Study methodologies

Of the 23 studies reviewed, five studies focused on adolescent cancer survivors,18–22 ten focused on adult (18 years and older) survivors of pediatric or AYA cancers,16–17,23–30 and eight focused on both.31–38 The reported ages of participating survivors ranged from 7 years21 to 71 years.29

Continue Reading

With regard to the heterogeneity of diagnoses, the majority of studies examined survivors of all cancers, with seven studies excluding central nervous system malignancies,17,19,22,25,31,34,38 and only one focusing on Hodgkin’s lymphoma survivors.29

The majority (number [n] =14) of studies examined both PTSS and PTSD outcomes; five assessed PTSD alone16,18,25,36,37 and four evaluated PTSS alone.22,27,34,38

Sample sizes ranged from 23–6,542, with ten samples of less than 100,17,19,23–25,30–32,36 eleven samples between 100 and 350,18,20,22,26,29,33–35,37,38 and two samples exceeding 6,500.16,28

Most studies employed a cross-sectional design, with only four studies using a longitudinal design.17,35,37,38 Only nine studies included a comparison or control group.16,17,21,28,29,31,34,36,37 Seven self-report measures, including the Adolescent Dissociative Experiences Scale-II (A-DES),39 Posttraumatic Stress Diagnostic Scale (PDS),40 University of California at Los Angeles Posttraumatic Stress Disorder Index (PTSDI),41 Posttraumatic Stress Checklist–Civilian Version (PCL-C),42

Child Posttraumatic Stress Reaction Index – Revision 2 (CPTS-RI),43 Trauma Symptom Checklist for Children – Alternate version (TSCC-A),44 Posttraumatic Stress Disorder Reaction Index (PTSD-RI),41 and one interview (Impact of Traumatic Stressors Interview Schedule [ITSIS])33 were used to measure PTSS.

Three of these measures (PDS, PTSD-RI, and PTSDI) were also used to indicate “likely” cases of PTSD when scores fell in the clinically significant range.

Five assessment instruments, all structured and semistructured interviews, were used to assess PTSD: Kiddie-Schedule for Affective Disorders and Schizophrenia–Epidemiologic Version–5 (K-SADS-E-5);45 Structured Clinical Interview for DSM-IV (SCID);46 Clinician-Administered PTSD Scale for Children and Adolescents (CAPS-CA);47 Structured Interview for PTSD (SI-PTSD);48 and Diagnostic Interview Schedule for Children (DISC 2.3).

Psychological distress and functional impairment were assessed using the Brief Symptom Inventory – 18 (BSI-18);50 Impact of Event Scale (IES);51 Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36);52 Short-Form Health Status questionnaire (SF-12);53 and Global Assessment of Functioning scale (GAF).54

The studies reviewed in this paper often identified the temporal relation between diagnosis/treatment and assessment of PTSD/PTSS (eg, how much time elapsed since diagnosis or treatment at time of assessment) rather than explicating whether PTSD/PTSS was linked to the cancer experience.

At least half (n=12) of the studies did not address whether or not PTSD symptoms were related to cancer, while three reported the number of cases in which the self-reported traumatic event related to the survivor’s cancer experience,17,29,36 four examined cancer-induced PTSD by having subjects respond to items on the basis of their cancer experience or adapting items to reflect the cancer experience (eg, by replacing the word “trauma” with “cancer experience”),18,21,31,33 three studies assumed that all participants were exposed to a traumatic event because the cancer diagnosis and treatment would constitute a traumatic event,16,32,35 and one study measured PTSS 2 months after a clinic visit to limit health factors from acting as traumatic stress triggers.38

Given the different methodological approaches used toward focusing survivors on cancer as a precipitating traumatic event,55 it is important that cancer-induced PTSD prevalence estimates are interpreted with this in mind.

As with most trauma research, it is important to consider the extent to which the participation of potentially eligible survivors affects prevalence estimates. Based on 21 studies reporting participation, survivor participation ranged from 38%–89%.17,25

Research procedures requiring participants to engage in emotionally burdening interviews may cause eligible survivors to resist revisiting potentially traumatic life events and may explain nonparticipation or attrition. 

Nonparticipation introduces potential selection bias and may influence the precision of prevalence estimates, the magnitude of observed risk, and the generalizability of study findings.