It can be stated that the radiographic classification of knee arthritis is based either on the degree of joint space narrowing, as with the Ahlbäck grading
, or on the presence of osteophytes, such as the classification proposed by Kellgren and Lawrence
. Some authors defend the latter
, arguing that marginal osteophytes are the most sensitive radiographic feature for the detection of osteoarthritis, more than joint space narrowing, subchondral sclerosis, and subchondral cysts. Both classifications show a good correlation
 and the Ahlbäck system seems easy to apply and suitable for the assessment of medial compartment arthritis of the knee, the most frequently affected compartment in knee arthritis; thus, it appears to be particularly useful for the orthopedic treatment of knee disorders
Although the relationship between radiographic knee OA and the need for TKA is not yet fully established, severe osteoarthritis and range of motion restriction eventually necessitate surgical intervention. It should be mentioned that in our study more than 18% of the patients in the waiting list for TKA had OA of the knee grade I according to the Ahlbäck classification
, and in grade I cases, the radiographic images were close to normality. On the other hand, if we take into account the HSS score at the start of the study, 112 patients showed a score >70 points, which translates as a mild clinical and functional involvement. Thus, the first finding of our study was the ascertainment that there is a group of patients showing no relevant radiographic findings which are included into the waiting list for TKA implantation. It is possible that other factors such as the social environment, a suboptimal communication between the patient and the physician, the difficulty of assessing pain or the lack of other resolutive therapeutic measures may modify the demand for TKA.
We have observed a good correlation between Ahlbäck grade
 and the HSS: whenever the former increases, the HSS score decreases. We haven't found in the literature any publications which may support or contradict this concordance. There are several studies that do not find any relationship between knee pain and radiographic changes. Larsson et al.
 state it is difficult to correlate chronic knee pain, the diagnosis of radiographic OA, and functional capacity. In this article, the proportion of patients with knee pain found to have radiographic OA ranged from 15–76%, whereas in the case of those with radiographic knee OA the proportion with pain ranged from 15–81%. Considerable variation occurred with x-ray view, pain definition, OA grading and demographic factors. Bedson and Croft
 mention that the results of knee x-rays should not be used in isolation when assessing individual patients with knee pain, and conclude that knee pain is an imprecise marker of radiographic knee osteoarthritis, although the more severe the radiographic osteoarthritis, the more likely there are to be accompanying symptoms.
The HSS scale we used does not assess only pain, but also function and range of motion. Although it cannot be described as a "quality of life scale", it does analyze some activities (such as stair climbing, the need of assistive walking devices, etc.) which are part of daily life. On the other hand, we have observed that there is only a very small difference in HSS score between Ahlbäck grades I and V (only 9.56 points). This would support the idea that the HSS scale does not discriminate radiologic involvement, meaning that a scarce number of radiographic lesions, or small ones, may nevertheless alter the clinical and functional status of the patient. The American Knee Society Score (AKSS) has been validated and is responsive and reproducible. However, it also suffers from high inter- and intraobserver variation when the assessments are performed by less experienced doctors and nurses
. Even though we could not get the SF-12 questionnaires from all the patients in our series, we observed that there was no relationship between the grade of radiographic involvement and the mental health component scale scores of the SF-12. Nor did we find such a relationship with the physical component scale, except if we grouped patients with grades I-IV versus patients with grade V according to the Ahlbäck classification. By performing this association, we did find a statistically significant difference. According to these findings, it seems that there is no relationship between radiographic classification and quality of life with the exception of the physical appearance of patients with more advanced radiographic arthritis.
We have not found differences in HSS scores between Ahlbäck grades when taking into consideration the age of patients. As for gender, differences were only present between men and women with grade III. We did not found any explanation for this finding. In the study by Muraki et al.
 knee pain was strongly associated with joint space narrowing, especially in men, while women tended to experience knee pain even without radiographic OA. The prevalence of knee pain was age-dependent in women, but not in men.
When comparing the status of patients who entered the study in 1994 and in 2010, we observed that nowadays patients are added to the TKA waiting list with lesser Ahlbäck grades and that their age has increased. The reasons behind these changes may have to do with an earlier indication for TKA both on the part of the surgeon and the patient, since it is known this procedure leads to positive results. The greater mean age may be due to the higher survival rate of the patients. In any case, both findings would require a more specific analysis.
Our study has some limitations. Its multicentric nature has the advantage of providing an overview of the actual clinical practice in hospitals and of a heterogeneous group of surgeons, although this may also become a disadvantage, for it could alter the homogeneity of the results and introduce bias in the measurement criteria. Even though all the participating surgeons were familiar with both the Ahlbäck classification
 and the HSS scale and used them on a regular basis, there is no way of guaranteeing the data have been properly collected. In our series, the radiographic studies were carried out homogeneously with full extension of the knee and resting limb. Even though some degree of flexion is recommended, since this is a multicentric study involving many surgeons, full extension seemed to be easier to reproduce and standardize. Since all radiographs were taken in this position, the overall results should not be compromised nor altered. If the study had been focused on knee OA progression, it would have been indispensable to perform the radiographic study with the knee slightly flexed
Knee OA assessment and grading is far from solved. We did find a relationship between Ahlbäck grades and the preoperative clinical score. Nevertheless, the range of variability of the HSS score between the different Ahlbäck grades is small, a fact that reduces its usefulness in clinical practice. Besides the scales analyzed in this study, other techniques are needed to discriminate the arthritic knee involvement taking into account the commonness of this process and the socioeconomic imbrications therapeutic decisions have on these patients.