Replies

The numbers are not estimates. When a doctor, coroner, medical examiner, funeral home director or other person records a death and signs a death certificate they assign a cause of death. Sometimes the cause of death is listed as unknown. Sometimes the kinfolk or insurance company contradict the cause of death. This is most often true is suicide. Then there is a process to review the cause of death.

The reality is that many people who die have several things wrong with them at time of death and it is the combination of multiple factors, and no single factor, that causes death. But our system is set up to list the PRIMARY cause of death.

Besides suicide, Condition Acquired in the Medical Facility is the most often fictionalized. A person enters the hospital with a bullet or knife wound, acquires SEPSIS in the hospital and dies of sepsis. The death certificate lists the bullet or knife wound as the cause of death. That is the bias of the person filling out the form. Often times a low paid clerk fills out the form and the doctor hardly reads what he signs. (Some would say purposely does not read).

The statistics have flaws. But for most causes of death they are statistically reliable. The states collect the raw data from those who sign the death certificate (or their oranizaiton). The CDC then collects the data the state colected. On rare occaisions there is an error in the transmission of the data. It has been known that one file layout B has n lines of cause of death. file layout A is an old, obsolete layout that has n-1 lines. And the extra line is in the middle and all lines after that are off.

This happens rarely, but happens. Most of the time the error is caught. But if an error was not caught and fixed, how would we know...it wasn’t caught. It probably is not statistically significant.

Unaware bias is likely the biggest problem.