Co-morbidity: lessons learned about post-traumatic stress disorder (PTSD) from developing PTSD scales for the MMPI

Results from efforts to develop and validate PTSD measures are promising, but a "gold standard" has not been achieved. Keane, Malloy, and Fairbank (1984) have developed an MMPI PTSD subscale that has been cross-validated with clinicians' classification of PTSD at acceptable levels of agreement, specificity, and sensitivity. There is, however, room for improvement. Empirical evidence is presented that indicates that the next round of efforts to increase reliability and validity of PTSD measures must account for the presence/absence of co-morbidity (i.e., the simultaneous occurrence of other psychiatric disorders). For example, differences are noted in MMPI group profiles and PTSD scales between psychiatric patients and substance abusers. Second, different MMPI items emerge as indicative of PTSD; these vary as a function of the presence of other Axis I disorders among groups of Vietnam combat veterans who seek treatment for substance abuse. Results substantiate that different MMPI items for classifying PTSD occur with groups that differ in co-morbidity. Improvements in PTSD scale development are more likely when the contributions of pre-existing or subsequently co-occurring psychiatric disorders are taken in account, as well as variations in level of personality maturity. The evidence suggests that a "family" of PTSD scales need to be developed that take into account co-morbidity differences.