Patients with cancer of the larynx who are treated at teaching and research hospitals that see high volumes of such patients are the least likely to die within a year of their diagnoses, researchers said at the annual meeting of the American Head and Neck Society.
According to a review of more than 30,000 cases of laryngeal cancer over nearly a decade using the National Cancer Database, laryngeal cancer patients at high-volume teaching and research medical centers had a 20% rate of death within a year-a significantly lower number than at any other type of hospital. It was even significantly lower than the rates at other teaching and research hospitals that see fewer cases of laryngeal cancer, according to an analysis by researchers at Emory University and the American Cancer Society in Atlanta.
The study shows a close tie between the type of institution and its volume of laryngeal cancer cases and the likelihood of whether a patient will live or die, which researchers say invites a closer look at how to alter the way patients with cancer of the larynx are cared for.
There are a lot of reasons why this may be happening, and it would be important to drill down a little bit further to figure out why we are having better outcomes at high-volume teaching research hospitals, so that those processes of care can be implemented in other facilities.-Amy Y. Chen, MD, MPH
Where one gets treated is clearly associated with survival, said Amy Y. Chen, MD, MPH, of the Emory Winship Cancer Institute. There have been lots of papers done on surgical volumes and outcomes, but most report 30-day mortality rates and none of them have looked at such a large cohort. There’s been little done in head and neck oncology and looking at one-year and 90-day outcomes.
Factors Related to Survival Rates
The findings were part of a broad look at many factors and how they relate to the survival rates of those with laryngeal cancer. The study examined access to quality care, demographic factors, as well as the facilities at which they were treated.
Patient factors that can affect survival include the clinical characteristics of the tumor, demographics, and the kind of insurance a patient has. Factors involving the physician include the training level, expertise at treating this type of cancer, and the volume of the cases seen.
A 2007 study published in the Journal of the National Cancer Institute (2007;99: 1171-7) showed that the five-year probability that a patient who had received a radical prostatectomy would remain free of a biochemical recurrence increased as the number of surgeries previously performed by that surgeon increased. The increase was sharpest from zero to about 250 surgeries, showing a steep learning curve.
But the researchers with the Emory/ American Cancer Society study said that there is little historical information as to how the quality of care of laryngeal cancer patients affects their outcomes.
Laryngeal cancer is relatively uncommon, with 12,250 cases in the United States in 2008, compared to more than 110,000 cases of colon cancer, 180,000 cases of breast cancer and more than 210,000 cases of lung cancer.
Researchers hypothesized that the volume of laryngeal care influences the survival rate of patients, with high-volume facilities having better odds for survival.
They plumbed the National Cancer Database, a hospital-based registry supported jointly by the American Cancer Society and the American College of Surgeons that includes numbers for more than 1400 facilities approved by the Commission on Cancer. It is estimated that the database collects information on 70% of the incident cancer cases in the United States.
Study Results
Researchers examined statistics from cases in which patients were diagnosed from 1996 to 2004 with Stage III and IV cancer. They ended up with 32,399 cases, having to exclude about 13% of the total because of missing information. They examined death rates for several factors, adjusting them for the other factors as they went.
Next to the 20% death rate within a year for teaching and research hospitals with high volumes of cases, comprehensive cancer centers with high volumes of cases had a 23% death rate, as did comprehensive cancer centers with a low volume. Then came teaching and research hospitals with a low volume and community hospitals with a high volume (both 24%), and finally community hospitals with a low volume of laryngeal cancer cases (26% death rate within a year).
Hazard ratios show that getting treated at a community hospital that treats a low number of laryngeal cancer patients results in a 33% increased risk of death than treatment at a teaching and research facility with a high volume.
The death rate within 90 days at teaching and research hospitals with a high volume was 3%, which was also superior to the other categories. All of the others had death rates of 4%, except community hospitals with low volumes of cases, which had a rate of 5%.
Variations Regarding Type of Treatment, Insurance
Researchers also examined the type of treatment. Patients who underwent a total laryngectomy had lower death rates after 90 days and one year, at 3% and 18%, respectively. The 90-day rate for both chemoradiation and radiation were 5%, and the death rate within a year was 27% for chemoradiation patients and 29% for radiation patients.
The hazard ratios showed that the chances of dying within one year were 50% higher for those getting chemoradiation than for those getting a total laryngectomy, researchers said. Dr. Chen cautioned, though, that the numbers might not be so straightforward because those rejected for laryngectomy would probably then be sent for chemoradiation.
The type of insurance also had a close tie to the outcomes of patients. Those with private insurance (16% death rate at one year) fared the best. They were followed by those patients 65 or older with Medicare (18% rate), those with government insurance (23%), those without insurance (24%), those on Medicaid (26%) and those on Medicare who were younger than 65 (27% death rate within a year).
Basically, if you don’t have private health insurance, you are at increased risk of death, Dr. Chen said.
There were little or no statistical differences for death rates when it came to race and education level, researchers found.
Despite the vast amount of information available in the National Cancer Database, there isn’t much information available for some types of institutions, Dr. Chen said.
There is little information on the quality of care of treatment of laryngeal cancer, particularly with the nonsurgical facilities, she said. We really don’t know about the care of the chemotherapy facilities or radiation facilities and, together, how that impacts outcomes. So we’ll not only need to evaluate surgical expertise and surgical outcomes, but we should also measure what happens when they don’t get treated by surgeons.
The ties between the death rates and the facility type and case volume are compelling, Dr. Chen said.
Recommendations for Action
There are a lot of reasons why this may be happening, and it would be important to drill down a little bit further to figure out why we are having better outcomes at high-volume teaching research hospitals, so that those processes of care can be implemented in other facilities, she said. And hopefully all patients can yield the same results across the board.
Dr. Chen was questioned on whether community hospitals should, in fact, try to take on more rare procedures. I’m not sure that community hospitals ought to be brought up to speed where they can maybe do rare surgical procedures, an audience member said. That may not be the most effective place to deliver that care. Perhaps we ought to focus on regional centers and tertiary centers for rare operations.
Dr. Chen agreed. I think that that is where things should be headed and I think more data like this may push the tide, she said. Right now, all I can say is that there is clearly a difference in outcomes depending on where you have your treatment, and those things are modifiable factors. To the extent that we can modify things that we can change and we can have control over, I think it would be very easy, and I would consider that low-hanging fruit for making sure quality is good.
©2009 The Triological Society