Military Service and Decision Quality in the Management of Knee Osteoarthritis

Military Service and Decision Quality in the Management of Knee Osteoarthritis Abstract Background Decision quality measures the degree to which care decisions are knowledge-based and value-aligned. Because military service emphasizes hierarchy, command, and mandates some healthcare decisions, military service may attenuate patient autonomy in healthcare decisions and lower decision quality. VA is the nation’s largest provider of orthopedic care. We compared decision quality in a sample of VA and non-VA patients seeking care for knee osteoarthritis. Methods Our study sample consisted of patients newly referred to our orthopedic clinic for the management of knee osteoarthritis. None of the study patients were exposed to a knee osteoarthritis decision aid. Consenting patients were administered the Hip/Knee Decision Quality Instrument (HK-DQI). In addition, they were surveyed about decision-making preferences and demographics. We compared results to a non-VA cohort from our academic institution’s arthroplasty database. Results The HK-DQI Knowledge Score was lower in the VA cohort (45%, SD = 22, n = 25) compared with the non-VA cohort (53%, SD = 21, n = 177) (p = 0.04). The Concordance Score was lower in the VA cohort (36%, SD = 49%) compared with the control cohort (70%, SD 46%) (p = 0.003). Non-VA patients were more likely to make a high-quality decision (p = 0.05). Non-VA patients were more likely to favor a shared decision-making process (p = 0.002). Conclusions Decision quality is lower in Veterans with knee osteoarthritis compared with civilians, placing them at risk for lower treatment satisfaction and possibly unwarranted surgical utilization. Our future work will examine if this difference is from conditioned military service behaviors or confounding demographic factors, and if conventional shared decision-making techniques will correct this deficiency. INTRODUCTION Military law requires “the subordination of the desires and interests of the individual to the needs of the Service.”1 This requirement extends to healthcare where medical decisions can be made for military patients based on the interests of Nation and Command.2,3 This relationship further accentuates the normal imbalance of power observed between physicians and patients. This system may create gaps in trust between patients and providers and reduce patients’ willingness to question their providers’ recommendations, thereby reducing patients’ understanding of proposed treatments.1 These conditioned behaviors of military service may persist after discharge and attenuate Veterans’ healthcare decision-making autonomy – a key component in patient-centered care. Patient-centered care occurs when the chosen treatment considers the needs, values, and preferences of the well-informed patient.4–7 Decision quality is a measurable indicator of patient-centered care.8,9 Patient-centered care may be achieved through measurement and enhancement of decision quality. High decision quality requires two endpoints: (1) patients’ decisions are based on the best evidence; and (2) treatment decisions are consistent with the treatment features that matter most to the informed individual patient.9,10 Low decision quality may be improved through shared decision-making, a process of deliberate, collaborative treatment selection based upon the best available evidence supplemented with decision aids.10 However, this process may be more difficult in the setting of prior military service. In this observational pilot study, we measured decision quality in VA and non-VA patients considering treatment for knee osteoarthritis to determine if associations exist between decision quality and prior military service. We hypothesized that decision quality in the VA cohort would be lower than in non-VA patients, given the prior exposure to military service and its command structure and healthcare mandates. MATERIALS AND METHODS Study Population After obtaining Institutional Review Board approval (Veteran’s IRB of New England, VINNE), we educated, consented, and enrolled newly referred Veteran patients referred to our VA orthopedic clinic with a chief complaint of knee pain and grade III or IV osteoarthritis on plain radiographs.11 Patients with an inflammatory or autoimmune arthritis diagnosis, concomitant hip pain, a prior total joint replacement or surgical intervention for osteoarthritis involving any joint, a contraindication for total joint replacement, or a sensory, cognitive, or other impairment that would preclude informed consent were excluded. No VA patients had exposure to a decision aid. We screened patients for enrollment in person or via telephone. Twenty-five Veterans meeting inclusion criteria were consented and enrolled between October 2015 and September 2016. There were 24 men and one woman (Table I). The mean age was 66 yr. Educational levels for the patients were: graduate degree, n = 4; undergraduate, 6; high school graduate, 11; some high school, 3; no answer, 1. The patients’ races were: White, 20; Hispanic (White), 1; Hispanic (Black), 1; American Indian or Alaska Native, 1; Asian or Pacific Islander, 1; no answer, 1. Lack of gender diversity in this cohort precluded gender-oriented inter-cohort analyses. Table I. Patient Demographic Characteristics VA (n = 25) Non-VA (n = 69) p-Value Age (mean (SD)) 65.5 (12.1) 65.0 (10.5) 0.85 Gender = male (%) 24 (96.0) 27 (39.1) <0.001 Race (%)  American Indian or Alaska Native 1 (4.0) 0 (0.0) 0.031  Asian or Pacific Islander 1 (4.0) 1 (0.6)  Black 0 (0.0) 0 (0.0)  Hispanic, Black 1 (4.0) 0 (0.0)  Hispanic, White 1 (4.0) 0 (0.0)  Prefer not to answer 1 (4.0) 0 (0.0)  White 20 (80.0) 68 (98.6) Education (%)  Bachelors degree 6 (24.0) 16 (23.2) <0.001  Graduate degree 4 (16.0) 18 (26.1)  High school degree or equivalent 13 (52.0) 17 (24.6)  High school without graduation 1 (4.0) 4 (5.8)  Some college 0 (0.0) 14 (20.3)  Prefer not to answer 1 (4.0) 0 (0.0) VA (n = 25) Non-VA (n = 69) p-Value Age (mean (SD)) 65.5 (12.1) 65.0 (10.5) 0.85 Gender = male (%) 24 (96.0) 27 (39.1) <0.001 Race (%)  American Indian or Alaska Native 1 (4.0) 0 (0.0) 0.031  Asian or Pacific Islander 1 (4.0) 1 (0.6)  Black 0 (0.0) 0 (0.0)  Hispanic, Black 1 (4.0) 0 (0.0)  Hispanic, White 1 (4.0) 0 (0.0)  Prefer not to answer 1 (4.0) 0 (0.0)  White 20 (80.0) 68 (98.6) Education (%)  Bachelors degree 6 (24.0) 16 (23.2) <0.001  Graduate degree 4 (16.0) 18 (26.1)  High school degree or equivalent 13 (52.0) 17 (24.6)  High school without graduation 1 (4.0) 4 (5.8)  Some college 0 (0.0) 14 (20.3)  Prefer not to answer 1 (4.0) 0 (0.0) Table I. Patient Demographic Characteristics VA (n = 25) Non-VA (n = 69) p-Value Age (mean (SD)) 65.5 (12.1) 65.0 (10.5) 0.85 Gender = male (%) 24 (96.0) 27 (39.1) <0.001 Race (%)  American Indian or Alaska Native 1 (4.0) 0 (0.0) 0.031  Asian or Pacific Islander 1 (4.0) 1 (0.6)  Black 0 (0.0) 0 (0.0)  Hispanic, Black 1 (4.0) 0 (0.0)  Hispanic, White 1 (4.0) 0 (0.0)  Prefer not to answer 1 (4.0) 0 (0.0)  White 20 (80.0) 68 (98.6) Education (%)  Bachelors degree 6 (24.0) 16 (23.2) <0.001  Graduate degree 4 (16.0) 18 (26.1)  High school degree or equivalent 13 (52.0) 17 (24.6)  High school without graduation 1 (4.0) 4 (5.8)  Some college 0 (0.0) 14 (20.3)  Prefer not to answer 1 (4.0) 0 (0.0) VA (n = 25) Non-VA (n = 69) p-Value Age (mean (SD)) 65.5 (12.1) 65.0 (10.5) 0.85 Gender = male (%) 24 (96.0) 27 (39.1) <0.001 Race (%)  American Indian or Alaska Native 1 (4.0) 0 (0.0) 0.031  Asian or Pacific Islander 1 (4.0) 1 (0.6)  Black 0 (0.0) 0 (0.0)  Hispanic, Black 1 (4.0) 0 (0.0)  Hispanic, White 1 (4.0) 0 (0.0)  Prefer not to answer 1 (4.0) 0 (0.0)  White 20 (80.0) 68 (98.6) Education (%)  Bachelors degree 6 (24.0) 16 (23.2) <0.001  Graduate degree 4 (16.0) 18 (26.1)  High school degree or equivalent 13 (52.0) 17 (24.6)  High school without graduation 1 (4.0) 4 (5.8)  Some college 0 (0.0) 14 (20.3)  Prefer not to answer 1 (4.0) 0 (0.0) Study Measures We measured our primary outcome, decision quality, using the Hip-Knee Decision Quality Instrument (HK-DQI).10 We also measured patient demographics (age, gender, education), the preferred level of decision-making engagement (patient and doctor equally, mostly patient, mostly doctor, totally patient, totally doctor), and patients comfort level in questioning the recommendations of their orthopedic provider.12 Observational Non-VA Control Cohort We drew our observational control cohort (non-VA) from a convenience sample derived from the prior 3 yr of our academic medical center’s orthopedic registry. This registry collects the HK-DQI, demographic, and decision-making preference data. We obtained separate Institutional Review Board approval for this query. The database contained 1,636 patients for the specified time period. Of these patients, 177 had undergone treatment for knee osteoarthritis and had completed the HK-DQI prior to treatment and had no exposure to a decision aid; these patients composed our unadjusted, bivariate analysis of decision quality (Fig. 1). Of these 177 patients, 69 had complete demographic and decision-making preference data sets that permitted multivariate logistic regression analyses. This final sample included 27 men and 42 women (Table I). We did not gender-match our cohort based upon prior work showing no gender-based differences in decision quality.13 The mean age was 65 yr (SD 11). Educational levels for this cohort were: graduate degree, n = 18; undergraduate, 16; some college, 14; high school graduate, 17; and some high school, 4. The patients’ races were: White, 68; Asian or Pacific Islander, 1. Lack of racial diversity in this cohort precluded race-oriented inter-cohort analyses. There were significant differences between the cohorts in terms of gender, race, and education. FIGURE 1. View largeDownload slide Composition of VA cohort and observational non-VA cohort for the current investigation. FIGURE 1. View largeDownload slide Composition of VA cohort and observational non-VA cohort for the current investigation. Sample Size Calculations We assessed sample size adequacy through power analysis for our primary outcome, decision quality, using available medical literature. Based on reported standard deviation values for decision quality scores with and without decision aid exposure, 18 and 19,10 respectively, we estimated that we would be able to detect a difference in decision quality scores of 12.0 points based on an unpaired t-test with a two-sided significance level of 0.05 and a power of 0.80 for samples of size 25. This calculation conservatively assumes that the intraclass correlation for the effect of matching is effectively zero. By comparison, the validation study of the HK-DQI demonstrated a difference of 14 between cohorts without and with shared decision-making,10 so our proposed detectable difference of 12 points seemed reasonable. Statistical Analysis Statistical analyses were conducted in R (version 3.3.0). Differences in means were calculated using a Student’s t-test or Chi-squared test. Associations between outcomes (TKS, CS) and covariates (race, education, decision-making preference) were quantified using univariate linear models between each covariate and each outcome. Compliance with Ethical Standards One author (ERH) received funding (V1CDA2015-34) that was in support of this project. RESULTS The mean HK-DQI TKS in the VA cohort (n = 25) was 45 (SD 21). This value was significantly lower than the non-VA cohort (n = 177) of 53 (SD 19) (p = 0.04). The mean HK-DQI Concordance Score (CS) in the VA cohort was 36 (SD 48). This value was significantly lower than the non-VA cohort 70 (SD 46) (p = 0.003). The mean HK-DQI DPS in the VA cohort was 51 (SD 40) (Table II). This value was not measured in our comparative cohort; however, its relation to values reported previously is considered in the discussion. The rate of VA patients making a high-quality decision (20%) was significantly lower than the rate of non-VA patients making a high-quality decision (69%) (p = 0.05). Table II. Unadjusted Univariate, Intercohort Comparison of HK-DQI Scores Hip-Knee Decision Quality Instrument Scores VA Cohort (n = 25) Non-VA Cohort (n = 177) p-Value Total Knowledge Score (%, SD) 45 (21) 53 (19) 0.04 Concordance Score (%, SD) 36 (48) 70 (46) 0.003 Decision Process Score (%, SD) 51 (40) NA NA High-Quality Decision (n, (%)) 5 (20) 69 (41) 0.05 Hip-Knee Decision Quality Instrument Scores VA Cohort (n = 25) Non-VA Cohort (n = 177) p-Value Total Knowledge Score (%, SD) 45 (21) 53 (19) 0.04 Concordance Score (%, SD) 36 (48) 70 (46) 0.003 Decision Process Score (%, SD) 51 (40) NA NA High-Quality Decision (n, (%)) 5 (20) 69 (41) 0.05 Table II. Unadjusted Univariate, Intercohort Comparison of HK-DQI Scores Hip-Knee Decision Quality Instrument Scores VA Cohort (n = 25) Non-VA Cohort (n = 177) p-Value Total Knowledge Score (%, SD) 45 (21) 53 (19) 0.04 Concordance Score (%, SD) 36 (48) 70 (46) 0.003 Decision Process Score (%, SD) 51 (40) NA NA High-Quality Decision (n, (%)) 5 (20) 69 (41) 0.05 Hip-Knee Decision Quality Instrument Scores VA Cohort (n = 25) Non-VA Cohort (n = 177) p-Value Total Knowledge Score (%, SD) 45 (21) 53 (19) 0.04 Concordance Score (%, SD) 36 (48) 70 (46) 0.003 Decision Process Score (%, SD) 51 (40) NA NA High-Quality Decision (n, (%)) 5 (20) 69 (41) 0.05 When we analyzed TKS with respect to patient age and educational levels factors (age, gender, race, and education), we found no significant differences between VA and non-VA cohorts, however, our study was powered insufficiently to support these analyses. When we analyzed CS with respect to patient factors, we again found no significant differences between VA and non-VA cohorts, but our study was powered insufficiently to support these analyses. We analyzed the VA cohort for variables predicting a high-quality decision without significant findings (Table III). Table III. Domains, Variables, and Bivariate Analyses for Predicting a High-Quality Decision* in VA Cohort. *TKS > 60 and positive concordance Domains and Variables High-Quality Decision 20% (n = 5) Low-Quality Decision 80% (n = 20) OR (95% CI) p-Value Age group 1.000  Age <60 (ref) 20% (1) 25% (5)    Age >60 80% (4) 75% (15) 0.638 (0.335–1.220) Gender 1.000  Female (ref) 0% (0) 5% (1)    Male 100% (5) 95% (19) 0.923 (0.502–1.685) Education 0.99  College or more (ref) 40% (2) 40% (8)    High school degree 60% (3) 50% (10) 0.914 (0.161–7.082)  High school unfinished or less 0% (0) 10% (2) 0.750 (0.085–7.503) Initial treatment preference 0.43  Non-surgical (ref) 40% (2) 15% (3)    Surgical 20% (1) 15% (3) 0.4286 (0.130–1.366)  Unsure/delay 40% (2) 65% (13) 0.132 (0.016–1.631)  No answer 0% (0) 5% (1) NA Final treatment selection 0.39  Non-surgical (ref) 80% (4) 95% (19)    Surgical 20% (1) 5% (1) NA  No answer 0% (0) 0% (0) NA Pre-visit comfort level questioning provider 0.22  Comfortable (ref) 80% (4) 70% (14)    Neutral 0% (0) 20% (4) 1.000 (3.6e-16 – 2.50)  Not comfortable 20% (1) 0% (0) NA  No answer 0% (0) 10% (2) NA Post-visit comfort level questioning provider 0.24  Comfortable (ref) 80% (4) 60% (14)    Neutral 20% (1) 5% (1) 3.500 (0.105–123.6)  Not comfortable 0% (0) 0% (0) 1.822 (0.422–12.415)  No answer 0% (0) 25% (5) NA Domains and Variables High-Quality Decision 20% (n = 5) Low-Quality Decision 80% (n = 20) OR (95% CI) p-Value Age group 1.000  Age <60 (ref) 20% (1) 25% (5)    Age >60 80% (4) 75% (15) 0.638 (0.335–1.220) Gender 1.000  Female (ref) 0% (0) 5% (1)    Male 100% (5) 95% (19) 0.923 (0.502–1.685) Education 0.99  College or more (ref) 40% (2) 40% (8)    High school degree 60% (3) 50% (10) 0.914 (0.161–7.082)  High school unfinished or less 0% (0) 10% (2) 0.750 (0.085–7.503) Initial treatment preference 0.43  Non-surgical (ref) 40% (2) 15% (3)    Surgical 20% (1) 15% (3) 0.4286 (0.130–1.366)  Unsure/delay 40% (2) 65% (13) 0.132 (0.016–1.631)  No answer 0% (0) 5% (1) NA Final treatment selection 0.39  Non-surgical (ref) 80% (4) 95% (19)    Surgical 20% (1) 5% (1) NA  No answer 0% (0) 0% (0) NA Pre-visit comfort level questioning provider 0.22  Comfortable (ref) 80% (4) 70% (14)    Neutral 0% (0) 20% (4) 1.000 (3.6e-16 – 2.50)  Not comfortable 20% (1) 0% (0) NA  No answer 0% (0) 10% (2) NA Post-visit comfort level questioning provider 0.24  Comfortable (ref) 80% (4) 60% (14)    Neutral 20% (1) 5% (1) 3.500 (0.105–123.6)  Not comfortable 0% (0) 0% (0) 1.822 (0.422–12.415)  No answer 0% (0) 25% (5) NA Bold values represents significance as a p value < 0.05. Table III. Domains, Variables, and Bivariate Analyses for Predicting a High-Quality Decision* in VA Cohort. *TKS > 60 and positive concordance Domains and Variables High-Quality Decision 20% (n = 5) Low-Quality Decision 80% (n = 20) OR (95% CI) p-Value Age group 1.000  Age <60 (ref) 20% (1) 25% (5)    Age >60 80% (4) 75% (15) 0.638 (0.335–1.220) Gender 1.000  Female (ref) 0% (0) 5% (1)    Male 100% (5) 95% (19) 0.923 (0.502–1.685) Education 0.99  College or more (ref) 40% (2) 40% (8)    High school degree 60% (3) 50% (10) 0.914 (0.161–7.082)  High school unfinished or less 0% (0) 10% (2) 0.750 (0.085–7.503) Initial treatment preference 0.43  Non-surgical (ref) 40% (2) 15% (3)    Surgical 20% (1) 15% (3) 0.4286 (0.130–1.366)  Unsure/delay 40% (2) 65% (13) 0.132 (0.016–1.631)  No answer 0% (0) 5% (1) NA Final treatment selection 0.39  Non-surgical (ref) 80% (4) 95% (19)    Surgical 20% (1) 5% (1) NA  No answer 0% (0) 0% (0) NA Pre-visit comfort level questioning provider 0.22  Comfortable (ref) 80% (4) 70% (14)    Neutral 0% (0) 20% (4) 1.000 (3.6e-16 – 2.50)  Not comfortable 20% (1) 0% (0) NA  No answer 0% (0) 10% (2) NA Post-visit comfort level questioning provider 0.24  Comfortable (ref) 80% (4) 60% (14)    Neutral 20% (1) 5% (1) 3.500 (0.105–123.6)  Not comfortable 0% (0) 0% (0) 1.822 (0.422–12.415)  No answer 0% (0) 25% (5) NA Domains and Variables High-Quality Decision 20% (n = 5) Low-Quality Decision 80% (n = 20) OR (95% CI) p-Value Age group 1.000  Age <60 (ref) 20% (1) 25% (5)    Age >60 80% (4) 75% (15) 0.638 (0.335–1.220) Gender 1.000  Female (ref) 0% (0) 5% (1)    Male 100% (5) 95% (19) 0.923 (0.502–1.685) Education 0.99  College or more (ref) 40% (2) 40% (8)    High school degree 60% (3) 50% (10) 0.914 (0.161–7.082)  High school unfinished or less 0% (0) 10% (2) 0.750 (0.085–7.503) Initial treatment preference 0.43  Non-surgical (ref) 40% (2) 15% (3)    Surgical 20% (1) 15% (3) 0.4286 (0.130–1.366)  Unsure/delay 40% (2) 65% (13) 0.132 (0.016–1.631)  No answer 0% (0) 5% (1) NA Final treatment selection 0.39  Non-surgical (ref) 80% (4) 95% (19)    Surgical 20% (1) 5% (1) NA  No answer 0% (0) 0% (0) NA Pre-visit comfort level questioning provider 0.22  Comfortable (ref) 80% (4) 70% (14)    Neutral 0% (0) 20% (4) 1.000 (3.6e-16 – 2.50)  Not comfortable 20% (1) 0% (0) NA  No answer 0% (0) 10% (2) NA Post-visit comfort level questioning provider 0.24  Comfortable (ref) 80% (4) 60% (14)    Neutral 20% (1) 5% (1) 3.500 (0.105–123.6)  Not comfortable 0% (0) 0% (0) 1.822 (0.422–12.415)  No answer 0% (0) 25% (5) NA Bold values represents significance as a p value < 0.05. In the subset with complete demographic and decision-making preference data, the mean TKS scores in the comparative cohort decreased from 53 to 47, which was higher than the VA cohort but no longer significant; CS was unchanged but no longer significant due to insufficient power (Table IV). The rate of non-VA patients making high-quality decisions also decreased. Survey results showed that no patients in the VA cohort reported preference for patient-only decision-making; 24% preferred mostly patient decision-making; 56% preferred equal patient-physician decision-making; no patients preferred mostly physician decision-making; one patient preferred physician-only decision-making; and 16% did not answer (Table IV). In the non-VA cohort, 13% of patients preferred patient-only or mostly patient decision-making; 84% preferred equal patient-physician decision-making; 3% of patients preferred mostly physician decision-making. The non-VA cohort had a significantly higher preference for a shared patient and physician decision-making process compared with the VA cohort (p = 0.002). The non-VA cohort also had a significantly higher initial preference for surgical treatment compared with the VA cohort (p < 0.001). Table IV. Comparison of Decision-Making Preferences Between VA Cohort and Non-VA Cohort with Complete Demographic and Decision-Making Preference Data Result VA Cohort (n = 25) Non-VA Cohort (n = 69) p-Value Decision-making engagement preference (%) n (%) n (%) 0.002  Physician-only or mostly physician 1 (4) 2 (3)  Equal patient-physician 14 (56) 58 (84)  Mostly patient or patient-only 6 (24) 9 (13)  No answer 4 (16) 0 (0) Initial treatment preference n (%) n (%) <0.001  Non-surgical 5 (20) 4 (6)  Surgical 4 (16) 49 (71)  Unsure/Delay 15 (60) 16 (23)  No answer 1 (4) 0 (0) Result VA Cohort (n = 25) Non-VA Cohort (n = 69) p-Value Decision-making engagement preference (%) n (%) n (%) 0.002  Physician-only or mostly physician 1 (4) 2 (3)  Equal patient-physician 14 (56) 58 (84)  Mostly patient or patient-only 6 (24) 9 (13)  No answer 4 (16) 0 (0) Initial treatment preference n (%) n (%) <0.001  Non-surgical 5 (20) 4 (6)  Surgical 4 (16) 49 (71)  Unsure/Delay 15 (60) 16 (23)  No answer 1 (4) 0 (0) Bold values represents significance as a p value < 0.05. Table IV. Comparison of Decision-Making Preferences Between VA Cohort and Non-VA Cohort with Complete Demographic and Decision-Making Preference Data Result VA Cohort (n = 25) Non-VA Cohort (n = 69) p-Value Decision-making engagement preference (%) n (%) n (%) 0.002  Physician-only or mostly physician 1 (4) 2 (3)  Equal patient-physician 14 (56) 58 (84)  Mostly patient or patient-only 6 (24) 9 (13)  No answer 4 (16) 0 (0) Initial treatment preference n (%) n (%) <0.001  Non-surgical 5 (20) 4 (6)  Surgical 4 (16) 49 (71)  Unsure/Delay 15 (60) 16 (23)  No answer 1 (4) 0 (0) Result VA Cohort (n = 25) Non-VA Cohort (n = 69) p-Value Decision-making engagement preference (%) n (%) n (%) 0.002  Physician-only or mostly physician 1 (4) 2 (3)  Equal patient-physician 14 (56) 58 (84)  Mostly patient or patient-only 6 (24) 9 (13)  No answer 4 (16) 0 (0) Initial treatment preference n (%) n (%) <0.001  Non-surgical 5 (20) 4 (6)  Surgical 4 (16) 49 (71)  Unsure/Delay 15 (60) 16 (23)  No answer 1 (4) 0 (0) Bold values represents significance as a p value < 0.05. VA patients generally expressed comfort in questioning the recommendations of their orthopedic providers with no significant changes following their initial visit (Table V). Table V. Pre- and Post-Visit Veteran Comfort with Questioning Provider Recommendations Result VA Cohort (n = 25) Pre-visit comfort questioning surgeon recommendations  Comfortable or very comfortable 18 (72)  Neutral 4 (16)  Not comfortable 1 (4)  No answer 2 (8) Post-visit comfort questioning surgeon recommendations  Comfortable or very comfortable 18 (72)  Neutral 2 (8)  Not comfortable 0 (0)  No answer 5 (20) Result VA Cohort (n = 25) Pre-visit comfort questioning surgeon recommendations  Comfortable or very comfortable 18 (72)  Neutral 4 (16)  Not comfortable 1 (4)  No answer 2 (8) Post-visit comfort questioning surgeon recommendations  Comfortable or very comfortable 18 (72)  Neutral 2 (8)  Not comfortable 0 (0)  No answer 5 (20) Table V. Pre- and Post-Visit Veteran Comfort with Questioning Provider Recommendations Result VA Cohort (n = 25) Pre-visit comfort questioning surgeon recommendations  Comfortable or very comfortable 18 (72)  Neutral 4 (16)  Not comfortable 1 (4)  No answer 2 (8) Post-visit comfort questioning surgeon recommendations  Comfortable or very comfortable 18 (72)  Neutral 2 (8)  Not comfortable 0 (0)  No answer 5 (20) Result VA Cohort (n = 25) Pre-visit comfort questioning surgeon recommendations  Comfortable or very comfortable 18 (72)  Neutral 4 (16)  Not comfortable 1 (4)  No answer 2 (8) Post-visit comfort questioning surgeon recommendations  Comfortable or very comfortable 18 (72)  Neutral 2 (8)  Not comfortable 0 (0)  No answer 5 (20) DISCUSSION Military law and healthcare decision-making are topics rarely, if ever, mentioned together. Military physicians’ first obligation is to the Nation, not to their patients.1,14 Because many aspects of their healthcare are mandated, without the opportunity for decision-making, we hypothesized that former military patients would have lower decision quality when considering treatment for knee osteoarthritis than a similar non-VA cohort.1 Whether these challenges are real has not been studied previously. Our goal was to begin the process of determining if former military service portends an inherent disadvantage with medical decision-making. Our results indicate that VA patients with knee osteoarthritis have lower decision quality than non-VA patients. High-quality treatment decisions are based upon the best available evidence, match treatment decisions with patients’ individual goals and beliefs, and establish realistic outcome expectations.4–6,9,10,15 A high-quality decision, therefore, requires the patient to be educated regarding positive and negative aspects of the proposed treatments, which is often achieved using decision aids. Decision aids have a demonstrated positive effect on decision quality.10,16 The current investigation focused on decision quality in the absence of decision aids in order to establish baseline decision quality in the VA. Our academic center’s orthopedic registry does contain patients with decision aid exposure. While these results were not explored in this study, prior work suggests that exposure to decision aids did result in a higher total knowledge score (57.2, SD 19) and concordance score (74.3, SD 44), although these differences were more modest than reported previously.10,16 Shared decision-making for hip and knee osteoarthritis has demonstrated benefits in the civilian population. These benefits include higher decision quality, engagement, satisfaction, and overall health in addition to lower rates of surgery, demographic-based treatment disparity, anxiety and decisional conflict, hospital readmission, and healthcare costs.4–6,10,16–25 There could be tangible benefits for Veterans and healthcare costs if shared decision-making was adopted in VA. In 2016 VA performed over 15,000 total hip and knee replacements. A formal shared decision-making program showed a 26% and 38% reduction in total hip and knee replacements respectively.17 Extrapolated to the volume of total hip and knee arthroplasty performed by VA this would represent a savings of $150 M in the private sector.26,27 Despite the high prevalence of osteoarthritis in VA and the important benefits of shared decision-making for patients undergoing treatment, the need for and feasibility of a shared decision-making program in the VA has not been assessed. We believe this deficiency is due most likely to a lack of shared decision-making awareness and training amongst orthopedic surgeons in the VA. There are limitations to consider in interpreting our findings. Socioeconomic status (SES) has known effects on decision quality and we were unable to control for SES in our statistical model.10 This is relevant to the current study because prior work demonstrated that Veterans had lower mean SES than civilians,28 however, these differences are lessened in our current Veteran population.29–32 Second, our non-VA cohort data do not include history of military service. This information is available for our academic institution as a whole but it does not allow us to exclude Veterans. We must, therefore, assume that approximately 6% of our control patients are Veterans, albeit these small differences are unlikely to change the conclusions drawn from our findings.33 In conclusion, in the absence of decision aid exposure, decision quality is lower in patients seeking treatment for knee osteoarthritis in VA compared with non-VA patients. This finding presents an opportunity to improve care practices while reducing treatment costs for patients with knee osteoarthritis in VA.4–6,10,17–25 A trial of decision aids for osteoarthritis is needed to determine if these positive effects seen in non-VA populations will be experienced similarly in VA. Acknowledgements The authors wish to thank Mondae Dupuis, LPN, for her dedication to this investigation. Funding BJK receives partial support through NIAMS (P60AR062799; P60AR048094). ERH received partial support through VA (V1CDA2015-34). AJT receives support through the NIH BD2K grant (T32LM012204). References 1 Maya H : Patient-physician relationships in the military . Health L Pol’y 2008 ; 2 ( 1 ): 70 – 81 . 2 Army Regulation 40–562, Immunizations and Chemoprophylaxis 8-3 (U.S. Dep’t of the Army 2006)(‘the FDA May Decide That Potential Recipients of a Drug Under a[N Emergency Use Authorization] Should Have the Option to Refuse It. the President May Waive This Option for Military Personnel.’).; 2014 : 1 – 1 . 3 Army Regulation 600–20, Army Command Policy 5-4(a) (U.S. Dep’t of the Army 2006)(“a Soldier on Active Duty or Active Duty for Training Will Usually Be Required to Submit to Medical Care Considered Necessary to Preserve His or Her Lef, Alleviate Undue Suffering, or Protect or Maintain the Health of Others.”). 2014 : 1 – 1 . 4 Bozic KJ , Chiu V : Emerging ideas: shared decision making in patients with osteoarthritis of the hip and knee . Clin Orthop Relat Res 2011 ; 469 ( 7 ): 2081 – 5 . doi:10.1007/s11999-010-1740-7 . Google Scholar CrossRef Search ADS PubMed 5 Bozic KJ , Lau E , Kurtz S , et al. : Patient-related risk factors for periprosthetic joint infection and postoperative mortality following total hip arthroplasty in medicare patients . J Bone Joint Surg 2012 ; 94 ( 9 ): 794 – 800 . doi:10.2106/JBJS.K.00072 . Google Scholar CrossRef Search ADS PubMed 6 Sepucha KR : Policy support for patient-centered care: the need for measurable improvements in decision quality . Health Aff 2004 . doi:10.1377/hlthaff.var.54 . 7 Strull WM , Lo B , Charles G : Do patients want to participate in medical decision making? JAMA 1984 ; 252 ( 21 ): 2990 – 4 . doi:10.1002/cbin.10492 . Google Scholar CrossRef Search ADS PubMed 8 Braddock CH III , Edwards KA , Hasenberg NM , Laidley TL , Levinson W : Informed decision making in outpatient practice: time to get back to basics . JAMA 1999 ; 282 ( 24 ): 2313 – 20 . Google Scholar CrossRef Search ADS PubMed 9 Elwyn G , O’Connor A , Stacey D , et al. : Developing a quality criteria framework for patient decision aids: online international Delphi consensus process . BMJ 2006 ; 333 ( 7565 ): 417 . doi:10.1136/bmj.38926.629329.AE . Google Scholar CrossRef Search ADS PubMed 10 Sepucha KR , Stacey D , Clay CF , et al. : Decision quality instrument for treatment ofhip and knee osteoarthritis: a psychometricevaluation . BMC Musculoskelet Disord 2011 ; 12 ( 1 ): 149 . doi:10.1186/1471-2474-12-149 . Google Scholar CrossRef Search ADS PubMed 11 Kellgren JH , Lawrence JS : Radiological assessment of osteo-arthrosis . Ann Rheum Dis 1957 ; 16 ( 4 ): 494 – 502 . Google Scholar CrossRef Search ADS PubMed 12 Garfield S , Smith F , Francis SA , Chalmers C : Can patients’ preferences for involvement in decision-making regarding the use of medicines be predicted? Patient Educ Counsel 2007 ; 66 ( 3 ): 361 – 7 . doi:10.1016/j.pec.2007.01.012 . Google Scholar CrossRef Search ADS 13 Sepucha K , Feibelmann S , Chang Y , et al. : Factors associated with the quality of patients’ surgical decisions for treatment of hip and knee osteoarthritis . J Am Coll Surg 2013 ; 217 ( 4 ): 694 – 701 . doi:10.1016/j.jamcollsurg.2013.06.002 . Google Scholar CrossRef Search ADS PubMed 14 Clark PA : Medical ethics at Guantanamo Bay and Abu Ghraib: the problem of dual loyalty . J Law Med Ethics 2006 ; 34 ( 3 ): 570 – 80 . doi:10.1111/j.1748-720×.2006.00071.x . Google Scholar CrossRef Search ADS PubMed 15 Strull WM , Lo B , Charles G : Do patients want to participate in medical decision making? JAMA 1984 ; 252 ( 21 ): 2990 – 4 . Google Scholar CrossRef Search ADS PubMed 16 Sepucha K , Atlas SJ , Chang Y , et al. : Patient decision aids improve decision quality and patient experience and reduce surgical rates in routine orthopaedic care: a prospective cohort study . J Bone Joint Surg Am 2017 ; 99 ( 15 ): 1253 – 60 . doi:10.2106/JBJS.16.01045 . Google Scholar CrossRef Search ADS PubMed 17 Arterburn D , Wellman R , Westbrook E , et al. . Introducing Decision Aids At Group : Health was linked to sharply lower hip and knee surgery rates and costs . Health Aff 2012 ; 31 ( 9 ): 2094 – 2104 . doi:10.1377/hlthaff.2011.0686 . Google Scholar CrossRef Search ADS 18 Youm J , Chenok KE , Belkora J , Chiu V , Bozic KJ : The emerging case for shared decision making in orthopaedics . Instr Course Lect 2013 ; 62 : 587 – 94 . Google Scholar PubMed 19 O’Connor AM , Llewellyn-Thomas HA , Flood AB : Modifying unwarranted variations in health care: shared decision making using patient decision aids . Health Aff 2004 . doi:10.1377/hlthaff.var.63 . 20 Griffin SJ : Effect on health-related outcomes of interventions to alter the interaction between patients and practitioners: a systematic review of trials . Ann Fam Med 2004 ; 2 ( 6 ): 595 – 608 . doi:10.1370/afm.142 . Google Scholar CrossRef Search ADS PubMed 21 Borkhoff CM , Hawker GA : How to Make Sure Your Patient with Osteoarthritis Gets the Best Care . Geriatrics and Aging 2008 ; 11 : 8 . 22 Weng HH , Kaplan RM , Boscardin WJ , et al. : Development of a decision aid to address racial disparities in utilization of knee replacement surgery . Arthritis Rheum 2007 ; 57 ( 4 ): 568 – 75 . doi:10.1002/art.22670 . Google Scholar CrossRef Search ADS PubMed 23 Stacey D , Hawker G , Dervin G , et al. : Decision aid for patients considering total kneearthroplasty with preference report for surgeons:a pilot randomized controlled trial . BMC Musculoskelet Disord 2014 ; 15 ( 1 ): 1 – 10 . doi:10.1186/1471-2474-15-54 . Google Scholar CrossRef Search ADS PubMed 24 Cohen S , Hartley S , Mavi J , Vest B , Wilson M : Veteran experiences related to participation in shared medical appointments . Mil Med 2012 ; 177 ( 11 ): 1287 – 92 . doi:10.7205/MILMED-D-12-00212 . Google Scholar CrossRef Search ADS PubMed 25 Lurie JD , Weinstein JN : Shared decision-making and the orthopaedic workforce . Clin Orthop Relat Res 2001 ; 385 : 68 – 75 . Google Scholar CrossRef Search ADS 26 Coyne F : A Study of Cost Variations for Knee and Hip Replacement Surgeries in the U.S. January 2015:1–9. 27 Losina E , Walensky RP , Kessler CL , et al. : Cost-effectiveness of total knee arthroplasty in the United States: patient risk and hospital volume . Arch Intern Med 2009 ; 169 ( 12 ): 1113 – 21 ; discussion 1121–2. doi:10.1001/archinternmed.2009.136 . Google Scholar CrossRef Search ADS PubMed 28 Villemez WJ , Kasarda JD : Veteran status and socioeconomic attainment . Arm Forces Soc 1976 . doi:10.1177/AFSA_2_3;subPage:string:Access . 29 Teachman J , Tedrow L : Joining up: Did military service in the early all volunteer era affect subsequent civilian income? Soc Sci Res 2007 ; 36 ( 4 ): 1447 – 74 . doi:10.1016/j.ssresearch.2007.03.002 . Google Scholar CrossRef Search ADS 30 Teachman J : Military service and educational attainment in the all-volunteer era . Sociol Educ 2007 ; 80.4 : 359 – 74 . 31 Teachman J : Military service during the Vietnam era: Were there consequences for subsequent civilian earnings? Soc Forces 2004 ; 83.2 : 709 – 30 . 32 Bailey AK : Military employment and spatial mobility across the life course. Life-course perspectives on military service , 2013 ; 185 – 99 . 33 Bagalman E : The number of veterans that use VA Health Care Services: A Fact Sheet . Congressional Research Service . 2014 . Author notes The views expressed are solely those of the authors and do not reflect the official policy or position or the VA or the U.S. Government. Published by Oxford University Press on behalf of the Association of Military Surgeons of the United States 2018. This work is written by (a) US Government employee(s) and is in the public domain in the US. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Military Medicine Oxford University Press

Military Service and Decision Quality in the Management of Knee Osteoarthritis

Military Medicine , Volume Advance Article (7) – Jun 28, 2018

Loading next page...
 
/lp/ou_press/military-service-and-decision-quality-in-the-management-of-knee-0epynpplgU
Publisher
Oxford University Press
Copyright
Published by Oxford University Press on behalf of the Association of Military Surgeons of the United States 2018.
ISSN
0026-4075
eISSN
1930-613X
D.O.I.
10.1093/milmed/usy104
Publisher site
See Article on Publisher Site

Abstract

Abstract Background Decision quality measures the degree to which care decisions are knowledge-based and value-aligned. Because military service emphasizes hierarchy, command, and mandates some healthcare decisions, military service may attenuate patient autonomy in healthcare decisions and lower decision quality. VA is the nation’s largest provider of orthopedic care. We compared decision quality in a sample of VA and non-VA patients seeking care for knee osteoarthritis. Methods Our study sample consisted of patients newly referred to our orthopedic clinic for the management of knee osteoarthritis. None of the study patients were exposed to a knee osteoarthritis decision aid. Consenting patients were administered the Hip/Knee Decision Quality Instrument (HK-DQI). In addition, they were surveyed about decision-making preferences and demographics. We compared results to a non-VA cohort from our academic institution’s arthroplasty database. Results The HK-DQI Knowledge Score was lower in the VA cohort (45%, SD = 22, n = 25) compared with the non-VA cohort (53%, SD = 21, n = 177) (p = 0.04). The Concordance Score was lower in the VA cohort (36%, SD = 49%) compared with the control cohort (70%, SD 46%) (p = 0.003). Non-VA patients were more likely to make a high-quality decision (p = 0.05). Non-VA patients were more likely to favor a shared decision-making process (p = 0.002). Conclusions Decision quality is lower in Veterans with knee osteoarthritis compared with civilians, placing them at risk for lower treatment satisfaction and possibly unwarranted surgical utilization. Our future work will examine if this difference is from conditioned military service behaviors or confounding demographic factors, and if conventional shared decision-making techniques will correct this deficiency. INTRODUCTION Military law requires “the subordination of the desires and interests of the individual to the needs of the Service.”1 This requirement extends to healthcare where medical decisions can be made for military patients based on the interests of Nation and Command.2,3 This relationship further accentuates the normal imbalance of power observed between physicians and patients. This system may create gaps in trust between patients and providers and reduce patients’ willingness to question their providers’ recommendations, thereby reducing patients’ understanding of proposed treatments.1 These conditioned behaviors of military service may persist after discharge and attenuate Veterans’ healthcare decision-making autonomy – a key component in patient-centered care. Patient-centered care occurs when the chosen treatment considers the needs, values, and preferences of the well-informed patient.4–7 Decision quality is a measurable indicator of patient-centered care.8,9 Patient-centered care may be achieved through measurement and enhancement of decision quality. High decision quality requires two endpoints: (1) patients’ decisions are based on the best evidence; and (2) treatment decisions are consistent with the treatment features that matter most to the informed individual patient.9,10 Low decision quality may be improved through shared decision-making, a process of deliberate, collaborative treatment selection based upon the best available evidence supplemented with decision aids.10 However, this process may be more difficult in the setting of prior military service. In this observational pilot study, we measured decision quality in VA and non-VA patients considering treatment for knee osteoarthritis to determine if associations exist between decision quality and prior military service. We hypothesized that decision quality in the VA cohort would be lower than in non-VA patients, given the prior exposure to military service and its command structure and healthcare mandates. MATERIALS AND METHODS Study Population After obtaining Institutional Review Board approval (Veteran’s IRB of New England, VINNE), we educated, consented, and enrolled newly referred Veteran patients referred to our VA orthopedic clinic with a chief complaint of knee pain and grade III or IV osteoarthritis on plain radiographs.11 Patients with an inflammatory or autoimmune arthritis diagnosis, concomitant hip pain, a prior total joint replacement or surgical intervention for osteoarthritis involving any joint, a contraindication for total joint replacement, or a sensory, cognitive, or other impairment that would preclude informed consent were excluded. No VA patients had exposure to a decision aid. We screened patients for enrollment in person or via telephone. Twenty-five Veterans meeting inclusion criteria were consented and enrolled between October 2015 and September 2016. There were 24 men and one woman (Table I). The mean age was 66 yr. Educational levels for the patients were: graduate degree, n = 4; undergraduate, 6; high school graduate, 11; some high school, 3; no answer, 1. The patients’ races were: White, 20; Hispanic (White), 1; Hispanic (Black), 1; American Indian or Alaska Native, 1; Asian or Pacific Islander, 1; no answer, 1. Lack of gender diversity in this cohort precluded gender-oriented inter-cohort analyses. Table I. Patient Demographic Characteristics VA (n = 25) Non-VA (n = 69) p-Value Age (mean (SD)) 65.5 (12.1) 65.0 (10.5) 0.85 Gender = male (%) 24 (96.0) 27 (39.1) <0.001 Race (%)  American Indian or Alaska Native 1 (4.0) 0 (0.0) 0.031  Asian or Pacific Islander 1 (4.0) 1 (0.6)  Black 0 (0.0) 0 (0.0)  Hispanic, Black 1 (4.0) 0 (0.0)  Hispanic, White 1 (4.0) 0 (0.0)  Prefer not to answer 1 (4.0) 0 (0.0)  White 20 (80.0) 68 (98.6) Education (%)  Bachelors degree 6 (24.0) 16 (23.2) <0.001  Graduate degree 4 (16.0) 18 (26.1)  High school degree or equivalent 13 (52.0) 17 (24.6)  High school without graduation 1 (4.0) 4 (5.8)  Some college 0 (0.0) 14 (20.3)  Prefer not to answer 1 (4.0) 0 (0.0) VA (n = 25) Non-VA (n = 69) p-Value Age (mean (SD)) 65.5 (12.1) 65.0 (10.5) 0.85 Gender = male (%) 24 (96.0) 27 (39.1) <0.001 Race (%)  American Indian or Alaska Native 1 (4.0) 0 (0.0) 0.031  Asian or Pacific Islander 1 (4.0) 1 (0.6)  Black 0 (0.0) 0 (0.0)  Hispanic, Black 1 (4.0) 0 (0.0)  Hispanic, White 1 (4.0) 0 (0.0)  Prefer not to answer 1 (4.0) 0 (0.0)  White 20 (80.0) 68 (98.6) Education (%)  Bachelors degree 6 (24.0) 16 (23.2) <0.001  Graduate degree 4 (16.0) 18 (26.1)  High school degree or equivalent 13 (52.0) 17 (24.6)  High school without graduation 1 (4.0) 4 (5.8)  Some college 0 (0.0) 14 (20.3)  Prefer not to answer 1 (4.0) 0 (0.0) Table I. Patient Demographic Characteristics VA (n = 25) Non-VA (n = 69) p-Value Age (mean (SD)) 65.5 (12.1) 65.0 (10.5) 0.85 Gender = male (%) 24 (96.0) 27 (39.1) <0.001 Race (%)  American Indian or Alaska Native 1 (4.0) 0 (0.0) 0.031  Asian or Pacific Islander 1 (4.0) 1 (0.6)  Black 0 (0.0) 0 (0.0)  Hispanic, Black 1 (4.0) 0 (0.0)  Hispanic, White 1 (4.0) 0 (0.0)  Prefer not to answer 1 (4.0) 0 (0.0)  White 20 (80.0) 68 (98.6) Education (%)  Bachelors degree 6 (24.0) 16 (23.2) <0.001  Graduate degree 4 (16.0) 18 (26.1)  High school degree or equivalent 13 (52.0) 17 (24.6)  High school without graduation 1 (4.0) 4 (5.8)  Some college 0 (0.0) 14 (20.3)  Prefer not to answer 1 (4.0) 0 (0.0) VA (n = 25) Non-VA (n = 69) p-Value Age (mean (SD)) 65.5 (12.1) 65.0 (10.5) 0.85 Gender = male (%) 24 (96.0) 27 (39.1) <0.001 Race (%)  American Indian or Alaska Native 1 (4.0) 0 (0.0) 0.031  Asian or Pacific Islander 1 (4.0) 1 (0.6)  Black 0 (0.0) 0 (0.0)  Hispanic, Black 1 (4.0) 0 (0.0)  Hispanic, White 1 (4.0) 0 (0.0)  Prefer not to answer 1 (4.0) 0 (0.0)  White 20 (80.0) 68 (98.6) Education (%)  Bachelors degree 6 (24.0) 16 (23.2) <0.001  Graduate degree 4 (16.0) 18 (26.1)  High school degree or equivalent 13 (52.0) 17 (24.6)  High school without graduation 1 (4.0) 4 (5.8)  Some college 0 (0.0) 14 (20.3)  Prefer not to answer 1 (4.0) 0 (0.0) Study Measures We measured our primary outcome, decision quality, using the Hip-Knee Decision Quality Instrument (HK-DQI).10 We also measured patient demographics (age, gender, education), the preferred level of decision-making engagement (patient and doctor equally, mostly patient, mostly doctor, totally patient, totally doctor), and patients comfort level in questioning the recommendations of their orthopedic provider.12 Observational Non-VA Control Cohort We drew our observational control cohort (non-VA) from a convenience sample derived from the prior 3 yr of our academic medical center’s orthopedic registry. This registry collects the HK-DQI, demographic, and decision-making preference data. We obtained separate Institutional Review Board approval for this query. The database contained 1,636 patients for the specified time period. Of these patients, 177 had undergone treatment for knee osteoarthritis and had completed the HK-DQI prior to treatment and had no exposure to a decision aid; these patients composed our unadjusted, bivariate analysis of decision quality (Fig. 1). Of these 177 patients, 69 had complete demographic and decision-making preference data sets that permitted multivariate logistic regression analyses. This final sample included 27 men and 42 women (Table I). We did not gender-match our cohort based upon prior work showing no gender-based differences in decision quality.13 The mean age was 65 yr (SD 11). Educational levels for this cohort were: graduate degree, n = 18; undergraduate, 16; some college, 14; high school graduate, 17; and some high school, 4. The patients’ races were: White, 68; Asian or Pacific Islander, 1. Lack of racial diversity in this cohort precluded race-oriented inter-cohort analyses. There were significant differences between the cohorts in terms of gender, race, and education. FIGURE 1. View largeDownload slide Composition of VA cohort and observational non-VA cohort for the current investigation. FIGURE 1. View largeDownload slide Composition of VA cohort and observational non-VA cohort for the current investigation. Sample Size Calculations We assessed sample size adequacy through power analysis for our primary outcome, decision quality, using available medical literature. Based on reported standard deviation values for decision quality scores with and without decision aid exposure, 18 and 19,10 respectively, we estimated that we would be able to detect a difference in decision quality scores of 12.0 points based on an unpaired t-test with a two-sided significance level of 0.05 and a power of 0.80 for samples of size 25. This calculation conservatively assumes that the intraclass correlation for the effect of matching is effectively zero. By comparison, the validation study of the HK-DQI demonstrated a difference of 14 between cohorts without and with shared decision-making,10 so our proposed detectable difference of 12 points seemed reasonable. Statistical Analysis Statistical analyses were conducted in R (version 3.3.0). Differences in means were calculated using a Student’s t-test or Chi-squared test. Associations between outcomes (TKS, CS) and covariates (race, education, decision-making preference) were quantified using univariate linear models between each covariate and each outcome. Compliance with Ethical Standards One author (ERH) received funding (V1CDA2015-34) that was in support of this project. RESULTS The mean HK-DQI TKS in the VA cohort (n = 25) was 45 (SD 21). This value was significantly lower than the non-VA cohort (n = 177) of 53 (SD 19) (p = 0.04). The mean HK-DQI Concordance Score (CS) in the VA cohort was 36 (SD 48). This value was significantly lower than the non-VA cohort 70 (SD 46) (p = 0.003). The mean HK-DQI DPS in the VA cohort was 51 (SD 40) (Table II). This value was not measured in our comparative cohort; however, its relation to values reported previously is considered in the discussion. The rate of VA patients making a high-quality decision (20%) was significantly lower than the rate of non-VA patients making a high-quality decision (69%) (p = 0.05). Table II. Unadjusted Univariate, Intercohort Comparison of HK-DQI Scores Hip-Knee Decision Quality Instrument Scores VA Cohort (n = 25) Non-VA Cohort (n = 177) p-Value Total Knowledge Score (%, SD) 45 (21) 53 (19) 0.04 Concordance Score (%, SD) 36 (48) 70 (46) 0.003 Decision Process Score (%, SD) 51 (40) NA NA High-Quality Decision (n, (%)) 5 (20) 69 (41) 0.05 Hip-Knee Decision Quality Instrument Scores VA Cohort (n = 25) Non-VA Cohort (n = 177) p-Value Total Knowledge Score (%, SD) 45 (21) 53 (19) 0.04 Concordance Score (%, SD) 36 (48) 70 (46) 0.003 Decision Process Score (%, SD) 51 (40) NA NA High-Quality Decision (n, (%)) 5 (20) 69 (41) 0.05 Table II. Unadjusted Univariate, Intercohort Comparison of HK-DQI Scores Hip-Knee Decision Quality Instrument Scores VA Cohort (n = 25) Non-VA Cohort (n = 177) p-Value Total Knowledge Score (%, SD) 45 (21) 53 (19) 0.04 Concordance Score (%, SD) 36 (48) 70 (46) 0.003 Decision Process Score (%, SD) 51 (40) NA NA High-Quality Decision (n, (%)) 5 (20) 69 (41) 0.05 Hip-Knee Decision Quality Instrument Scores VA Cohort (n = 25) Non-VA Cohort (n = 177) p-Value Total Knowledge Score (%, SD) 45 (21) 53 (19) 0.04 Concordance Score (%, SD) 36 (48) 70 (46) 0.003 Decision Process Score (%, SD) 51 (40) NA NA High-Quality Decision (n, (%)) 5 (20) 69 (41) 0.05 When we analyzed TKS with respect to patient age and educational levels factors (age, gender, race, and education), we found no significant differences between VA and non-VA cohorts, however, our study was powered insufficiently to support these analyses. When we analyzed CS with respect to patient factors, we again found no significant differences between VA and non-VA cohorts, but our study was powered insufficiently to support these analyses. We analyzed the VA cohort for variables predicting a high-quality decision without significant findings (Table III). Table III. Domains, Variables, and Bivariate Analyses for Predicting a High-Quality Decision* in VA Cohort. *TKS > 60 and positive concordance Domains and Variables High-Quality Decision 20% (n = 5) Low-Quality Decision 80% (n = 20) OR (95% CI) p-Value Age group 1.000  Age <60 (ref) 20% (1) 25% (5)    Age >60 80% (4) 75% (15) 0.638 (0.335–1.220) Gender 1.000  Female (ref) 0% (0) 5% (1)    Male 100% (5) 95% (19) 0.923 (0.502–1.685) Education 0.99  College or more (ref) 40% (2) 40% (8)    High school degree 60% (3) 50% (10) 0.914 (0.161–7.082)  High school unfinished or less 0% (0) 10% (2) 0.750 (0.085–7.503) Initial treatment preference 0.43  Non-surgical (ref) 40% (2) 15% (3)    Surgical 20% (1) 15% (3) 0.4286 (0.130–1.366)  Unsure/delay 40% (2) 65% (13) 0.132 (0.016–1.631)  No answer 0% (0) 5% (1) NA Final treatment selection 0.39  Non-surgical (ref) 80% (4) 95% (19)    Surgical 20% (1) 5% (1) NA  No answer 0% (0) 0% (0) NA Pre-visit comfort level questioning provider 0.22  Comfortable (ref) 80% (4) 70% (14)    Neutral 0% (0) 20% (4) 1.000 (3.6e-16 – 2.50)  Not comfortable 20% (1) 0% (0) NA  No answer 0% (0) 10% (2) NA Post-visit comfort level questioning provider 0.24  Comfortable (ref) 80% (4) 60% (14)    Neutral 20% (1) 5% (1) 3.500 (0.105–123.6)  Not comfortable 0% (0) 0% (0) 1.822 (0.422–12.415)  No answer 0% (0) 25% (5) NA Domains and Variables High-Quality Decision 20% (n = 5) Low-Quality Decision 80% (n = 20) OR (95% CI) p-Value Age group 1.000  Age <60 (ref) 20% (1) 25% (5)    Age >60 80% (4) 75% (15) 0.638 (0.335–1.220) Gender 1.000  Female (ref) 0% (0) 5% (1)    Male 100% (5) 95% (19) 0.923 (0.502–1.685) Education 0.99  College or more (ref) 40% (2) 40% (8)    High school degree 60% (3) 50% (10) 0.914 (0.161–7.082)  High school unfinished or less 0% (0) 10% (2) 0.750 (0.085–7.503) Initial treatment preference 0.43  Non-surgical (ref) 40% (2) 15% (3)    Surgical 20% (1) 15% (3) 0.4286 (0.130–1.366)  Unsure/delay 40% (2) 65% (13) 0.132 (0.016–1.631)  No answer 0% (0) 5% (1) NA Final treatment selection 0.39  Non-surgical (ref) 80% (4) 95% (19)    Surgical 20% (1) 5% (1) NA  No answer 0% (0) 0% (0) NA Pre-visit comfort level questioning provider 0.22  Comfortable (ref) 80% (4) 70% (14)    Neutral 0% (0) 20% (4) 1.000 (3.6e-16 – 2.50)  Not comfortable 20% (1) 0% (0) NA  No answer 0% (0) 10% (2) NA Post-visit comfort level questioning provider 0.24  Comfortable (ref) 80% (4) 60% (14)    Neutral 20% (1) 5% (1) 3.500 (0.105–123.6)  Not comfortable 0% (0) 0% (0) 1.822 (0.422–12.415)  No answer 0% (0) 25% (5) NA Bold values represents significance as a p value < 0.05. Table III. Domains, Variables, and Bivariate Analyses for Predicting a High-Quality Decision* in VA Cohort. *TKS > 60 and positive concordance Domains and Variables High-Quality Decision 20% (n = 5) Low-Quality Decision 80% (n = 20) OR (95% CI) p-Value Age group 1.000  Age <60 (ref) 20% (1) 25% (5)    Age >60 80% (4) 75% (15) 0.638 (0.335–1.220) Gender 1.000  Female (ref) 0% (0) 5% (1)    Male 100% (5) 95% (19) 0.923 (0.502–1.685) Education 0.99  College or more (ref) 40% (2) 40% (8)    High school degree 60% (3) 50% (10) 0.914 (0.161–7.082)  High school unfinished or less 0% (0) 10% (2) 0.750 (0.085–7.503) Initial treatment preference 0.43  Non-surgical (ref) 40% (2) 15% (3)    Surgical 20% (1) 15% (3) 0.4286 (0.130–1.366)  Unsure/delay 40% (2) 65% (13) 0.132 (0.016–1.631)  No answer 0% (0) 5% (1) NA Final treatment selection 0.39  Non-surgical (ref) 80% (4) 95% (19)    Surgical 20% (1) 5% (1) NA  No answer 0% (0) 0% (0) NA Pre-visit comfort level questioning provider 0.22  Comfortable (ref) 80% (4) 70% (14)    Neutral 0% (0) 20% (4) 1.000 (3.6e-16 – 2.50)  Not comfortable 20% (1) 0% (0) NA  No answer 0% (0) 10% (2) NA Post-visit comfort level questioning provider 0.24  Comfortable (ref) 80% (4) 60% (14)    Neutral 20% (1) 5% (1) 3.500 (0.105–123.6)  Not comfortable 0% (0) 0% (0) 1.822 (0.422–12.415)  No answer 0% (0) 25% (5) NA Domains and Variables High-Quality Decision 20% (n = 5) Low-Quality Decision 80% (n = 20) OR (95% CI) p-Value Age group 1.000  Age <60 (ref) 20% (1) 25% (5)    Age >60 80% (4) 75% (15) 0.638 (0.335–1.220) Gender 1.000  Female (ref) 0% (0) 5% (1)    Male 100% (5) 95% (19) 0.923 (0.502–1.685) Education 0.99  College or more (ref) 40% (2) 40% (8)    High school degree 60% (3) 50% (10) 0.914 (0.161–7.082)  High school unfinished or less 0% (0) 10% (2) 0.750 (0.085–7.503) Initial treatment preference 0.43  Non-surgical (ref) 40% (2) 15% (3)    Surgical 20% (1) 15% (3) 0.4286 (0.130–1.366)  Unsure/delay 40% (2) 65% (13) 0.132 (0.016–1.631)  No answer 0% (0) 5% (1) NA Final treatment selection 0.39  Non-surgical (ref) 80% (4) 95% (19)    Surgical 20% (1) 5% (1) NA  No answer 0% (0) 0% (0) NA Pre-visit comfort level questioning provider 0.22  Comfortable (ref) 80% (4) 70% (14)    Neutral 0% (0) 20% (4) 1.000 (3.6e-16 – 2.50)  Not comfortable 20% (1) 0% (0) NA  No answer 0% (0) 10% (2) NA Post-visit comfort level questioning provider 0.24  Comfortable (ref) 80% (4) 60% (14)    Neutral 20% (1) 5% (1) 3.500 (0.105–123.6)  Not comfortable 0% (0) 0% (0) 1.822 (0.422–12.415)  No answer 0% (0) 25% (5) NA Bold values represents significance as a p value < 0.05. In the subset with complete demographic and decision-making preference data, the mean TKS scores in the comparative cohort decreased from 53 to 47, which was higher than the VA cohort but no longer significant; CS was unchanged but no longer significant due to insufficient power (Table IV). The rate of non-VA patients making high-quality decisions also decreased. Survey results showed that no patients in the VA cohort reported preference for patient-only decision-making; 24% preferred mostly patient decision-making; 56% preferred equal patient-physician decision-making; no patients preferred mostly physician decision-making; one patient preferred physician-only decision-making; and 16% did not answer (Table IV). In the non-VA cohort, 13% of patients preferred patient-only or mostly patient decision-making; 84% preferred equal patient-physician decision-making; 3% of patients preferred mostly physician decision-making. The non-VA cohort had a significantly higher preference for a shared patient and physician decision-making process compared with the VA cohort (p = 0.002). The non-VA cohort also had a significantly higher initial preference for surgical treatment compared with the VA cohort (p < 0.001). Table IV. Comparison of Decision-Making Preferences Between VA Cohort and Non-VA Cohort with Complete Demographic and Decision-Making Preference Data Result VA Cohort (n = 25) Non-VA Cohort (n = 69) p-Value Decision-making engagement preference (%) n (%) n (%) 0.002  Physician-only or mostly physician 1 (4) 2 (3)  Equal patient-physician 14 (56) 58 (84)  Mostly patient or patient-only 6 (24) 9 (13)  No answer 4 (16) 0 (0) Initial treatment preference n (%) n (%) <0.001  Non-surgical 5 (20) 4 (6)  Surgical 4 (16) 49 (71)  Unsure/Delay 15 (60) 16 (23)  No answer 1 (4) 0 (0) Result VA Cohort (n = 25) Non-VA Cohort (n = 69) p-Value Decision-making engagement preference (%) n (%) n (%) 0.002  Physician-only or mostly physician 1 (4) 2 (3)  Equal patient-physician 14 (56) 58 (84)  Mostly patient or patient-only 6 (24) 9 (13)  No answer 4 (16) 0 (0) Initial treatment preference n (%) n (%) <0.001  Non-surgical 5 (20) 4 (6)  Surgical 4 (16) 49 (71)  Unsure/Delay 15 (60) 16 (23)  No answer 1 (4) 0 (0) Bold values represents significance as a p value < 0.05. Table IV. Comparison of Decision-Making Preferences Between VA Cohort and Non-VA Cohort with Complete Demographic and Decision-Making Preference Data Result VA Cohort (n = 25) Non-VA Cohort (n = 69) p-Value Decision-making engagement preference (%) n (%) n (%) 0.002  Physician-only or mostly physician 1 (4) 2 (3)  Equal patient-physician 14 (56) 58 (84)  Mostly patient or patient-only 6 (24) 9 (13)  No answer 4 (16) 0 (0) Initial treatment preference n (%) n (%) <0.001  Non-surgical 5 (20) 4 (6)  Surgical 4 (16) 49 (71)  Unsure/Delay 15 (60) 16 (23)  No answer 1 (4) 0 (0) Result VA Cohort (n = 25) Non-VA Cohort (n = 69) p-Value Decision-making engagement preference (%) n (%) n (%) 0.002  Physician-only or mostly physician 1 (4) 2 (3)  Equal patient-physician 14 (56) 58 (84)  Mostly patient or patient-only 6 (24) 9 (13)  No answer 4 (16) 0 (0) Initial treatment preference n (%) n (%) <0.001  Non-surgical 5 (20) 4 (6)  Surgical 4 (16) 49 (71)  Unsure/Delay 15 (60) 16 (23)  No answer 1 (4) 0 (0) Bold values represents significance as a p value < 0.05. VA patients generally expressed comfort in questioning the recommendations of their orthopedic providers with no significant changes following their initial visit (Table V). Table V. Pre- and Post-Visit Veteran Comfort with Questioning Provider Recommendations Result VA Cohort (n = 25) Pre-visit comfort questioning surgeon recommendations  Comfortable or very comfortable 18 (72)  Neutral 4 (16)  Not comfortable 1 (4)  No answer 2 (8) Post-visit comfort questioning surgeon recommendations  Comfortable or very comfortable 18 (72)  Neutral 2 (8)  Not comfortable 0 (0)  No answer 5 (20) Result VA Cohort (n = 25) Pre-visit comfort questioning surgeon recommendations  Comfortable or very comfortable 18 (72)  Neutral 4 (16)  Not comfortable 1 (4)  No answer 2 (8) Post-visit comfort questioning surgeon recommendations  Comfortable or very comfortable 18 (72)  Neutral 2 (8)  Not comfortable 0 (0)  No answer 5 (20) Table V. Pre- and Post-Visit Veteran Comfort with Questioning Provider Recommendations Result VA Cohort (n = 25) Pre-visit comfort questioning surgeon recommendations  Comfortable or very comfortable 18 (72)  Neutral 4 (16)  Not comfortable 1 (4)  No answer 2 (8) Post-visit comfort questioning surgeon recommendations  Comfortable or very comfortable 18 (72)  Neutral 2 (8)  Not comfortable 0 (0)  No answer 5 (20) Result VA Cohort (n = 25) Pre-visit comfort questioning surgeon recommendations  Comfortable or very comfortable 18 (72)  Neutral 4 (16)  Not comfortable 1 (4)  No answer 2 (8) Post-visit comfort questioning surgeon recommendations  Comfortable or very comfortable 18 (72)  Neutral 2 (8)  Not comfortable 0 (0)  No answer 5 (20) DISCUSSION Military law and healthcare decision-making are topics rarely, if ever, mentioned together. Military physicians’ first obligation is to the Nation, not to their patients.1,14 Because many aspects of their healthcare are mandated, without the opportunity for decision-making, we hypothesized that former military patients would have lower decision quality when considering treatment for knee osteoarthritis than a similar non-VA cohort.1 Whether these challenges are real has not been studied previously. Our goal was to begin the process of determining if former military service portends an inherent disadvantage with medical decision-making. Our results indicate that VA patients with knee osteoarthritis have lower decision quality than non-VA patients. High-quality treatment decisions are based upon the best available evidence, match treatment decisions with patients’ individual goals and beliefs, and establish realistic outcome expectations.4–6,9,10,15 A high-quality decision, therefore, requires the patient to be educated regarding positive and negative aspects of the proposed treatments, which is often achieved using decision aids. Decision aids have a demonstrated positive effect on decision quality.10,16 The current investigation focused on decision quality in the absence of decision aids in order to establish baseline decision quality in the VA. Our academic center’s orthopedic registry does contain patients with decision aid exposure. While these results were not explored in this study, prior work suggests that exposure to decision aids did result in a higher total knowledge score (57.2, SD 19) and concordance score (74.3, SD 44), although these differences were more modest than reported previously.10,16 Shared decision-making for hip and knee osteoarthritis has demonstrated benefits in the civilian population. These benefits include higher decision quality, engagement, satisfaction, and overall health in addition to lower rates of surgery, demographic-based treatment disparity, anxiety and decisional conflict, hospital readmission, and healthcare costs.4–6,10,16–25 There could be tangible benefits for Veterans and healthcare costs if shared decision-making was adopted in VA. In 2016 VA performed over 15,000 total hip and knee replacements. A formal shared decision-making program showed a 26% and 38% reduction in total hip and knee replacements respectively.17 Extrapolated to the volume of total hip and knee arthroplasty performed by VA this would represent a savings of $150 M in the private sector.26,27 Despite the high prevalence of osteoarthritis in VA and the important benefits of shared decision-making for patients undergoing treatment, the need for and feasibility of a shared decision-making program in the VA has not been assessed. We believe this deficiency is due most likely to a lack of shared decision-making awareness and training amongst orthopedic surgeons in the VA. There are limitations to consider in interpreting our findings. Socioeconomic status (SES) has known effects on decision quality and we were unable to control for SES in our statistical model.10 This is relevant to the current study because prior work demonstrated that Veterans had lower mean SES than civilians,28 however, these differences are lessened in our current Veteran population.29–32 Second, our non-VA cohort data do not include history of military service. This information is available for our academic institution as a whole but it does not allow us to exclude Veterans. We must, therefore, assume that approximately 6% of our control patients are Veterans, albeit these small differences are unlikely to change the conclusions drawn from our findings.33 In conclusion, in the absence of decision aid exposure, decision quality is lower in patients seeking treatment for knee osteoarthritis in VA compared with non-VA patients. This finding presents an opportunity to improve care practices while reducing treatment costs for patients with knee osteoarthritis in VA.4–6,10,17–25 A trial of decision aids for osteoarthritis is needed to determine if these positive effects seen in non-VA populations will be experienced similarly in VA. Acknowledgements The authors wish to thank Mondae Dupuis, LPN, for her dedication to this investigation. Funding BJK receives partial support through NIAMS (P60AR062799; P60AR048094). ERH received partial support through VA (V1CDA2015-34). AJT receives support through the NIH BD2K grant (T32LM012204). References 1 Maya H : Patient-physician relationships in the military . Health L Pol’y 2008 ; 2 ( 1 ): 70 – 81 . 2 Army Regulation 40–562, Immunizations and Chemoprophylaxis 8-3 (U.S. Dep’t of the Army 2006)(‘the FDA May Decide That Potential Recipients of a Drug Under a[N Emergency Use Authorization] Should Have the Option to Refuse It. the President May Waive This Option for Military Personnel.’).; 2014 : 1 – 1 . 3 Army Regulation 600–20, Army Command Policy 5-4(a) (U.S. Dep’t of the Army 2006)(“a Soldier on Active Duty or Active Duty for Training Will Usually Be Required to Submit to Medical Care Considered Necessary to Preserve His or Her Lef, Alleviate Undue Suffering, or Protect or Maintain the Health of Others.”). 2014 : 1 – 1 . 4 Bozic KJ , Chiu V : Emerging ideas: shared decision making in patients with osteoarthritis of the hip and knee . Clin Orthop Relat Res 2011 ; 469 ( 7 ): 2081 – 5 . doi:10.1007/s11999-010-1740-7 . Google Scholar CrossRef Search ADS PubMed 5 Bozic KJ , Lau E , Kurtz S , et al. : Patient-related risk factors for periprosthetic joint infection and postoperative mortality following total hip arthroplasty in medicare patients . J Bone Joint Surg 2012 ; 94 ( 9 ): 794 – 800 . doi:10.2106/JBJS.K.00072 . Google Scholar CrossRef Search ADS PubMed 6 Sepucha KR : Policy support for patient-centered care: the need for measurable improvements in decision quality . Health Aff 2004 . doi:10.1377/hlthaff.var.54 . 7 Strull WM , Lo B , Charles G : Do patients want to participate in medical decision making? JAMA 1984 ; 252 ( 21 ): 2990 – 4 . doi:10.1002/cbin.10492 . Google Scholar CrossRef Search ADS PubMed 8 Braddock CH III , Edwards KA , Hasenberg NM , Laidley TL , Levinson W : Informed decision making in outpatient practice: time to get back to basics . JAMA 1999 ; 282 ( 24 ): 2313 – 20 . Google Scholar CrossRef Search ADS PubMed 9 Elwyn G , O’Connor A , Stacey D , et al. : Developing a quality criteria framework for patient decision aids: online international Delphi consensus process . BMJ 2006 ; 333 ( 7565 ): 417 . doi:10.1136/bmj.38926.629329.AE . Google Scholar CrossRef Search ADS PubMed 10 Sepucha KR , Stacey D , Clay CF , et al. : Decision quality instrument for treatment ofhip and knee osteoarthritis: a psychometricevaluation . BMC Musculoskelet Disord 2011 ; 12 ( 1 ): 149 . doi:10.1186/1471-2474-12-149 . Google Scholar CrossRef Search ADS PubMed 11 Kellgren JH , Lawrence JS : Radiological assessment of osteo-arthrosis . Ann Rheum Dis 1957 ; 16 ( 4 ): 494 – 502 . Google Scholar CrossRef Search ADS PubMed 12 Garfield S , Smith F , Francis SA , Chalmers C : Can patients’ preferences for involvement in decision-making regarding the use of medicines be predicted? Patient Educ Counsel 2007 ; 66 ( 3 ): 361 – 7 . doi:10.1016/j.pec.2007.01.012 . Google Scholar CrossRef Search ADS 13 Sepucha K , Feibelmann S , Chang Y , et al. : Factors associated with the quality of patients’ surgical decisions for treatment of hip and knee osteoarthritis . J Am Coll Surg 2013 ; 217 ( 4 ): 694 – 701 . doi:10.1016/j.jamcollsurg.2013.06.002 . Google Scholar CrossRef Search ADS PubMed 14 Clark PA : Medical ethics at Guantanamo Bay and Abu Ghraib: the problem of dual loyalty . J Law Med Ethics 2006 ; 34 ( 3 ): 570 – 80 . doi:10.1111/j.1748-720×.2006.00071.x . Google Scholar CrossRef Search ADS PubMed 15 Strull WM , Lo B , Charles G : Do patients want to participate in medical decision making? JAMA 1984 ; 252 ( 21 ): 2990 – 4 . Google Scholar CrossRef Search ADS PubMed 16 Sepucha K , Atlas SJ , Chang Y , et al. : Patient decision aids improve decision quality and patient experience and reduce surgical rates in routine orthopaedic care: a prospective cohort study . J Bone Joint Surg Am 2017 ; 99 ( 15 ): 1253 – 60 . doi:10.2106/JBJS.16.01045 . Google Scholar CrossRef Search ADS PubMed 17 Arterburn D , Wellman R , Westbrook E , et al. . Introducing Decision Aids At Group : Health was linked to sharply lower hip and knee surgery rates and costs . Health Aff 2012 ; 31 ( 9 ): 2094 – 2104 . doi:10.1377/hlthaff.2011.0686 . Google Scholar CrossRef Search ADS 18 Youm J , Chenok KE , Belkora J , Chiu V , Bozic KJ : The emerging case for shared decision making in orthopaedics . Instr Course Lect 2013 ; 62 : 587 – 94 . Google Scholar PubMed 19 O’Connor AM , Llewellyn-Thomas HA , Flood AB : Modifying unwarranted variations in health care: shared decision making using patient decision aids . Health Aff 2004 . doi:10.1377/hlthaff.var.63 . 20 Griffin SJ : Effect on health-related outcomes of interventions to alter the interaction between patients and practitioners: a systematic review of trials . Ann Fam Med 2004 ; 2 ( 6 ): 595 – 608 . doi:10.1370/afm.142 . Google Scholar CrossRef Search ADS PubMed 21 Borkhoff CM , Hawker GA : How to Make Sure Your Patient with Osteoarthritis Gets the Best Care . Geriatrics and Aging 2008 ; 11 : 8 . 22 Weng HH , Kaplan RM , Boscardin WJ , et al. : Development of a decision aid to address racial disparities in utilization of knee replacement surgery . Arthritis Rheum 2007 ; 57 ( 4 ): 568 – 75 . doi:10.1002/art.22670 . Google Scholar CrossRef Search ADS PubMed 23 Stacey D , Hawker G , Dervin G , et al. : Decision aid for patients considering total kneearthroplasty with preference report for surgeons:a pilot randomized controlled trial . BMC Musculoskelet Disord 2014 ; 15 ( 1 ): 1 – 10 . doi:10.1186/1471-2474-15-54 . Google Scholar CrossRef Search ADS PubMed 24 Cohen S , Hartley S , Mavi J , Vest B , Wilson M : Veteran experiences related to participation in shared medical appointments . Mil Med 2012 ; 177 ( 11 ): 1287 – 92 . doi:10.7205/MILMED-D-12-00212 . Google Scholar CrossRef Search ADS PubMed 25 Lurie JD , Weinstein JN : Shared decision-making and the orthopaedic workforce . Clin Orthop Relat Res 2001 ; 385 : 68 – 75 . Google Scholar CrossRef Search ADS 26 Coyne F : A Study of Cost Variations for Knee and Hip Replacement Surgeries in the U.S. January 2015:1–9. 27 Losina E , Walensky RP , Kessler CL , et al. : Cost-effectiveness of total knee arthroplasty in the United States: patient risk and hospital volume . Arch Intern Med 2009 ; 169 ( 12 ): 1113 – 21 ; discussion 1121–2. doi:10.1001/archinternmed.2009.136 . Google Scholar CrossRef Search ADS PubMed 28 Villemez WJ , Kasarda JD : Veteran status and socioeconomic attainment . Arm Forces Soc 1976 . doi:10.1177/AFSA_2_3;subPage:string:Access . 29 Teachman J , Tedrow L : Joining up: Did military service in the early all volunteer era affect subsequent civilian income? Soc Sci Res 2007 ; 36 ( 4 ): 1447 – 74 . doi:10.1016/j.ssresearch.2007.03.002 . Google Scholar CrossRef Search ADS 30 Teachman J : Military service and educational attainment in the all-volunteer era . Sociol Educ 2007 ; 80.4 : 359 – 74 . 31 Teachman J : Military service during the Vietnam era: Were there consequences for subsequent civilian earnings? Soc Forces 2004 ; 83.2 : 709 – 30 . 32 Bailey AK : Military employment and spatial mobility across the life course. Life-course perspectives on military service , 2013 ; 185 – 99 . 33 Bagalman E : The number of veterans that use VA Health Care Services: A Fact Sheet . Congressional Research Service . 2014 . Author notes The views expressed are solely those of the authors and do not reflect the official policy or position or the VA or the U.S. Government. Published by Oxford University Press on behalf of the Association of Military Surgeons of the United States 2018. This work is written by (a) US Government employee(s) and is in the public domain in the US.

Journal

Military MedicineOxford University Press

Published: Jun 28, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off