**4. Discussion**

Overall, we observed a significant improvement in ODI and leg and back VAS at the last follow-up. CB11 analysis highlighted a high level of satisfaction after surgery, confirming the results of previous studies, which reported positive outcomes after surgical therapy for adult spine deformity [21,22].

The correlation between ODI and age, BMI, or ASA was moderate at the one-year follow-up, but the strength of these correlations was reduced at the two-year follow-up. The correlation between leg and back VAS and age, BMI, and ASA showed similar trends to those observed for the ODI: back pain weakly correlated with age and BMI before surgery and at the 1 year follow-up, but no significant correlation was observed at 2 years, or with ASA at any follow-up. Leg pain showed a weak-to-moderate correlation with all parameters and at all follow-up, except with ASA at the last follow-up. Similar trends were also observed for CB11. These data confirmed that older age and poorer overall health condition may have a moderate negative impact on the level of complications and disability or pain after surgery [23–25], but this negative influence dissipates over time. Thus, these patients can also expect positive outcomes after long spine fusion [26–29], but have to be adequately informed that a poorer preoperative health status correlates with longer recovery time. Surgeons, however, need to consider that obesity and age or comorbidities have a relevant impact on intraoperative blood loss, length of surgery, and complication rate; thus, preoperative BMI and ASA should still be considered when planning long spine fusion [30–32].

Length of surgery, estimated blood loss, and length of hospital stay showed no or only weak correlation with ODI, VAS, and CB11. This aspect is also key for the informed consent of the patients and their attitude toward the recovery process, as a prolonged hospital stay does not have a negative impact on the long-term outcomes of surgery.

Analyzing the correlation of ODI, VAS, or CB with the extent of the instrumentation, we found that the level of the UIV did not affect any of the outcomes of interest. Given the relative limited mobility of the thoracic spine [33], these data are not surprising. It is however striking that the moderate correlation between ODI and LIV at the one-year follow-up was further reduced at the two-year follow-up. Similar results were obtained in other studies observing different PROMs and the ability of patients to perform determined activities after spinal fusion: over time, a gradual ODI improvement could be observed even in patients with fusion to the pelvis [10,34]. The explanation for this finding may lie in the postoperative movement restrictions required by many surgeons after fusion (e.g., avoiding forward bending or heavy lifting), which then ease over time, or in the fact that patients adapt to the movement restrictions imposed by the instrumentation and develop strategies to overcome them. This topic requires further investigation: if the developing of these strategies is the key in reducing postoperative disability after spine fusion, specific pre- or postoperative physiotherapy programs may be implemented to support patients and improve their quality of life after surgery.

Overall, the ODI, VAS, and CB parameters showed multiple moderate and strong correlations amongs<sup>t</sup> each other, confirming how different aspects of a patient's health, quality of life, and satisfaction regarding treatment are interconnected [35]. Regarding the ODI, a strong correlation was observed between pre- and postoperative disability levels; this suggests that patients starting with high ODI values have lower chances of achieving a low ODI postoperatively. This represents a key factor in planning the timing of surgery. Different to what was observed for the ODI, the preoperative VAS only weakly to moderately associated with levels of back and pain level at the two-year follow-up. Thus, even patients with a high preoperative pain level can expect an improvement with respect to the painful symptoms two years after surgery. Unsurprisingly, the level of satisfaction with the treatment (CB11) correlated with ODI and VAS both at the one- and two-year follow-ups. However, while the correlation with pain level was of moderate intensity and declined at the two-year follow-up, the correlation to disability was strong at both followups. A similar correlation between patients' satisfaction and PROMs was also observed by another study group [35].

This study is not without limitations, the main one being its retrospective nature. The relationship between ODI, pain, and satisfaction with treatment and pre- and perioperative data proved to be a complex, and further research on a wider patient cohort will be required to investigate it. Furthermore, the patients in our cohort presented different types of instrumentations (e.g., different types or levels of interbody implants) and deformity correction techniques. While it was not possible to investigate the effect of different surgical techniques on the outcome of interest due to the limited number of observations, this topic deserves further analysis in the future.
