We conducted a retrospective cohort study of patients seen by the SCKO unit, attached to the section of Rheumatology of the Hospital of Jaén. Data were obtained from electronic medical records, which collected homogeneous clinical parameters in all patients being followed. A specialist in rheumatology was trained to complete the survey and entered data into a workbook designed for electronic data collection regarding the study, which had a verification system. Data entered into the database were validated against the source data from the medical records. Patients selected to complete the study had to meet the selection criteria listed below. Included in the analysis were all cases in the database. The ethical review was requested from the Clinical Research Committee of the Hospital of Jaén study and the classification of the study from the Spanish Drug Agency, which called it a postmarketing study with other designs (EPA-OD), directed at gathering information on drugs or medical devices from the survey data. Due to the retrospective and anonymous nature of the study, no consent was obtained from patients in order to include their data in the analysis.

The criteria for selecting patients for the study were: (1) referral to the SCKO; (2) diagnosis of RA was done according to the clinical protocol based on the Consensus recommendations of the Spanish Society of Rheumatology2; (3) no other osteoarthritis mechanical problems in different joints: torsion of cartilage, ligaments affected due to trauma, severe varus or valgus deformity, terminal bone–bone disease; (4) inclusion in the study once for the involvement of the joints; (5) availability of a minimum set of data: age, sex, severity of RA (Kellgren–Lawrence), date of first visit, date of entry into the waiting list for prosthesis, date of prosthetic intervention, last known state of the patient, date of replacement of the prosthesis; (6) the patient should not have had any contraindication to the use of viscosupplementation (hyaluronic acid, NASHA: DUROLANE®) during the follow-up period in the SCKO; (7) viscosupplementation therapy in patients who received it was always with the same product: NASHA hyaluronic acid, DUROLANE®; and (8) the minimum period of follow-up in patients at the SCKO should have been at least one year.

Criteria for referral of the patient to Trauma for knee prosthesis surgery were homogeneous for all cases: (1) patients with radiographic knee arthritis grade IV, except those that due to disease or treatment were contraindicated concomitant surgery, and patients older than 75 years of age, or (2) patients with knee osteoarthritis radiographic grade III and who had disabling conditions or disease refractory to standard medical treatment (analgesics, NSAIDs, steroid injections or disease modifying drugs for OA e) provided that they were 60 years of age or older.

We collected demographic and relevant clinical data that could influence the status and prognosis of the patient as potential comfounders or effect modifiers, and data on the affected joint as potential predictors. We registered the origin of the case as being either from primary care or the trauma consultation, as well as the birth date, gender, physical activity (work, sports or non-working if the patient was unemployed), the presence of some other disease and its relationship with patient outcome or response to treatment of joint pain and medications received for the treatment of these diseases. Regarding clinical data, we recorded the date on which the patient was included in the SCKO unit for monitoring, the date of onset of symptoms of knee OA, the affected joint (left, right or both), the location of the injury (patellofemoral or femorotibial), the presence or absence of traumatic injury of the joint, the date of completion of the last arthroscopy, the presence of other local condition of the joint and the classification of osteoarthritis upon inclusion into the SCKO clinic, as well as the severity criteria of Kellgren–Lawrence (grade 0–4, from lowest to highest radiographic severity). Pain intensity was recorded at the time of incorporation into the SCKO and the last follow-up visit through a visual analog scale (VAS) measuring pain intensity on a scale of 0–100mm, where the value 0 means the absence of pain and the value 100 the worst pain imaginable. We registered the treatment that patients had received before entering the SCKO clinic and the general data of the treatments given to the patient as well as its follow-up until the most recent visit. We collected the patient's condition in his last follow-up visit with regard to vital status, if lost to follow-up, or if derived for knee replacement surgery, and the date on which these events occurred. We also collected the date of entry into the waiting list for prosthesis and the date of the intervention. Clinical data were collected from patients in the final state, in relation to radiographic severity of osteoarthritis (Kellgren–Lawrence criteria), pain intensity measured with a VAS and treatments and recommendations that the patient received at the time. We recorded whether the patient was referred from the SCKO unit for follow up to primary care or trauma if not referred to the waiting list for knee replacement.

No differences in demographic data were seen by gender, the origin of the case (primary care or trauma), or depending on whether the patient received viscosupplementation during follow up at the unit.

There were statistically significant differences (P=.005) in the time evolution of joint pain according to patient gender, being significantly higher in women, 5.7 years (95%, 5, 3, 6.2) and 47 years in men (95%, 4, 1–5.2).The mean difference was 1.1 years (95% CI, 0.3–1.9).

Patients receiving viscosupplementation during follow up were classified, in a higher proportion (51.5%), in grade 3 joint involvement according to the Kellgren–Lawrence scale than patients who did not receive viscosupplementation (18.2%), P<.0001.The proportion of patients classified as level 1 of joint involvement was significantly higher (P<.0001) in patients not receiving viscosupplementation during follow-up (36.4%) than those who received this treatment (6%).

No differences by gender or treated with or without viscosupplementation group were seen regarding the proportion of patients who were treated with postural measures (100%), analgesics and/or NSAIDs (100%), SYSADOA (1.8%), arthrocentesis (1.8%), pre-entry viscosupplementation (0.9%), or the proportion of patients referred to primary care (97.3%) or trauma (2.7%).

No significant difference in the proportion of patients in the final stage of function were seen according to whether they did or did not receive viscosupplementation during follow up (P=.258).

Patients included in the study began their follow up in the SCKO unit between 5 April 2006 and 29 April 2009. The percentage of missing data was less than 1% for most of the variables studied.

We included 224 patients whose demographic data are shown in Table 1. The mean age of the patients was 65.7 years (95% CI, 64–67) with a median of 66.1 years and a minimum age of 34 years and maximum of 89 years.

90.2% of patients (202 cases) were treated with viscosupplementation (hyaluronic acid, NASHA: DUROLANE®) at some point of SCKO follow up, and 9.8% did not receive this treatment (22 cases).

84.8% (190) of patients was receiving treatment for associated medical conditions.

52.7% of patients had both knees affected and 25% of cases had only the left knee affected.

48.2% of patients were classified as grade 3 Kellgren–Lawrence, 25.5% grade 2 and 15.3% grade 4.

The mean follow-up of patients included in the study was 374 days (95% CI, 323–425), with a median of 279 days, a minimum of 0 days and a maximum of 1547.

Forty patients (17.9%) were referred for knee replacement surgery. The mean follow-up of these patients until prosthetic surgery was performed was 328 days (95% CI, 232–424), with a median of 279 days. Twenty of the 39 patients referred for surgery were operated, with a range from the date they joined the SCKO unit of 584 days (95% CI, 487–680), with a median of 596 days. The average time from patient referral to surgery was 188 days (95% CI, 119–258), with a median of 100 days, with a minimum of 27 days and maximum of 813 days. The mean time spent on the waiting list was 165 days (95% CI, 147–184), with a median of 169 days, minimum 94 days and maximum of 343 days.

9.1% (2 cases of 22) of patients who did not receive viscosupplementation were referred for surgery and 11.9% of patients receiving viscosupplementation during follow-up in the unit were referred for surgery (38 case of 202). In total, 11.6% of patients were referred for knee replacement surgery.

We analyzed the length of stay of patients in the SCKO unit and the factors that influenced the final referral of patients for prosthetic surgery.

The factors examined were patient age at admission into SCKO, gender of the patient, the presence of comorbidity, number of joints affected, the severity of joint involvement, the time evolution of joint involvement, pain intensity on admission and viscosupplementation treatment during follow up.

We observed a median survival of 1225 days (95% CI, 942–1507), when 50% of patients were referred to the unit for prosthetic surgery. A differentiated survival curve was observed generally for patients receiving no treatment or viscosupplementation (Fig. 1). The mean time to surgery in patients who did not receive viscosupplementation was 694 days (95% CI, 548–839) and 1093 days (95% CI, 980–1206) if they received treatment with viscosupplementation.

Fig. 1.

Time from referral of the patient to knee replacement surgery, since their inclusion in the SCKO unit was based on viscosupplementation treatment.

(0.3MB).

Because the proportion of patients in each radiologic severity group was not homogeneous among patients who received treatment with viscosupplementation compared to the non-treated, further analysis was stratified by radiographic severity of osteoarthritis to determine times until the surgery depending on whether or not they were receiving treatment.

In the stratified analysis, treatment with viscosupplementation extended time until surgery in the group of patients with grade 3 joint involvement upon joining SCKO, close to statistical significance (P=.064). The median survival of patients with grade 3 lesions and viscosupplementation was 1278 days (95%, 474–2081) and for those not receiving treatment it was 596 days (95% CI, 14–1179).

No differences in time to referral to surgery were overall associated to gender (P=.741). When performing a stratified analysis according to the severity of the joint involved and the gender of the patient, we observed that the time until surgery was significantly higher in women with grade 3 lesions who received viscosupplementation (P=.029). There were no significant differences between genders in the other degrees of joint involvement.

We explored whether age at admission influenced the time to patient referral for surgery and to that end we divided the patients into higher or equal to 65 and less than 65 years of age groups. There was a longer time to surgery in patients older than 65 years (P=.024), with a median survival in this group of 1278 days (95% CI, 600–1956), and, in patients younger than 65 years, of 945 days (95% CI, 769–1121). See the survival curves in Fig. 2. No differences between patients in the different strata were seen based on whether they were receiving treatment with viscosupplementation or not.

Fig. 2.

Time from referral of the patient to knee replacement surgery, since their inclusion in the SCKO unit depended on the age of the patient.

(0.35MB).

We analyzed if patients with one or both affected joints were derived earlier to surgery and found that patients with one affected joint (47.3%) were derived earlier than those with 2 joints affected (P<.0001). The median survival of patients with only one joint affected was 778 days (95% CI, 622–934), and patients with two affected joints did not reach the median survival (Fig. 3).

Fig. 3.

Time from referral of the patient to knee replacement surgery, since their inclusion in the SCKO unit depended on the number of joints affected.

(0.32MB).

When performing an analysis stratified by the two previous variables, we found that the time until surgery was higher in patients with two affected joints, whether the patients were younger than 65 years (P=.002) or if they were over 65 years (P=.008).

Multivariate analysis was performed using Cox regression to assess the variables related to the time before referring the patient to surgery jointly (Table 2). The gender of the patient, the presence of comorbidity, number of joints affected, the age at which the patient was admitted to the SCKO unit, the severity of joint involvement at the last follow-up visit, the time of disease progression, the intensity of pain at the last follow up visit and treatment with viscosupplementation into the equation were included. The following were statistically significant: (a) the number of affected joints (P<.0001), with less time to surgery if only one joint was affected (OR=0.267, 95% CI, 0.130–0.549), (b) the severity of osteoarthritis (P=.048), higher severity, shorter time to surgery (OR=0.410, 95%, 0.169–0.992), and (c) pain intensity measured on the visual analog scale (P<.0001), with an OR of 1.091 (95% CI, 1.044–1.141), higher pain intensity, longer time to surgery.

The authors declare that procedures conformed to the ethical standards of the committee responsible for human experimentation and were in accordance with the World Medical Association Declaration of Helsinki.

The authors declare that they have followed the protocols of their workplace on the publication of data from patients and all patients included in the study have received sufficient information and gave their written informed consent to participate in this study.

The authors have obtained informed consent from patients and/or subjects referred to in the article. This document is in the possession of the corresponding author.