1 Introduction

Patient reported outcome measures (PROMs) have been increasingly utilized to understand factors leading to patients’ perceived pain relief, joint functionality, and overall satisfaction following total joint arthroplasty (TJA) of the hip and knee.^1,2 Several clinical variables have been shown to be important factors affecting patient-reported outcomes after TJA, many of which are patient-related (e.g., socioeconomic status, medical comorbidities, preoperative expectations, etc.).¹ The veteran population which receives care through the United States (US) Veterans Affairs (VA) healthcare system represents a unique patient population which has been shown to be characteristically different from non-VA (i.e., civilian) patient populations, with VA patients having poorer sociodemographic characteristics and worse health status.^3–5 Relatively little is known regarding PROMs after TJA in the veteran population, and data from civilian populations may not be generalizable to the VA population, due to differences in many important patient-related factors affecting TJA outcomes. For example, compared to the general population, the veteran population historically has lower socioeconomic status as well as higher burdens of medical and mental illness; they also tend to present with self-reported physical and cognitive disabilities.^5,6 Even after controlling for demographics and medical comorbidities, VA patients are more likely to have a readmission event, prolonged length of stay (>four days), or experience a complication within 30 days of surgery.⁷

Within the VA system, veterans can apply for disability benefits through service connection (SC). Relatively analogous to the Workers' Compensation (WC) system of civilian medicine, service-connected veterans must be able to document/prove a SC. SCs can be granted for any injury, illness, and/or medical or psychiatric condition caused or exacerbated by a veteran's military service. SC conditions and mental health are negatively associated with pain outcomes postoperatively in the elderly, and an increased risk of premature death due to external causes in young veterans.^8,9

Although not specifically supported in the literature, medical professionals are commonly concerned by patients’ secondary gain of not losing disability benefits as an influencing factor in long-term outcomes.^8–10 The discrepancy in clinical outcomes after elective orthopaedic surgery between WC and non-WC patients has been established, yet there is a relative lack of studies specific to outcomes of veterans receiving SC-disability benefits following TJA.^11–14 For example, patients receiving WC are less likely to report symptom relief and improvement in quality of life after treatment of a herniated lumbar disc,¹¹ are more likely to have inferior functional outcomes and PROMs after rotator cuff repair,¹² and have significantly lower outpatient satisfaction scores¹⁴ when compared with patients not receiving WC. Relatively less is known about the relationship between SC-disability status and outcomes in the veteran population.

Therefore, we sought to determine what influence, if any, SCs might have on clinical outcomes after TJA in a veteran population in this retrospective cohort study aimed at augmenting the literature surrounding PROMs, the VA population, and SCs. The study compares pre- and postoperative Hip disability and Osteoarthritis Outcome Score for Joint Replacement (HOOS-JR)¹⁵ and Knee injury and Osteoarthritis Outcome Score for Joint Replacement (KOOS-JR)¹⁶ scores between SC and non-service connected (NSC) veterans up to one year following TJA in order to determine the effects of military SC-disability status on patient outcomes. We hypothesize that HOOS-JR/KOOS-JR scores will be lower in SC compared to NSC veterans.

2 Methodology

The relevant local institutional review boards approved all study procedures. Veterans were included if they underwent primary total hip arthroplasty (THA) or total knee arthroplasty (TKA) for hip/knee osteoarthritis at a regional VA Medical Center between July 2019 and December 2020. Revision TJAs and TJAs performed for indications other than osteoarthritis (e.g., fractures, failed internal fixation) were excluded. Patients were also excluded for more than one primary arthroplasty performed or revision arthroplasty within the 1-year interval of the index procedure. All THA procedures were performed by a single surgeon using ceramic-on-polyethylene components (ACTIS® Total Hip System, DePuy Synthes, Warsaw, IN). All TKA procedures were performed by one of two surgeons using the same implant system (Persona®, Zimmer Biomet, Warsaw, IN). Patients were grouped into two cohorts according to SC status (no SC: NSC cohort; any SC: SC cohort), and the specific type of SC (mental health, musculoskeletal) was subanalyzed.

A trained orthopaedic staff member administered HOOS-JR and KOOS-JR surveys to patients at preoperative, 2-month, and 12-month postoperative follow-up visits, and assisted patients with completion as needed. If a patient did not complete a survey in person for any reason, either the primary surgeon or the staff member conducted the survey by telephone. Covariates included for analysis were age, sex, body mass index (BMI), history of mental illness (depression, posttraumatic stress disorder [PTSD], schizophrenia), and Charlson Comorbidity Index (CCI).

3 Results

The THA sample at baseline included n = 67 patients (SC: n = 40; NSC: n = 27); Eleven SC THA patients (28% reduction vs. baseline) and 11 NSC THA patients (41% reduction vs. baseline) were lost to follow-up. The TKA sample at baseline included n = 142 patients (SC: n = 95; NSC: n = 47); Seventeen SC TKA patients (18% reduction vs. baseline) and 11 NSC TKA patients (28% reduction vs. baseline) were lost to follow-up. None of the groups lost to follow-up differed significantly from baseline groups for any covariate, rates of SC status, or joint function scores (for each, P ≥ 0.180). Except for sex in THA (P < 0.001) and age in TKA (P = 0.003), the SC and NSC cohorts did not differ for baseline covariates or joint function (see Table 1).

3.1

3.1 Service connection

For THA, neither the time*SC interaction (P = 0.223) nor the SC main effect (P = 0.900) were significant. The time main effect was significant (P < 0.001), with both cohorts exhibiting substantial baseline-to-follow-up improvement in HOOS-JR scores (mean change of 34.7 in SC and 41.7 in NSC at 12 months; see Fig. 1). Proportions of THA patients achieving SCB were similar in the SC (72.4%) and NSC (75.0%) cohorts (P = 0.851; see Fig. 2). Improvement in HOOS-JR scores during follow-up was not significantly affected by the quartile*SC interaction (P = 0.646) or SC (P = 0.832). Baseline quartile influenced improvement in HOOS-JR (P < 0.001), with the lowest quartile (i.e., worst function) at baseline exhibiting significantly greater improvement than each of the other quartiles (for each pairwise comparison, P ≤ 0.014; no other pairwise differences: for each, P ≥ 0.969; see Fig. 2). All quartiles on average achieved SCB at 12-month follow-up.

Close

Fig. 1 Mean Hip disability and Osteoarthritis Outcome Score for Joint Replacement (HOOS-JR) scores (circles) with 95% confidence intervals (error bars) at preoperative baseline (0 months) and at 2-month and 12-month follow-up. Black circles represent the Service Connected (SC) cohort, and white circles represent the Non-Service Connected (NSC) cohort. Only the time main effect was statistically significant (P < 0.001). Asterisks (*) above a time point indicate that the mean HOOS-JR score at that time point was significantly different from the previous time point (for each, P < 0.001).

Export to PPT

Close

Fig. 2 A: Percent of patients achieving Hip disability and Osteoarthritis Outcome Score for Joint Replacement (HOOS-JR) substantial clinical benefit (SCB) (ΔHOOS-JR≥22; black bar segments) vs. not achieving SCB (ΔHOOS-JR<22; white bar segments) in the Service Connected (SC) and Non-Service Connected (NSC) cohorts. B: Distributions of ΔHOOS-JR by baseline HOOS-JR quartile. Boxes represent the interquartile range (25th to 75th percentiles), horizontal lines in the boxes are the medians, and error bars represent the 10th and 90th percentiles. The horizontal grey line represents SCB at the ΔHOOS-JR value of 22. The asterisk (*) indicates that quartile one (Q1) differed significantly from all other quartiles for ΔHOOS-JR (for each, P ≤ 0.014).

Export to PPT

For TKA, the time*SC interaction was not significant (P = 0.700), but the main effects of time (P < 0.001) and SC (P = 0.036) were. Both TKA cohorts exhibited improved KOOS-JR scores over the course of follow-up (mean change of 21.6 in SC and 25.3 in NSC at 12 months). The NSC cohort had slightly, yet statistically significantly higher scores than the SC cohort at all time points (see Fig. 3). Proportions of TKA patients achieving SCB were similar in the SC (52.6%) and NSC (58.8%) cohorts (P = 0.541; see Fig. 4). Improvement in KOOS-JR scores during follow-up was not significantly affected by the quartile*SC interaction (P = 0.679), but was affected by SC (P = 0.031) and baseline quartile (P < 0.001). The SC cohort had significantly less improvement than the NSC cohort, although both on average achieved SCB (ΔKOOS-JR of 21.3 ± 18.7 vs. 25.0 ± 18.8, respectively). All baseline quartiles differed significantly from each other (for each, P ≤ 0.039), except for the second and third quartiles (P = 0.571; see Fig. 4). Quartiles one through three (Q1-Q3) on average achieved SCB by 12 months, but quartile four (Q4) did not. Within KOOS-JR Q4, only 18% of patients achieved SCB and 35% experienced negative ΔKOOS-JR.

Close

Fig. 3 Mean Knee injury and Osteoarthritis Outcome Score for Joint Replacement (KOOS-JR) scores (circles) with 95% confidence intervals (error bars) at preoperative baseline (0 months) and at 2-month and 12-month follow-up. Black circles represent the Service Connected (SC) cohort, and white circles represent the Non-Service Connected (NSC) cohort. The SC (P = 0.034) and time main effects were statistically significant (P < 0.001). The NSC cohort had higher KOOS-JR scores than the SC cohort across all time points. Asterisks (*) above a time point indicate that the mean KOOS-JR score (within the combined SC and NSC cohorts) at that time point was significantly different from the previous time point (for each, P < 0.001).

Export to PPT

Close

Fig. 4 A: Percent of patients achieving Knee injury and Osteoarthritis Outcome Score for Joint Replacement (KOOS-JR) substantial clinical benefit (SCB) (ΔKOOS-JR≥20; black bar segments) vs. not achieving SCB (ΔKOOS-JR<20; white bar segments) in the Service Connected (SC) and Non-Service Connected (NSC) cohorts. B: Distributions of ΔKOOS-JR by baseline KOOS-JR quartile. Boxes represent the interquartile range (25th to 75th percentiles), horizontal lines in the boxes are the medians, and error bars represent the 10th and 90th percentiles. The horizontal grey line represents SCB at the ΔKOOS-JR value of 20. An asterisk (*) below a quartile's box means that quartile differed significantly from all other quartiles for ΔKOOS-JR (for each, P ≤ 0.039).

Export to PPT

4 Discussion

The current study sought to determine whether military SC-disability status influenced postoperative PROMs after primary THA and TKA in a US veteran population. Overall, both SC and NSC veterans undergoing THA or TKA achieved significant clinical improvements over the follow-up period, as measured by HOOS-JR and KOOS-JR. However, results suggest that mental health SC status may have a negative influence on perceived TKA outcomes. On the other hand, SC status does not seem to influence perceived THA outcomes.

Veterans undergoing THA tended to have favorable subjective outcomes regardless of SC status, and most had successful outcomes. This is consistent with published data regarding THA outcomes in civilian patients receiving WC. Hostin and colleagues reported on 22 patients receiving WC benefits (related to their hip disorder) matched with 22 patients not receiving WC who underwent primary THA, and found that the percentage of patients with good or excellent results (Harris hip score ≥80) did not differ significantly between WC and non-WC groups (83% versus 88%, respectively).¹⁸

Results among veterans undergoing TKA, on the other hand, differed. While both SC and NSC groups achieved significant improvements in KOOS-JR scores postoperatively, significant differences in KOOS-JR scores were present between SC and NSC cohorts across all time points. Significantly less improvement was observed among the SC cohort when compared with the NSC cohort (i.e., the mean change in KOOS-JR from preoperatively to 1-year postoperatively was significantly less in the SC cohort). Finally, although the proportion of patients who achieved SCB after TKA was not significantly different between cohorts, the proportions were notably lower for TKA compared to THA. Taken together, these findings indicate that veterans receiving SC benefits tended to have inferior subjective outcomes after TKA when compared with those not receiving SC benefits. The results of TKA among veterans seem to be generally less reliable than those of THA, regardless of SC status, with only 50–60% of veterans experiencing SCB after TKA, compared to ∼75% following THA. This is consistent with previously published studies regarding the influence of WC on outcomes after TKA in civilian patient populations, which have generally reported suboptimal outcomes among WC patients.^19–22

Mental health SCs, but not musculoskeletal SCs, led to significantly lower KOOS-JR scores at all time points when compared with veterans with other types of SC. Poor mental health has been shown to be a significant predictor of dissatisfaction after TJA and is associated with increased risk of persistent pain, inferior function, and lower postoperative PROMs.^23–26 The influence of poor mental health seems to be most significant when considering pain and other subjective outcome measures, as objective functional outcomes have been less affected by mental health status.²⁴

The results of the current study indicate that mental health indeed plays an important role in subjective outcomes after TJA in the US veteran patient population, consistent with data from other patient populations.^23–26 However, it is important to note that although veterans with mental health SCs had lower PROMs at all time points, they still achieved significant improvements in their outcome scores postoperatively. These findings agree with those of previously published studies demonstrating that although poor mental health may result in inferior outcomes after TJA, these patients can still expect clinically significant improvements compared to preoperative status.^23,24

Another notable finding from this study was the influence of baseline functional status on postoperative improvement and rates of achieving SCB following TJA, which differed somewhat between THA and TKA. When veterans were separated into quartiles based upon preoperative functional status (as determined by preoperative HOOS-JR/KOOS-JR), the group with the poorest baseline function achieved significantly greater improvements than the other groups, for both THA and TKA. For veterans undergoing THA, this was the only difference, as the other three quartiles achieved similar postoperative functional improvements. In addition, for the THA groups, all quartiles achieved SCB by 1-year postoperatively.

For veterans undergoing TKA, on the other hand, postoperative improvement appeared to be more dependent upon preoperative status, and achievement of SCB was less predictable. Although the lowest quartile (i.e., poorest preoperative status) still experienced the largest gains (a nearly 40-point improvement in KOOS-JR on average), improvements in the other quartiles seemed to occur in a graded manner, with the middle two quartiles achieving modest gains and the highest quartile (i.e., best preoperative status) achieving very little improvement. Furthermore, just 18% of patients in the highest quartile achieved SCB after TKA, whereas SCB was reliably achieved in all other quartiles. These findings are intuitive as those patients with the lowest baseline function likely stand to gain the most, whereas those with higher baseline function likely have less room for improvement, underscoring the importance of adequate preoperative counseling and management of expectations.

Limitations of this study are the relatively small sample size and retrospective nature of the investigation. Although statistically underpowered, this study reported single-digit effect sizes of HOOS-JR and KOOS-JR (3-point differences on 100-point scales). The authors find this to carry low clinical meaningfulness irrespective of statistical power. The rate of patient loss to follow up was also relatively high. Those patients who were lost to follow up did not differ significantly for baseline characteristics from the rest of the cohort, so what impact, if any, this might have had on the results of the study is unclear. Finally, we did not stratify SCs based on the level/degree of disability (e.g., 50% SC versus 100% SC) or the origin of the SC (e.g., condition incurred in, versus aggravated by, military service), which could have played some role in outcomes. Despite these limitations, this study provides new insights regarding the effect of SC-disability compensation status on outcomes following THA and TKA in the US veteran population, and it is the first study, to the authors’ knowledge, to examine this topic.

5 Conclusions

Significant differences in postoperative outcomes between SC and NSC veterans were noted only among patients undergoing TKA. Lower KOOS-JR scores were attributed to the designation of a mental health SC. SC status did not influence the outcomes of THA or the likelihood of achieving SCB after either THA or TKA. Veterans with mental health SCs and high preoperative functional status undergoing TKA should remain a population of particular attention to surgeons, although most veterans can expect significant clinical improvement relative to their baseline status following TJA.

Funding/Sponsorship

The research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Informed consent

Informed consent was obtained.

Institutional ethical committee approval

Institutional Review Board approval was obtained.

Authors contribution

GGV – Conceptualization, Methodology, Data Curation, Writing - Original Draft, Writing - Review & Editing.

DAB – Conceptualization, Writing - Review & Editing.

JGL – Data Curation, Writing - Original Draft, Writing - Review & Editing.

AWF – Conceptualization, Methodology, Formal Analysis.

ABK – Conceptualization, Methodology, Supervision.