There are five trials [9, 14–17] that compare surgical with conservative interventions, either being conservative management (three trials), prolonged conservative management with optional surgery (one trial), or steroid injections (one trial). Unfortunately, these trials are heterogeneous regarding interventions, some have a high risk of bias, and some suffer from poor reporting which prohibits quantitative analysis [18]. The general conclusion after evaluation of these trials does not support either surgery or nonoperative intervention [18]. However, surgery appeared to result in faster recovery in at least two of the trials. This implies that the choice for surgery over continuation of conservative care comprises balancing fast recovery against surgical risks, cost, and burden.

The three studies [14–16] that compared surgery with conservative management showed conflicting results. One older, high risk of bias study including 126 patients [14] demonstrated that diskectomy was significantly better regarding patient and observer ratings than conservative treatment at 1 year. Twenty-four of the 66 patients (36 %) in the conservative care group versus 39 of the 60 patients (65 %) in the surgery group reported a good outcome. This difference disappeared after 4 and 10 years. One high risk of bias trial including 56 patients [15] found no significant differences for leg pain or back pain and subjective disability throughout the 2 years of follow-up. VAS leg pain scores, however, improved more rapidly in the diskectomy group; 6-week score in the surgery group was 12 (SD 20) versus 25 (SD 27) in the conservative group. The per-protocol analysis demonstrated no statistically significant differences. A large low risk of bias trial including 501 patients showed that both the surgery and the conservative treatment group improved substantially over 2 years for all primary and secondary outcome measures [16]. The intention-to-treat analysis showed no statistically significant differences for any of the primary outcome measures. There was considerable crossover: 50 % of the patients randomized to surgery and 30 % of the patients randomized to conservative treatment. After 2 years this was 45 and 40 %.

The sciatica trial randomized 283 patients with severe sciatica for 6–12 weeks to early surgery or prolonged conservative treatment followed by surgery if needed [9]. In this study, crossover was anticipated and offered surgery to patients that did not improve after 6 months. After 2 weeks, 89 % of patients randomized to early surgery underwent micro-diskectomy, while 39 % of patients randomized to conservative treatment underwent surgery after a mean of 19 weeks. Relief of leg pain was faster for patients assigned to early surgery. Intention-to-treat analysis showed statistically significant more leg pain relief was found in favour of early surgery compared with prolonged conservative care at 3 months (MD −17.70, 95 % CI −23.1 to −12.3). There was no significant overall difference between the two groups regarding disability scores during the first year. The median time to recovery was 4.0 weeks (95 % confidence interval (CI), 3.7–4.4) for early surgery and 12.1 weeks (95 % CI, 9.5–14.9) for prolonged conservative treatment. During the first year, early surgery achieved a faster rate of perceived recovery with a hazard ratio of 1.97 (95 % CI 1.72–2.22, P < 0.001). At 1 year of follow-up, however, 95 % of patients in both treatment groups had experienced satisfactory recovery, and no subsequent differences were found. This lack of a difference between groups was maintained after 2 years and also after 5 years [19]. The same pattern was found for the subset of 150 patients with a motor deficit [20]. Motor deficit recovered significantly faster in patients randomized to early surgery, but there were no differences found after 1 year.

One high risk of bias trial including 100 patients compared micro-diskectomy with epidural steroid injection [17]. Patients undergoing diskectomy had the most rapid decrease in their symptoms. The decrease in leg pain in the diskectomy group was significantly greater than epidural steroid injection group at 3- and 6-month follow-up intervals, but not beyond 1 year. There were no significant differences between groups for back pain throughout the follow-up. Twenty-seven of 50 patients receiving a steroid injection had a subsequent micro-diskectomy. Outcomes in this crossover group were similar to those of the surgery group.

The current status of evidence does not support a definite choice for conservative or surgery, at least for the indications that were studied in these trials. The general consensus from guidelines is to wait for at least 6–8 weeks before considering surgery [5]. Also, some of the choosing wisely initiatives provide definitive advice not to proceed with surgery too early but also not to wait too long (more than 9 months) [21]. Between these margins, choice for either surgery or conservative intervention should be based on preferences of well-informed patients (Fig. 17.1). Only if the patients have information about the advantages and disadvantages of surgery on the short term and the equivalent outlook of both interventions on the longer term, an informed choice can be made [9]. Decision tools can guide these decisions [22].

The most common type of surgery is microscopic diskectomy, which is defined as the surgical removal of part of the disk, performed with the use of an operating microscope or other magnifying tools. Most studies refer to Caspar [23], Yasargil [24], and Williams [25] when diskectomy is performed with microscope and to Foley and Smith [26] or Greiner-Perth et al. [27] when diskectomy is performed with tubular, muscle-splitting, retractor systems and endoscope. However, some have returned to using a microscope while retaining the less invasive muscle-splitting approach of Foley and Smith [26]. The result is an array of surgical approaches for which it is difficult to acquire sufficient evidence from randomized trials comparing all techniques.

There are eight trials that have compared open diskectomy with minimal invasive techniques, including microscopic diskectomy, video-assisted microscopic diskectomy, automated percutaneous micro-diskectomy, or micro-endoscopic diskectomy.

Six trials compared the classical open diskectomy, also called standard diskectomy or macro-diskectomy, with microscopic diskectomy [28–33]. There is a consistent finding in these studies that microscopic diskectomy leads to an increased operating time with a pooled effect of 12 min (95 % CI 2.20–22.3; p = 0.02; moderate quality of evidence). No differences were found for length of stay, which was only reported in five studies with a total of 452 patients. The mean difference was 0.18 days in favour of open diskectomy (95 % CI −0.09 to +0.45 days; p = 0.47; moderate quality of evidence). Blood loss was reported in two studies; in one study with 119 patients, microscopic diskectomy resulted in less blood loss [31], while in the other study with 60 patients, there was no difference [28]. The quality of evidence for blood loss was “very low”. The length of incision was reported in three studies with together 353 patients and found to be shorter for microscopic diskectomy in two studies [30, 33]. The quality of evidence for incision was “low”. Leg pain was reported in four studies with together 453 patients and was significantly less for microscopic diskectomy by 2.01 mm (95 % CI 0.57–3.44; p = 0.006; moderate quality of evidence), while this can hardly be regarded as a clinical relevant difference. Further outcomes (pain, return to work) were found to be comparable, except for a higher return to work at 4 weeks for microscopic diskectomy [33] in one study with 114 patients where two other studies with together 140 patients found no difference at 10.4 weeks [28] and 14.9 months [29]. It should be noted that all but one of these trials was associated with a high risk of bias.

Two trials compared open diskectomy with micro-endoscopic diskectomy [32, 34]. Huang et al. [34] reported results of a very small, high risk of bias, trial with only 22 patients. The micro-endoscopic diskectomy group had shorter postoperative hospital stay and less intraoperative blood loss compared with the open diskectomy group, but duration of the operation was longer. There were no differences in pain severity and MacNab criteria between the groups. Teli et al. [32] showed in a larger trial including 220 patients that the micro-endoscopic group compared to open and microscopic diskectomy suffered more dural tears (7 %, 3 %, 3 %, respectively), root injuries (3 %, 0 %, 0 %, respectively), and a recurrent herniation (7 %, 4 %, 3 %, respectively).

One low risk of bias trial with 60 patients found that patients who had received video-assisted arthroscopic micro-diskectomy had similar satisfactory outcomes compared with open laminotomy and diskectomy, but patients who had had an arthroscopic micro-diskectomy had a shorter duration of postoperative disability and used narcotics for a shorter period [35].

There is evidence on the comparative effectiveness of the different minimal invasive techniques for diskectomy such as endoscopic diskectomy, video-assisted diskectomy, percutaneous transforaminal diskectomy, etc.

Eight trials with an accumulative 1047 patients evaluated different approaches for less invasive diskectomy, such as micro-endoscopic diskectomy, tubular microscopic diskectomy, microscopic-assisted percutaneous nucleotomy, minimal access trocar/microsurgical micro-diskectomy, percutaneous endoscopic diskectomy, or sequestrectomy. We analysed the comparisons between these techniques, keeping the differences muscle damage and differences in use of microscope or endoscope in mind. The results of these trials are given in Table 17.1.

Seven (six high risk of bias) trials with 923 patients compared tubular diskectomy with conventional microscopic diskectomy [32, 36–41]. Of these, four used an endoscope [32, 36, 37, 39]. There was low to moderate quality of evidence for incision length and this was consistently shorter for tubular diskectomy in all three studies (n = 260) that reported this outcome [32, 36, 39]. However, results could not be pooled due to sparse data on variation (SD). The quality of evidence for the remaining outcome parameters was “low’ to “very low”, so no further meta-analyses could be performed. Inconsistent results were found for operative morbidity. Two studies (n = 368) of the six studies (n = 718) reporting operative time found a longer duration for tubular diskectomy [32, 36], while one study (n = 100) found a shorter duration [38]. No differences were found for blood loss in three studies. Length of stay was longer (2 h) for conventional microscopic diskectomy in only one of four studies [36]. One study found a faster improvement in pain scores for tubular diskectomy before discharge [37], while the only low risk of bias study found a slightly better pain score for conventional diskectomy at 2 years [41]. All other outcomes for pain as measured with VAS, for Oswestry or Roland-Morris score, or for SF36 scores were not significantly different between the two surgical techniques. For Shin et al. [37], baseline values for back pain were not comparable. In one trial, the postoperative analgesic consumption was significantly less in the tubular diskectomy group [40].

One high risk of bias trial [42] with 40 patients compared percutaneous endoscopic diskectomy (cannula inserted into the central disk) with microscopic diskectomy. This trial showed comparable clinical outcomes after the two procedures but contained a small sample size.

Evidence regarding techniques that are applied for the prevention of scar tissue is relatively sparse. Recent trials of an interposition gel covering the dural sheet, fat, preservation of the ligamentum flavum, and use of a drain show promising effects in reducing epidural scar formation, but no effect on clinical outcomes.

Thirteen studies considered the effect of different techniques to prevent formation of intra-spinal scarring following diskectomy, as assessed by magnetic resonance imaging or enhanced computerized tomography. Ten studies evaluated the use of an interposition membrane. The types of membrane used are autologous free fat graft or commercially available gels. The results of these trials are given in Table 17.1.

Four high risk of bias studies compared the use of fat graft versus no fat graft [43–46]. These studies failed to show any improvement in clinical outcomes following use of fat. Three studies evaluated fibrous tissue formation on CT or MRI, two found a decrease for fat graft [44, 46], and one small subsample of MacKay et al. [43] found no difference. The pooled effect with a moderate quality of evidence yielded a significant decrease in scar tissue for fat graft (OR 0.22 (95 % CI 0.08–0.62). One study reported a lesser number of painful episodes 1 year after surgery [46], but this was evaluated by the surgeon.