Postoperative pain following posterior iliac crest bone graft harvesting in spine surgery: a prospective, randomized...

Postoperative pain at the site of bone graft harvest for posterior spine fusion is reported to occur in 6%-39% of cases. However, the area around the posterior, superior iliac spine is a frequent site of referred pain for many structures. Therefore, many postoperative spine patients may have pain in the vicinity of the posterior iliac crest that may not in fact be caused by bone graft harvesting. The literature may then overestimate the true incidence of postoperative iliac crest pain.

We performed a prospective study testing the hypothesis that patients will not report significantly higher visual analog scores over the graft harvest site when compared with the contralateral, non-harvested side.

This is a prospective, randomized cohort study.

Patients aged 18-75 years undergoing elective spinal fusion of one to two levels between L4 and S1 for spinal stenosis and spondylolisthesis were randomized to left-sided or right-sided iliac crest bone graft (ICBG) donor sites and blinded to the side of harvest.

Primary outcome was a 10-point visual analog scale (VAS) for pain over the left and right posterior superior iliac spine.

Bone graft was harvested via spinal access incisions without making a separate skin incision over the crest. Each patient's non-harvested side served as an internal control. Data points were recorded by patients on their study visit sheets preoperatively and at 6 weeks, 3 months, 6 months, and 1 year postoperatively.

Forty patients were enrolled in the study (23 females) with an average follow-up of 8.1 months (1.5-12 months). Mean age was 51.7 years (23-77 years). Left- and right-side ICBG harvesting was performed equally between the 40 patients. The average volume of graft harvested from the left was 35.3 mL (15-70 mL) and 36.1 mL (15-60 mL) from the right. There was no statistical difference between preoperative VAS score on the harvested side compared with the non-harvested side (p=.415). Postoperatively, there were consistently higher VAS scores on the operative side; however, these differences were not statistically significant at 6 weeks (p=.111), 3 months (p=.440), 6 months (p=.887), or 12 months (p=.240). Both groups did, however, show statistically significant improvements in VAS scores over time within the operative and nonoperative sides (p<.05). Graft volume had no effect on the VAS scores (p=.382).

The current literature does not adequately illuminate the incidence of postoperative pain at the site of harvest and the relative magnitude of this pain in comparison with the patient's residual low back pain. This is the first study to blind the patient to the laterality of bone graft harvesting. Our randomized investigation showed that although pain on the surgical side was slightly higher, it was neither clinically nor statistically different from the nonsurgical side. Our conclusion supports surgeons' use of autologous bone graft, which offers a cost-effective, efficacious spinal fusion supplement.