Many cancer patients have other diseases, and those ailments can influence their chance of survival and response to treatment.
Although physicians take a patient’s overall health into account when determining prognosis and making treatment recommendations, that information is not recorded in tumor registries and used in medical research, according to Jay F. Piccirillo, M.D., associate professor of otolaryngology, of medicine and of occupational therapy.
Piccirillo and his team have developed a program to train cancer registrars to track patients’ other diseases, called comorbidities. Five centers around the country, including the Siteman Cancer Center, have been learning and implementing the program since 1999.
Piccirillo recently discussed his group’s methods at the Centers for Disease Control and Prevention’s 2003 Cancer Conference in Atlanta. At the same conference, another team member and research assistant, Irene Costas, outlined findings from the five centers.
“Comorbidity is increasingly recognized as an important feature of the patient with cancer,” Piccirillo said. “Some people are skeptical about whether it’s practical to train registrars to code comorbidity information, but our research shows that the method we’ve developed is easy for them to learn and to do well.”
He believes recording comorbidities will provide more comprehensive profiles of patients and improve cancer research. For example, a patient’s chance of survival depends not only on his or her cancer type and severity but also on the patient’s other health problems. Therefore, in comparing the success of two treatments, it may not be accurate to simply compare all cancer patients undergoing treatment.
Similarly, comorbidities can cloud quality-of-care assessments across different institutions because certain institutions tend to see sicker patients. Cancer staging helps control for the differences in tumor size, but no attempt is made to control for differences in the overall health of the patient.
In 1996, Piccirillo’s team developed the “Comorbidity Education Program,” which includes a training manual, video and data-collection forms. The program incorporates the 27 most common comorbid ailments, including the level of severity of these conditions on a four-point scale: no disease, mild, moderate or severe. This information is combined with data already being collected by registrars, including tumor size and type.
Cancer registrars trained in this program yielded consistent results when validated by a trained research assistant who evaluated the program’s success. The average time spent recording information from medical records only increased by about two minutes per record, and post-program questionnaires revealed that registrars found the coding method relatively simple and easy to learn.
The information gathered over the past several years is already being evaluated. For example, patients with severe comorbidity are almost three times more likely to die than patients with no comorbid ailments, even after controlling for tumor site and size and patient age, race and gender.
Because of the importance of comorbidity information in determining treatment and prognosis, the American College of Surgeons’ Commission on Cancer recently mandated the collection of this information and the method for its collection. But according to Piccirillo, the commission’s approach, which is based on the ICD-9 coding system used for medical bills, has several key flaws.
First, it does not include information about disease severity. Therefore, an individual with mild diabetes that can be controlled by dietary restrictions receives the same code as an individual with severe diabetes who takes daily insulin shots.
Medical bills may also omit certain diseases for socio-political reasons (mental illness or AIDS) and often are written with the intent of maximizing reimbursement rather than capturing the most severe comorbidities.
“The Commission on Cancer based its approach on the theory that training registrars on a new coding system is time-consuming and impractical,” Piccirillo said. “But we’ve taken one example of a new method and we’ve studied it, we’ve documented it and we’ve published it: Cancer registrars can learn comorbidity coding.”
Having improved and refined their training protocol during the first several years of its use, the “Comorbidity Education Program” takes 10 hours to learn. The team is finalizing a Web-based training program based on their experiences with in-person training seminars.
The National Health Service in London is planning to incorporate this program into its current cancer registry system throughout the United Kingdom.
The program will also be used in the five-year Cancer Care Outcomes Research and Surveillance Consortium project, sponsored by the National Cancer Institute. This initiative will enroll 10,000 patients with newly diagnosed lung or colorectal cancers and address how characteristics of patients, providers and the systems delivering care affect what services patients receive for the management of cancer and its consequences.
In addition to training other cancer centers about its program, Piccirillo’s team is working on developing mathematical models that identify important comorbidity elements in elderly cancer patients.
Those models may provide a more comprehensive and informative way to determine prognosis for elderly cancer patients.