To our knowledge, this is the first meta-analytic review of stroke or TIA-induced PTSD. We found an overall prevalence of 13% among stroke/TIA survivors, with 23% prevalence in the first year post-stroke and 11% after the first year. Given that about 85% of the 795,000 patients who experience a stroke each year in the US 
survive at least 30 days 
, and up to an additional 500,000 suffer from a TIA 
, these results suggest that 297,850 stroke and TIA survivors in the United States alone will develop PTSD symptoms due to the event annually. Within 90 days of the index stroke, 6% of survivors experience a recurrent stroke 
, and recurrence rates increase substantially over longer term follow up 
. We do not yet know if stroke or TIA-induced PTSD is associated with increased risk for a recurrent cerebrovascular attack, but given the doubling of ACS recurrence risk due to ACS-induced PTSD 
and recent research showing that PTSD is strongly associated with medication nonadherence 
, such research is sorely needed.
Important caveats to this review include the fact that only studies with relatively small sample sizes have addressed stroke-induced PTSD, and one of the studies represented half of all of the participants included in this review. Also, none of the included studies used population-based sampling, and all but one study 
relied on cross-sectional designs (and we used only the earliest point estimate, 1 month after stroke). Further, the PTSD screening questionnaires used in many of the studies have not been validated against clinical diagnosis of PTSD in survivors of stroke/TIA, though they perform very well in other patient populations. Thus, although this meta-analysis was able to quantify more precisely the prevalence of stroke-induced PTSD symptoms, a clear need for additional research remains. It is also important to note that we were very strict in our inclusion of only studies that measured PTSD specifically due to the stroke event,
so these estimates may represent an underestimate of the total burden of PTSD symptoms in these participants in so much as PTSD due to other types of events may have been present.
The estimated prevalence of stroke-induced PTSD was lower when PTSD was assessed by clinical interview. This may be due to the fact that the prevalence of psychiatric disorders is often higher when measured using diagnostic instruments as compared to clinical psychiatric interviews Nevertheless, research on the association between ACS-induced PTSD and ACS recurrence and mortality risk suggests that even elevated symptoms of PTSD, not clinical diagnosis, is associated with increased risk. Hence, the estimate of the prevalence of PTSD symptoms based on diagnostic instruments may still be clinically relevant 
Prevalence estimates were limited to patients who could participate in PTSD assessments and hence should be extrapolated to patients with severe cognitive impairment or aphasia with caution. Similarly, these results should be interpreted with awareness that the mean age of the participants was slightly low relative to the entire population of stroke survivors and PTSD is associated with younger age in ACS-induced PTSD 
, and therefore our rate estimate may not generalize to older stroke survivors. We were unable to determine whether the prevalence of PTSD differed if due to TIA or stroke, as the only study to include TIA survivors 
did not record the proportion of participants who experienced either condition. Moreover, none of the studies reported on the severity of the stroke event in terms of the types of critical care that may have been necessitated such as intensive care unit (ICU) admission.
Only one of the studies included in this review evaluated for stroke lesion localization 
, suggesting the need to further examine whether stroke location (i.e. hemisphere, etc) or the nature of the associated deficits (i.e. paresis/paralysis, aphasia, anosagnosia) further differentiate PTSD risk. Such knowledge would allow clinicians to identify those at higher risk for PTSD to implement early intervention. Similarly, we opted to retain one study that focused on PTSD due to subarachnoid hemorrhage, which clinicians and epidemiologists view as a distinct condition from stroke/TIA, because of the similarities in patient experience and because the aggregate estimate was not sensitive to its exclusion. Future research should determine whether PTSD due to subarachnoid hemorrhage is rightly grouped with PTSD due to stroke/TIA.
It is also important to note the difference in prevalence estimates reported in studies that assessed PTSD by screening questionnaire versus clinical interview. Although screening questionnaires are often used to identify patients who should be further assessed for PTSD diagnosis–and therefore may overestimate prevalence–reliance on screening questionnaires for estimates of stroke-induced PTSD prevalence may further inflate those estimates due to overlap in symptoms of stroke and PTSD. Very little is known about the validity of PTSD questionnaire items in stroke survivors, but the evidence that exists suggests that PTSD questionnaires are valid in this population. In one study, confirmatory factor analysis of responses to the PTSD Checklist-specific for stroke conformed to the established factor structure for PTSD due to other types of events 
. In another small study, PTSD severity was not related to either lesion site or neurologic and memory deficits that could be expected to inflate PTSD symptoms were the scale not valid 
. Further research on the influence of neurologic and cognitive deficits on PTSD symptom expression is sorely needed. Finally, future research should also consider concomitance with other comorbid stroke factors including depression, however, it is important to note that the effect of ACS-induced PTSD on ACS recurrence and mortality is independent of depression 
Across published studies estimating the prevalence of stroke-induced PTSD, two general conclusions can be drawn: (1) stroke-induced PTSD is relatively common, with approximately one in four experiencing PTSD in the first year after stroke and one in nine experiencing chronic PTSD over a year later, and (2) based on a single study, stroke-induced PTSD symptoms appear to influence important secondary prevention behaviors such as medication adherence 
, independent of depression. Though no data now exist to address the issue, if the risk of stroke or other cardiovascular event recurrence associated with stroke-induced PTSD is comparable to that found for PTSD after ACS, stroke-induced PTSD may be a significant, novel risk factor for recurrent stroke, especially given its high prevalence after stroke. Screening for PTSD in a large, population-based prospective cohort study would yield definitive prevalence and secondary risk estimates, and should be a priority for researchers. In the meantime, clinicians should be mindful that PTSD can be a devastating mental health condition and should consider screening for PTSD in stroke survivors.