Stress reactions are common in parents in the aftermath of a life-threatening medical event of their child. However, a minority of parents develop chronic mental health problems [1]. Most common mental health problem after experiencing highly stressful events is posttraumatic stress disorder (PTSD), which is characterized by intrusive distressing memories, avoidance, emotional numbing and hyperarousal [2]. Other mental health problems may also be seen such as depression, anxiety disorder, sleep disturbances, and substance abuse [2]. Identification of parents with mental health problems after a child's life-threatening illness or injury is important. Once these parents are identified, psychological support can be offered at an early stage, aimed at minimizing chronic mental health problems and preserving their competence as caregivers. Consequently, parents will be able to support their child's recovery trajectory and adjustment in their best possible way. Therefore, improving identification and raise awareness of PTSD is a necessary first step in pediatrics.

Prevalence rates of mental health problems in parents vary widely after different life-threatening medical events. Research has mainly focused on cancer, diabetes and accidents with rates ranging from 10% to 40% for PTSD, anxiety and depression in parents [3–5]. Overall, women seem to have a higher risk of developing PTSD than men. Studies in heterogeneous pediatric intensive care treatment (PICU) populations have identified PTSD in approximately 13-27% of parents [6–10]. Prevalence rates of general psychological distress in PICU parents are even higher and exceed rates of distress of parents with children in general wards [7, 10].

Risk factors for parental mental health problems are scarcely studied within PICU. Studies suggest that parental PTSD is not strongly related to objective characteristics of the PICU treatment but is related to parents' perceptions of the life threat for their child and to acute stress reactions in the PICU [6–10]. Research within pediatrics and traumatic stress studies has shown that responses immediately following the stressful event can help to predict the course of PTSD over time. For example, findings show that coping styles such as avoidance coping and passive reaction pattern have been linked to more mental health problems in pediatrics [11–13]. Furthermore, a recent meta-analysis of PTSD predictors in adults after interpersonal violence, combat and accidents suggested that peritraumatic psychological processes and peritraumatic dissocation are the strongest predictors of PTSD [14].

Peritraumatic dissocation is a state of limited or distorted awareness during and immediately after the stressful event. Examples of symptoms of peritraumatic dissocation are reduced awareness, time distortion, derealisation, amnesia or emotional numbing [15]. It has been suggested that such symptoms reflect a defensive response related to immobilization (freezing) in animals [16]. In addition, high levels of peritraumatic dissocation in adults during a stressful event such as interpersonal violence or burn injury, may also predict symptoms of psychopathology, such as anxiety and depression [17, 18].

So far, only five studies examined prevalence of parental PTSD in heterogeneous PICU populations [6–10]. Anxiety and depression prevalence rates after PICU treatment have hardly been studied yet. Furthermore, until now no research has been conducted on whether coping styles or peritraumatic dissocation of parents after PICU treatment are risk factors of mental health problems such as PTSD, anxiety and depression. Therefore, the first aim of the present study was to describe the prevalence of mental health problems (PTSD, anxiety and depression) in parents three months after discharge from the PICU. The second aim of the study was to examine if coping styles and peritraumatic dissocation shortly after the stressful event are related to mental health problems in parents.

This explorative study shows that 12.2% of parents were likely to meet diagnostic criteria for PTSD and on top of that 24.3% were likely to meet criteria for subclinical PTSD three months after PICU treatment. Respectively, 23.4% and 15.6% of parents reported possible clinically significant anxiety and depression. Mothers reported significantly more PTSD and anxiety than fathers did. Peritraumatic dissocation was related to mental health outcomes in general. Avoidance coping was primarily associated with PTSD. Furthermore, passive reaction pattern and comforting thoughts were significantly associated with anxiety.

The prevalence rate of parents that were likely to meet diagnostic criteria for clinical PTSD is similar to earlier research at PICU [6–10]. In addition, almost one quarter of parents were likely to meet criteria for subclinical PTSD which can lead to clinically meaningful levels of functional impairment as well [25]. Furthermore, in line with most studies on PTSD, the results show higher rates of PTSD in women than in men [26]. To our knowledge, the prevalence rates of anxiety and depression in parents of unexpectedly PICU admitted children have not been studied before. Compared to findings of a large national representative survey in The Netherlands the prevalence rates of mental health problems in the present study are considerably higher. In general, the 1-month prevalence of anxiety disorders in The Netherlands is 9.7%. Anxiety disorders are more prevalent than mood disorders. The 1-months prevalence of mood disorders is 3.9% in The Netherlands [27].

The avoidance coping strategy was strongly associated with symptoms of PTSD. PTSD symptoms increased as a function of using avoidance coping. This effect of avoidance coping has also been found in several earlier studies after cancer, and in general stress literature [11–13, 26]. Interestingly, avoidance coping strategy was not related to anxiety and depression, indicating that avoidance coping may pose increased risk for specific posttraumatic stress reactions [11]. However, causality has not been established and avoidance coping may reflect a representation of the same underlying construct (e.g. overlap with avoidance symptoms of PTSD). Next to avoidance coping, peritraumatic dissocation also turned out to be significantly associated with symptoms of PTSD, as well as with symptoms of anxiety and depression. However, some recent studies suggest viewing the relationship between peritraumatic dissocation and PTSD as an artefact of confounding variables. In other words, peritraumatic dissocation is related to PTSD because it is associated with other risk factors such as prior mental health problems [28, 29]. In sum, there seems to be a strong relationship between peritraumatic dissocation and mental health problems. Yet, this should not be interpreted as proof for a causal relationship and further prospective research is necessary to disentangle this connection [30].

The passive coping strategy was associated with anxiety and depression. The relationship between passive coping and mental health problems has been found in previous research after cancer, and in general stress literature as well [12, 26]. Once again, this association may reflect a shared, underlying construct, or it may indicate a causal relationship with either distress affecting coping or coping affecting distress. If the association between passive coping and anxiety or depression is direct, this coping strategy could be seen as maladaptive. Passive coping may be related to the concepts of learned helplessness and locus of control [31]. These theories propose that perceived absence of control over the situation will lead to more negative mental health outcome. Helping parents manage the PICU period by regaining perceived control might be effective in reducing these outcomes (e.g. involve parents in the care for the child).

Some limitations of the study should be addressed. The first limitation of this study pertains to sampling issues. A considerable number of parents and children were lost due to non-response. Fifty families (37%) did not visit the outpatient follow-up clinic. This may have biased the results, even though similar response rates were found in earlier studies. The sample size for the regression analyses was even smaller. The timing of our study was not optimal since medical staff had to get used to the new standardized protocol of transfer out of the PICU. Our centre implemented this transfer protocol, of which the questionnaires were part of, in January 2006. Furthermore, we suspect that few parents were motivated to complete questionnaires immediately after PICU discharge due to possibly continuing stress of the hospital admission. Consequently, this small sample size raises questions about the generalizability of study findings and the degree to which study participants are representative of typical PICU populations. Therefore, findings of this study are preliminary and exploratory. Besides, it minimized the number of risk variables that could be included in the analyses of our study. Therefore, gender differences in peritraumatic dissocation and coping could not be analyzed. Second, a structured clinical interview can be regarded as the best measurement for mental disorders. The use of self-reports only gives an indication for the diagnosis of mental disorders and cut-off scores should be used with caution. Self-reports can lead to an overestimation of cases with mental health problems. Nevertheless, good diagnostic agreement between the SRS-PTSD self-report measure and clinical interviews for PTSD has been reported [19, 20]. Third, in identifying risk factors for mental health problems of parents, other risk factors might be relevant, such as initial mental health problems, perceived life threat or previous stressful events. Future research should investigate multiple risk factors and their interactions in order to unravel the mechanisms underlying longer-term mental health problems.

Notwithstanding the limitations, the present study is one of the first longitudinal follow-up studies on anxiety, depression and PTSD in a relatively large number of mothers and fathers of children after unplanned PICU treatment, examining peritraumatic dissocation and coping. The results of this pilot study support that many parents experience symptoms of PTSD i.e. subclinical PTSD. Approximately one third of parents show clinical levels of PTSD, anxiety and depression after unexpected admission of their child at PICU for which adequate psychological support is necessary. Avoidance coping, passive reaction pattern and particularly peritraumatic dissocation were associated with mental health problems in parents. The presence of the variety of emotional reactions in the sample underscores the need for medical staff and psychosocial professionals to identify parents at risk and intervene in an early stage to minimize chronic and pathologic mental health problems. A next step is to replicate these findings in a larger sample of parents and explore possible other risk factors for mental health problems. So, parents at risk can be identified and monitored in an early phase and referred for psychological support if necessary.

Finally, this study can have some clinical implications for early identification of those parents at risk. The strong relationship between peritraumatic dissocation symptoms at PICU and parental mental health problems at three months suggests that pediatric health care providers in the hospital should ask parents about these peritraumatic dissocation symptoms. Inquiring parents during PICU admission about reduced awareness (do you ever lose track of what is going on around you?), time distortion (do you ever feel as though you are disoriented, as though you are uncertain about where you are of what time it is?), derealisation (do you ever feel as though you are a spectator, watching what is happening to you as if you were an outsider?), amnesia (can you remember everything of the PICU admision?) as well as emotional numbing (do you feel a restricted range of affect?) may help to identify those who are in need for further assessment and psychosocial support. This assessment is particularly warranted when the parent also applies passive and avoidance coping styles. In addition, a set of informational materials for use by pediatric health care providers has recently been developed: the medical traumatic stress toolkit [32]. This toolkit includes a preventative intervention model suggesting that the health care team provide every family with general information and basic support, and regularly screen for acute stress symptoms and risk factors to determine which children and families might need more support. This toolkit should be made accessible for parents and children at PICU and should be evaluated in future research for its effects on preventing or reducing PTSD, depression and anxiety http://www.nctsn.org/medtoolkit.