The Carer Support Needs Assessment Tool (CSNAT) for Use in Palliative and End-of-life Care at Home: A Validation Study

Transcription

1 Vol. 46 No. 3 September 2013 Journal of Pain and Symptom Management 395 Original Article The Carer Support Needs Assessment Tool (CSNAT) for Use in Palliative and End-of-life Care at Home: A Validation Study Gail Ewing, BSc, PhD, RGN, Caroline Brundle, BSc, Sheila Payne, BA, PhD, DipN, RN, CPsychol, and Gunn Grande, BA (Hon), MPhil, PhD, on behalf of the National Association for Hospice at Home Centre for Family Research (G.E.), University of Cambridge, Cambridge; School of Nursing, Midwifery & Social Work (C.B., G.G.), University of Manchester, Manchester; and International Observatory on End of Life Care (S.P.), Lancaster University, Lancaster, United Kingdom Abstract Context. Family carers need to be supported in their central role of caring for patients at the end of life, but brief practical tools to assess their support needs have been missing. To address this gap, we developed a brief evidence-based Carer Support Needs Assessment Tool (CSNAT) suitable for everyday practice. Objectives. To assess face, content, and criterion validity of the CSNAT and measure sensitivity to change over time. Methods. Participants were 225 adult carers of patients from six U.K. Hospice Home Care services. Carers were surveyed at baseline and at four-week follow-up using self-completed questionnaires, including CSNAT, standard measures (distress, strain, positive appraisals, preparedness, and global health), help provided with activities of daily living, and patients symptom levels. Qualitative feedback on CSNAT was sought through 10 pilot carer interviews and professional and carer advisory group input. Results. The CSNAT has good face, content, and criterion validity. CSNAT domains comprehensively covered carer support needs. CSNAT scores showed clear and consistent positive correlations with strain and distress and negative correlations with preparedness for caregiving and global health. There also were clear correlations with help with activities of daily living and some relationships with positive appraisals and symptom burden. The CSNAT s sensitivity to change in relevant domains was similar to other measures. Conclusion. The CSNAT is a valid tool for the direct measurement of carers support needs. It combines comprehensiveness of content with feasibility of administration and has utility both as a research tool and a tool for everyday palliative care practice. J Pain Symptom Manage 2013;46:395e405. Ó 2013 U.S. Cancer Pain Relief Committee. Published by Elsevier Inc. All rights reserved. Address correspondence to: Gail Ewing, BSc, PhD, RGN, Centre for Family Research, University of Cambridge, Free School Lane, Cambridge, CB2 3RQ, United Kingdom. ge200@cam.ac.uk Ó 2013 U.S. Cancer Pain Relief Committee. Published by Elsevier Inc. All rights reserved. Accepted for publication: September 19, /$ - see front matter

2 396 Ewing et al. Vol. 46 No. 3 September 2013 Key Words Carers, caregivers, needs assessment, palliative care, validation Introduction Family carers play a central role in palliative and end-of-life (EOL) care, enabling patients to be cared for and to die in the place of their choice. 1,2 As a part of their role, carers frequently take on responsibilities for managing symptoms 3,4 and personal care 5,6 as well as having to attend to routine household duties. 4 However, taking on the role of a carer is associated with considerable cost, including psychosocial distress, 7 impact on health, 8,9 pressure on finances, 10 and, in some cases, increased mortality. 11 Ensuring that carers are adequately supported can ameliorate these effects; 12e14 it is, therefore, important that their support needs are assessed and addressed. 15,16 However, on reviewing the literature before the study, we were unable to identify a suitable existing tool to assess carer support needs in EOL care practice at home; 17,18 most carer research tools found were indirect measures of caregiving difficulty or, if they were needs measures, they were too lengthy to be feasible for use in health care practice. 19e21 Review of practice tools showed a lack of consistent comprehensive approaches to carer assessment used by health care providers themselves. 17 Consequently, we developed an evidencebased tool suitable for use in health care practice that would fill the gap between validated research tools that have little relevance to actual practice and more ad hoc practice-based assessment forms that lack a proper evidence base. Development of the Carer Support Needs Assessment Tool Development of the Carer Support Needs Assessment Tool (CSNAT) is described in detail in an earlier publication. 17 In brief, carers perspectives of key aspects of support needed during the provision of EOL care at home were obtained through qualitative focus groups and interviews with 75 adult bereaved carers who were family members/friends of the patients referred to Hospice at Home services. Carers needs fell into two distinct groups of support areas or domains: support to enable them to provide care for their relative and more direct personal support for themselves. From these qualitative data, 14 support domains were identified and used to structure a direct measure of carers support needs (Table 1). In addition to the 14 support domains, the tool includes one optional additional question to capture anything else not already covered. All questions follow the format: Do you need more support with (e.g., Do you need more support with looking after your own health?). Four response categories (no; a little more; quite a bit more; and very much more) allow carers to indicate support needed with any of the domains. The CSNAT was designed as a screening tool to identify carers key support needs that require further detailed assessment, thus enabling it to be short but comprehensive in approach. Although the CSNAT is evidencebased, specifically developed for EOL care, there is still a need to test whether such a brief, pragmatic tool performs sufficiently well in terms of face, content, and criterion validity and sensitivity to change. The objective of this study was to assess the validity of the CSNAT and its sensitivity to change over time. Methods This was a self-completed survey testing the CSNAT alongside standard measures of impact of caregiving, preparedness, and health, with a sample of carers who were currently caring for a family member/friend towards the end of life. The study was reviewed and approved by the University of Manchester Committee on the Ethics of Research on Human Beings (Reference number 09044). Piloting Initial feedback on face and content validity of the tool was provided by a carer advisory group comprising eight carers, who had participated in the original qualitative study; they

3 Vol. 46 No. 3 September 2013 Carer Support Needs Assessment Tool Validation 397 Table 1 CSNAT Support Domains Support domains to enable the carer to care Do you need more support with.. understanding your relative s illness?. managing your relative s symptoms, including giving medicines?. providing personal care for your relative (e.g., dressing, washing, and toileting)?. knowing who to contact if you are concerned about your relative (for a range of needs, including at night)?. equipment to help care for your relative?. talking with your relative about his or her illness?. knowing what to expect in the future when caring for your relative? More direct support domains for carers themselves Do you need more support with.. having time for yourself in the day?. your financial, legal, or work issues?. dealing with your feelings and worries?. looking after your own health (physical problems)?. your beliefs or spiritual concerns?. practical help in the home?. getting a break from caring overnight? CSNAT ¼ Carer Support Needs Assessment Tool. commented through either or telephone discussions on the content of the CSNAT domains, wording of items, and the tool s presentation. A similar feedback procedure took place at a meeting of the Research Steering Group comprising health-care professionals from the National Association for Hospice at Home (NAHH). Procedures and study materials were then piloted in one Hospice Home Care (HHC) service. In the U.K., HHC services are specialist providers of care and support to palliative care patients and their carers at home. Ten participants in the pilot study (current carers) were interviewed by telephone (by G. E.) to gain further feedback on the tool. They were asked for their views on the CSNAT items, any support areas not covered by the tool, whether they had any difficulty with the scoring system, and whether they had been upset by completing the tool. Survey Recruitment and Data Collection Procedures Six HHC services within the NAHH participated. Service administrators identified all active cases in their databases and subsequent new referrals plus the designated main carer for each patient until a minimum sample size of 200 carers was achieved. A study recruitment pack (including consent forms and questionnaire booklet) was assigned to each carer. All adult carers (aged 18 years or older) were eligible for the study unless the service had concerns about their ability to cope with research because of exceptionally high levels of distress. Each service reviewed all identified carers before the packs were distributed. Piloting had demonstrated the need for flexible distribution arrangements: services could choose for nurses to hand over the packs at home visits or to mail the packs directly to carers. Each service kept a recruitment log monitoring distribution and any reasons for exclusion. On receipt of a study pack, participating carers completed the questionnaire booklet and consent forms and returned them directly to the researcher. The researcher then sent a follow-up questionnaire booklet to the carer four weeks thereafter. Deaths were monitored throughout data collection to ensure that follow-up booklets were not sent to carers of patients who had died. A final audit of patient deaths also was conducted by each participating service at the end of the study. Data Collected Patient and carer demographic information was collected at baseline. The questionnaire booklet comprising the CSNAT and standard measures was administered both at baseline and follow-up. Need for more support on the CSNAT was scored as: no (0), a little more (1), quite a bit more (2), and very much more (3). Three standard measures were administered: subscales of strain, distress, and positive appraisals from the Family Appraisal of Caregiving Questionnaire; 22 the Preparedness for Caregiving Scale; 23 and the Global Health/Quality of Life scales from the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire-Core Information on help provided for patients with activities of daily living (ADLs) was collected through six set questions and information on patients symptoms through seven set questions on how much the patient has been bothered by symptoms (see Tables 4 and 5 for items). Response categories were not at all (0); a little (1); quite a bit (2); and very much (3). In addition to the information collected through the questionnaire booklets, anonymized

4 398 Ewing et al. Vol. 46 No. 3 September 2013 information on the patient s gender, age, diagnosis and date of death, and the carer s gender and age was obtained from the HHC database records for the study period to enable the assessment of representativeness of the study participants. Statistical Analysis To understand how representative participants were of the total sample approached, differences in demographic variables between participating and nonparticipating carers and patients were investigated using Chi-squared tests for categorical variables and t-tests for interval variables. Survey scores and percentage of carers indicating the need for support on the CSNAT were explored to assess content validity. To test criterion validity, Spearman s rank correlation was used to investigate the relationships between CSNAT items and standard measures, help given with ADLs, and patients symptom levels. Nonparametric tests were used because all CSNAT and most of the other variables were ordinal and/or skewed. For sensitivity to change, data from carers who completed both booklets were analyzed for changes between baseline and follow-up in CSNAT variables, standard measures, symptoms, and help with ADLs. Nonparametric Wilcoxon signed-rank tests were used, with the exception of the Family Appraisal of Caregiving Questionnaire for Palliative Care strain and distress subscales. Responses to these scales were normally distributed, so paired sample t-tests were used. Results Sample Six participating HHC services had 1559 referrals during the study period. The main carers of 953 (61.1%) patients were invited to participate, and 606 (38.9%)patients were excluded. The main reasons for exclusions included patient death before the study started or during the recruitment period (26%), high levels of carer distress (19%), patient discharged from service/referral not accepted (16%), and no identified main carer (13%). Of those invited, 235 (24.7%) agreed to participate. Ten carers were found to have completed the study booklet shortly after the patient s death and were omitted from the analyses; the study was aiming to assess support needs and experiences of current carers. Of the 225 carers included in the analysis, 141 (62.7%) completed both baseline and followup booklets. The 84 who did not complete the follow-up included 49 cases, in which the patient had died and a follow-up booklet, therefore, was not sent. A larger proportion of patients of participating carers were still alive 19 weeks after study entry than patients of nonparticipating or excluded carers (52.1% vs. 45.7% and 29.5%, respectively; c 2 ¼ 38.38; df ¼ 2; P < ). Patients of participating carers were less likely to be aged 85 years or older compared with patients of nonparticipating or excluded carers (c 2 ¼ 17.37; df ¼ 8; P ¼ 0.026). Otherwise, no differences were found in available variables, suggesting participants were reasonably representative of the total sample approached. Sample details are provided in Table 2. Face and Content Validity Initial User Feedback on the CSNAT From the Carer Advisory Group. Feedback from the carer advisory group was very positive on content of the CSNAT domains; the group felt that all main issues were covered: I think you ve pitched it right, it hits the button for support needs. One group member said she felt she could have been heard through this. The group identified the importance of keeping the tool simple, with short items. Not having too many words was important at this time and short sentences made it accessible. The group also offered specific feedback on wording and presentation. For example, knowing who to contact was very relevant, but it also was noted that different contacts may be required for different needs and the wording should reflect this. In the personal care domain, providing examples in the format of dressing or bathing was thought to suggest that providing personal care was limited to only these activities. Similar feedback also came from health care professionals. Feedback From NAHH Research Steering Group. The group appreciated the simplicity of the tool for use in practice. Comprehensiveness of the domains covered also was noted. The

5 Vol. 46 No. 3 September 2013 Carer Support Needs Assessment Tool Validation 399 Table 2 Study Sample Carer Participants (N ¼ 225) Female 148 (65.8%) Married/living as married 206 (92.8%) White 223 (99.6%) Retired 138 (61.3%) Other caring responsibilities 47 (21.1%) Provides practical or physical support 205 (91.9%) Mean SD Age Age completed full-time education 17 5 How many months provided practical or physical support Patient Characteristics Median IQR Relationship to carer Spouse or partner 179 (79.9%) Parent 38 (16.9%) Cancer diagnosis 192 (86.9%) Mean SD Age Median IQR How many months been ill IQR ¼ interquartile range. 82 years and all had a cancer diagnosis. Interviews indicated that the CSNAT was straightforward to complete; carers found questions clear and had no problems with scoring. They also reported that the CSNAT support domains were relevant to their situation: I think that you ve got pretty well everything that you would think about really (ID 6062). None of the carers was upset by the content of any of the CSNAT questions. For the pilot study, the CSNAT items on support to enable them to care for the patient were grouped together on one page and the items on personal support for carers themselves were similarly grouped on the next. This appeared to make some carers dismiss support for themselves as a whole by completing a set of no responses for each item on the personal support page. Therefore, CSNAT items were subsequently mixed to encourage consideration of all support domains. group agreed on the wording of 12 items and advised revision of two items, extending examples given in the personal care item to include toileting and rewording talking about spiritual concerns to beliefs or spiritual concerns. Carers Views on the CSNAT From the Pilot Study. Ten carers provided feedback on the CSNAT: seven women and three men, aged 21e81 years, caring for a spouse/partner (5), parent (3), sibling (1), or son-in-law (1). The age of those whom they cared for ranged from 23 to Responses to CSNAT Items at Baseline From Main Study. Fig. 1 shows scores of 14 CSNAT items at baseline ordered according to expressed need (i.e., any scoring of need for more support other than no ). Most carers wanted more support with knowing what to expect in the future (64.3%) and dealing with their feelings and worries (52.7%). Just less than half wanted more support with having time for themselves in the day and with understanding their relative s illness. All items were used, indicating no item was superfluous. Although only a small percentage of carers wanted more Fig. 1. Percentage of carers expressing need for more support with each Carer Support Needs Assessment Tool domain (N ¼ 216e222).

6 400 Ewing et al. Vol. 46 No. 3 September 2013 support with beliefs and spiritual concerns (6.5%), this still represented 14 carers in the sample. Where carers wanted more support, most only wanted a little more support. Nevertheless, the CSNAT was able to identify 63 (28.5%) carers who wanted quite a bit or very much more help regarding: what to expect in the future, 40 (18.1%) daytime respite, 37 (16.7%) own feelings and worries, and 35 (15.8%) understanding their relative s illness. A final optional question asked carers if there was anything else they needed help with, not covered in the 14 items. Only five (5.3%) carers identified an additional need that was not fully covered by the CSNAT, but these were nevertheless very closely related to the existing CSNAT items: information on services available in the area, help understanding all agencies involved in care and support, weekend respite, preparation for bereavement, and help providing emotional support to the patient. Thus, existing items comprehensively covered carer support needs. Criterion Validity If the CSNAT is measuring something meaningful and relevant within the caregiving context, it also should be significantly related both to the measures of how the carer is feeling and to the patient s underlying illness burden. Table 3 shows the relationship between CSNAT items and scores on standard measures. All CSNAT items were negatively correlated with preparedness for caregiving; that is, the greater the need for support the less prepared the carer felt for the caregiving role. All CSNAT items were positively correlated with caregiver strain and distress; that is, the greater the need for support the more strain and distress the caregiver felt. All but two CSNAT items were negatively correlated with quality of life; the greater the need for support the worse the quality of life. Negative correlations between positive caregiving appraisals and some CSNAT items were weaker but still significant (P < 0.05). CSNAT items also were positively correlated with patients need for help with ADLs (Table 4), particularly need for support with personal care, equipment, and overnight and daytime respite, followed by the need for practical help in the home. Items that showed Table 3 Correlations Between CSNAT Items and Standard Measures of Preparedness for Caregiving, Health-Related Quality of Life, Caregiver Strain, Caregiver Distress and Positive Appraisals of Caregiving (N ¼ 215e222) CSNAT Items Preparedness for Caregiving Global Health/Quality of Life Caregiver Strain Caregiver Distress Positive Caregiving Appraisals Understanding relative s illness (P < ) (P ¼ 0.005) (P < ) Time for yourself in the day (P < ) (P < ) (P < ) (P < ) (P ¼ 0.019) Managing relative s symptoms (P < ) (P < ) (P < ) (P < ) (P ¼ 0.017) Financial, legal, or work issues (P ¼ 0.014) (P ¼ 0.001) (P < ) (P ¼ 0.018) Providing personal care (P < ) (P < ) (P < ) (P < ) (P ¼ 0.008) Dealing with your feelings and worries (P < ) (P < ) (P < ) (P < ) Knowing who to contact if concerned (P < ) (P ¼ 0.050) (P < ) (P < ) Looking after your own health (P ¼ 0.001) (P < ) (P < ) (P < ) Equipment to help care for your relative (P ¼ 0.026) (P ¼ 0.010) (P ¼ 0.008) (P < ) Your beliefs or spiritual concerns (P ¼ 0.018) (P ¼ 0.004) (P ¼ 0.019) Talking with relative about their illness (P < ) (P ¼ 0.002) (P < ) (P < ) Practical help in the home (P < ) (P < ) (P < ) (P < ) (P ¼ 0.014) Knowing what to expect in the future (P < ) (P < ) (P < ) (P < ) (P ¼ 0.048) Getting break from caring overnight (P < ) (P < ) (P < ) (P < ) CSNAT ¼ Carer Support Needs Assessment Tool. Significant Spearman s rank correlations, two-tailed test, all relationships at P # 0.05 reported.

7 Vol. 46 No. 3 September 2013 Carer Support Needs Assessment Tool Validation 401 Table 4 Correlations Between CSNAT Items and Activities of Daily Living (N ¼ 202e212) CSNAT Items Eating Dressing Walking Washing Toileting Getting Up at Night Understanding relative s illness (P ¼ 0.047) (P ¼ 0.027) Time for yourself in the day (P < ) (P ¼ 0.006) (P ¼ 0.016) (P < ) (P ¼ 0.006) Managing relative s symptoms (P < ) (P ¼ 0.026) (P ¼ 0.002) Financial, legal, or work issues Providing personal care (P < ) (P < ) (P ¼ 0.023) (P < ) (P < ) (P ¼ 0.001) Dealing with your feelings and worries (P ¼ 0.039) Knowing who to contact if concerned (P ¼ 0.014) (P ¼ 0.020) Looking after your own health (P ¼ 0.037) (P ¼ 0.004) (P < ) Equipment to help care for your relative (P ¼ 0.033) (P ¼ 0.001) (P ¼ 0.008) (P < ) (P < ) (P ¼ 0.047) Your beliefs or spiritual concerns Talking with relative about their illness Practical help in the home (P ¼ 0.001) (P ¼ 0.004) (P ¼ 0.009) (P ¼ 0.026) Knowing what to expect in the future Getting break from caring overnight (P < ) (P < ) (P < ) (P < ) (P < ) CSNAT ¼ Carer Support Needs Assessment Tool. Significant Spearman s rank correlations, two-tailed test, all relationships at P # 0.05 reported. weaker relationships with ADLs were dealing with feelings and worries, knowing who to contact, understanding of the illness, managing symptoms, and looking after own health. Table 5 shows the relationship between the CSNAT and patients symptom levels. Nearly all CSNAT items showed positive correlations (mostly at P < 0.01) with patients level of anxiety and/or depression. Apart from anxiety and depression, only pain and constipation were positively associated with the need for more support with symptom control. Pain also was associated with more need for support with feelings and worries and, to a lesser extent, to knowing what to expect and providing personal care. Apart from the mental health variables, breathlessness and constipation were the symptoms most closely correlated with support needs. Sensitivity to Change This analysis considers whether the CSNAT is sufficiently sensitive to pick up changes over time; it relates to 141 carers who completed both baseline and follow-up data collection booklets. However, not all carers responded to all questions. Table 6 shows significant changes from baseline to follow-up both for the CSNAT and accompanying measures. The CSNAT picked up increases in need for support with providing personal care and with financial, legal, or work issues. There was a corresponding increase in caregiver strain and in the patient s need for help with walking and toileting and the symptom of diarrhea. This analysis suggests that the CSNAT is able to pick up change in relevant domains to a similar extent as other measures. As caregiver strain increases and care-demanding patient problems such as mobility, toileting, and diarrhea increase, there is also registration within the CSNAT of increased need for support with personal care and financial, legal, and work issues that may be expected to accompany increased effort and resources spent on caring. Discussion This article reports on the validation of the CSNAT in hospice home care with 225 carers currently providing care and support to

8 402 Ewing et al. Vol. 46 No. 3 September 2013 Table 5 Correlations Between CSNAT Items and Patients Symptoms (N ¼ 192e207) CSNAT Items Pain Nausea/Vomiting Constipation Diarrhea Breathlessness Anxiety Depression Understanding relative s illness (P ¼ 0.007) (P ¼ 0.021) Time for yourself in the day (P ¼ 0.003) (P ¼ 0.043) Managing relative s symptoms (P < ) (P ¼ 0.008) (P < ) (P ¼ 0.006) Financial, legal, or work issues (P ¼ 0.041) (P ¼ 0.040) (P ¼ 0.025) Providing personal care (P ¼ 0.029) (P ¼ 0.017) (P ¼ 0.019) (P ¼ 0.002) (P ¼ 0.001) Dealing with your feelings and worries (P ¼ 0.002) (P ¼ 0.040) (P ¼ 0.001) (P < ) (P < ) Knowing who to contact if concerned (P ¼ 0.004) (P ¼ 0.002) Looking after your own health (P ¼ 0.027) (P ¼ 0.019) (P ¼ 0.005) Equipment to help care for your relative (P ¼ 0.008) Your beliefs or spiritual concerns (P ¼ 0.026) (P ¼ 0.047) Talking with relative about their illness (P ¼ 0.050) (P ¼ 0.003) (P < ) (P < ) Practical help in the home (P ¼ 0.006) (P ¼ 0.046) (P ¼ 0.001) (P < ) Knowing what to expect in the future (P ¼ 0.025) (P < ) (P < ) Getting break from caring overnight (P ¼ 0.036) (P ¼ 0.004) (P < ) (P < ) CSNAT ¼ Carer Support Needs Assessment Tool. Significant Spearman s rank correlations, two-tailed test, all relationships at P # 0.05 reported. a family member/friend toward the end of life. Findings demonstrate that CSNAT has good face, content, and criterion validity and shows sensitivity to change over time. Grande et al. 12 have identified as a priority validated tools for eliciting, assessing, and monitoring carers ongoing needs. The CSNAT meets this requirement: it is an evidence-based, validated measure of carer support needs and as such has potential for use in both research and practice. Its value lies in it being a direct measure of carers support needs, whereas other carer tools are more often measures of caregiving difficulty and thus, only broad indicators of carers need for support. 25 Face validity, particularly in a tool to be used in practice, is important and should never be treated as trivial. 26 Qualitative feedback on acceptability of the tool and its items came from current carers, those who had previously been carers, and health care professionals from hospice home care. Content validity is demonstrated through survey findings. Scores of need for support with the 14 CSNAT domains showed that none of the items were superfluous and the existing items comprehensively covered carer support needs, indicating that CSNAT has good content validity. The 14 tool items encompass assessment of physical, practical, social, financial, psychological, and spiritual support needs that current policy guidance indicates should be delivered to carers during EOL care. 15,16 Criterion validation used standard measures, 22e24 information about patients symptoms, and help provided for patients with ADLs. Clear correlations between the CSNAT items and other variables that should be related to carer support needs suggest that CSNAT has good criterion validity. CSNAT scores showed clear and consistent correlations with preparedness for caregiving, healthrelated quality of life, caregiver strain and distress, and also showed clear correlations with patients need for help with ADLs. There was also some correlation between CSNAT items and patients symptoms, but these patterns were less consistent. In particular, findings on anxiety and depression should be interpreted with caution because research suggests that carers assessment of patients mental state may relate more closely to how the carer is feeling than how the patient is

9 Vol. 46 No. 3 September 2013 Carer Support Needs Assessment Tool Validation 403 CSNAT Items and Accompanying Measures Table 6 Changes Between Baseline and Follow-up at P # 0.05 Follow-up Mean (SD) Baseline Mean (SD) (Four Weeks After Baseline) Z/t P N CSNAT items Providing personal care 0.22 (0.50) 0.33 (0.65) Z ¼ Financial, legal, or work issues 0.34 (0.73) 0.47 (0.88) Z ¼ Standard measures Caregiver strain 2.85 (0.82) 2.94 (0.77) t ¼ ADLs Help with walking 0.92 (1.01) 1.08 (1.05) Z ¼ Help with toileting 0.65 (1.05) 0.87 (1.17) Z ¼ Symptoms Constipation 0.83 (1.03) 0.66 (0.98) Z ¼ Diarrhea 0.43 (0.76) 0.65 (0.91) Z ¼ CSNAT ¼ Carer Support Needs Assessment Tool; ADL ¼ activities of daily living. CSNAT, symptoms and ADLs: Wilcoxon signed rank test; caregiver strain: paired sample t-test; all relationships at P # 0.05 reported. feeling. 27 These correlations, therefore, may be due to a third underlying variable, for example, carer distress that affects both support need and perceptions of patients mental state. Breathlessness may both be a sign of deterioration and an anxiety-provoking symptom. Constipation also poses challenges, although its correlation with need for more support with spiritual concerns may possibly be ascribed to statistical chance. On the whole, patterns of relationships between patient symptoms and carer support needs are less convincing than for other variables considered. The CSNAT was not subjected to reliability testing. Reliability in terms of internal consistency is not appropriate for CSNAT because the intention is not to derive a total score with items all contributing to a particular construct. 28,29 Test-retest reliability, which assesses the stability of a measure over time and where the time period between measurements is critical, is problematic in palliative care where rapid and sudden changes in the care situation can take place. Thus, instead, sensitivity of the CSNAT to change over time was examined. If standard measures of caregiving and measures of the patient s underlying disease are able to pick up changes over time, but the CSNAT fails to do so, we may question CSNAT s ability to register change. Findings indicate that CSNAT was able to pick up change in relevant domains to a similar extent as other measures. It must be noted, however, that this was a relatively crude analysis, given that we had limited information on quality and level of service support provided. It also is possible that the patient s condition may deteriorate, yet CSNAT scores remain the same, depending on the level of service input provided. Additional work is needed to explore these relationships further. Although CSNAT has demonstrated good validity and sensitivity to change, there are limitations in both the samples and analyses carried out. Patients of participating carers had better survival and were less likely to be aged 85 years or older, but otherwise the included carers appeared representative of the total sample approached. Even with advances in research methodology, challenges remain in recruiting to palliative care studies, 30 particularly gatekeeping by practitioners. 31 The study sample, however, does reflect national patterns in which carers are predominantly women, spouses, often retired, and caring for older patients 32 and in which people with nonmalignant disease and from black and ethnic minorities are underrepresented within palliative care. 33 In particular, further work is indicated to test the CSNAT with carers of patients with other lifelimiting conditions, such as end-stage respiratory or heart disease. Further aspects of analysis also have been explored. 34 In conclusion, we have demonstrated that CSNAT is a valid tool for direct measurement of carers support needs. Its great strength is its short screening format, which combines comprehensiveness of content with feasibility of administration. For the EOL care research community, the CSNAT has utility as a research tool for administration alongside other standard measures, allowing examination of how support needs impact on caregiving experiences; its 14-item content can be administered with minimal additional burden to carer participants. Crucially, the CSNAT is also feasible for carer assessment in everyday practice. It

10 404 Ewing et al. Vol. 46 No. 3 September 2013 provides practitioners working in palliative and EOL care with an evidence-based tool with which to engage carers and together to assess their support needs. For successful integration into practice, however, there are additional issues to address beyond the development and validation of a tool. These include identifying how and when CSNAT is best introduced to carers, how to integrate action plans with the assessment process, and addressing mechanisms for recording and reviewing assessments completed. Work on integrating the CSNAT into hospice home care practice is currently ongoing and will be reported in due course, including translation of lessons learned about implementation into an education and training package for services. Disclosures and Acknowledgments This work was supported by a grant from Dimbleby Cancer Care Research Fund. The funders were not involved in the conduct of the study or in the development of this article. All authors declare no conflicts of interest. The authors thank the carers who gave their time to take part in the study and the nursing and administrative staff of the participating HHC services. They are grateful to the developers of the Family Appraisal of Caregiving Questionnaire (Dr. Brian Cooper) and the Preparedness Scale (Professor Patricia Archbold) for permission to use the scales in this study. The authors benefited from advice from a number of sources and wish to thank our Carer Advisory Group; the Research Steering Group from the NAHH; Professor John Rust, The Psychometrics Centre, University of Cambridge; and Professor Susan Golombok and Helen Statham, Centre for Family Research, University of Cambridge. For further information about use of the CSNAT, please contact Dr. Gail Ewing at the University of Cambridge (ge200@cam.ac.uk) or Professor Gunn Grande at the University of Manchester (gunn.grande@manchester.ac.uk). References 1. Gomes B, Higginson IJ. Factors influencing death at home in terminally ill patients with cancer: systematic review. BMJ 2006;332:515e Grande G, Ewing G. Death at home unlikely if informal carers prefer otherwise: implications for policy. Palliat Med 2008;22:971e Armes PJ, Addington-Hall JM. Perspectives on symptom control in patients receiving community palliative care. Palliat Med 2003;17:608e Aranda SK, Hayman-White K. Home caregivers of the person with advanced cancer: an Australian perspective. Cancer Nurs 2001;24:300e Bee PE, Barnes P, Luker KA. A systematic review of informal caregivers needs in providing homebased end-of-life care to people with cancer. J Clin Nurs 2008;18:1379e Stajduhar KI. Examining the perspectives of family members involved in the delivery of palliative care at home. J Palliat Care 2003;19:27e Payne S, Smith P, Dean S. Identifying the concerns of informal carers in palliative care. Palliat Med 1999;13:37e Mangan PA, Taylor KL, Yabroff KR, Fleming DA, Ingham JM. Caregiving near the end of life: unmet needs and potential solutions. Palliat Support Care 2003;1:247e Stenberg U, Ruland CM, Miaskowski C. Review of the literature on the effects of caring for a patient with cancer. Psychooncology 2010;19:1013e Aoun SM, Kristjanson LJ, Currow DC, Hudson PL. Caregiving for the terminally ill: at what cost? Palliat Med 2005;19:551e Schulz R, Beach SR. Caregiving as a risk factor for mortality: the Caregiver Health Effects Study. JAMA 1999;282:2215e Grande G, Stajduhar K, Aoun S, et al. Supporting lay carers in end of life care: current gaps and future priorities. Palliat Med 2009;23:339e Rossi Ferrario S, Cardillo V, Vicario F, Balzarini E, Zotti AM. Advanced cancer at home: caregiving and bereavement. Palliat Med 2004;18: 129e Christakis NA, Iwashyna TJ. The health impact of health care on families: a matched cohort study of hospice use by decedents and mortality outcomes in surviving, widowed spouses. Soc Sci Med 2003;57: 465e National Institute for Clinical Excellence. Guidance on cancer services. Improving supportive and palliative care for adults with cancer. The manual. London, UK: National Institute for Clinical Excellence, Department of Health. End of life care strategy: Promoting high quality care for all adults at the end of life. London, UK: Department of Health, Ewing G, Grande G. Development of a Carer Support Needs Assessment Tool (CSNAT) for endof-life care practice at home: a qualitative study. Palliat Med 2013;27:244e256.

11 Vol. 46 No. 3 September 2013 Carer Support Needs Assessment Tool Validation Hudson PL, Trauer T, Graham S, et al. A systematic review of instruments related to family caregivers of palliative care patients. Palliat Med 2010;24:656e Hileman JW, Lackey NR, Hassanein RS. Identifying the needs of home caregivers of patients with cancer. Oncol Nurs Forum 1992;19:771e Osse BHP, Vernooij-Dassen MJFJ, Schade EM, Grol RPTM. Problems experienced by the informal caregivers of cancer patients and their needs for support. Cancer Nurs 2006;29:378e Salmon JR, Kwak J, Acquaviva KD, Egan KA, Brandt K. Validation of the caregiving at life s end questionnaire. Am J Hosp Palliat Care 2005;22: 188e Cooper B, Kinsella GJ, Picton C. Development and initial validation of a family appraisal of caregiving questionnaire for palliative care. Psychooncology 2006;15:613e Archbold PG, Stewart BJ, Greenlick MR, Harvath T. Mutuality and preparedness as predictors of caregiver role strain. Res Nurs Health 1990; 13:375e Aaronson NK, Ahmedzai S, Bergman B, et al. The European Organization for Research and Treatment of Cancer QLQ-C30: a quality-of-life instrument for use in international clinical trials in oncology. J Natl Cancer Inst 1993;85:365e Stajduhar KI, Funk L, Toye C, et al. Part 1: homebased family caregiving at the end of life: a comprehensive review of published quantitative research (1998e2008). Palliat Med 2010;24:573e Rust J, Golombok S. Modern psychometrics: The science of psychological assessment, 3rd ed. London, UK: Routledge, 2009: McPherson CJ, Wilson KG, Lobchuk MM, Brajtman S. Family caregivers assessment of symptoms in patients with advanced cancer: concordance with patients and factors affecting accuracy. J Pain Symptom Manage 2008;35:70e Kottner J, Streiner DL. Internal consistency and Cronbach s alpha: a comment on Beeckman et al. Int J Nurs Stud 2010;47:926e Cronbach LJ, Shavelson RJ. My current thoughts on coefficient alpha and successor procedures. Educ Psychol Meas 2004;64:391e Christakis NA. Advances in palliative care research methodology. Palliat Med 2006;20:725e Ewing G, Rogers M, Barclay S, et al. Recruiting patients into a primary care based study of palliative care: why is it so difficult? Palliat Med 2004;18: 452e Buckner L, Yeandle S. Older carers in the UK. London: Carers UK, Walshe C, Todd C, Caress A, Chew-Graham C. Patterns of access to community palliative care services: a literature review. J Pain Symptom Manage 2009;37:884e Grande G, Ewing G, Sawatzky R. Supporting family carers: the relative importance of different support domains in explaining negative and positive impacts from caregiving. [abstract]. Palliat Med 2012;26:408.