INTRAARTICULAR KNEE TEMPERATURE CHANGES: ICE VS CRYOTHERAPY DEVICE

Kurt P. Spindler, MD, Todd A. Warren, ACNP, and Eric C. McCarty, MD

Objective:
Ice and cryotherapy devices are ubiquitous in treatment of knee injuries; however, no studies in normal or injured knees, in the absence of surgery, have evaluated if intraarticular (IA) temperature changes. This is the first evaluation of IA knee temperature comparing ice vs cryotherapy device in normal knees during and after treatment.

Methods:
After IRB approval and consent, thermocouple probes were placed via 18G needle intraarticularly in the suprapatellar pouch (SP) and on the skin of right and left knees in twelve normal subjects. Treatment was crushed ice in a plastic bag secured by plastic wrap applied to the anterior aspect of one knee with a cryotherapy device (Cryocuff“) applied to the anterior aspect of the contralateral knee. Both knees were treated for one hour, then both methods of cryotherapy were removed for the second hour. Temperatures were simultaneously recorded every minute for a total of two hours. Subjects were asked at six points during the study to indicate discomfort levels of both knees through the use of a 10cm visual analog pain scale (VAS).

Results:
As expected, ice significantly lowered skin temperature vs cryotherapy device (30, 60, 90 min p<0.001) (see Table). IA temperatures were significantly lowered from baseline in both groups with treatment. However, ice dropped the temperature significantly greater than cryotherapy device at 60*, 90*, 120** min. By VAS pain scores ice was more painful at 30*** and 60* min. Pain was inversely correlated with IA temperature in the ice group at 120 minutes (rho=-0.65, p=0.02).

Conclusions:
Both methods of cryotherapy showed significant declines in skin and intraarticular temperatures in healthy knees. Ice produced significantly lower temperatures and higher pain scores than the cryotherapy device. The increased pain scores, however, are only clinically meaningful at 60 min (difference 2.4). Further, pain was inversely correlated with IA temperature in the ice group. We hypothesize a critical pain threshold may exist within the normal knee.

*p<0.001 **p=0.013 ***p=0.01