1 Introduction

The traditional notion of pain, Cartesian notion, is that of one-to-one relation between tissue damage and the pain experienced by the patient. However, this sensory account of pain was inconsistent with the variability of pain reported by patients with verifiable physiological insult. As a result, the gate control theory was proposed in 1965 which conceptualised pain as a complex blend of sensory, emotional, cognitive–evaluative, interpersonal and cultural factors.¹ Since, a number of factors have been identified from psychiatric and psychological research as determinants of the pain experience. In particular, for musculoskeletal patients, pain catastrophising, fear avoidance and poor expectations for recovery.²

Pain catastrophising is defined as a negative cognitive-affective response to anticipated or actual pain and has been associated with a number of important pain-related outcomes.^3–5 In other words, catastrophising pain is an irrationally negative forecast of future events and is conceived as a set of exaggerated and negative cognitive and emotional experiences during actual or anticipated painful stimulation.⁶ The Pain Catastrophising Scale (PCS) is a self-reported measure that assesses three domains believed to comprise much of the pain catastrophising concept; helplessness, rumination and magnification.⁷ PCS consists of 13 items scored from 0 to 4, resulting in a total possible score of 52 points. The higher the score, the more catastrophising thoughts are present. Although a cut-off of more than 30 points thought to be associated with clinical relevance in validation studies.^8,9

In a recent systematic review of 35 studies of psychological factors affecting the outcome of total hip and total knee arthroplasty, the evidence suggests that low preoperative mental health and pain catastrophising had a negative influence on outcome after knee arthroplasty.¹⁰ However, the effect of pain catastrophising on hip arthroplasty patients is less clear. The aim of this pilot study was to investigate the relationship between pain catastrophising and patient-reported clinical outcomes following primary total hip arthroplasty (THA). Specifically, whether a low preoperative pain catastrophising score can predict worse patient-reported outcomes following primary THA.

2 Materials and Methods

This was a prospective consecutive study of all patients who underwent primary THA and also had completed preoperative PCS scores, preoperative and 12-month postoperative Oxford hip score (OHS). The study period was between December 2015 and March 2017.

In our institution, data of hip arthroplasty patients are routinely collected in a prospective database including demographic and surgical data, patients reported outcome measures (PROMS) including Oxford Hip Score (OHS)¹¹ and the EuroQol Questionnaire-Visual Analogue Scale (EQ-VAS).¹²

For the purposes of this study, following a local ethical study approval, we introduced the PCS score as part of the preoperative assessment of all arthroplasty patients. Patients informed consent was obtained.

Statistical analysis: data expressed in means, standard deviations and range with changes in mean scores analysed by paired t-test. The relationship between preoperative PCS score and postoperative OHS was investigated. The primary independent variable was the preoperative PCS score. Potential cofounders were also considered including patient factors (age, gender, BMI, smoking status, opiate use, residential status), medical factors (ASA grade, number of comorbidities), and surgical factors (side of surgery, anaesthetic type, indication for surgery, type of component fixation, length of procedure, length of inpatient stay, surgical approach).

Correlation analysis was carried out between the improvement in OHS (mean difference between pre- and postoperative scores) and the mean preoperative PCS score using the Pearson's r rank test. Multiple linear regression was then performed using the postoperative OHS as the outcome variable against a number of predictor variables; namely PCS score. Unadjusted analysis was performed to identify any other variables that had an impact on the OHS outcome including pre-op OHS, BMI, number of comorbidities and length of hospital stay. This allowed for identification of possible confounding variables to the outcome.

Statistical significance was set at a P-value of <0.05, and analysis was performed using StatsDirect software version 3.1.1.

3 Results

A total of 103 patients met the inclusion criteria and were included in the analysis. There were 37 males and 66 females with average age of 60.5 years (range 22–84). End-stage osteoarthritis was the most common indication for surgery (83.9%). Other indications included sequalae of developmental hip dysplasia, osteonecrosis, previous septic arthritis, and post-traumatic arthritis. The most common comorbidity was obesity, defined as body mass index (BMI) > 30 (kg/m²), seen in 36.9% of patients, followed by hypertension (34.0%) (Table 1).

Mean preoperative PCS score was 16.3 (±13.6; range 0–49), indicating a mild pain catastrophising status of the cohort. PROMS data were collected preoperatively and at 12-months follow up for all 103 included patients. Mean preoperative OHS was 16.5 (±3.5) which had improved at 12-months postoperatively to a mean 38.1 (±11.1). The difference was statistically significant (P < 0.0001). Similarly, preoperative EQ-VAS 64.9 (±21.4) had improved at 12-months follow up to a mean 78.2 (±18.1) (P < 0.0001) (Table 2).

Surgical complications: overall rate of complication in this series was 3.8% (4 patients); 1 patient had an intra-operative calcar fracture and 3 patients had acute postoperative infections.

Correlation and regression analysis: Preoperative PCS scores were correlated with the OHS improvement at 12-months which revealed a weak negative correlation Pearson's correlation coefficient r = - 0.248 (P = 0.0114) (Table 3, Fig. 1). Further, regression analysis was carried out considering possible predictors of postoperative OHS including preoperative OHS, BMI, number of comorbidities, length of hospital-stay and preoperative PCS score. The only variable that had a statistically significant relationship with the postoperative OHS was PCS (P = 0.0207). The regression coefficient for the PCS was −0.25, therefore for each unit increase in the preoperative PCS score there was a 0.25 unit decrease in the postoperative OHS score (Table 4).

Close

Fig. 1 Scatter plot of the correlation between PCS and OHS improvement (r = - 0.248, P = 0.0114).

Export to PPT

4 Discussion

The concept of pain catastrophising emerged in the 1980s from extensive research on coping strategies in patients with chronic pain as the tendency to “catastrophise” in response to pain was associated with poorer adjustment to chronic pain particularly for patients with low back pain.¹³ The perceived control over pain, catastrophising, and perceived disability were consistently related to physical functioning.¹⁴ In the field of pain medicine, there has been a recent shift towards procedure-focused interventional pain medicine where cognitive factors, including pain catastrophising, predict pain-related distress and disability to a greater extent than medical status.¹⁵ High levels of pain catastrophising are associated with poorer outcomes following lumbar spine surgery,¹⁶ post-surgical knee pain¹⁷ and even in healthy individuals after a laboratory pain induction procedure.¹⁸

The reported prevalence of persistent postoperative pain after total hip arthroplasty ranged between 27% and 38% of the patients, depending on the definition used.^19–21 Fortunately, however, only a minority of patients report persistent hip pain of moderate or severe intensity. In this study, we examined the relationship between preoperative pain catastrophising and a widely-used postoperative patient-reported outcome measure (OHS). Our findings suggest a negative correlation between the two. Similar studies are scarce in the literature for direct comparison, however these findings are in keeping with those found in knee arthroplasty populations.^10,17,19 A number of possible cofounders were also explored in this study and weak correlations were found including BMI, number of comorbidities and length of inpatient stay. These findings add to the growing evidence that high BMI is a predictor of chronic postoperative pain.^19–21 However, its relation to pain catastrophising is not entirely clear.

Patients satisfaction rate following hip arthroplasty is higher than knee arthroplasty. In a recent systematic review of 33 studies, average satisfaction rate following hip arthroplasty was 93%.²² Kahlenberg et al.²³ reviewed 208 studies (95,560 patients) and found 83% of studies reported 80% patient satisfaction following primary knee arthroplasty. Hence, historically, the research focus has been on knee arthroplasty populations. Forsythe et al.²⁴ found that preoperative PCS score, in primary knee arthroplasty patients, predicted the presence of pain at 24 months postoperatively. Vissers et al.,¹⁰ and despite variability in the evidence reviewed (35 studies), found moderate level evidence for pain catastrophising as an independent predictor of chronic pain following knee arthroplasty. Others have also reported similar findings in knee arthroplasty patients.²⁵ Our findings add to the limited evidence in the hip arthroplasty population.

The research literature on chronic pain indicates that cognitive variables, including pain catastrophising, can predict functional/physical impairment to a larger extent than medical factors.¹⁵ More importantly, however, participation in a comprehensive rehabilitation programme can result in significant decreases in pain catastrophising, and this decrease was found to be a significant mediator of the corresponding improvement in treatment outcomes.²⁶ Research is limited on whether perioperative interventions can improve clinical outcomes in arthroplasty patients with high levels of pain catastrophisation. We are unaware of any trials in the hip arthroplasty patients. However, few interventions have been studied in knee arthroplasty patients. In a double-blinded randomised controlled trial (RCT), Lunn et al. evaluated the use of escitalopram (selective serotonin reuptake inhibitor) on reducing early postoperative pain in knee arthroplasty patients with high PCS scores (n = 120; PCS>21). Pain with ambulation 24 h after surgery was the primary outcome. Secondary outcomes included overall pain during ambulation up to 6 days. The found that pain upon ambulation and at rest was lower in the escitalopram group on postoperative days 2 through 6.²⁷ Rakel et al., in their RCT (n = 317 patients), evaluated the use of transcutaneous electrical nerve stimulation (TENS) in reducing early postoperative pain and hyperalgesia in knee arthroplasty patients and its relation to PCS score. They found that TENS-group patients with low PCS scores had a greater reduction in pain with range of motion at 6 weeks compared to TENS-group patients with high PCS scores.²⁸

Other interventions for patients with high levels of pain catastrophising include biopsychosocial treatment approaches such as specialist physiotherapy, cognitive behaviour therapy and other multidisciplinary interventions with pain specialists.^29,30 Therefore, from an arthroplasty surgeon's point of view, identifying patients with high levels of preoperative pain catastrophising and early referral to pain specialists may help improve their long-term outcomes following hip arthroplasty.

The findings of this study are limited by the small sample size and the relatively short follow up period at 12-months. However, in this pilot study we demonstrated a relationship between pain catastrophisation and postoperative OHS, a widely used PROM for hip arthroplasty and a routinely collected parameter as part of the National Joint Registry data collection. We also demonstrated that PCS score can be introduced as part of the preoperative assessment process. Future long-term studies are needed to evaluate whether the effect of pain catastrophising on functional outcomes persist beyond short-to medium term. Secondly, to identify a cut-off point for hip arthroplasty patients that would warrant further assessment and management. Finally, further research is needed to identify clinical interventions, pharmacological or biopsychosocial, that may improve the outcomes of patients with high levels of pain catastrophisation undergoing hip arthroplasty.

To conclude, chronic pain post total hip arthroplasty is multifactorial. However, high level of pain catastrophising appear to predict poorer postoperative patient-reported outcome measures. Further research is needed to evaluate the value of early identification of high-risk patients and the role of preoperative involvement of pain specialists and its effects on postoperative outcomes.