By R Carter
In Part 2 of this series I looked at Box 32 Part 1, where certifiers fill in the causes of death on a death certificate. And I looked at the basic rules they follow to ensure that one cause has a medical link to a preceding. More importantly I looked at how the entire chain is needed to have an accurate context for a cause of the death. As key requirements they have a bearing on how data is collected when compiling statistics. Finally we established how taking a single data point out of context, leads to inflating estimates on the cause of death.
This is an important concept, as the data has one meaning to doctors and researchers but as used, leads to misunderstanding and confusion by the general population.
|A Doctor’s Judgement|
Part 3 of this series is subdivided but all parts come under the heading of “Weaponizing the Data”. I’ll look at the how, what, when and where for data used in compiling statistics. It’s not an exhaustive effort nor a comprehensive one, the subject is too large for a single researcher. But, I’m not the only one asking these questions. As I discover others sources I’ll post links.
Let’s say a patient is a chronic pain patient of 14 years taking opioids without incident or complications and he has developed liver cancer. The patient eventually slips into a coma and dies. Under current definitions, the patient could be counted as an opiate related death. But in practice opiates played little and likely no part in the death.
As is often the case, cancer patients are given opiates in high doses during the final stages of cancer to alleviate suffering. But even this wouldn’t constitute a cause of death. If it did, then under that logic we would also have to include chemotherapy as a cause, knowing that these classes of drugs are far more toxic than opiates.
It’s these types of fine distinctions which are often exploited by those reporting on opiate overdose deaths. Another example would be comparing opiate overdose deaths with auto accident deaths. While an interesting comparison it has no real meaning as driving an automobile is not a disease, where as addiction is. Nor is treating a pain condition with opiates in anyway related to automobile deaths. To connect the two, would require first proving the individual was impaired, a very difficult claim to make as opiates do not produce the same level or type of mental impairment as do things such as alcohol or sedatives.
The science of pharmacologically induced mental impairment is complicated and requires a medical background. It’s also highly dependent on many factors, age, body mass, dosage amount, other comorbid conditions. And for this reason, when opiates are singled out as a cause, those reporting such statistics are in fact taking a very liberal position, sometimes fabricating evidence to support an ideological point of view.
These facts should be taken into account before blindly accepting as fact some headlines and statements made.
|Where Judgement Presents an Opportunity|
Whether or not this person is counted as an opiate death depends on three things.
- First, the diagnosis codes used by the certifier completing the death certificate. While chronic pain treated with opiates is one of the diagnoses, it’s up the judgment of the certifier to include that information on the death certificate. It’s also up to the certifier to decide the weight of this information in contributing to the final cause of death.
- Second, a researcher polling data may blindly collect all deaths whose diagnosis codes include opiates. Or the researcher may exclude some based on other criteria. In our example opiates played no direct role in contributing to the cause of death. Using this record from the database is a judgement call. Often subjectively influenced based on the researchers point of view or experiences.
- Third, large scale studies like this one, 2016 Guidelines from the CDC, references multiple research studies. A researcher must look at each paper and make a judgement on whether or not the data has real value for their efforts.
|Weaponizing the Data|
These two later points play a critical role in the final published statistics. Because data is often inconclusive, what’s included or excluded from a study, is dependent upon the judgement of researcher. Subjective motivations can come into play in the final outcome. While it’s expected a researcher will remain open and objective, researchers who review the research process in publications report up to 60% of all medical research fails to follow scientific methodologies which are required to qualify as reliable research.
Researchers whose moral and ethical values allow cutting corners, fudging so to speak, often let their ideological views or other pressures, dissuade them from reporting honest facts or making logical conclusions.
This is why research is peer reviewed. Peer review could be repeating the same experiment in an effort to produce the same results. The most frequent method is having experts in the same discipline review the research for errors, omissions or bad logic in drawing conclusions.
|PEER Review Process|
Peer review is also vulnerable to subjective judgment calls, usually by selecting peers whose views favor the desired conclusions of the original researcher. Peer review in the strictest sense, should include multiple experts. Include those who are pro, con and uncommitted to the conclusions of the original research.
It’s my opinion that the peer review process is where data weaponization occurs most often when building an anti-opiate message for chronic pain.
For the 2016 CDC Guideline on Chronic Pain, there are two telling facts. First the majority of the selected expert panel was known to have an anti-opiate stance. Second, the research studies selected for inclusion, were by the CDC’s own admission, inconclusive for use in chronic pain. So instead the panel selected data using opiate naïve patients. By all the definitions of good science, this was a major failure.
In Part 3.2 I will take a look at Liberalizing the Weaponization Process and Widening the Scope of Weaponization.