Abstract
Background
Unrecognized posttraumatic stress disorder (PTSD) is common and may be an important factor in treatment-resistant depression. Brief screens for PTSD have not been evaluated for patients with depression.
Objective
The objective was to evaluate a 4-item screen for PTSD in patients with depression.
Design
Baseline data from a depression study were used to evaluate sensitivity, specificity, and likelihood ratios (LRs) using the PTSD checklist (PCL-17) as the reference standard.
Subjects
Subjects are 398 depressed patients seen in Veterans Affairs (VA) primary care clinics.
Measures
The Patient Health Questionnaire (PHQ) for depression, PCL-17, and 4-item screen for PTSD were used.
Results
Patients had a mean PHQ score of 14.8 (SD 3.7). Using a conservative PCL-17 cut point “(>50)”, the prevalence of PTSD was 37%. PCL-17 scores were strongly associated with PHQ scores (r = 0.59, P < 0.001). Among the 342 (86%) patients endorsing trauma, a score of 0 on the remaining 3 symptom items had a LR = 0.21, score of 1 a LR = .62, score of 2 a LR = 1.36, and score of 3 a LR = 4.38.
Conclusions
Most depressed VA primary care patients report a history of trauma, and one third may have comorbid PTSD. Our 4-item screen has useful LRs for scores of 0 and 3. Modifying item rating options may improve screening characteristics.
KEY WORDS: posttraumatic stress disorder, depression, screening, primary care, veterans
INTRODUCTION
Posttraumatic stress disorder (PTSD) is common in primary care settings and often under-recognized1,2. PTSD is especially common in veteran populations3,4; for individuals who faced combat, estimates of lifetime prevalence of PTSD are as high as 20 to 30%5. Recent studies suggest that PTSD is commonly comorbid with depression and that patients with both disorders have a poorer prognosis, poorer response to depression treatment, and greater health care utilization2,6,7.
Several brief screens for PTSD have recently been developed for use in primary care8–12. However, the 4-item Primary Care PTSD Screen (PC-PTSD) has had conflicting results regarding its test characteristics9,13,14, and Breslau’s 7-item PTSD screen12 was evaluated using a sample of 134 patients11 and needs confirmation with a larger sample. Furthermore, test characteristics of these and other screens have not been established for patients with depression.
Our objectives were to assess the prevalence of PTSD in patients with depression symptoms in VA primary care clinics and to evaluate the sensitivity, specificity, and likelihood ratios (LR) of a 4-item screen using the PTSD checklist (PCL-17)15 as the reference standard.
METHODS
Setting and Sample Between January 2002 and September 2003, we recruited clinicians and patients from 5 primary care clinics at 1 VA medical center for a randomized controlled trial of a collaborative intervention for depression in primary care (DEP-PC)16. DEP-PC was approved by the medical center’s Institutional Review Board, and participating clinicians and patients gave informed consent.Potential participants for DEP-PC were identified using lists of patients due to see their primary care clinicians within a month and whose primary care clinicians (n = 41) were participating in DEP-PC. We excluded patients who received treatment from a mental health clinician within the prior 6 months or had Alzheimer’s disease, cognitive problems, psychotic symptoms, bipolar disorder, or terminal illness noted in their medical records. We did not exclude patients with dysthymia, substance abuse, or anxiety disorders. Patients not excluded at this stage (n = 5,434) were mailed study introduction letters, and 3,103 (57%) completed subsequent phone-screening including the Patient Health Questionnaire for depression (PHQ)17, our 4-item PTSD screen, and the Short Blessed test for dementia18. Patients scoring ≥10 on the Short Blessed were excluded. Patients with PHQ scores 10–25, suggesting moderate or greater depression severity19, were invited to an enrollment interview. Patients with PHQ >25 or active dangerous ideation were referred for urgent care and excluded. Of 560 eligible patients, 398 (71%) completed enrollment interviews. A trained interviewer administered the PCL-17 and repeat PHQ 2 to 5 weeks after phone screening. Baseline data from these 398 patients were used for the current study.
Measures The 9-item PHQ can be used to diagnose depression and measure depression severity in primary care settings17,19. Cut points of 10 and 15 have sensitivities of 88 and 68% and specificities of 88 and 95% for depression diagnosis, respectively, when compared to clinician-administered interview19.The 17-item PTSD checklist (PCL-17)15 has been well studied and often used for PTSD screening. Respondents rate each item for severity over the past month using a 5-point Likert scale ranging from 1 = not at all to 5 = extremely. We scored the PCL-17 by summing responses to the 17 items. Higher scores indicate greater symptom severity. Cut points between 30 and 50 have been evaluated in prior research to define PTSD20,21. For Vietnam veterans, a cut point of ≥50 has been recommended22.In 1999, 1 of the authors (SKD) helped construct a 4-item PTSD screen for use in routine screening in primary care (Table 1). The screen was derived from the DSM-IV trauma definition and main symptom categories for PTSD diagnosis23. When patients endorse the trauma stem-item, the remaining 3 items (re-experiencing, avoidance, and hyper-arousal) are administered. Patient responses are dichotomous (No = 0, Yes = 1). Total screen scores are calculated for patients who endorse trauma (stem item) by summing responses across the 3 symptom items. Screen scores range from 0 to 3.
Table 1.
Four-item PTSD screen
| Items | |
|---|---|
| Stem item | Have you witnessed or experienced an event that involved threatened or actual serious injury or death? |
| Symptom items | Within the past month, has the memory of this event troubled you to the point that it interferes with your sleep, concentration or relationships? |
| Within the past month, have you felt distant or cut off from other people? | |
| Within the past month, have you been ‘super alert’; or watchful or on guard? |
All items are scored no = 0 or yes = 1. The symptom items are administered only if the trauma stem item is endorsed.
Analysis We used means, standard deviations, and percentages to describe patient characteristics. Prevalence rates for PTSD were calculated using PCL-17 cut points of 30, 44, and 50. We calculated sensitivity, specificity, LRs and 95% confidence intervals for the 4-item PTSD screen using a conservative cut point for the PCL-17 (≥50) as the reference standard for diagnosis of PTSD22. Sensitivity is the true positives rate [true positives/(true positives + false negatives)]. Specificity is the true negative rate [true negatives/(true negatives + false positives)]. The LR is the likelihood that a given score occurs in patients with PTSD compared to those without PTSD24.Because 2 items from the screen are similar to those in the PCL-17, we also calculated sensitivity, specificity, and LRs using a 15-item PCL (removing the 2 overlapping items) to determine to what extent the 2-item overlap affected test characteristics. SPSS v11.5 and v13.0 were used for the analyses.
RESULTS
Characteristics of the 398 patients are presented in Table 2. The demographic characteristics resemble local and national VA patient populations25. Patients were mostly male and had a mean age of 56.9 (11.3) years. Fifty-three percent were Vietnam veterans. Patients had a mean PHQ score of 14.8 (SD 3.7), indicating moderate depression severity,19 and mean PCL-17 score of 45.2 (12.8).
Table 2.
Characteristics of the sample (N = 398)
| Characteristic | n (%)* |
|---|---|
| Demographic | |
| Mean age, years (SD) | 56.9 (11.3) |
| Female | 28 (7.0) |
| Married† | 202 (52) |
| Vietnam veteran | 210 (53) |
| Mean PHQ score (SD) | 14.8 (3.7) |
| Mean PCL-17 score (SD) | 45.2 (12.8) |
| PCL-17 score ≥30 | 350 (88) |
| PCL-17 score ≥44 | 207 (52) |
| PCL-17 score ≥50 | 147 (37) |
| Reported trauma (positive stem) | 342 (86) |
| Positive PTSD screen (trauma plus 1 or more symptoms) | 279 (70) |
| Positive PTSD screen (trauma plus 2 or more symptoms) | 161 (40) |
Potential range for the PHQ scores is 0–27, the PCL-17 is 17–85, and the 4-item PTSD screen is 0–3.
PHQ Patient Health Questionnaire, PTSD posttraumatic stress disorder, PCL-17 PTSD checklist
*Number and percentage except when noted that the mean and SD are used
†Marital status data were missing for 6 (1.5%) patients.
Using a conservative PCL-17 score of ≥50 as the cut point for diagnosis, 147 (37%) patients had PTSD. PTSD symptom severity was strongly associated with depression severity based on PHQ scores (r = 0.59, P < .001). PTSD was observed in 24% of patients with PHQ scores 10–14 (moderate depression), 56% of patients with scores 15–19 (moderately severe depression), and 83% of patients with scores 20–25 (severe depression).
Using a PCL-17 cut point score ≥50 as the reference standard, the trauma stem item alone had a sensitivity of 87% and specificity of 15% (trauma stem item in Table 3). Among the 342 patients who endorsed the trauma item, LRs for the PTSD screen were 4.38, 1.36, 0.62, and 0.21 for scores of 3, 2, 1 and 0, respectively, based on responses to the 3 symptom items. Many depressed patients (86%) reported a history of trauma. If a cut point of trauma plus 1 or more symptoms were used to define a positive screen (sensitivity 95%, specificity 20%), 279 (70%) depressed patients would have a positive screen (Table 2). If a cut point of trauma plus 2 or more symptoms were used (sensitivity 70%, specificity 66%), 161 (40%) depressed patients would have a positive screen.
Table 3.
Likelihood ratios for the 4-item PTSD screen (N = 398)
| Trauma item result/screen result | PTSD (PCL-17 ≥ 50) | Likelihood ratio (95 %CI) | |
|---|---|---|---|
| Present n (%) | Absent n (%) | ||
| Trauma stem item | |||
| Trauma (positive) | 128 (87) | 214 (85) | 1.02 (0.94, 1.11) |
| No trauma (negative) | 19 (13) | 37 (15) | 0.88 (0.53, 1.47) |
| Totals | 147 (100) | 251 (100) | – |
| PTSD screen results for patients reporting trauma (n = 342) | |||
| 3 symptoms | 45 (35) | 18 (8) | 4.38 (2.53, 6.90) |
| 2 symptoms | 44 (34) | 54 (25) | 1.36 (0.98, 1.90) |
| 1 symptom | 32 (25) | 86 (40) | 0.62 (0.44, 0.87) |
| 0 (Trauma, but no symptoms) | 7 (5) | 56 (26) | 0.21 (0.10, 0.45) |
| Totals | 128 (100) | 214 (100) | – |
Sensitivity, specificity, and area under the curve (AUC) did not differ significantly using PCL-17 cut points of ≥50, ≥44, and ≥30 (Appendix 1). AUCs ranged from 0.73 to 0.78. In addition, these test characteristics did not differ substantially when the 15-item PCL was used as the reference standard (Appendix 2).
DISCUSSION
A majority of patients with depressive symptoms in a VA primary care setting report a history of trauma and over one-third are likely to have comorbid PTSD. Our 4-item PTSD screen has LRs that may be useful in deciding who should be targeted for further evaluation for PTSD. For example, if the prescreen probability of PTSD is approximately 40% in patients with depressive symptoms (estimated from prevalence of PTSD based on PCL-17 ≥50), individuals with PTSD screen scores of 0 (LR = 0.21) would have a post-screen probability of PTSD of approximately 12%. These individuals would not need further evaluation. Conversely, for individuals with PTSD screen scores of 3 (LR = 4.38), their post-screen probability of PTSD would be approximately 75%, indicating a need for further evaluation for PTSD.
Forty percent of patients in our sample had a positive 4-item screen using a cut point of trauma plus 2 or more symptom items. These results are consistent with findings from a recent study using the 4-item Primary Care PTSD Screen (PC-PTSD)9 where 36% of depressed primary care patients had PTSD6. These estimates are higher than estimates of PTSD in general VA primary care populations. Magruder et al.3 found that 11.5% of patients in VA primary care clinics had PTSD using diagnostic interviews with the Clinician Administered PTSD Scale (CAPS), and Spiro et al.7 found that 20% of VA primary care patients had PTSD using the PCL-17. The difference in these latter 2 estimates of PTSD may be caused by the method of assessment (CAPS vs PCL-17).
Although our 4-item screen is likely to be useful, further research is need. Variability in the test characteristics of screening measures and PTSD prevalence rates across patient populations points to the need for careful study of screening instruments before their widespread use. Test characteristics of brief screens are likely to vary by patient population, the reference standard used, use of a trauma stem item with skip-out approach, and symptom rating scale (e.g., yes/no vs 5-point Likert scale).
Our study had 3 important limitations. First, 2 items in our brief screen were similar to items in the PCL-17. However, our results did not differ substantially using the 15-item version of the PCL as the reference standard. Second, we did not use gold-standard clinician-administered measures to diagnose PTSD or depression. Use of such measures would help to clarify distinctions among the disorders and better define the study population and test characteristics of brief screens. Finally, our sample consisted of VA primary care patients with depression symptoms. The test characteristics of the 4-item screen may not generalize to non-depressed or non-veteran populations.
Screening for PTSD in primary care and veteran populations has become increasingly important. Indeed, recent reports indicate that rates of mental health problems including PTSD and depression are especially common among OEF/OIF veterans4,26,27. Our results suggest that patients with depression are at a high risk for comorbid PTSD and should be screened. Brief screens that can be administered rapidly and easily in primary care clinics, such as our 4-item PTSD screen, may facilitate this process. More research is needed to determine whether modifying item rating options improves test characteristics and to assess test characteristics in other patient samples using clinician-administered diagnostic measures as reference standards.
Acknowledgment
The authors wish to acknowledge Jeffrey Solodky BA and Nancy Cuilwik BS for assistance with reviewing the literature and for screening patients for the DEP-PC study. The research reported here was supported by the Department of Veterans Affairs, Veterans Health Administration, Health Services Research and Development Service Project MHI 20-020. The views expressed in this manuscript are those of the authors and do not necessarily represent the views of the Department of Veterans Affairs.
Conflict Of Interest Statement All 3 authors received grant support from the Department of Veterans Affairs Health Services Research and Development Service. Aside from this support, there are no potential conflicts of interest.
Appendix A
Table 4.
Sensitivity, specificity and area under the curve of 4-item screen at various PCL-17 cut points for patients who reported trauma (N = 342)
| Sensitivity | Specificity | AUC | |
|---|---|---|---|
| PCL-17 ≥ 30 as reference standard | 0.78 (SE = 0.03) | ||
| PTSD screen = 1 | 0.85 | 0.48 | |
| PTSD screen = 2 | 0.52 | 0.93 | |
| PTSD screen = 3 | 0.21 | 1.0 | |
| PCL-17 ≥ 44 as reference standard | 0.74 (SE = 0.03) | ||
| PTSD screen = 1 | 0.94 | 0.32 | |
| PTSD screen = 2 | 0.64 | 0.72 | |
| PTSD screen = 3 | 0.30 | 0.94 | |
| PCL-17 ≥ 50 as reference standard | 0.73 (SE = 0.03) | ||
| PTSD screen = 1 | 0.95 | 0.26 | |
| PTSD screen = 2 | 0.70 | 0.66 | |
| PTSD screen = 3 | 0.35 | 0.92 |
Appendix B
Table 5.
Sensitivity, specificity and area under the curve of 4-item screen at various PCL-15 cut points for patients who reported trauma (N = 342)
| Sensitivity | Specificity | AUC | |
|---|---|---|---|
| PCL-15 ≥ 30 as reference standard | 0.77 (SE = 0.03) | ||
| PTSD screen = 1 | 0.88 | 0.40 | |
| PTSD screen = 2 | 0.57 | 0.88 | |
| PTSD screen = 3 | 0.23 | 0.99 | |
| PCL-15 ≥ 44 as reference standard | 0.72 (SE = 0.03) | ||
| PTSD screen = 1 | 0.94 | 0.26 | |
| PTSD screen = 2 | 0.68 | 0.65 | |
| PTSD screen = 3 | 0.35 | 0.91 | |
| PCL-15 ≥ 50 as reference standard | 0.75 (SE = 0.03) | ||
| PTSD screen = 1 | 0.96 | 0.22 | |
| PTSD screen = 2 | 0.80 | 0.61 | |
| PTSD screen = 3 | 0.41 | 0.88 |
References
- 1.Lecrubier Y. Posttraumatic stress disorder in primary care: a hidden diagnosis. J Clin Psychiatry. 2004;65(Suppl 1):49–54. [PubMed]
- 2.Liebschutz J, Saitz R, Brower V, et al. PTSD in urban primary care: high prevalence and low physician recognition. J Gen Intern Med. 2007;22(6):719–26. [DOI] [PMC free article] [PubMed]
- 3.Magruder KM, Frueh BC, Knapp RG, et al. Prevalence of posttraumatic stress disorder in Veterans Affairs primary care clinics. Gen Hosp Psych. 2005;27(3):169–79. [DOI] [PubMed]
- 4.Seal KH, Bertenthal D, Miner CR, Sen S, Marmar C. Bringing the war back home: mental health disorders among 103,788 US veterans returning from Iraq and Afghanistan seen at Department of Veterans Affairs facilities. Arch Intern Med. 2007;167(5):476–82. [DOI] [PubMed]
- 5.Hoge CW, Castro CA, Messer SC, McGurk D, Cotting DI, Koffman RL. Combat duty in Iraq and Afghanistan, mental health problems, and barriers to care. N Engl J Med. 2004;351(1):13–22. [DOI] [PubMed]
- 6.Campbell DG, Felker BL, Liu CF, et al. Prevalence of depression-PTSD comorbidity: implications for clinical practice guidelines and primary care-based interventions. J Gen Intern Med. 2007;22(6):711–8. [DOI] [PMC free article] [PubMed]
- 7.Spiro A, 3rd, Hankin CS, Mansell D, Kazis LE. Posttraumatic stress disorder and health status: the veterans health study. J Ambul Care Manage. 2006;29(1):71–86. [DOI] [PubMed]
- 8.Meltzer-Brody S, Churchill E, Davidson JR. Derivation of the SPAN, a brief diagnostic screening test for post-traumatic stress disorder. Psychiatry Res. 1999;88(1):63–70. [DOI] [PubMed]
- 9.Prins A QP, Kimmerling R, Camerond RP, Hugelshofer DS, Shaw-Hegwer J, Thrailill A, Gusman FD, Sheikh JI. The primary care PTSD screen (PC-PTSD): development and operating characteristics. Prim Care Psychiatry. 2004;9(1):9–14. [DOI]
- 10.Kimerling R, Trafton JA, Nguyen B. Validation of a brief screen for post-traumatic stress disorder with substance use disorder patients. Addict Behav. 2006;31(11):2074–9. [DOI] [PubMed]
- 11.Kimerling R, Ouimette P, Prins A, et al. Brief report: utility of a short screening scale for DSM-IV PTSD in primary care. J Gen Intern Med. 2006;21(1):65–7. [DOI] [PMC free article] [PubMed]
- 12.Breslau N, Peterson EL, Kessler RC, Schultz LR. Short screening scale for DSM-IV posttraumatic stress disorder. Am J Psychiatry. 1999;156(6):908–11. [DOI] [PubMed]
- 13.Bliese PD WK, Adler AB, Thomas JL. Psychological screening validation with soldiers returning from combat. In: Roy MJ, ed. Novel Approaches to the Diagnosis and Treatment of Posttraumatic Stress Disorder. Amsterdam: IOS Press; 2006.
- 14.Greenburg DL, Roy MJ. In the shadow of Iraq: posttraumatic stress disorder in 2007. J Gen Intern Med. 2007;22(6):888–9. [DOI] [PMC free article] [PubMed]
- 15.Weathers F, Ford, J. Psychometric review of PTSD Checklist (PCL-C, PCL-S, PCL-M, PCL-PR). In: Stamm BH, ed. Measurement of Stress, Trauma, and Adaptation. Lutherville: Sidran Press; 1996.
- 16.Dobscha S, Corson K, Hickam D, Perrin N, Kraemer D, Gerrity M. Depression decision support in primary care: a cluster randomized trial. Ann Intern Med. 2006;145:477–87. [DOI] [PubMed]
- 17.Spitzer RL, Kroenke K, Williams JB. Validation and utility of a self-report version of PRIME-MD: the PHQ primary care study. Primary care evaluation of mental disorders. Patient Health Questionnaire. JAMA. 1999;282(18):1737–44. [DOI] [PubMed]
- 18.Katzman R, Brown T, Fuld P, Peck A, Schechter R, Schimmel H. Validation of a short orientation-memory-concentration test of cognitive impairment. Am J Psychiatry. 1983;140(6):734–9. [DOI] [PubMed]
- 19.Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med. 2001;16(9):606–13. [DOI] [PMC free article] [PubMed]
- 20.Blanchard EB, Jones-Alexander J, Buckley TC, Forneris CA. Psychometric properties of the PTSD checklist (PCL). Behav Res Ther. 1996;34(8):669–73. [DOI] [PubMed]
- 21.Dobie DJ, Kivlahan DR, Maynard C, et al. Screening for post-traumatic stress disorder in female Veteran’s Affairs patients: validation of the PTSD checklist. Gen Hosp Psych. 2002;24(6):367–74. [DOI] [PubMed]
- 22.Forbes D, Creamer M, Biddle D. The validity of the PTSD checklist as a measure of symptomatic change in combat-related PTSD. Behav Res Ther. 2001;39(8):977–86. [DOI] [PubMed]
- 23.American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, 4th Edition. Washington DC: American Psychiatric Association; 1994.
- 24.Kraemer H. Evaluating Medical Tests. Newbury Park, CA: Sage; 1992.
- 25.U.S. Department of Commerce. Economics and Statistics Administration, US Census Bureau. Census 2003 Brief. http://www1.va.gov/vetdata/page.cfm?pg=1. 2003: Accessed on 6/4/07.
- 26.Kang HK, Natelson BH, Mahan CM, Lee KY, Murphy FM. Posttraumatic stress disorder and chronic fatigue syndrome-like illness among Gulf War veterans: a population-based survey of 30,000 veterans. Am J Epidemiol. 2003;157(2):141–8. [DOI] [PubMed]
- 27.Hoge CW, Auchterlonie JL, Milliken CS. Mental health problems, use of mental health services, and attrition from military service after returning from deployment to Iraq or Afghanistan. JAMA. 2006;295(9):1023–32. [DOI] [PubMed]