|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2003 American Society for Clinical Oncology Identifying Psychosocial Risk Indicative of Subsequent Resource Use in Families of Newly Diagnosed Pediatric Oncology Patients
From the Division of Oncology, The Children’s Hospital of Philadelphia, Philadelphia, PA. Address reprint requests to Anne E. Kazak, PhD, ABPP, Division of Oncology, The Children’s Hospital of Philadelphia, Room 1486 CHOP North, 34th St and Civic Center Blvd, Philadelphia, PA 19104; email: kazak{at}e-mail.chop.edu.
Purpose: The primary purpose of this prospective study was to identify the level of risk for psychosocial distress in families of children newly diagnosed with cancer. Additional study aims were to examine concordance among family and staff reports of psychosocial risk, changes in risk status over time, and to predict the use of psychosocial resources during the first months of treatment. Patients and Methods: Caregivers of 125 children newly diagnosed with cancer completed the Psychosocial Assessment Tool (PAT) at diagnosis (t1) and 3 to 6 months later (t2). Primary oncologists and nurses completed an analogous measure of perceived family psychosocial risk at t1 and t2. At t2, oncology social workers reported types and intensity of psychosocial interventions provided. Results: The PAT identified three subsets of families who presented with increasing levels of psychosocial risk at diagnosis. In general, there was moderate concordance among family, oncologist, and nurse reports of psychosocial risk. PAT scores at t1 predicted t2 PAT scores and psychosocial resource use at t2 beyond demographic or disease factors. Conclusion: A brief screening tool (PAT) is valuable in identifying psychosocial risk factors at diagnosis and is predictive of later use of psychosocial resources. As a next step in this research, the development of psychosocial interventions to match family risk level may be an effective and cost-efficient approach to working with families to address their concerns and promote short- and long-term adjustment.
THE INTENSE, complex, and enduring demands of treatment for childhood cancer on the family are well established. National and international recommendations for comprehensive care emphasize the importance of psychosocial services.1–5 In practice, providing the recommended range of psychosocial assessment, intervention, and consultation services in oncology programs is difficult and variable across treatment centers. There is no consensus about the most effective models for providing this care. Changes in healthcare in the United States in the 1990s contributed to a reduction in available psychosocial services.6 Psychosocial care is sometimes perceived as expensive and less essential than other evaluations and interventions. The costs of these services often are not reimbursable. However, psychosocial difficulties affect medical treatment. For example, families experiencing multiple stressors may have more difficulty making effective decisions and adhering to treatment recommendations; those with limited financial resources may be less able to comply with expensive treatment. Families facing multiple challenges may strain pediatric oncology teams by requiring more healthcare personnel time to deal with nonmedical concerns. Although there are effective interventions that can minimize interferences that psychosocial factors may pose to treatment (Spirito and Kazak, manuscript in preparation), there are no known studies that empirically demonstrate the link between psychosocial factors and healthcare costs and resource use. Most healthcare resource data address length of stay and hospital charges without considering the expenditure of professional resources caused by the heavy demand on medical, nursing, and psychosocial personnel when family difficulties persist or escalate during the course of treatment. The psychological functioning of most families of medically ill children is within normal limits. Well controlled studies have documented more similarities than differences between them and never-ill controls7; however, subsets of families experience ongoing and sometimes escalating distress. Kupst et al8,9 demonstrated that families with the most difficulty at diagnosis continue to experience the highest levels of distress, even after treatment ends. Our group has also shown that parental anxiety at diagnosis is predictive of ongoing distress.10,11 If oncology teams could predict the course of psychosocial adjustment at diagnosis, they could target psychosocial intervention accordingly in a systematic way. Maximizing care early in a family’s treatment course could minimize the toll of psychosocial factors on the patient, family, and healthcare team. We propose that an evidence-based classification of level of psychosocial risk for families will allow for optimal use of the healthcare team’s resources. Using an adaptation of the National Institute of Mental Health prevention framework,12,13 we hypothesize that we can classify families into three psychosocial risk groups and that they will require different levels of care. The first group, the majority of families with children newly diagnosed with cancer, contains normally functioning families who are well suited to what we term universal care. After the initial shock of their child’s diagnosis and treatment, they will likely respond in a psychologically adaptive manner with the general psychosocial care typically provided by medical and psychosocial oncology teams. A smaller group of families have experiences that place them at increased risk for distress and may require more intensive and focused intervention (selected care). The smallest subset of families presents multiple risk factors evident at diagnosis and warrants the most intensive and ongoing psychosocial care (targeted care). Targeted families, although few in number, are most costly to help. As the first step in establishing an efficient, empirically supported approach to assessing and addressing psychosocial risk in families of children newly diagnosed with cancer, we developed a 20-item screening questionnaire, the Psychosocial Assessment Tool (PAT). In an initial pilot study of 109 families, we concluded that the PAT was a feasible screening tool; families were amenable to completing it and to giving the medical team full access to the information. In addition, families endorsed a range of risks. The majority of families had few psychosocial risk factors, although a subset (5% to 30% depending on the type of risk) responded in a pattern indicative of elevated psychosocial risk.14 The current prospective study represents the second step in evaluating the utility of the PAT. Four specific questions were identified. Question 1: What are the most common psychosocial risks at diagnosis (t1), and can patterns of reported risks be used to characterize families using the framework of low (universal), moderate (selected), and high (targeted) risk? Question 2: Do families and staff agree in their report of psychosocial risk? Question 3: Does psychosocial risk change between the time of diagnosis (t1) and the first months of treatment (t2, at 3 to 6 months postdiagnosis)? Question 4: Can assessment of psychosocial risk at t1 predict use of psychosocial resources during the first months of treatment (t2) beyond demographic and disease-related variables?
Participants and Procedure Participants were caregivers of 125 children newly diagnosed with cancer, recruited from the Division of Oncology at the Children’s Hospital of Philadelphia (CHOP). The children ranged in age from 2 months to 19 years (mean [M] = 8.11 years; standard deviation, 5.38); 60% were male. Regarding ethnicity, 78.4% of the children were white, 16.0% were African-American, 2.4% were Hispanic-American, and 3.2% were of multiracial decent. Forty-eight percent of the patients were diagnosed with leukemias, including acute lymphocytic leukemia, acute myeloid leukemia, or acute promyelocytic leukemia; 17.6% had brain tumors. Other diagnoses included neuroblastoma (8%), lymphoma (7.2%), Wilms’ tumor (3.2%), osteosarcoma (1.6%), Ewing’s sarcoma (4%), Hodgkin’s disease (3.2%), and other tumors (7.2%; rhabdomyosarcoma, desmoplastic tumor, carcinoma, sarcoma, primitive neuroectodermal tumor, hepatoblastoma, and rhabdoid tumor). These demographic and diagnosis distributions are representative of the oncology population at CHOP. Each child’s primary oncologist (attending or fellow, n = 33) and nurse (n = 55) also completed questionnaires assessing their perceptions of the family’s psychosocial risk factors. Social workers (n = 7) reported on the families’ use of social work services.
Procedure Caregivers completed the PAT questionnaire again during the early months of treatment (Mt2 = 4.43 months; range, 3 to 6 months) during an inpatient stay, an outpatient visit, or returned it by mail. Seventy-nine families completed t2 within 3 to 6 months of diagnosis. Four children died before t2. Fourteen families did not return the form within 6 months of diagnosis, and 28 families did not return their forms, despite phone and personal follow-up requests. Oncologists, nurses, and social workers also received an IRB-approved description of the study with instructions for completing their rating forms. Each patient’s primary oncologist and nurse completed staff PATs at t1 and t2. Each child’s primary social worker completed a Social Work Activity Form (SWAF) at t2 (4 to 5 months postdiagnosis).
Measures Staff PAT. The staff PAT was also constructed for this study by our multidisciplinary research team to assess oncologist and nurse ratings of family psychosocial functioning and risk. It is a 14-item Likert scale that requires less than 5 minutes to complete, and includes items that correspond to domains on the PAT, as well as a three-level categorical rating of risk level (universal, selected, and targeted) on the basis of the National Institute of Mental Health risk prevention framework. The staff PAT yielded two scores per family: a quantitative total score, calculated by summing the 14 items for each staff respondent, and a qualitative rating of each family in terms of the three risk levels. SWAF. The SWAF was developed for this study to record the social work services (types and intensity) provided to each participating family. For each participant family, a social worker rated the categories of activity that they provided from t1 to t2 (supportive counseling, medical care issues, community or financial organizations, education or school contact, mental health or family problems, child welfare, and legal issues). They also rated each category of activity in terms of the intensity of service provided on a scale of 1 to 5. The SWAF, which took 5 to 10 minutes to complete, was scored by summing the number of activities carried out within each category, weighting each subtotal by multiplying it by the intensity rating (on a scale of 1 to 5) for that category, and then summing the weighted scores for a total SWAF score. The SWAF can be completed in 5 to 10 minutes. Treatment intensity ratings. Each child’s medical treatment was rated by an oncologist (B.J.L.) using four categories, similar to our prior research.15 On the basis of data abstracted from the medical chart, treatment was rated as follows: level 1, surgery or enucleation only or surgery plus short-term chemotherapy protocols; level 2, brain tumor surgery and low- or standard-risk chemotherapy protocols; level 3, multiple modalities of treatment (eg, chemotherapy and radiation) and high-risk protocols; and level 4, metastatic disease, relapse protocols, and bone marrow transplantation. When disease course changed between t1 and t2 (eg, refractory disease, relapse, metastatic disease), the ratings reflected the most intense treatment protocol.
Question 1: Frequency and Distribution of Risk Factors The frequencies of risk factors were calculated on the basis of item responses on the PAT. The mean number of risk items endorsed was 3.49 (range, 0 to 16 of a possible 31 items) at t1 (Fig 1  ). The most commonly endorsed risk factors were financial difficulties (43.2%), followed by having more than three children living in the home (18.4%), a history of emotional problems in the family (18.4%), and single parenthood (17.6%; Table 1). As anticipated, three distinct psychosocial risk groups emerged (F2,122 = 277.84; P = .001). The distribution of risk factors indicated that 6.4% of families reported no risk factors; 59.2% of the sample reported a total risk score at or below the median (three items), likely comprising the universal risk category (PAT score mean, 1.76). An additional 33.6% reported four to seven items, indicating a selected category (M = 5.07); 7.2% reported eight to 16 items, indicating the subset of targeted families (M = 10.33).
Families were also rated by their primary oncologists and nurses on perceived level of psychosocial risk and subsequent need for services. On the staff PAT forms, universal was defined as "This family experienced initial distress and has adjusted well to their child’s illness and treatment; I would recommend standard psychosocial care focused on general themes relevant to families of children with cancer." Selected was defined as "This family came to us with some difficulties that might indicate ongoing psychosocial difficulties; I would recommend a more intensive level of psychosocial care, such as groups, educational programs, regular access to social workers and child life specialists for specific concerns." Targeted was defined as "This family is at risk for ongoing and escalating psychological distress; I would recommend an intensive level of psychosocial care, such as consultation with a psychologist, including assistance to staff in working with this family."
Consistent with the conceptual framework, most families were rated as universal (52.8% by MDs; 51.2% by RNs), fewer families were rated as selected (24.8% by MDs; 14.4% by RNs), and a small group of families were perceived to be targeted (5.6% by MDs; 6.4% by RNs; Fig 2
Question 2: Correlations Between and Within Family and Staff Reports There was moderate concordance between staff and parent report of psychosocial risk at t1 and t2. PAT and staff PAT scores were significantly correlated for oncologists (rt1 =.41, P < .001; rt2 =.35, P < .01) and nurses (rt1 = .43, P < .001; rt2 =.33, P < .01). Oncologist and nurse ratings were also significantly associated (rt1 =.42, P < .001; rt2 =.59, P < .001).
Question 3: Change in Psychosocial Risk Over Time
Question 4: Psychosocial Risk Predicting Use of Social Work Resources To examine whether PAT scores could predict social work resource use beyond demographic and disease-related variables, and medical staff ratings, a multiple linear regression model was established. Age, sex, ethnicity, and treatment intensity ratings entered during step 1 predicted 6.7% of the variance in SWAF scores (F4,64 = 2.23; P < .10). For step 2, t1 physician and nurse PAT were entered. Together, they accounted for an additional 16.7% of the variance (likelihood ratio test, 7.97; P = < .01). For step 3, t1 PAT was entered and an additional 8.4% of the variance in SWAF was explained (likelihood ratio test, 8.63; P < .01). A total of 31.8% of the variance in social work resource use was explained by this model (F7,61 = 5.53; P < .001).
Providing comprehensive care to pediatric oncology patients and their families necessitates the integration of multidisciplinary psychosocial teams and interventions into treatment plans. There is a long history of social work partnership in pediatric oncology and an increasing presence of psychologists. Although this care is viewed as necessary and largely successful, current economic concerns require data-based evidence of the effectiveness of psychosocial services. A systematic approach to assessing families at diagnosis and understanding their psychosocial concerns during the course of treatment can facilitate the development and evaluation of interventions targeted toward specific family needs. This study is the second in which we have used a brief psychosocial screening tool, the PAT, to assess risk factors in families with a child with newly diagnosed cancer.14 The data supported the existing literature which indicates that most families of children with cancer continue to function well as they face the extraordinary challenges of their child’s diagnosis and treatment. As proposed, staff and family ratings support the conceptualization of a three-tiered psychosocial risk framework. Specifically, the PAT data show that the majority of families (ie, 59.2%) evidenced a minimal number of risks, a smaller number reported several risk factors (ie, 33.6%), and a minority (ie, 7.2%) reported multiple risk factors. Understanding commonly endorsed psychosocial risk factors for families (ie, financial difficulties, marital status, family size, and a family history of emotional problems) may aid in the development of targeted interventions for newly diagnosed families. Oncologists and nurses showed good agreement in their perception of psychosocial risk. Consistent with our theoretical model, both groups agreed that the majority of families had few risk factors (universal), one quarter to one third showed indications of greater psychosocial concerns (selected), and less than 10% were seen as being at highest risk (targeted). In addition, the correlation between specific staff and family PAT scores indicated that oncologists’ and nurses’ perceptions are consistent with family report of specific concerns at diagnosis, and that this accuracy is maintained several months into treatment. Oncologists and nurses were also concordant with each other regarding the specific risks they believe families experience. Although these correlations were statistically significant, they were moderate in magnitude. Thus, although these staff measures are not ideal predictors of family risk in isolation, they indicate that oncology team members form valid impressions of family risk. Prior research indicates that family distress peaks during the first months after diagnosis, then decreases and stabilizes at about 1 year postdiagnosis.16 Knowing a family’s initial level of risk appears to be important in understanding their risk several months later. t2 PAT scores can be predicted reliably from t1 measures, supporting the utility of this brief screening device at diagnosis. Furthermore, family and staff reports of higher risk at diagnosis predicted social work activity across the first months of treatment, providing more support for the utility of PAT as a brief early screening tool. Dropouts from this study offer important information. Families that completed t1 but not t2 had significantly higher initial PAT scores than those who remained enrolled in the study. It is likely that these families continued to have difficulties and were less connected with their treatment team. These families may require more immediate and targeted services because it requires more time and skilled efforts to sustain their involvement and engagement with the team over time. This finding also supports the importance of a routine brief screening tool at diagnosis rather than a wait-and-see approach to addressing psychosocial factors in pediatric cancer. It is plausible that the predictive ability of the PAT for t2 would have been even stronger were these higher risk families retained in the study. Correlated ratings among staff and between staff and families raise an important question about other ways in which completing the PAT and staff PAT may be helpful to medical and psychosocial teams. It is possible that simply completing the PAT or staff PAT forms may have served as an intervention to facilitate communication and heighten awareness of families’ psychosocial needs. Communication among the medical team members and between the team and families may be responsible for the concordances seen here. For example, in the days immediately following diagnosis, staff members typically communicate regarding apparent risk factors that may predispose a family to greater distress, particularly when concerns (eg, preexisting emotional problems, significant financial strain) are raised. It is also possible that, in this study, the use of the PAT facilitated communication about important risk items. Whether PAT adds a structure to facilitate timely and focused communication that might not occur (or would occur differently) without such an instrument is a question for future study. In future work, it is recommended that more attention be given to systematically evaluating the role of demographic characteristics such as socioeconomic status and ethnicity. Additional exploration of how the PAT was perceived by families of ethnic minority backgrounds would be helpful to ensure that the procedure reflects the concerns of all families and that the process reflects values and beliefs of minority families. The PAT has well-established feasibility and is helpful in reliably assessing psychosocial risk at diagnosis and over the course of the first months of treatment for families of children with cancer. Clinically, identifying psychosocial risk soon after diagnosis is vital in the timely implementation of services. Indeed, scores on this brief screening tool were able to predict future social work intervention. Additional steps in this work will include estimates of additional costs, such as the time and effort required of oncologists, nurses, other mental health professionals, and hospital administrators in addressing psychosocial needs of families. These data have implications for the development of empirically supported recommendations for psychosocial services in pediatric oncology.1–5 In a climate of limited resources a systematic identification of higher risk families could help to target the type and intensity of services needed by families. The use of a brief psychosocial screening tool is recommended because it creates a structure for identifying the needs of families early in treatment. However, it is essential that all families of children with cancer have access to the range of services necessary. Additional research is necessary before specific treatment recommendations can be linked to PAT scores. Even families that appear to be coping well (level I) should receive psychosocial care. Identification and delivery of more preventatively oriented interventions (eg, procedural distress, family counseling, and linkages with schools and community resources) are essential. Although it is likely that a small number of families will continue to be the highest users of psychosocial resources, no family should go without access to quality psychosocial interventions. Ultimately, a comprehensive approach to the costs associated with providing psychosocial care in pediatric oncology must include outcomes such as patient and family satisfaction, and thoughtful means of calculating the broader costs to hospitals of untreated or undertreated psychosocial problems. With more children surviving cancer and the associated medical late effects that include and influence psychological health, the long-term adjustment of survivors and their families are essential outcomes.
The data collection period spanned the terrorist activities of September 11, 2001. These events (in New York, Pennsylvania, and Washington, DC) were in relatively close proximity to Philadelphia. We immediately developed a Recent Events Checklist (REC; unpublished measure, available from the authors) and obtained IRB approval to add this 14-item questionnaire to assess families’ exposure to, and perceived impact of, the terrorist events to assess whether the distress reported by families was associated with the events of September 11. No families were excluded on the basis of their date of diagnosis, although families were not approached on September 11 to 13, 2001. The REC assesses exposure to each of 15 situations associated with the events of September 11 (eg, "I was in the NY or Washington area on September 11"; "I was injured"; "I know someone who died"). The mean number of exposure items endorsed was 2.4 (range, 0 to 7 items of 15 possible items). Participants’ mean estimate of the current perceived impact of the terrorist events on them was 2.2, on a scale of 1 (very little) to 4 (very great), and mean future impact was 2.0. There were no significant differences between families diagnosed before September 11 and those diagnosed afterward in Mt1 or Mt2 PAT scores (t = -0.02, not significant; t = -0.13, not significant, respectively) or anxiety scores (t = -0.77, not significant), indicating that these families were not additionally stressed beyond the distressing impact of their child’s diagnosis.
We thank the families who participated in this study and the staff of the Division of Oncology for their participation and interest in this work.
Supported by the University of Pennsylvania Cancer Center and the United States Child and Maternal Health Bureau Leadership Education in Neurodevelopmental Disabilities Training Grant at The Children’s Hospital of Philadelphia.
1. American Academy of Pediatrics: Guidelines for the pediatric cancer center and the role of such centers in diagnosis and treatment. Pediatrics 99:139–140, 1997 2. Kazak A, Meadows A: Integrating psychosocial research and clinical practice in a pediatric hospital, in Drotar D (ed): Handbook of Research in Pediatric and Clinical Child Psychology: Practical Strategies and Methods. New York, NY, Plenum, 2000, pp 511–528 3. Noll R, Kazak A: Psychosocial care, in Ablin A (ed): Supportive Care of Children with Cancer: Current Therapy and Guidelines from the Children’s Cancer Group. Baltimore, MD, Johns Hopkins Press, 1997, pp 263–273 4. Noll R, Kazak A: Psychosocial care, in Ablin A (ed): Supportive Care of Children with Cancer (ed 3). Baltimore, MD, Johns Hopkins Press, in press 5. Jankovic M, Van Dongen-Melman J, Vasilatou-Hosmidis H, et al: Improving the quality of life for children with cancer. Tumori 85:273–279, 1999[CrossRef][Medline] 6. Drotar, D: Anticipating the future of collaboration, in Consulting with Pediatricians: Psychological Perspectives. New York, NY, Plenum, 1995, pp 209–220 7. Kazak A, Rourke M, Crump T: Families and other systems in pediatric psychology, in Roberts M (ed): Handbook of Pediatric Psychology (ed 3). New York, NY, Guilford, 2003
8. Kupst M, Schulman A: Long-term coping with pediatric leukemia. J Pediatr Psychol 13:7–22, 1988
9. Kupst M, Natta M, Richardson C, et al: Family coping with pediatric leukemia. J Pediatr Psychol 20:601–617, 1995
10. Kazak A, Barakat L: Parenting stress and quality of life during treatment for childhood leukemia predicts child and parent adjustment after treatment ends. J Pediatr Psychol 22:749–758, 1997 11. Best M, Streisand R, Catania L, et al: Parental distress during pediatric leukemia and parental posttraumatic stress symptoms (PTSS) after treatment ends. J Pediatr Psychol 26:299–307, 2002 12. National Institute of Mental Health: Priorities for Prevention Research: A National Advisory Council Workgroup on Mental Health Disorders Prevention Research. Bethesda, MD, NIMH publication #98-4321, 1998 13. Power T, DuPaul G, Shapiro E, et al: Promoting Children’s Health: Integrating School, Family and Community. New York, NY, Guilford, 2003 14. Kazak A, Prusak A, McSherry M, et al: The Psychosocial Assessment Tool (PAT): Pilot data on a brief screening instrument for identifying high risk families in pediatric oncology. Fam Syst Health 19:303–317, 2001
15. Hobbie W, Stuber M, Meeske K, et al: Symptoms of posttraumatic stress in young adult survivors of childhood cancer. J Clin Oncol 18:4060–4066, 2000
16. Kazak A, Penati B, Brophy P, et al: Pharmacologic and psychologic interventions for procedural pain. Pediatrics 102:59–66, 1998 Submitted December 26, 2002; accepted May 13, 2003.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2003 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
