Assessment of Needs of Hospitalized Cancer Patients with Advanced Cancer

Background: Needs assessment of patients with advanced cancer (ACPs) is essential for optimal care. This study evaluated the psychometric properties of the Needs Evaluation Questionnaire (NEQ) and assessed the supportive care needs of hospitalized ACPs with solid tumors. Methods: The validated Greek version of the NEQ along with demographic and clinical data of 95 consecutive breast, colon and lung ACPs hospitalized in the University Hospital of Heraklion, Crete, Greece, were used to assess their supportive care needs. The NEQ score was subsequently rescaled to 0-100. Results: NEQ displayed adequate psychometric properties in validity and reliability tests. The average number of needs reported was 8.4(4.1). Female and younger patients reported a higher score of unmet needs than their male (40.3 versus 30.0, p=0.005) and elder (40.1 versus 29.2, p=0.001) counterparts. Patients reported higher needs in receiving information about their future (73.7%), treatments (56.8%), examinations (51.6%) and for the need "to speak with people who had the same experience" (53.7%). In contrast, lower scores were observed in the assistance and treatments needs regarding intimacy (11.6%), "better attention from nurses" (15.8%), "more help with eating, dressing, and going to the bathroom" (18.9%). Lung ACPs reported more assistance and treatment needs than colon and breast ACPs (p<0.05). Conclusions: ACPs reported many unmet needs, mainly informational, that were related to gender, age, and type of cancer. NEQ is a useful tool in everyday clinical practice for obtaining information for supportive care needs. Health care personnel has to address these needs for implementing effective patient-centered care.


Introduction
The optimal supportive care of patients with advanced cancer (ACPs) aims to maximize the benefits of different treatments and interventions (Glare, 2013;Peppercorn et al., 2011).This can be achieved with effective communication between patients, their caregivers and health professionals, accurate assessment of patients' needs, alleviation of symptoms, improvement of everyday functionality and quality of life, and patients' satisfaction (Bredart et al., 2013;Abernethy, Wheeler, & Currow, 2010).
Continuous, systematic and individualized needs assessment is the core point in planning ACPs care (Richardson, Medina, Brown, & Sitzia, 2007).This care usually focuses on the physical, psychological, social, sexual, spiritual, informational, financial and other daily living needs of patients and their important ones (Moghaddam, Coxon, Nabarro, Hardy, & Cox, 2016;Ugalde et al., 2012;Pearce, Coan, Herndon, Koenig, & Abernethy, 2012;Marcusen, 2010;Bahti, 2010;Harrison et al., 2009).Effective needs' assessment requires an in depth and increased patient's involvement.Patients' perceptions and views contribute to better identify the type of assistance required and adaptation of interventions according to ACPs' individual needs (Pearce et al., 2012;Carey et al., 2012;Elkin, Kim, Casper, Kissane, & Schrag, 2007).Many tools assessing the different and diverse supportive care needs of ACPs have been developed (Carlson, Waller, & Mitchell, 2012;Abernethy et al., 2010;Richardson et al., 2007).Some of them are quite specialized and extended, covering most of the ACPs needs.However, these tools often don't take into account the limited or non-existing palliative care services, the high workload and the reduced resources of Greek hospitals (Efstathiou, Ameen, & Coll, 2008).In such environments the optimum choice would be the implementation of short, easy-to-use questionnaires in the daily clinical practice in order to screen and document urgent or unmet needs.
Limited research on the unmet ACPs needs, especially regarding their psychological distress, symptom management, as well as the burden of caregivers, has been conducted in Greece (Govina et al., 2015;Alexopoulos, Koutsogiannou, Moratis, Mestousi & Jelastopulu, 2011;Mystakidou et al., 2006).Thus, the aim of the present study was the assessment of supportive care needs with the Greek version of Needs Evaluation Questionnaire in a group of hospitalized ACPs with solid tumors.

Translation of the NEQ
After obtaining the permission for using the NEQ, its translation and cultural adaptation was performed according to the Minimal Translation Criteria (Medical Outcomes Trust, 1997).Based on this procedure, two independent bilingual translators, native speakers of Greek with advanced knowledge of the Italian language, translated the questionnaire into Greek (forward translation).With the contribution of a third reviewer, a reconciliation meeting was conducted to develop a consensus version (reconciliated Greek version).This version was then back-translated by a native Italian speaker, not familiar with the original questionnaire.Using all the information from the two phases and after thorough discussion with translators and reviewers, the first author acting as adjudicator approved the final version of the NEQ.Translation and back translation did not reveal substantial problems and the research team solved very minor differences.As the last step of the translation procedure, a cognitive debriefing process was used for the cultural adaptation of the questionnaire (Medical Outcomes Trust, 1997).As part of the cultural adaptation process, the questionnaire was administered to five ACPs and in-depth interviews were implemented regarding the respondents' level of understanding of the questionnaire with the purpose of revealing inappropriately interpreted items and translation alternatives.The participants expressed their views regarding the clarity of each item, the relevance of the content to their situation, the comprehensiveness of the instructions and their ability to complete it on their own.They were also encouraged to make suggestions whenever necessary.The comments made by the participants were discussed in a debriefing session ending to the final Greek version of the NEQ.

Test-retest reliability
Test-retest reliability was measured in 8 ACPs with an interval of one week.From the affirmative answers of the 23 needs (Yes=1, No=0) it became the aggregation and calculation of the total concentration for each patient's needs (range 0-23).The test-retest analysis did not reveal any significant change in the mean number of needs in the pre and post testing period of seven days (Wilcoxon signed ranks test: 9.38±5.53vs. 9.63±5.98,respectively, p=0.414 and Spearman's correlation coefficient: r=0.994, p<0.001).

Reliability analysis
Out of the 95 ACPs, 91 patients fully completed the NEQ and four of them provided missing responses in four items (see Table 2 for missing data).The missing values were then replaced by the mode values in each of the four items.The reliability coefficient of Kuder -Richardson was also estimated giving an estimation of r20 equal to 0.752.

Validity
An exploratory factor analysis (EFA) was performed to determine the structure of the questionnaire and how the individual items grouped together.Principal component analysis (PCA) was used with and Varimax with Kaiser normalization as the rotation method (oblique).Kaiser-Meyer-Olkin (ΚΜΟ) measure of sampling adequacy was estimated at 0.628 (middling/mediocre fit) and the Bartlett's test of sphericity was employed to determine the appropriateness of the dataset for factor analysis and to verify the degree of interrelation between the items (χ 2 =454.5, d.f.=253, p<0.001).Eight factors (groups of needs) were found as reported by the developers of the NEQ (Tamburini et al., 2000), with eigenvalues ≥1.00 and 61.9% of total variance explained.A minimum loading criterion of 0.40 was adopted for scale items to be maintained in each factor.

Final Questionnaire
In the final questionnaire, in addition to NEQ, a number of demographic questions regarding the patients' gender, marital status, place of residence, year of birth, and education level were included.Clinical data including patient's disease, presence or absence of metastases, time of initial diagnosis, number of hospital admissions in the last two months and type of treatments were extracted from the medical files.Satisfaction with care was recorded in a scale from 0 (not satisfied) to 10 (very satisfied).The mean duration of completing the questionnaire was 10-15 minutes.

Participants
Ninety five consecutive breast, lung and colon ACPs cared for as inpatients (41.1%) or at the day clinic (58.9%) of the Department of Medical Oncology of the University Hospital of Heraklion, Crete, Greece, were enrolled between November 2007 -February 2008.The sample size was based on the number of in-and day clinic patients cared for, during the month preceding the research.They were randomly selected according to the number of the three common types of cancer of the two services (multistage proportional stratified probability sampling).The inclusion criteria were age over 18 years, ability to speak and write in Greek, treatment with at least one chemotherapy cycle or at least two months' time period from the disease's diagnosis and granting of informed consent.Patients completed the questionnaire only once, although most of them (73.4%)visited the hospital repeatedly during the study period.

Ethical Considerations
The study was approved by the Hospital's Ethical Committee and all patients were thoroughly informed about the aims of the study before completing the questionnaires.

Statistical Analysis
Data were analyzed with the use of the SPSS software (IBM SPSS Statistics for Windows, Version 23.0 Armonk, NY: IBM Corp).A normality test was performed of the total score of the needs according to the Blom's method (QQ plot), without revealing any declination.The frequencies of the descriptive characteristics of 95 cancer patients were estimated, while the mean levels of needs score were compared in these categories throughout the methods of Student t, Kruskal Wallis or Mann Whitney and Chi-square tests.
Each need (item or question) was separately coded into a binary variable (0=absence, 1=presence).A composite score was then calculated by adding up the number of individual needs (min 0 and max 23 needs), whereas the accumulation of greater than the median value (≥8 needs) was considered to be as a high frequency of needs.Following recent research, this score was subsequently rescaled to 0-100 to be more easily interpretable and understood, with higher values indicating higher level of needs (Linardakis et al., 2015).
The overall mean score of patients' needs was 36.1(±17.9)and their mean number was equal to 8.4(±4.1)Females reported a higher score of unmet needs than males (40.3 versus 30.0, p=0.005) and more than 8 needs (high score) (60.7% versus 35.9%, p=0.022).In a similar way, younger ACPs (<66 years) reported a higher score of unmet needs than older patients (40.1 versus 29.2, p=0.001) and more than 8 needs (60.0%versus 34.3%, p=0.02).Furthermore, the group of "satisfied" (≥8/10) patients from the care provided (74.2%) had a lower mean score of needs than those "not satisfied" (30.6 versus 40.3, p=0.009) and less high number of needs (40.6% versus 75.0%, p=0.005).Concerning other variables, such as marital status, education, place of residence, cancer type and department of care, statistical significant differences between groups were not detected.
Patients reported higher levels of needs with regard to receiving information about their future condition (73.7%), treatments (56.8%), examinations (51.6%) and diagnosis (50.5%) (Table 2).In addition, their need "to speak with people who had the same experience" was also reported as high too (53.7%).In contrast, lower reported needs were observed in group assistance and treatments needs regarding "better respect for my intimacy" (11.6%), "better attention from nurses" (15.8%), "more help with eating, dressing, and going to the bathroom" (18.9%).By dividing patients into 2 subgroups (up to 65 and ≥66 years old) it was found that in both groups the most prevalent needs were informational (items 2 & 4 for younger patients and 2 & 3 for older), while the less frequently expressed need was for both groups the one for intimacy (item11).Significant differences were observed among younger patients in the 8 items of their needs (items 11-16, 22, 23).
Associations between the socio-demographic variables, the overall satisfaction with care and the type of cancer with the percentage of positive answers in at least one need for each group of ACPs are described in Table 3. Females reported more needs in assistance and treatment, support and relational domain of needs than males.Younger patients reported more needs for assistance and treatment, structure, financial and support domain of needs and less communicative need than the older ones.Patients living near the hospital revealed less examinations and treatment informative needs than those living in other cities/areas.Patients cared for in a oncology ward had more examination/treatment informative needs and more structure needs than those cared for in the day clinic.Patients who had five or more visits to hospital over the last two months had higher relational needs than the other two subgroups.Patients less satisfied with the care provided, declared more communicative, assistance and treatment and structure needs than patients who declared satisfied.Finally, lung ACPs revealed more assistance and treatment needs than colon and breast ACPs (p<0.05).a With arrow are presented the five higher (↑) or lower (↓) needs according to percentages.
b Significant differences between age groups (p<0.05,Chi-square tests).

Discussion
The present study has translated and validated the Greek version of the NEQ and subsequently used it to assess the needs of hospitalized ACPs in a large tertiary regional Hospital.The NEQ displayed adequate psychometric properties and was easily handled both by the health professionals and by the patients.ACPs have considerable level of supportive care needs, arising from their disease and the therapeutical interventions provided (Rainbird et al., 2009;Bahti, 2010).The present findings have shown a mean number of needs equal to 8.4.Other authors have reported higher needs in a group of breast ACPs (Uchida et al., 2011), while in another group with mixed cancer population, lower number of needs was reported (Annunziata, Muzzatti, & Bidoli, 2010).Systematic reviews regarding unmet care needs have identified that ACPs have needs across multiple domains, while most prevalent have been those of physical, economic, informational, communicational, spiritual and psychological spectrum (Moghaddam et al., 2016;Harrison et al., 2009).High levels of unmet needs suggest that health professionals don't adequately recognize issues relevant to ACPs care (Elkin et al., 2007), or that patients may not request palliative care services (Schenker et al., 2014) or that the interventional procedures have limited effectiveness (Carey et al., 2012).
Over the last years there is growing interest to systematically explore patients' prevalent needs in routine clinical practice with the use of specific needs assessment tools (Carlson et al., 2012;Richardson et al., 2007).With these procedures, early implementation of palliative care, appropriate use of supportive care services and improvement of the patients' well-being can occur (Schenker et al., 2014;Glare, 2013;Peppercorn et al., 2011).
In a holistic patients-centered approach, demographic and clinical variables such as gender, age, as well as type of cancer should be taken into account (Kirkova, Rybicki, Walsh, & Aktas, 2012;Li & Girgis, 2006;Walsh et al., 2000).In the present study gender related differences of needs have been observed, with females expressing more needs than males.Females have required more effective care and reported more prevalent symptoms than males (Kirkova et al., 2012;Wessels et al., 2010).Younger ACPs reported higher number of unmet needs in many domains than the older ones.Previous studies have come to different conclusions.In particular, some studies have also shown that younger ACPs were more likely to have more symptoms, greater suffering and distress and probably more unmet needs (Walsh et al., 2000;Kirkova et al., 2012;Kirkova et al., 2010).However, another study, using the NEQ in a mixed cancer population, did not find significant differences between younger and older patients, except for needs regarding patients' intimacy and support (Romito et al., 2011).
Additionally, in the present population, more than half of the breast and lung ACPs reported more than eight needs, while lung ACPs reported significantly higher unmet supportive care needs in assistance and treatment domain.Previous studies have shown a large number of unmet needs in lung ACPs with more prevalent those of the psychological/emotional, physical and communicational domains (Ugalde et al., 2012;Li & Girgis, 2006).
Our findings for high assistance and treatment needs including better control of symptoms, more help for eating, dressing, better attention from nurses and others may be due to poor life expectancy and many serious problems related to lung cancer as compared to other neoplastic diseases (Li & Girgis, 2006).
Four of the five most prevalent unmet needs reported in this study were informational, related to diagnosis, prognosis, results of examinations and type of treatment.Similar high informational needs were found in patients with different types of cancer and disease stages using the same questionnaire (Romito et al., 2011;Tamburini et al., 2003;Tamburini et al., 2000).Furthermore, high unmet needs in the communicational/informational, physical and psychological domains have been reported by other authors using different questionnaires (Edib et al., 2016;Kritsotakis, Koutis, Kotsori, Alexopoulos, & Philalithis, 2010;Rainbird et al., 2009).Several studies have been carried out investigating the type, sources, appropriate time and level of information provided and its influence on the patients' quality of life (Marcusen, 2010;Wong et al., 2002).Patients were found to seek more information about symptom management, their prognosis, quality of life, clinical trials and home palliative care resources (Danesh, Belkora, Volz, & Rugo, 2014;Wong et al., 2002).Unfortunately, in everyday practice there are many factors obstructing satisfactory information (Marcusen, 2010;Bahti, 2010).For example, many patients believe that some problems are inevitable with cancer and don't ask, while others still feel that the information is unclear by its nature (Bahti, 2010;Tamburini et al., 2003).Additionally, many times there are requests from family members to conceal information about prognosis (Hancock et al., 2007).On the other hand some health professionals often do not have enough communication skills or they don't understand the patient's unmet needs.Hence, they focus only on the management of symptoms expressed by the patients, while they work under pressure due to time restrictions that exist in demanding clinical settings (Marcusen, 2010;Rainbird et al., 2009;Hancock et al., 2007;Tamburini et al., 2000).Filling the gaps of informational needs helps the patient to understand the disease, diminish uncertainty, have realistic expectations, decrease anxiety and better participate in the disease's management (Husson, Mols, & van de Poll-Franse, 2011;Hurny, 2002).
In the present study ACPs had expressed the desire to speak with people who have had the same experience.This may be an alternative way to find sources of information and to satisfy their unmet informational needs.Additionally, interaction with patients having the same experience helps them to acquire knowledge about successful management of the difficulties and the fears they may be experiencing, receive support, encouragement and guidance, and therefore, to better cope with their disease (Rini et al., 2007).
The present findings have shown an overall low need for attention from nurses, respect of intimacy and for need for help with every day activities.This is probably due to two reasons.The study was conducted in a well-established oncology department, where, although nurses work excessively under intense conditions, they are well trained and experienced to manage situations patients consider as nursing tasks.Additionally, family members are always present, supporting the health care professionals, as part of the caring team.
It is of note that most of present patients have expressed an overall satisfaction despite a number of unmet needs.
The less satisfied patients expressed communicative, assistance and treatment and structure needs.Satisfaction is of utmost importance, since it can be associated with better outcome (Bredart et al., 2013).Hence, it is important to identify specific unmet needs among unsatisfied ACPs and examine how they influence satisfaction.
The present study has some limitations.Firstly, the data were collected only from the island of Crete and, therefore, generalization should be made with caution.Nevertheless, the study had been conducted in the only major oncology center, serving patients from different areas of the island with diverse socioeconomic backgrounds.Secondly, the Greek version of NEQ was used for first time in hospitalized ACPs with solid tumors.Thirdly, due to its design, the NEQ doesn't cover the full spectrum of needs and the burden that may have on ACPs.However, such questionnaires are used more as screening tools to reveal needs and find subgroups of patients for more in detail interventions.Finally, a relative limitation is that the study was carried out a few years ago.However, improvements in treatment and increased survival over the last years, did not significantly change nursing care and the appearance and frequency of symptoms in patients with advanced cancer.

Conclusions
The Greek version of NEQ displayed adequate psychometric properties and was easily handled by both the health professionals and patients.This study reported a high prevalence of informational unmet needs with younger and female patients reporting more needs than their older and male counterparts.Most patients expressed overall satisfaction despite a number of unmet needs.Thus, the NEQ is a promising, useful tool in everyday clinical practice for obtaining information for supportive care needs.Integrated and effective patient-centered care should recognize and address each and every patient's needs; this will help to reduce the burden of symptoms, improve supportive interventions and maintain high quality of care.

a
Score from 0 to 100; higher score means more needs.The overall mean number of needs was 8.4 (±4.1).The cut-off value of low or high frequency was the median value of needs (≥8 needs).b Student t and Kruskall Wallis tests were used.Post hoc comparisons were tested by Mann Whitney test (no significant differences were found).c Chi-square tests.d Satisfaction from care was recorded in a scale from 0 (not satisfied) to 10 (very satisfied).Eight corresponds to median value.

Table 1 .
Descriptive characteristics, score and frequency of needs of the 95 cancer patients.

Table 2 .
Answers distribution of Needs Evaluation Questionnaire (NEQ) in 95 cancer patients