At first we noted that when rats were undisturbed and sitting quietly, they held their incised paw off of the floor of the cage, tending not to rest it on the plastic mesh. We graded this guarding behavior by observing the position of the incised paw on the mesh for a one-hour period and compared it to the contralateral unincised hindpaw . Before surgery, there was little difference in the position of the hindpaws. Immediately after surgery, the guarding behavior was greatest; this behavior gradually subsided over 3–4 days (Fig. 1A). It was eliminated by hindpaw denervation, local anesthetic infiltration and intrathecal morphine administration. Because the intensity was greatest immediately after surgery, the nociceptive responses decreased each day and resolved on postoperative day 3 or 4. This behavior has some similarities to the time course of pain at rest in patients treated with patient-controlled analgesia .
In order to understand the role of skin injury to guarding pain (Fig. 1A), we developed the method to incise skin only and limit the damage to the underlying fascia [31, 32]. We attempted to minimize muscle injury. Very little guarding pain could be elicited in rats that underwent skin incision only. Any guarding resolved by postoperative day 1, whereas those animals that had deep tissue injury (incision to skin, fascia, and muscle) had marked guarding. These results indicate that skin incision is insufficient to elicit unprovoked guarding pain in the plantar incision model. We hypothesized that skin incision would produce minimal ongoing activity in nociceptive pathways.
We examined dorsal horn neurons for spontaneous activity before and immediately after plantar incision [25, 28, 36]. Incision increased spontaneous activity of approximately 40 to 50% of dorsal horn neurons that received input from the plantar region [25, 28, 36]. In general, the amount of spontaneous activity was extremely low, averaging 1 to 2 imp/sec 1 hour after incision. This small amount of activity occurs at a time when guarding pain is greatest and we hypothesized spontaneous activity should be high. We suggested that examining the dorsal horn neurons in normal skin largely selects for neurons that receive predominately cutaneous primary afferent input . Thus, only low levels of spontaneous activity would be expected after incision in normal skin.
Similarly, incision in normal skin did not increase spontaneous activity in nociceptive primary afferent fibers . Our data, in addition to the work of Perl and colleagues [3, 7] and that of Campbell and colleagues , support the notion that incision of skin produces very little spontaneous activity of primary afferent fibers innervating the skin.
Additional electrophysiology experiments support the concept that cutaneous injury elicits a little spontaneous activity in dorsal horn neurons (Fig. 2A-C) and primary afferent fibers. We recorded dorsal horn neurons from rats that underwent sham incision and from rats that underwent skin incision only . Thirty percent of dorsal horn neurons had spontaneous activity in sham-operated animals and 1 day after skin incision, approximately 50% of dorsal horn neurons had spontaneous activity (Fig. 2B). There was no difference between sham and skin incision in the percentage of dorsal horn neurons with spontaneous activity and average rate of spontaneous activity. In contrast, a greater proportion, approximately 80%, of dorsal horn neurons from rats that had undergone an incision that included deep tissue injury had spontaneous activity. The average rate of spontaneous activity tended to be greater in rats that underwent deep tissue incision (Fig. 2C).
We also recorded primary afferent fibers from rats that underwent sham incision and from rats that underwent skin incision only (Fig. 3A-C) . Thirteen percent of afferents had spontaneous activity in sham-operated animals and 1 day after skin incision, approximately 17% of nociceptive fibers that had spontaneous activity (Fig. 3B). There was no difference between sham and skin incision in the percentage of afferents with spontaneous activity and average rate of spontaneous activity. In contrast, a greater proportion, approximately 80%, of nociceptors from rats that had undergone an incision that included deep tissue injury, had spontaneous activity. Although there was much variability, the average rate of spontaneous activity tended to be greater in afferents from rats that underwent deep tissue incision (Fig. 3C).
Two human studies also indicate a very low amount of ongoing pain after skin incision. Kawamata et al.  made a forearm incision in human volunteers. Significant spontaneous pain was only present for 30 minutes after the forearm incision. One hour later, there was no difference between a local anesthetic-treated, control group and the group that underwent forearm skin incision. Fimer et al.  also measured ongoing pain after a forearm incision. On a scale of 1–100, pain ratings were less than 1 two minutes after the incision and pain ratings remained less than 1 for three hours and thereafter.
Reduced mechanical withdrawal responses are easily elicited after hindpaw incision . The reduced mechanical threshold is greatest immediately after incision and sustained for up to 7 to 10 days afterwards . These mechanical responses are decreased by hindpaw denervation , local anesthetic infiltration  and parenteral and intrathecal morphine  and thus are pain related.
Experiments evaluating mechanical withdrawal threshold after a sham incision, skin incision only, or skin, fascia, and muscle incision were undertaken to better understand the role of injury to these tissues in this behavior . While sham incision produced no reduced withdrawal threshold, incision of either skin only or skin plus deep tissue incision produced similar withdrawal thresholds for 2–3 days afterwards (Fig. 1B). The duration of the mechanical hyperalgesia was greater when skin plus deep tissue incision were made. These data indicate that early after incision, skin injury alone is sufficient to produce decreased mechanical withdrawal thresholds, which is not different than that which includes deep tissue injury.
The studies in dorsal horn neurons tend to separate those neurons which are capable of encoding exaggerated mechanical responses from those dorsal horn neurons which develop high spontaneous activity. We had shown in 2002  that when incision was made in the receptive fields of dorsal horn neurons, 7 of 10 neurons gained a small amount of spontaneous activity and most of them were less than 2 imp/sec. These 10 WDR neurons had a marked increase in mechanical responses 1 hour after incision.
Further studies indicated that reduced mechanical withdrawal thresholds were signaled by WDR neurons that did not have spontaneous activity after incision . One day after incision, WDR neurons could be separated into two groups; those without spontaneous activity and markedly coded for decreased mechanical withdrawal threshold and those with spontaneous activity that did not respond well to these mechanical stimuli.
Data on primary afferents are in agreement that mechanical hyperalgesia can be transmitted by incision to the receptive fields of cutaneous primary afferent fibers . Aδ and C fiber thresholds were recorded from rats that had undergone sham control surgery, skin incision or skin, fascia, and muscle incision one day earlier (Fig. 4A-B). Of note, the number of fibers is small in some groups. Thresholds were lower in rats from both the skin incision group and the skin, fascia, and muscle incision group compared to fibers from rats that recovered from incision. When mechanical hyperalgesia resolved, the Aδ and C fiber thresholds were similar to the sham control group.
Studies by Kawamata et al.  in the forearm skin incision produced very similar withdrawal thresholds to those seen in our rat model. Very high mechanical pain thresholds, greater than 150 mN, were present before incision. These pain thresholds were reduced to less than 50 mN for several hours after forearm incision. The mechanical pain threshold gradually recovered by the 3rd or 4th day. These data are very similar to those generated in rat by skin incision only further supporting that skin incision is sufficient to produce the reduced mechanical withdrawal threshold in the early period after injury.
Punctate mechanical testing can also be used to measure the area of hyperalgesia after incisional injury. Indeed, after plantar incision, reduced withdrawal threshold is present in undamaged tissue approximately 1 cm from the incision . Several studies have examined the area of cutaneous sensitivity and wound hyperalgesia  to understand plasticity, central sensitization and persistent pain [10, 21]. This area of hyperalgesia in clinical postoperative pain appears to have commonalities to some aspects of plasticity that has been extensively characterized in preclinical studies like the area of the receptive field of dorsal horn neurons. The area of wound hyperalgesia may be decreased by specific analgesic treatments, but inconsistently associated with reduction of clinical pain [13, 14].
Although patients may complain of burning pain after surgery, we did not examine heat withdrawal responses after hindpaw incision because we were uncertain of the role of heat responses in clinical postoperative pain. Many investigators were concerned that we had not studied heat withdrawal latency; these studies were eventually completed several years after developing the model. We discovered that heat withdrawal latencies were markedly reduced after hindpaw incision; the decreased latency was greatest immediately after incision and gradually increased back to baseline within 7 to 10 days . The heat withdrawal latency is increased by low doses of parenteral morphine,  parenteral capsaicin receptor antagonists,  nerve growth factor blockade  and in capsaicin receptor knockout mice .