the United States each year as a result of a preventable hospital error. Despite the precautions hospitals take to minimize the likelihood of a mistake, many are still harming patients through avoidable missteps.

Several measures can help in evaluating how well hospitals are run in a given area, including readmission rates of patients after they are discharged and patient mortality rates. Another indicator is unnecessary hospitalizations of patients who can be treated by ambulatory care. The hospital safety score from the Leapfrog Group also measures quality of care and is based on procedures hospitals put in place to prevent errors. Using a combination of these measures, 24/7 Wall St. reviewed hospitals in the country’s large metropolitan areas to determine the 10 cities where you don’t want to get sick.

According to Bruce Boissonnault, CEO of the Niagara Health Quality Coalition (NHQC), there is no “ultimate definition of quality.” However, there is no denying that health care results vary considerably from hospital to hospital. The Binghamton, New York metro area is the worst place to get sick in the country, while the Ames, Iowa hospital system reported some of the best outcomes.

One primary factor behind the different outcomes among hospitals is human error. The best hospitals have systems in place to “ensure that, at every decision point along the continuum of care for each patient, [doctors and nurses are doing] what leads to the best results,” Boissonnault said.

Click here to see the 10 cities where you don’t want to get sick. 

Different factors contribute to different outcomes. A high readmission rate for example, is often the symptom of an inefficient system. Boissonnault gave the Catholic Health system in Buffalo as an example where the opposite is the case. Hospitals in the area have a readmission rate of 15%, better than the national readmission rate. For Boissonnault, this is likely because area hospitals require doctors and specialists to have “a data system that allows data interchange,” so that an area hospital “can easily see the electronic medical record” of a referred patient. In Buffalo, the system increases communication and decreases the likelihood of error.

Inefficient hospital systems, on the other hand, can lead to mistakes, which can increase readmission rates. In each of the cities with the worst hospitals, the readmission rate was higher than the national average of 15.2%.

Another important factor is efficiency of care — the efficient use of medical care resources on those patients that require them. “All things being equal, [hospital systems] that do more actually have worse outcomes,” Boissonnault said. While it may be counter intuitive to some, overuse of medical care can be harmful, “because you’re exposing a patient to radiation that they don’t need, or some procedure that has a risk that they don’t need.” Even if unnecessary testing or treatment does not directly harm the patient, the resources used on that unneeded care “could have been spent on effective care for another patient.”

Hospitalization in cases where the patient could have been treated with an outpatient service is much more likely in areas where care is administered inefficiently. In six of the 10 cities with the worst hospitals, the number of unnecessary hospitalizations was greater than the national average.

Ultimately, however, people’s hospital choice may not be determined only by the quality of care at that facility, but also by other, more subjective, factors. “If your grandmother is in the last 11 days of her life, you might decide that the most important quality factor is which care provider is closest to your home so you can see her as many times as possible before she dies,” Boissonnault said.

Click here to see our full methodology.

These are the 10 cities where you don’t want to get sick.

10. Jonesboro, AR
> Readmission rate: 16.0%
> Avg. 30 day mortality rate: 13.5%
> Hospital safety score grade: C-
> Preventable hospitalizations: 71.5 per 1,000

In the Jonesboro metro area, 16.0% of hospital patients needed to be readmitted within 30 days of their initial discharge, slightly higher than the national patient readmittance rate of 15.2%. The average mortality rate within 30 days after admission was 13.5%, 2 percentage points higher than the corresponding national rate of 11.5%.

Located near the center of the city, Saint Bernard’s Medical Center is the facility of choice for many Jonesboro residents. The hospital, which claims to specialize in cancer treatment, received abysmal grades from U.S. News and World Report, scoring low marks for nurse-to-patient ratio and recurring safety errors.

9. Odessa, TX
> Readmission rate: 15.9%
> Avg. 30 day mortality rate: 13.6%
> Hospital safety score grade: B-
> Preventable hospitalizations: 77.0 per 1,000

In Odessa, nearly 17% of heart attack patients died within 30 days of admission, one of the highest such rates of any metro area in the country. Similarly, nearly 14% of people admitted for pneumonia died within 30 days of admission, a rate more than 2 percentage points higher than the corresponding national rate of 11.5%.

Odessa hospitals also had one of the highest unnecessary hospitalization rates in the country. Each year, about 77 patients for every 1,000 Medicare enrollees are unnecessarily hospitalized for conditions treatable by ambulatory care, significantly higher than the national rate of 59.3 unnecessary hospitalizations in 1,000. According to the Leapfrog Group, a nonprofit organization that aggregates data to rank hospital safety, one of the area’s two major hospitals received a safety grade of C. The third was not ranked by Leapfrog. Safety scores measure unsafe processes and negative outcomes in hospitals, important in evaluating patient safety and health care quality.

8. Greenville, NC
> Readmission rate: 17.2%
> Avg. 30 day mortality rate: 12.8%
> Hospital safety score grade: D-
> Preventable hospitalizations: 57.9 per 1,000

Scoring a D-, the metro area’s hospital received the second lowest safety score of any U.S. city hospital system. Hospital safety scores measure the safety of the hospital’s procedures and processes as well as the patient outcomes. The share of discharged patients who are readmitted within 30 days — the readmission rate — was particularly high in Greenville. At 17.2%, Greenville’s readmission rate was a full 2 percentage points higher than the national rate and nearly the highest in the nation. However, certain departments do not fare as poorly. Greenville hospital had the fifth lowest heart attack mortality rate, better than some of the best hospitals in the country. Despite this, 12.8% of Greenville hospital admissions resulted in death within 30 days, 1.3 percentage points higher than the national hospital mortality rate.

7. Bowling Green, KY
> Readmission rate: 17.3%
> Avg. 30 day mortality rate: 11.8%
> Hospital safety score grade: C
> Preventable hospitalizations: 101.5 per 1,000

Of all U.S. metro areas, Bowling Green had the highest unnecessary hospitalization rate. With 101.5 patients unnecessarily hospitalized per 1,000 Medicare enrollees, Bowling Green was the only area where there were more than 100 patients hospitalized for ambulatory treatable conditions per 1,000 patients, significantly higher than the national rate of 59.3 per 1,000. Furthermore, with 17.3% of those discharged from area hospitals readmitted within 30 days, the Bowling Green metro area had the fourth highest readmission rate in the country.

Data show that area hospitals had especially poor records in treating chronic obstructive pulmonary disease (COPD). Nearly 10% of patients admitted for COPD died within 30 days, one of the highest such rates in the country and higher than the national rate of 7.7%. Those attending the city’s largest hospital, the Medical Center at Bowling Green, will face subpar safety standards. The center received a C for safety from the Leapfrog Group, with staff earning particularly bad marks for alertness to breathing problems after a surgery and preventing infection.

6. Madera, CA
> Readmission rate: 16.4%
> Avg. 30 day mortality rate: 13.9%
> Hospital safety score grade: F
> Preventable hospitalizations: 49.0 per 1,000

Madera was the only metro area to receive an F for safety from the Leapfrog Group, the worst safety rating for a hospital system in the country. In California, 42.7% of the state’s hospitals received an A grade for safety — the seventh highest such percentage in the nation — which makes Madera’s poor rating a significant outlier. In Madera, 13.9% of hospital admissions resulted in death within 30 days, a mortality rate 2.4 percentage points higher than the national rate. Despite the low scores, Madera hospitals had a relatively low unnecessary hospitalization rate. Madera hospitals admitted only 49 patients per 1,000 Medicare enrollees for conditions that could have been treated outside the hospital, about 10 fewer unnecessary admissions than the national rate.

5. Bloomsburg-Berwick, PA
> Readmission rate: 17.0%
> Avg. 30 day mortality rate: 13.4%
> Hospital safety score grade: C
> Preventable hospitalizations: 51.2 per 1,000

The Bloomsburg-Berwick metro area hospitals had a relatively low rate of hospitalizations of patients treatable by outpatient care. About 51 patients were unnecessarily hospitalized per 1,000 Medicare enrollees, lower than the national rate. Still, area hospitals had below average performances in several other measures. Nearly 19% of patients admitted for a stroke died within 30 days, one of the highest such rates in the nation and about 4 percentage points higher than the national rate of 14.8%. Also, 17% of people admitted to a metro area hospital needed to be readmitted within 30 days, indicative of hospital error.

The Geisinger Bloomsburg Hospital, the area’s smallest hospital, received a C in safety from the Leapfrog Group. Particularly, the hospital’s staff received a below average score in adhering to safety measures surrounding surgical procedures.

4. Hot Springs, AR
> Readmission rate: 15.8%
> Avg. 30 day mortality rate: 14.3%
> Hospital safety score grade: C-
> Preventable hospitalizations: 66.7 per 1,000

In Hot Springs, 14.3% of patients who were admitted to hospitals died within 30 days — the fifth highest such rate in the country and 2.8 percentage points higher than the corresponding national mortality rate. Of the five cities with the worst hospitals in the country, however, Hot Springs hospitals had the lowest readmission rate. At 15.8%, Hot Springs hospitals’ readmission rate was much closer to the national rate of 15.2%. Only 7.1% of Arkansas hospitals received an A for safety, the fourth lowest percentage of any state in the country. In Hot Springs, only one hospital received a B, the highest score in the metro area.

3. Kingston, NY
> Readmission rate: 17.5%
> Avg. 30 day mortality rate: 13.3%
> Hospital safety score grade: C
> Preventable hospitalizations: 64.2 per 1,000

Upstate New York was home to two of the three cities with the worst hospitals. The Kingston metro area hospitals had the second highest readmission rate of any city hospitals in the country. Nearly 17.5% of patients had to be readmitted within 30 days of their initial discharge from a local hospital.

Area hospitals had an especially bad record in treating COPD. More than 10% of patients admitted for COPD died within 30 days, one of the highest rates in the country and much higher than the national rate of 7.7%. The HealthAlliance of the Hudson Valley – Broadway Campus received a safety score of C from the Leapfrog Group. Particularly, the hospital’s staff received a below average score in adhering to safety measures surrounding surgical procedures.

2. Morgantown, WV
> Readmission rate: 17.0%
> Avg. 30 day mortality rate: 13.3%
> Hospital safety score grade: C-
> Preventable hospitalizations: 88.6 per 1,000

Hospitals vary in medical specialty. While a hospital may excel in cancer treatment, it may trail far behind national standards in stroke recovery. Morgantown hospitals perform relatively well in treating COPD and pneumonia, but are among the worst in the country in treating heart attacks and strokes. However, high mortality rates resulting from heart attacks and strokes are not the reason for Morgantown’s poor hospital ranking. Morgantown had a particularly high number of preventable hospitalizations, with 88.6 hospitalizations per 1,000 Medicare enrollees for conditions treatable by ambulatory care. This was far higher than the national rate of 59.3 preventable hospitalizations per 1,000 Medicare enrollees. Morgantown hospitals also had one of the highest readmission rates in the country, with 17% of discharged patients returning to the hospital within 30 days.

1. Binghamton, NY
> Readmission rate: 17.3%
> Avg. 30 day mortality rate: 14.2%
> Hospital safety score grade: C-
> Preventable hospitalizations: 59.4 per 1,000

Unlike most poorly ranked hospitals, Binghamton hospitals performed inadequately in every measure of hospital quality. As many as 17.3% of discharged patients were readmitted within 30 days, nearly the highest such rate in the country. Similarly, 14.2% of patients died within 30 days of being admitted to a hospital in Binghamton, one of the highest mortality rates of any metro area’s hospitals. This mortality rate accounts for the inherent risks of death associated with the diseases treated. Only in one measure did Binghamton hospitals achieve an average rate — unnecessary patient hospitalizations, as only 59.4 patients per 1,000 Medicare enrollees were hospitalized for conditions otherwise treatable by ambulatory care. This was similar to the national rate of 59.3 patients per 1,000 Medicare enrollees.

Correction: Due to a fact checking error, an earlier version of this article incorrectly stated that the Geisinger Bloomsburg Hospital was one of the area’s largest. In fact, it is the smallest hospital in the Bloomsburg-Berwick, PA metro area. However, because the area’s score was calculated independently, this error had no impact on the metro area’s performance in our ranking.

Additional correction: Due to a fact checking error, an earlier version of this article incorrectly stated that two of three major hospitals in the Odessa, TX area received a safety score of C from Leapfrog. In fact, only one of two major hospitals ranked by Leapfrog received a C. The third hospital was not considered by Leapfrog.

Methodology:

To determine the cities with the worst hospitals, 24/7 Wall St. created an index using data from the Centers for Medicare and Medicaid Services (CMS) and other sources. The CMS produces statistics on numerous health measures for most hospitals in the country. We grouped hospitals by their respective metropolitan areas and aggregated each measure to the city level.

Each datapoint was converted into an index using the min-max normalization method. The composite index was calculated from each individual index. Mortality and readmission rates were each given 40% weights. Preventable hospitalizations was given a 15% weight and the Leapfrog Group’s safety score was given a 5% weight.

First, we calculated a weighted average of 30-day mortality rates for heart attacks, COPD, heart failure, pneumonia, and stroke to find the average percentage of people who died within 30 days of being admitted to a hospital. For many experts, 30 days is an appropriate window in which a hospital, rather than individual behavior, is at fault for a patient’s death. All mortality rates are risk-adjusted.

Second, we included the rate at which individuals were readmitted to a hospital within 30 days of being discharged. Like mortality rates, readmission within 30 days speaks to a health system’s low quality of care and poor communication between hospitals and ambulatory care facilities, rather than individual behavior.

We also included the Leapfrog Group’s hospital grading, which considers a host of objective measures related to a hospital’s care delivery, as well as responses of patient surveys. Grades range from A to F. To quantify each measure, we assigned A grade a score of 90, B grade a score of 80, and so on.

Lastly, we used the share of hospitalizations for conditions that could have been treated with outpatient or ambulatory care — often referred to as the unnecessary hospitalization rate. Included in this measure is hospitalization for conditions such as asthma, dehydration, or hypertension. This data came from County Health Rankings.

Thank you for reading! Have some feedback for us?
Contact the 24/7 Wall St. editorial team.