Scientific Bias in Studies

[anthony]

From some recent conversations here it has resurfaced surrounding bias within scientific studies, and I most certainly concur with this. The one that makes me write this now is the latest media escapade on "concussion increases PTSD risk within soldiers."

Now you don't have to look hard for this one, just go to Google News or the like and do a quick search and you will find it. The reason I and many say these studies are biased is because the study goes looking for a particular result, not taking something and testing it across a wide variety of aspects and concluding what comes out the other end.

The study mentioned, they put all these soldiers who suffered concussion during service and measured how many had PTSD or PTSD attribute. Well, astonishingly.... lots did, because they endured a trauma to get concussion in the first place. So they put another group in as a placebo effect, those who haven't suffered concussion yet experienced operational service, funny enough the amount is lesser.

This is a very basic breakdown of the scenario, though you don't need be a rocket scientist to discover what is referred as bias towards a study. If you canvas a group of people, identifying specific attributes within them, you immediately bais that study to a group that firstly must have x, y & z to obtain this result. Some of which is relevant at times, some of which is not.

The obvious factor here is that if you suffer concussion you will get PTSD. Its not because of the concussion, but because the trauma endured to be concussed in the first place. I just hate reading some studies.... they are stupid before they begin.

 

[Lisa]

Brilliant point to raise.

When I'm studying articles and research for my degree, I am constantly having to critically evaluate in order to test it out as something that truly measures what it purports to measure.

Research studies are not to be taken as the Gospel unless critically looked at. Not everything published (and be very wary of something unpublished) is necessarily true. Every researcher could be argued to be 'biased' in that researchers develop a hypothesis(es) and then test it... however these days a decent rationale based on previous research is required for any study... but for relatively 'new' areas such as the concussion one you mention, Anthony, the study needs to be even more critically evaluated to see if it really is measuring what it says it is (in this case, that concussion causes trauma - which it doesn't. Trauma causes PTSD.).

Another thing to be wary of is that not all researchers are as honest as they like to say they are. A good example can always be found in history. 'The Bell Curve' is a classic example of a biased study, cheating, and institutionalised racism.

This isn't to say that all studies are not to be trusted... psychology, psychiatry and science have learned from their mistakes, and are much better at distinguishing the studies that deserve publication. Though there will always be studies way off mark. It's only through critical evaluation, testing, retesting, and alternative theories to compare to ,that research pushes forward knowledge and understanding.

It's something that can be applied to everything. Knowledge is power. But just because a research article says something it doesn't mean we must take it as the truth. It means it must be critically analysed so that no obvious faults or biases are present. Generally, peer reviewed articles do this, but it's something everyone can do when they know what to look for. This is the very nature of science in some ways.

(That philosophy of science essay is really coming in handy!).

 

[ruddy]

Scientific research is generally based on some form of statistical analysis. I've become quite cynical about statistics. Hell, I've spent half of my 25 year career as an accountant developing creative convoluted logic and statistical analysis that supports the forgone conclusion. It is naive to trust every statistic presented as fact without looking at sample size, parameters, and methodology used. Unfortunately that information is rarely shared with the end user.

Motivation for the end result also needs to be considered.

• A scientist or researcher who is anxious to get published may be willing to "push the envelope" to make a name for him/herself.
• Who is funding the research? The tobacco industry isn't going to tell us that smoking kills and the American Cancer Society isn't going to tell us it's safe.
• Do future profits hinge upon research results? I'm thinking about new products and drugs hitting the marketplace. Of course they're safe and effective??

 

[anthony]

Well, a good point you raise ruddy surrounding Tobacco, is that tobacco actually have the 10 fat fingers and toes in the nicotine patch market. Nicotine is proven to be completely gone from your system within 7 days... no traces what so ever. So what is so addictive then to give up smoking? Is it needed that one must continue putting an addictive substance into their body over a 3 month or longer period so that they remove that addictive substance, or more so it remains within them and they have a chance of collapsing back to cigarettes? The later is the idea.

They do this to prolong the craving, because the craving is not for nicotine after the first 7 days, it is the hand to mouth habit that one forms and the brain becomes used to as a coping method. This is why so many replace it with food and lead themselves down another dangerous path with obesity.... so what do they do? Go back to smoking as it keeps them slim, because of habitual issues, not because of the addictive nicotine.

Tobacco have their fingers in everything, even helping people give up smoking to hinder the process. I did research on this not long ago, and the scientific conclusions all round say nicotine is addictive, nicotine however is fully gone from the human system within a week. Some found trace elements at 9 days, though nothing by 10 days, though the studies print 7 days as the majority factor for complete nicotine withdrawal. Hell, even if you say 10 days, as nobody is was found to have it after that, not even a trace element, na da.... so why would anyone want to continue putting nicotine into their system through nicotine patches then? The tobacco companies do.... because they get their money one way or another and increase the chance of getting you back smoking.

Interesting what studies do though.... bias, hide facts about company interests, pharmaceuticals are the worst IMHO. They fund studies to show bias towards their drugs.... how corrupt. Even scientist will corrupt to money in order to continue their research. If they have to print one biased document that the world will grab to fund study for another two years, they will do just that.

As Lisa said above.... be careful what you read in just one study, or the same study retyped and spread across the web. Go looking for independent and unique studies based over a period of time and look for consistencies, as that is how you find more accurate results IMHO.

 

[Beachbum]

Anthony wrote: 'The obvious factor here is that if you suffer concussion you will get PTSD. Its not because of the concussion, but because the trauma endured to be concussed in the first place.'
I beg to differ: you won't automatically get PTSD if you get concussion BUT with concussion you are have received a brain injury, therefore unless they do ALL necessary tests that rule out brain injury (this is difficult as 'diffuse' damage like brain shear often doesn't show on MRI), then they will put all your symptoms down as PTSD when in fact they may well have a PHYSICAL reason that needs DIFFERENT treatment/rehab. Of course if have experienced a truly terrifying incident aswell = PTSD very likely. But the more I learn about brain injury & PTSD the more I believe that damage to the bain is far more common and unrecognised than people realise. Far too much brain injury is wrongly termed 'just' PTSD. Just my humble view here!

 

[anthony]

Hi Beachbum... your point is valid, though are you referring this to the study being spoken about? It is about concussed veterans from operations, not typical concussion to just anyone, but concussion to veterans from trauma related instances! Nor did I say it is "automatic" because that is not so, and I concur with your statement about that. Your presuming possibly?

The study took a group of veterans who suffered concussion during operational experience. Now you understand what makes concussion during war I would assume? Mortars, explosions... traumatic events is the point!

If your taking that one statement outside of those realms, then that is not what this is about and what your saying is absolutely correct. Normal concussion is first priority a brain matter, not a traumatic matter. Even in operations it can be a brain matter, though concussion on operations is not; get concussed and go to the doctor. It is get concussed, get the shit shaken out of you by another soldier to get up and keep fighting or retreat, not go to the doctor. Trauma in soldiers concussion is the majority disposition to the act. Hence why the study about this is nothing short of biased nonsense.

 

[lrs]

bias - "a partiality that prevents objective consideration of an issue or situation" from wordnet princeton.
****************
dishonest - "Dishonesty is a term which in common usage may be defined as the act of being dishonest; to act without honesty; a lack of probity, to cheat, lying or being deliberately deceptive; lacking in integrity; to be knavish, perfidious, corrupt or treacherous; charlatanism or quackery." from widipedia
****************
I almost expect the element of bias, and can accept it, to a small degree. It's part of the human equation.
It is the dishonesty I have a problem with.

 

[ruddy]

I almost expect the element of bias, and can accept it, to a small degree. It's part of the human equation.
It is the dishonesty I have a problem with.

The line between bias and dishonest can be pretty thin. IMO in practical terms the biggest difference is that complete dishonesty might be criminal and result in a jail term. On the other hand bias studies can be honest, mathematically correct, and still yield useless or unreliable results.

 

[anthony]

I agree with that statement, being a thin line. If you relate study bias to pharmaceuticals... well, that thin line disappears with profit through dishonesty overruling all other factors of the study... I just don't believe one study on any pharmaceutical aspect. I must have many, so I can compare them for similarities and dismiss the rest.

 

[lrs]

I work as a clinical pharmacist. I encounter outright dishonesty often enough. It's like if people smell money, many lose their bearings. I see it all levels.

 

[ruddy]

IRS,

I've read that women are typically underrepresented in clinical trials and studies because pharmeceutical companies prefer not to deal with whatever effect fluctuating hormone levels might have. If the study does not include the portion of the population the individual belongs to (race, age, gender, pregnancy status, ect.) how worthwhile is that study to the individual making the decision?

I guess I'm interested in hearing your professional opinion - if you care to share.

Ruddy

 

[lrs]

Honestly, I have never heard that assertion being made.