KNEE
The natural history of pain and neuropathic pain after knee replacement A PROSPECTIVE COHORT STUDY OF THE POINT PREVALENCE OF PAIN AND NEUROPATHIC PAIN TO A MINIMUM THREE-YEAR FOLLOW-UP J. R. A. Phillips, B. Hopwood, C. Arthur, R. Stroud, A. D Toms From Exeter Knee Reconstruction Unit, Princess Elizabeth Orthopaedic Centre, Royal Devon and Exeter Hospital, Exeter, United Kingdom
A small proportion of patients have persistent pain after total knee replacement (TKR). The primary aim of this study was to record the prevalence of pain after TKR at specific intervals post-operatively and to ascertain the impact of neuropathic pain. The secondary aim was to establish any predictive factors that could be used to identify patients who were likely to have high levels of pain or neuropathic pain after TKR. A total of 96 patients were included in the study. Their mean age was 71 years (48 to 89); 54 (56%) were female. The mean follow-up was 46 months (39 to 51). Pre-operative demographic details were recorded including a Visual Analogue Score (VAS) for pain, the Hospital Anxiety and Depression score as well as the painDETECT score for neuropathic pain. Functional outcome was assessed using the Oxford Knee score The mean pre-operative VAS was 5.8 (1 to 10); and it improved significantly at all time periods post-operatively (p < 0.001): (from 4.5 at day three to five (1 to 10), 3.2 at six weeks (0 to 9), 2.4 at three months (0 to 7), 2.0 at six months (0 to 9), 1.7 at nine months (0 to 9), 1.5 at one year (0 to 8) and 2.0 at mean 46 months (0 to 10)). There was a high correlation (r > 0.7; p < 0.001) between the mean VAS scores for pain and the mean painDETECT scores at three months, one year and three years post-operatively. There was no correlation between the pre-operative scores and any post-operative scores at any time point. We report the prevalence of pain and neuropathic pain at various intervals up to three years after TKR. Neuropathic pain is an underestimated problem in patients with pain after TKR. It peaks at between six weeks and three-months post-operatively. However, from these data we were unable to predict which patients are most likely to be affected. Cite this article: Bone Joint J 2014;96-B:1227–33.
J. R. A. Phillips, MBChB, BSc, FRCS (Tr&Orth), Knee Fellow B. Hopwood, MBBS, MRCA, MSc, Associate Specialist Anaesthetics And Pain Management C. Arthur, MB ChB, FRCS (Tr&Orth), Knee Fellow R. Stroud, MCSP, Extended Scope Physiotherapist A. D Toms, MB ChB, MSc, FRCS (Tr&Orth), Consultant Orthopaedic Surgeon Royal Devon and Exeter Hospital, Exeter Knee Reconstruction Unit, Barrack Road, Exeter, UK. Correspondence should be sent to Mr J. R. A. Phillips; e-mail: [email protected] ©2014 The British Editorial Society of Bone & Joint Surgery doi:10.1302/0301-620X.96B9. 33756 $2.00 Bone Joint J 2014;96-B:1227–33. Received 2 January 2014; Accepted after revision 06 June 2014
The incidence of chronic pain and disability after total knee replacement (TKR) is reported to be between 8% and 34%.1,2 National Joint Registry data have revealed rates of dissatisfaction one year after TKR of more than 18%.3,4 Many causes of chronic pain after TKR have been described.5 Efforts have been made to improve the outcome by technological developments such as the use of navigation and custom-made cutting blocks,6-9 as well as methodology, such as by using kinematic rather than mechanical alignment.10,11 However, there still appears to be a group of patients with a technically satisfactory TKR and without evidence of infection who remain dissatisfied. Dissatisfaction is clearly multifactorial but a contributing factor is pain.12 However, revision has been found to have limited success when performed for pain alone.13 Neuropathic pain has been described after TKR. Buvanendran et al14 identified a rate of neuropathic pain of 5% six months postoperatively in a prospective series of
VOL. 96-B, No. 9, SEPTEMBER 2014
120 patients after TKR. Wylde et al2 identified rates of neuropathic pain of 6% three to four years post-operatively in a retrospective study of 632 patients after TKR. Why rates of neuropathic pain are so high has not been explained. Neuropathic pain is defined as “pain arising as a direct consequence of a lesion or disease affecting the somatosensory system”.15 Common findings in neuropathic pain are a sensory deficit or hypersensitivity in the painful area and a gradual increase in pain following repetitive stimulation, and paroxysms of pain.16 Patients with neuropathic pain report having pain of longer duration and have greater interference from pain and significantly poorer quality of life than those with nociceptive pain.17 Neuropathic pain does not usually respond to simple analgesics and specific management strategies are needed.18 There is a lack of evidence about the natural history of pain and, specifically, neuropathic pain following TKR. Most studies on 1227
1228
J. R. A. PHILLIPS, B. HOPWOOD, C. ARTHUR, R. STROUD, A. D TOMS
Table I. Pre-operative scores Pre-operative score (n = 96)
Mean
Range
VAS pain 5.8 1 to 10 painDETECT 4.9 0 to 20 HADS anxiety 3.8 0 to 17 HADS depression 5.4 1 to 14 Oxford Knee Score 20 7 to 45 VAS, visual analogue scale; HADS, Hospital Anxiety and Depression Score
pain focus on the early post-operative period with few continuing beyond three months after TKR.19-21 We therefore designed a prospective study involving a cohort of patients undergoing TKR for osteoarthritis evaluating pain and any neuropathic element to it, at regular intervals, for a minimum of three years post-operatively. The primary aim was to establish the prevalence of pain at specific intervals post-operatively, and to ascertain the impact of neuropathic pain. A secondary aim was to establish any predictive factors that could be used to identify which patients were likely to experience high levels of pain or neuropathic pain after TKR.
Patients and Methods A total of 96 patients with primary osteoarthritis of the knee were recruited to the study between July 2009 and May 2010. These were consecutive patients seen in a physiotherapist-led pre-operative education class, of whom 54 (56%) were female. Their mean age was 70.6 years (48 to 89) and all underwent primary TKR. Baseline details included demographics, co-morbidities and the use of medication including anti-depressants and anti-neuropathics were recorded. Pain was assessed using a 0 to 10 Visual Analogue Scale (VAS); this is a well evaluated tool.22 The 11-point Likert scale uses zero to represent no pain and ten to represent the worst possible pain. Improvements of > 30% (i.e. three points on the scale) are thought to be significant.23 The VAS has also been interpreted as either no pain (VAS = 0), mild (VAS = 1 to 3), moderate (VAS = 4 to 6) or severe (VAS = 7 to 10).24 Neuropathic pain was self-assessed using the painDETECT questionnaire.25 This has also been evaluated in a mixed pain clinic population, and we felt that was a better tool than others such as the self-administered version of the Leeds Assessment of Neuropathic Signs and Symptoms (S-LANSS),26 as it allowed patients to grade the severity of their symptoms and a number could be recorded which could be correlated with a VAS pain score. A painDETECT score 0 to 12 implies neuropathic pain is unlikely, 13 to 18 implies neuropathic pain is unclear (possible) and 19 to 38 means neuropathic pain is likely to be present. Pre-operative anxiety and depression was evaluated using the HADS assessment tool.27 Scores < 7 points are negative, 8 to 10 to be possible and > 11 points indicates anxiety and/ or depression. Functional outcome was assessed using the Oxford Knee Score (OKS; 0 to 48 worst to best).28
painDETECT > 19 n = 1 > 11 n = 6 > 11 n = 7
Patients were reviewed in hospital on days three to five post-operatively and were contacted by post at six weeks, three months, six months, nine months, 12 months and finally at a minimum of three years post-operatively. Those with high pain scores in the early post-operative period were referred to the acute pain service. Patients were also asked to make free comments which could be used in the interpretation of the results. At one year and at final follow-up, patients were asked their interpretation of the success of their operation (excellent, very good, good, fair or poor), and their satisfaction with surgery (delighted, pleased, mostly satisfied, mixed, mostly dissatisfied, unhappy or terrible). In addition, further surgery or interventions were recorded. The use of medication was recorded at each stage to aid the interpretation of results but not as a surrogate marker as it would be unreliable due to the many factors that affect the use of medication including intolerance and inefficacy. Those who failed to respond were followed-up with a telephone call but if not successful data were omitted for that period. All clinical interventions, whether related or not to the TKR, were recorded. The study had ethical approval and consent was granted by Pfizer (Tadworth, Surrey, United Kingdom) to use their pre-printed painDETECT forms. Statistical analysis. This was performed using GraphPad Prism version 6.00 for Windows, (GraphPad Software, La Jolla, California). Paired results were compared using paired Wilcoxon tests (non-parametric data). Correlation was determined using the Spearman correlation test for non-parametric data. R > 0.7 is deemed to be highly correlated, and r = 0.5 to 0.7 moderately correlated. A p-value of < 0.05 was considered to be statistically significant.
Results The mean follow-up was 46 months (39 to 51). The rate for completion of the documentation was high, with 89 patients (94%; one patient had died) responding at 12 months, and 80 (92%; seven patients had died) responding at final follow-up, at which time 19 patients (20%) had undergone contralateral TKR. Pre-operative scores are shown in Table I. Pain scores improved up to the time of final follow-up (p < 0.001 for all data points compared with the pre-operative VAS for pain; Wilcoxon matched-pairs signed rank test; Fig. 1). The prevalence of pain at all points is shown in Table II and Figure 2. THE BONE & JOINT JOURNAL
THE NATURAL HISTORY OF PAIN AND NEUROPATHIC PAIN AFTER KNEE REPLACEMENT
1229
VAS scale
10
5
0 Day 3 wks 3 mths 6 mths 9 mths 1 yr Preoperative 3 to 5
4 yrs
Time since surgery Fig. 1 Plot graph showing pain according to the mean visual analogue scale (VAS) with 95% confidence intervals at each time interval after knee replacement.
Table II. Mean visual analogue score (VAS) for pain at each review
VAS Mean Range
Pre-operative
Days three to five
Six weeks
Three months
Six months
Nine months
12 months
Final follow-up mean 46 months
5.8 1 to 10
4.5 1 to 10
3.2 0 to 9
2.4 0 to 7
2.0 0 to 9
1.7 0 to 9
1.5 0 to 8
2.0 0 to 10
Pre-operatively, the painDETECT scores revealed a prevalence of possible neuropathic pain in five patients (5%) and likely neuropathic pain in only one patient (Table III). Neuropathic pain became most prevalent six weeks postoperatively with 23 of 85 respondents (27%) having possible neuropathic pain, and seven (8%) having likely neuropathic pain. Three months post-operatively, 16 of 82 respondents (17%) had possible and two (3%) had likely neuropathic pain. After peaking between six weeks and three months, the proportion of patients having neuropathic pain then reduced to eight of 87 respondents (9%) with possible and two (2%) with likely neuropathic pain one year post-operatively. At final follow-up, six of 79 respondents (7%) had possible neuropathic pain and five (6%) had likely neuropathic pain. The changes in painDETECT scores are shown in Figures 3 and 4. The intensity of neuropathic pain varied with time. Of the seven patients with likely neuropathic pain six weeks post-operatively, only one had likely neuropathic pain at final follow-up (with one lost to follow-up). Of the five patients at final follow-up with likely neuropathic pain, only one had this pre-operatively, two had it six-weeks post-operatively (with no data collected for one patient at this time) and one had it three months post-operatively. Of the five patients with possible neuropathic pain preoperatively, one had developed likely neuropathic pain one year post-operatively (one was lost to follow-up). At VOL. 96-B, No. 9, SEPTEMBER 2014
three years post-operatively one new patient had developed likely neuropathic pain, two were lost to follow up and the other patient with neuropathic pain had died. The single patient with likely neuropathic pain preoperatively had high scores suggestive of neuropathic pain at every stage and at 40 months post-operatively. Preoperatively she had no anxiety or depression, her VAS for pain was 10/10 and she was already on treatment for neuropathic pain. She had high levels of pain post-operatively and was treated with opiate patches, pregabalin and amitryptiline. She underwent two manipulations under anaesthesia and at the latest follow-up at 40 months she had a VAS pain score of 8/10, an OKS score of 17 and a painDETECT score of 21. Despite this, she thought the outcome of her TKR was ‘fair’ and she had ‘mixed’ satisfaction levels. Patients with high painDETECT scores at final followup underwent secondary review of their medical records and radiographs. This revealed no features of malalignment or incorrect sizing that could explain the symptoms. Patient satisfaction. Satisfaction one year post-operatively and at final follow-up is shown in Figures 5 and 6. At final follow-up, 11 of 79 respondents (14%) had mixed feelings about their TKR or were dissatisfied. In all, ten of this group had VAS pain scores of > 7 on walking. Two had possible neuropathic pain and three had likely neuropathic pain. Of the 11 dissatisfied patients, six had undergone further surgery and one is awaiting arthroscopy to debride
1230
J. R. A. PHILLIPS, B. HOPWOOD, C. ARTHUR, R. STROUD, A. D TOMS
Vas score (%)
100
No pain
90
Mild
80
Moderate
70
Severe
60 50 40 30 20 10 0 PreDay 3 to 5 operative
6 wks
3 mths
6 mths
9 mths
1 yr
3 yrs
Time since surgery Fig. 2 Bar chart showing the prevalence and severity of pain before and after knee replacement using the visual analogue score (VAS). (0: no pain, 7 to 10: severe pain).
Table III. The numbers of patients according to the ranges of the visual analogue score (VAS) for pain and painDETECT scores for neuropathic pain at each review
Pre-operative (%) Day three post-op (%) Six weeks post-op (%) Three months post-op (%) Six months post-op (%) Nine months post-op (%) 12 months post-op (%) Mean 46 months post-op (%)
n=
PAIN none PAIN mild VAS = 0 VAS = 1 to 3
PAIN moderate PAIN severe VAS = 4 to 6 VAS = 7 to 10
n=
painDETECT painDETECT painDETECT 0 to 12 negative 13 to 18 possible 19 to 38 likely
96 74 85 83 90 86 86 80
0 3 4 (5) 15 (18) 29 (32) 28 (32) 35 (41) 36 (45)
39 (41) 35 (47) 33 (39) 21 (21) 14 (16) 14 (16) 12 (14) 12 (15)
96 74 85 82 90 86 87 79
90 (94) 68 (92) 55 (65) 63 (77) 72 (80) 70 (81) 77 (89) 68 (87)
9 (9) 22 (34) 40 (47) 42 (51) 42 (47) 41 (48) 36 (42) 26 (33)
patellofemoral scar tissue, causing crepitus. No significant factors were identified pre-operatively that predicted dissatisfaction at final assessment. Of the five patients with likely neuropathic pain at final follow-up, two were not satisfied with their operation, one had mixed feelings and two were satisfied. Of the five patients with VAS pain scores of > 7 at final follow-up, two were not satisfied with their operation, one reported a mixed outcome and two were satisfied. A total of ten patients had high anxiety or depression scores pre-operatively (HADS > 11); three had high scores for both anxiety and depression; four had high scores for depression alone and three for anxiety alone. One patient died and two others were lost to follow-up. Only one patient, with depression, developed likely neuropathic pain at final follow-up; all other patients had painDETECT scores of < 12, all had VAS pain scores of < 5, and all but one patient had better than ‘mixed’ satisfaction levels.
48 (50) 14 (19) 8 (9) 5 (10) 5 (6) 3 (4) 3 (3) 6 (7)
5 (5) 4 (5) 23 (27) 16 (19) 16 (17) 11 (13) 8 (9) 6 (7)
1 (1) 2 (3) 7 (8) 3 (4) 2 (3) 5 (6) 2 (2) 5 (6)
There was no correlation between the pre-operative VAS pain scores and post-operative painDETECT scores at any stage (r < 0.5). Similarly, there was no correlation between the pre-operative painDETECT scores and the postoperative VAS pain scores at any stage (r < 0.5). However, at other intervals there were either moderate or high correlations between the VAS pain scores and the painDeTECT scores at the same times post-operatively. VAS pain and painDETECT scores correlated at three months (Spearman correlation r = 0.74, 95% confidence intervals (CI) 0.61 to 0.82), nine months (r = 0.76, 95% CI 0.65 to 0.84), 12 months (r = 0.72, 95% CI 0.59 to 0.81) and at final follow-up (r = 0.76, 95% CI 0.65 to 0.85). The painDETECT scores at six weeks and six months postoperatively correlated only moderately with VAS pain scores at final follow-up (r = 0.53 and r = 0.50). At final follow-up, a high (negative) correlation was identified between the OKS and the VAS pain (r = -0.86, 95% CI -0.78 to -0.91), and painDETECT (r = -0.77, 95% THE BONE & JOINT JOURNAL
THE NATURAL HISTORY OF PAIN AND NEUROPATHIC PAIN AFTER KNEE REPLACEMENT
1231
20
Delighted
15 Patient satisfaction
painDETECT score
100
10
5
90
Pleased
80
Mostly satisfied mixed
70 60 50
Mostly dissatisfied unhappy
40 30 20
Terrible
10 0
0 Day Preoperative 3 to 5
3 mths 6 mths 9 mths 1 yr
6 wks
1 yr
4 yrs
3 yrs Time since surgery
Time since surgery Fig. 5 Graph showing mean painDETECT scores pre- and post-operatively, with 95% confidence intervals.
0 to 12 Neuropathic pain unlikely
100 pain DETECT score (%)
90 80 70
13 to 18 Neuropathic pain possible
60 50 40 30 20
> 19 Neuropathic pain likely
10 0 Pre- Day 3 to 5 6 wks 3 mths 6 mths 9 mths 1 yr operative Time since surgery
3 yrs
Bar chart showing satisfaction one and three years after knee replacement.
100
Patient interpretation of success (%)
Fig. 3
80 60 40 20 0 1 yr Excellent
3 yrs Very good
Good
Fair
Poor
Time since surgery
Fig. 4
Fig. 6
Bar chart showing prevalence of neuropathic pain pre- and postoperatively using the painDETECT score.
Bar chart showing the patients’ interpretation of the level of success one and three years after knee replacement.
CI -0.85 to -0.65) scores. There was no correlation between the pre-operative HADS scores and any other scores. Complications and further surgery. A total of 11 patients (11%) had a complication or underwent further surgery. Two patients underwent revision surgery, one for instability and one for deep infection; three patients sustained traumatic extensor mechanism injuries (one open avulsion of the patellar tendon, one avulsion of the quadriceps tendon and one patellar fracture, treated non-operatively); three patients underwent manipulation under anaesthesia within four months post-operatively; two underwent arthroscopic arthrolysis for stiffness and one underwent secondary patellar resurfacing. The mean VAS pain score at final follow-up in these patients, two of whom were lost to follow-up was 4.3 (1.3 to 10).The mean painDETECT score was 11.7 (2 to 21) and two had scores of > 18. Six of the remaining ten patients reported a ‘fair’ or ‘poor’ outcome, and seven reported their satisfaction as being ‘mixed’ or ‘worse’. VOL. 96-B, No. 9, SEPTEMBER 2014
Discussion This study confirms that there is a significant decrease in pain after TKR. However, for some, pain persists at three years post-operatively; with 12 of 80 patients (15%) having moderate pain and six of 80 (7%) having severe pain at a mean of 46 months. This is consistent with the findings of Wylde et al2 who retrospectively identified 632 patients who had undergone TKR and sent out questionnaires between three and four years post-operatively, achieving a 73% response rate. Using the Western Ontario and McMaster Universities Arthritis Index score,29 17% had moderate persistent pain and 15% had severe persistent pain. We found significant improvements of pain after TKR at all points (p < 0.001) and levels of satisfaction that are consistent with United Kingdom patient reported outcome measures (PROMS).3,30 Previous studies by Wylde et al2 and Buvenendran et al14 found levels of neuropathic pain to be 5% at six months
1232
J. R. A. PHILLIPS, B. HOPWOOD, C. ARTHUR, R. STROUD, A. D TOMS
and 6% between three and four years post-operatively. Using the painDETECT score, we found that the rate of neuropathic pain was at its highest six-weeks postoperatively, with 8% having likely, and 27% having possible neuropathic pain. This rate then decreased and 3% had likely and 17% had possible neuropathic pain six months post-operatively. At final follow-up, 6% had likely and 7% had possible neuropathic pain. Neuropathic pain adversely affects the quality of life.17 We have demonstrated that potentially up to 35% of patients have symptoms of neuropathic pain six weeks after TKR. The question arises whether these patients should receive specific treatment for these symptoms. It is possible that many of the symptoms experienced in the early postoperative period may be neuropathic in origin as part of the normal, inflammatoy process. Significant symptoms can be treated in accordance with the National Institute for Health and Clinical Excellence (NICE) guidelines for the pharmacological management of neuropathic pain in adults.31 Amitriptyline or pregabalin are recommended as first line treatment, but when symptoms cannot be controlled, early referral to a specialist pain service is suggested.31 This study has identified a correlation between pain as assessed using a VAS and neuropathic pain as assessed using the painDETECT score. There was a high correlation (r > 0.7) between these two at three months, one year and a mean 46 months after operation. Pre-operative screening with painDETECT scores is unlikely to be useful in the absence of a correlation between the pre-operative scores and those at a minimum of three years. However, a moderate correlation was identified between the painDETECT scores at six weeks and three months and those at final follow-up. Thus, early intervention in these patients might be helpful. There was no correlation between any pre-operative scores (VAS, painDETECT or HADS), and the postoperative scores. Depression has been reported to affect the outcome after TKR,32,33 and Wylde et al2 recently identified a relationship between the number of painful sites elsewhere in the body and chronic pain. We did not find any significant influence of depression on post-operative outcome. Only one patient had high levels of neuropathic pain preoperatively and continued with high levels of neuropathic pain up to three years post-operatively. It is unclear whether such patients should have the neuropathic element treated before being considered for TKR. More than 80% of the dissatisfied patients had high levels of pain, and 45% had possible neuropathic pain. The dissatisfied patients had a high rate of post-operative intervention. The challenge for surgeons is to identify remediable faults in those with persistent pain and dissatisfaction after TKR. The study involved patients listed for primary TKR by consultant surgeons, and seen in a pre-operative clinic run by a physiotherapist. The education and explanations received by these patients may have helped alleviate anxieties about pain, and the fact that they were included in
a study investigating pain after TKR may have biased the reports of pain according to the ‘Hawthorne effect’.34 We found no correlation between pre-operative levels of pain, neuropathic pain, anxiety or depression with outcome at a minimum of three years post-operatively. We did, however, find a moderate correlation between pain (VAS) and painDETECT scores six weeks, three months and three years post-operatively. Overall, TKR is a very successful operation. We have been able to demonstrate that by nine months, a third of patients had no pain and a half only had mild pain. At final follow-up at a mean 46 months, almost half (45%) had no pain at all and a third experienced only mild pain. However, patients should be informed as part of the consent process that at three years post-operatively some may still have moderate or severe pain, for which no mechanical cause can be found. The authors wish to thanks the consultants of the Exeter Knee Reconstruction Unit for permission to include their patients in the study. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. This article was primary edited by G. Scott and first proof edited by J. Scott.
References 1. Beswick AD, Wylde V, Gooberman-Hill R, Blom A, Dieppe P. What proportion of patients report long-term pain after total hip or knee replacement for osteoarthritis? A systematic review of prospective studies in unselected patients. BMJ Open 2012;2:000435. 2. Wylde V, Hewlett S, Learmonth ID, Dieppe P. Persistent pain after joint replacement: prevalence, sensory qualities, and postoperative determinants. Pain 2011;152:566–572. 3. Baker PN, van der Meulen JH, Lewsey J, Gregg PJ; National Joint Registry for England and Wales. The role of pain and function in determining patient satisfaction after total knee replacement. Data from the National Joint Registry for England and Wales. J Bone Joint Surg [Br] 2007;89-B:893–900. 4. Baker PN, Rushton S, Jameson SS, et al. Patient satisfaction with total knee replacement cannot be predicted from pre-operative variables alone: A cohort study from the National Joint Registry for England and Wales. Bone Joint J 2013;95B:1359–1365. 5. Mandalia V, Eyres K, Schranz P, Toms AD. Evaluation of patients with a painful total knee replacement. J Bone Joint Surg [Br] 2008;90-B:265–271. 6. Hetaimish BM, Khan MM, Simunovic N, et al. Meta-analysis of navigation vs conventional total knee arthroplasty. J Arthroplasty 27:1177–1182. 7. Woolson ST, Harris AH, Wagner DW, Giori NJ. Component alignment during total knee arthroplasty with use of standard or custom instrumentation: a randomized clinical trial using computed tomography for postoperative alignment measurement. J Bone Joint Surg [Am] 2014;96-A:366–372. 8. Gøthesen O, Espehaug B, Havelin LI, et al. Functional outcome and alignment in computer-assisted and conventionally operated total knee replacements: a multicentre parallel-group randomised controlled trial. Bone Joint J 2014;96-B:609–618. 9. Krishnan SP, Dawood A, Richards R, Henckel J, Hart AJ. A review of rapid prototyped surgical guides for patient-specific total knee replacement. J Bone Joint Surg [Br] 2012;94-B:1457–1461. 10. Abdel MP, Oussedik S, Parratte S, Lustig S. Coronal alignment in total knee replacement: historical review, contemporary analysis, and future direction. Bone Joint J 2014;96-B:857–862. 11. Dossett HG, Estrada NA, Swartz GJ, LeFevre GW, Kwasman BG. A randomised controlled trial of kinematically and mechanically aligned total knee replacements: Two-year clinical results. Bone Joint J 2014;96-B:907–913. 12. Drexler M, Dwyer T, Chakravertty R, Farno A, Backstein D. Assuring the happy total knee replacement patient . Bone Joint J 2013;95-B(11SupplA):120–123. 13. Mont MA, Serna FK, Krackow KA, Hungerford DS. Exploration of radiographically normal total knee replacements for unexplained pain. Clin Orthop Relat Res 1996;331:216–220. 14. Buvanendran A, Kroin JS, Della Valle CJ, et al. Perioperative oral pregabalin reduces chronic pain after total knee arthroplasty: a prospective, randomized, controlled trial. Anesth Analg 2010;110:199–207. THE BONE & JOINT JOURNAL
THE NATURAL HISTORY OF PAIN AND NEUROPATHIC PAIN AFTER KNEE REPLACEMENT
15. Treede RD, Jensen TS, Campbell JN, et al. Neuropathic pain: redefinition and a grading system for clinical and research purposes. Neurology 2008;70:1630–1635. 16. Jensen TS. Neuropathic pain; definition and screening. Eur J Pain 2007;11:S7. 17. McDermott AM, Toelle TR, Rowbotham DJ, Schaefer CP, Dukes EM. The burden of neuropathic pain: results from a cross-sectional survey. Eur J Pain 2006;10:127–135. 18. Freynhagen R, Bennett MI. Diagnosis and management of neuropathic pain. BMJ 2009;339:3002. 19. Wylde V, Bruce J, Beswick A, Elvers K, Gooberman-Hill R. Assessment of chronic postsurgical pain after knee replacement: a systematic review. Arthritis Care Res (Hoboken) 2013;65:1795–1803. 20. Marques EM, Jones HE, Elvers KT, et al. Local anaesthetic infiltration for perioperative pain control in total hip and knee replacement: systematic review and metaanalyses of short- and long-term effectiveness. BMC Musculoskelet Disord 2014;15:220.(Epub). 21. Chan EY, Blyth FM, Nairn L. Fransen M. Acute postoperative pain following hospital discharge after total knee arthroplasty. Osteoarthritis Cartilage 2013;9:1257–1263. 22. Dworkin RH, Turk DC, Farrar JT, et al. Core outcome measures for chronic pain clinical trials: IMMPACT recommendations. Pain 2005;113:9–19. 23. Farrar JT, Young JP Jr, LaMoreaux L, Werth JL, Poole RM. Clinical importance of changes in chronic pain intensity measured on an 11-point numerical pain rating scale. Pain 2001;94:149–158. 24. Collins SL, Moore RA, McQuay HJ. The visual analogue pain intensity scale: what is moderate pain in millimetres? Pain 1997;72:95–97. 25. Freynhagen R, Baron R, Gockel U, Tölle TR. painDETECT: a new screening questionnaire to identify neuropathic components in patients with back pain. Curr Med Res Opin 2006;10:1911–1920.
VOL. 96-B, No. 9, SEPTEMBER 2014
1233
26. Bennett MI, Smith BH, Torrance N, Potter J. The S-LANSS score for identifying pain of predominantly neuropathic origin: validation for use in clinical and postal research. J Pain 2005;6:149–158. 27. Zigmond AS, Snaith RP. The hospital anxiety and depression scale. Acta Psychiatr Scand 1983;67:361–370. 28. Murray DW, Fitzpatrick R, Rogers K, et al. The use of the Oxford hip and knee scores. J Bone Joint Surg [Br] 2007;89-B:1010–1014. 29. Bellamy N, Buchanan WW, Goldsmith CH, Campbell J, Stitt LW. Validation study of WOMAC: a health status instrument for measuring clinically important patient relevant outcomes to antirheumatic drug therapy in patients with osteoarthritis of the hip or knee. J Rheumatol 1988;15:1833–1840. 30. No authors listed. PROMS team at Health and Social Care Information Centre. Finalised Patient Reported Outcome Measures in England, April 2011 to March 2012. http://www.hscic.gov.uk/catalogue/PUB11359/final-proms-eng-apr11-mar12-finreport-v2.pdf (date last accessed 17 June 2014). 31. No authors listed. National Institute for Health and Clinical Excellence: NICE guideline 96: Neuropathic pain: The pharmacological management of neuropathic pain in adults in non-specialist settings, March 2010. http://guidance.nice.org.uk/CG96 (date last accessed 17 June 2014). 32. Brander V, Gondek S, Martin E, Stulberg SD. Pain and depression influence outcome 5 years after knee replacement surgery. Clin Orthop Relat Res 2007;464:21–26. 33. Hanusch BC, O'Connor DB, Ions P, Scott A, Gregg PJ. Effects of psychological distress and perceptions of illness on recovery from total knee replacement. Bone Joint J 2014;96-B:210–216. 34. Landsberger HA. Hawthorne revisited. Social Forces 1959;37:361–364.
The natural history of pain and neuropathic pain after knee replacement A PROSPECTIVE COHORT STUDY OF THE POINT PREVALENCE OF PAIN AND NEUROPATHIC PAIN TO A MINIMUM THREE-YEAR FOLLOW-UP J. R. A. Phillips, B. Hopwood, C. Arthur, R. Stroud, A. D Toms From Exeter Knee Reconstruction Unit, Princess Elizabeth Orthopaedic Centre, Royal Devon and Exeter Hospital, Exeter, United Kingdom
A small proportion of patients have persistent pain after total knee replacement (TKR). The primary aim of this study was to record the prevalence of pain after TKR at specific intervals post-operatively and to ascertain the impact of neuropathic pain. The secondary aim was to establish any predictive factors that could be used to identify patients who were likely to have high levels of pain or neuropathic pain after TKR. A total of 96 patients were included in the study. Their mean age was 71 years (48 to 89); 54 (56%) were female. The mean follow-up was 46 months (39 to 51). Pre-operative demographic details were recorded including a Visual Analogue Score (VAS) for pain, the Hospital Anxiety and Depression score as well as the painDETECT score for neuropathic pain. Functional outcome was assessed using the Oxford Knee score The mean pre-operative VAS was 5.8 (1 to 10); and it improved significantly at all time periods post-operatively (p < 0.001): (from 4.5 at day three to five (1 to 10), 3.2 at six weeks (0 to 9), 2.4 at three months (0 to 7), 2.0 at six months (0 to 9), 1.7 at nine months (0 to 9), 1.5 at one year (0 to 8) and 2.0 at mean 46 months (0 to 10)). There was a high correlation (r > 0.7; p < 0.001) between the mean VAS scores for pain and the mean painDETECT scores at three months, one year and three years post-operatively. There was no correlation between the pre-operative scores and any post-operative scores at any time point. We report the prevalence of pain and neuropathic pain at various intervals up to three years after TKR. Neuropathic pain is an underestimated problem in patients with pain after TKR. It peaks at between six weeks and three-months post-operatively. However, from these data we were unable to predict which patients are most likely to be affected. Cite this article: Bone Joint J 2014;96-B:1227–33.
J. R. A. Phillips, MBChB, BSc, FRCS (Tr&Orth), Knee Fellow B. Hopwood, MBBS, MRCA, MSc, Associate Specialist Anaesthetics And Pain Management C. Arthur, MB ChB, FRCS (Tr&Orth), Knee Fellow R. Stroud, MCSP, Extended Scope Physiotherapist A. D Toms, MB ChB, MSc, FRCS (Tr&Orth), Consultant Orthopaedic Surgeon Royal Devon and Exeter Hospital, Exeter Knee Reconstruction Unit, Barrack Road, Exeter, UK. Correspondence should be sent to Mr J. R. A. Phillips; e-mail: [email protected] ©2014 The British Editorial Society of Bone & Joint Surgery doi:10.1302/0301-620X.96B9. 33756 $2.00 Bone Joint J 2014;96-B:1227–33. Received 2 January 2014; Accepted after revision 06 June 2014
The incidence of chronic pain and disability after total knee replacement (TKR) is reported to be between 8% and 34%.1,2 National Joint Registry data have revealed rates of dissatisfaction one year after TKR of more than 18%.3,4 Many causes of chronic pain after TKR have been described.5 Efforts have been made to improve the outcome by technological developments such as the use of navigation and custom-made cutting blocks,6-9 as well as methodology, such as by using kinematic rather than mechanical alignment.10,11 However, there still appears to be a group of patients with a technically satisfactory TKR and without evidence of infection who remain dissatisfied. Dissatisfaction is clearly multifactorial but a contributing factor is pain.12 However, revision has been found to have limited success when performed for pain alone.13 Neuropathic pain has been described after TKR. Buvanendran et al14 identified a rate of neuropathic pain of 5% six months postoperatively in a prospective series of
VOL. 96-B, No. 9, SEPTEMBER 2014
120 patients after TKR. Wylde et al2 identified rates of neuropathic pain of 6% three to four years post-operatively in a retrospective study of 632 patients after TKR. Why rates of neuropathic pain are so high has not been explained. Neuropathic pain is defined as “pain arising as a direct consequence of a lesion or disease affecting the somatosensory system”.15 Common findings in neuropathic pain are a sensory deficit or hypersensitivity in the painful area and a gradual increase in pain following repetitive stimulation, and paroxysms of pain.16 Patients with neuropathic pain report having pain of longer duration and have greater interference from pain and significantly poorer quality of life than those with nociceptive pain.17 Neuropathic pain does not usually respond to simple analgesics and specific management strategies are needed.18 There is a lack of evidence about the natural history of pain and, specifically, neuropathic pain following TKR. Most studies on 1227
1228
J. R. A. PHILLIPS, B. HOPWOOD, C. ARTHUR, R. STROUD, A. D TOMS
Table I. Pre-operative scores Pre-operative score (n = 96)
Mean
Range
VAS pain 5.8 1 to 10 painDETECT 4.9 0 to 20 HADS anxiety 3.8 0 to 17 HADS depression 5.4 1 to 14 Oxford Knee Score 20 7 to 45 VAS, visual analogue scale; HADS, Hospital Anxiety and Depression Score
pain focus on the early post-operative period with few continuing beyond three months after TKR.19-21 We therefore designed a prospective study involving a cohort of patients undergoing TKR for osteoarthritis evaluating pain and any neuropathic element to it, at regular intervals, for a minimum of three years post-operatively. The primary aim was to establish the prevalence of pain at specific intervals post-operatively, and to ascertain the impact of neuropathic pain. A secondary aim was to establish any predictive factors that could be used to identify which patients were likely to experience high levels of pain or neuropathic pain after TKR.
Patients and Methods A total of 96 patients with primary osteoarthritis of the knee were recruited to the study between July 2009 and May 2010. These were consecutive patients seen in a physiotherapist-led pre-operative education class, of whom 54 (56%) were female. Their mean age was 70.6 years (48 to 89) and all underwent primary TKR. Baseline details included demographics, co-morbidities and the use of medication including anti-depressants and anti-neuropathics were recorded. Pain was assessed using a 0 to 10 Visual Analogue Scale (VAS); this is a well evaluated tool.22 The 11-point Likert scale uses zero to represent no pain and ten to represent the worst possible pain. Improvements of > 30% (i.e. three points on the scale) are thought to be significant.23 The VAS has also been interpreted as either no pain (VAS = 0), mild (VAS = 1 to 3), moderate (VAS = 4 to 6) or severe (VAS = 7 to 10).24 Neuropathic pain was self-assessed using the painDETECT questionnaire.25 This has also been evaluated in a mixed pain clinic population, and we felt that was a better tool than others such as the self-administered version of the Leeds Assessment of Neuropathic Signs and Symptoms (S-LANSS),26 as it allowed patients to grade the severity of their symptoms and a number could be recorded which could be correlated with a VAS pain score. A painDETECT score 0 to 12 implies neuropathic pain is unlikely, 13 to 18 implies neuropathic pain is unclear (possible) and 19 to 38 means neuropathic pain is likely to be present. Pre-operative anxiety and depression was evaluated using the HADS assessment tool.27 Scores < 7 points are negative, 8 to 10 to be possible and > 11 points indicates anxiety and/ or depression. Functional outcome was assessed using the Oxford Knee Score (OKS; 0 to 48 worst to best).28
painDETECT > 19 n = 1 > 11 n = 6 > 11 n = 7
Patients were reviewed in hospital on days three to five post-operatively and were contacted by post at six weeks, three months, six months, nine months, 12 months and finally at a minimum of three years post-operatively. Those with high pain scores in the early post-operative period were referred to the acute pain service. Patients were also asked to make free comments which could be used in the interpretation of the results. At one year and at final follow-up, patients were asked their interpretation of the success of their operation (excellent, very good, good, fair or poor), and their satisfaction with surgery (delighted, pleased, mostly satisfied, mixed, mostly dissatisfied, unhappy or terrible). In addition, further surgery or interventions were recorded. The use of medication was recorded at each stage to aid the interpretation of results but not as a surrogate marker as it would be unreliable due to the many factors that affect the use of medication including intolerance and inefficacy. Those who failed to respond were followed-up with a telephone call but if not successful data were omitted for that period. All clinical interventions, whether related or not to the TKR, were recorded. The study had ethical approval and consent was granted by Pfizer (Tadworth, Surrey, United Kingdom) to use their pre-printed painDETECT forms. Statistical analysis. This was performed using GraphPad Prism version 6.00 for Windows, (GraphPad Software, La Jolla, California). Paired results were compared using paired Wilcoxon tests (non-parametric data). Correlation was determined using the Spearman correlation test for non-parametric data. R > 0.7 is deemed to be highly correlated, and r = 0.5 to 0.7 moderately correlated. A p-value of < 0.05 was considered to be statistically significant.
Results The mean follow-up was 46 months (39 to 51). The rate for completion of the documentation was high, with 89 patients (94%; one patient had died) responding at 12 months, and 80 (92%; seven patients had died) responding at final follow-up, at which time 19 patients (20%) had undergone contralateral TKR. Pre-operative scores are shown in Table I. Pain scores improved up to the time of final follow-up (p < 0.001 for all data points compared with the pre-operative VAS for pain; Wilcoxon matched-pairs signed rank test; Fig. 1). The prevalence of pain at all points is shown in Table II and Figure 2. THE BONE & JOINT JOURNAL
THE NATURAL HISTORY OF PAIN AND NEUROPATHIC PAIN AFTER KNEE REPLACEMENT
1229
VAS scale
10
5
0 Day 3 wks 3 mths 6 mths 9 mths 1 yr Preoperative 3 to 5
4 yrs
Time since surgery Fig. 1 Plot graph showing pain according to the mean visual analogue scale (VAS) with 95% confidence intervals at each time interval after knee replacement.
Table II. Mean visual analogue score (VAS) for pain at each review
VAS Mean Range
Pre-operative
Days three to five
Six weeks
Three months
Six months
Nine months
12 months
Final follow-up mean 46 months
5.8 1 to 10
4.5 1 to 10
3.2 0 to 9
2.4 0 to 7
2.0 0 to 9
1.7 0 to 9
1.5 0 to 8
2.0 0 to 10
Pre-operatively, the painDETECT scores revealed a prevalence of possible neuropathic pain in five patients (5%) and likely neuropathic pain in only one patient (Table III). Neuropathic pain became most prevalent six weeks postoperatively with 23 of 85 respondents (27%) having possible neuropathic pain, and seven (8%) having likely neuropathic pain. Three months post-operatively, 16 of 82 respondents (17%) had possible and two (3%) had likely neuropathic pain. After peaking between six weeks and three months, the proportion of patients having neuropathic pain then reduced to eight of 87 respondents (9%) with possible and two (2%) with likely neuropathic pain one year post-operatively. At final follow-up, six of 79 respondents (7%) had possible neuropathic pain and five (6%) had likely neuropathic pain. The changes in painDETECT scores are shown in Figures 3 and 4. The intensity of neuropathic pain varied with time. Of the seven patients with likely neuropathic pain six weeks post-operatively, only one had likely neuropathic pain at final follow-up (with one lost to follow-up). Of the five patients at final follow-up with likely neuropathic pain, only one had this pre-operatively, two had it six-weeks post-operatively (with no data collected for one patient at this time) and one had it three months post-operatively. Of the five patients with possible neuropathic pain preoperatively, one had developed likely neuropathic pain one year post-operatively (one was lost to follow-up). At VOL. 96-B, No. 9, SEPTEMBER 2014
three years post-operatively one new patient had developed likely neuropathic pain, two were lost to follow up and the other patient with neuropathic pain had died. The single patient with likely neuropathic pain preoperatively had high scores suggestive of neuropathic pain at every stage and at 40 months post-operatively. Preoperatively she had no anxiety or depression, her VAS for pain was 10/10 and she was already on treatment for neuropathic pain. She had high levels of pain post-operatively and was treated with opiate patches, pregabalin and amitryptiline. She underwent two manipulations under anaesthesia and at the latest follow-up at 40 months she had a VAS pain score of 8/10, an OKS score of 17 and a painDETECT score of 21. Despite this, she thought the outcome of her TKR was ‘fair’ and she had ‘mixed’ satisfaction levels. Patients with high painDETECT scores at final followup underwent secondary review of their medical records and radiographs. This revealed no features of malalignment or incorrect sizing that could explain the symptoms. Patient satisfaction. Satisfaction one year post-operatively and at final follow-up is shown in Figures 5 and 6. At final follow-up, 11 of 79 respondents (14%) had mixed feelings about their TKR or were dissatisfied. In all, ten of this group had VAS pain scores of > 7 on walking. Two had possible neuropathic pain and three had likely neuropathic pain. Of the 11 dissatisfied patients, six had undergone further surgery and one is awaiting arthroscopy to debride
1230
J. R. A. PHILLIPS, B. HOPWOOD, C. ARTHUR, R. STROUD, A. D TOMS
Vas score (%)
100
No pain
90
Mild
80
Moderate
70
Severe
60 50 40 30 20 10 0 PreDay 3 to 5 operative
6 wks
3 mths
6 mths
9 mths
1 yr
3 yrs
Time since surgery Fig. 2 Bar chart showing the prevalence and severity of pain before and after knee replacement using the visual analogue score (VAS). (0: no pain, 7 to 10: severe pain).
Table III. The numbers of patients according to the ranges of the visual analogue score (VAS) for pain and painDETECT scores for neuropathic pain at each review
Pre-operative (%) Day three post-op (%) Six weeks post-op (%) Three months post-op (%) Six months post-op (%) Nine months post-op (%) 12 months post-op (%) Mean 46 months post-op (%)
n=
PAIN none PAIN mild VAS = 0 VAS = 1 to 3
PAIN moderate PAIN severe VAS = 4 to 6 VAS = 7 to 10
n=
painDETECT painDETECT painDETECT 0 to 12 negative 13 to 18 possible 19 to 38 likely
96 74 85 83 90 86 86 80
0 3 4 (5) 15 (18) 29 (32) 28 (32) 35 (41) 36 (45)
39 (41) 35 (47) 33 (39) 21 (21) 14 (16) 14 (16) 12 (14) 12 (15)
96 74 85 82 90 86 87 79
90 (94) 68 (92) 55 (65) 63 (77) 72 (80) 70 (81) 77 (89) 68 (87)
9 (9) 22 (34) 40 (47) 42 (51) 42 (47) 41 (48) 36 (42) 26 (33)
patellofemoral scar tissue, causing crepitus. No significant factors were identified pre-operatively that predicted dissatisfaction at final assessment. Of the five patients with likely neuropathic pain at final follow-up, two were not satisfied with their operation, one had mixed feelings and two were satisfied. Of the five patients with VAS pain scores of > 7 at final follow-up, two were not satisfied with their operation, one reported a mixed outcome and two were satisfied. A total of ten patients had high anxiety or depression scores pre-operatively (HADS > 11); three had high scores for both anxiety and depression; four had high scores for depression alone and three for anxiety alone. One patient died and two others were lost to follow-up. Only one patient, with depression, developed likely neuropathic pain at final follow-up; all other patients had painDETECT scores of < 12, all had VAS pain scores of < 5, and all but one patient had better than ‘mixed’ satisfaction levels.
48 (50) 14 (19) 8 (9) 5 (10) 5 (6) 3 (4) 3 (3) 6 (7)
5 (5) 4 (5) 23 (27) 16 (19) 16 (17) 11 (13) 8 (9) 6 (7)
1 (1) 2 (3) 7 (8) 3 (4) 2 (3) 5 (6) 2 (2) 5 (6)
There was no correlation between the pre-operative VAS pain scores and post-operative painDETECT scores at any stage (r < 0.5). Similarly, there was no correlation between the pre-operative painDETECT scores and the postoperative VAS pain scores at any stage (r < 0.5). However, at other intervals there were either moderate or high correlations between the VAS pain scores and the painDeTECT scores at the same times post-operatively. VAS pain and painDETECT scores correlated at three months (Spearman correlation r = 0.74, 95% confidence intervals (CI) 0.61 to 0.82), nine months (r = 0.76, 95% CI 0.65 to 0.84), 12 months (r = 0.72, 95% CI 0.59 to 0.81) and at final follow-up (r = 0.76, 95% CI 0.65 to 0.85). The painDETECT scores at six weeks and six months postoperatively correlated only moderately with VAS pain scores at final follow-up (r = 0.53 and r = 0.50). At final follow-up, a high (negative) correlation was identified between the OKS and the VAS pain (r = -0.86, 95% CI -0.78 to -0.91), and painDETECT (r = -0.77, 95% THE BONE & JOINT JOURNAL
THE NATURAL HISTORY OF PAIN AND NEUROPATHIC PAIN AFTER KNEE REPLACEMENT
1231
20
Delighted
15 Patient satisfaction
painDETECT score
100
10
5
90
Pleased
80
Mostly satisfied mixed
70 60 50
Mostly dissatisfied unhappy
40 30 20
Terrible
10 0
0 Day Preoperative 3 to 5
3 mths 6 mths 9 mths 1 yr
6 wks
1 yr
4 yrs
3 yrs Time since surgery
Time since surgery Fig. 5 Graph showing mean painDETECT scores pre- and post-operatively, with 95% confidence intervals.
0 to 12 Neuropathic pain unlikely
100 pain DETECT score (%)
90 80 70
13 to 18 Neuropathic pain possible
60 50 40 30 20
> 19 Neuropathic pain likely
10 0 Pre- Day 3 to 5 6 wks 3 mths 6 mths 9 mths 1 yr operative Time since surgery
3 yrs
Bar chart showing satisfaction one and three years after knee replacement.
100
Patient interpretation of success (%)
Fig. 3
80 60 40 20 0 1 yr Excellent
3 yrs Very good
Good
Fair
Poor
Time since surgery
Fig. 4
Fig. 6
Bar chart showing prevalence of neuropathic pain pre- and postoperatively using the painDETECT score.
Bar chart showing the patients’ interpretation of the level of success one and three years after knee replacement.
CI -0.85 to -0.65) scores. There was no correlation between the pre-operative HADS scores and any other scores. Complications and further surgery. A total of 11 patients (11%) had a complication or underwent further surgery. Two patients underwent revision surgery, one for instability and one for deep infection; three patients sustained traumatic extensor mechanism injuries (one open avulsion of the patellar tendon, one avulsion of the quadriceps tendon and one patellar fracture, treated non-operatively); three patients underwent manipulation under anaesthesia within four months post-operatively; two underwent arthroscopic arthrolysis for stiffness and one underwent secondary patellar resurfacing. The mean VAS pain score at final follow-up in these patients, two of whom were lost to follow-up was 4.3 (1.3 to 10).The mean painDETECT score was 11.7 (2 to 21) and two had scores of > 18. Six of the remaining ten patients reported a ‘fair’ or ‘poor’ outcome, and seven reported their satisfaction as being ‘mixed’ or ‘worse’. VOL. 96-B, No. 9, SEPTEMBER 2014
Discussion This study confirms that there is a significant decrease in pain after TKR. However, for some, pain persists at three years post-operatively; with 12 of 80 patients (15%) having moderate pain and six of 80 (7%) having severe pain at a mean of 46 months. This is consistent with the findings of Wylde et al2 who retrospectively identified 632 patients who had undergone TKR and sent out questionnaires between three and four years post-operatively, achieving a 73% response rate. Using the Western Ontario and McMaster Universities Arthritis Index score,29 17% had moderate persistent pain and 15% had severe persistent pain. We found significant improvements of pain after TKR at all points (p < 0.001) and levels of satisfaction that are consistent with United Kingdom patient reported outcome measures (PROMS).3,30 Previous studies by Wylde et al2 and Buvenendran et al14 found levels of neuropathic pain to be 5% at six months
1232
J. R. A. PHILLIPS, B. HOPWOOD, C. ARTHUR, R. STROUD, A. D TOMS
and 6% between three and four years post-operatively. Using the painDETECT score, we found that the rate of neuropathic pain was at its highest six-weeks postoperatively, with 8% having likely, and 27% having possible neuropathic pain. This rate then decreased and 3% had likely and 17% had possible neuropathic pain six months post-operatively. At final follow-up, 6% had likely and 7% had possible neuropathic pain. Neuropathic pain adversely affects the quality of life.17 We have demonstrated that potentially up to 35% of patients have symptoms of neuropathic pain six weeks after TKR. The question arises whether these patients should receive specific treatment for these symptoms. It is possible that many of the symptoms experienced in the early postoperative period may be neuropathic in origin as part of the normal, inflammatoy process. Significant symptoms can be treated in accordance with the National Institute for Health and Clinical Excellence (NICE) guidelines for the pharmacological management of neuropathic pain in adults.31 Amitriptyline or pregabalin are recommended as first line treatment, but when symptoms cannot be controlled, early referral to a specialist pain service is suggested.31 This study has identified a correlation between pain as assessed using a VAS and neuropathic pain as assessed using the painDETECT score. There was a high correlation (r > 0.7) between these two at three months, one year and a mean 46 months after operation. Pre-operative screening with painDETECT scores is unlikely to be useful in the absence of a correlation between the pre-operative scores and those at a minimum of three years. However, a moderate correlation was identified between the painDETECT scores at six weeks and three months and those at final follow-up. Thus, early intervention in these patients might be helpful. There was no correlation between any pre-operative scores (VAS, painDETECT or HADS), and the postoperative scores. Depression has been reported to affect the outcome after TKR,32,33 and Wylde et al2 recently identified a relationship between the number of painful sites elsewhere in the body and chronic pain. We did not find any significant influence of depression on post-operative outcome. Only one patient had high levels of neuropathic pain preoperatively and continued with high levels of neuropathic pain up to three years post-operatively. It is unclear whether such patients should have the neuropathic element treated before being considered for TKR. More than 80% of the dissatisfied patients had high levels of pain, and 45% had possible neuropathic pain. The dissatisfied patients had a high rate of post-operative intervention. The challenge for surgeons is to identify remediable faults in those with persistent pain and dissatisfaction after TKR. The study involved patients listed for primary TKR by consultant surgeons, and seen in a pre-operative clinic run by a physiotherapist. The education and explanations received by these patients may have helped alleviate anxieties about pain, and the fact that they were included in
a study investigating pain after TKR may have biased the reports of pain according to the ‘Hawthorne effect’.34 We found no correlation between pre-operative levels of pain, neuropathic pain, anxiety or depression with outcome at a minimum of three years post-operatively. We did, however, find a moderate correlation between pain (VAS) and painDETECT scores six weeks, three months and three years post-operatively. Overall, TKR is a very successful operation. We have been able to demonstrate that by nine months, a third of patients had no pain and a half only had mild pain. At final follow-up at a mean 46 months, almost half (45%) had no pain at all and a third experienced only mild pain. However, patients should be informed as part of the consent process that at three years post-operatively some may still have moderate or severe pain, for which no mechanical cause can be found. The authors wish to thanks the consultants of the Exeter Knee Reconstruction Unit for permission to include their patients in the study. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. This article was primary edited by G. Scott and first proof edited by J. Scott.
References 1. Beswick AD, Wylde V, Gooberman-Hill R, Blom A, Dieppe P. What proportion of patients report long-term pain after total hip or knee replacement for osteoarthritis? A systematic review of prospective studies in unselected patients. BMJ Open 2012;2:000435. 2. Wylde V, Hewlett S, Learmonth ID, Dieppe P. Persistent pain after joint replacement: prevalence, sensory qualities, and postoperative determinants. Pain 2011;152:566–572. 3. Baker PN, van der Meulen JH, Lewsey J, Gregg PJ; National Joint Registry for England and Wales. The role of pain and function in determining patient satisfaction after total knee replacement. Data from the National Joint Registry for England and Wales. J Bone Joint Surg [Br] 2007;89-B:893–900. 4. Baker PN, Rushton S, Jameson SS, et al. Patient satisfaction with total knee replacement cannot be predicted from pre-operative variables alone: A cohort study from the National Joint Registry for England and Wales. Bone Joint J 2013;95B:1359–1365. 5. Mandalia V, Eyres K, Schranz P, Toms AD. Evaluation of patients with a painful total knee replacement. J Bone Joint Surg [Br] 2008;90-B:265–271. 6. Hetaimish BM, Khan MM, Simunovic N, et al. Meta-analysis of navigation vs conventional total knee arthroplasty. J Arthroplasty 27:1177–1182. 7. Woolson ST, Harris AH, Wagner DW, Giori NJ. Component alignment during total knee arthroplasty with use of standard or custom instrumentation: a randomized clinical trial using computed tomography for postoperative alignment measurement. J Bone Joint Surg [Am] 2014;96-A:366–372. 8. Gøthesen O, Espehaug B, Havelin LI, et al. Functional outcome and alignment in computer-assisted and conventionally operated total knee replacements: a multicentre parallel-group randomised controlled trial. Bone Joint J 2014;96-B:609–618. 9. Krishnan SP, Dawood A, Richards R, Henckel J, Hart AJ. A review of rapid prototyped surgical guides for patient-specific total knee replacement. J Bone Joint Surg [Br] 2012;94-B:1457–1461. 10. Abdel MP, Oussedik S, Parratte S, Lustig S. Coronal alignment in total knee replacement: historical review, contemporary analysis, and future direction. Bone Joint J 2014;96-B:857–862. 11. Dossett HG, Estrada NA, Swartz GJ, LeFevre GW, Kwasman BG. A randomised controlled trial of kinematically and mechanically aligned total knee replacements: Two-year clinical results. Bone Joint J 2014;96-B:907–913. 12. Drexler M, Dwyer T, Chakravertty R, Farno A, Backstein D. Assuring the happy total knee replacement patient . Bone Joint J 2013;95-B(11SupplA):120–123. 13. Mont MA, Serna FK, Krackow KA, Hungerford DS. Exploration of radiographically normal total knee replacements for unexplained pain. Clin Orthop Relat Res 1996;331:216–220. 14. Buvanendran A, Kroin JS, Della Valle CJ, et al. Perioperative oral pregabalin reduces chronic pain after total knee arthroplasty: a prospective, randomized, controlled trial. Anesth Analg 2010;110:199–207. THE BONE & JOINT JOURNAL
THE NATURAL HISTORY OF PAIN AND NEUROPATHIC PAIN AFTER KNEE REPLACEMENT
15. Treede RD, Jensen TS, Campbell JN, et al. Neuropathic pain: redefinition and a grading system for clinical and research purposes. Neurology 2008;70:1630–1635. 16. Jensen TS. Neuropathic pain; definition and screening. Eur J Pain 2007;11:S7. 17. McDermott AM, Toelle TR, Rowbotham DJ, Schaefer CP, Dukes EM. The burden of neuropathic pain: results from a cross-sectional survey. Eur J Pain 2006;10:127–135. 18. Freynhagen R, Bennett MI. Diagnosis and management of neuropathic pain. BMJ 2009;339:3002. 19. Wylde V, Bruce J, Beswick A, Elvers K, Gooberman-Hill R. Assessment of chronic postsurgical pain after knee replacement: a systematic review. Arthritis Care Res (Hoboken) 2013;65:1795–1803. 20. Marques EM, Jones HE, Elvers KT, et al. Local anaesthetic infiltration for perioperative pain control in total hip and knee replacement: systematic review and metaanalyses of short- and long-term effectiveness. BMC Musculoskelet Disord 2014;15:220.(Epub). 21. Chan EY, Blyth FM, Nairn L. Fransen M. Acute postoperative pain following hospital discharge after total knee arthroplasty. Osteoarthritis Cartilage 2013;9:1257–1263. 22. Dworkin RH, Turk DC, Farrar JT, et al. Core outcome measures for chronic pain clinical trials: IMMPACT recommendations. Pain 2005;113:9–19. 23. Farrar JT, Young JP Jr, LaMoreaux L, Werth JL, Poole RM. Clinical importance of changes in chronic pain intensity measured on an 11-point numerical pain rating scale. Pain 2001;94:149–158. 24. Collins SL, Moore RA, McQuay HJ. The visual analogue pain intensity scale: what is moderate pain in millimetres? Pain 1997;72:95–97. 25. Freynhagen R, Baron R, Gockel U, Tölle TR. painDETECT: a new screening questionnaire to identify neuropathic components in patients with back pain. Curr Med Res Opin 2006;10:1911–1920.
VOL. 96-B, No. 9, SEPTEMBER 2014
1233
26. Bennett MI, Smith BH, Torrance N, Potter J. The S-LANSS score for identifying pain of predominantly neuropathic origin: validation for use in clinical and postal research. J Pain 2005;6:149–158. 27. Zigmond AS, Snaith RP. The hospital anxiety and depression scale. Acta Psychiatr Scand 1983;67:361–370. 28. Murray DW, Fitzpatrick R, Rogers K, et al. The use of the Oxford hip and knee scores. J Bone Joint Surg [Br] 2007;89-B:1010–1014. 29. Bellamy N, Buchanan WW, Goldsmith CH, Campbell J, Stitt LW. Validation study of WOMAC: a health status instrument for measuring clinically important patient relevant outcomes to antirheumatic drug therapy in patients with osteoarthritis of the hip or knee. J Rheumatol 1988;15:1833–1840. 30. No authors listed. PROMS team at Health and Social Care Information Centre. Finalised Patient Reported Outcome Measures in England, April 2011 to March 2012. http://www.hscic.gov.uk/catalogue/PUB11359/final-proms-eng-apr11-mar12-finreport-v2.pdf (date last accessed 17 June 2014). 31. No authors listed. National Institute for Health and Clinical Excellence: NICE guideline 96: Neuropathic pain: The pharmacological management of neuropathic pain in adults in non-specialist settings, March 2010. http://guidance.nice.org.uk/CG96 (date last accessed 17 June 2014). 32. Brander V, Gondek S, Martin E, Stulberg SD. Pain and depression influence outcome 5 years after knee replacement surgery. Clin Orthop Relat Res 2007;464:21–26. 33. Hanusch BC, O'Connor DB, Ions P, Scott A, Gregg PJ. Effects of psychological distress and perceptions of illness on recovery from total knee replacement. Bone Joint J 2014;96-B:210–216. 34. Landsberger HA. Hawthorne revisited. Social Forces 1959;37:361–364.
