Screening for birth-related PTSD: psychometric properties of the Turkish version of the Posttraumatic Diagnostic Scale in postpartum women in Turkey

Figures & data

Table 1. PDS item descriptions, Item mapping for the tested models, EFA factor loadings for the Model F, CFA standardized factor loadings and Squared multiple correlations for the Model F.

	Proposed factor structures for PTSD						EFA factor loadings			Communalities		CFA factor loadings
PDS items	Model A	Model B	Model C	Model D	Model E	Model F	RA	NDA	DAA	In	Ex	RA	NDA	DAA	R^2
B1: intrusive thoughts	R	R	R	R	RA	RA	.67	−.05	.13	.58	.61	.72			.52
B2: recurrent dreams	R	R	R	R	RA	RA	.56	−.03	.10	.40	.39	.75			.55
B3: flashbacks	R	R	R	R	RA	RA	.70	.03	−.16	.37	.36	.64			.41
B4: emotional reactivity	R	R	R	R	RA	RA	.49	−.08	.23	.44	.42	.67			.45
B5: physiological reactivity	R	R	R	R	RA	RA	.73	.03	−.08	.44	.47	.63			.40
C1: avoidance of thoughts	A	A	A	AN	RA	RA	.63	−.04	.30	.75	.73	.89			.80
C2: avoidance of reminders	A	A	A	AN	RA	RA	.75	.03	.14	.74	.73	.87			.76
C3: amnesia	N	D	N	AN	RA	RA	.65	.10	−.14	.31	.33	.46			.21
C4: loss of interest	N	D	N	AN	NA	NDA	.31	.56	−.15	.36	.40		.55		.30
C5: detachment	N	D	N	AN	NA	NDA	.03	.77	−.13	.44	.54		.47		.23
C6: feeling numb	N	D	N	AN	NA	NDA	−.00	.59	.02	.32	.35		.48		.23
C7: hopelessness	N	D	N	AN	NA	NDA	−.10	.76	.08	.48	.59		.70		.49
D1: sleeping difficulty	DA	D	HA	HA	NA	NDA	−.09	.49	.38	.42	.46		.69		.47
D2: irritability	DA	D	HA	HA	NA	NDA	−.07	.54	.27	.39	.44		.64		.41
D3: difficulty concentrating	DA	D	HA	HA	NA	DAA	−.02	.02	.77	.49	.58			.61	.37
D4: overly alert	AA	HA	HA	HA	NA	DAA	.24	−.04	.58	.57	.56			.76	.57
D5: easily startled	AA	HA	HA	HA	NA	DAA	.27	−.02	.53	.57	.54			.88	.78
Factor correlations in EFA
RA								.27	.70
NDA									.38
Factor correlations in CFA
RA								.57	.77
NDA									.45

PDS = Posttraumatic Diagnostic Scale; EFA = Exploratory factor analysis; Model F = EFA-derived model; CFA = Confirmatory factor analysis; PTSD = Posttraumatic stress disorder; Model A = five-factor model (Elhai et al., 2011); Model B = four-factor model (Simms et al., 2002); Model C = four-factor model (King et al., 1998); Model D = DSM-IV-TR model (APA, 2000); Model E = two-factor model (Ayers et al., 2009); RA = Re-experiencing and avoidance; NDA = Numbing and dysphoric-arousal; DAA = Dysphoric-arousal and anxious-arousal; In = Initial; Ex = Extraction; R² = Squared multiple correlations; R = Re-experiencing; A = Avoidance; AN = Avoidance and numbing; N = Numbing; D = Dysphoria; NA = Numbing and arousal; DA = Dysphoric-arousal; HA = Hyper-arousal; AA = Anxious-arousal. The highest loadings are boldfaced.

Table 2. Correlations of the Posttraumatic Diagnostic Scale (PDS) with validity measures (N = 829).

Scale	M	SD	1	2	3	4	5	6	7
1. PDS Total	6.57	7.03	__
2. PDS Re-experience	2.17	2.50	.64	__
3. PDS Avoidance	2.19	2.93	.80	.33	__
4. PDS Hyper-arousal	2.19	2.60	.77	.27	.52	__
5. EPDS Total	7.86	5.07	.60	.31	.56	.46	__
6. HADS Total	11.93	6.78	.61	.29	.58	.47	.68	__
7. HADS Anxiety	7.16	3.83	.49	.32	.42	.39	.54	.83	__
8. HADS Depression	4.77	4.26	.52	.19	.54	.38	.59	.84	.43

All correlation coefficients were significant, p < .001. EPDS = Edinburgh Depression Scale; HADS = Hospital Anxiety and Depression Scale.

Table 3. Agreement of PDS and SCID diagnoses for birth-related PTSD.

	SCID PTSD diagnosis
PDS diagnosis of PTSD	Absent	Present	Total
Negative	62 (46.3%)	4 (3.0%)	66 (49.3%)
Positive	20 (14.9%)	48 (35.8%)	68 (50.7%)
Total	82 (61.2%)	52 (38.8%)	134 (100.0%)

Table 4. Psychometric properties of PDS relative to the SCID diagnosis of PTSD after birth.

Statistical measures	PDS DSM-IV-TR diagnostic criteria	95% CI
Sensitivity (%)	92	80.5–97.5
Specificity (%)	76	64.6–84.1
False-positive rate (%)	24	14.8–33.2
False-negative rate (%)	8	.6–15.4
Positive predictive value (%)	70	58.1–80.7
Negative predictive value (%)	93	84.4–98.0
Overall accuracy (%)	82	69.5–94.5
Kappa (%)	64	51.7–76.9
Positive likelihood ratio	3.78	2.56–5.58
Negative likelihood ratio	.10	.03–.26

Table 5. Fit indices for the six proposed models.

Factor model	χ^2 (a)	Df	CFI^(b)	TLI^(c)	GFI^(d)	RMSEA^(e)	SRMR^(f)	AIC^(g)
Five-factor model (Elhai et al., 2011)	330.83	109*	.93	.91	.92	.070	.071	418.83
Four-factor model (Simms et al., 2002)	352.95	113*	.92	.91	.91	.072	.070	432.95
Four-factor model (King et al., 1998)	493.38	113*	.88	.85	.86	.090	.081	573.38
DSM-IV-TR model (American Psychiatric Association, 2000)	607.95	116*	.84	.81	.82	.101	.085	681.95
Two-factor model (Ayers et al., 2009)	576.73	118*	.85	.82	.83	.097	.074	646.73
EFA-derived model	294.75	116*	.94	.93	.92	.061	.053	368.75

^(a) The χ² values for all models were significant suggesting that a substantial proportion of the variance is unexplained by the model (Kline, 2005); ^(b) CFI = Comparative Fit Index; ^(c) TLI = Tucker-Lewis Index; ^(d) GFI = Goodness-of-Fit Index; ^(e) RMSEA = Root Mean Square Error of Approximation; ^(f) SRMR = Standardized Root Mean Square Residual; ^(g) AIC = Akaike Information Criterion. CFI, TLI and GFI values > .9 indicates a good fit; SRMR and RMSEA < .08 indicates an adequate fit (Hu & Bentler, 1999). The model with the lowest AIC was considered to be the best-fit model among others. * p < .001.

Figure 1. Group differences in PDS and its subscales between women with PTSD (n = 52) and without (n = 82) diagnosed by SCID. Bars represent means of total symptom severity for PDS and each subscale.

Figure 2. Group differences between participants with and without PTSD according to PDS at 6-months postpartum. Group differences in EPDS and HADS along with its subscales between women with PTSD (n = 94) and without (n = 735) diagnosed by PDS. Bars represent means of total symptom severity for EPDS and HADS.

Elhai, J. D., Biehn, T. L., Armour, C., Klopper, J. J., Frueh, B. C., & Palmieri, P. A. (2011). Evidence for a unique PTSD construct represented by PTSD’s D1-D3 symptoms. Journal of Anxiety Disorders, 25(3), 340–345. doi:10.1016/j.janxdis.2010.10.007

 PubMed Web of Science ®Google Scholar

Simms, L. J., Watson, D., & Doebbelling, B. N. (2002). Confirmatory factor analyses of posttraumatic stress symptoms in deployed and nondeployed veterans of the Gulf War. Journal of Abnormal Psychology, 111(4), 637–647. doi:10.1037/0021-843x.111.4.637

 PubMed Web of Science ®Google Scholar

King, D. W., Leskin, G. A., King, L. A., & Weathers, F. W. (1998). Confirmatory factor analysis of the clinician-administered PTSD scale: Evidence for the dimensionality of posttraumatic stress disorder. Psychological Assessment, 10(2), 90–96. doi:10.1037/1040-3590.10.2.90

 Web of Science ®Google Scholar

American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text revision). Washington, DC: American Psychiatric Association.

 Google Scholar

Ayers, S., Harris, R., Sawyer, A., Parfitt, Y., & Ford, E. (2009). Posttraumatic stress disorder after childbirth: Analysis of symptom presentation and sampling. Journal of Affective Disorders, 119(1–3), 200–204. doi:10.1016/j.jad.2009.02.029

 Web of Science ®Google Scholar

Kline, R. B. (2005). Principles and practices of structural equation modelling. New York: Guildford Press.

 Google Scholar

Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6(1), 1–55. doi:10.1080/10705519909540118

 Web of Science ®Google Scholar