The electronic search identified 1312 studies from databases and conference abstracts. Despite our thorough search, no additional records were identified through the manual reference search. This might be due to the comprehensive nature of the initial database search, which captured the relevant studies available at the time. Additionally, the field of interest might have a limited number of eligible studies, further reducing the likelihood of finding unpublished or unindexed records. After duplicate removal, 625 papers were screened. Among them, 607 papers were excluded as not English or Chinese language, not RCTs, preclinical papers, or different topics. At the end of the selection process, thirteen studies fulfilled the inclusion criteria and were included in the meta-analysis [14, 26,27,28,29,30,31,32,33,34,35,36,37]. The Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) flow chart summarizing the process of selection is shown in Fig. 1.

Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) flow chart of the selection process

The baseline characteristics of thirteen included RCTs are shown in Table 1. These studies are published between 1997 and 2023, and the total sample size is 938. Among the included RCTs, Ketorolac is administered at the dose of 15 mg [26, 27, 30], 20 mg [32], 30 mg [14, 27,28,29,30,31, 33,34,35,36,37]. The routes of administration include intravenous injection [14, 26, 27, 29, 33,34,35,36,37], intramuscular injection [28, 30], intraoperative local infiltration [31], and oral administration [32]. Its adjunctive drugs include bupivacaine [31], morphine [31], epinephrine [31], paracetamol [31] and pregabalin [31]. The timing of administration includes preoperative [29, 32, 36], postoperative [14, 26,27,28, 33, 34, 37], and intraoperative periods [31, 35].

Among the thirteen included RCTs, seven studies report pain scores at 0–6 h [14, 27, 29,30,31, 36, 37], eight studies report pain scores at 6–12 h [27, 29, 31, 36, 37], six studies report length of hospital stay [26, 27, 31,32,33,34], ten studies report postoperative morphine requirements [14, 26,27,28,29,30,31,32,33, 36], eight studies report nausea [28, 35,36,37], six studies report vomiting [29, 30, 32, 35,36,37], four studies report pruritus [28,29,30, 35], and three studies report constipation [28, 29, 32]. Jadad scores of the four included studies vary from 3 to 5, and all thirteen studies have high-quality based on the quality assessment.

Due to differences in race, administration methods, and timing of administration across the studies, the random-effect model was used for the analysis of postoperative pain scores. In this systematic review, all studies used the same scale for the primary outcome, such as the Visual Analog Scale (VAS) for pain, making the use of mean difference (MD) generally more intuitive and meaningful.

The results indicate that, compared to the control group, Ketorolac is associated with significantly lower postoperative pain scores within the first 0–6 h following lumbar spinal surgery. This is demonstrated by a mean difference (MD) of -1.42 on the pain scale, with a 95% confidence interval (CI) ranging from − 2.03 to − 0.80, indicating a statistically significant reduction in pain (P < 0.0001). The studies show moderate heterogeneity with an I2 value of 65% (P < 0.01, Fig. 2), suggesting some variability in the effect sizes across the included studies. The reduction of − 1.42 points on the VAS represents a notable improvement in pain relief. To provide context, the MCID for VAS pain scores is typically around 1.2 to 2.0 points. The observed reduction exceeds the lower end of this range, suggesting that the pain relief provided by Ketorolac is not only statistically significant but also clinically meaningful. However, it is important to note that there is moderate heterogeneity among the studies, with an I2 value of 65% (P < 0.01), indicating some variability in the effect sizes across the included studies.