The Inaccuracy of Digital Meters

[KN4SMF]

I'm tired of lawyer asbestos and drug commercials on TV. People die. But this fishing for plaintiffs for class action suits is a complete disgrace. They're dredging the bottom for the dregs of society who are likely to say anything.

 

[Theo Sulphate]

...
An additional subtlety relates to training.
...

You've given a very interesting perspective in many areas. I just thought of another subtlety: given this entire process, from the writing of the SOP, to calibration of instruments, training, etc., is there any sampling performed to observe the final results of the process? There should be some sampling in order to evaluate how good or bad the entire process is.

 

[Alan Edward Klein]

Making it longer would not improve your ability to distinguish the difference between 3.14159 and 3.142857 ...unless the graduations were available engraved to 3 digits right of the decimal.
So domestic tranquility will continue in your household.

What you're describing is what every measuring instrument is concerned with - accuracy vs. resolution. If a thermometer gives you the temperature down to hundreths of a degrees but is accuracy is a half a degree, it is less valuable than a thermometer than only give it down to tenths of a degree but it's more accurate at 2/10ths of a degree. All that resolution in the first case has no value. The resolution provides no valuable information. Better buy the thermometer with 2/10's of a degree accuracy and lower resolution.

 

[Luckless]

Making it longer would not improve your ability to distinguish the difference between 3.14159 and 3.142857 ...unless the graduations were available engraved to 3 digits right of the decimal.
So domestic tranquility will continue in your household.

Why would I stop at 3 digits for the decimal scales? ... Isn't that missing the entire point?

 

[alanrockwood]

You've given a very interesting perspective in many areas. I just thought of another subtlety: given this entire process, from the writing of the SOP, to calibration of instruments, training, etc., is there any sampling performed to observe the final results of the process? There should be some sampling in order to evaluate how good or bad the entire process is.

Clinical laboratories are required to subject a method to a rigorous validation procedure before the method is offered for clinical testing. Among other things this includes evaluation of experimental accuracy and precision over the full analytical measurement range and over a period of time.

Clinical laboratories are also required to participate in proficiency testing programs. In most cases this means that an independent body (such as the College of American Pathologists) sends samples to the laboratory, and the laboratory's performance on the proficiency test is evaluated. Proficiency testing is repeated on a regular schedule, generally several times per year, and if a laboratory consistently fails proficiency testing for a given analyte they will no longer be allowed to offer that test until they have corrected the problem. Furthermore, if misconduct occurs then the person in the laboratory who has overall clinical responsibility for the laboratory is at risk of criminal prosecution and could serve jail time.

There is also the matter of quality control and quality assurance. To give a taste of this aspect, the generally accepted standard for quality control includes the use of "Westgard rules" which are designed to assure that a clinical test stays within an acceptable accuracy range over a period of time. The use of Westgard rules is not universal, but they are are widely used.

There is also the matter of laboratory inspections. At a large laboratory that serves a national clientele there are multiple inspection authorities. The most important of these are the College of American Pathologists and the New York Department of Health. Both engage in very rigorous inspections of the laboratories, and if problems are found then deficiencies may be cited. If deficiencies are not corrected then serious consequences may follow.

There is also a matter of ISO certification. There is a growing trend for clinical laboratories to seek and obtain ISO certification, and this is quite an involved process.

There are also matters of certification of laboratory workers. For example, California has stringent requirements on the education and training of clinical laboratory staff, such as medical technologists, and if a laboratory performs testing on samples originating in California then they should (I believe they must) follow California's regulations in this realm. At a higher level of the lab, most labs have several medical directors that oversee sections of the laboratory. They are generally required to have certifications. For example, a PhD medical director who oversees a special chemistry section would typically have a DABCC certification, or if it is an MD medical director they would typically be a board certified pathologist, or in some cases may have DABCC certification or other equivalent level of certification. (I am DABCC certified.)

I don't know what regulations relate to preparation of drug doses, and the regulations may vary from country to country. For the sake of patient safety I hope the regulations are strict.

 

[wiltw]

Why would I stop at 3 digits for the decimal scales? ... Isn't that missing the entire point?

Not missing the point, merely stopping where the DIFFERENCE in the two values needs to be distinguishable so as to resolve one value vs. the other one.

 

[wiltw]

What you're describing is what every measuring instrument is concerned with - accuracy vs. resolution. If a thermometer gives you the temperature down to hundreths of a degrees but is accuracy is a half a degree, it is less valuable than a thermometer than only give it down to tenths of a degree but it's more accurate at 2/10ths of a degree. All that resolution in the first case has no value. The resolution provides no valuable information. Better buy the thermometer with 2/10's of a degree accuracy and lower resolution.

Yes 'precision' (or resolution) vs. 'accuracy'...very different things.

 

[alanrockwood]

Yes 'precision' (or resolution) vs. 'accuracy'...very different things.

Also, precision and resolution are not quite the same thing. To take an example from mass spectrometry, at one time there was a lot of excitement among mass spectrometrists that used certain types of ion traps because they figured out how to do high resolution measurements. The peak widths were narrow, so they could resolve different species that had closely spaced masses. However, precision was poor, i.e. the repeatability of mass measurement between runs was poor.

 

[Arklatexian]

You don't need to clutter up the scale with junk that any professional should already know. Any time a professional, and especially one who's mistakes can kill people, gets a new piece of equipment, they need to know how to use it. It's the same with any profession. And really, that rule should apply to everyone, not just professionals. But professionals should be held to a higher standard, because it's what they do for a living. Besides, putting that information on the front isn't going to help because if he didn't care enough to look it up in the manual, he probably wouldn't care enough to read any printed warnings on the faceplate. Plus, there's probably more to the scales specifications than would fit on a faceplate anyway. So the only logical thing to put next to the display would be "read and understand manual" which he shouldn't have to be told to do, to do.

And lastly, you can't tell me that if the scale was that far off that he didn't notice it bouncing all over the place as he approached 105mg. I've used inaccurate scales before. I know how to tell if it's off. That should have tipped him off that something was wrong. But rather than go the extra effort to investigate and resolve the issue, he just ignored it, killed some people, and then hired a lawyer to figure out a way to blame someone else for his laziness.

Your first paragraph tells us something I learned working with cameras over 50 years ago and that was/is "when all else fails, Read The Instructions". After using a Leica for over a year, I read the instructions that came with the camera and discovered many, many things that Leica had built into that camera that I did not know about. That is when I discovered that "it was not just "another" camera....Learned the same thing about a 4x5 Graphic.....Regards!