Table 1.

Study question and conclusions.

Author	Study Design	Level of Evidence	Study Question	Study Methods	Study Results
Ali et al.15	Prospective observational study	II	Is there a correlation between preoperative psychological factors and postoperative satisfaction after TKA?	186 patients who underwent primary TKA filled out Hospital Anxiety and Depression Scale, Visual Analog Pain Scale (VAS) (0–100), and Knee Injury and Osteoarthritis Outcome Score (KOOS) preoperatively and 4 years postoperatively.	15% of the included patients reported that they were dissatisfied or uncertain with the result of their TKA. Patients with preoperative anxiety or depression had 6 times higher risk of being dissatisfied after TKA compared to those patients without (P < 0.001).
Bierke et al.16	Prospective observational study	II	What is the influence of pain catastrophizing, anxiety and depression symptoms, and somatization dysfunction on the outcome of TKA at mid-term follow-up?	150 patients were asked to fill out preoperative questionnaire assessing mental parameters such as pain catastrophizing (Pain Catastrophizing Scale), anxiety (State-Trait Anxiety Inventory) (STAI), depressive symptoms and somatization dysfunction (Patient Health Questionnaire). They were then assessed on postoperative pain on numerical rating scale (NRS).	At 5-year follow-up, patients with depressive symptoms and somatization dysfunction had significantly higher pain level both at rest and while walking (P < 0.001) compared to patients without these symptoms.
Bistolfi17	Prospective observational study	II	How does depression influence pain perception after total knee arthroplasty at short term follow-up?	67 patients who underwent primary TKA were evaluated preoperatively for depression using the Hamilton depression rating scale (HDRS) and pain using NRS. Postoperative outcomes were measured using Hospital for Special Surgery (HSS) knee score.	Patients who scored in the mild depression range on the HDRS scale were found to be more likely to report more pain (P < 0.001) and have lower HSS score at 1 year postoperatively (P < 0.001) compared to those who scored in the normal range on the HDRS scale.
Blackburn et al.18	Prospective observational study	II	Is there an association between anxiety/depression and knee pain/function after TKA?	40 patients undergoing primary TKA completed Hospital Anxiety and Depression (HAD) scale and Oxford Knee Scores (OKS) preoperatively and postoperatively at 3 months and 6 months.	There was significant improvement in HAD scores postoperatively at 3 months and 6 months. Severity of preoperative anxiety and depression was associated with higher levels of knee disability preoperatively (P = 0.009) and the improved level of anxiety and depression postop was associated with reduction in knee disability at 3 months (P = 0.003) and 6 months (P = 0.006).
Duivenvoorden et al.19	Retrospective database study	III	What is the relationship between preoperative anxiety and depression on patient-reported outcome measures (PROMs) after THA and TKA?	149 THA and 133 TKA patients were asked to fill out HAD scale, KOOS, or Hip disability and Osteoarthritis Outcome Score (HOOS) preoperatively and postop 3 months and 12 months after TJA.	The prevalence of anxiety decreased significantly from 27.9% to 10.8% at 12 months postoperatively in hip patients (P < 0.0001), and from 20.3% to 14.8% in knee patients (P < 0.05). Depressive symptoms decreased significantly from 33.6% to 12.1% at 12 months postoperatively in hip patients (P < 0.0001), and from 22.7% to 11.7% in knee patients (P < 0.0001). However, in both THA and TKA patients, preoperative depressive symptoms predicted smaller changes in HOOS or KOOS subscales and patients were less satisfied 12 months postoperatively.
Etcheson et al.20	Prospective observational study	II	What is the pain perception and opioid consumption among patients with and without major depressive disorder (MDD) who underwent THA or TKA?	48 patients with MDD underwent THA and 68 patients with MDD underwent TKA. Opioid consumption and pain levels of these patients were compared to 45 patients without MDD who underwent THA and 61 patients without MDD who underwent TKA.	Patients with MDD who underwent THA or TKA rated higher pain intensity than those without MDD but this was not statistically significant. Patients with MDD who underwent TJA consumed significantly more opioids than those without MDD who underwent TJA (P < 0.05). The length of stay was similar between the two groups.
Fehring et al.21	Retrospective database study	III	Is depression a modifiable risk factor that needs to be corrected prior to TJA?	Patients scheduled for TKA or THA were routinely complete the patient health questionnaire (PHQ-9) preoperatively and 1 year postoperatively. This data was retrospectively analyzed.	65 of 282 patients in the study period scored greater than 10 on PHQ-9, indicating moderate depressive symptoms. Of these patients, 88% improved to <10 postop (P = 0.0012). 10 patients had severe depressive symptoms preoperatively and 9 of them improved to <10 postop (P = 0.10). There were no significant differences in postoperative HOOS and KOOS scores in depressed vs nondepressed patients. Depression does not need to be treated to a certain threshold prior to TJA based on the results of this study.
Filardo et al.22	Retrospective analysis of routinely collected registry data	III	What is the relationship between kinesiophobia and outcome after TKA? Are the outcomes driven by anxiety and depression?	200 patients who underwent TKA were evaluated for kinesiophobia using the Tampa scale and they were evaluated for depression and anxiety the Beck Depression Inventory (BDI) and STAI respectively. Their postoperative outcomes were assess using NRS for pain and clinical outcomes using the Western Ontario and McMaster Universities Arthritis Index (WOMAC) score at 12 months.	While age, body mass index (BMI), education level, number of painful joints, and years of symptoms did not affect outcomes, depressive symptoms and kinesiophobia were synergistically correlated with negative outcome on WOMAC scale (P < 0.0005) at 12-month follow up. This study suggests that kinesiophobia is an independent risk factor for negative outcome after TKA but also has a synergistic effect with depression.
Gold et al.23	Prospective observational study	II	What is the relationship between depression and risk of readmission after TJA?	Retrospective data from California Healthcare Cost and Utilization Project database were analyzed for odds of 90-day readmission after hospital discharge for primary TKA or THA. 132,422 TKA patients and 65,071 THA patients were included for analysis.	Overall 90-day readmission rates were 8% for both THA and TKA. The odds of readmission were 21–24% higher for depressed patients (odds ratio for TKA: 1.21, odds ratio for THA: 1.24, P < 0.001 for both) compared to non-depressed patients.
Gold et al.24	Retrospective cohort study	III	What is the relationship between mental health disorder (substance use, alcohol use, or depression) and postoperative complications after TKA?	A total of 11,403 TKA patients were identified in the database and 2073 of them (18%) had mental health disorders. Univariate and multivariate regression analysis was performed to determine if there was an association between mental health disorders and complications.	Patients with depression were more likely to experience mechanical failure of implants (2.3% vs 1.1% P < 0.001). Patients with depression and substance use disorder were found to have four-fold increased risk for prosthetic joint infection (PJI) compared to substance use disorder alone (P < 0.001).
Halawi et al.25	Retrospective analysis of routinely collected data	III	What is the association between depression and patient reported outcomes after TKA or THA?	469 patients who had undergone elective primary TJA completed the Short Form-12 (SF-12) for assessment of mental health and WOMAC scores at 4 and 12 months postoperatively were analyzed.	Patients with depression but good baseline mental health achieved gains in PROMs similar to those of normal controls (P > 0.05). There was no difference in WOMAC gains between the group with no depression and good mental health and the group with depression and good mental health at 4 and 12 months. The group with no depression and poor mental health had the highest WOMAC gains (p < 0.001). Compared with the group with no depression and poor mental health, the WOMAC gains in the group with depression and poor mental health were significantly lower at 12 months (p = 0.007).
Halawi et al.26	Retrospective analysis of routinely collected registry data	III	Are there any differences in outcomes after TJA between treated and untreated depressed patients?	749 patients who under TKA or THA were asked to self-report a history of depression at the time of surgery. The primary outcomes were measured using SF-12, WOMAC, and University of California Los Angeles (UCLA) activity rating scale 12 months postoperatively.	Compared to patients who reported being treated for depression, those with untreated depression had lower baseline SF-12 mental health scores but there were no differences in PROMs at 12 months postoperatively. Depression treatment did not affect patients' perception of activity limitation.
Hassett et al.27	Retrospective analysis of routinely collected registry data	III	Are there changes in depressive and anxiety symptoms after lower extremity TJA?	1448 patients who underwent primary TJA were evaluated for pain intensity using the Brief Pain Inventory (BPI), functional status using WOMAC, and depressive and anxiety symptoms using HAD scale preoperatively and 6 months postoperatively.	Improvement in pain and physical function from baseline to 6 months postoperatively was associated with improvement in depressive and anxiety symptoms (P < 0.001). Greater decreases in overall body pain were associated with lower depression scores at 6 months (P < 0.001). Even though THA patients reported significantly higher overall body pain scores than TKA patients preoperatively, TKA patients reported significantly higher overall body pain scores than THA patients 6 months postoperatively (P < 0.001).
Hirschmann et al.28	Prospective observational study	II	Do preoperative psychological factors affect subjective and objective outcomes after TKA at short term follow-up?	104 patients undergoing primary TKA were investigated for depression and anxiety using BDI and STAI respectively. The Knee Society Clinical Rating System (KSS) and WOMAC scores were measured preoperatively and postop 12 months after surgery.	More depressed patients showed higher pre and postoperative WOMAC pain scores (P < 0.05) but there was no difference in change in scores compared to non-depressed patients.
Jones et al.29	Prospective observational study	II	Do anxiety and depression levels of patients undergoing TKA affect their outcomes in the mid and long term?	104 patients undergoing TKA were asked to fill out HAD scale to record psychological status and OKS and KSS to record functional outcomes after surgery at 1 year and 7 years postop.	44% of the included patients had abnormal anxiety or depression score. Mean anxiety and depression scores improved at 6 weeks and 1 year postoperatively (P < 0.001). They deteriorated at 7 years postoperatively but not to baseline level prior to surgery (P < 0.001). Similarly, KSS and OKS improved at 6 weeks and 1 year postoperatively and slightly deteriorate at 7 years postop but not to baseline levels. This study showed that TKA positively influences all outcome measures rather than recovery being negatively influenced by preoperative states.
Kohring et al.30	Prospective observational study	II	Is there a difference in patient reported outcome measures between nondepressed, medically treated depressed, and untreated depressed patients undergoing TJA?	A retrospective review of 271 TJA patients was performed from 2014 to 2016. Patient reported outcome measurement information system (PROMIS) scores were measured for patients preoperatively and postoperatively at 1 year.	Untreated depressed patients experienced smaller gains in physical function scores compared to nondepressed patients (P = 0.020). Furthermore, untreated depressed patients experienced smaller gains in physical function compared to treated depressed patients (P = 0.015). This study concluded that medical treatment of depressed patients may lead to outcomes equivalent to nondepressed patients after TJA.
Namba et al.31	Retrospective cohort study	III	What preoperative factors increase the risk of increased opioid consumption after TKA?	A retrospective review of 23,726 patients who underwent TKA was conducted. Multivariate regression analysis was performed to determine risk factors for greater opioid use.	Depression, in addition to other factors, was associated with higher number of opioid prescriptions (Odds ratio: 1.20, P-value < 0.001) at 6–9 months postoperatively.
Pan et al.32	Retrospective cohort study	III	What is the relationship between psychiatric disorders and postoperative outcomes in patients undergoing primary TKA?	National Inpatient Sample (NIS) data was used to analyze 7,153,750 who underwent TKA in the US between 2002 and 2014. Multiple logistic regression analysis was used to determine whether psychological comorbidities were independent risk factors for postoperative complications and pain.	Patients with depression had higher cost compared to patients without ($45,883.02 vs $44,573.65, P < 0.001). Patients with depression had a shorter hospital stay than nondepressed patients (3.42 vs 3.41, P < 0.001). Depression patients had higher odds ratios (ORs) for acute renal failure, anemia, myocardial infarction, pulmonary embolism, pneumonia, and stroke but lower ORs for cardiac and central nervous system complications and inpatient mortality compared to nondepressed patients (P < 0.05 for all).
Perez-Prieto et al.33	Retrospective analysis of registry data	III	What is the quality of life, function, pain, and satisfaction in patients with and without depression undergoing TKA surgery at short term follow-up?	716 patients were enrolled into depressed (200) and nondepressed (516) cohorts. Patient outcomes were measures using Short Form-36 (SF-36), KSS, WOMAC, and VAS scores preoperatively and 1 year postop.	Depressed patients had poorer baseline scores on almost every scale and had lower function scores (KSS, WOMAC) at 1 year postop compared to non-depressed patients (P = 0.001). Depressed patients had higher VAS scores at 1 year postop compared to nondepressed patients (P = 0.002). 86.8% of depressed patients had lower scores on the Geriatric Depression Scale (GDS) 1 year after surgery, suggesting that they “had recovered” from their depression (P < 0.001).
Petrovic et al.34	Prospective cohort study	II	What are the risk factors for development of postoperative pain following THA?	90 patients with complaints of postoperative pain following cemented THA were divided into two groups based on postop pain: patients scoring >5 on NRS, patients scoring less than 5 on NRS (control). Hamilton scale for anxiety and depression were used for psychological evaluation.	Factors associated with development of postoperative pain were depressive symptoms (OR 7.33, P = 0.030), in addition to anxiety (OR 6.01, P = 0.009), preoperative severe pain (OR 2.64, P < 0.001), female gender (OR 4.91, P < 0.001) and type D personality (OR 2.81, P = 0.030).
Rasouli et al.35	Retrospective case control study	III	What is the impact of concomitant psychiatric disorders on hospitalization charges and complications in patients undergoing TJA?	TJA database was queried for patients who underwent THA or TKA in 2009. Records of patients with clinical diagnosis of depression or anxiety were analyzed for hospital charges, length of stay and complications.	There were 128 TKA patients and 62 THA patients who had preoperative clinical diagnosis of anxiety or depression. Rate of complications was higher in patients with preoperative anxiety or depression compared to controls (29% vs 15.5%, P < 0.001). These complications included postoperative anemia, infection, hematoma, wound dehiscence, and device-related issues. There was no difference in length of stay between the two groups. Hospitalization charges were higher in depressed TKA patients vs nondepressed TKA patients ($55,670 vs $52,270, P < 0.001) but there was no difference in charges for THA patients with or without depression.
Riddle et al.36	Retrospective analysis of routinely collected database data	III	Are specific psychological disorders or pain-related beliefs associated with poor outcome after TKA?	140 patients undergoing TKA were asked to complete PHQ-8 for evaluation of depression severity and Generalized Anxiety and Pain Disorder modules from PRIME-MD. Patients also completed Tampa scale for kinesiophobia, Arthritis Self-efficacy Scale, Pain Catastrophizing Scale, WOMAC pain and function questionnaires preoperatively and 6 months postoperatively.	Pain catastrophizing was the only consistent predictor of poor WOMAC pain outcome (P = 0.005). Psychological disorders such as depression and anxiety were not associated consistently with poor WOMAC pain and no psychological predictors were associated with poor WOMAC function.
Riediger et al.37	Prospective observational study	II	What is the relationship between depression, somatization, and pain beliefs on outcomes after THA?	79 patents who were scheduled for elective THA were asked to fill out HAD scale, pain beliefs questionnaire (PBQ), Screening of Somatoform Disorders (SOMS-2), SF-36, and WOMAC preoperatively and 8 weeks postoperatively.	Depressed patients had lower WOMAC preoperatively (30 vs 45, P < 0.05) and postoperatively (72 vs 85, P < 0.05) compared to nondepressed patients but the change in WOMAC scores was similar between the two groups.
Schwartz et al.38	Prospective observational study	II	Does psychotherapy prior to surgery have an effect on outcomes after THA in depressed patients?	Retrospective review was performed of 3 patient cohorts who underwent THA: no depression (158,427), depression and psychotherapy (1919), and depression without psychotherapy (10,912). Outcomes were analyzed to assess resource utilization, surgical and medical complications, narcotic use, and 1 and 3-year revision rates.	Depressed patients who did not receive psychotherapy were more likely to be discharged to inpatient rehab facility (IPR) (OR 1.28, P < 0.001), require 2 or more narcotic prescriptions (OR 1.2, P = 0.004), have continued narcotic requirements 1 year after surgery (OR 1.23, P < 0.001), undergo revision at 1 year (OR 1.74, P = 0.006) and 3 years (OR 1.92, P = 0.021) compared to depressed patients who received psychotherapy. When compared to nondepressed patients, patients with depression (whether or not they received psychotherapy) had greater odds of presenting to ED, nonhome discharge. Depressed patients without psychotherapy had higher odds of readmission at 30 days and 90 days and prolonged length of stay compared to nondepressed patients (P < 0.05). This was not true for depressed patients with psychotherapy.
Stundner et al.39	Retrospective cohort study	III	What is the relationship between depression/anxiety and postoperative outcomes after TJA?	National Inpatient Sample data was used to analyze 1,212,493 patients who underwent TJA between 2000 and 2008. Multiple regression analysis was performed to determine if depression or anxiety were risk factors for certain outcome measures.	Patients with depression had higher hospital charges ($14,438 vs $13,981, P < 0.001), greater length of stay (4.03 vs 3.93, P < 0.001), and higher rates of non-routine discharge (P < 0.0001) compared to patients without depression. The depressed group preoperatively had lower odds of in-hospital mortality compared to nondepressed group (OR 0.53, P = 0.0147). The risk of developing a major complication was also lower in patients with depression compared to those without (OR 0.95, P = 0.0738).
Tarakji et al.40	Retrospective analysis of registry data	III	Does arthroplasty improve patients' depressive symptoms by relieving pain?	146 patients who underwent primary TKA or THA were retrospectively analyzed. They were grouped into depressed and nondepressed groups based on their preoperative mental component summary (MCS), with MCS < 42 defining depression. Their outcomes were evaluated using the SF-36 results at 3 months and 1 year postoperatively.	The MCS scores for depressed group improved from their level, with mean change of 10.76 (P < 0.001). Proportion of depressed group with MCS <42 decreased from 100% to 33.3% 1 year after surgery. In contrast, the change in MCS score for nondepressed group was not significant. The depressed group had significantly lower physical function subscale and role physical subscale scores on SF-36 compared to nondepressed group (P < 0.001) but the magnitude of change in scores after surgery remained the same between the two groups.
Torres-Claramunt et al.41	Retrospective cohort study	III	Do depressed patients feel more pain in the immediate postoperative period after TKA compared to non-depressed patients?	803 patients undergoing primary TKA were asked to complete Geriatric Depression Scale (GDS) preoperatively and they were divided into depressed (GDS > 5) or nondepressed (GDS < 5) groups. The patients completed KSS, VAS, WOMAC, SF-36 forms preoperatively and 1 year after surgery.	48 patients out of 803 included were depressed. Depressed patients scored worse in functional outcome scores preoperatively and 1 year after surgery (P = 0.00) but the improvement obtained was similar between the two groups with the exception of mental domain of SF-36. Depressed patients obtained more improvement in the mental domain of SF-36 than nondepressed patients (P < 0.001).
Vakharia et al.42	Prospective cohort study	II	Do depressed patients undergoing primary TKA have longer inpatient length of stay (LOS), higher readmission rates, medical complications, and higher costs compared to nondepressed patients?	23,061 depressed patients were matched to 115,015 nondepressed patients. Primary outcomes such a LOS, 90-day readmission rates, medical complications, and costs were analyzed for the two groups.	Patients with depressive disorders had longer LOS (6.2 days vs 3.1 days, P < 0.0001), higher odds of readmission (15.5% vs 12.1%, P < 0.001), medical complications (5% s 1.6%, P < 0.0001), higher day of surgery costs ($12,356.59 vs $10,487.71, P < 0.0001), and 90-day costs ($23,386.17 vs $22,201.43, P < 0.0001) compared to nondepressed patients.
Visser et al.43	Retrospective analysis of routine collected database data	III	What is the impact of major depressive disorder on functional outcomes after TKA?	260 patients undergoing TKA were asked to complete SF-36, WOMAC, KSS, PHQ preoperatively and 12 months postoperatively. Multiple regression was performed to see if there was any correlation between depressive symptoms and outcomes.	Both depressed and nondepressed patients had similar improvement in functional scores at 1 year postoperatively from baseline. Depression was associated with poorer preoperative and postoperative KSS and WOMAC scores (P < 0.05).
	Prospective cohort study	II