New progress in an old debate? Applying the DSM‐5 criteria to assess traumatic events and stressor‐related disorders in cancer survivors

The rather broad definition of medical trauma within DSM‐IV has contributed to long‐lasting debates on the applicability of Posttraumatic Stress Disorder (PTSD) in oncological patients and its differentiation from Adjustment Disorder (AjD) which results from non‐traumatic critical life events. The DSM‐5 criteria have introduced a narrower definition of medical traumatization. However, studies on updated prevalence rates in cancer patients are missing.


| BACKGROUND
Starting at the early stage of detecting symptoms and receiving the diagnosis, 1 cancer and its treatment encompass various stressors which may be potentially traumatic. 2,3Consequently, Posttraumatic Stress Disorder (PTSD) has often been considered a relevant issue in cancer survivors. 3[6] Nevertheless, the question whether cancer-related experiences are traumatic events leading to PTSD has always been controversially discussed. 3Debates started after the Diagnostic and Statistical Manual of Mental Disorders -Fourth Edition (DSM-IV) had introduced life-threatening diseases such as cancer as potentially traumatic event (A1 criterion). 7Alter and colleagues were the first to investigate the traumatic character of cancer in a small set of cancer survivors in the PTSD Field Trials, 8 supporting proposed changes to the DSM-IV.Even though research confirmed the general existence of cancer-related PTSD, 1,[9][10][11][12] the rather broad definition of medical trauma within DSM-IV enabled any aspect relating to the cancer disease to potentially trigger PTSD. 1,9Thus, even the mere anticipation of events such as waiting for test results or aftercare checkups were considered potentially traumatic.However, conceptual issues on the applicability of cancer-related PTSD have not been resolved to date. 3 Meanwhile, the DSM-5 has modified the requirements for medical trauma. 13Life-threatening diseases are no longer considered necessarily traumatic, and medical traumata need to be "sudden, catastrophic events (e.g.waking during surgery, anaphylactic shock)". 14Whether cancer diagnosis disclosure can be classified as potentially traumatic according to DSM-5 is debatable, as it is usually a predictable situation following a phase of diagnostic procedures, but is experienced by some patients as sudden and catastrophic.
Adjustment disorder (AjD) is now considered the major psychological response to medical diseases 14 and can be diagnosed, if the full symptomatology of PTSD is met, but not the trauma criterion. 14Information on differential-diagnostic implications of these modifications within the cancer setting is very sparse, currently limited to one study among young adult cancer patients, 15 but is needed for developing tailored psychological interventions 12,13 and enabling effective oncological care.
Among oncological populations, hematological cancer survivors are particularly exposed to stressful events.Some hematological cancer types require intensive treatment, [16][17][18] others are chronic and being treated with oral therapies. 19A subgroup is treated with autologous or allogeneic stem cell transplantation (SCT), 20 which requires spatial isolation 21 and bears the risk of life-threatening complications, 20 such as Graft-versus-Host-Disease in case of allogeneic SCT. 22Despite the variety of stressors within this group, relevant findings are highly inconsistent and methodologically limited, with most studies being based on small samples and selfreport questionnaires. 23ven the gaps in existing knowledge outlined above, we conducted a study with hematological cancer survivors using the Structured Clinical Interview for DSM-5.The research objectives were to (i) gain information on the prevalence of cancer-related stressors, (ii) examine the degree to which these stressors meet trauma criteria according to DSM-5 and (iii) provide updated prevalence rates of PTSD, AjD and cancer-related PTSD via clinical interview data.

| Design and participants
Data for this cross-sectional study were collected between April hematopoietic and related tissue (ICD-10: D47) or myelodysplastic syndrome (ICD-10: D46), (ii) were receiving/had received oncological treatment, (iii) were aged ≥18 years at recruitment, (iv) were ≤70 years at recruitment as this was considered the maximum age for SCT, (v) were able to provide written informed consent, (vi) were fluent in German and (vii) had no medical reasons for re-admission into the inpatient setting.
Patients were assessed at a minimum of 6 weeks after the end of the oncological treatment to ensure that the exposure to treatmentassociated stressors had ended.For patients with SCT, we additionally applied a maximum time of 2 years since undergoing SCT in order to capture relevant treatment-specific stressors.As an exception, patients receiving permanent treatment, such as tyrosine kinase inhibitors, could be assessed directly after study inclusion.Participants received a paper-pencil form with validated questionnaires regarding distress 24 and were assessed though a structured clinical interview.Interviews were conducted either in person or by phone.
Eligible patients who declined to participate were asked to provide their reason for non-participation.Patients were reminded regularly, when the questionnaire was not returned, depending on their respective clinical status, for example, hospitalization.
All participants provided written informed consent.Participants who completed the study received 20 Euros.The study was approved by the ethics committee of the Medical Faculty at the University of Leipzig (case number: 447/17-ek).Deviations from the study protocol 23 not mentioned in the manuscript are provided in the supplement (Table S1).

| Clinical interview
Participants were assessed with the Structured Clinical Interview for DSM-5 -Clinician Version (SCID-5-CV). 25As primary outcomes, we assessed PTSD and AjD.We also assessed disorders that have a substantial symptom overlap with PTSD and AjD, that is, major depression, persistent depressive episode, panic disorder, agoraphobia, specific phobia, and generalized anxiety disorder.These disorders were solely assessed for differential-diagnostic purpose and to prevent a bias towards AjD (see respective section below).
For our study, the lifetime trauma history checklist within SCID-5 was extended with a cancer-specific item.Having assessed the traumatic events within the official trauma list, we additionally asked patients "Did you experience any cancer-related events which you consider as traumatic or particularly distressing?"If patients confirmed, they were asked to further describe the number and type of event(s).
Each of these cancer-related stressors was checked for meeting the DSM-5 criteria for medical traumata, that is, whether the stressor was (i) a discrete event in the past in order to clearly differentiate PTSD-intrusions from future-oriented ruminations, (ii) of sudden and catastrophic character and (iii) is accompanied by actual or threatened death or serious injury. 14PTSD symptoms were assessed for the event rated as worst, and a PTSD diagnosis required meeting both trauma and symptom criteria. 25e diagnostic procedure for AjD requires to assess symptoms related to a stressor identified during the course of the clinical interview.Since cancer was the most prominent stressor for most interviewed patients, AjD symptoms were by default assessed in relation to cancer, unless otherwise reported.According to DSM-5, AjD can also be diagnosed if the symptom criteria for PTSD are met, but not the trauma criterion. 14Nevertheless, AjD can only be diagnosed if no other DSM-5 disorder assessed within the interview has occurred within the previous 6 months.

| Interview training and quality assurance
Initial training was conducted by a psychotherapist experienced with SCID-assessment.Concerns regarding updated criteria were resolved with the help of one of the translators of the German SCID-5.Differential-diagnostic issues were discussed within the team.
After the initial raining, all interviewers that were involved in data collection underwent a standard training.This encompassed (i) introduction to the SCID-5 by an experienced interviewer including a case study, (ii) work shadowing during at least two interviews and (iii) two supervised interviews.

| Sociodemographic and medical information
Age, gender, type of diagnosis, time since first diagnosis, treatment with SCT (yes/no) and type of SCT (allogeneic/autologous) were obtained from the medical records.Other characteristics were obtained via self-report.

| Statistical analyses
We ran descriptive statistics for patient characteristics and reasons for non-participation.Among the patients filtered in the outpatient clinic, we were able to conduct responder and non-responder-analyses on age, time since diagnosis (Mann-Whitney U-tests) as well as gender, type of diagnosis and treatment with SCT (Chi-square tests).
We examined the type and frequency of the cancer-related stressors and the traumatic events included on the original SCID-5 trauma list.Cancer-related stressors were analyzed in as much detail as possible (e.g., if patients were distressed by loss of hair during chemotherapy, loss of hair and chemotherapy were handled as separate events).
We subsequently looked at the proportion of cancer-related stressors meeting the criterion for medical traumata according to DSM-5.The categorization of cancer-related stressors and evaluation of their traumatic character was initially undertaken by PE and then reviewed by FS.
As a next step, we calculated the prevalence rates of PTSD, AjD and cancer-related PTSD.Finally, we identified the number of cases that met all symptoms for PTSD, but not the trauma criteria, and assessed whether these patients were to be diagnosed with AjD, if no other DSM-5 disorder was present.
Tests for group comparisons were two-sided, with a significance level alpha of 0.05.All analyses were performed using IBM SPSS Statistics for Windows, Version 27 (IBM Corporation, Armonk, 2020).

| RESULTS
In total, 291 hematological cancer patients participated in the study (Figure 1).Of the patients filtered in the outpatient clinic, 211 out of 363 eligible patients participated (response rate: 58%).The most frequent reasons for non-participation were no interest and being too burdened.Responders and non-responders did not differ in age (p = 0.10), sex (p = 0.11), type of diagnosis (p = 0.12), time since diagnosis (p = 0.40) or treatment with SCT (p = 0.85) (Table 1).
The mean age was 54 years, 60% were male (Table 1).The most frequent diagnoses were myeloid leukemia and multiple myeloma, and more than half of the sample had received some form of SCT.In total, 46% of the patients indicated a clinically relevant level of distress (≥5 Distress Thermometer), which is comparable to other studies, 27 indicating no bias towards less or more distressed patients in our study.285 patients provided interview data and were considered for further analyses.
More than half of the sample (59%) reported at least one cancerrelated stressor, one out of ten reported more than one (Table 2).A total of 213 stressors were reported.The most frequent cancerrelated stressors were the moment of the cancer diagnosis disclosure, physical side effects, treatment complications, and witnessing distressing events affecting other patients.
Out of all reported cancer-related stressors, 8% met criteria for a medical trauma according to DSM-5.Among those, the most frequent medical traumata were experiencing complications such as anaphylactic shock, sepsis or pneumothorax and witnessing distressing events including anaphylactic shock or death.When traumatic event criteria were broadened beyond DSM-5 and cancer diagnosis disclosure was also considered as traumatic, 38% of the stressors then met criteria for medical trauma.In addition, more than two thirds of the patients (70%) reported at least one non-cancer related traumatic event from the original SCID-5 trauma list, one fourth had experienced more than one such event.
Five (1.8%) and 15 (5.3%) patients met criteria for current and lifetime PTSD according to DSM-5, respectively.Among all PTSD cases three (1.1%) were cancer-related, of which two had current

| Main findings
In a sample of hematological cancer survivors, more than half reported cancer-related stressors, with the most frequent event being the cancer diagnosis disclosure.Around ten percent of those stressors met requirements for a medical trauma according to DSM-5.About half of the current PTSD cases were cancer-related.Some cases showed all symptoms for cancer-related PTSD but did not meet the DSM-5 trauma criteria.

| Integration into previous research
More than half of the sample reported cancer-related stressors, which is in the mid-range of previous results ranging considerably from 12% (2141 mixed cancer patients) 12 to 92% (127 breast cancer patients). 1This discrepancy might be explained by different approaches to establishing stressors, for example, at which severity a stressor needs to be assessed or whether side effects, such as fears, are defined as stressors.Furthermore, stressors depend on sample specifics, which limits comparability: Our population required intensive treatment, heightening the risk for treatment complications and physical side effects; furthermore, stressors such as spatial isolation are unique to those treated with SCT.Nevertheless, our study fits with the findings of previous research with regard to demonstrating T A B L E 2 Frequency and type of cancer-related stressors and (medical) traumatic events among 285 hematological cancer survivors assessed with SCID-5.

Events according to official trauma list
At least one traumatic event 175 (70) More than one traumatic event 65 (26)   Distribution among all reported traumatic events (n = 271) Witnessing death or severe injury of another person 110 (41) Experiencing a life-threatening situation 56 (21)   Experiencing physical or sexual attack 53 (20)   Witnessing physical or sexual attack 39 (14)   Learning that traumatic event occurred to family member/friend that receiving the diagnosis is the most frequent cancer-related stressor. 1,9,28evious research based on DSM-IV found that 59% 1 and 92% 12 of cancer-related stressors met trauma-criteria, which is considerably higher than in our sample.However, our results are in line with the study among young adult cancer patients that identified fewer SPRINGER ET AL.
-1621 traumatic events when applying DSM-5 criteria compared with DSM-IV. 15This above-mentioned discrepancy in research findings might thus directly result from the modified criteria for medical trauma.
According to DSM-IV, a potentially traumatic event (A1 criterion) needed to be accompanied by an intense feeling of fear, helplessness or horror (A2-criterion) to be traumatic. 7However, the rather broad definition of medical trauma allowed the cancer disease as a whole and thus all inherent aspects to be potentially traumatic, for example, waiting for results or surgery 1,9 or uncertainty about the future. 1 Given the serious threat of cancer, these A1-events usually also met the A2-criterion resulting in high rates of cancer-related traumatic events.The DSM-5 in turn is more concise and targeted in the definition of medical traumata, 14 which may have contributed to a lower percentage of cancer-related traumatic events.
About half of the current PTSD cases in our study were cancerrelated, which largely fits with the findings of other studies showing that cancer-related PTSD occurs as frequently or even more frequently than non-cancer related PTSD. 1,5,9,11In contrast, a large study (n = 2141) revealed a very low prevalence of cancer-related PTSD (0.2%) representing only 9% of the current PTSD cases. 12veral explanations are possible.Even though the assessment of cancer-related stressors was similar in both studies, the aforementioned study assigned only a subsample of patients (67%) to the diagnostic interview, which might have biased their results.In addition, this sample was a mixed cancer sample, within a shorter time since diagnosis which might have decreased rates of PTSD, and the study aimed to capture all types of mental disorders, whereas other studies, including ours focused specifically on (cancer-related) PTSD.
Future research may investigate whether such a focus may unconsciously affect the assessment, for example, by putting more emphasis on asking about stressors and PTSD-related symptoms.
11]29 When considering the cancer diagnosis disclosure as traumatic, which is debatable according to DSM-5 definition, the proportion of traumatic events in the context of cancer increases, nevertheless cancer-related PTSD remains rare compared to DSM-IV studies.
Previous evidence on prevalence rates of PTSD and AjD is mostly based on studies among breast cancer patients, 29,30 limiting comparability with our results.Nevertheless, a few studies provided rates for hematological cancer survivors. 12,31Compared to these, our prevalence of PTSD is higher, 12 whereas our rate of AjD is considerably lower. 27Even though our small sample sizes do not allow for strong conclusions, our data contrasts the theoretical assumption that DSM-5 criteria may result in a decrease in PTSD rates and an increase of AjD rates within the cancer setting. 13ere stressor-related symptoms do not meet the full DSM-5 criteria for PTSD, patients may nevertheless suffer distressing psychological symptoms that require psychological treatment.It may thus be of great value to investigate subthreshold PTSD in future studies, that is, traumatic events not leading to full symptomatology.
This may be particularly prevalent among cancer patients since avoidance of cancer-related stressors is often difficult.High comorbidity with other mental disorders may potentially lead to subthreshold PTSD often going unrecognized and untreated.

| Clinical implications
In light of the ongoing debate about whether cancer can be considered traumatic, our data suggest that diagnosing cancer-related PTSD remains possible within the DSM-5 framework but does not apply to a significant proportion of this population.Even though these results have been found in previous studies based on DSM-IV, 3 we think that the discussion on trauma in cancer may considerably change with DSM-5.Cancer is not necessarily traumatic and only a fraction of stressful events during the disease are considered traumatic and can lead to PTSD.Previous debates focused mostly on the rather broad definition of medical trauma in DSM-IV.The DSM-5 in turn provides a more concise definition: Even if the disease trajectory of cancer may indeed contain specific traumatic events leading to PTSD, it is never the disease as a whole which qualifies as a trauma.
Events such as cancer treatment may therefore be non-traumatic if it proceeds as planned, but may be traumatic if unexpected treatment complications occur with a catastrophic impact on the patient's health.It is therefore important to explore stressful events in more detail to understand their potentially traumatic character.We therefore suggest that clinicians and researchers should further investigate the traumatic character of reported cancer-related stressors in order to advance ongoing discussions on the applicability of PTSD in oncological settings.This may further implicate that a particular traumatic event during the disease is needed.
We also recommend that future studies report PTSD rates among cancer patients always in conjunction with its stressors so that findings can be validated and compared.This would also have therapeutic im- An important finding for physicians is that the most frequently reported cancer-related stressors was the moment of cancer diagnosis.
Therefore, special attention should be paid to communication skills when informing patients about the diagnosis and to effective medical care planning to avoid further psychological distress for the patient.

| Study strengths and limitations
To our knowledge, our study is one of the first that applied DSM-5 criteria of PTSD and AjD in an oncological population.We assessed stressors in a more detailed way than previous studies, enabling a comprehensive picture of stressors and a thorough differentialdiagnostic process.In addition, data were generated through diagnostic interviews that applied stringent diagnostic criteria for differential diagnostic purposes.
However, we note that hematological cancer survivors, including SCT recipients, report specific stressors and thus findings need to be replicated among other cancer populations.Treatment with SCT in hematological cancer may impact the prevalence of PTSD.Unfortunately, the sample size was smaller than originally intended, which did not allow for data stratification by treatment with SCT.However, this research question was investigated via self-report questionnaires and reported in another publication. 26There was no discernible impact of SCT on PTSD and AjD symptomatology.Furthermore, our detailed assessment of cancer-related stressors may have inflated the amount of events reported.This limits comparability with prevalence rates of non-cancer samples.However, this approach was necessary to explore all stressors faced by patients during their disease, which may help to improve our understanding of existing stressors in patients.In addition, some patients remained vague in the description of stressors and thus the proportion of events meeting the trauma criterion could have been underestimated.
Finally, no data on interviewer reliability was available and future studies need to replicate our findings.

| CONCLUSIONS
2019 and September 2021.The study was coordinated by the Department of Medical Psychology and Medical Sociology at the University Medical Center of Leipzig in cooperation with the Clinic for Hematology, Cellular Therapy and Hemostaseology.At the beginning of the study, attending physicians at the Clinic for Hematology asked survivors whether they would be interested in being contacted by the study team; interested patients were subsequently approached by members of the research team.Additionally, patients were recruited from hematological practices or by advertising in publications of patient networks.During the study period, due to the COVID-19 pandemic, recruitment was adapted to minimize personal contact, reduce additional burden on physicians and to ensure comprehensive recruitment: The study team filtered all patients aged between 18 and 70 years who were scheduled for the hematological outpatient clinic on the days of recruitment to be checked for eligibility criteria.All eligible patients were directly approached by mail or phone to be informed about the study.We included patients who (i) were diagnosed with any hematological malignancy (ICD-10: C81-96), other neoplasms of lymphoid, SPRINGER ET AL.

E 1 T A B L E 1
Flow chart.This flow-diagram has been published in a similar form elsewhere26 ; Abbreviations: SCT, stem cell transplantation; † referred by physicians at the study center or practices and self-initiated contact of interested patients; ‡ all patients between 18 and 70 years who were scheduled for the outpatient clinic at the days of recruitment; § no medical information in the records, for example, stem cell donors or patients from other institutions seeking second medical opinion; ¶ already recruited by physicians within first recruitment period (left arm).Sample characteristics of hematological cancer survivors.

A
significant part of hematological cancer survivors reported cancerrelated stressors, but only a fraction qualified as traumatic according to DSM-5.The new criteria still enable a diagnosis of PTSD in the context of cancer but provide a more concise definition of what may constitute a medical trauma.To advance conceptual debates, we propose to report cancer-related PTSD cases in conjunction with their stressors.