Table 3.
Study results.
| Author | Groups | Post-op Pain | Post-op Clinical Outcomes | Results |
|---|---|---|---|---|
| Phillips 2014 | No categorization (linear modeling) | VAS-pain | NR |
Linear correlation (pre-op painDETECT and post op-VAS) VAS post-op not reported according to neuropathic pain status. No correlation between pre-op painDETEC and post-op VAS-pain (r < 0.5) PainDETECT scores did not predict dissatisfaction postoperatively (p > 0.05). |
| Kurien 2018 | Neuropathic pain (PainDETECT score ≥19) | VAS (6 months): 4 cm (range: 0.75–7) | NR |
Group comparison (VAS scores of neuropathic vs unclear/unlikely pain) The neuropathic pain group reported higher postoperative VAS pain scores at 6 months after TKR surgery compared with the unclear/unlikely group (p = 0.0003) |
| Unclear/Unlikely (PainDETECT score <19) | VAS (6 months): 0 cm (range: 0–1) | |||
|
Linear Regression (PainDETECT vs post-op pain) A linear stepwise regression, including preoperative significant associated parameters, found that preoperative PainDETECT was the only independent factor associated with postoperative pain (p = 0.006) | ||||
| Soni 2018 | Nociceptive pain group (EPIONE) | Epione Study Long term pain (VAS≥ 3 at 12 months post-op) = 6/42 (14 %) |
OKS score (2 months post-op): 39 (IQR: 29–4) OKS score (12 months post-op): 43 (IQR: 38–46) |
Logistic Regression (PainDETECT vs post-op pain) (Univariable & multivariable model) EPIONE Unclear pain = OR: 3.7 & 3.4 (p = 0.05 & 0.08) neuropathic pain = OR: 3.4 & 2.6 (p = 0.08 & 0.23) COASt Unclear pain = OR: 2.3 & 2 (p = 0.001 & 0.006) neuropathic pain = OR: 3.6 & 3 (p = 0.001 & 0.001) Group comparison (unclear vs nociceptive vs neuropathic pain post-op pain severity) Epione Cohort: The unclear pain group prior to surgery were significantly (p = 0.05) more likely to report moderate to severe long-term pain after arthroplasty at 12-months post-operatively, compared to the nociceptive group. No significant difference after adjusting for confounding factors No significant difference when a higher cut-off value was used to define moderate to severe long-term pain after arthroplasty COASt Cohort: Patients in the unclear and neuropathic groups were significantly more likely to report moderate to severe long-term pain after arthroplasty, when compared to the nociceptive group (p < 0.05). This relationship remained significant after adjusting for confounding factors and when a higher threshold for moderate to severe long-term pain after arthroplasty was used |
| Nociceptive pain group (COASt) | COAst Cohort Long term pain (NRS≥ 4 at 12 months post-op) = 53/219 (24 %) |
OKS score (12 months post-op): 42 (IQR: 35–46) | ||
| Unclear pain (EPIONE) | Epione Cohort Loong term pain (VAS≥ 3 at 12 months post-op) = 6/16 (38 %) |
OKS score (2 months post-op): 35 (IQR: 29–40) OKS score (12 months post-op): 43.5 (IQR: 35–44) |
||
| Unclear pain (COASt) | COAst Cohort (NRS≥ 4 at 12 months post-op) = 44/107 (41 %) | OKS score (12 months post-op): 39 (IQR: 29.5–44) | ||
| Neuropathic pain (EPIONE) | Epione Cohort ong term pain (VAS≥ 3 at 12 months post-op) = 5/14 (36 %) | OKS score (2 months post-op): 32 (IQR: 18–41) OKS score (12 months post-op): 39 (IQR: 32–43) |
||
| Neuropathic pain (COASt) | COAst Cohort Long term pain (NRS≥ 4 at 12 months post-op) = 29/58 (50 %) |
OKS score (12 months post-op): 37 (IQR: 25–43) | ||
| Soni 2019 | Nociceptive pain group | OKS pain subscale (12 months): 26.0 (IQR: 24.0–32.0) Moderate-to-severe long-term pain (12 months): 0 (0 %) |
OKS function (12 months): 20 (IQR: 20.0–28.6) OKS (12 months): 46 (IQR: 40.0–47.0) Patient-acceptable symptom state (12 months): 9 (90 %) HADS anxiety (12 months): 0.5 (IQR: 0.0–2.0) HADS depression (12 months): 1.0 (IQR: 0.0–3.0) STAI state anxiety (12 months): 24.0 (SD: 10.2) STAI trait anxiety (12 months): 28.0 (SD: 5.5) PCS (12 months): 5 (IQR: 0–6) PSQI (12 months): 7.8 (SD: 2.9) |
Group comparison (neuropathic vs nociceptive pain severity) The neuropathic-like pain group presented a higher proportion of patients with moderate-to-severe long-term pain after arthroplasty, compared to the nociceptive pain group (p = 0.0356). The neuropathic-like pain group presented a higher HAD anxiety and PCS score after arthroplasty, compared to the nociceptive pain group (p < 0.05). |
| Neuropathic-like + unclear pain group | OKS pain subscale (12 months): 36.0 (IQR: 20.0–52.0) Moderate-to-severe long-term pain (12 months): 4 (44 %) |
OKS function (12 months): 31.7 (IQR: 21.5–37.2) OKS (12 months): 40 (IQR: 33.0–48.0) Patient-acceptable symptom state (12 months): 5 (56 %) HADS anxiety (12 months): 3.0 (IQR: 1.0–7.0) HADS depression (12 months): 1.0 (IQR: 0.0–7.0) STAI state anxiety (12 months): 33.0 (SD: 15.8) STAI trait anxiety (12 months): 33.9 (SD: 12.7) PCS (12 months): 14 (IQR: 2–17) † PSQI (12 months): 8.7 (SD: 4.3) |
||
| Hasegawa 2019 | Unclear pain (PainDETECT score 13–18) | NRS (4.7 years): 5 Dispersion NR |
(4.7 years) Moderate to severe pain %: 85.7 Flexion angle: 121 KSS symptoms: 13.5 KSS satisfaction: 20.7 KSS functional: 46 KSS alignment: 20 KSS mediolateral/anterolateral instability: 12.4/10 Dispersion NR |
Group comparison (unclear vs nociceptive NRS and KSS scores) Mean NRS scores were higher among patients with unclear pain than with nociceptive pain. Patients with unclear pain had lower symptom scores, patient satisfaction, patient expectations, And functional activities. NRS score: p < 0.5 KSS symptoms: p < 0.5 KSS satisfaction: p < 0.5 KSS functional: p < 0.5 |
| Nociceptive pain (PainDETECT score <13) | NRS (4.7 years): 1.9 Dispersion NR |
(4.7 years) Flexion angle: 124 KSS symptoms: 20.1 KSS satisfaction: 26.8 KSS functional: 62.5 KSS alignment: 24.3 KSS mediolateral/anterolateral instability: 12.4/9.9 Dispersion NR |
||
| Larsen 2021 | No categorization (linear modeling) | VAS-pain | NR |
Linear modeling (PainDETECT vs post-op pain intensity) Positive correlation between preoperative PainDETECT and postoperative pain intensity (VAS-pain) (r = 0.298)(p = 0.001)(R2 = 0.0889) Multiple regression analysis including PainDETECT score was significant (p < 0.001) but PainDETECT score itself was not a significant independent predictor (p = 0.58) in the model. |
| Hasegawa 2021 | Possible neuropathic pain (PainDETECT score >13) | NRS (6 months): 1.5 (NR) | Flexion angle: 113 (NR) |
Group comparison (neuropathic vs nociceptive pain) No differecence in post-op NRS scores and knee flexion angle between the possbile neuropathic paiin group and nociceptive pain group (p = 0.15 and 0.94) There were no differences in preoperative pain in relation to sex, age, BMI, KL grade, and preoperative Flexion angle The preoperative presence of possible neuropathic pain might be associated with the development of persistent postoperative pain following TKA. |
| Nociceptive pain group (painDETECT score <13) | NRS (6 months): 1.2 (NR) | Flexion angle: 117 (NR) | ||
| Shemesh 2023 | PainDETECT <19 | VAS-pain (4 weeks): 7.68 (SD: 2.4) | Distance walked POD1: 82.3 (SD: 70.6) Maximal distance walked before discharge: 168.5 (SD: 101.5) Length of stay (hours): 73.6 (SD: 45.4) Discharge destination (Home): 57 (77 %) Discharge destination (Acute rehabilitation): 4 (5.4 %) Discharge destination (Rehabilitation): 13 (17.5 %) Early post-operative complications: 6 (12 %) 30-days readmission rate: 1 (1.35 %) 90-days readmission rate: 1 (1.35 %) |
Group comparison (Paindetect ≥ 19 vs < 19 clinical outcomes) The group with PainDETECT scores ≥19 showed statistically higher VAS (4 weeks) compared the <19 PainDETECT scores. (p = 0.006) The existence of neuropathic pain did affect the discharge destination, with a significantly higher proportion of patients necessitating referral to rehabilitation than patients with PainDETECT score <19 (47.3 % vs. 17.5 %, p = 0.02). Linear modeling (PainDETECT vs post-op outcomes PainDETECT was not a predictor of length of stay, discharge destination, and early postoperative complications (p > 0.05) |
| PainDETECT≥ 19 | VAS-pain (4 weeks): 9.3 (SD: 8) | Distance walked POD1: 65 (SD: 72) Maximal distance walked before discharge: 120 (SD: 83.3) Length of stay (hours): 70.8 (SD: 28.7) Discharge destination (Home): 10 (52.6 %) Discharge destination (Acute rehabilitation): 0 (0 %) Discharge destination (Rehabilitation): 9 (47.3 %) Early post-operative complications: 6 (12 %) 30-days readmission rate: 0 (0 %) 90-days readmission rate: 0 (0 %) |
||
| Vigotsky 2024 | NRS modeled as a function of PainDETECT scores | NRS (0-10) 3 months(β = 0.012, p = 0.052, pmvt = 0.162) 6 months (β = 0.024,p < 0.001,pmvt = 0.002) 12 months (β = 0.021,p = 0.004, pmvt = 0.014) |
NR |
Linear modeling (PainDETECT vs post-op pain) PainDETECT scores were significantly predictive of pain at 6 and 12 months after TKA. Preoperative neuropathic pain scores captured 30 % and 20 % of the variance in postoperative pain at 6 and 12 months, respectively. |
| Kim 2024 | Group 1: Central sensitisation and neuropathic pain | VAS pain (2 year post-op): 2.9 (SD: 1.0) | WOMAC (2 yr post-op): 42.4(SD: 14.2) WOMAC pain (2 yr post-op): 7.4 (SD: 4.1) Satisfaction = 63.6 % |
Group comparison (CS, NP, CS + NP clinical outcomes) Group 1 (central sensitisation and neuropathic pain) showed inferior WOMAC pain, function and total scores compared with the other groups (p < 0.05 for all). Groups 2 (central sensitisation) and 3 (neuropathic pain) had worse WOMAC pain, function and total scores at 2 years after surgery compared with group 4 (no CS nor neuropathic pain) (p < 0.05 for all) |
| Group 2: Central sensitisation only | VAS pain (2 year post-op): 2.8 (SD: 1.1) | WOMAC (2 yr post-op): 33.3 (SD: 14.6) WOMAC pain (2 yr post-op): 4.9 (SD: 3.5) Satisfaction = 73.5 % |
||
| Group 3: Neuropathic pain only | VAS pain (2 year post-op): 2.3 (SD: 1.6) | WOMAC (2 yr post-op): 30.1 (SD: 18.6) WOMAC pain (2 yr post-op): 4.3 (SD: 4.4) Satisfaction = 74.3 % |
||
| Group 4: Neither central sensitisation nor neuropathic pain | VAS pain (2 year post-op): 1.6 (SD: 0.9) | WOMAC (2 yr post-op): 16.4 (SD: 11.1) WOMAC pain (2 yr post-op): 2.1 (SD: 2.1) Satisfaction 88 % |
||
| Fitzsimmons 2018 | Unsuspected neuropathic pain (S-LANS <12) | ICOAP (1 month): 11.7 (SD: 8.5) ICOAP (6 months): 5 (SD: 6.1) |
PCS (1 month): 5.3 (SD: 6.2) PCS (6 months): 2.3 (SD: 4.3) PHQ-9 (1 month): 4.4 (SD: 4.1) PHQ-9 (6 months): 1.9 (SD: 4) |
Group comparison of neuropathic pain and non-neuropathic pain ICOAP, pain catostrophizing and depression scores Those with suspected neuropathic pain had higher scores for ICOAP total pain (p = 0.05), pain catastrophizing (p < 0.01), and depression (p < 0.01) at each assessment. Linear Regression (neuropathic pain identification vs ICOAP scores) After adjusting for potential confounding, pre-TKA suspected neuropathic pain did not predict ICOAP total pain or PHQ-9 depression scores at 6 months. |
| Suspected neuropathic pain (S-LANSS ≥12) | ICOAP (1 month): 22.7 (SD: 7.4) ICOAP (6 months): 19 (SD: 12.5) |
PCS (1 month): 15.4 (SD: 11.6) PCS (6 months): 13.6 (SD: 13.2) PHQ-9 (1 month): 10 (SD: 7.4) PHQ-9 (6 months): 8.1 (SD: 7.3) |
||
| Lee 2022 | Nonneuropathic pain | NR | WOMAC (1 year): 11.82 (SD: 11.28) EQ-5D health (1 year): 75 (SD: 13.86) EQ-5D (1 year): 6.37 (SD: 1.6) |
Group comparison (neuropathic vs non-neuropathic WOMAC, EQ-5D scores) There was no difference between the two groups in WOMAC, EQ-5D and EQ-5D health scores 1 year after surgery |
| Neuropathic pain | NR | WOMAC (1 year): 8.36 (SD: 9.24) EQ-5D health (1 year): 74.77 (SD: 10.74) EQ-5D (1 year): 6.18 (SD: 1.18) |
||
| Kim 2015 | Low CSI score group (<40) | VNRS (1 month): 3 (IQR: 2–5) VNRS (3 month): 2 (IQR: 2–4) VNRS≥ 5 (n%) (1 month) = 25 % VNRS≥ 5 (n%) (3 month) = 6 % Pain relief (poor/fair/good/excellent) = 1(2 %)/8(17 %)/28(59 %)/10(21 %) |
Functional improvement (poor/fair/good/excellent) 1(2 %)/6(12 %)/29(61 %)/11(23 %) Knee flexion (1 month): 90 (IQR: 80–100) Knee flexion (3 month): 120 (IQR: 110–120) |
Group comparison (low CSI pain scores post op vs high CSI pain post op) High CSI score group presented statistically higher pain scores, greater proportion of patients presenting moderate-to-severe pain (≥5 VNRS) at 1 and 3 months of follow-up, and less pain relief rates at 3 months, compared to low CSI score group (p < 0.05) Other outcomes were not statistically significant (p > 0.05) Logistic Regression (CSI vs post-op pain) In multivariate analysis, a preoperative CSI score ≥40 was the strongest determinant with 5.091 of the highest odds ratio (95 % CI 1.324 to 19.523, p = 0.016) for predicting a persistent pain 3 months after surgery among demographic and pain-related variables. |
| High CSI score group (≥40) | VNRS (1 month): 4 (IQR: 4–6) VNRS (3 month): 4 (IQR: 3–5) VNRS≥ 5 (n%) (1 month) = 47 % VNRS≥ 5 (n%) (3 month) = 27 % Pain relief (poor/fair/good/excellent) = 4(9 %)/17(38 %)/19(43 %)/5(11 %) |
Functional improvement (poor/fair/good/excellent) = 2(4 %)/7(15 %)/25(56 %)/10(22 %) Knee flexion (1 month): 90 (IQR: 80–100) Knee flexion (3 month): 120 (IQR: 100–120) |
||
| Koh 2020 | CS |
2 years post-op VAS: 2.3 Changes from preoperative status (pain VAS): 3.5 Proportion of MCID (pain VAS) (%): 47 (86 %) Proportion of persistent pain (%): 21 (38 %) Dispersion NR |
2 years post-op Flexion contracture (°): 0.3 Further flexion (°): 128.7 Hip-knee-ankle axis (°): 0.7 KSS: 165.3 Changes from preoperative status (knee Society score): 58.6 Proportion of MCID (knee Society score) (%): 45 (89 %) WOMAC score: 25.2 Changes from preoperative status (WOMAC): 29.5 Proportion of MCID (WOMAC) (%): 44 (80 %) Dispersion NR |
Group comparison (CS vs Non-CS VAS, WOMAC, KSS, satisfaction scores) The CS group presented statistically higher pain VAS, WOMAC scores, and proportion of persistent pain compared to the Non-CS group (p < 0.05). Non-CS group had statistically higher knee Society scores, changes from preoperative WOMAC status, and the proportion of patients achieving MCID for WOMAC compared to the CS group (p < 0.05). Non-CS patients were more satisfied with all kinds of activities (sitting, lying in bed, getting out of bed, light household duties, leisure recreational activities, and total overall satisfaction) than CS patients (p < 0.05) (data not extracted due to absence of table). Logistic Regression (CSI vs post-op dissatisfaction) Multivariate regression analyses revealed that preoperative CSI score, postoperative pain VAS, and WOMAC score were risk factors for dissatisfaction following TKA (P < 00.01) |
| Non-CS |
2 years post-op VAS: 1 Changes from preoperative status (pain VAS): 4.2 Proportion of MCID (pain VAS) (%): 154 (92 %) Proportion of persistent pain (%): 19 (11 %) Dispersion NR |
2 years post-op Flexion contracture (°): 0.1 Further flexion (°): 128.1 Hip-knee-ankle axis (°): 0.8 KSS: 177.6 Changes from preoperative status (knee Society score): 65.2 Proportion of MCID (knee Society score) (%): 157 (94 %) WOMAC score: 15.4 Changes from preoperative status (WOMAC): 42.1 Proportion of MCID (WOMAC) (%): 150 (90 %) Dispersion NR |
||
| Hasegawa 2022 | CSI ≥14 | NRS (6 months post-op): 1.8 (p = 0.025) |
Multivariate regression analysis of post-op knee pain Pre-op CSI ≥14 - OR: 2.5 (p = 0.438) Correlation between preop CSI and walking and standing score - (r = −0.422, p= <0.001) Correlation between pre-op CSI and Post-op NRS (r-0.311,p = 0.018) |
Group Comparison The CSI ≥14 group presented higher post-op NRS than the group with <14 in CSI (p = 0.025). Linear modeling (pre-op CSI score with pain) There was a weak positive correlation with pre-op CSI and post-op NRS (r = 0.311)(p = 0.018). Preoperative CSI correlated negatively with postoperative walking and standing score (r = −0.442)(p < 0.001). Using a multivariate analysis, CSI was not a significant factor in postoperative pain (p = 0.438). |
| CSI <14 | NRS (6 months post-op): 0.9 (p = 0.025) |
BMI: Body Mass Index; CSI: Central Sensitisation Inventory; CS: Central Sensitisation; DN4: Douleur Neuropathique 4; EQ-5D: EuroQol 5-Dimensions Questionnaire; HADS: Hospital Anxiety and Depression Scale; ICOAP: Intermittent and Constant Osteoarthritis Pain Questionnaire; IQR: Interquartile Range; KL: Kellgren-Lawrence Grade; KSS: Knee Society Score; LOS: Length of Stay; MCID: Minimal Clinically Important Difference; n: Number of Participants; NR: Not Reported; NRS: Numeric Rating Scale; OKS: Oxford Knee Score; PCS: Pain Catastrophizing Scale; PHQ-9: Patient Health Questionnaire-9; POD1: Postoperative Day 1; PSQI: Pittsburgh Sleep Quality Index; SD: Standard Deviation; S-LANSS: Self-reported Leeds Assessment of Neuropathic Symptoms and Signs Scale; TKA: Total Knee Arthroplasty; VAS: Visual Analog Scale; VNRS: Verbal Numeric Rating Scale; WOMAC: Western Ontario and McMaster Universities Osteoarthritis Index; OR: Odds Ratio.