Table 3.

Quality assessment of the studies included using the QUADAS-2 tool [8]

Study	Risk of bias			Applicability concerns
	Patient selection	Biopsy	Flow and timing	Patient selection	Biopsy
Ahuja and Sharma [10]	High	Low	Unclear	Unclear	Low
Terreaux et al. [11]	High	Low	Unclear	Unclear	Low
Gras et al. [12]	High	Low	Unclear	Unclear	Low
Kim et al. [13]	High	Low	Unclear	Unclear	Low
Gasbarrini et al. [14]	High	Low	Unclear	Unclear	Low
Lora-Tamayo et al. [15]	High	Low	Unclear	Unclear	Low
De Lucas et al. [16]	High	Low	Unclear	Unclear	Low
Friedman et al. [17]	Low	Unclear	Unclear	Unclear	Unclear

The following signaling questions were used to assess the risk of bias and applicability concerns (which were then scored as high risk, low risk, or unclear):

Risk of bias:

1. Patient selection. Did most patients with negative initial biopsy cultures undergo a repeat biopsy? Was it reported why patients were selected for repeat biopsy?

2. Biopsy. Could the conduct or interpretation of biopsy have introduced bias?

3. Flow and timing. Was MRI performed within 2 months before tissue biopsy? Was the repeat biopsy performed within 1 month of the initial biopsy and was no therapy administered between the initial and repeat biopsies?

Applicability concerns:

4. Patient selection. Were patients with a previous history of spondylodiscitis excluded? Were patients with positive blood cultures before biopsy excluded? Was MRI performed before biopsy and were the criteria for positivity reported? Which patients underwent a repeat biopsy after a negative initial biopsy?

5. Biopsy. Was fluoroscopic or CT guidance used? What needle size was used? How many biopsy samples were acquired?