Psychosocial needs assessment among an underserved, ethnically diverse cancer patient population†‡
Presented at Exploring Models to Eliminate Cancer Disparities Among African American and Latino Populations: Research and Community Solutions, Atlanta, GA, April 21–22, 2005.
This article is dedicated to the late Carole Morgan, PhD, whose contribution to understanding the needs of the underserved is immeasurable. In addition, we thank Arlene Caban, Jeanne Carter, Anne Fatone, Gia LaRuffa, Anne Skamai, and Jennifer Usher for data collection and Melanie Harris, Ruth Santizo, and Rita Guimaraes for their invaluable editorial assistance.
Little attention has been directed towards identifying and addressing the psychosocial needs of ethnic minority and underserved cancer patients. This study describes the development of a psychosocial needs survey and patterns and predictors of need among an ethnically diverse underserved cancer patient population in Bronx, New York. A 34-item psychosocial needs assessment survey was developed to assess 4 categories of need: Informational, Practical, Supportive, and Spiritual. A total of 248 oncology outpatients (48% non-Hispanic whites, 25% African Americans; 19% Hispanic) completed the survey in oncology clinic waiting rooms. The survey demonstrated high internal consistency and face validity. Ethnicity was the sole predictor of needs (P < .02), even after controlling for education, time since diagnosis, treatment status, marital status, and age. The mean percentage of needs endorsed by African Americans, Hispanics, and non-Hispanic whites respectively was 81%, 85%, 70% for Informational; 63%, 68%, 36% for Practical; 69%, 73%, 48% for Supportive; and 49%, 60%, 31% for Spiritual needs. This needs assessment offers clear directions in which to focus QOL intervention efforts among underserved and ethnic minority cancer patients. Cancer 2007. © 2006 American Cancer Society.
Substantial research has documented socioeconomic and ethnic disparities in cancer outcomes with regard to disease stage, treatment, and survival. The underserved, or those from the lowest socioeconomic strata, are more likely to have the most adverse cancer outcomes.1 In comparison to non-Hispanic whites, African American and Hispanic populations are also at greater risk for advanced disease and poorer medical outcomes,2, 3 which are not always explainable by socioeconomic factors.4 As a result, efforts to reduce disparities have followed, with a focus on interventions to promote cancer education, early detection, access to care, and clinical trials participation among these high-risk populations.5 Comparatively little attention, however, has been directed towards identifying and addressing these “socio-ethnic” disparities with regard to another important area, namely, psychosocial needs and services.6
The impact of cancer and its treatment on quality of life (QOL) has been well documented, as reflected in the more than 500 review articles on the subject found in Medline between 1996 and 2006. Pain, fatigue, anxiety, depression, sexual changes, role losses, and financial strain are all potential “side-effects” of living with a cancer diagnosis.7–12 Further, research suggests that ethnic minority and/or undeserved populations may experience specific vulnerabilities in QOL after cancer.13 Hispanic patients, for example, appear to be at risk for high levels of psychological distress,14–17 whereas African Americans seem to be more vulnerable to social disruption and inadequate support.18–20 Although a plethora of behavioral interventions to address QOL concerns have been developed and studied among cancer patients,21–23 only a few of these studies have focused on ethnic minority and underserved cancer patients.24–26 Given the limited enrollment, retention, or efficacy outcomes of these studies, it is clear that more information is needed to develop culturally tailored interventions.
Psychosocial needs assessments are a direct method to identify the specific services and assistance most desired by patients. They have been utilized frequently in the area of cancer as a first step in designing needs-tailored interventions.27–36 The importance of assessing needs is underscored by several findings: 1) unmet needs have been associated with poorer QOL37; 2) attention to psychosocial needs has been linked with higher patient satisfaction with medical care38; and 3) provision of “unwanted” interventions that do not fit with personal needs may actually have detrimental effects.39
In the area of cancer, needs assessments have identified several major areas of need, including emotional, physical, daily living, patient and family support, financial, informational, and spiritual. Unfortunately, generalizability of previous needs assessments is limited, as most studies were conducted outside of the United States28–32, 34, 39–42 or with a distinct U.S. subpopulation (e.g. Asian/Pacific Islanders43). Further, other needs assessment approaches focused on identifying “problems” or “concerns” rather than self-reported needs,27, 33, 35, 44 or addressed patients' desire for physician-related interventions only.45 Lastly, existing needs assessment measures have not been developed for use with underserved, low-literate, and English/Spanish-speaking populations.
Although psychosocial needs assessments can provide a critical starting point for developing tailored interventions, their use among those most in need, that is, the underserved and ethnic minority cancer patients, is glaringly lacking. There are several arrows pointing to areas of need among these vulnerable groups, but a clear roadmap for intervention has yet to be developed. In an effort to help contribute to this roadmap, this article will describe the results of a Psychosocial Needs Assessment conducted to identify areas of need and potential intervention among underserved white, African American, and Hispanic patients. The specific aims of this study are: 1) to describe the development of a Psychosocial Needs Assessment Survey; and 2) to describe patterns and predictors of psychosocial needs among an underserved, ethnically diverse cancer patient population.
This study takes place in the Bronx, New York, which has a population of nearly 1.4 million and the highest ethnic diversity rating among 19 major counties across the United States.46 The major racial/ethnic groups represented are Hispanics (48%), blacks (36%), and non-Hispanic whites (15%).47 The subpopulations include African American, Puerto Rican, West Indian, Dominican, and Eastern European. According to the 2000 U.S. Census data, the Bronx is the poorest borough in New York City, with 31% living below the poverty line. As with most disadvantaged minority populations in the United States, the Bronx suffers from high cancer morbidity and mortality rates. According to the American Cancer Society, over 5000 new cases are diagnosed in the Bronx annually, with incidence rates rising.48 Based on population-based statistics, the Bronx is designated as a medically underserved community.49
MATERIALS AND METHODS
Oncology outpatients were recruited from 2 outpatient oncology clinics of a university medical center in the Bronx, New York. Eligible subjects were 18 years and older. The characteristics of the 248 participants are displayed in Table 1. As shown, the sample was ethnically diverse, with 48% white, 25% African American, 19% Hispanic, and 4.5% Other (included Asian/Pacific Islander, American/Alaskan Native, Mixed) participants. The mean age of patients was 56. There were slightly more women (59%), with most patients (68%) having earned up to, but no greater than, a high school degree. There was wide variability in cancer diagnosis and time since diagnosis, which ranged from 1 month to 22 years (M = 2.93 yr + 3.95), with about half of the participants currently receiving treatment.
Table 1. Sociodemographic and Medical Characteristics of Needs Assessment Survey Participants (N = 248)
| Male||96 (39)|
| Female||145 (59)|
| White||118 (48)|
| African American||62 (25)|
| Hispanic||48 (19)|
| Other/missing||20 (8)|
| Catholic||113 (46)|
| Jewish||38 (15)|
| Protestant||30 (12|
| Other||51 (21)|
| None||7 (3)|
| Elementary school or less||22 (9)|
| High school||96 (39)|
| Some college||49 (20)|
| College||42 (17)|
| Graduate/professional school||30 (12)|
| Married/cohabitating||135 (54)|
| Separated/divorced||34 (14)|
| Single||38 (15|
| Widowed||29 (12)|
| Breast||47 (19)|
| Other solid tumors||81 (33)|
| Hematologic||78 (32)|
| Two or more cancers||11 (4)|
| Yes||59 (24)|
| No||146 (59)|
| Yes||115 (46)|
| No||83 (34)|
|Age, mean ± SD (range)||55.82 ± 14.14 (18–85) y|
|Time since diagnosis, mean ± SD (range)||35.14 ± 47.34 (1–264) m|
|2.93 ± 3.95 (0.1–22) y|
During a 2-month period, trained psychology graduate students approached patients in the oncology clinic waiting rooms, where they were asked to fill out a brief 5-page survey to “better understand what our patients need so we can provide more services and programs for you in the future.” After providing written informed consent, patients completed the confidential self-report survey, with the data collector remaining nearby to answer any questions. The needs assessment survey took approximately 10 min to complete.
Data Analytic Approach
Data analyses were conducted in 5 stages: 1) preliminary analyses examined missing data and their correlates, as well as sociodemographic and medical differences between ethnic subgroups; 2) reliability analyses (i.e. Kuder-Richardson 20) were run to examine internal consistency of the needs assessment subscales; 3) descriptive (i.e. percentages, means) and univariate (i.e. ANOVA) statistics were employed to examine prevalence of needs in the group as a whole: 4) multivariate analyses (i.e. hierarchical regression) were run to identify predictors of needs within specified domains; and 5) further descriptive and univariate analyses examined patterns of needs endorsement by ethnicity, education, and treatment status. All data were analyzed using the Statistical Package of the Social Sciences (SPSS®) – Version 13 software.
A total of 248 patients completed the needs assessment survey. Although an exact account of refusals was not documented, based on retrospective account, an estimated 30% refused participation for various reasons, the most common including illness-related physical discomfort, a language barrier, or time restrictions. In a few cases, emotional discomfort or family refusal precluded participation. Fifty-six patients had extensive missing data on the survey and were therefore excluded from multivariate analyses. This subsample tended to report being off-treatment (χ2(1) = 8.75, P < .003), further out from diagnosis (χ2(3) = 8.54, P < .04), and in remission (χ2(1) = 10.34, P < .001). They did not differ from full-data respondents in regard to age, gender, ethnicity, religion, or marital status. Based on list-wise deletions pertaining to missing data on both predictor variables and subscale scores, the final analyzable sample size for the multivariate analyses reported below ranged from 144 to 159 (dependent upon the needs subscale under analysis).
Among respondents, differences among the ethnic subgroups emerged on 3 sociodemographic/medical variables: education, age, and time since diagnosis. The proportion of patients who had college or graduate education was 60% of whites, 47% of African Americans, 32% of Hispanics, 73% of Others, and 0% Missing (χ2(4) = 17.18, P < .002). Caucasian patients were the oldest (M = 59.15), following by African American (M = 55.62), Hispanic (M = 50.60), Missing (M = 48.50), and then Others (M = 46.20) as the youngest (F(4234) = 4.99, P < .001). A similar pattern of differences for years since diagnosis was observed: Non-Hispanic whites (M = 2.83), African Americans (M = 2.57), Hispanics (M = 2.32), Missing (M = 1.75), and Others (M = 2.00; (F(4201) = 3.23, P < .014). As a result of these differences, education, age, and time since diagnosis were included in subsequent multivariate analyses in predicting needs.
To describe the development of a Psychosocial Needs Assessment Survey, A self-report Psychosocial Needs Assessment survey was developed based on published research33, 35, 36, 41, 50, 51 and the investigators' clinical and research experience with underserved and ethnic minority cancer patients of the Bronx, New York. Special attention was given to the limited literacy of the population and time constraints of a clinic-based assessment method. As such, the presence rather than level of need (which requires Likert-scaling or additional time and processing by the respondent) was the target of this assessment. Guided by strategies for developing cancer education materials for low-literate populations,52 the survey utilized brief and simple statements that were based on common language and displayed in large, generously spaced font.
Four major categories of needs were identified: Informational needs focused on gaining knowledge and help regarding cancer, treatment, symptom management, and lifestyle change. Practical items assessed needs around finances, transportation, family care, and personal assistance. Supportive needs addressed desire for emotional/coping support, stress management, and connection with others. Spiritual needs elicited patients' interest in finding meaning, hope, peace, and spiritual resources, as well as discussing death and dying. Thirty-four items were generated, after which they were translated into Spanish, piloted on 6 patients, and revised to improve comprehension. Questions related to sociodemographic and medical characteristics were included at the end of the survey. On the basis of prior experience with receiving largely incomplete data for income-related inquiries, education was used as an indicator of socioeconomic status.
Patients were instructed to indicate if they would like to know more about, help with, or someone to talk to about each psychosocial issue. Responses were dichotomized as follows: (A) Yes or Yes, but not now and (B) No or Does not apply, and examined individually as well as in summary score format. One Informational item around quitting smoking was dropped from analyses due to missing data for 77% of the sample. Reliability analyses based on the Kuder-Richardson 20 statistic confirmed high internal consistency for each of the Needs Assessment subscales: informational (13 items; KD20 = 0.90), practical (8 items; KD20 = 0.86), supportive (6 items; KD20 = 0.83), and spiritual (6 items; KD20 = 0.90). Subscale correlations ranged from 0.57 to 0.82.
To assess the patterns and predictors of psychosocial needs among an underserved, ethnically-diverse cancer patient population, the percentage of patients who endorsed informational needs ranged from 50% (help with making lifestyle changes) to 91% (what to eat to help with treatment side effects), with the mean rate of endorsement being 77%. Across the 13 items, a mean of 9.54 (±3.40 SD) needs was endorsed. For the 8 practical needs, the mean endorsement rate was 47%, ranging from 35% (help with managing finances) to 79% (help with a place to call with questions/emergencies), with a mean of 3.79 (±2.58 SD) needs endorsed. Among the 6 supportive needs, the mean endorsement rate was 59%, ranging from 52% (help with worries about family) to 76% (know about relaxation/stress management), and a mean of 3.43 (±2.08 SD) items endorsed. For the 6 spiritual needs, the mean endorsement rate was 42%, ranging from 27% (someone to talk to about death and dying) to 52% (help with finding hope), with a mean of 2.52 (±2.36 SD) needs endorsed.
Hierarchical stepwise multiple regression analyses were performed to identify predictors of needs with education, age, time since diagnosis, treatment status, and marital status entered on the first step, and ethnicity entered on step two. Ethnicity emerged as the main predictor of needs for all 4 subscales: informational, R = 0.19, F(1143) = 5.51, P < .02; practical, R = 0.37, F(1145) = 23.14, P < .001; supportive, R = 0.27, F(1157) = 12.50, P < .001, and spiritual, R = 0.28, F(1158) = 12.93, P < .0001. Even after controlling for education, age, time since diagnosis, marital status, and treatment status, ethnicity was the only significant predictor of needs across all domains.
To better understand the ethnic variation with regard to psychosocial needs, and to more closely examine needs by socioeconomic and medical subgroups, Tables 2–5 provide an overview of each subscale's distributions by major ethnic groups, education, and treatment status. As illustrated in Table 2, with regard to informational needs, African American and Hispanic patients reported a mean endorsement rate of 81% and 85% respectively, in comparison to 70% among white patients. Those on treatment reported 79% endorsement versus 70% for those off treatment. Endorsement was similar (75%) among those with or without post-high school education.
Table 2. Informational Needs by Ethnicity, Education, and Treatment Status
|I would like to know more about:|| |
| My type of cancer||73.8||91.2||93.0||83.3||81.1||84.5||75.7|
| Immediate side-effects of treatment||73.7||81.8||86.0||79.5||80.2||79.8||73.9|
| Long-term side-effects of treatment||82.0||82.6||90.0||84.9||84.0||84.8||83.7|
| Information hotlines||62.7||79.6||79.1||51.0||49.0||78.9||57.5|
| Alternative/natural treatments||74.0||89.3||88.6||83.3||81.5||87.2||71.0|
| Genes and cancer||74.0||83.3||94.9||80.2||82.2||84.3||75.5|
| Reduce risk of second cancer||86.0||91.4||93.6||91.2||88.0||90.3||89.9|
| Foods to help with side-effects||87.2||91.5||97.7||92.2||89.7||90.0||90.4|
| Tests to find cancer||62.4||87.3||80.0||69.2||78.3||73.4||67.1|
| Pain management||67.5||83.3||85.0||77.0||76.2||82.0||60.4|
| Healthy eating||75.5||85.0||91.7||82.1||80.2||82.0||78.5|
|I would like help with:|
| Making lifestyle changes||41.3||52.1||68.2||45.5||54.3||48.0||47.9|
| Keeping lifestyle changes||46.9||58.3||63.4||50.5||57.4||61.5||44.4|
|Mean subscale score (range 0–13)||8.67||10.00||10.43||9.49||9.56||9.67||9.02|
Table 3. Practical Needs by Ethnicity, Education, and Treatment Status
|I would like to know more about:|| |
| Transportation to medical visits||38.2||77.8||81.8||64.0||56.2||62.9||46.2|
|I would like help with:|
| Managing finances||20.6||47.9||51.4||33.3||37.1||32.3||35.2|
| Making plans if become ill||49.5||75.5||82.2||61.9||64.6||59.0||64.0|
| Someone with me during tx||17.7||47.7||62.2||45.3||28.2||31.4||34.0|
| Someone with me after tx||17.9||44.4||62.2||43.7||28.9||32.9||32.7|
| Place to call with questions||68.5||88.7||85.4||80.0||77.5||77.8||74.3|
| Food preparation||35.8||66.7||54.8||49.5||46.8||46.6||45.8|
| Family care when I am weak/sick||40.4||56.1||67.5||54.3||50.5||53.6||49.2|
|Mean subscale score (range 0–8)||2.69||4.55||5.10||3.89||3.68||3.62||3.54|
Table 4. Supportive Needs by Ethnicity, Education, and Treatment Status
|I would like to know more about:|| |
| Connecting with other patients||38.0||73.6||63.6||57.9||49.5||56.2||44.6|
| Stress Management||71.2||79.2||84.8||73.6||79.5||72.7||77.0|
|I would like help with:|
| Coping with sadness||48.1||73.6||72.7||65.7||56.5||60.8||56.2|
| Sharing thoughts/feelings||36.8||66.0||72.7||58.1||48.1||49.5||52.6|
| Worries about family||42.5||54.7||70.5||48.0||55.9||56.0||43.8|
| Overcoming my fears||52.4||68.0||72.5||60.6||62.9||60.0||60.0|
|Mean subscale score (range 0–6)||2.83||3.94||4.19||3.44||3.44||3.47||3.17|
Table 5. Spiritual Needs by Ethnicity, Education, and Treatment Status
|I would like to help with:|| |
| Finding spiritual resources||32.7||54.2||69.8||49.5||43.6||47.6||40.6|
| Finding meaning in life||35.5||60.9||60.5||44.3||49.5||46.2||43.7|
| Finding hope||38.7||65.9||72.5||59.6||46.2||55.7||44.6|
|I would like someone to talk to about:|| |
| The meaning of life||21.4||35.1||47.9||33.6||27.1||31.0||28.6|
| Death and dying||18.0||32.1||41.7||28.4||24.8||29.5||19.7|
| Finding peace of mind||38.1||47.5||66.7||52.4||47.6||46.9||43.8|
|Mean subscale score (range 0–6)||1.85||2.96||3.76||2.68||2.36||2.56||2.23|
In Table 3, the endorsement rate of practical needs among African American (63%) and Hispanic (68%) patients was almost twice that of white (36%) patients. The mean endorsement rate was midway between these groups for patients on treatment (50%), off treatment (48%), post-high school educated (49%), and high school or below (54%).
Endorsement of supportive needs was also significantly higher among African Americans (69%) and Hispanics (73%) compared with that of whites (48%), as seen in Table 4. All treatment status and educational groups fell within the 56–61% endorsement range for supportive needs.
On average, twice as many Hispanic (60%) patients reported spiritual needs as white (31%) patients, with African Americans falling somewhere in the middle (49%; Table 5). For spiritual needs, all treatment status and educational groups fell within the 37–45% endorsement range.
This article presents a first look into the self-reported psychosocial needs of the underserved cancer patient. Similar to other needs assessments, patients seemed to have the greatest need for information/education, followed by support.29, 31, 36 Nearly half of the sample reported a minimum of 10 (of 13) informational and 3 (of 6) supportive needs.
Surprisingly, this underserved population reported only moderate levels of practical needs, with half endorsing a minimum of 3 (of 8) needs. Spiritual needs were the least endorsed, as in other studies,36 although nearly one half endorsed 2 or more needs (of a possible 6). Examination of individual items sheds light on salient areas of need that the majority of patients endorsed, including a place to call with questions; help with overcoming fears; knowledge about stress management; making plans if I become ill; and information about cancer, side effects, healthy eating, hotlines, alternative treatments, and other medically related issues.
Across all categories of psychosocial need, ethnicity emerged as the key predictor above and beyond education, diagnosis/treatment factors, marital status, and age. Hispanic and African American cancer patients reported greater informational, practical, supportive, and spiritual needs than non-Hispanic whites. These “need disparities” are not surprising, given the factors that generally contribute to ethnic disparities, including deficient socioeconomic resources and medical access.5 Although the ethnic minority groups in this assessment study were less educated and more recently diagnosed, education and treatment status alone were not predictive of needs. Ethnic minority patients, however, were younger, a factor that may have contributed to their greater needs, as shown in other studies.30
Sociocultural coping factors may also be considered in the understanding of the ethnic differences found in psychosocial needs. Across all domains, Hispanic cancer patients reported the highest number of needs. Previous literature points to the substantially high levels of distress14–17 and use of “venting” as a coping strategy53 among Hispanic cancer patients. This distress-venting reaction may be reflective of a cultural style that fits with the assertive expression of needs. Although African American patients also expressed more psychosocial needs than non-Hispanic white patients, the basis of their needs may be related to a deficiency in social support that is frequently observed.18–20 Whether distress or social factors are underlining their needs, one thing is clear: African Americans and Hispanics are asking for support and services to manage the psychosocial challenges of living with cancer.
The purpose of this needs assessment initiative was to provide an initial roadmap for developing interventions. As such, it has lead to the initiation of a culturally informed psychosocial intervention study geared towards the underserved and ethnic minority cancer patients. Funded by the American Cancer Society, the Mind-Body Cancer Research Program is a randomized intervention study comparing a Psycho-Educational to a Psycho-Spiritual group intervention on QOL and survival among underserved, advanced stage cancer patients. The rationale to direct attention to the spiritual/existential needs of cancer patients is strong on several fronts. First, a number of studies highlight the central role that spirituality/religiosity plays in coping with cancer for ethnic minority and disadvantaged populations.54–63 These cultural variations are mirrored in surveys of complementary medicine use and practices where “spiritual healing” was one of the most common modalities used among African Americans (36–57%) and Hispanics (26%).64, 65 Second, the need for meaning is universal; it provides a sense of purpose and connection in people's lives. Finally, there is some evidence that spiritual and religious beliefs/practices may be positively associated with health outcomes.66, 67
The use of the support group format was based on patients' need for more support, as reported in the current study, as well as the documented use of formal and informal support groups in coping and information gathering among minority cancer patients.68 Further, insufficient social support has been found to be associated with higher distress levels among patients from various cultural backgrounds.64, 69–72
There are a number of limitations to be noted. First, although the needs assessment survey shows strong internal consistency for each of its domains, it is conceivable that a statistically-based factor analysis may have produced alternative need categories. Further, this needs assessment did not examine additional variables that could provide convergent validity, including QOL and social support. Nonetheless, the survey was developed and analyzed based on the initial constructs on which it was derived, the results of which reinforce its face validity. Second, there was a substantial level of missing data. Although missing data can pose significant limitations in validity and generalizability, it can also provide indirect information about the topic at hand. In this survey, those who left many of the survey items blank tended to be longer-term survivors, not on treatment, and in remission. This data pattern may be reflective of these survivors having fewer needs in comparison to patients on active disease. In accordance, other studies have reported high psychosocial needs among recently diagnosed patients28 and those with metastatic disease.39 Third, the use of education as the only socioeconomic status indicator may not be fully informative. Although income may be difficult data to access, information on health insurance would have offered additionally insight into the “underserved” level of this population. The inclusion of this information is suggested for future assessments.
Insufficient attention has been given to the psychosocial needs of ethnic minority and underserved cancer patients.6 The time has come to address these disparities and provide the support and services that such vulnerable patients need. This psychosocial needs assessment provides a first step towards this goal. Future research is needed to tell us what interventions and for whom those interventions are most effective in addressing the information, emotional, practical, and spiritual needs of these patients.