Skip to main content
NCBI home page
Journal of Child & Adolescent Trauma logoLink to Journal of Child & Adolescent Trauma
. 2018 Jun 28;13(1):1–10. doi: 10.1007/s40653-018-0222-z

Posttraumatic Stress Symptoms and Disorders in Children and Adolescents with Chronic Physical Illnesses: a Meta-Analysis

Martin Pinquart 1,✉
PMCID: PMC7163825  PMID: 32318223

Abstract

The aim of the present meta-analysis was to compare levels of posttraumatic stress symptoms (PTSS) and the prevalence of posttraumatic stress disorders (PTSD) in individuals with and without pediatric chronic physical illnesses, and to analyze correlates of these symptoms. In total, 150 studies were identified that provided relevant data. On average, 11.5% of the participants with pediatric chronic physical illnesses met the criteria of PTSD (Odds Ratio 2.70). PTSS were also more common in this group than in control groups without chronic physical illnesses. While the PTSS levels did not differ across physical diseases, we found positive associations of PTSS with illness severity and duration/intensity of treatment, as well as negative associations with duration of illness, time since last treatment, treatment adherence, and family functioning. We conclude that individuals with pediatric chronic physical illnesses who experienced traumatic events should be screened for PTSS and receive psychological interventions when needed.

Electronic supplementary material

The online version of this article (10.1007/s40653-018-0222-z) contains supplementary material, which is available to authorized users.

Keywords: Posttraumatic stress disorder, Pediatric medical traumatic stress, Chronic illness, Meta-analysis, Burn, Cancer, Heart disease, Traumatic brain injury

Introduction

Being diagnosed with a severe chronic illness and undergoing highly stressful experiences within the context of the illness can be characterized as traumatic events, and the resulting invasive treatment could also be labeled as trauma (Barakat et al. 1997; Meentken et al. 2017). The present meta-analysis compared the levels of posttraumatic stress symptoms (PTSS) and the prevalence of posttraumatic stress disorders (PTSD) in individuals diagnosed with chronic physical illnesses in childhood and adolescence to peers without such diagnoses or to community norms, respectively. We also analyzed correlates of these symptoms in young people with chronic illnesses. PTSS/PTSD in young people with chronic physical illnesses is a relevant topic because PTSS and PTSD are associated with impaired psychological well-being, as well as possible negative effects on later physical functioning and school functioning (O’Connor et al. 2012), and, on the course of the physical illness itself, if avoidance symptoms restrict treatment adherence and intention to seek follow-up care (Kerkar et al. 2006). Thus, knowledge on these topics is relevant for psychological assessment, as well as for prevention and intervention.

The 5th edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5; American Psychiatric Association [APA] 2013) lists 20 PTSD symptoms divided into four clusters: Intrusion (e.g., flashbacks, intrusive thoughts), avoidance (e.g., staying away from reminders of the traumatic event), negative alterations in cognition and mood (e.g., persistent and negative beliefs about oneself or the world), and alterations in arousal and reactivity (e.g., feeling easily startled). A defined number of symptoms in the four clusters must be present over a period of one month in order to qualify for a PTSD diagnosis. In the 4th edition of the DSM, being confronted with an event or events that involved actual or threatened death, serious injury, or threat to the physical integrity of oneself or others have been listed as traumatic events that may lead to PTSD (APA 1994). Having a chronic illness meets the exposure criterion because it threatens the physical integrity and could – in a number of diseases – lead to death. The DSM-5 narrowed down the list of possible traumatic events and discarded threats to the physical integrity; therefore, chronic illnesses no longer meet the exposure criterion for a PTSD diagnosis if they are not associated with increased mortality.

In contrast to PTSD, PTSS refer to subclinical symptoms of posttraumatic stress that do not qualify for a PTSD diagnosis; they are more prevalent than the clinical diagnosis itself (e.g., Barakat and Wodka 2006). Some authors also use the term pediatric medical traumatic stress when referring to acute stress symptoms and PTSS that result from pain, injury, serious illness, medical procedures, and invasive or frightening treatment experiences (e.g., Price et al. 2016).

There are several sources of traumatic stress that can result from chronic physical diseases (Barakat et al. 1997; Meentken et al. 2017): First, getting the information that one has been diagnosed with cancer or another potentially fatal disease is a traumatic stressor if individuals are aware of the associated life threat (Buckley et al. 2004). Second, traumatizing experiences within the context of the illness include disturbing symptoms (e.g., dyspnea in the case of asthma attacks, severe pain), stays in the emergency room, side effects of intense and invasive treatment, encountering severely ill patients during a hospital stay, or death of fellow patients. Such experiences are not limited to individuals with potentially fatal diseases.

A meta-analysis on eight studies addressing pediatric chronic illnesses (mainly cancer and liver diseases) reported that, on average, 12% of the affected children and adolescents fulfilled the criteria of PTSD, with rates varying between 0 and 32% (Kahana et al. 2006). In an additional analysis on young people with heterogeneous injuries, 19.8% fulfilled the criteria of PTSD (range 0–37.5%). Two narrative reviews indicated that the current prevalence of PTSD in children with cancer varies between 4 and 29% (Bruce 2006; Yalug et al. 2011). However, a number of individual studies showed no elevated PTSS or PTSD rates in survivors of pediatric cancer compared to healthy controls or population norms (Kazak et al. 1997; Phipps et al. 2014). Sources of the heterogeneity of the results of the individual studies have not yet been systematically analyzed.

Given the fact that the number of studies on PTSS and PTSD in children and adolescents with chronic physical illnesses has considerably increased since the first meta-analysis by Kahana et al. (2006), the goal of the present meta-analysis was to provide an update and extension of the previous meta-analysis. Because larger numbers of young people with chronic illnesses show PTSS without meeting the diagnostic criteria of PTSD (e.g., Price et al. 2016), we broadened the focus of the previous meta-analysis by adding studies on PTSS. The first research question asked whether children and adolescents with chronic physical illnesses report higher levels of PTSS and higher rates of PTSD than their healthy peers. The other research questions addressed sources of heterogeneity across the individual studies. More concretely, the second research question asked whether the size of between-group differences varies by study characteristics, such as the type of illness. Finally, the third research question asked for correlates of PTSS in young people with chronic physical illnesses.

Variables Affecting the Size of Differences in PTSS/PTSD between Young People with and without Chronic Illnesses

Type of Illness

Given the stricter criteria for illness-related traumatic events in the DSM-5 (APA 2013), we asked whether PTSS/PTSD would be more prevalent in young people diagnosed with cancer and HIV-infection/AIDS than in those with diseases that are not associated with strong increases in mortality, such as juvenile diabetes and asthma. Although comparisons across illnesses have rarely been made in the field of pediatric PTSS, studies found higher rates of PTSS in children with cancer than in children with asthma (Landolt et al. 2003) and arthritis (Graziano et al. 2016).

Age at Diagnosis/Assessment

The stage of development (or age as a proxy variable) at exposure to a traumatic event is relevant for understanding the imminent threat to one’s life. Young children have limited cognitive capacities for evaluating the threat of the illness to their lives (Yalug et al. 2011). Thus, PTSD has rarely been diagnosed in the preschool years (APA 2013). Although the DSM-5 has defined a preschool subtype of PTSD with more behaviorally anchored criteria, studies that assessed this subtype in preschoolers with chronic physical illnesses are still lacking. The present meta-analysis tested whether smaller elevations of PTSS would be found if the chronic illness was diagnosed and PTSS were assessed in the preschool years rather than at a later age.

Assessment of PTSD Versus PTSS

We asked whether the effect sizes vary between studies that assessed PTSD rates and studies with continuous PTSS measures. Effect sizes tend to be slightly smaller when using a dummy variable (such as receiving or not receiving a PTSD diagnosis) rather than a continuous variable due to variance restrictions (Lipsey and Wilson 2001).

Study Quality

Sum measures of study quality should not be used for a moderator analysis because some lower quality aspects may lead to underestimations of effect sizes, while others could lead to overestimations (Higgins and Green 2009). Thus, we asked whether between-group differences in PTSS vary by four individual criteria derived from the Modified Quality Index that has been previously used in research on pediatric chronic conditions (Ferro and Speechley 2009). The criteria address (a) external validity (whether the participants were representative of the entire population from which they were recruited), (b) internal validity (valid and reliable assessment of PTSS/PTSD); (c) whether young people with and without chronic illnesses did not differ in third variables or whether the analysis adequately adjusted for confounding effects of third variables, and (d) sufficient statistical power. With regard to statistical power, we assessed whether the sample size was sufficient for detecting an effect of d = .50 that would be found when comparing the PTSD rate of 12% in young people with chronic illnesses from the meta-analysis by Kahana et al. (2006) to the PTSD rate in the general adolescent U.S. population (Kilpatrick et al. 2003). Because non-significant effects may be less likely to be published than significant effects (the file-drawer problem; Lipsey and Wilson 2001), we also tested whether elevations of PTSS would be lower in unpublished studies than in published studies.

Correlates of PTSS in Young People with Chronic Illnesses

The third research question addressed correlates of PTSS in children and adolescents with chronic physical illnesses. We mainly focused on correlates related to the chronic illnesses and their treatments, as well as those related to social resources that may serve as protection from the development of PTSS.

Age at Diagnosis and Assessment

We tested whether absolute levels of PTSS would vary by age at diagnosis and assessment. Note that the meta-analysis by Kahana et al. (2006) found no association of age with PTSS in children with pediatric illnesses or injuries.

Time since Being Diagnosed or Treated

Most studies on single traumatic events report a decline in PTSS over time, although a minority shows persistent symptoms (e.g., Le Brocque et al. 2010). However, it is not clear whether this could be generalized to chronic illnesses, because the physical illness and associated consequences may persist or even deteriorate over time. While the review by Bruce (2006) indicated that time since diagnosis or treatment cessation would not be related to PTSS in cancer patients, the meta-analysis by Kahana et al. (2006) found that a longer period of time since the last treatment was associated with lower levels of PTSS in individuals with pediatric chronic diseases or injuries. Thus, we asked whether the conclusions of Bruce (2006) or Kahana et al. (2006) could be replicated in the present meta-analysis.

Illness Severity

A more severe illness is likely to have more traumatic features, such as a threat of mortality, pain, or other symptoms (Barakat et al. 1997; Meentken et al. 2017). Thus, we asked whether a more severe chronic illness is associated with more PTSS. Four of five studies reviewed by Kahana et al. (2006) found that stronger trauma severity was associated with more PTSS. In contrast, Bruce (2006) reported that the severity of cancer failed to predict PTSS.

Treatment Duration/Intensity

Invasive and frightening treatment experiences are possible sources of PTSS (Price et al. 2016). Thus, we asked whether higher treatment intensity/longer treatment duration would be associated with higher levels of PTSS.

Adherence

Because avoidance of reminders of the traumatic event is a core symptom of PTSD, PTSS may be related to lower adherence, such as missing appointments with the physician and not seeking follow-up care (Kerkar et al. 2006). Nonetheless, this association may be weaker in pediatric patients than in adult patients because parents take some responsibility for their child’s adherence. Thus, we asked whether a negative association between PTSS and adherence could be found in young people with chronic physical illnesses.

Social Resources

Social resources, such as positive family functioning and social support, have been suggested as factors that protect pediatric patients from developing PTSS/PTSD (Bruce 2006; Kazak et al. 2004). Surprisingly, the meta-analysis by Kahana et al. (2006) found a significant positive association of social support with PTSS in young people with pediatric illnesses. The present meta-analysis tested whether this result can be replicated.

Methods

Sample

Studies were identified in a search using electronic databases [PSYCINFO, MEDLINE, Google Scholar, PSYNDEX (an electronic database of psychological literature from German-speaking countries)] using the search terms: (posttraumatic stress or post-traumatic stress or pediatric medical traumatic stress) AND (children or adolescents or pediatric) AND (illness or disease). In addition, the references sections of the identified papers were checked for additional studies. Unpublished studies (e.g., dissertations, master’s theses) were identified as part of the systematic search using electronic databases (PSYCINFO, Google Scholar, PSYNDEX) and cross-referencing, and were included if they met the inclusion criteria.

Criteria for inclusion of studies in the present meta-analysis were:

  1. the studies assessed present PTSS or PTSD in individuals who have been diagnosed with chronic physical illnesses during childhood or adolescence

  2. the studies provided sufficient information for a comparison of levels of PTSS or rates of PTSD with established normative data or a similar group of physically healthy individuals, or reported correlations of PTSS/PTSD of individuals diagnosed with chronic physical illnesses in childhood or adolescence with age at diagnosis, age at assessment, time since being diagnosed or treated, illness severity, treatment duration/intensity, adherence, family functioning, and/or social support.

  3. the studies were published or presented before November 2017.

We did not restrict the age at assessment of PTSS/PTSD. Thus, the former pediatric patients could already have reached adulthood when they had their PTSS/PTSD assessed. In regard to the question whether a condition is defined as chronic, we followed the suggestion by Thompson and Gustafson (1996), stating that a chronic illness can be defined as a condition that is associated with functional impairment, persists for more than three months in a year, and/or necessitates a period of continuous hospitalization for more than a month. Based on this suggestion, we also included patients with pediatric burns and pediatric traumatic brain injuries, because they tend to be confronted with symptoms that persist over time, such as changes in appearance and cognitive abilities, respectively (Anderson and Knight 2010). The literature search was completed on October 30, 2017. We identified 590 papers. After screening and assessing for eligibility, we were able to include 150 studies in the meta-analysis. A flow chart of the search for studies is provided in Appendix A1, and the studies included are listed in Appendix A2 (see supplementary online material).

All studies were coded by the author, and a random sample of 30 studies was also coded by a psychologist with experience in meta-analyses. Differences between the two coders were resolved by discussion. Validity of the PTSS/PTSD measures was coded as high if support for the validity of the measures was available and if established structured clinical interviews were used for the diagnosis of PTSD. In contrast, validity of PTSD status was coded as limited if only cut-offs of screening questionnaires were used for estimating PTSD rates, or if information on the diagnostic process was lacking.

We entered the number of individuals with pediatric chronic illnesses (intraclass correlation coefficient [ICC] = 1.0), number of individuals in the control condition (ICC = .88), mean age at assessment (ICC = .95), mean age at initial diagnosis of the physical illness (ICC = .90), time since diagnosis (ICC = .92), indicator of illness severity (inter-rater agreement 96%), indicator of treatment intensity/duration (inter-rater agreement 90%), indicator of social support (inter-rater agreement 100%), indicator of family functioning (inter-rater agreement 100%), publication status (1 = published, 2 = unpublished; ICC = 1.0), type of illness (inter-rater agreement 100%), representativeness of the sample (1 = yes, 0 = no/not reported; ICC = .90) equivalence of patient and control group (1 = yes, 2 = not tested, 3 = no; inter-rater agreement 93%), available support for the validity and reliability of the measures (1 = yes, 0 = limited/no; ICC = .92), method for assessing PTSS/PTSD (inter-rater agreement 97%), method for assessing the correlates of parenting stress (inter-rater agreement 93%), the standardized size of between-group differences in PTSS (ICC = .94), the percentage of individuals with PTSD diagnoses (ICC = 1.0), the Odds Ratio of being diagnosed with PTSD (ICC = .90), correlations of PTSS with the child’s age at assessment (ICC = .94), age at diagnosis of a chronic physical illness (ICC = .96), illness duration (ICC = .94), illness severity (ICC = .95), treatment duration/intensity (ICC = .93), social support (ICC = 1.0), and family functioning (ICC = 1.0). Whether the study had sufficient test power for identifying moderate effect sizes (1 = yes, 0 = no) was assessed based on the sample sizes. If more than one measure of PTSD/PTSS was used, results were averaged across the measures.

Measures

PTSS/PTSD were assessed by use of structured clinical interviews, such as the Schedule for Affective Disorders and Schizophrenia for School-Age Children (Kaufman et al. 1997; 50 studies), the Childhood Posttraumatic Stress Reaction Index (Frederick et al. 1992; 27 studies), various versions of the Impact of Events Scale (Weiss and Marmar 1997; 25 studies), the UCLA Posttraumatic Stress Disorder Reaction Index (Rodriguez et al. 1998; 16 studies), and related instruments (46 studies). Illness severity was measured by the number or severity of symptoms and impairments associated with the disease (23 studies, such as amount of burned surface, 8 studies). Treatment intensity and duration were assessed via the length of stay in a hospital (2 studies) and ratings on treatment intensity, such as the Treatment Rating scale 2.0 (Werba et al. 2007; 6 studies). Adherence was assessed via clinician-reports (3 studies) or self-reports (Self-Care Inventory-Revised; La Greca 2004; 1 study). Social support was assessed with the Multidimensional Scale of Perceived Social Support (Zimet et al. 1988; 1 study) and other related instruments (3 studies). Finally, family functioning was measured with the Family Environment Scale (FES) (Moos and Moos 1994; 2 studies) and other measures (4 studies).

Statistical Integration of the Findings

Calculations for the meta-analysis were performed in seven steps with Comprehensive Meta-Analysis (Borenstein et al. 2005), using random-effects models and the method of moments.

  1. If the study did not include a control group of young people without chronic physical conditions, we searched for a comparison group from the same country. With electronic databases (PSYCINFO, Google Scholar), we first tried to find a comparison group of young people with similar ages, without chronic physical illnesses, that has been assessed with the same PTSD measure as the target sample. If this was not possible, we searched for a community sample of young people for which no information about physical diseases was available. Because we did not want to compare children with chronic physical illnesses with otherwise highly distressed samples (e.g., victims of child abuse or earthquakes), community samples were only considered if they were not defined by confrontation with another traumatic event. If more than one sample was available as a potential control condition, we selected the sample that was demographically most similar to the target sample. If no comparison group could be identified for an individual sample, the study had to be excluded from the meta-analysis. This was the case with regard to 11 studies (Appendix A4).

  2. We computed effect sizes d as the difference in levels of PTSS between the sample with pediatric chronic physical illnesses and the control sample, divided by the pooled standard deviation (SD). Logged Odds-Ratios (OR) were computed for comparing the numbers of individuals with PTSD in samples with and without chronic physical illnesses. Associations of PTSS with age and aspects of illness, treatment, and social relations were coded as correlation coefficients. Outliers that were more than two SD’s from the mean effect size were recoded to the value at two SD’s, based on Lipsey and Wilson (2001).

  3. The effect sizes d were transformed to Hedges’ g, in order to correct for bias due to overestimation of the population effect size in small samples. The correlation coefficients were transformed using Fisher’s r-to-z transformation.

  4. Weighted mean effect sizes and 95%-confidence intervals were computed. The significance of the mean was tested by dividing the weighted mean effect size by the standard error of the mean. To compare the mean z-scores with the correlation coefficients reported in the individual studies, the mean z-scores were later converted to the original metric of product–moment correlations. The alpha level was set at a conservative p < .01. To interpret the practical significance of the results we used Cohen’s criteria (Cohen 1992): Effect sizes of g = .20 to .49 and r = .10 to .29 are small, g = .50 to .79 and r = .30 to .49 are medium, and g ≥ 0.8 and r ≥ .5 are large.

  5. Homogeneity of effect sizes was tested by use of the Q statistic.

  6. In order to test the influence of moderator variables, we used an analogue of an analysis of variance. A significant Q score indicates heterogeneity of the effect sizes between the compared conditions.

  7. Egger’s test and the trim-and-fill algorithm by Duval and Tweedie (2000) were used in order to check whether the results may have been influenced by publication bias.

Results

The 150 studies provided data on 37,382 individuals with pediatric chronic physical illnesses. The participants most often had TBI (N = 17,209), cancer (N = 13,765), heart disease (N = 1090), and burns (N = 1007). They had a mean age of 14.10 years (SD = 6.04); 48% were female, and 21% were members of ethnic minorities. At the time of the study, they had the illness, on average, for 5.93 years (SD = 5.86). If the treatment had been completed (e.g., organ transplantation), the average time since last treatment was 7.48 years (SD = 4.82, based on 46 studies).

Comparing Levels of PTSS and Rates of PTSD in Individuals with and without Pediatric Chronic Illnesses

To answer the first research question, we compared levels of PTSS and rates of PTSD in individuals with and without pediatric chronic physical illnesses. Studies that only provided ORs of PTSD were also included after transforming the ORs into a d-score based on Lipsey and Wilson (2001). Separate analyses on PTSD only included those studies that reported numbers of individuals who were diagnosed with PTSD or scored above the clinical cut-off on the related screening instrument. We first checked whether effect sizes differ between studies that had included a control group of children without chronic physical illness and studies without such a control group. Because the effect sizes did not differ between these conditions (Q(1) = .06, p < .80), both groups of studies were combined in the following analyses.

Individuals with pediatric chronic physical illnesses reported higher levels of PTSS than community norms or healthy control groups (g = .50) – a moderate difference (Cohen 1992; Table 1). Egger’s test found significant funnel plot asymmetry (t = 3.29, p < .002). The trim-and-fill procedure indicated that 26 effect sizes would have to be added to produce a symmetrical funnel plot. Adding these effect sizes led to a decrease in size of between-group differences compared to the original analysis of g = .35 (95%-confidence interval [CI] .26 to .43, Z = 7.69, p < .001). Nonetheless, the elevation of PTSS remained statistically significant.

Table 1.

PTSS in individuals with pediatric chronic illness

Group k g 95%-CI Z Q
Total group 177 .50 .41 .58 11.49** 191.59
Type of illness 13.96
 asthma 6 .64 .17 1.12 2.67* 6.56
 burns 15 .74 .43 1.05 4.72** 6.44
 cancer 65 .54 .40 .68 7.69** 55.22
 diabetes 4 1.15 .60 1.70 4.09** 6.04
 epilepsy 7 .44 .03 .85 2.10 6.92
 heart disease 9 .35 −.02 .72 1.88 24.14*
 HIV infection/AIDS 6 .47 −.01 .96 1.92 1.43
 liver diseases 4 .51 −.08 1.11 1.69 4.47
 sickle cell disease 4 .37 −.12 .94 1.25 3.93
 TBI 29 .25 .03 .46 2.28 40.13
 Other/mixed diseases 28 .49 .28 .71 4.46** 20.50
Age at diagnosis of chronic illness .53
  ≤ 6 years 25 .42 .17 .67 3.33* 25.88
 6–12 years 74 .52 .38 .67 7.05** 73.24
  > 12 years 19 .54 .24 .84 3.52** 16.40
Age at assessment of PTSS/PTSD 1.36
  ≤ 6 years 12 .61 .25 .97 3.34** 5.10
 6–12 years 51 .57 .40 .74 6.56** 46.23
  > 12 years 111 .47 .35 .58 8.13** 118.05
Assessment of PTSS/PTSD 1.90
 PTSS 78 .43 .29 .56 6.28** 89.15
 PTSD 99 .55 .43 .67 9.00** 83.99
Representativeness of the sample 2.30
 Yes 24 .32 .07 .56 2.52 11.35
 No/not reported 152 .52 .42 .62 10.42** 155.77
Equivalence of patient and control group .31
 Yes 29 .47 .25 .70 4.13** 26.92
 Not tested 141 .49 .39 .60 9.65** 143.63
 No 7 .61 .18 1.04 2.77* 2.57
Valid and reliable PTSS/PTSD measure 1.09
 Yes 138 .47 .37 .57 9.06** 124.10
 No/limited 39 .58 .40 .77 6.08** 49.01
Sufficient test power 1.94
 Yes 119 .45 .35 .56 8.57** 131.59
 No 58 .61 .43 .77 6.82** 40.85
Publication status .02
 Published 164 .50 .41 .59 10.51** 167.78
 Unpublished 13 .47 .12 .82 2.61* 5.18
Open in a new tab

k = number of effect sizes that were included in the analysis, g = weighted effect size (positive scores indicate higher PTSS in individuals with pediatric chronic illness); Z = test for significance. 95%-CI = lower and upper limits of 95%-confidence interval; Q = test for homogeneity of effect sizes. *p < .01, **p < .001

On average, 11.5% of the participants with pediatric chronic physical illnesses fulfilled the criteria of PTSD. The analysis of the ORs indicated that individuals diagnosed with chronic physical illnesses in childhood or adolescence had a 2.70-fold higher chance of meeting the criteria of PTSD CI 2.20 to 3.31; Z = 9.51, p < .001; Q = 117.16 n.s., based on 102 effect sizes).

Illness-specific analyses were computed if at least three effect sizes were available for individual illnesses. The Q-test indicates that the levels of PTSS did not vary between the compared illnesses. Significant elevations of PTSS were observed in individuals with pediatric diabetes (g = 1.15), burns (g = .74), asthma (g = .64), and cancer (g = .55). Elevations of PTSS did not reach statistical significance in the case of HIV-infection/AIDS (g = .47), possibly due to the low number of available effect sizes. Rates of PTSD were highest in individuals with diabetes (28.3%, based on 1 study), followed by pediatric asthma (22.7%, based on 5 studies), HIV-infection/AIDS (17.3%, based on 3 studies), burns (16.6%, based on 14 studies), liver disease (15.8%, based on 2 studies), cancer (13.8%, based on 27 studies), sickle cell disease (11.2%, based on 3 studies), heart disease (6.9%, based on 6 studies), TBI (5.9%, based on 17 studies), and epilepsy (5.5%, based on 6 studies). Thus, we found no evidence for higher elevations of PTSS in individuals diagnosed with life-threatening illnesses as compared to other illnesses.

The effect sizes did not differ between studies with preschoolers and studies with older children or adolescents. Studies that used continuous PTSS measures had effect sizes similar to those with categorical PTSD measures. In addition, the assessed criteria of study quality were not related to the size of the observed between-group differences.

Correlates of PTSS in Individuals Diagnosed with Chronic Physical Illnesses in Childhood or Adolescence

With regard to the third research question, we found that those individuals who were older when they were diagnosed with chronic physical diseases reported higher levels of PTSS (Table 2). In contrast, levels of PTSS did not vary by age at assessment of PTSS. However, a longer time since the diagnosis of the physical illness and a longer time since active treatment were both related to lower levels of PTSS.

Table 2.

Correlates of PTSS in individuals with a pediatric chronic illness

Correlate k r 95%-CI Z Q
Age at diagnosis of the chronic illness 21 .16 .05 .26 2.79* 20.83
Age at assessment of PTSS 30 .01 −.05 .06 .18 30.20
Time since being diagnosed 30 −.09 −.14 −.04 −3.50** 30.63
Time off treatment 11 −.18 −.26 −.09 −3.99** 11.63
Illness severity 27 .25 .18 .31 7.13** 27.52
Treatment duration/intensity 10 .26 .19 .33 6.94** 14.07
Adherence 4 −.51 −.71 −.24 −3.37** 13.97*
Social support 4 −.20 −.39 −.00 −1.99 13.46*
Family functioning 6 −.25 −.36 −.13 −4.33** 5.49
Open in a new tab

k = number of effect sizes that were included in the analysis, r = weighted effect size; Z = test for significance of r. 95% CI = lower and upper limits of 95%-confidence interval; Q = test for homogeneity of effect sizes. *p < .01, **p < .001

Higher illness severity and higher treatment intensity and/or duration were associated with higher PTSS. In addition, there was a strong negative association between PTSS and treatment adherence. Finally, higher levels of perceived family functioning were related to lower levels of PTSS. According to Cohen (1992), these associations can be interpreted as small (age at diagnosis, time since last treatment, illness intensity, treatment intensity/duration, family functioning) or very small (time since diagnosis). Given the limited number of effect sizes on correlates of PTSS, and the fact that almost all effect sizes were homogeneous, we did not compute moderator analyses.

Discussion

The present meta-analysis found small to moderate elevations of PTSS in individuals who were diagnosed with chronic physical illnesses in childhood or adolescence. On average, 11.5% of these persons fulfilled the criteria for PTSD used in the individual studies. Elevated PTSS were observed in the cases of asthma, burns, cancer, and diabetes. Egger’s test and trim-and-fill analysis indicated that there may be a possible publication bias. In addition, children who were older at diagnosis reported higher PTSS, while a longer duration since the initial diagnosis and longer time since completion of treatment were associated with a small decline in PTSS. Furthermore, higher severity of the physical illness and higher treatment intensity/longer duration were related to higher PTSS. Among individuals still in treatment, PTSS were associated with lower adherence. Finally, higher levels of perceived family functioning were related to lower PTSS.

The observed overall rate of PTSD in individuals diagnosed with chronic physical illnesses in childhood or adolescence was similar to the rate reported in a previous meta-analysis (Kahana et al. 2006). Our data indicate small to moderate elevations of PTSS in individuals with pediatric chronic physical illnesses, with about 3.0% (after adjustment for possible file-drawer problem) or 5.9% (without such adjustment) of the interindividual variance in present PTSS being explained by the pediatric physical illness (Lipsey and Wilson 2001). Thus, most patients with pediatric chronic physical illnesses do not develop moderate or even high levels of PTSS or PTSD (Tillery et al. 2016), or have overcome many symptoms by the time of assessment. One reason for not developing PTSS or PTSD could be the fact that many chronic illnesses and their therapies are not considered to be traumatic events, in a narrow sense (APA 2013). However, we did not find stronger elevations of PTSS in individuals with (potentially) fatal diseases – such as cancer and HIV infection – than in the cases of nonfatal illnesses if well-treated (asthma, diabetes). As a second possible explanation, some young patients may not have been aware of potentially fatal consequences of their illnesses and may, therefore, not have perceived their illnesses as traumatic events. For example, Tillery et al. (2016) reported that 47% of survivors of pediatric cancer did not perceive their illnesses as being their most traumatic event. Many parents, particularly those of younger children, tend to protect their children from negative information about their illnesses (Jalmsell et al. 2015). Third, because vulnerabilities and protective factors play a role in the development of PTSS -- such as prior stressful life events, family functioning, and social support (Kazak et al. 2004) -- our data indicate that most young people with chronic physical illnesses do not have severe vulnerabilities to PTSS/PTSD and/or have sufficient protective factors.

The slightly higher rates of PTSS/PTSD observed in patients with diabetes and asthma as compared to those with illnesses with clear elevations of mortality (cancer, HIV-infection/AIDS) should not be over-interpreted as the differences between illnesses were not statistically significant. Nonetheless, the results indicate that illnesses which are usually nonfatal can be associated with illness-related traumatic experiences, for example, if an asthma attack caused a fear of dying or if children are taught that nonadherence to their diabetic regime could lead to death.

Results tended to be similar when using categorical measures of PTSD and continuous measures of PTSS, and with regard to the assessed characteristics of study quality as well. Our data indicate that the results are robust in terms of these study characteristics. The trim-and-fill analysis found some evidence for a possible file drawer problem. The possibility of file drawer problems and moderator effects of study quality had not been addressed in the previous meta-analysis by Kahana et al. (2006).

The present meta-analysis identified correlates of PTSS. Lower rates of PTSS in individuals diagnosed at a younger age could indicate that younger children are less aware of traumatic characteristics of their diseases (e.g., mortality risk; Jalmsell et al. 2015) and that the studies had not yet adjusted the criteria of assessing PTSS to the age of younger children (APA 2013). In contrast, age at assessment of PTSS did not play a role when analyzing relative elevations of PTSS compared to healthy peers or community norms, because cognitive limitations and problems with assessing PTSS/PTSD in young children also reduce PTSS/PTSD in the control group.

In contrast to Bruce (2006), but in line with Kahana et al. (2006), we found some evidence for a decline in PTSS when the duration of the illness and amount of time since treatment increased. The narrative review by Bruce (2006) probably could not identify this effect because small to very small correlations are only statistically significant if individual studies have sufficient statistical power.

While the results of previous reviews have been inconclusive on whether greater illness severity would be related to more PTSS, the present meta-analysis found clear evidence for such an association. The same was true for associations with treatment duration/intensity, thus indicating that more severe illnesses, as well as longer and more intense treatments, have more traumatic features that promote PTSS (Barakat et al. 1997; Meentken et al. 2017; Price et al. 2016). The strongest correlation was observed between PTSS and treatment adherence, indicating that avoidance as a core component of PTSS inhibits behavior that reminds patients of their traumatic illness- or treatment-related experiences (such as seeking follow-up care; Kerkar et al. 2006), or that general psychological distress associated with PTSS interferes with adherent behavior.

While Kahana et al. (2006) found a surprisingly positive association between PTSD and social support, the present meta-analysis could not replicate this finding. More positive family relations were even related to fewer PTSS. This result indicates that positive relations within the family protect young people with chronic illnesses from developing PTSS, or that more PTSS negatively affect social relations with family members. In fact, a longitudinal study, not specific to pediatric chronic illnesses, found that PTSS negatively affects the social support system (Kaniasty and Norris 2008).

Limitations and Conclusions

Some limitations of the present study must be mentioned. First, limited numbers of studies were available on some diseases, such as sickle cell disease, diabetes, and liver disease, as well as some correlates of PTSS (e.g., adherence and social support). For individual diseases, we could not compute analyses on subgroups of patients (such as cancer patients who were successfully treated versus those who relapsed) because separate information on such subgroups was missing in most individual studies. Second, some studies on PTSD used screening instruments rather than structured clinical interviews, which are the gold standard of clinical assessment. Nonetheless, studies with well-validated measures reported similar results to other studies, thus indicating that this source of bias did not affect our main results. Third, most available studies did not compare levels of PTSS in young people with pediatric chronic illnesses with a socio-demographically matched healthy control group. Thus, general population norms or healthy control groups from other studies had to be used. A few studies had to be excluded because we could not identify a control group. Fourth, because avoidance of reminders of the traumatic event could lead to refusal of participation in psychological studies, available studies may have underestimated the prevalence of PTSD. Nonetheless, the trim-and-fill procedure indicates that we might have missed some studies with small elevations of PTSS rather than those with high elevations. Fifth, we addressed only a limited number of correlates of PTSS. For example, valid data on mental health from the time before being diagnosed with a chronic physical disease are usually not available if the psychological assessment takes place years after being diagnosed with the physical illness. We also did not include parental variables (such as levels of parental PTSS) because they will be addressed in a separate meta-analysis. Finally, we could not test causal relations between having a chronic physical disease and PTSS/PTSD. Although there are good arguments for an effect of chronic physical illnesses on the development of PTSS/PTSD, and individuals with PTSD often refer to their chronic illnesses as traumatic (Tillery et al. 2016), PTSD could also affect the development of some illnesses, mediated by hormonal changes associated with chronic stress or reduced natural killer cell activity (Boscarino 2004).

Despite these limitations, several conclusions can be drawn from the present meta-analysis. Because PTSD and PTSS are more prevalent in individuals diagnosed with chronic physical diseases in childhood or adolescence than in their healthy peers, individuals who experienced a traumatizing event in the context of their illness should be screened for PTSS and PTSD. The present results indicate that the need to screen for PTSS/PTSD does not only apply to diseases associated with a threat to mortality (based on the PTSD criteria in the DSM-5; APA 2013), but also to other chronic diseases, such as diabetes or asthma. Because many commonly used broadband scales that measure emotional and behavioral problems in the field of pediatric psychology – such as the Child Behavior Checklist and the Youth Self-Report (Achenbach 1991) – do not assess PTSS, assessments should include a measure of PTSS/PTSD.

As prevalence rates of PTSS did not differ between potentially fatal and nonfatal diseases, it must be questioned whether narrowing the DSM-5 criteria to the confrontation with actual or threatened death was the best choice. Young distressed people with nonfatal physical diseases may no longer receive psychotherapy if the PTSD diagnoses are made based on the DSM-5.

Interventions are needed for chronically ill young people with PTSD and elevated levels of PTSS. A meta-analysis on treatment of PTSD in children and adolescents found strong average improvements of PTSS, with trauma-focused cognitive behavior therapy showing the strongest effects (Morina et al. 2015). Intervention effects on PTSS will be less prominent for individuals in the subclinical range of PTSS, and when psychological interventions do not directly target PTSS. For example, Sahler et al. (2005) found only small declines of PTSS among young cancer patients who participated in a problem-solving intervention.

With regard to future research needs, more studies are recommended on PTSS/PTSD among those with rarely studied chronic diseases, such as epilepsy or HIV-infection. Given the fact that most young people with pediatric chronic physical illnesses do not meet the criteria of PTSD, more research is recommended on factors that protect them against PTSS/PTSD. Because we found some evidence of a possible publication bias, the publication of non-significant results (indicating that some patients do not show elevated PTSS) should also be encouraged.

Electronic Supplementary Material

ESM 1 (83.3KB, docx)

(DOCX 83 kb)

Compliance with Ethical Standards

Conflict of Interest

The author states that there is no conflict of interest.

References

  1. Achenbach TM. Integrative guide for the 1991 CBCL/4_18, YSR, and TRF profiles. Burlington: Department of Psychiatry, University of Vermont; 1991. [Google Scholar]
  2. American Psychiatric Association . Diagnostic and statistical manual of mental disorders. 4. Washington, DC: Author; 1994. [Google Scholar]
  3. American Psychiatric Association . Diagnostic and statistical manual of mental disorders. 5. Washington, DC: Author; 2013. [Google Scholar]
  4. Anderson TM, Knight RG. The long-term effects of traumatic brain injury on the coordinative function of the central executive. Journal of Clinical and Experimental Neuropsychology. 2010;32:1074–1082. doi: 10.1080/13803391003733560. [DOI] [PubMed] [Google Scholar]
  5. Barakat LP, Wodka EL. Posttraumatic stress symptoms in college students with a chronic illness. Social Behavior and Personality. 2006;34:999–1006. doi: 10.2224/sbp.2006.34.8.999. [DOI] [Google Scholar]
  6. Barakat LP, Kazak AE, Meadows AT, Casey R, Meeske K, Stuber ML. Families surviving childhood cancer: A comparison of posttraumatic stress symptoms with families of healthy children. Journal of Pediatric Psychology. 1997;6:843–859. doi: 10.1093/jpepsy/22.6.843. [DOI] [PubMed] [Google Scholar]
  7. Borenstein M, Higgins L, Hedges J, Rothstein H. Comprehensive meta-analysis, version 2. Englewood: Biostat; 2005. [Google Scholar]
  8. Boscarino JA. Posttraumatic stress disorder and physical illness: Results from clinical and epidemiologic studies. In: Yehuda R, McEwen B, editors. Biobehavioral stress response: Protective and damaging effects. New York: New York Academy of Sciences; 2004. pp. 141–153. [DOI] [PubMed] [Google Scholar]
  9. Bruce M. A systematic and conceptual review of posttraumatic stress in childhood cancer survivors and their parents. Clinical Psychology Review. 2006;26:233–256. doi: 10.1016/j.cpr.2005.10.002. [DOI] [PubMed] [Google Scholar]
  10. Buckley TC, Green BL, Schnurr PP. Trauma, PTSD and physical health: Clinical issues. In: Wilson JP, Keane TM, editors. Assessing psychological trauma and PTSD. 2. New York: Guilford; 2004. pp. 441–465. [Google Scholar]
  11. Cohen J. A power primer. Psychological Bulletin. 1992;112:155–159. doi: 10.1037/0033-2909.112.1.155. [DOI] [PubMed] [Google Scholar]
  12. Duval S, Tweedie R. Trim and fill: A simple funnel-plot-based method of testing and adjusting for publication bias in meta-analysis. Biometrics. 2000;56:455–463. doi: 10.1111/j.0006-341X.2000.00455.x. [DOI] [PubMed] [Google Scholar]
  13. Ferro MA, Speechley KM. Depressive symptoms among mothers of children with epilepsy: A review of prevalence, associated factors, and impact on children. Epilepsia. 2009;50:2344–2354. doi: 10.1111/j.1528-1167.2009.02276.x. [DOI] [PubMed] [Google Scholar]
  14. Frederick C, Pynoos R, Nader K. Childhood posttraumatic stress reaction index—a copyrighted instrument. Los Angeles: University of California; 1992. [Google Scholar]
  15. Graziano S, Rossi A, Spano B, Petrocchi M, Biondi G, Ammaniti M. Comparison of psychological functioning in children and their mothers living through a life-threatening and non-life-threatening chronic disease: A pilot study. Journal of Child Health Care. 2016;20:174–184. doi: 10.1177/1367493514563854. [DOI] [PubMed] [Google Scholar]
  16. Higgins, P. P., & Green, S. (2009). Cochrane handbook for systematic reviews of interventions. Version 5.02. Available online at http://handbook.cochrane.org/v5.0.2/ (retrieved 2017-05-23).
  17. Jalmsell L, Kontio T, Stein M, Henter JI, Kreicbergs U. On the child's own initiative: Parents communicate with their dying child about death. Death Studies. 2015;39:111–117. doi: 10.1080/07481187.2014.913086. [DOI] [PubMed] [Google Scholar]
  18. Kahana S, Feeny NC, Youngstrom ER, Drotar D. Posttraumatic stress in youth experiencing illnesses and injuries: An exploratory meta-analysis. Traumatology. 2006;12:148–161. doi: 10.1177/1534765606294562. [DOI] [Google Scholar]
  19. Kaniasty K, Norris FH. Longitudinal linkages between perceived social support and posttraumatic stress symptoms: Sequential roles of social causation and social selection. Journal of Traumatic Stress. 2008;21:274–281. doi: 10.1002/jts.20334. [DOI] [PubMed] [Google Scholar]
  20. Kaufman J, Birmaher B, Brent D, Rao U, Flynn C, Moreci P, et al. Schedule for affective disorders and schizophrenia for school-age children-present and lifetime version (K-SADS-PL): Initial reliability and validity data. Journal of the American Academy of Child and Adolescent Psychiatry. 1997;36:980–988. doi: 10.1097/00004583-199707000-00021. [DOI] [PubMed] [Google Scholar]
  21. Kazak AE, Barakat LP, Meeske K, Christakis D, Meadows AT, Casey R, Penati B, Stuber ML. Post-traumatic stress, family functioning, and social support in survivors of childhood leukemia and their mothers and fathers. Journal of Consulting and Clinical Psychology. 1997;65:120–129. doi: 10.1037/0022-006X.65.1.120. [DOI] [PubMed] [Google Scholar]
  22. Kazak AE, Alderfer MA, Streisand R, Simms S, Rourke MT, Barakat LP, et al. Treatment of posttraumatic stress symptoms in adolescent survivors of childhood cancer and their families: A randomized clinical trial. Journal of Family Psychology. 2004;18:493–504. doi: 10.1037/0893-3200.18.3.493. [DOI] [PubMed] [Google Scholar]
  23. Kerkar N, Annunziato RA, Foley L, Schmeidler J, Rumbo C, Emre S, Shemesh E. Prospective analysis of nonadherence in autoimmune hepatitis: A common problem. Journal of Pediatric Gastroenterology and Nutrition. 2006;43:629–634. doi: 10.1097/01.mpg.0000239735.87111. [DOI] [PubMed] [Google Scholar]
  24. Kilpatrick DG, Ruggiero KJ, Acierno R, Saunders BE, Resnick HS, Best CL. Violence and risk of PTSD, major depression, substance abuse/dependence, and comorbidity: Results from the National Survey of adolescents. Journal of Consulting and Clinical Psychology. 2003;71:692–700. doi: 10.1037/0022-006X.71.4.692. [DOI] [PubMed] [Google Scholar]
  25. La Greca A. Manual for the self care inventory. Coral Gables: Author; 2004. [Google Scholar]
  26. Landolt MA, Vollrath M, Ribi K, Gnehm HE, Sennhauser FH. Incidence and associations of parental and child posttraumatic stress symptoms in pediatric patients. Journal of Child Psychology and Psychiatry. 2003;44:1199–1207. doi: 10.1111/1469-7610.00201. [DOI] [PubMed] [Google Scholar]
  27. Le Brocque RM, Hendrikz J, Kenardy JA. The course of posttraumatic stress in children: Examination of recovery trajectories following traumatic injury. Journal of Pediatric Psychology. 2010;35:637–645. doi: 10.1093/jpepsy/jsp050. [DOI] [PubMed] [Google Scholar]
  28. Lipsey MW, Wilson DB. Practical meta-analysis. Thousand Oaks: Sage; 2001. [Google Scholar]
  29. Meentken MG, van Beynum IM, Lagerstee JS, Helbing WA, Utens WJ. Medically related post-traumatic stress in children and adolescents with congenital heart defects. Frontiers in Pediatrics. 2017;5:20. doi: 10.3389/fped.2017.00020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Moos R, Moos B. Family environment scale manual: Development, applications, research. 3. Palo Alto: Consulting Psychologist Press; 1994. [Google Scholar]
  31. Morina N, Koerssen R, Pollet TV. Interventions for children and adolescents with posttraumatic stress disorder: A meta-analysis of comparative outcome studies. Clinical Psychology Review. 2015;47:41–54. doi: 10.1016/j.cpr.2016.05.006. [DOI] [PubMed] [Google Scholar]
  32. O’Connor SS, Zatzick DF, Wang J, Temkin N, Koepsell TD, Jaffe KM, et al. Association between posttraumatic stress, depression, and functional impairments in adolescents 24 months after traumatic brain injury. Journal of Traumatic Stress. 2012;25:264–271. doi: 10.1002/jts.21704. [DOI] [PubMed] [Google Scholar]
  33. Phipps S, Klosky JL, Long A, Hudson MM, Huang Q, Zhang H, Noll RB. Posttraumatic stress and psychological growth in children with cancer: Has the traumatic impact of cancer been overestimated? Journal of Clinical Oncology. 2014;32:641–646. doi: 10.1200/JCO.2013.49.8212. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Price J, Kassam-Adams N, Alderfer MA, Christofferson J, Kazak AE. A reevaluation and update of the integrative (trajectory) model of pediatric medical traumatic stress. Journal of Pediatric Psychology. 2016;41:86–97. doi: 10.1093/jpepsy/jsv074. [DOI] [PubMed] [Google Scholar]
  35. Rodriguez N, Steinberg A, Pynoos RS. UCLA post-traumatic stress disorder reaction index for DSM-IV, child, adolescent, and parent versions. Los Angeles: University of California at Los Angeles; 1998. [Google Scholar]
  36. Sahler OJ, Fairclough DL, Phipps S, Mulhern RK, Dolgin MJ, Noll RB, et al. Using problem-solving skills training to reduce negative affectivity in mothers of children with newly diagnosed cancer: Report of a multisite randomized trial. Journal of Consulting and Clinical Psychology. 2005;73:272–283. doi: 10.1037/0022-006X.73.2.272. [DOI] [PubMed] [Google Scholar]
  37. Thompson, R. J., & Gustafson, K. E. (1996). Adaptation to chronic childhood illness. Washington, DC: American Psychological Association.
  38. Tillery R, Sharp KMH, Okado Y, Long A, Phipps S. Profiles of resilience and growth in youth with cancer and healthy comparisons. Journal of Pediatric Psychology. 2016;41:290–297. doi: 10.1093/jpepsy/jsv091. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Weiss DS, Marmar CR. The impact of events scale—Revised. In: Wilson JP, Keane TM, editors. Assessing psychological trauma and PTSD. New York: Guilford; 1997. pp. 399–411. [Google Scholar]
  40. Werba BE, Hobbie W, Kazak AE, Ittenbach RF, Reilly AF, Meadows AT. Classifying the intensity of pediatric cancer treatment protocols: The intensity of treatment rating scale 2.0 (ITR-2) Pediatric Blood and Cancer. 2007;48:673–677. doi: 10.1002/pbc.21184. [DOI] [PubMed] [Google Scholar]
  41. Yalug I, Tufan AE, Doksat K, Yalug K. Post-traumatic stress disorder and post-traumatic stress symptoms in parents of children with cancer: A review. Neurology, Psychiatry and Brain Research. 2011;17:27–31. doi: 10.1016/j.npbr.2011.02.007. [DOI] [Google Scholar]
  42. Zimet Gregory D., Dahlem Nancy W., Zimet Sara G., Farley Gordon K. The Multidimensional Scale of Perceived Social Support. Journal of Personality Assessment. 1988;52(1):30–41. doi: 10.1207/s15327752jpa5201_2. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ESM 1 (83.3KB, docx)

(DOCX 83 kb)

Articles from Journal of Child & Adolescent Trauma are provided here courtesy of Springer

RESOURCES