An audit of death certificates from Norway has found that one third of all deaths attributed to prostate cancer were actually caused by something else.
However, there was also an underreporting of prostate deaths in the study — wherein some men who were reported to have died of other causes were considered by the researchers to have actually died of prostate cancer.
Overall, the correlation between the registered (on death certificates) and observed (upon review by the investigators) prostate cancer deaths was 81%.
The study was published online January 11 in the Scandinavian Journal of Urology.
The overreporting of 32% was partially offset by a 23% rate of underreporting, but the “net result was a considerable over-registration of prostate cancer deaths,” say the authors, led by Sven Löffeler, MD, a urologist at the University of Bergen, Norway.
Overreporting of prostate cancer deaths — particularly among elderly men — “may not be an exclusively Norwegian phenomenon and may affect prostate cancer mortality statistics in other countries,” they comment.
The authors say their investigation is important because “official cancer mortality statistics strongly influence the perception of different cancers.”
Dr Löffeler observed that the medical literature routinely reinforces such perceptions. “Virtually every article that is published on prostate cancer starts with something like, ‘Every year in the US so-and-so many men die of prostate cancer,’ suggesting that reducing this number is what we are aiming at,” he said. “But what if these numbers are unreliable?”
Study Details
“Our study is one of the very few that has studied an entire prostate cancer death population and their medical history in detail,” Dr Löffeler told Medscape Medical News.
The investigators reviewed the records of all deceased men from 2009 to 2014 in Vestfold County, 1 of the 18 counties in Norway. Vestfold is small (only 230,000 residents) but is representative of the socioeconomics of Norway’s population.
The study’s review committee consisted of three urologists and one oncologist, all of whom treat patients with prostate cancer. They did a careful, three-part review of patient history, checking death certificates against medical records to determine the likelihood of a prostate cancer–related death.
The team identified 328 men who had prostate cancer as the immediate cause of death (part I of the death certificate) and 126 who had prostate cancer as another significant condition (part II of the death certificate).
There was an additional group of 310 men who died but did not have a diagnosis of prostate cancer on their death certificate; a crosscheck using the Norwegian Cancer Registry and the Norwegian Institute of Public Health revealed an earlier prostate cancer diagnosis.
In the first group (prostate cancer as the immediate cause of death), 32% of patients died of another cause (n = 105).
In the second group (prostate cancer as another significant condition but not the cause of death), the investigators determined that 18% (n = 23) had actually died of prostate cancer.
In the third group (prostate cancer not listed on death certificate but in medical records), the team determined that 5% (n = 16) had actually died of prostate cancer.
As noted above, this amounted to 32% overreporting and 23% underreporting of prostate cancer as the cause of death.
In general, the likelihood of misattribution increased significantly with increasing age and decreasing Gleason score, the investigators note.
The authors emphasized the importance of advancing age in their results.
“Prostate cancer mortality statistics in Norway are relatively accurate for patients aged <75 years at death. However, overall accuracy of cause of death assignment is significantly reduced by misattribution among older patients (>75 years), who represent the large majority of prostate cancer deaths,” they write.
The authors stress that their study design is one of the study’s strengths. Other studies of prostate cancer deaths using death certificates in Finland, Sweden, and the United Kingdom have demonstrated “high accuracy” of those certificates. (A study in the United States revealed less accuracy.)
But the authors point out that those studies were based on data from large clinical trials, which are known for having patients with prostate cancer who are “significantly younger at death.”
On the other hand, in a general population, such as in this Norwegian study, most prostate cancer deaths are among older men, who have more competing causes of death and thus are more likely to have a misattributed cause.
“You have to remember that in Western societies literally no one dies officially of old age. You have to put a diagnosis on the death certificate…this can be very tricky and the potential for misrepresentation is huge,” said Dr Löffeler.
You have to put a diagnosis on the death certificate…this can be very tricky and the potential for misrepresentation is huge.
In Line With Previous Reports
However, another researcher who has studied this issue did not see the new Norwegian study as exceptional. Katja Fall, MD, PhD, an epidemiologist at the Karolinska Institutet in Stockholm, has previously reported on the reliability of death certificates in patients with prostate cancer in Sweden (Scand J Urol Nephrol. 2008;42:352-357) and was approached for comment by Medscape Medical News.
“My impression is that the studies report findings that point in the same direction: Our overall correlation was 86% and the corresponding figure in the Norwegian study was 81%. Both studies also suggest that the accuracy is higher in younger men (younger than age 65 in our study and younger than age 75 in the Norwegian) than in older.”
However, the rate of overreported prostate cancer deaths was higher in the Norwegian study than in her study of Swedish men.
Both Dr Fall and the authors also pointed out that the higher misattribution rates seen in the Norwegian study (vs the study in nearby Sweden) could be due to the more recent timeframe, with higher underlying prevalence rates of prostate cancer.
Prostate cancer prevalence doubled in the past decade in Norway, as prostate-specific antigen testing became much more common.
In any case, Dr Fall also agreed with the Norwegian authors that determining prostate cancer as a cause of death is challenging “because of the long expected survival (relative to other cancer types) and many competing causes of death.”
In view of the challenges of determining a prostate cancer death, Dr Löffeler and coauthors suggest using another metric for determining the ultimate impact of the disease on a population: an account of castrate-resistant and metastatic disease.
“It is the end-stage of prostate cancer and it is certainly something we would want to avoid in most patients. But it could in theory be easily and reliably reported and followed over time,” said Dr Löffeler.
The study was supported by a research grant from Vestfold Hospital Trust. The authors and Dr Fall have disclosed no relevant financial relationships.
Scand J Urol. Published online January 11, 2018. Full text
Follow Medscape senior journalist Nick Mulcahy on Twitter: @MulcahyNick
For more from Medscape Oncology, follow us on Twitter: @MedscapeOnc
Tidak ada komentar:
Posting Komentar