Abstract
Background:
Psychological distress is presumed to be an important factor that can adversely impact the outcome of orthopaedic procedures. The Distress and Risk Assessment Method (DRAM) is an evaluation tool which assesses psychological distress in patients with low back pain. The purpose of this prospective study was to assess the influence of preoperative psychological distress, as determined by the DRAM score, on the functional outcomes of total joint arthroplasty (TJA).
Materials and methods:
A prospective study of 61 TJAs was performed at a single institution. The DRAM questionnaire and a variety of functional measures (12-Item Short Form Health Survey [SF-12], visual analogue scale [VAS], Oxford Hip Score [OHS], Oxford Knee Score [OKS], and Oswestry Disability Questionnaire [ODQ]) were administered to the patient at baseline, 1 month, and 6 months postoperatively. Mixed model regressions and Mann-Whitney tests were utilised to evaluate the relationship of the DRAM score with functional outcomes.
Results:
The summed quantitative DRAM score was predictive of functional outcomes. With each 1 point increase in psychological distress, VAS pain increased by 0.023 (p = 0.015), OKS decreased by 0.34 (p = 0.01), ODQ increased by 0.065 (p = 0.02), and MCS decreased by 0.14 (p = 0.015). In addition, patients with lower preoperative distress scores had higher rates of improvement than patients with higher preoperative distress scores for VAS pain (p = 0.034).
Discussion:
Psychological distress was associated with decreased baseline mental health and function in the early postoperative period, which has important implications for bundled payments. However, patients with psychological distress still demonstrated functional improvements and TJA should thus not be contraindicated in these patients.
Introduction
Total joint arthroplasty (TJA) remains 1 of the most successful procedures in orthopaedics and has been proven to achieve durable pain relief and improve function and quality of life. While TJA provides many patients with symptomatic relief, a minority of patients will achieve less than optimal results.1,2 As the osteoarthritic population increases and more patients opt for surgical intervention, it has become increasingly important to identify risk factors that may compromise the outcome of this otherwise successful surgical procedure.
There are numerous factors that are known to influence the outcome of surgical procedures. These include demographics (age, gender), vocational issues (history of disability, jobs that require heavy lifting, job satisfaction), lifestyle (weight, smoking status, fitness), sociologic factors, medical comorbidities, and surgical factors (individual surgeon, technique). 3 Preoperative psychological status has also been found to be one of the strongest determinants of functional outcomes following total hip and knee replacement.4,5 The contribution of preoperative psychological factors, particularly distress, has been a particular point of interest in orthopaedics. 6 Psychological distress is a general term used to describe a variety of psychological symptoms including depression, anxiety, poor coping, and somatization. 7
The Distress and Risk Assessment Method (DRAM) score was developed as a practical means of assessing the degree of psychological distress in patients with low back pain. 7 Based on prior studies, the DRAM score has been shown to be accurate and useful in the prediction of outcomes following operative and non-operative management of patients with low back pain.8–11 However, while psychological distress had been frequently studied for spinal surgery, its usefulness in total joint arthroplasty (TJA) is unclear.3,12 The impact of psychological distress as a predictor of outcome following TJA has not been well studied. Currently used questionnaires to determine psychological distress, pain and function scores demonstrate inconsistent and discordant findings.12–15 The DRAM score has been used to correlate higher degrees of distress with lower preoperative modified Harris Hip Scores and Hip Outcome score-sports scores in patients with femoroacetabular impingement, 16 but, to our knowledge, it has not been utilised to characterise the degree of distress in TJA patients. The aim of this prospective study was to assess the influence of preoperative psychological distress, as determined by the DRAM score, on functional outcomes of TJA.
Materials and methods
Following institutional review board approval, a single institution, prospective observational study of primary total knee and hip arthroplasty patients was performed in 2014. Patients with avascular necrosis, septic arthritis, or other non-osteoarthritic etiologies, simultaneous bilateral arthroplasties, drug abuse, less than 18 years of age, and inability to complete the survey due to language barriers or insufficient mental capacity were excluded. Standard TJA postoperative care was given to all patients. Patients undergoing TKA were given routine prescriptions for outpatient PT unlike THA.
Data collection procedures and data elements
Independent research assistants recruited eligible patients preoperatively at a single institution. All components of the questionnaire were administered by the independent researchers preoperatively and at 1 month and 6 months postoperatively. In patients that were unable to attend their scheduled postoperative appointment, telephone interviews were conducted to administer the appropriate questionnaires. The questionnaires completed by the participants included the 12-Item Short Form Health Survey (SF-12), DRAM, visual analogue scale (VAS), Oxford Hip and Knee Score (OHS, OKS) and Oswestry Disability Questionnaire (ODQ). A higher score for VAS is indicative of greater pain, while lower scores on the ODQ, and increased scores on the OKS, OHS, SF-12 Physical Health Composite Scale (PCS), and Mental Health Composite Scale (MCS) reflect greater functional outcomes. The MCS and PCS portions of the SF12 were tabulated and analysed separately. Variables including age, sex, and body mass index (BMI) were recorded for each patient (Table 1). Such variables were not further substratified and analysed.
Patient demographics.
DRAM, Distress and Risk Assessment Method; TJA, total joint arthroplasty; TKA, total knee arthroplasty; BMI, body mass index; SD, standard deviation; ODQ, Oswestry Disability Questionnaire; OHS, Oxford Hip Score; OKS, Oxford Knee Score.
The DRAM survey is a validated 45-item questionnaire consisting of 2 parts, the Modified Somatic Perception Questionnaire (MSPQ) and the Modified Zung Depression Index (MZDI). 7 It is scored on a 0 to 100 scale with higher scores reflecting greater psychological distress. Scores can be further classified into “normal” and “abnormal” (here “abnormal” includes “at risk”, “distressed depressive” and “distressed somatic”). Analysis was performed using the aforementioned tiers and the quantitative DRAM score consisting of the sum of MSPQ and MZDI. Furthermore, statistical analysis was performed for differences between “normal” and “abnormal” (distressed/at-risk) scoring patients.
Statistical analysis
Power analysis was based on a moderate to large effect size, where the difference is 70% of the standard deviation within groups, a 2-sided t-test, and a p-value of 0.025 (0.05/2) to compensate for 2 comparisons. We therefore needed 40 patients to obtain power of 80% for this study. We included more patients in the study to cover possible loss to follow-up. The prior study by Hobby et al. 7 that found improvement in mean of preoperative DRAM score by -14.1 (Zung) and -5.3 (MSPQ) does not include the standard deviations necessary for a more exact power analysis.
Using R software 2.15.1 (R Foundation for Statistical Computing, Vienna, Austria), the statistical analysis was performed when the outcomes were available for 53 subjects including those lost to follow-up as per a priori power analysis. An alpha level of 0.10 was used to determine significance. Multiple linear regressions were performed to evaluate the relationship of the quantitative DRAM score, controlling for baseline DRAM score, with each of the 6 analysed postoperative endpoints: VAS, ODI, OKS, OHS, MCS, PCS. In addition, a series of mixed models with repeated measures (baseline to 6 month) were run to determine if the rate of change of functional outcomes varied significantly based on the DRAM baseline score.
Finally, within the 2-tier categorisation of the DRAM baseline score (“normal” and “abnormal”), Mann-Whitney non-parametric tests were used to compare functional outcomes, with an MZDI < 17 as the cutoff for “normal” patients.
Results
Between May 2014 and August 2014, 61 patients were recruited into the study with 53 patients completing both preoperative and all postoperative questionnaires. The summed quantitative DRAM score was predictive of functional outcomes at most postoperative endpoints. Per 1 point increase in the overall psychological distress score, VAS pain increased by 0.023 (p = 0.015, Figure 1), OKS decreased by 0.34 (p = 0.01, Figure 2), ODI increased by 0.065 (p = 0.02, Figure 3), and MCS decreased by 0.14 (p = 0.015, Figure 4). However, there was no relationship between PCS (-0.193, p = 0.166) and OHS (0.024, p = 0.366) with the overall DRAM score. Furthermore, the preoperative DRAM score was also predictive of the postoperative DRAM component MZDI (p = 0.009). In addition, patients with lower distress scores had higher rates of improvement than patients with higher distress scores for VAS pain (p = 0.034).
6-month postoperative VAS score as a function of baseline DRAM score.
6-month postoperative OKS score as a function of baseline DRAM score.
6-month postoperative ODQ score as a function of baseline DRAM score.
6-month postoperative MCS score as a function of baseline DRAM score.
When comparing patients with “normal” and “abnormal” (distressed/at risk) MZDI baseline scores, patients with “normal” distress levels had marginally significant higher functional outcomes than patients with “abnormal” DRAM scores for ODQ (p = 0.096) and MCS (p = 0.048) at the 1 month postoperative period, and OKS (p = 0.083) at the 6-month postoperative period. In addition, patients with abnormal scores tend to have less improvement in MCS (p = 0.079) but more improvement in the MZDI score (0.011).
Discussion
Utilising a well-validated metric of assessing a psychological distress questionnaire, namely the DRAM, we aimed to better elucidate the influence of psychological distress on pain and function following total joint arthroplasty. Controlling for baseline DRAM score, the amount of distress, assessed by the DRAM, was predictive of the majority of functional outcomes at the 6-month postoperative period. It is important to note that both distressed and non-distressed patients demonstrated marked functional and psychological improvement postoperatively. Furthermore, even patients with “abnormal” distress status, as measured by the DRAM, who had a markedly lower Mental Health Composite Scale (MCS) scores at 1 month (p = 0.048), seemed to normalise their MCS by 6 months postoperatively. The latter finding may indicate that patients with a higher distress score are likely to “struggle” in the postoperative time period and hence require more resources. These patients, however, ultimately reach the same mental health outcome as those without high levels of distress.
The effect of psychological distress has been examined extensively among the spine surgery patient population. Vialle et al. 11 examined a cohort of patients undergoing 1- or 2-level lumbar fusions and used the DRAM to evaluate levels of psychological distress. While postoperative outcomes improved among patients receiving the procedure, those with “abnormal” DRAM scores had significantly lower functional outcomes when compared to patients who did not display signs of psychological distress. Pollock et al. 10 used the DRAM to conclude that patients who are diagnosed as distressed somatic preoperatively have less favorable outcomes from posterior lumbar interbody fusion. Daubs et al. 17 concluded that patients with personality disorders, higher depression and neuroticism should be treated conservatively in lieu of fusion. In contrast, Hobby et al. 7 has suggested that lumbar discectomy outcomes at 6 months are not unfavourably affected by preoperative psychological disturbance as measured by DRAM. Okoro and Sell 18 reported that patients who are identified as “somatized” based on the DRAM score, had similar improvements in the ODI compared to the “non-somatized” patients. A review on psychosocial variables on outcome of lumbar surgery and spinal cord stimulation suggests that psychological factors are predictors of poor outcome, but that there is little empirical evidence that psychological screening helps to improve treatment outcome. 17
While the influence of psychological factors on clinical outcomes in TJA is less well known than in the spine literature, a few studies have been performed. These studies showed that patients with symptoms of anxiety, depression, and poor coping skills experience less improvement after TJA.4,13,19 However, many psychologically distressed patients may experience some improvement in mental distress following surgery. Lingard and Riddle 15 demonstrated that patients with psychological distress have a substantial decrease in stress following surgery despite the fact that postoperative pain may be slightly elevated for up to 2 years following hip arthroplasty. Hossain et al. 12 concluded that patients with elevated psychological distress did not have any difference in patient satisfaction compared to patients who did not exhibit significant levels of distress.
The findings of our study should be interpreted with caution due to the following limitations. In order to examine a variety of outcomes, the numerous questionnaires administered required a significant time commitment from the patients involved. Thus, the extensive length of the surveys could have caused some patients to elect not to continue with the study and also could have contributed to the lack of patient follow-up in some instances. Also, the relatively small sample size limited our ability to dissect all factors that might have contributed to the patients’ underlying psychological distress. Specifically, the small sample size limited our ability to further break down patients into four individual tiers (“normal”, “at risk”, “distressed depressive” and “distressed somatic”) as defined by DRAM scoring guidelines. In particular, we had minimal numbers of patients in the “distressed depressed” and “distressed somatic” groups which limits the conclusions that we can reach about the heavily distressed patients. In addition, while numerous interactions were found between psychological distress score and functional outcomes for both THA and TKA, interactions between DRAM scores and THA specific outcomes were more prevalent. In particular, a point increase in psychological distress score was associated with a decrease in OKS, but not with OHS. However, this specific outcome was most likely due to relatively small sample size for THA in our study compared to TKA as our sample size was further stratified (23 versus 30, respectively). In addition, THA generally is associated with better outcomes so there may be a “ceiling effect,” particularly with SF-12 scores. Additionally, only short-term outcomes were obtained and the impact of psychological distress on outcomes of longer duration cannot be extrapolated.
In summary, the present study shows a relationship between psychological distress and decreased baseline mental health and function in the early postoperative period raising the possibility that more resources may need to be allocated to these patients at such time periods during rehabilitation. This finding may be very important as bundling of care is being introduced for patients undergoing TJA. Based on the data retrieved by this study we speculate that patients with a higher distress score preoperatively may require more attention during the early postoperative period. These patients may benefit from frequent office visits or phone calls to provide assurance at minimum or possibly even benefit from psychiatric services.
Despite the lower Mental Health Composite Scale (MCS), distressed patients stand to benefit from total joint arthroplasty. However, the summed quantitative DRAM score does suggest that patients with higher dram scores may be at risk for lesser gains in pain and function. Since both groups demonstrated significant improvements following TJA, psychological distress should not be a contraindication to joint arthroplasty in these patients. Further research is needed to ascertain if any of the subsets of psychiatric patients in particular predispose patients to poor outcomes as well as to understand if preoperative or postoperative psychological intervention improves outcome in the appropriate patient.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship and/or publication of this article.