In recent years, concern has increased regarding the importance of cervical sagittal parameters in clinical prognoses [1, 4, 5]. Therefore, accurately and rapidly evaluating cervical sagittal balance requires good cervical parameters.
We found that T1s, C2–7 lordosis and C2–7 SVA were the three most important sagittal parameters in the cervical spine. A study by Kim et al.  showed that patients with a large T1s required increased cervical lordosis and increased energy consumption in the upper cervical spine to maintain their head weight. A study by Oe et al.  showed that when T1s > 40°, the probability of failure was higher for deformity corrections performed in the cervical vertebra. Many studies have also noted that maintaining surgical intervention for C2–7 lordosis has a positive effect on a patient’s prognosis, perhaps because less energy is consumed by the neck muscles and ligaments [11,12,13]. Tang et al.  also suggested that when C2–7 SVA > 40 mm, NDI scores will be worse.
In this study, from Table 1, we could conclude that the surgery mainly changed the SA (p = 0.000) and K-line tilt (p = 0.000), but by analyzing the correlations among Δ values (Table 2), we concluded that surgery indirectly changed other sagittal parameters by directly changing the SA and K-line tilt. These data clearly indicated that ACDF surgery maintains global lordosis mainly by increasing regional lordosis.
Among the indicators measured, the correlation between C2–7 lordosis and NDI scores was not obviously significant. The T1 vertebral body is not a direct surgical segment because it cannot be directly changed by surgery but must instead be indirectly changed. These changes were relatively small. The good cervical sagittal observation index can evaluate the curvature of the entire cervical vertebrae. As the most active cervical vertebra, it is impossible to display the complete cervical vertebrae by simply evaluating T1s, but the K-line tilt (a line connecting the centers of the C2 and C7 spinal canals) can better evaluate sagittal balance. Furthermore, the T1s is not easy to measure because the thoracic spine is difficult to correctly identify on a simple lateral X-ray film due to anatomical interference from the shoulder contour density, especially in obese people with thick thoraces . Although T1s minus C2–7 lordosis (T1s-CL) is also a good cervical parameter, it is more difficult to measure and use. The K-line can be measured by identifying and connecting the two midpoints of the C2 and C7 spinal canals. Therefore, K-line tilt detection is simple and convenient and can be performed quickly and intuitively.
Of the indicators evaluated in our study, we found that both K-line tilt and C2–7 SVA were substantially correlated with NDI scores (r = 0.832 vs 0.756), and K-line tilt was positively correlated with C2–7 SVA (r = 0.707, p = 0.008) and T1s (r = 0.501, p = 0.036). During the operation, the angle is more intuitive than the length because there is no need to consider the effect of scale. Therefore, we believe that K-line tilt may provide a more advantageous assessment method. Linear regression analysis was used to analyze the relationship between K-line tilt and NDI scores, and the results indicated that when the follow-up K-line tilt was > 23.75°, the NDI scores were > 25 (R2 = 0.737, p = 0.000). In this regard, we believe that if the patient’s K-line tilt is larger before surgery, it may be considered to increase the size of the cage (5, 6, and 7 mm are commonly used) during surgery and to correct the regional lordosis, which will yield a better prognosis for the patient.
This study only included a small number of sample statistics. Moreover, this study was retrospective, and therefore, some unintended biases could exist, such as selection bias and information bias. NDI scores were used to evaluate the clinical prognosis and as a quality of life index, and the JOA scores, SF-36, and VAS scores were not measured. We chose adjacent two-level ACDF because the single-segment surgery (single-segment ACCF or disc replacement, etc.) resulted in little change in sagittal parameters before and after surgery. There were more patients with two-level ACDF in our hospital. In principle, if there are sufficient segments to perform multi-segment ACDF, this approach is best. In this study, we only had one experienced doctor to measure the data, which inevitably leads to errors. We hope that a follow-up study will enhance and verify the value of this article by increasing the sample size and improving scoring standards.