Just how precise, really?

[BrentW]

I wonder if others could address an issue that has been troubling me for some time: what is the confidence limits on the PSA test? In other words, if we are presented with a PSA reading, which seems extremely precise, being at the nearest 0.1 (in my case), how precise is that really? Should there not be a +/- attached to that? If we knew that, might it help alleviate anxieties among those of us with high Prostate Specific Anxiety levels? When I mentioned this issue to my (previous) doctor, he looked clueless -- and a little concerned, given my knowledge of statistics.

[CantChoose]

uptodate.com/contents/scree...

I have no idea if this is helpful, but leaving here in case it is. Screening as part of trends seems less accurate.

If no one can answer it, I could ask my husband to look into it. He has inside sources.

[Tall_Allen]

Day-to-day fluctuations may be ±20%. It's also a good idea to stick with one lab.

[BrentW in reply to Tall_Allen]

Thank you, TA. That's most reassuring.

[FCoffey]

I've seen my PSA fall 50% in one week and 80% in two weeks, or jump over 100% in a month.

The long-term trend matters, the number after the decimal point are all but meaningless unless the number before the decimal is zero.

[BrentW in reply to FCoffey]

Why do you write that of numbers <1.0, FC?

[FCoffey in reply to BrentW]

Because when my PSA is 57, it doesn't make any difference whether the lab result is 57.0, 57.4, or 56.8.

But when my PSA is less than one, there is a big difference between 0.0, 0.4, or 0.8. In practice, only men with no prostate (RP and some RT) will have such low scores. If you have no prostate then PSA rising from 0.4 to 0.8 might concern you, because the PSA is almost certainly coming from mets.

If I have a prostate and my PSA is less than one, I don't worry about PSA at all.

[Magnus1964]

You have to look at trends not fluctuations. If your PSA is rising with every test that is a trend. If it is bouncing up and down within a range you can ignore it.

[tango65]

If I understood correctly you are referring to the inter-assay precision of the test in one particular lab and not to the fluctuations day to day or variations due to treatment etc. it seems that the coeficient of variation (standard deviation/Mean x 100) could be 5 to 10%.

zrtlab.com/images/documents...

[PhilipSZacarias]

Hello Brent, I have been trying to obtain standard deviation data from my hospital and local testing lab for some time without success. The testing lab did indicate that the COV can be about 20% at the low end (0.1 to 0.2 ng/ml). I have seen linear increases in PSA from 0.1 to 0.2 over a period of 7 months which have correlation coefficients of 0.97 (results from 2 labs) which could be a fluke. For the last 9+ months the PSA has been stable with an average of 0.1 and COV ~8%, again from multiple labs. IMO the PSA testing from the two labs I use (Toronto, Canada) have been fairly accurate. Assuming a standard deviation of .008, then a change of 0.015 - 0.02 may indicate a change in conditions. Cheers, Phil

[Break60]

My take on the subject of Psa tests :

Regular Psa tests are to nearest tenth of a point (i.e. 0.2. , 1.2 etc.) . Ultra sensitive Psa tests are to neared thousandth of a point ( i.e. 0.002 , 1.234 etc. ) . So with a regular Psa test any reading below 0.1 would show as <0.1 and be considered “ undetectable”.

After RP , two successive readings of .2 or more had traditionally been considered recurrence and actionable with SRT .

Nowadays however ultra sensitive tests are considered appropriate following RP because .03 is considered to be the threshold for SRT not .2

There’s no reason for the ultra sensitive test thereafter because a tenth , hundredth or thousandth of a point change won’t impact treatment decision one way or another.

Tell me if I’m wrong please.

Bob

[Hidden in reply to Break60]

My docs use the regular only. Ive had tests elsewhere and instead of 0.0 with the regular test on lupron and Zytiga the result is .05. That’s only 2 weeks apart.

[Break60 in reply to Hidden]

Like I said you don’t need ultra sensitive Psa test.

[Hidden in reply to Break60]

Here at ku Med they use the sensitive test. I had my radiation and now the rest of my treatments in Houston’s m d Anderson. They only use the regular. I didn’t need the question in my own mind. Good enough for those guys good enough for me.

[j-o-h-n in reply to Hidden]

"Good enough for those guys good enough for me.

youtube.com/watch?v=LxLhytQ...(Not working as expected?)

Good Luck, Good Health and Good Humor.

j-o-h-n Saturday 05/18/2019 12:05 PM DST

[tango65 in reply to Break60]

You may be interested in this article by Dr. Partin from JHopkins.

"When the investigators tested the samples with AccuPSA, they found that at three months after surgery, all patients who ultimately had a rise in PSA had an AccuPSA level of 0.003 ng/ml or greater. By standard measures, this number would be considered undetectable. Among patients whose PSA levels never went back up, 75 percent had AccuPSA levels less than 0.003 ng/ml. These initial findings suggest that you could have an AccuPSA test three months after surgery. If your PSA level is less than 0.003 ng/ml, you could be confident that all your cancer has been removed. On the other hand, if your PSA level is greater than 0.003 ng/ml, you might choose to be monitored more closely for PSA recurrence in the immediate years following your surgery."

hopkinsmedicine.org/brady-u...

[Break60 in reply to tango65]

Thanks. I thought it was .03 not .003. At any rate after RP is as I tried to say the only time that ultra sensitive Psa is necessary.

[j-o-h-n]

Gotta be careful sometimes that decimal point positioned on the paper is really fly shit. Just make sure you blow on the paper first.....

Good Luck, Good Health and Good Humor.

j-o-h-n Saturday 05/18/2019 12:00 PM DST

[Hidden in reply to j-o-h-n]

I like cyndi. You are definitely spending too much time on your computer.

[j-o-h-n in reply to Hidden]

Cyndi agrees too........

Good Luck, Good Health and Good Humor.

j-o-h-n Saturday 05/18/2019 8:36 PM DST

[monte1111]

I've also been at 0.1 for several months. Also wondered if the lab being used went below 0.1 Nurse did not know. Doesn't really matter a lot, been on a downward/level slide for 26 months. Sooner or later it will go the other direction. Xtandi still working. CT/bone scan in month or two may change everything. (Have extensive bone mets.)

[milonguero]

In my experience, not very precise. At my last PSA check recently, the PSA was accidentally measured and reported twice, separately but by the same reference lab using the same assay method. The the two numbers, taken from the same blood sample, differed by nearly ten percent (10%)! (The actual numbers were 5.3 and 5.8)