Abstract
Background
The rising prevalence of total knee arthroplasty (TKA) parallels increasing rates of anxiety and/or depression (Anx/Dep), yet the impact on recovery remains poorly defined. This study evaluates how Anx/Dep influences recovery and patient-reported outcomes after TKA, and whether Patient-Reported Outcomes Measurement Information System Global Health Short Form (10 items) (PROMIS-10) Mental Health T-scores provide additional insight.
Methods
A retrospective review of 624 primary TKA patients (July 2023–October 2024) was performed. Patients were grouped by Anx/Dep diagnosis (309 with, 315 without) and stratified by PROMIS-10 Mental Health T-scores into below average mental health T-score (BA-MHT), average mental health T-score (A-MHT), and above average mental health T-score (AA-MHT) groups. PROMIS-10 physical, PROMIS-10 mental, and Knee injury and Osteoarthritis Outcome Score for Joint Replacement (KOOS, JR) scores were collected preoperatively and at 3, 6, and 12 months. Opioid use, pain, length of stay (LOS), and readmission rates were compared. Baseline demographic differences between groups were present.
Results
Anx/Dep patients had significantly longer LOS, greater opioid use before discharge, and higher early postoperative pain. Although KOOS, JR scores were similar at 1 year, Anx/Dep patients reported consistently lower PROMIS-10 mental and physical scores (P < .0001). Mental health stratification revealed BA-MHT patients, regardless of diagnosis, had worse outcomes, including higher opioid use, longer LOS, and lower KOOS, JR and PROMIS-10 scores across all intervals. In contrast, AA-MHT patients demonstrated more favorable recovery.
Conclusions
Anx/Dep is associated with increased early postoperative burden following TKA. PROMIS-10 mental health scores better capture variation in recovery, suggesting that mental health influences outcomes along a continuum rather than a binary diagnosis. Routine PROMIS-10 screening may help identify at-risk patients and guide personalized care strategies.
Keywords: Total knee arthroplasty, Anxiety and depression, Patient-reported outcomes, Opioid consumption, Mental health stratification
Introduction
As total knee arthroplasty (TKA) continues to evolve in safety and efficiency, demand remains high with US annual case volume projected to exceed 1.26 million by 2030, despite growing uncertainty in long-term forecast due to the shifting guidelines and disease prevalence [1,2]. Simultaneously, the prevalence of anxiety and/or depression (Anx/Dep) has surged nationwide, raising concern as these mental health conditions have been linked to poorer surgical outcomes, including lower patient-reported outcomes measures (PROMs), and may increasingly intersect with the expanding TKA population [3,4].
The influence of Anx/Dep diagnoses on postoperative surgical outcomes, specifically in orthopaedics, is a growing area of interest that has brought forth new findings in total joint arthroplasty [5]. Arthroplasty procedures like TKA are linked to poorer postoperative pain, reduced function, higher costs, increased persistent opioid use, more complications, and lower satisfaction in patients with Anx/Dep symptoms compared to those without [[5], [6], [7], [8], [9], [10], [11]]. Consequently, these disparities are further worsened by the growing TKA and Anx/Dep intersection, as the TKA patient population has been shown to experience a twofold increase in preoperative anxiety and a threefold increase in depression [12].
The spike in preoperative Anx/Dep in TKA, coupled with poorer reported outcomes for these patients, demonstrates how closely associated these afflictions are and the great need for more attention around the relationship between Anx/Dep and TKA. Moreover, this focus is especially important given reports of higher dissatisfaction despite improved pain and function with this population and evidence that negative outcomes may be mitigated when mental health illness is appropriately treated postoperatively [10,13]. Despite this, mental health illnesses may be overlooked in TKA when establishing surgical expectations and postoperative follow-up, leaving significant gaps in care for this patient population that may have greater social, psychological, and physical needs.
Understanding both the individual and combined impacts of Anx/Dep on physical and mental health is crucial for enhancing patient recovery. This study examines the impact of mental health diagnoses on recovery from postoperative day 1 through 1 year, focusing on the relationship between Anx/Dep and TKA outcomes using PROMs. Patients are stratified by preoperative Anx/Dep diagnosis and Patient-Reported Outcomes Measurement Information System Global Health Short Form (10 items) (PROMIS-10) Global Mental Health scores, allowing assessment of mental health as both binary and continuous variables [14]. We hypothesize that PROMIS-10 scores will more accurately reflect recovery differences than diagnosis alone. Ultimately, this study highlights the importance of including mental health in recovery planning across the postoperative timeline.
Material and methods
Following institutional review board approval (TOH269e), a single-site electronic medical record (EMR) was queried for patients who underwent primary TKA (Current Procedural Terminology Code - CPT: 27447) between July 1, 2023, and October 1, 2024 (n = 1836). Demographic data including age, sex, race, ethnicity, body mass index, insurance type, hospital length of stay (LOS), 30- and 90-day readmission status, marital status, preoperative opioid usage (yes/no), and current mental health diagnoses were extracted from the EMR. Perioperative data were collected from a local surgical database, which included in-hospital opioid consumption (morphine milligram equivalents) prior to discharge, and pain at 2 hours postsurgery and at discharge.
PROMs, specifically, PROMIS-10 physical, PROMIS-10 mental, and Knee injury and Osteoarthritis Outcome Score for Joint Replacement (KOOS, JR) [15], were collected preoperatively, 3 months, 6 months, and 1 year postoperative. These PROMs were matched with the EMR and perioperative data. Patients were excluded from further analysis if they underwent bilateral TKA (n = 8), were under the age of 18 (n = 0), had an unreported anxiety and depression status (n = 844), had no perioperative data (n = 198), and no PROMs from any time point (n = 162). After applying the exclusion criteria, 624 patients remained in the final analysis.
Patients were analyzed using 2 stratification approaches. First, patients with a diagnosis of Anx/Dep (Anx/Dep cohort, n = 309) were compared with those without either condition (no Anx/Dep cohort, n = 315). Demographic and mental health data for both cohorts are presented in Supplement Tables 1 and 2. Second, patients were further stratified based on their PROMIS-10 Mental Health T-scores (MHT) using established cutoffs [16]: scores <40 were categorized as below average (BA-MHT Anx/Dep; n = 70) (BA-MHT no Anx/Dep; n = 25), scores between 40 and 50 as average (A-MHT Anx/Dep; n = 143) (A-MHT no Anx/Dep; n = 79), and scores >50 as above average (AA-MHT Anx/Dep; n = 79) (AA-MHT no Anx/Dep; n = 189).
Comparisons were performed between the Anx/Dep and no Anx/Dep cohorts, followed by subgroup analyses based on MHT stratification. Descriptive statistics were calculated for all variables of interest. Continuous variables with normal distribution were summarized using means and standard deviations and compared using independent t-tests or one-way analysis of variance, as appropriate. Categorical variables were analyzed using Fisher’s exact test or chi-square tests. Statistical significance was defined as P < .05. All descriptive statistics and tests were calculated using Python and Excel.
Results
A total of 624 patients were included in this study, with 309 patients (49.5%) in the Anx/Dep group and 315 (50.5%) in the no Anx/Dep group. Baseline demographic differences were observed between groups including sex and marital status (Table 1).
Table 1.
Comparison of demographics among the two cohorts.
| Variable | Anx/Dep n = 309 | No Anx/Dep n = 315 | P value | ||
|---|---|---|---|---|---|
| Age, y | 67.18 | 43-89 | 67.98 | 37-91 | .2604 |
| BMI | 33.32 | 17.8-60.4 | 32.32 | 14.8-51.2 | .0690 |
| Sex | |||||
| Male | 79 | 25.57 | 147 | 46.67 | .0001 |
| Female | 230 | 74.43 | 168 | 53.33 | |
| Race | .6050 | ||||
| White | 222 | 71.84 | 227 | 72.06 | |
| Black or African American | 37 | 11.97 | 32 | 10.16 | |
| Hispanic | 23 | 7.44 | 32 | 10.16 | |
| Asian | 7 | 2.27 | 9 | 2.86 | |
| Unknown | 16 | 5.18 | 13 | 4.13 | |
| Other | 4 | 1.29 | 2 | 0.63 | |
| Ethnicity | .6364 | ||||
| Hispanic | 25 | 8.09 | 32 | 10.16 | |
| Not Hispanic | 263 | 85.11 | 264 | 83.81 | |
| Declined | 21 | 6.8 | 19 | 6.03 | |
| Marital status | .0019 | ||||
| Divorced | 43 | 13.92 | 31 | 9.84 | |
| Married | 179 | 57.93 | 232 | 73.65 | |
| Partner | 2 | 0.65 | 1 | 0.32 | |
| Separated | 7 | 2.27 | 0 | 0 | |
| Single | 33 | 10.68 | 25 | 7.94 | |
| Widowed | 43 | 13.92 | 26 | 8.25 | |
| Unknown | 2 | 0.65 | 0 | 0 | |
| Insurance | .4306 | ||||
| Commercial | 96 | 31.07 | 98 | 31.11 | |
| Health exchange/other | 7 | 2.27 | 7 | 2.22 | |
| Medicaid | 7 | 2.27 | 2 | 0.63 | |
| Medicare | 192 | 62.14 | 204 | 64.76 | |
| Self-pay | 0 | 0 | 1 | 0.32 | |
| Other | 7 | 2.27 | 3 | 0.95 | |
| Mental health diagnosis | a | ||||
| Anxiety only | 118 | 38.19 | 0 | 0 | |
| Depression only | 91 | 29.45 | 0 | 0 | |
| Anxiety and depression | 100 | 32.36 | 0 | 0 | |
Age, body mass index (BMI), and length of stay reported as mean (range).
All other data reported as frequency (percentage).
BMI, body mass index.
Values in bold represent statistical significance.
Not included in P value.
Patients in the Anx/Dep cohort had a modestly longer hospital stay compared to those without Anx/Dep (P < .05) while 30- and 90-day readmission rates were similar between the 2 groups (Table 2). The Anx/Dep cohort also demonstrated higher in-hospital opioid consumption and greater early postoperative pain at 2 hours following surgery (both P < .05). Pain scores at discharge did not differ between cohorts.
Table 2.
Comparison of early postoperative outcomes following TKA in Anx/Dep and no Anx/Dep.
| Variable | Anx/Dep n = 309 | No Anx/Dep n = 315 | P value | ||
|---|---|---|---|---|---|
| Preoperative L use | |||||
| Yes (%) | 104 (35.62) | 119 (39.02) | .3906 | ||
| No (%) | 188 (64.38) | 186 (60.98) | |||
| Total MMEs prior to discharge | 52.60 ± 51.34 (45; n = 295) | 41.07 ± 72.57 (39; n = 284) | .0274 | ||
| Pain at 2 h (VAS) | 5.98 ± 2.44 (3; n = 306) | 5.22 ± 2.76 (4; n = 314) | .0003 | ||
| Pain at discharge (VAS) | 5.30 ± 2.20 (3; n = 307) | 4.99 ± 2.43 (4; n = 314) | .0971 | ||
| Avg. length of stay (d) | 1.03 | 0.21-7.29 | 0.90 | 0.12-11.08 | .0359 |
| ≥1 d | 184 | 59.55 | 170 | 53.97 | .1596 |
| <1 d | 125 | 40.45 | 145 | 46.03 | |
| Readmissions | n = | % | n = | % | |
| 30 d | 3 | 0.97 | 3 | 0.95 | 1.0 |
| 90 d | 20 | 6.47 | 26 | 8.25 | .3945 |
Current display: Average ± Stdev (interquartile range; count).
MMEs, morphine milligram equivalents; VAS, visual analog scale.
Values in bold represent statistical significance.
At 1 year postoperatively, KOOS, JR scores were comparable between Anx/Dep and no Anx/Dep patients (PRO response rate of 76 and 74%, respectively), indicating similar joint-specific functional recovery (P > .05; Table 3). In contrast, PROMIS-10 physical and mental health scores were consistently lower in the Anx/Dep cohort at 3, 6, and 12 months (all P < .001).
Table 3.
PROMs of patients in two cohorts.
| Interval | Anx/Dep | No Anx/Dep | P value |
|---|---|---|---|
| KOOS, JR | |||
| Preop | 47.31 ± 12.70 (n = 277) | 50.62 ± 11.74 (n = 281) | .0015 |
| 3 mo | 67.59 ± 13.96 (n = 247) | 69.55 ± 13.42 (n = 263) | .1070 |
| 6 mo | 69.79 ± 14.65 (n = 247) | 72.71 ± 15.25 (n = 254) | .0282 |
| 12 mo | 72.77 ± 15.90 (n = 234) | 74.57 ± 14.99 (n = 234) | .2088 |
| PROMIS-10 physical | |||
| Preop | 38.51 ± 6.81 (n = 292) | 42.27 ± 6.66 (n = 293) | < .0001 |
| 3 mo | 44.12 ± 7.94 (n = 260) | 47.10 ± 7.94 (n = 284) | < .0001 |
| 6 mo | 44.40 ± 7.90 (n = 260) | 47.87 ± 7.89 (n = 278) | < .0001 |
| 12 mo | 44.57 ± 8.49 (n = 246) | 48.20 ± 8.85 (n = 255) | < .0001 |
| PROMIS-10 mental | |||
| Preop | 44.80 ± 7.94 (n = 292) | 52.24 ± 8.41 (n = 293) | < .0001 |
| 3 mo | 46.34 ± 7.88 (n = 260) | 52.51 ± 8.66 (n = 284) | < .0001 |
| 6 mo | 46.39 ± 8.41 (n = 260) | 52.89 ± 8.35 (n = 278) | < .0001 |
| 12 mo | 46.93 ± 8.52 (n = 246) | 53.19 ± 8.61 (n = 255) | < .0001 |
Current display: Average ± Stdev (count).
Values in bold represent statistical significance.
MHT subgroup analysis
Patients in both cohorts were further stratified into above average (AA-MHT), average (A-MHT), and below average (BA-MHT) mental health groups based on PROMIS-10 T-scores and demonstrated different recovery outcomes. Within the Anx/Dep cohort, early postoperative outcomes, in-hospital opioid consumption, pain scores, and 30- and 90-day readmission rates, did not differ significantly across mental health strata; however, LOS was significantly longer in the BA-MHT group (Table 4).
Table 4.
Comparison of early postoperative outcomes following TKA for the three Anx/Dep subcohorts.
| Variable | Above-average MHT (n = 79) | Average MHT (n = 143) | Below-average MHT (n = 70) | P value | |||
|---|---|---|---|---|---|---|---|
| Preoperative opioid use | |||||||
| Yes (%) | 35 (44.87) | 41 (30.83) | 23 (32.86) | .1193 | |||
| No (%) | 43 (55.13) | 92 (69.17) | 43 (61.43) | ||||
| Total MMEs | 46.89 ± 55.21 (35; n = 79) | 51.94 ± 52.18 (36.55; n = 143) | 61.23 ± 48.66 (34.35; n = 70) | .2537 | |||
| Pain at 2 h (VAS) | 5.49 ± 2.68 (4; n = 79) | 6.12 ± 2.41 (3; n = 143) | 6.23 ± 2.19 (3; n = 70) | .1211 | |||
| Pain at discharge (VAS) | 5.26 ± 2.29 (4; n = 79) | 5.40 ± 2.19 (3; n = 143) | 5.23 ± 2.13 (3; n = 70) | .8631 | |||
| Avg. length of stay (d) | 0.99 | 27.05 | 0.97 | 48.97 | 1.17 | 23.97 | .2961 |
| ≥1 d | 43 | 54.43 | 77 | 53.85 | 50 | 71.43 | .0112 |
| <1 d | 36 | 45.60 | 66 | 46.15 | 20 | 28.57 | |
| Readmissions | n = | % | n = | % | n = | % | |
| 30 d | 1 | 1.27 | 0 | 0 | 2 | 2.86 | .1471 |
| 90 d | 8 | 10.13 | 7 | 4.90 | 5 | 7.14 | .3337 |
Current display: Average ± Stdev (interquartile range; count).
MMEs, morphine milligram equivalents; VAS, visual analog scale.
Values in bold represent statistical significance.
Among patients without Anx/Dep, PROMIS-10 mental health stratification revealed differences in early postoperative utilization. Patients in the BA-MHT subgroup had longer hospital stays, higher pain scores at 2 hours, and higher in-hospital opioid consumption compared with those with A- or AA-MHT (both P < .01), while readmission rates were similar across strata (Table 5).
Table 5.
Comparison of early postoperative outcomes following TKA for the three no Anx/Dep subcohorts.
| Variable | Above-average MHT (n = 189) | Average MHT (n = 79) | Below-average MHT (n = 25) | P value | |||
|---|---|---|---|---|---|---|---|
| Preoperative opioid use | |||||||
| Yes (%) | 70 (39.33) | 31 (40.26) | 11 (45.83) | .7631 | |||
| No (%) | 113 (63.48) | 46 (59.74) | 13 (54.17) | ||||
| Total MMEs | 35.09 ± 41.67 (33.13; n = 189) | 37.86 ± 25.86 (33.38; n = 79) | 94.86 ± 213.62 (45.5; n = 25) | .0004 | |||
| Pain at 2 h (VAS) | 4.95 ± 2.75 (5; n = 189) | 5.37 ± 2.70 (3; n = 79) | 6.46 ± 2.62 (5; n = 25) | .0180 | |||
| Pain at discharge (VAS) | 4.72 ± 2.36 (4; n = 189) | 5.28 ± 2.49 (4; n = 79) | 5.50 ± 2.41 (3.25; n = 25) | .1084 | |||
| Avg. length of stay (d) | 0.82 | 0.29–2.42 | 0.86 | 0.12- 1.54 | 1.52 | 0.17-11.08 | < .0001 |
| ≥1 d | 92 | 48.68 | 44 | 55.70 | 18 | 78.26 | .0231 |
| <1 d | 97 | 51.32 | 35 | 44.30 | 5 | 21.74 | |
| Readmissions | n = | % | n = | % | n = | % | |
| 30 d | 1 | 0.53 | 1 | 1.27 | 1 | 4.00 | .2609 |
| 90 d | 14 | 7.41 | 7 | 8.86 | 5 | 20.00 | .1148 |
Current display: Average ± Stdev (interquartile range; count).
MMEs, morphine milligram equivalents; VAS, visual analog scale.
Values in bold represent statistical significance.
PROM analyses within the Anx/Dep cohort demonstrated graded differences across mental health strata. Patients with AA-MHT scores reported higher PROMIS-10 physical and mental scores, as well as higher KOOS, JR scores, across all postoperative time points (PRO response rate varying from 68 to 100%) compared with the A- and BA-MHT groups (all P < .001; Table 6). In the no Anx/Dep cohort, PROMIS-10 mental health stratification similarly demonstrated higher PROMIS-10 mental scores among patients with AA-MHT at all time points (P < .001). By 1 year postoperatively, KOOS, JR and PROMIS-10 physical scores did differ significantly across mental health strata, indicating persistent differences in long-term joint-specific functional recovery (P < .05; Table 7).
Table 6.
PROMs of patients in three Anx/Dep subcohorts.
| Interval | Above-average MHT | Average MHT | Below-average MHT | P value |
|---|---|---|---|---|
| KOOS, JR | ||||
| Preop | 50.12 ± 13.65 (n = 75) | 48.04 ± 12.11 (n = 140) | 42.25 ± 11.54 (n = 62) | .0008 |
| 3 mo | 73.70 ± 14.42 (n = 70) | 68.19 ± 12.39 (n = 124) | 60.52 ± 11.99 (n = 53) | < .0001 |
| 6 mo | 76.53 ± 15.04 (n = 68) | 69.47 ± 13.35 (n = 123) | 63.48 ± 13.37 (n = 53) | < .0001 |
| 12 mo | 78.19 ± 16.05 (n = 66) | 72.31 ± 15.04 (n = 114) | 67.07 ± 15.80 (n = 49) | .0007 |
| PROMIS-10 physical | ||||
| Preop | 43.27 ± 6.09 (n = 79) | 38.99 ± 4.83 (n = 143) | 32.15 ± 6.20 (n = 70) | < .0001 |
| 3 mo | 48.98 ± 7.69 (n = 72) | 44.68 ± 6.39 (n = 125) | 37.33 ± 6.36 (n = 59) | < .0001 |
| 6 mo | 49.45 ± 7.08 (n = 72) | 45.08 ± 5.99 (n = 125) | 36.98 ± 7.27 (n = 56) | < .0001 |
| 12 mo | 49.57 ± 7.94 (n = 71) | 44.81 ± 6.76 (n = 117) | 37.84 ± 8.10 (n = 54) | < .0001 |
| PROMIS-10 mental | ||||
| Preop | 54.06 ± 3.95 (n = 79) | 44.70 ± 2.81 (n = 143) | 34.55 ± 5.19 (n = 70) | < .0001 |
| 3 mo | 51.99 ± 6.55 (n = 72) | 47.08 ± 5.67 (n = 125) | 51.99 ± 6.55 (n = 70) | < .0001 |
| 6 mo | 52.63 ± 7.50 (n = 73) | 46.69 ± 5.81 (n = 125) | 37.79 ± 6.93 (n = 56) | < .0001 |
| 12 mo | 52.27 ± 7.78 (n = 71) | 47.44 ± 6.77 (n = 117) | 39.08 ± 6.89 (n = 54) | < .0001 |
Current display: Average ± Stdev (count).
Values in bold represent statistical significance.
Table 7.
PROMs of patients in three no Anx/Dep subcohorts.
| Interval | Above-average MHT | Average MHT | Below-average MHT | P value |
|---|---|---|---|---|
| KOOS, JR | ||||
| Preop | 53.46 ± 10.65 (n = 181) | 46.66 ± 11.88 (n = 76) | 41.79 ± 11.64 (n = 24) | < .0001 |
| 3 mo | 71.70 ± 12.90 (n = 168) | 66.62 ± 12.36 (n = 70) | 61.01 ± 17.38 (n = 20) | .0003 |
| 6 mo | 74.44 ± 15.15 (n = 165) | 70.41 ± 13.44 (n = 64) | 65.85 ± 16.72 (n = 18) | .0237 |
| 12 mo | 76.33 ± 15.04 (n = 152) | 72.21 ± 12.94 (n = 60) | 65.57 ± 17.20 (n = 17) | .0069 |
| PROMIS-10 physical | ||||
| Preop | 44.79 ± 6.06 (n = 189) | 38.45 ± 4.94 (n = 79) | 35.30 ± 4.83 (n = 25) | < .0001 |
| 3 mo | 49.37 ± 7.26 (n = 182) | 43.86 ± 6.25 (n = 73) | 40.02 ± 9.95 (n = 22) | < .0001 |
| 6 mo | 49.81 ± 7.37 (n = 182) | 44.88 ± 6.05 (n = 67) | 41.37 ± 9.46 (n = 21) | < .0001 |
| 12 mo | 50.08 ± 8.14 (n = 165) | 45.62 ± 8.18 (n = 65) | 41.46 ± 10.79 (n = 20) | < .0001 |
| PROMIS-10 mental | ||||
| Preop | 57.16 ± 5.38 (n = 189) | 45.36 ± 2.56 (n = 79) | 36.74 ± 3.90 (n = 25) | < .0001 |
| 3 mo | 55.98 ± 7.30 (n = 182) | 47.02 ± 5.71 (n = 73) | 42.84 ± 8.80 (n = 22) | < .0001 |
| 6 mo | 55.77 ± 7.57 (n = 182) | 47.88 ± 5.43 (n = 67) | 44.62 ± 8.08 (n = 21) | < .0001 |
| 12 mo | 56.27 ± 7.52 (n = 165) | 48.41 ± 5.74 (n = 65) | 43.43 ± 10.31 (n = 20) | < .0001 |
Current display: Average ± Stdev (count).
Values in bold represent statistical significance.
Discussion
This study examined the relationship between Anx/Dep and TKA outcomes, comparing postoperative in-hospital opioid consumption, pain levels, and PROMs among patients with and without these mental health diagnoses. Patients with a diagnosis of Anx/Dep experienced longer hospital stays following TKA compared to those without an Anx/Dep diagnosis. However, there were no differences in readmission rates at 30 or 90 days, suggesting that while Anx/Dep may be associated with increased postoperative care, it does not predict a higher rate of long-term hospital visits. Additionally, patients in the Anx/Dep cohort consumed a higher percentage of opioids postoperatively prior to discharge and reported greater pain levels at 2 hours postoperation. By the time of discharge, pain scores were similar between groups; however, this was achieved with higher in-hospital opioid consumption in the Anx/Dep cohort.
Functionally, both groups demonstrated improvements in knee function, as reflected by similar KOOS, JR scores at 1 year postoperatively. However, patients without an Anx/Dep diagnosis consistently reported higher PROMIS-10 mental and physical health scores throughout the year following surgery, indicating that patients with Anx/Dep may view their overall recovery less favorably despite positive functional outcomes. These findings can help orthopaedic surgeons reassure patients with anxiety or depression that their diagnosis is unlikely to negatively impact long-term pain or functional outcomes following TKA.
Our results both support and differ from existing literature on postoperative outcomes among patients with mental health diagnoses in orthopaedic populations. A 2024 study evaluating depression in TKA reported longer hospital LOS along with higher odds of 7- and 30-day readmission and increased medical and surgical complications among depressed patients [11]. Conversely, Tamate et al. reported no significant differences in opioid consumption or hospital LOS between patients with high and low Global Mental Health scores following total hip arthroplasty, and satisfaction scores at 6 weeks postoperation were similar between groups [17]. Furthermore, their study identified that patients who were diagnosed with anxiety or depression reported similar postoperative PROMs. Consistent with our broader interpretation that mental health influences recovery experience beyond joint-specific function, a 2024 study on the impact of patient-reported Anx/Dep on TKA found that patients reporting both preoperative and postoperative Anx/Dep experienced increased rates of dissatisfaction following TKA, despite improvements in pain or function [13].
Our findings further show that mental health exists on a continuum and its influence on postoperative outcomes may depend on a variety of external factors within the context of an Anx/Dep diagnosis. Among patients with an Anx/Dep diagnosis, those with below average mental health scores reported significantly worse functional and mental health outcomes at 3, 6, and 12 months post-TKA compared to patients with average or above average mental health scores. Interestingly, even among patients without an Anx/Dep diagnosis, mental health status remained a consistent predictor those with average and above average PROMIS-10 mental health scores reported higher functional and mental health outcomes throughout the postoperative year. By 12 months, functional status differed significantly across mental health strata, with patients in the BA-MHT group demonstrating worse functional outcomes regardless of Anx/Dep diagnosis, indicating that baseline mental health status remains an important determinant of long-term recovery following TKA.
Notably, even within the no Anx/Dep group, patients in the below average mental health cohort had longer hospital stays and consumed more opioids postoperatively prior to discharge, mirroring trends seen in the diagnosed Anx/Dep cohort. These findings support emerging literature advocating for a shift away from viewing mental health as a clear-cut predictor of surgical outcomes; rather, individualized assessments using PROMIS-10 scores may offer more accurate insight into expected recovery paths. For example, Kostyun et al. reported that among patients with a known mental illness undergoing total hip arthroplasty, those with above average or average mental health scores consumed 14–28% less opioids during their inpatient stay compared to patients with below average mental health scores [16]. Their study, like ours, challenges the traditional assumption that all patients with mental illness experience poorer postoperative outcomes, and instead highlights the usefulness of the PROMIS-10 score in guiding postoperative care.
This study is not without limitations. First, because our final cohort represented only a portion of the initially identified TKA patients, in part due to incomplete PRO compliance, selection bias cannot be fully excluded, and this may limit the generalizability of our findings. One important consideration is the need to identify protective factors that contribute to a patient’s above-average mental health score, such as greater access to psychosocial support and behavioral health services. Additionally, our analysis did not account for surgical variables, which could have influenced postoperative in-hospital opioid consumption and pain levels. Preoperative opioid use was recorded as a binary variable, and detailed information regarding dosage, duration, or prescriptions was not available, which may limit the interpretation of opioid-related findings. The limited diversity within our cohort may further restrict applicability to broader patient populations. We also did not record patients’ psychiatric medication use, which could have impacted the effects of mental health on recovery trajectories. Lastly, our study did not capture the full range of mental health conditions that may influence postoperative outcomes, highlighting the need for future research to explore other mental health diagnoses and the corresponding effect on outcomes.
Despite these limitations, our study adds to the growing awareness within the orthopaedic community of the complexity of mental health diagnoses and how they can impact patient-reported outcomes. As the Centers for Medicare & Medicaid Services increasingly integrates PROMs into its quality metrics, understanding a patient’s mental health becomes crucial, not only to optimize recovery but also to help establish realistic expectations for the postoperative course. In this context, our findings suggest that baseline PROMIS-10 mental health scores may provide more meaningful risk stratification for Centers for Medicare & Medicaid Services-reported functional improvement than a binary anxiety or depression diagnosis alone. By proactively assessing and addressing mental well-being, clinicians may better identify patients at risk and tailor care strategies that support both physical and psychological recovery. Nevertheless, our study reinforces that a mental health diagnosis of Anx/Dep alone does not necessarily predict worse patient-reported outcomes, supporting the importance of a more case-by-case approach to both preoperative evaluation and postoperative care.
Conclusions
Our study shows that PROMIS-10 mental health scores are reliable, individualized indicators of postoperative outcomes such as opioid use prior to discharge, hospital LOS, and PROMs. Patients with lower mental health scores, regardless of a formal anxiety or depression diagnosis, tended to have longer hospitalizations, higher in-hospital opioid consumption, and lower physical and mental scores following TKA. These findings support our hypothesis that PROMIS-10 scores better reflect variation in recovery than binary diagnostic status alone. Moreover, the results reinforce the idea that orthopaedic providers should routinely assess mental health using validated tools like the PROMIS-10 and consider mental well-being when discussing recovery plans with patients.
CRediT authorship contribution statement
Mathew Mendoza: Writing – review & editing, Writing – original draft, Visualization, Methodology, Investigation, Formal analysis, Data curation. Dallas F. Bennett: Writing – review & editing, Writing – original draft, Visualization, Methodology, Investigation, Formal analysis, Data curation. Joshua H. Taylor: Writing – review & editing, Writing – original draft, Visualization, Methodology, Investigation, Data curation. Bryce Palumbo: Writing – review & editing, Writing – original draft, Investigation, Data curation. Ugonna N. Ihekweazu: Writing – review & editing, Supervision, Project administration, Investigation. Corey F. Hryc: Writing – review & editing, Supervision, Project administration, Methodology, Investigation, Formal analysis, Conceptualization.
Conflict of interest
C.F. Hryc received research support from J&J MedTech. U.N. Ihekweazu received royalties from Fondren LTD; is on the speakers bureau/paid presentations for J&J MedTech; is a paid consultant for J&J MedTech and Zimmer Biomet (C3); has stock or stock options in VISIE, RevelAI, and ROMTech; received research support from J&J MedTech; is a member of the medical/orthopaedic publications editorial/governing board at the Journal of Arthroplasty; and is a board member/committee appointments for AAHKS Board of Directors and the Arthritis Foundation; all other authors declare no potential conflicts of interest.
For full disclosure statements refer to https://doi.org/10.1016/j.artd.2026.102001.
Footnotes
Funding: The Fondren Orthopedic Research Institute at the Texas Orthopedic Hospital supported a portion of the study team.
Appendix
Supplement Table 1.
Comparison of demographics among the three Anx/Dep subcohorts.
| Variable | Above-average MHT (n = 79) | Average MHT (n = 143) | Below-average MHT (n = 70) | P value | |||
|---|---|---|---|---|---|---|---|
| Age, y | 68.89 | 44-85 | 66.87 | 44-89 | 65.34 | 43-84 | .2098 |
| BMI (kg/m2) | 33.05 | 18.5-60.4 | 32.90 | 17.8-59.4 | 35.30 | 19.5-55.2 | .0556 |
| Sex | .4344 | ||||||
| Male | 18 | 22.78 | 35 | 24.48 | 22 | 31.43 | |
| Female | 61 1 | 77.22 | 108 | 75.52 | 48 | 68.57 | |
| Race | .010 | ||||||
| White | 67 | 84.81 | 104 | 72.73 | 38 | 54.29 | |
| Black or African American | 6 | 7.59 | 14 | 9.79 | 16 | 22.86 | |
| Asian | 0 | 0 | 3 | 2.10 | 3 | 4.29 | |
| Hispanic | 4 | 5.06 | 10 | 6.99 | 8 | 11.43 | |
| Unknown | 2 | 2.53 | 10 | 6.99 | 3 | 4.29 | |
| Other | 0 | 0 | 2 | 1.40 | 2 | 2.86 | |
| Ethnicity | .3376 | ||||||
| Hispanic | 4 | 5.06 | 10 | 6.99 | 8 | 11.43 | |
| Not Hispanic | 73 | 92.4 | 123 | 86.02 | 59 | 84.30 | |
| Declineda | 2 | 2.53 | 10 | 6.99 | 3 | 4.29m | |
| Marital status | .0629 | ||||||
| Divorced | 9 | 11.39 | 20 | 13.99 | 12 | 17.14 | |
| Married | 50 | 63.29 | 85 | 59.44 | 33 | 47.14 | |
| Partner | 0 | 0 | 1 | 0.70 | 1 | 1.43 | |
| Separated | 1 | 1.27 | 1 | 0.70 | 5 | 7.14 | |
| Single | 5 | 6.33 | 14 | 9.79 | 12 | 17.14 | |
| Widowed | 14 | 17.72 | 21 | 14.69 | 6 | 8.57 | |
| Unknown | 0 | 0 | 1 | 0.70 | 1 | 1.43 | |
| Insurance | .0009 | ||||||
| Commercial | 17 | 21.52 | 54 | 37.76 | 17 | 24.29 | |
| Health exchange/other | 1 | 1.27 | 5 | 3.50 | 1 | 1.43 | |
| Medicaid | 0 | 0 | 1 | 0.70 | 6 | 8.57 | |
| Medicare | 60 | 75.95 | 80 | 55.94 | 43 | 61.43 | |
| Self-pay | 0 | 0 | 0 | 0 | 0 | 0 | |
| Other | 1 | 1.27 | 3 | 2.10 | 3 | 4.29 | |
| Mental health diagnosis | < .0001 | ||||||
| Anxiety only | 37 | 46.84 | 57 | 39.86 | 16 | 22.86 | |
| Depression only | 27 | 34.18 | 46 | 32.17 | 14 | 20.00 | |
| Anxiety and depression | 15 | 18.99 | 40 | 27.97 | 40 | 57.14 | |
Age, body mass index (BMI), and length of stay reported as mean (range).
All other data reported as frequency (percentage).
BMI, body mass index.
Values in bold represent statistical significance.
Not included in P-value.
Supplement Table 2.
Comparison of demographics among the three no Anx/Dep subcohorts.
| Variable | Above-average MHT (n = 189) | Average MHT (n = 79) | Below-average MHT (n = 25) | P value | |||
|---|---|---|---|---|---|---|---|
| Age, y | 68.28 | 37-91 | 66.91 | 44-84 | 69.04 | 55-81 | .1825 |
| BMI (kg/m2) | 31.98 | 14.8-49.1 | 32.62 | 15.4-49.5 | 33.78 | 26.1-45.7 | .4191 |
| Sex | .6005 | ||||||
| Male | 86 | 45.50 | 41 | 48.10 | 11 | 44 | |
| Female | 103 | 54.50 | 38 | 51.90 | 14 | 56 | |
| Race | .0009 | ||||||
| White | 152 | 80.42 | 47 | 59.49 | 15 | 60 | |
| Black or African American | 12 | 6.35 | 13 | 16.46 | 4 | 16 | |
| Asian | 7 | 3.70 | 2 | 2.53 | 0 | 0 | |
| Hispanic | 13 | 6.88 | 12 | 15.19 | 2 | 8 | |
| Unknown | 3 | 1.59 | 5 | 6.33 | 4 | 16 | |
| Other | 2 | 2.12 | 0 | 0 | 0 | 0 | |
| Ethnicity | .0011 | ||||||
| Hispanic | 13 | 6.88 | 12 | 15.19 | 2 | 8 | |
| Not Hispanic | 173 | 91.53 | 62 | 78.48 | 19 | 76 | |
| Declineda | 3 | 1.59 | 5 | 6.33 | 4 | 16 | |
| Marital status | .3373 | ||||||
| Divorced | 13 | 6.88 | 12 | 15.19 | 3 | 12 | |
| Married | 149 | 78.84 | 53 | 67.09 | 17 | 68 | |
| Partner | 0 | 0 | 1 | 1.27 | 0 | 0 | |
| Separated | 0 | 0 | 0 | 0 | 0 | 0 | |
| Single | 13 | 6.88 | 5 | 6.33 | 3 | 12 | |
| Widowed | 14 | 7.41 | 8 | 10.13 | 2 | 8 | |
| Unknown | 0 | 0 | 0 | 0 | 0 | 0 | |
| Insurance | |||||||
| Commercial | 61 | 32.28 | 30 | 37.97 | 5 | 20 | .6058 |
| Health exchange/other | 2 | 1.06 | 2 | 2.53 | 1 | 4 | |
| Medicaid | 1 | 0.56 | 0 | 0 | 0 | 0 | |
| Medicare | 123 | 65.08 | 45 | 56.69 | 19 | 76 | |
| Self-pay | 1 | 0.53 | 0 | 0 | 0 | 0 | |
| Other | 1 | 0.53 | 2 | 2.52 | 0 | 0 | |
Age, body mass index (BMI), and length of stay reported as mean (range).
All other data reported as frequency (percentage).
BMI, body mass index.
Values in bold represent statistical significance.
Not included in P-value.
Appendix A. Supplementary data
References
- 1.Sloan M., Premkumar A., Sheth N.P. Projected volume of Primary Total Joint Arthroplasty in the U.S., 2014 to 2030. J Bone Joint Surg Am. 2018;100:1455–1460. doi: 10.2106/JBJS.17.01617. [DOI] [PubMed] [Google Scholar]
- 2.Goudie E.B., Robinson C., Walmsley P., Brenkel I. Changing trends in total knee replacement. Eur J Orthop Surg Traumatol. 2017;27:539–544. doi: 10.1007/s00590-017-1934-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Chen S., Huang W., Zhang M., Song Y., Zhao C., Sun H., et al. Dynamic changes and future trend predictions of the global burden of anxiety disorders: analysis of 204 countries and regions from 1990 to 2021 and the impact of the COVID-19 pandemic. EClinicalMedicine. 2025;79:103014. doi: 10.1016/j.eclinm.2024.103014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Ghoneim M.M., O’Hara M.W. Depression and postoperative complications: an overview. BMC Surg. 2016;16:5. doi: 10.1186/s12893-016-0120-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Wood T.J., Gazendam A.M., Kabali C.B., Wood T.J., Petruccelli D.T., Tushinski D.M., et al. Postoperative outcomes following total hip and knee arthroplasty in patients with pain catastrophizing, anxiety, or depression. J Arthroplasty. 2021;36:1908–1914. doi: 10.1016/j.arth.2021.02.018. [DOI] [PubMed] [Google Scholar]
- 6.Pan X., Wang J., Lin Z., Dai W., Shi Z. Depression and anxiety are risk factors for postoperative pain-related symptoms and complications in patients undergoing primary total knee arthroplasty in the United States. J Arthroplasty. 2019;34:2337–2346. doi: 10.1016/j.arth.2019.05.035. [DOI] [PubMed] [Google Scholar]
- 7.Diei C., Mehdipour S., Wall P.V., Gabriel R.A. The association of depression and anxiety with postoperative opioid use following total joint arthroplasty. Heliyon. 2023;9:e18813. doi: 10.1016/j.heliyon.2023.e18813. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Paredes A.C., Costa P., Costa M., Oliveira P., Varanda P., Almeida A., et al. Differences in the relationship between pain and anxiety in total knee and hip arthroplasty: a longitudinal cross-lagged analysis mediated by depression and pain catastrophizing. Brit J Pain. 2024;18:482–496. doi: 10.1177/20494637241273905. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Duivenvoorden T., Vissers M.M., Verhaar J.A.N., Busschbach J.J.V., Gosens T., Bloem R.M., et al. Anxiety and depressive symptoms before and after total hip and knee arthroplasty: a prospective multicentre study. Osteoarthr Cartilage. 2013;21:1834–1840. doi: 10.1016/j.joca.2013.08.022. [DOI] [PubMed] [Google Scholar]
- 10.Fillingham Y.A., Hanson T.M., Leinweber K.A., Lucas A.P., Jevsevar D.S. Generalized anxiety disorder: a modifiable risk factor for pain catastrophizing after total joint arthroplasty. J Arthroplasty. 2021;36:S179–S183. doi: 10.1016/j.arth.2021.02.023. [DOI] [PubMed] [Google Scholar]
- 11.Kotzur T., Singh A., Peng L., Lundquist K., Peterson B., Young W., et al. The effect of depression on total knee arthroplasty. JAAOS Glob Res Rev. 2024;8 doi: 10.5435/JAAOSGlobal-D-24-00098. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Singh J.A., Lewallen D.G. Time trends in the characteristics of patients undergoing primary total knee arthroplasty. Arthritis Care Res. 2014;66:897–906. doi: 10.1002/acr.22233. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Heijbel S., W-Dahl A., E-Naili J., Hedström M. Patient-Reported anxiety or depression increased the risk of dissatisfaction despite improvement in pain or function following total knee arthroplasty: a Swedish register-based observational Study of 8,745 patients. J Arthroplasty. 2024;39:2708–2713. doi: 10.1016/j.arth.2024.04.071. [DOI] [PubMed] [Google Scholar]
- 14.Hays R.D., Bjorner J.B., Revicki D.A., Spritzer K.L., Cella D. Development of physical and mental health summary scores from the patient-reported outcomes measurement information system (PROMIS) global items. Qual Life Res. 2009;18:873–880. doi: 10.1007/s11136-009-9496-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Lyman S., Lee Y.-Y., Franklin P.D., Li W., Cross M.B., Padgett D.E. Validation of the KOOS, JR: a Short-form Knee Arthroplasty Outcomes Survey. Clin Orthop Relat Res. 2016;474:1461–1471. doi: 10.1007/s11999-016-4719-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Kostyun R.O., Hackett A., Lucchio P., Witmer D.K., Solomito M.J. Positive deviance and exploring the role of mental health on early postoperative recovery after total hip arthroplasty: a retrospective analysis using the PROMIS-10 global health. J Am Acad Orthop Surg. 2025;33:34–41. doi: 10.5435/JAAOS-D-23-01230. [DOI] [PubMed] [Google Scholar]
- 17.Tamate T., Nishioka S., Ry L.D., Weldon R.H., N A.S., Nakasone C.K. The influence of mental health on early outcomes following total hip arthroplasty. Arch Orthop Trauma Surg. 2023;144:1773–1779. doi: 10.1007/s00402-023-05159-4. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.