I've seen some misinterpretations of the radiological report (pp. 371-377), so I wanted to walk people through the methodology and the math on Table 2.
The Soviets are using a measurement unit for beta contamination of disintegrations per minute per 150 square centimeters. (This is a bit strange; I've recently acquired American military surplus contamination probes, and they use 100 cm^2 as the area.)
The test equipment is an array of four STS-6 Geiger tubes in a "lead house." People in the US typically use the term "lead castle" - it's just an enclosure shielded on all sides by lead. Because the goal is to count decays from the fabric, anything that can be done to reduce the natural background hitting the Geiger tubes will make the instrument more sensitive to the sample placed inside it due to decreased background noise.
The report mentions the background radiation as being 90 pulses per minute before washing, and 100 pulses per minute after washing. My assumption is that they probably measured all the samples before washing at one time, and all the samples after washing at another time, taking one background reading for each time. Background radiation is not constant. Weather can affect radon concentration, for example. I imagine atmospheric density and the sunspot cycle will have an effect, too. To determine the background number, they simply run the detector with the chamber empty.
Let's look at item #10, Dubinina's brown sweater. Before washing, it measured 640 counts per minute, which appears in the "Total cpm" column. However, 90 is the background number, so we take 640 - 90 = 550 counts per minute.
Now there is an adjustment factor of 8.9 mentioned in the report. A Geiger tube isn't always going to respond with a pulse if a beta particle hits it because it is not 100% sensitive. Also, there are geometry issues. A radioactive sample is going to emit radiation uniformly in all directions. If you have four tubes sitting above the sample, not every particle will strike a tube. So that means not every beta emitted by the sample will be counted. It appears this detector will measure 1 beta for every 8.9 emitted. So we take 550 counts / minute * 8.9 and arrive at 4895 actual disintegrations from the sample. This appears in the report as 4900, clearly rounded, in the "Radioactivity of contaminated area" column.
Next, it must be noted that the sweater sample size was only 75 cm^2 according to the "Area cm^2" column. But the standard for activity is disintegrations per minute per 150 cm^2. So we need to multiply the result by (150 / 75) or 2. And 4895 * 2 = 9790. The report lists 9900 in the "Radioactivity of contaminated area in terms of 150 cm^2" column. Again, rounded. (This is a guess, but I'm wondering if they used a slide rule, in which case the numbers will be close but not exact.)
Looking at the values after washing, we get:
(390 - 100) * 8.9 = 2581 (the report shows 2600)
2581 * (150 / 75) = 5162 (the report shows 5200)
Now I need to note that lines 6 and 7 have typos in the English translation on dyatlovpass.com. They list the counts per minute after washing as 11 and 77, respectively. These make no sense because background after washing is 100. But looking at the report image, I see this is 111 and 177 respectively.)
Samples 2, 3, and 4 list 2 pairs of numbers for total counts per minute. I'm guessing that the before washing and after washing activities may have been measured twice. It seems only the first number was used for further computation, except for sample 4 after washing, in which case the last was used.
Now the number 5000 decays / minute / 150 cm^2 has been used as a threshold in the report. First, I need to be clear that this is NOT background. As shown above, background was subtracted early on. So any number substantially >0 means there is contamination above natural background.
This 5000 number appears to be an occupational safety limit. In other words, if you worked at a Soviet nuclear facility, and at the end of your shift your clothes had 6000 decays per min per 150 cm^2, that would be considered an unacceptable problem.
But that does NOT mean that, say, ending your shift with 4000 on your clothes is in any way normal! It still means your clothing is contaminated. The contamination is just not at a level that would be unacceptable for a nuclear worker. Also, that nuclear worker will not be wearing clothing contaminated at that level home! If there is a potential of contamination, then they're going to change into different clothing for their shift.
But we're talking about hikers in the woods, not workers at a nuclear facility, so anything obviously above 0 is going to be unusual. The fact that many of the results do not exceed the safety standards for a nuclear worker doesn't mean they should be considered normal.
The waistband of Kolevatov's sweater, the bottom part of Kolevatov's trousers, and Dubinina's brown sweater are all >5000. But that doesn't mean they are the only contaminated clothes. All of the 9 samples, before washing, were substantially above background. The least contaminated of the nine samples was the bottom part of Thibault Brignoles' trousers, at 600. Everything else was >1000.
So ALL of the clothing samples tested on ALL four hikers in the ravine was contaminated with beta emitting isotope(s).