Validation of a screening instrument for identification of patients, who have a risk to develop a Post Traumatic Stress Disorder (PTSD), based on data of severely injured accident victims

Status:

completed 04/2007

Aims:

The object of the project was the validation of a screening instrument (Freiburger PTSD-Screener) for identification of patients, who have a risk to develop a post traumatic stress disorder. Of special interest was the prognostic relevance to identify PTSD in severely injured accident victims six month after an accident.

Activities/Methods:

For validation we used the data of different psycho diagnostic instruments (revised version of the Impact of Event Scale IES-r, German translation of the Posttraumatic Stress Diagnostic Scale PDS-d1) and of the screening instrument from 130 severely injured accident victims (four hospitals): - Time of measurement - Admission on ward and six month after the accident, at both times examination with the same set of questionnaires - Comparison of the screener with the data of the other diagnostic instruments (at both time points) - Evaluation of the prognostic ability of the screener: criterion validity, prognostic validity, sensitivity and specification for different cut of values, and divergent validity

Results:

The purpose of the present study was the identification of risk patients for the development of a Post Traumatic Stress Disorder (PTSD) based on a group of severely injured accident victims. Of special interest was the identification of risk patients for the development of PTSD six months after accident. As variables we calculated sensitivity, specivity and the optimal cut-off-value as well as Youden-Index and Cross-Product-Ratio. Furthermore, we were interested in the values of divergent and convergent validity, to assess how far the screening-questionnaire is able to identify patients with PTSD and to which extent it is able to discriminate between PTSD and other mental health problems. In the context of a DFG-funded study we tested 113 surgically treated patients with multiple injuries after an accident for the existence of mental health problems as the result of the accident. Additionally, we used the screening questionnaire. Times of measurement were the admission on ward as well as 6 month after accident. All patients on ward got psychotherapy, 50% of them got supplemental ambulant psychotherapy subsequent to a phase of rehabilitation. For validation we could use the data of 59 of the 113 study patients. To evaluate the prognostic validity we considered patients with PTSD as well as patients with partial PTSD. For the calculation of divergent validity, we considered the existence of symptoms of anxiety and depression. On ward 55.4% of patients showed symptoms of mental health problems, a total of 33.2% showed symptoms of PTSD or partial PTSD. Six months after accident 39.5% of the examined patients still suffered from mental health problems, 28.2% showed symptoms of PTSD or partial PTSD. As an optimal cut-off-value for the screening questionnaire we calculated a value of three. For this cut-off-value the values of sensitivity and specivity were 86.7% (Sens) and 69.2% (Spec), the value of the Youden-Index was 0.56 and the value of the Cross-Product-Ratio 14.7. The variance shared between screening questionnaire and BDI and the subscales of the BSI amounted from 19% to 35% depending on the questionnaire or subscale used. The shared variance between BDI and screening questionnaire and between the subscale "social incertitude" of the BSI and the screening questionnaire were greater than 50%. However, requested values are as small as possible. Vice versa the shared variance between the screening questionnaire and the sum score of IES-R as well as PDS-d1 was 26.7% and 21.6%. Overall the values of the prognostic validity were satisfying; the values of divergent and convergent validity were nonsatisfying. That means, it cannot be guaranteed that the screening questionnaire only identifies patients with PTSD. It should not be used as a stand-alone instrument for identifying risk patients for PTSD and for an indication of symptom specific PTSD-treatment.

Last Update: