Relative and absolute test-retest reliabilities of pressure pain threshold in patients with knee osteoarthritis

2.1 Participants

Twenty KOA patients, ranging from 59 to 80 years participated in this study (Table 1). The KOA patients were recruited from a centralized database at the Centre for Clinical and Basic Research (CCBR) and diagnosed according to American College of Rheumatology classification. Clinical data was collected from patients, including radiological evaluation, medication and pain duration. The inclusion criteria were aged 18–80 years, clinically diagnosed with knee OA, Kellgren-Lawrence X-ray grade ≥2, pain while walking according to the patient and BMI <35. No subjects had used any analgesics within the previous 24 h prior to the experiment. The subjects were given a detailed written and verbal explanation and they signed an informed consent form before participating in the study. The study had been approved by the local Ethics Committee (VN-20160081) and conducted according to the Declaration of Helsinki.

2.2 Experimental protocol

The experiment consisted of two sessions was conducted at CCBR. Session 1 was followed 2 weeks±3 days later by session 2. All the subjects were asked to rate the maximum intensity of pain on a 0–10 cm horizontal visual analog scale (VAS) ranging from 0 to 10 for the worst pain from the most painful knee during the past 24 h and after 5 min rest prior to the PPT recordings. The VAS was anchored with 0=no pain and 10=worst pain imaginable, and the distance (cm) of the subject’s mark on the horizontal line was the outcome measure. The most painful knee was selected for assessment for all the patients. The subjects were lying comfortably in a supine position. A handheld algometer (type II, Somedic AB, Sweden) with a tip area of 1 cm² was used to assess PPTs from ten anatomical locations in a randomized order. All assessments were made by the same examiner. A mechanical pressure stimulus with an ascending pressure gradient of 30 kPa/s was applied continuously until the subjects felt pain and pressed a stop-button. The pressure indicated on the algometer display corresponding to the PPT where noted by the examiner. The PPT recordings were performed three times and the mean value was used for statistical analysis. To avoid temporal integration [9], [25], a 1-min interval was observed between the PPT assessments. To make sure the subject was confident with the method, a test PPT measurement was tested prior to recordings on the bulk of the tibialis anterior on the contralateral leg. Additionally, each subject completed a knee-directed The Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) survey and a PainDetect survey. WOMAC is a standardized subject-rated instrument for measuring pain severity and pain interference scores [26]. The WOMAC questionnaire consists of 24 questions focusing on pain (score range 0–20), stiffness (score range 0–8), and functional limitations (score range 0–68) of the OA knee during the past 48 h [26].

PainDetect is a questionnaire to assess the neuropathic components of chronic pain conditions [27]. A total score of ≤12 indicate a neuropathic pain component is unlikely, and a total score of ≥19 indicate that a neuropathic pain component is likely. A total score of 13–18 are uncertain [27].

2.3 Pressure pain threshold

Eight PPT sites were marked over the knee in the peripatellar region and two control points according to our previous study [8]. The eight test points are defined as (1) 2 cm distal to the inferomedial edge of patella; (2) 2 cm distal to the inferolateral edge of patella; (3) 3 cm lateral to the center of the lateral edge of patella; (4) 2 cm proximal to the superolateral edge of patella; (5) 2 cm proximal to the superior edge of patella; (6) 2 cm proximal to the superomedial edge of patella; (7) 3 cm medial to the center of the medial edge of patella; (8) on the center of patella. The two extra segmental sites were (9) the belly of the tibialis anterior muscle and (10) in the center of brachioradialis ipsilateral to the selected knee (Fig. 1). Points 9 and 10 enabled to assess from external sites to evaluate generalized sensitization.

Fig. 1:

Location of the assessed eight pressure pain threshold (PPT) points on the knee (A) and the two control points (B and C).

PPT maps were generated by interpolating mean PPT values using an inverse distance-weighted interpolation to compute PPT values of the unknown locations. This was obtained by using mean scores from the known PPT values and locations [8], [9].

2.4 Data analysis

The data were analyzed using a repeated measure of variance (ANOVA) with the factor sites (10 test sites) and sessions (two sessions). A student t-test for pair-wise comparisons was used as post hoc test to compare PPT’s between the sessions, in case of significant main effect. A paired sample t-test was performed to test for differences in VAS between the sessions and for differences in WOMAC sub scores between the sessions. The relative and absolute reliability across the sessions were computed using intra-class correlation coefficients (ICC_2,1 for absolute agreement), SEM and MDC. ICC were interpreted according to Landis and Koch in which an ICC between 0.00 and 0.20 is considered “poor”, 0.21–0.40 is “fair”, 0.41–0.60 is “moderate”, 0.61–0.80 is “substantial”, and 0.81–1.00 is “almost perfect” [28]. The inter-class correlation coefficient was calculated with the coefficient of variance (CV) and reflects the overall variability of the model. CV is expressed in % and is calculated as:

CV = SD μ × 100 %

SEM represents measurement error, which means it represent if a patient’s measurements vary if the test is repeated without any underlying change in the patient. SD is the standard deviation of the PPT from patients in both sessions and the ICC are the relative reliability. SEM has the same unit as pain sensitivity (kPa) and are calculated with the following formula [29]:

SEM = SD 1 − ICC

MDC indicates the minimum value for which a difference can be considered as real. The formula for MDC is the following [30]:

MDC = SEM  × 1 .96  × 2

Furthermore, Bland Altman plot of differences between sessions against their mean and limits of agreement (LOA) were used to assess the disagreement between the PPT’s in the two sessions. A real change can be considered, in case a difference between sessions are seen outside LOA [29]. The Spearmans correlation test between PainDetect and WOMAC sub-scores and PainDetect and PPT’s for both sessions were analyzed. All the statistical analysis was performed using the statistical program software IBM SPSS version 25, and p<0.05 was considered significant.

3.1 Intra- and inter-session reliability

The intra-session reliability indicators at each PPT location are reported in Table 2 and inter-session reliability are reported in Table 3. The ICCs of the 10 anatomical locations were “almost perfect” according to the definition [28] ranging from 0.85 to 0.96. The inter-individual variation (CV) was smaller in session 1 (44.2–57.6%) than session 2 (47.9–62.7%). SEM values ranged from 34 to 71 kPa and MDC ranged from 93 to 196 kPa and the mean PPT ranged from 273.5 to 367.7 kPa in session 1 and 268.1–331.3 kPa in session 2. The analysis of the Bland-Altman plot showed that no apparent systematic bias was present in the data and that zero was included in the 95% confidence interval. Upper-lower LOA were 216 and −156 kPa while the bias was 30 kPa. Differences between the sessions can be seen outside LOA, which indicates a real change (Fig. 2).

Fig. 2:

Bland and Altman analysis plotted for the differences of patient’s pressure pain threshold (PPT) in the eight locations around the knee for the mean of session 1 and session 2. Solid line represents mean difference and the dashed lines the limits of agreement (LOA). Upper LOA, 216 kPa; lower LOA, −156 kPa; bias, 30 kPa. PPT locations around the knee: 1: , 2: , 3: , 4: , 5: , 6: , 7: , 8: .

3.2 Pressure pain threshold

PPT maps showed that the patients were most sensitive at the location of superolateral edge of patella (PPT 4), superomedial edge of patella (PPT 6) and medial to the center of the medial edge of patella (PPT 7) (Fig. 3). The PPT’s from the 10 anatomical locations were not significantly different from session 1 to session 2 (p=0.074) though lower PPT was seen in session 2 (Table 2). Likewise, the VAS scores for the past 24 h and after resting in 5 min were not significantly different between the sessions (Fig. 4).

Fig. 3:

Anterior view of mean pressure pain threshold (PPT) maps from the eight anatomical locations on the knee (kPa) in session 1 (A) and session 2 (B). No significant difference was seen between the sessions (n=20).

Fig. 4:

Mean±SD visual analog scale (VAS) values for the maximum pain in the most painful knee in the past 24 h and after resting for 5 min. X-axis indicates the session and Y-axis indicates VAS score (0–10).

No significant correlations were found between PainDetect and WOMAC for pain, stiffness, function or total score and between PainDetect and PPTs when evaluating data from session 1.

3.3 Pain questionnaire surveys

The PainDetect questionnaire was calculated with a maximum possible score of 38 and a minimum score of −1. At screening, 11 subjects had a score ≤12, and only two had a score ≥19 in session 1 and in session 2, 13 subjects had a score ≤12, and two subjects had a score ≥19 indicating that most of the subjects had pain, but unlikely to have neuropathic component. For the WOMAC sub scores, a significant difference was seen for stiffness between the two sessions (p=0.002), but no difference in total score for pain or physical function were seen between the two sessions (Table 4).