Miscalculating morbidity: Deaths from cutaneous T-cell lymphoma often misclassified, study shows

Key Points

Providence, R.I. - The incidence of cutaneous T-cell lymphoma is on a steady rise; however, according to a recent study, fewer than half the deaths of patients who succumb to this mysterious disease are properly reported as such, raising serious issues on disease classification in respect to mortality.

"We wanted to see if we could reasonably assess if there is a mortality increase, and what we discovered was that mortality data is seriously flawed, in the sense that the data is very unreliable.

"Like many other disorders, we looked at trends of death by looking at death certificates. Unfortunately, when looking at cutaneous T-cell lymphoma, the numbers simply do not match up," says Martin A. Weinstock, M.D., of the department of dermatology at Rhode Island Hospital, Brown University.

They found that the number of deaths that they estimated were attributable to cutaneous T-cell lymphoma were, in fact, much greater than the number of deaths that were actually attributed to cutaneous T-cell lymphoma on the death certificates.

Misclassifications

According to Dr. Weinstock, one reason for this inaccuracy is the substantial amount of misclassification on death certificates between cutaneous T-cell lymphoma and other types of lymphoma.

He suspects that the physicians who complete the death certificate may just say "lymphoma" without specifying which type, leading to the gross inaccuracies. A large number of these patients were general lymphoma patients as opposed to cutaneous T-cell lymphoma, and many of them had co-morbidities that could have led to their deaths, including heart or lung disease as well as other forms of cancer.

"The consequence of this is that when looking at the impact of cutaneous T-cell lymphoma, and when looking at death certificates, one can serious(ly) underestimate what the impact would be, and one could be deceived in terms of other characteristics, such as age, gender and geographic region, of the mortality from cutaneous T-cell lymphoma by simply looking at death certificates," Dr. Weinstock tells Dermatology Times.

Dr. Weinstock says steps need to be taken in order to improve the process and get the numbers straight.

One approach could be to provide more detail on the death certificates of these patients and coding of individual death certificates, as well as on how physicians actually code the underlying cause of death.

Dr. Weinstock estimated the proportion of the population-based cohort of patients who actually died from cutaneous T-cell lymphoma, and found that the 10- and 20-year relative survival rates were 79 and 72 percent, respectively.

The 72 percent survival rate over 20 years means that in the cohort of cutaneous T-cell lymphoma patients a little more than 1 out 4 (28 percent) who ordinarily would have survived 20 years did not survive, and those are the deaths that are attributed to the cutaneous T-cell lymphoma.

According to death certificate data in this particular cohort, there were 124 deaths at 20 years follow-up in this particular group of people who had cutaneous T-cell lymphoma. But the estimated actual death at 20 years that could be attributed to the cutaneous T-cell lymphoma was 289 patients.

That means that only 43 percent of the actual deaths were, in fact, listed as due to cutaneous T-cell lymphoma on the death certificate.

"The system is flawed and needs to be revamped. We also need to learn more about cutaneous T-cell lymphoma. And to do that, we have to look at risk factors and trends of morbidity in the population with respect to age, gender and geography, among other parameters," Dr. Weinstock says.

According to Dr. Weinstock, the accuracy varied by time after diagnosis, and the greatest accuracy or the least biased by numbers seemed to be five to 10 years following diagnosis.

Dr. Weinstock speculated that those patients who died in the five- to 10-year period after diagnosis more obviously died because of cutaneous T-cell lymphoma, and it is, therefore, more accurately recorded.

Possibly, the patients who died later were more vulnerable to infections or system failures, which were the actual cause of death.