Abstract
The Patient Reported Outcomes Measurement System (PROMIS) offers standardized assessment measures of clinically relevant patient reported outcomes. This study evaluated the reliability and validity of select PROMIS measures with United States Military Veterans following orthopedic surgery. Data for the current study were collected as part of a pilot randomized control trial assessing the efficacy of a one day Acceptance and Commitment Therapy workshop on persistent post-surgical pain in at-risk Veterans undergoing orthopedic surgery. Sixty-seven participants completed surveys three months after surgery. Participants completed the following PROMIS instruments: PROMIS Anxiety Short Form 8a, PROMIS Depression Short Form 8b, and PROMIS pain interference short form 8a. PROMIS measures were compared to the Generalized Anxiety Disorder 7-item scale (GAD-7), the Patient Health Questionnaire 9-item scale (PHQ-9), and the Brief Pain Inventory (BPI) pain interference subscale (respectively). All three PROMIS measures demonstrated excellent internal consistency (Cronbach’s alphas ranged from .93 - .96) and each loaded onto a single factor. The PROMIS measures were moderately correlated with their respective comparison measures (r = .69 - .76). The PROMIS anxiety and PROMIS depression measures were highly correlated to one another (r = .91). Findings highlight the potential utility of these PROMIS measures in Veterans following orthopedic surgery and the overlap between the PROMIS depression and anxiety measures in this sample.
Introduction
Brief, comparable, standardized measures of pain, anxiety, and depression are needed in medical settings, to maximize comparability of findings across studies and patient populations. The patient-reported outcomes measurement information system (PROMIS), developed by the National Institutes of Health (NIH), provides self-report assessments of multiple mental, physical, and social domains. These measures were designed to standardize measurement of patient outcomes across a range of chronic conditions (NIH, 2018).
United States Military Veterans in medical settings may represent a unique medical population. Veterans have a heightened risk for distress-based conditions, including depressive and anxious disorders, (Pickett et al., 2015) and chronic pain is prevalent in Veterans (Haskell, Heapy, Reid, Papas, & Kerns, 2006). As such, validating brief measures of pain, depression, and anxiety in Veteran populations can contribute to increased comparability between future study findings.
Mounting evidence indicates that pain, depression, and anxiety impact orthopedic surgery outcomes (Lewis, Rice, McNair, & Kluger, 2015; Noiseux et al., 2014). Consistent measurement of these variables in the perioperative period is vital. PROMIS measures have been used in multiple studies about surgical care including orthopedic surgeries (Jones & Stukenborg, 2016). However, PROMIS short-form measures have not been validated in orthopedic patients having undergone surgeries that pose a risk for persistent postsurgical pain.
The current study aims to validate the reliability (internal consistency) and validity of three PROMIS instruments in Veterans 3 months following orthopedic surgery. Specifically, pain interference, anxiety, and depression were assessed in relation to well established self-report surveys.
Methods
Study Design.
Data for the current study were collected as part of a pilot randomized control trial assessing the efficacy of a one day Acceptance and Commitment Therapy workshop on persistent post-surgical pain in at-risk orthopedic patients (Dindo et al., 2018). This study was approved by the local institutional review board. Participants completed the measures for this study three months after surgery with a blinded assessor.
Participants
Participants were United States military Veterans having undergone orthopedic surgeries requiring at least 6 weeks of rehabilitation. Participants were recruited from a single site orthopedic clinic and compensated for their time. Inclusion in the parent study required elevated preoperative risk for persistent post-surgical pain. Preoperative risk criteria were defined based on findings that preoperative pain, depression, and anxiety infer risk for persistent post-surgical pain (Lewis et al., 2015; Noiseux et al., 2014). The specific criteria for inclusion in the parent study were: severe preoperative pain (> 7 on a 0–10 scale) or greater than mild pain (> 3 on a 0–10 scale) with elevated anxiety or depression (assessed with the Hamilton Anxiety and Depression Scales: (Hamilton, 1959, 1960). (See the parent study for full CONSORT diagram and exclusion criteria: (Dindo, 2018).
PROMIS Instruments
Anxiety.
Anxiety was assessed with the PROMIS Anxiety Short Form 8a. This scale assesses self-reported fear, anxious misery, hyperarousal, and somatic symptoms related to arousal across 8 items. Each item is rated on a scale of 1–5 (Never, Rarely, Sometimes, Often, Always). Items were then summed and raw scores were converted into T-scores (NIH, 2018). Validity and reliability of this measure have been demonstrated in patients with knee osteoarthritis (Driban, Morgan, Price, Cook, & Wang, 2015), multiple sclerosis (Marrie et al., 2018), and lumbar degenerative disease (Purvis, Neuman, Riley, & Skolasky, 2018).
Depression.
Depression was assessed with the PROMIS Depression Short Form 8b. This scale assesses negative mood, views of self, social cognitions, and decreased positive affect or engagement. This 8-item scale assesses depressive symptoms across the past 7 days. Each item is rated on a scale of 1–5 (Never, Rarely, Sometimes, Often, Always). Items were then summed and raw scores were converted into T-scores (NIH, 2018). Validity and reliability of this measure have been demonstrated in patients with knee osteoarthritis (Driban et al., 2015), multiple sclerosis (Marrie et al., 2018), and lumbar degenerative disease (Purvis et al., 2018)
Pain Interference
Pain interference was assessed with the PROMIS Pain Interference Short Form 8a. This instrument assesses the self-reported impact of pain on daily activities, social, emotional, physical, and recreational activities. This measure assesses pain interference over the past seven days. Each of 8 items is rated on a scale of 1–5 (Not at all, A little bit, Somewhat, Quite a bit, Very much), for a maximum total of 40. Items were then summed and raw scores converted into T-scores (NIH, 2018). The PROMIS pain interference short-form has demonstrated concurrent validity and comparative responsiveness in patients undergoing cervical and lumbar spine surgeries (Sharma, Ugiliweneza, Beswick, & Boakye, 2018) .
Comparison Instruments
Anxiety
Anxiety was assessed with the Generalized Anxiety Disorder 7-Item (GAD-7) scale. This self-report measure screens for generalized anxiety. Each item is scored on a 0–3 (Not at all, Several Days, More than half the days, Nearly every day), with a maximum total of 21 (Spitzer, Kroenke, Williams, & Lowe, 2006). A large, multisite study including 2,740 adult patients in primary care settings demonstrated that this measure has good reliability (Cronbach α = .92 and test-re-test intraclass correlation = 0.83) and both criterion and construct validity (demonstrating expected relationships with multiple other measures including a correlation of 0.75 on a mental health scale) (Spitzer et al., 2006). This measure has been recommended to assess anxiety following surgery (in coronary artery bypass patients: (Tully & Baker, 2012)).
Depression
Depression was assessed with the Patient Health Questionnaire 9-Item (PHQ-9) scale. Scores on this self-report measure range from 0–27; each of 9 items is scored as 0–3 (Not at all, Several Days, More than half the days, Nearly every day) (Kroenke, Spitzer, & Williams, 2001). A validity study including 6,000 medical patients demonstrated that this is a reliable (Cronbach’s α of 0.89) and valid measure of depression severity, demonstrating significant associations to multiple self-report measures, sick days, and healthcare utilization (Kroenke et al., 2001). This measure has also been validated in chronic spinal pain patients compared to a semi-structured diagnostic interview (Choi, Mayer, Williams, & Gatchel, 2014).
Pain Interference
Pain interference was assessed by the pain interference sub-scale of the Brief Pain Inventory (BPI). This seven item sub-scale assesses pain interference in the domains of general activities, mood, physical activities, relationships, and sleep (Cleeland, 1991). Each of the 7 items is scored on a 0–10 (0 = does not interfere, 10 = completely interferes) and then the average score is taken, for total scores ranging from 0–10. This widely used measure has demonstrated good reliability and validity (Cleeland, 1991). In surgical cardiac patients, this measure demonstrated good internal consistency both prior to surgery (Cronbach’s α = 0.91) and six months following surgery (Cronbach’s α = 0.94) and demonstrated responsiveness between pre- and post-surgical time points (Gjeilo, Stenseth, Wahba, Lydersen, & Klepstad, 2007).
Statistical Analyses.
Descriptive statistics were calculated using frequencies for categorical variables and mean (standard deviation) or median (interquartile range) for continuous variables. Individuals with missing data on a measure were excluded from analyses with that measure. Internal consistency, or the degree to which items are interrelated, of the PROMIS measures was assessed using Cronbach’s alpha and item-total correlations. One aspect of construct validity, or the degree to which an instrument measures what it claims to be measuring, was examined with factor analyses to determine whether the PROMIS measures were each assessing a single construct. Factor analyses allow for the identification of the number of underlying constructs within a given measure. Because the surveys involved frequency scale responses, polychoric correlations were used to maximize the robustness and accuracy of the factor analysis findings. To further assess construct validity, the PROMIS measures were correlated (Pearson’s) to the comparison measures to assess for convergent validity (the degree to which two tests reportedly measuring the same construct are related). The depression and anxiety PROMIS measures were then cross-compared (PROMIS depression to GAD-7 anxiety and vice-versa) in Pearson’s correlation analyses to assess for discriminant validity. The correlation between the PROMIS depression and anxiety measures were also calculated.
Results
Sample Description.
This study included 67 participants who had completed self-report questionnaires three months following orthopedic surgery (table 1). The sample was 93% male and the mean age was 63.7 years (SD: 8.9). The majority of the sample were married (57%), Caucasian (85%), had undergone total knee arthroplasty (67%), were not currently receiving opioids (69%), and had an income of $40,000 per year or less (56%). The sample reported relatively low levels of depression, anxiety, and pain interference. For anxiety, the mean PROMIS score was 49.83 (scores above 60 are more than one SD above the mean on the PROMIS measures), for depression it was 49.30, and for pain interference it was 58.34. Both of the index anxiety and depression measures demonstrated means (GAD-7 mean: 3.61, PHQ-9 mean: 6.15) below the recommended cut-off score of 10 indicating at least moderate symptoms (Kroenke et al., 2001; Spitzer et al., 2006). Similarly, the mean score on the index pain interference measure (BPI-PI) was only 3.18 on a 0–10 scale (with higher scores indicating greater interference).
Table 1.
Demographics and Descriptive Statistics (N = 67).
| Variable | Mean (SD) or Count (%) |
|---|---|
| Sex (Male) | 62 (93%) |
| Age | 63.7 (8.9) |
| Married (n = 65) | 37 (57%) |
| Racial background (Caucasian) | 57 (85%) |
| Receiving opioids | 21 (31%) |
| Surgery type | |
| Back or Neck | 4 (6%) |
| Total Hip Arthroplasty | 6 (9%) |
| Total Knee Arthroplasty | 45 (67%) |
| Shoulder | 12 (18%) |
| Education (n = 65) | |
| Some high school/high school graduate | 22 (34%) |
| Some college | 22 (34%) |
| College graduate/Post graduate education | 21 (32%) |
| Income (n = 64) | |
| < 20,000/year | 14 (22%) |
| $20,000 – $40,000/year | 22 (34%) |
| $40,000 – $60,000/year | 14 (22%) |
| > $60,000/year | 14 (22%) |
| PROMIS Measures | |
| Anxiety | 49.83 (8.67) |
| Depression | 49.30 (8.77) |
| Pain Interference | 58.34 (7.70) |
| Comparison Measures | |
| Anxiety (GAD-7) | 3.61 (4.47) |
| Depression (PHQ-9) | 6.15 (5.02) |
| Pain Interference (BPI: n = 65) | 3.18 (2.60) |
Internal Consistency
All PROMIS measures demonstrated good internal consistency. PROMIS anxiety Cronbach’s α was 0.93. Only deletion of item 1 would increase the Cronbach’s α to 0.94. Individual item-to-total correlations ranged from 0.54 – 0.87, with the average being 0.76 (see Table 2). The inter-item correlations ranged from 0.35 – 0.85. PROMIS depression Cronbach’s α was 0.95 and there was no single item that if deleted would improve the overall Cronbach’s α for the scale. Individual item-to-total correlations ranged from 0.74 – 0.88, with the average being 0.81 (see Table 2). The inter-item correlations ranged from 0.52 – 0.83. PROMIS pain interference Cronbach’s α was 0.96 and there was no single item that if deleted would improve the overall Cronbach’s α for the scale. Individual item-to-total correlations ranged from 0.81 – 0.88, with the average being 0.85 (see Table 2). The inter-item correlations ranged from 0.65 – 0.82.
Table 2.
PROMIS Measure Internal Consistency.
| Anxiety | Depression | Pain Interference |
||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| M | SD | Item-to-total correlation coefficient (r) | Cornbach’s α if item deleted | M | SD | Item-to-total correlation coefficient (r) | Cornbach’s α if item deleted | M | SD | Item-to-total correlation coefficient (r) | Cornbach’s α if item deleted | |
| Item 1 | 1.70 | 0.80 | 0.54 | 0.94 | 1.52 | 0.84 | 0.79 | 0.94 | 2.79 | 1.07 | 0.86 | 0.96 |
| Item 2 | 1.67 | 0.84 | 0.82 | 0.91 | 1.45 | 0.80 | 0.81 | 0.94 | 2.75 | 1.09 | 0.84 | 0.96 |
| Item 3 | 1.47 | 0.66 | 0.69 | 0.92 | 1.66 | 0.88 | 0.84 | 0.94 | 2.45 | 1.21 | 0.85 | 0.96 |
| Item 4 | 1.90 | 0.82 | 0.87 | 0.91 | 2.09 | 0.87 | 0.75 | 0.95 | 2.52 | 1.13 | 0.87 | 0.95 |
| Item 5 | 1.94 | 0.80 | 0.86 | 0.91 | 1.60 | 0.87 | 0.87 | 0.94 | 2.96 | 1.30 | 0.88 | 0.95 |
| Item 6 | 1.57 | 0.84 | 0.73 | 0.92 | 1.94 | 0.97 | 0.88 | 0.94 | 2.57 | 1.17 | 0.83 | 0.96 |
| Item 7 | 1.90 | 0.89 | 0.78 | 0.92 | 2.04 | 0.91 | 0.74 | 0.95 | 2.64 | 1.15 | 0.86 | 0.96 |
| Item 8 | 1.97 | 0.90 | 0.77 | 0.92 | 1.51 | 0.80 | 0.82 | 0.95 | 2.28 | 1.11 | 0.81 | 0.96 |
| Overall Scale | 0.76 | 0.93 | 0.81 | 0.95 | 0.85 | 0.96 | ||||||
Construct Validity
For the PROMIS anxiety measure, a one-factor solution was indicated (eigenvalue = 6.29), accounting for 79% of total variance; individual item loadings ranged from 0.70 – 0.96 . For the PROMIS depression measure, a one-factor solution was indicated (eigenvalue = 6.80), accounting for 85% of total variance; individual item loadings ranged from 0.85 – 0.96. For the PROMIS pain interference measure, a one-factor solution was also indicated (eigenvalue = 6.83), accounting for 85% of total variance; individual item loadings ranged from 0.90 – 0.95. See Table 3 for Eigenvalues of differing factor solution per measure.
Table 3.
PROMIS Measure Factor Analysis.
| Anxiety | Depression | Pain Interference |
||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Eigenvalue | Difference | Proportion | Cumulative | Eigenvalue | Difference | Proportion | Cumulative | Eigenvalue | Difference | Proportion | Cumulative | |
| 1 | 6.29 | 5.63 | 0.79 | 0.79 | 6.80 | 6.17 | 0.85 | 0.85 | 6.83 | 6.49 | 0.85 | 0.85 |
| 2 | 0.66 | 0.34 | 0.83 | 0.87 | 0.63 | 0.43 | 0.08 | 0.93 | 0.34 | 0.10 | 0.04 | 0.90 |
| 3 | 0.32 | 0.02 | 0.04 | 0.91 | 0.21 | 0.09 | 0.03 | 0.96 | 0.24 | 0.06 | 0.03 | 0.93 |
| 4 | 0.30 | 0.12 | 0.04 | 0.95 | 0.12 | 0.02 | 0.01 | 0.97 | 0.18 | 0.04 | 0.02 | 0.95 |
| 5 | 0.18 | 0.06 | 0.02 | 0.97 | 0.10 | 0.02 | 0.01 | 0.98 | 0.15 | 0.03 | 0.02 | 0.97 |
| 6 | 0.12 | 0.03 | 0.02 | 0.99 | 0.08 | 0.03 | 0.01 | 0.99 | 0.12 | 0.02 | 0.01 | 0.98 |
| 7 | 0.09 | 0.07 | 0.01 | 1.00 | 0.05 | 0.04 | 0.01 | 1.00 | 0.10 | 0.05 | 0.01 | 0.99 |
| 8 | 0.02 | 0.00 | 1.00 | 0.02 | 0.00 | 1.00 | 0.05 | 0.01 | 1.00 | |||
Concurrent Validity
The PROMIS anxiety measure correlated moderately with the GAD-7 (r = .69, p < .01, 95% CI: 0.53 – 0.79) and the PROMIS depression measure correlated moderately with the PHQ-9 (r = .69, p < .01, 95% CI: 0.54 – 0.80). The PROMIS pain interference measure correlated strongly with the BPI pain interference subscale (r = .76, p < .01, 95% CI: 0.64 – 0.85).
Discriminant Validity
The PROMIS anxiety measure correlated moderately with the PHQ-9 (r = .63, p < .01, 95% CI: 0.46 – 0.76) and the BPI pain interference subscale (r = .57, p < .01, 95% CI: 0.38 – 0.71). The PROMIS depression measure correlated moderately with the GAD-7 (r = .69, p < .01, 95% CI: 0.53 – 0.80) and the BPI pain interference subscale (r = .60, p < .01, 95% CI: 0.42 – 0.73).. The PROMIS depression and PROMIS anxiety measures were highly correlated (r = .91, p < .01, 95% CI: 0.85 – 0.94).
Discussion
This study contributes to the ongoing body of work validating the PROMIS measures (NIH, 2018) as part of the broader effort to standardize measures of pain and distress across various healthcare populations. Specifically, this study examined measures of anxiety, depression, and pain interference in Veterans three months following orthopedic surgery. These results demonstrated that the three PROMIS short forms (PROMIS Anxiety Short Form 8a, PROMIS Depression Short Form 8b, and PROMIS Pain Interference Short Form 8a) had good homogeneity in a postoperative Veteran population. Factor analyses confirmed single structure models for all three measures and concurrent validity was adequate. Given the low levels of depression and anxiety in this sample, and the low levels of variability in the PROMIS measures in this sample, concurrent validity was good. Discriminant validity for the PROMIS depression and anxiety measures indicated significant overlap.
The relatively strong concurrent validity between the PROMIS pain interference measure and the BPI – pain interference subscale (our index measure) is consistent with previous work. A previous study using the computer adapted version of the PROMIS pain interference measure, confirmed construct validity by distinguishing between a sample of patients with pain (related to knee osteoarthritis: OA) and a general population sample (Broderick, Schneider, Junghaenel, Schwartz, & Stone, 2013). In that study, participants with OA scored above the 80th percentile on average on the PROMIS pain interference measure (Broderick et al., 2013).
The correlation between the PROMIS depression measure and the PHQ-9 was not as high as expected. However, this moderate correlation is consistent with previous work in a symptomatic knee OA sample where the same PROMIS depression short-form correlated with the Beck Depression Index – II (Beck & Steer, 1996) at only r = 0.70 (Driban et al., 2015). The correlation between the PROMIS depression measure and the PHQ-9 was higher (r = 0.79) in preoperative patients with lumbar degenerative disease undergoing decompression (Purvis et al., 2018). The moderate correlations may reflect subtle differences in the ways that depression is measured by the PROMIS measure compared to the PHQ-9. The PHQ-9 (Kroenke et al., 2001) has more items indicating physical functioning and cognitive-related symptoms (i.e., sleep, energy, appetite, concentration, psychomotor retardation), compared to the PROMIS depression short form measure. Previous work has also compared the diagnostic utility of both the PHQ-9 and PROMIS depression scale as compared to a structured diagnostic clinical interview (Structured Clinical Interview for DSM-IV-TR Axis I Diagnoses: SCID) and demonstrated similar accuracy between the measures, but greater sensitivity of the PHQ-9 and greater specificity of the PROMIS depression measure in a sample of patients with multiple sclerosis (Marrie et al., 2018). Taken together, these findings suggest that the PROMIS depression short-form may function slightly differently than the PHQ-9.
The correlation between the PROMIS anxiety measure and the GAD-7 was also moderate in the current study. These measures have also demonstrated moderate correlations in a sample of pre-operative patients with lumbar degenerative disease (r = 0.71) (Purvis et al., 2018) . The current finding is also consistent with a study in a sample of patients with knee OA, where a different PROMIS anxiety short-form (7a) demonstrated a moderate correlation (r = 0.71) with the Perceived Stress Scale (Driban et al., 2015). In a comparison of the diagnostic utility of both the PROMIS anxiety scale and the GAD-7 compared to a structured diagnostic interview (SCID) for anxiety disorders in patients with multiple sclerosis, the GAD-7 and PROMIS anxiety scale demonstrated similar accuracy; the area under the curve for both was 0.85 (Marrie et al., 2018).
Depression and anxiety measures are often highly correlated, arguably measuring in part an underlying construct of distress (Driban et al., 2015; Purvis et al., 2018; Xie et al., 2012). This could also reflect the shared variance of negative affect between anxiety and depression (Griffith et al., 2010). This overlap was seen in the current study, where the PROMIS measures did not demonstrate strong discriminant validity between anxiety and depression. Specifically, the PROMIS depression measure had the same correlations between the index depression and anxiety measures and the PROMIS anxiety measure demonstrated a slightly higher correlation to the index anxiety measure than the index depression measure. Further, the PROMIS depression and anxiety measures were highly correlated to one another, suggesting that they are both measuring an underlying construct of distress in the current sample.
This study had some limitations. The relatively small sample size precluded doing more extensive modeling to explore the overlap of anxiety and depressive symptoms. Further, the generalizability of results might be impacted by the fact that the participants in this study agreed to participate in the parent study (Dindo, 2018).
In summary, this study uniquely demonstrated preliminary validity of three PROMIS measures (anxiety, depression and pain interference) with 67 Veterans three months following orthopedic surgery. However, while the PROMIS depression and anxiety measures demonstrated excellent internal validity, they appeared to be largely measuring the same construct. Replication of these findings is needed in a larger sample. Further validation (including predictive validity) of these PROMIS short form measures for a surgical orthopedic Veteran population is also needed. Overall, the current findings suggest potential utility of these PROMIS measures with postoperative orthopedic patients, but suggest a high degree of overlap between the PROMIS depression and anxiety measures.
Acknowledgments
This work was supported by the National Institutes of Health NCCIH 5R34AT008349–02 to PI Barbara Rakel and the National Institute of Neurological Disorders and Stroke (T32 NS045549). Data were collected with REDCap electronic data capture tools hosted at University of Iowa (supported by NIH 54TR001013).
Footnotes
The authors have no conflicts of interest.
References
- Beck AT, & Steer RA (1996). Manual for the Beck Depression Inventory – Second Edition Manual. Texas: The Psychological Corporation. [Google Scholar]
- Broderick JE, Schneider S, Junghaenel DU, Schwartz JE, & Stone AA (2013). Validity and reliability of patient-reported outcomes measurement information system instruments in osteoarthritis. Arthritis Care Res (Hoboken), 65(10), 1625–1633. doi: 10.1002/acr.22025 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Choi Y, Mayer TG, Williams MJ, & Gatchel RJ (2014). What is the best screening test for depression in chronic spinal pain patients? Spine J, 14(7), 1175–1182. doi: 10.1016/j.spinee.2013.10.037 [DOI] [PubMed] [Google Scholar]
- Cleeland CS (1991). The Brief Pain Inventory User Guide. [Google Scholar]
- Dindo L (2018). Acceptance and Commitment Therapy for Prevention of Chronic Post-surgical Pain and Opioid Use in At-Risk Veterans: A Pilot Randomized Controlled Study. doi: 10.1016/j.jpain.2018.04.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dindo L, Zimmerman MB, Hadlandsmyth K, StMarie B, Embree J, Marchman J, … Rakel B (2018). Acceptance and Commitment Therapy for Prevention of Chronic Post-surgical Pain and Opioid Use in At-Risk Veterans: A Pilot Randomized Controlled Study. J Pain. doi: 10.1016/j.jpain.2018.04.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Driban JB, Morgan N, Price LL, Cook KF, & Wang C (2015). Patient-Reported Outcomes Measurement Information System (PROMIS) instruments among individuals with symptomatic knee osteoarthritis: a cross-sectional study of floor/ceiling effects and construct validity. BMC Musculoskelet Disord, 16, 253. doi: 10.1186/s12891-015-0715-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gjeilo KH, Stenseth R, Wahba A, Lydersen S, & Klepstad P (2007). Validation of the brief pain inventory in patients six months after cardiac surgery. J Pain Symptom Manage, 34(6), 648–656. doi: 10.1016/j.jpainsymman.2007.01.010 [DOI] [PubMed] [Google Scholar]
- Griffith JW, Zinbarg RE, Craske MG, Mineka S, Rose RD, Waters AM, & Sutton JM (2010). Neuroticism as a common dimension in the internalizing disorders. Psychol Med, 40(7), 1125–1136. doi: 10.1017/S0033291709991449 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hamilton M (1959). The assessment of anxiety states by rating. Br J Med Psychol, 32(1), 50–55. [DOI] [PubMed] [Google Scholar]
- Hamilton M (1960). A rating scale for depression. J Neurol Neurosurg Psychiatry, 23, 56–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Haskell SG, Heapy A, Reid MC, Papas RK, & Kerns RD (2006). The prevalence and age-related characteristics of pain in a sample of women veterans receiving primary care. J Womens Health (Larchmt), 15(7), 862–869. doi: 10.1089/jwh.2006.15.862 [DOI] [PubMed] [Google Scholar]
- Jones RS, & Stukenborg GJ (2016). Patient-Reported Outcomes Measurement Information System (PROMIS) Use in Surgical Care: A Scoping Study. J Am Coll Surg. doi: 10.1016/j.jamcollsurg.2016.11.015 [DOI] [PubMed] [Google Scholar]
- Kroenke K, Spitzer RL, & Williams JB (2001). The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med, 16(9), 606–613. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lewis GN, Rice DA, McNair PJ, & Kluger M (2015). Predictors of persistent pain after total knee arthroplasty: a systematic review and meta-analysis. Br J Anaesth, 114(4), 551–561. doi: 10.1093/bja/aeu441 [DOI] [PubMed] [Google Scholar]
- Marrie RA, Zhang L, Lix LM, Graff LA, Walker JR, Fisk JD, … Bernstein CN (2018). The validity and reliability of screening measures for depression and anxiety disorders in multiple sclerosis. Mult Scler Relat Disord, 20, 9–15. doi: 10.1016/j.msard.2017.12.007 [DOI] [PubMed] [Google Scholar]
- NIH. (2018). Patient-Reported Outcomes Measurement Information System (PROMIS) Available at: http://www.nihpromis.org/.
- Noiseux NO, Callaghan JJ, Clark CR, Zimmerman MB, Sluka KA, & Rakel BA (2014). Preoperative predictors of pain following total knee arthroplasty. J Arthroplasty, 29(7), 1383–1387. doi: 10.1016/j.arth.2014.01.034 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Pickett T, Rothman D, Crawford EF, Brancu M, Fairbank JA, & Kudler HS (2015). Mental Health Among Military Personnel and Veterans. N C Med J, 76(5), 299–306. [DOI] [PubMed] [Google Scholar]
- Purvis TE, Neuman BJ, Riley LH 3rd, & Skolasky RL (2018). Discriminant Ability, Concurrent Validity, and Responsiveness of PROMIS Health Domains Among Patients With Lumbar Degenerative Disease Undergoing Decompression With or Without Arthrodesis. Spine (Phila Pa 1976). doi: 10.1097/BRS.0000000000002661 [DOI] [PubMed] [Google Scholar]
- Sharma M, Ugiliweneza B, Beswick J, & Boakye M (2018). Concurrent Validity and Comparative Responsiveness of PROMIS-SF Versus Legacy Measures in the Cervical and Lumbar Spine Population: Longitudinal Analysis from Baseline to Postsurgery. World Neurosurg. doi: 10.1016/j.wneu.2018.04.131 [DOI] [PubMed] [Google Scholar]
- Spitzer RL, Kroenke K, Williams JB, & Lowe B (2006). A brief measure for assessing generalized anxiety disorder: the GAD-7. Arch Intern Med, 166(10), 1092–1097. doi: 10.1001/archinte.166.10.1092 [DOI] [PubMed] [Google Scholar]
- Tully PJ, & Baker RA (2012). Depression, anxiety, and cardiac morbidity outcomes after coronary artery bypass surgery: a contemporary and practical review. J Geriatr Cardiol, 9(2), 197–208. doi: 10.3724/SP.J.1263.2011.12221 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Xie J, Bi Q, Li W, Shang W, Yan M, Yang Y,… Zhang H (2012). Positive and negative relationship between anxiety and depression of patients in pain: a bifactor model analysis. PLoS One, 7(10), e47577. doi: 10.1371/journal.pone.0047577 [DOI] [PMC free article] [PubMed] [Google Scholar]