| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From the Abramson Cancer Center of the University of Pennsylvania, Philadelphia, PA (S.C.P., J.C.C., A.D.); and the University of Stellenbosch, Cape Town, South Africa (A.K.).
Address reprint requests to Steven C. Palmer, PhD, University of Pennsylvania/HUP, 3400 Spruce St./11 Gates, Philadelphia, PA 19104. E-mail: stpalmer{at}mail.med.upenn.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESACKNOWLEDGMENTSREFERENCES |
|---|
METHODS: We assessed 115 consecutive patients with breast cancer in the waiting room of a large comprehensive cancer center using measures of general distress, posttraumatic stress symptoms, and a semistructured diagnostic interview.
RESULTS: A substantial minority (41%) reported responding to cancer with intense fear, helplessness, or horror (DSM-IV A2 criterion). However, cancer-related PTSD was uncommon (4%), and meeting the A2 criterion was a poor indicator of PTSD. Psychological distress was common (38%) and was strongly associated with A2, but was a poor predictor of PSTD.
CONCLUSIONS: Although an intense negative emotional reaction to breast cancer was common, PTSD had low prevalence. Results suggest that using a trauma framework to understand the experience of most patients with cancer may be inaccurate.
Key Words: posttraumatic stress disorder, • cancer, • DSM-IV, • trauma, • distress.
Abbreviations: PTSD = posttraumatic stress disorder;; SCID = Structured Clinical Interview for DSM-IV;; MDD = major depressive disorder;; GAD = generalized anxiety disorder;; HSCL-25 = Hopkins Symptom Checklist;; IES = Impact of Event Scale.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESACKNOWLEDGMENTSREFERENCES |
|---|
DSM-III-R (3) had required that an event be "outside the range of usual human experience" (p. 250) to be considered a traumatic stressor. In contrast, DSM-IV requires the combination of exposure to an event involving "actual or threatened death or serious injury, or a threat to the physical integrity of self or others" (criterion A1) and that "the person’s response involved intense fear, helplessness, or horror" (criterion A2; 4). The text accompanying the stressor criteria identifies life-threatening illness as a potentially qualifying event, and many have suggested that the experience of cancer diagnosis and treatment typify this (eg, 5, 6), even granting that cancer differs in many ways from more typical traumatic stressors (eg, rape, combat, childhood sexual abuse).
The changes in stressor criteria introduced in DSM-IV were somewhat arbitrary and controvertible (7, 8). Neel (1) reported that the inclusion of life-threatening illness as a potential stressor in DSM-IV resulted from work by Pelcovitz et al. (9). Yet the work of Pelcovitz et al. (9) did not address personal experience of illness, but focused on vicarious traumatization of parents of children with cancer. Regardless, alterations in diagnostic criteria have led to changes in the epidemiology of PTSD. Revision of the A1 criterion increased the number of events considered traumatic stressors in community samples by approximately 59%, although the subjective A2 criterion limits the net increase to 22% (10). The DSM-IV PTSD field trial data suggested that lifetime prevalence of PTSD was elevated among patients with cancer (22%; 11). However, these data were based on a small (N = 27) sample in which only 42% of the patients approached participated.
In the absence of agreement about a gold standard, evaluation of the appropriateness of diagnostic criteria becomes a complicated process of deciding whether relevant clinical phenomena are encompassed or whether revision of diagnostic criteria is in order. Uncritical expansion of diagnostic criteria, or bracket creep (12), risks rendering the established research data and clinical recommendations based on more restricted criteria inapplicable and threatens our ability to generalize new data to the populations in which the diagnostic criteria were originally developed. On the other hand, setting overly restrictive diagnostic criteria or arbitrarily limiting their clinical application risks denying persons in need the benefit of relevant research and effective intervention.
Examining the applicability of PTSD criteria to patients with cancer contributes to an evaluation of the appropriateness of DMS-IV diagnostic criteria. However, there are implications for our understanding of the experience of patients with cancer as well. If cancer is usefully viewed as a traumatic experience, as some have suggested (13, 14), distress and other diffuse psychological phenomena that do not meet threshold for a diagnosis may be viewed within a spectrum of posttraumatic stress responses, presumably as subsyndromal PTSD or PTSD-like symptoms. Indeed, arguments have been advanced that symptoms of anxiety and depression among patients with cancer should be viewed as posttraumatic stress responses (15), and there are increasing references to subsyndromal and PTSD-like symptoms in the cancer literature (5, 14, 16).
A literature has begun to accumulate examining PTSD, both syndromal and subsyndromal, among patients with cancer (see 17 for a detailed review). Studies have produced estimates of the prevalence of cancer-related current PTSD ranging from 0% (18) to 32% (19) and lifetime cancer-related PTSD from 3.5% (20) to 35.1% (18). Most reports have focused on patients with breast cancer (16, 18–26) , and some rely on the same or overlapping samples in ways that make summaries and comparisons difficult. Sample sizes have been small (eg, N = 37 [18]; N = 31 [19]); entry criteria have sometimes been so stringent as to restrict generalizability (eg, no surgery, chemotherapy, or radiotherapy for 6 to 72 months and remission of cancer [23]); assessment techniques have relied on instruments that fail to establish the A2 criterion before assessing symptoms of PTSD (eg, 22), and assessment of psychiatric comorbidity has been limited.
The present study examined PTSD among a waiting room sample of patients with breast cancer at a university-affiliated cancer center. We were interested in adding to the limited literature concerning the prevalence of PTSD among breast cancer patients using DSM-IV criteria, but also in exploring issues concerning the relationship of the A2 criterion to symptoms of PTSD among patients meeting the traumatic stressor criterion but not receiving a PTSD diagnosis. Moreover, we were interested in the relationship of the cancer experience to psychiatric comorbidity and nonspecific psychological distress, and the ability of brief screening instruments to uncover PTSD among women diagnosed with breast cancer.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESACKNOWLEDGMENTSREFERENCES |
|---|
2[3] < 2.64, NS). Women who were unavailable for interview were younger (mean = 50.0 years) than women who completed interviews (mean = 55.6 years, t[136] = 2.09, p < .05).
Measures
Structured Clinical Interview for DSM-IV
Diagnoses were obtained using the Structured Clinical Interview for DSM-IV (SCID)-I/NP (27), a semistructured interview that provides for evaluation of caseness for psychiatric morbidity according to DSM-IV criteria (2). Because of its modular construction, the SCID can be adapted for use in studies in which only particular diagnoses are of interest. Modules for PTSD, current major depressive disorder (MDD) and past MDD, and generalized anxiety disorder (GAD) were telephone-administered in the present study by trained, doctoral-level psychologists. PTSD was assessed only in relation to cancer and its treatment as stressors, and diagnosis was based on the period from diagnosis of cancer to the present. Interviewers were blind to participant responses on questionnaires.
Previous studies have shown adequate concordance between telephone-administered diagnostic interviews and face-to-face interviews (28–31). A random sample of 15 of the interviews was audiotaped and scored by second raters for reliability analysis. Agreement on the presence of symptoms between initial and secondary raters was high (95%; 
= 0.81).
Psychological Distress
The 25-item version of the Hopkins Symptom Checklist (HSCL-25; 32) was used to assess psychological distress. The scale incorporates 10 items from the HSCL-90 anxiety cluster, 13 items from the depression cluster, and 2 additional somatic symptoms (poor appetite, difficulty falling asleep or staying asleep). The same items appear with inconsequential differences in wording on the Symptom Checklist-90 (33). Hough et al. (32) found that the HSCL-25 was comparable or superior to the Center for Epidemiologic Studies Depression Scale (CES-D) (34) in detecting psychiatric disorder, depending on the criterion used. In the current sample, internal consistency was high (0.93).
Posttraumatic Stress Symptoms
The 15-item Impact of Event Scale (IES; 35) was used to assess symptoms of posttraumatic stress. Although this scale was not designed for the identification of caseness for PTSD diagnoses (36), it does provide an estimate of the extent to which people experience intrusive thoughts and avoidant behaviors related to a specific event across the preceding 7 days. It has been used previously with patients with breast cancer (eg, 37–39) and women at high risk for breast cancer (40). This measure was chosen over the revised IESs (41) because it has produced a more complete psychometric literature, provides greater comparability across studies, and allows for defining clinically important symptom levels. Internal consistency for our sample was calculated at 0.91 for the overall scale.1
Chart Abstraction
Information regarding demographics, staging of cancer at diagnosis, date of diagnosis, and occurrence and date of detected metastases and recurrences was abstracted from clinic charts by a trained predoctoral research assistant.
Procedure
Participants were recruited after approval was obtained from the Institutional Review Board of the University of Pennsylvania. Patients were approached in the waiting area by a research assistant during routine medical appointments for treatment of breast cancer and asked to participate in a study of the impact of cancer on emotional and social well-being. Patients who were unable to understand English well enough to provide informed consent or complete the instruments and diagnostic interview were excluded from study. After providing informed consent, participants were asked to complete a brief screening instrument on site (HSCL-25) and were given a packet of questionnaires to complete and mail to the investigators within 1 week. At that time, patients were scheduled for a telephone SCID interview to take place within 2 weeks. If questionnaire materials were not received before the scheduled interview, participants received a reminder phone call, and the interview was rescheduled.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESACKNOWLEDGMENTSREFERENCES |
|---|
A2 Criterion and the Prevalence of Posttraumatic Stress Disorder
We followed the strategy of Breslau and Kessler (10) in assuming that all subjects met the objective component of the stressor criterion (A1) requiring that the person "experienced, witnessed, or was confronted with an event or events that involved actual or threatened death or serious injury, or a threat to the physical integrity of self or others" (2) by virtue of their experience with cancer. The A2 criterion of the DSM-IV PTSD diagnosis, requiring that the person’s response involve "intense fear, helplessness, or horror" (2), was endorsed by 41% of the sample during the SCID interview. To determine which women were more likely to endorse the A2 criterion, endorsement was regressed on age at diagnosis, staging of neoplasm, level of education, and marital status. Although the overall equation was marginally significant (F[4,110] = 2.21, p = .07), the only significant individual predictor was age, with younger women being more likely to endorse the A2 criterion (ß = -0.28, p = .009).
The prevalence of PTSD was much lower than the prevalence of A2 endorsement, with 4% (confidence interval, 2%–9%) meeting the full diagnostic criteria. Although there was a significant relationship between meeting A2 criterion and receiving a diagnosis of PTSD (2[1] = 7.73, p = .005), 38% of those without PTSD met the A2 criterion, and only 11% of those endorsing A2 met criteria for PTSD.
As can be seen in Table 1, among people who met the A2 criterion without having PTSD,2 approximately 24% fulfilled criteria for avoidance cluster symptoms, and 48% fulfilled criteria for arousal cluster symptoms. However, 81% of the sample meeting the A2 criterion also met criteria for the re-experiencing cluster. Symptoms differentiating those receiving a PTSD diagnosis from those not receiving a PTSD diagnosis primarily involved avoiding thoughts and activities associated with the cancer experience, detachment, sleep difficulties, and irritability.
|
0.10, NS), as was disease stage(F[3,38] = 0.47, NS).
A2 Criterion and Comorbid Psychiatric Disorders
Overall, the prevalence of other psychiatric disorders was moderate, with 17% of the sample meeting criteria for at least 1 diagnosis: MDD, GAD, or PTSD. MDD was present in 9% (confidence interval, 5%–16%) and GAD in 6% (3%–12%) of the sample. Past MDD was reported by 30% (confidence interval, 22%–39%) of the sample. The A2 criterion predicted these disorders approximately as efficiently as it did PTSD. The probability of meeting diagnostic criterion for MDD, GAD, or past MDD given the A2 criterion was 17%, 11%, and 39%, respectively. These disorders, however, demonstrated high comorbidity with PTSD: 20% of those with PTSD were also diagnosed with GAD and 40% with MDD, and 100% met criteria for past MDD. Such comorbidity did not account for the nonspecificity of A2. If only people without comorbid PTSD were considered, the predictive utility of A2 for other disorders decreased but remained similar, with values of 9%, 12%, and 27% for MDD, GAD, and past MDD, respectively.
Posttraumatic Stress Symptoms, Distress, and Posttraumatic Stress Disorder
The mean score on the IES was below the cut-point of 20 (42, 43) designating clinically elevated distress (mean = 17.18, SD = 16.09). Using recommended criteria, 38% of the sample reported low, 24% moderate, and 38% elevated posttraumatic stress symptoms. To simplify interpretation, analyses using the IES as a categorical variable combined the low and moderate groups.
People with IES elevations were more likely to meet A2 criterion than those without (2[1] = 17.85, p < .001). Yet such symptoms were a poor predictor of PTSD. Although IES elevation related to PTSD diagnosis (2[1] = 6.81, p < .001) and the IES was highly sensitive to PTSD (100%), the probability that a person with an elevated IES had PTSD was only 11%. Similarly, although no people with nonelevated IES scores had PTSD, 36% of those without PTSD produced an elevated IES score.
The IES predicted MDD, GAD, and past MDD as well (all 2[1] 
3.94, p < .05), and had at least as much utility in predicting these disorders as in predicting PTSD. The probability that a person above the IES cut-point had MDD was 14%; GAD, 11%; and past MDD, 48%.
Overall, HSCL-25 distress scores were moderate (mean = 40.48, SD = 11.26) and below the cut-point of 44 establishing clinically significant distress, although 29% of the sample did report elevated distress. Such elevations were related to meeting the A2 criterion (2[1] = 15.22, p < .001). Not surprisingly, the HSCL-25 and IES were related (r = 0.57, p < .001), and people who exceeded cut-points on 1 tended to do so on the other (2[1] = 8.33, p < .005). General distress proved to be a poor predictor of PTSD. Although elevated HSCL-25 scores were moderately predictive of PTSD (80%), 20% of those with PTSD were missed, and the probability that a person with elevated distress had PTSD was only 12%.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESACKNOWLEDGMENTSREFERENCES |
|---|
Our data indicate that cancer-related PTSD was experienced by a small proportion of our sample, with confidence intervals suggesting that the population value is between 2% and 9%. This rate is similar to rates found by other investigators examining breast cancer patients using structured clinical interviews (eg, 23, 25), although lower than the estimated lifetime rate of cancer-related PTSD found by Alter et al. (11) in the DSM-IV field trial. This confidence interval overlaps with the lifetime prevalence of PTSD found for women in the general population in the National Comorbidity Survey (44) and is much lower than the 18.2% lifetime prevalence among women reported by Breslau (45) from the Detroit sample. Moreover, we found that the conditional probability of meeting the A2 criterion given breast cancer (41%) was less than half that found by Breslau and Kessler (10) in a representative sample of community-residing women exposed to either traumatic events overall (85.7%) or life-threatening illness (85.2%) in particular. The probability of developing PTSD among our sample is similar to or lower than that of the general population of women in which exposure to stressors is more variable, suggesting that breast cancer may not present a significant risk for PTSD.
Yet it is important that we not minimize the consequences to the minority of patients for whom cancer was traumatic. Those patients endorsing the A2 criterion appear similar to other populations in conditional risk of PTSD after exposure to traumatic stressors. Breslau and Kessler (10) report a conditional probability for PTSD of 12% given the A2 criterion, which is very similar to our own findings. However, A2 proved approximately equally useful as a predictor of MDD, GAD, past MDD or PTSD. The A2 criterion may be better thought of as a correlate of elevated distress in this population than as a marker for PTSD.
Symptoms from the re-experiencing cluster were common, and most participants who met the A2 criterion also met the single-symptom criteria for the re-experiencing cluster. More than 80% of those who met the A2 criterion reported intrusive, distressing thoughts or memories of the breast cancer experience. This rate is similar to that found among people who experienced the Northridge, California, earthquake (46) and may suggest that re-experiencing cluster symptoms do not indicate pathology among breast cancer survivors so much as they represent normative cognitive processing of a stressor. Research exploring aspects of these symptoms could play an important role in distinguishing PTSD from benign conditions. That is, cognitive processing of an ongoing health condition that holds the realistic threat of a progression or recurrence may be qualitatively different from unbidden thoughts about an experience securely in the past. Consistent with this interpretation, Deimling et al. (47) report that arousal and intrusive thoughts among cancer survivors are related to current cancer symptoms and enduring side effects of treatment. These authors suggested that such thoughts may be interpreted as appropriate vigilance and efforts to manage physical symptoms rather than as maladaptive responses. Future studies of intrusive thoughts may profitably incorporate assessment of the content and temporal focus of thoughts in differentiating normal processing from PTSD.
There was greater variability in arousal cluster symptoms between people with and without cancer-related PTSD. However, criteria for this cluster were met by 48% of those who met the A2 criterion but did not meet criteria for PTSD. At the level of individual symptoms, the 2 most common arousal symptoms were relatively nonspecific items—sleep disturbance and difficulty concentrating—which may reflect diffuse distress rather than PTSD. These were also very common among women with PTSD, as has been found among people who have experienced other stressors (eg, 48). In a heterogeneous group of patients with cancer, Deimling et al. (47) found that arousal symptoms such as these were best predicted by current cancer-related illness symptoms and postulated that they may relate to heightened monitoring of illness. Given that our sample was drawn from a group of women awaiting treatment or monitoring of their breast cancer, relatively high levels of arousal symptoms may reflect such cancer monitoring concerns, although our design does not allow us to differentiate vigilance from distress.
Consistent with research in other populations (eg, 46), participants were much less likely to meet criteria for the avoidance/numbing symptoms. Indeed, these symptoms are specific enough that Breslau et al. (49) devised a brief, 7-item screen for PTSD in which 5 of the items come from the avoidance/numbing cluster. Clinicians should be made aware that presence of multiple symptoms from this cluster may warrant a more in-depth inquiry into psychosocial functioning. However, a sense of foreshortened future, endorsed by all participants with PTSD, was acknowledged by 69% of those who met the A2 criterion but were not diagnosed with PTSD. This high rate of endorsement brings into question whether foreshortened future should be considered a symptom of PTSD or recognition of a real threat to physical integrity because of current disease, treatment, or recurrence among women being treated for cancer.
Meeting the A2 criterion only weakly predicted PTSD but was related to elevated distress and the presence of posttraumatic stress symptoms, suggesting that the A2 criterion is nonspecific and more indicative of general psychological distress than of PTSD. People endorsing A2 criterion were more than twice as likely to have elevated distress as those who did not. At the same time, people endorsing the A2 criterion were more than 4.5 times as likely to have elevated distress as they were to meet criteria for PTSD. This finding raises the question whether subsyndromal PTSD or posttraumatic stress symptoms (eg, 17) are better viewed as representing an adjustment disorder (2) than a traumatic stress response. Although this may be an applicable description of the clinical phenomena, there has been little research examining the validity of adjustment disorder as a diagnostic category (50), particularly among populations in which there is an ongoing threat of illness after resolution of an initial medical stressor. As well, the treatment implications for this broad diagnostic category are unknown, and it remains more an administrative than an empirical diagnosis.
Our findings also indicate that the IES is a poor predictor of PTSD. People scoring within the clinical range of the IES were approximately 5.8 times more likely to report increased distress on the HSCL-25 than they were to meet criteria for PTSD. This finding brings into question whether such symptoms may be more accurately viewed as representing diffuse emotional distress than indicating symptoms of PTSD.
Framing the psychological distress experienced by patients with breast cancer in terms of PTSD and trauma shapes the expectations of clinicians and patients about the adjustment expected and implies a treatment structure for managing difficulties in adjustment. However, a substantial proportion of patients who experience distress after a cancer diagnosis is likely to prefer interventions that are informational or educational to those that are psychiatric in nature (51) and may be more likely to engage in and benefit from such interventions. Most patients with cancer who experience increased distress are likely to experience a remission in distress soon after the initial period after diagnosis (52). As well, providing psychosocial interventions for distress is not a risk-free endeavor, because aggressive intervention may disrupt a natural coping process and relates to increased mortality among women recovering from myocardial infarction (53). Providing distressed patients with cancer the choice of interventions that directly address their need for information may be a conservative approach that broadly benefits the majority of patients.
A few caveats should be considered in interpreting these data. Our sample consisted mainly of middle-class, European American women who had access to financial and informational resources and medical care at a comprehensive cancer center. The results are most confidently generalized to this segment of the population with breast cancer. As well, factors on which our sample was relatively homogenous, such as access to financial and informational resources and social and environmental stability, likely act to moderate the impact of breast cancer on adaptation, and women from less enriched environments may differ in their risk for difficulties in adjustment. Although our sample is as large as that of most studies of psychiatric morbidity among patients with cancer, representativeness and sample size were limited, and future research aimed at establishing prevalence rates of PTSD in populations with breast cancer will require larger, more diverse samples. Our retention rate was high, but people who failed to complete interviews were younger than completers. Although these people did not differ from completers in distress, we found a relationship between age and endorsement of the A2 criterion, and younger women have been found to be at increased risk of breast cancer-related distress in other studies (eg, 54). Other issues arise from operationalization decisions. Our use of the SCID did not allow for differentiation of current from past PTSD. However, differentiation would have decreased overall prevalence and would have made positive predictive power even lower for screens. We assumed that all patients met the A1 criterion, and only those who also met the A2 criterion provided further responses to the PTSD interview. Thus, we were unable to make predictions about the proportion of women who perceived the cancer experience as potentially life-threatening, and were unable to make statements about the presence of PTSD spectrum symptoms among patients who did not acknowledge responding with fear, helplessness, and horror to their diagnosis and treatment.
Recognizing these limitations, our findings still have implications about how breast cancer is construed by our sample. Although diagnosis and treatment were distressing, this distress amounted to PTSD in few cases. This is not to suggest that these few cases be overlooked, only that trauma and PTSD not be used to frame the experiences of most patients. Symptoms of re-experiencing and arousal are relatively common among women who respond to breast cancer with feelings of fear and helplessness. Although our design did not allow assessment of these features among all women, it is reasonable for health care providers to expect these reactions among patients and to be prepared to normalize the experience. Given that all women with PTSD had a history of MDD and the relationship of PTSD to exposure to previous trauma (eg, 21, 55), assessment of past MDD and trauma history as risk factors may be fruitful. Similarly, given the high rate of A2 endorsement and its poor predictive value, future research may attempt to disentangle "intense fear, helplessness, or horror" to improve prediction. Given the normative fear associated with a diagnosis of cancer, a reaction of helplessness or horror may be more discriminating and predictive of difficulties in adjustment. Prediction may also be enhanced among patients reporting multiple symptoms from the avoidance/numbing cluster. Regardless, it may be less appropriate to conceptualize patients’ experience in terms of trauma and a need for specialized mental health services than to consider it a psychosocial stress reaction more responsive to supportive, nonpsychiatric intervention; renewed focus on psychosocial services that strengthen personal and social resources may be the more appropriate strategy to adopt.
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESACKNOWLEDGMENTSREFERENCES |
|---|
|
NOTES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESACKNOWLEDGMENTSREFERENCES |
|---|
2Participants not meeting the A2 criterion were not administered the remaining PTSD module, resulting in a smaller N for analyses involving PTSD criteria.
Received for publication August 1, 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESACKNOWLEDGMENTSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  A. L. Stanton Psychosocial Concerns and Interventions for Cancer Survivors J. Clin. Oncol., November 10, 2006; 24(32): 5132 - 5137. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. A. Hamann, T. J. Somers, A. W. Smith, S. S. Inslicht, and A. Baum Posttraumatic Stress Associated With Cancer History and BRCA1/2 Genetic Testing Psychosom Med, September 1, 2005; 67(5): 766 - 772. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Okamura, S. Yamawaki, T. Akechi, K. Taniguchi, and Y. Uchitomi Psychiatric Disorders Following First Breast Cancer Recurrence: Prevalence, Associated Factors and Relationship to Quality of Life Jpn. J. Clin. Oncol., June 1, 2005; 35(6): 302 - 309. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
