Hi Rhiannon, no need to spend any money for an apparatus. Try to get a thick wall styrofoam box from colleagues that receive cooled stuff (5 cm wide will well do it). It has to be 30 x 30 x 30 cm wide (inside). Fill with crushed ice up to 5 cm from top corners and fill with cold water to make a slush mixture. That will maintain the ice/water mix at close to 0 °C. Subject should put their hand into to wrist level. And that's all you need. We follow pain ratings at 10 s intervals  on a 0-100 numeriacal rating scale (up to cut-off time = 3 min). When hand is withdrawn if tolerance limit is reached earlier we put in 100 for the remaining time. Can be well repeated with no change in response (pain ratings or time to tolerance limit). By the way, best discrimation is reached at around 30 s in ice water. Good luck.

Fantastic! Thanks Walter. Always a bonus to find a low cost solution...! 

I use a mobicool C40 'frost-box' and a simple submersion pump (Reich, 10 L/min, 0.5 bar) running of a 12v battery. All of which are relatively inexpensive (less than £200 in total) and can be purchased online or in shops which sell camping equipment.

...you will need to keep the water in circulation during CPTesting ... this is important!

The mobicool box is thermostate controlled and with a bit of fiddling, you can get it just right where it keeps the water at 0-2 Celsius. You will need a regular thermometer to confirm the temperature. Also, until you get the thermostat setting right -- do not leave the thermometer in the water overnight as you are likely to find a block of ice in there on the first few attempts, and you don't want a cracked thermometer.

For easy cleaning, I keep the water in a plastic contained inside the mobicool box.

I place a neodynium magnet on the outside of the plastic contained to hold the submersible pump in place near the bottom (on the inside) of the plastic container ... otherwise it will start moving about once it is turned on and it will eventually get close to the surface and spray water all over the place.

I have cut a small groove in the upper edge of the mobicool box to pass the flex/wire for the pump into the box and still allow it to close and seal completely. This helps keep the temperature just right.

For VAS scale: below is a link to an online VAScale I have made, which is simple to use -- it samples the VAS score every second and can be controlled with a regular computer mouse, albeit I use a trackball mouse instead. After use it reports the time, max-score, time-to-max-score and raw data. I'd like som feedback if you find it useful: http://smerteforskning.dk/tools/freeVAS/scale.php

I have found that participants who will not tolerate CPT normally quite within 60 seconds. Most who can tolerate it for 1+ minute will tolerate it for as long as you care to keep testing.

Agree with everything Soren just wrote. We are just not circulating the water, but let the crushed ice do the job (and we have plenty of it, and will keep the temperature low while melting). You can actually combine the cold pressor test with an easy method to estimate the activation of endogenous pain control by testing  the pressure pain threshold with a hand-held algometer (increasing the pressure constantly by 50 kPa/s, which equals 0.5 kg/s on a 1 cm² surface; c.f. Rolke Eur J Pain 2006 and PAIN 2006, Magerl PAIN 2010) before and immediately (!!) after the cold water exposure (conditioned pain modulation CPM). Might be an interesting parameter to be obtained.

 You could probably also just instruct the participants to keep moving their hand back and forth in the water. It is simply to avoid a layer of heated water forming around the hand.

I have tried using a simple spring clamp to test for mechanical pain sensitivity before/after CPT and that seems to work as well. Others would suggest that the test pain stimulus should be thermal (heat), but I wouldn't expect the CPM response to be stimulus modality specific.

So I would tend to agree with Walter: what ever your test stimulus, the important thing is not to wait too long as the CPM effect can fade out quickly. I would certainly also agree that CPM very often is an interesting effect to assess (albeit also a little more complex to grasp for readers from other research fields).

 Søren , Nice to see that you have found that moving the hand does to effectively brake the shell of warmed water around the Hand. The idea of the spring also works and we had tried that before, however, since the force varies with the opening angle the thickness of the tissue (e.g. the thickness of the interdigital web or skin fold or diameter of finger) has a significant Impact. But if you just test for before vs. after this is still a valid method.

We have tested all modalities of pain (cold, heat, blunt pressure and pinprick) to assess CPM. Works with any. Pressure pain is recommended since is has by far the best reliability (see the Rolke papers, side to side correlation typically > r=0.90; before vs after correlation - with no Intervention - r>0.86), heat and cold have the least reliability !!

You can also induce it whatever you like provided that it is painful and the magniotude of pain is no significant predictor (!!! We have data collapsed over >300 subjects and the correlation of perceived pain and CPM is exactly r=0.02 !!).

By the way, I tried your VAS. Very neat, we will try to use it for teaching purposes (labcourse on pain and Inflammation). Is there any reference to it??

We use a different method using the mouse (DADISP, a shareware introduced by Brian Turnqvist, a colleague from Johns Hopkins University, but needs aparticular AD board). With this Software you do not have to touch the scale but just move and it also has speed component, i.e. if you move the mouse slowly there is a very small increase of the scale, but a big one if you move fast. This allows very high Resolution at threshold downto a 0.1 resolution on a 0-100 scale

 Hi Walter,

what you say about the spring and differences in tissue thickness is true and a weakness of that model (see: attached). It has several other limitations, but also a couple of benefits: it's very cheap, quick and safe and you can get them in all sizes and strengths. I wouldn't suggest it as a replacement for the algometer in a pain laboratory, but perhaps it has clinical application. (see: the other attached)

I am surprised you found such a weak correlation between pain intensity and CPM response. Can I read the paper?

I have used the online VAS in a number of studies, but I have not validated it separately. You also do not have to touch the scale on my online version and it is also open source. I have coded the time resolution to be 1 Hz, but that can be changed. Similarly the spatial resolution can be changed. You a welcome to have a copy of the code if you like.

 Hi all,

Going back to Rhiannon's original question and Soren's answers, I'd be grateful for some advice about the safety issues arising from using water in a frostbox connected to the electricity supply.  Rhiannon - I'd be very interested to know what you ended up doing.

Many thanks for your views.