Supportive Care in Cancer

, Volume 15, Issue 10, pp 1125–1144

Patients’ needs assessment in cancer care: a review of assessment tools

Authors

    • Florence Nightingale School of Nursing and MidwiferyKing’s College London
  • Jibby Medina
    • Florence Nightingale School of Nursing and MidwiferyKing’s College London
  • Vivienne Brown
    • Worthing and Southlands Hospitals NHS Trust
  • John Sitzia
    • Worthing and Southlands Hospitals NHS Trust
Review Article

DOI: 10.1007/s00520-006-0205-8

Cite this article as:
Richardson, A., Medina, J., Brown, V. et al. Support Care Cancer (2007) 15: 1125. doi:10.1007/s00520-006-0205-8

Abstract

Background

The assessment of patients’ needs for care is a critical step in achieving patient-centred cancer care. Tools can be used to assess needs and inform care planning. This review discusses the importance of systematic assessment of needs in routine care and the contribution tools can make to this process.

Method

A rapid appraisal was undertaken to identify currently available tools for patient assessment in cancer care through searches conducted with Medline and CINHAL databases. It focused on tools for the systematic assessment of individual patients’ needs for help, care or support, to be used for clinical purposes—not for research or other purposes. Tools that focused on a single domain of care such as psychosocial needs were excluded, as were studies of patient satisfaction. A wide list of search terms was used, with references stored and managed using bibliographic software.

Results

In all, 1,803 papers were identified from the initial search, with 91 papers found to be relevant; although 36 tools were identified, only 15 tools were found to fit our criteria. These were appraised for their validity, reliability, responsiveness to change and feasibility, including acceptability to patients. The process of their development and psychometric properties were reasonably well documented, but data on how feasible they were to use in practice was scarce. Each tool met some but not all the widely accepted criteria for validity, reliability, responsiveness and burden. None were found to be complete for all dimensions of needs assessment. Most have not been sufficiently well tested for use in routine care.

Conclusion

There is a need to continue to develop and test tools that have the attributes necessary for effective practice and to research their effects on the quality of supportive cancer care.

Keywords

Systematic reviewCare needs assessmentCancer carePatient assessment

Introduction

Patient-centred care is unlikely to occur without a good understanding of patient needs and factors that influence them. There is little disagreement that having a system for routine assessment offers a rich opportunity for health professionals to understand and respond to patients’ needs, and there is growing interest in improving needs assessment [1]. There is some enthusiasm for using tools during the process of assessment to systematize the assessment of needs experienced by patients for help, care and support.

Supportive care can be defined as care ‘that helps the patient and their family to cope with cancer and treatment of it—from pre-diagnosis, through the process of diagnosis and treatment, to cure, continuing illness or death into bereavement. It helps the patient to maximize the benefits of treatment and to live as well as possible with the effects of the disease’ (p. 18) [2]. The broad aspects of need addressed by supportive care include physical, emotional, spiritual, environmental, social, sexual, financial and cultural. Seen from a patient’s perspective, the quality of supportive care can be considered to be the extent to which needs are addressed and met, and if they are to be addressed adequately, must first be identified.

Although evidence suggests that patients want health care professionals to ask about their physical and emotional needs, current assessment is not systematic and professionals frequently do not capture accurately what patients are trying to tell them [3]. The fact that some patients and families are dissatisfied with the care received and feel that they are not getting what they need [46] may, in part, be related to inadequate assessment. Healthcare professionals vary widely in their ability to elicit relevant information [7, 8], and equally, patients vary in their ability to voice their concerns and anxieties. This can be due to beliefs that certain problems are inevitable with cancer or that clinicians do not wish to address them.

A careful assessment of patients’ needs is central to the process of providing care. Recent guidance on supportive and palliative care in the UK [2] argued that rigorous and systematic assessment should be the critical first step in supportive care, leading to the delivery—and, in some cases, development—of appropriate services. Such assessment should be an ongoing process with particular attention at key points such as at the time of diagnosis, start of treatment and completion of the primary treatment plan, when incurability is recognized and when dying is diagnosed. It was suggested in the guidance that care teams should develop a common approach to assessment, examining the potential of common tools and systems to ensure that results are shared across and between teams. The review presented in this paper was part of a project conducted to inform the development of a common approach to assessment in England, UK. Its aim was to identify and appraise tools currently available to support patient assessment in routine cancer care. It focused on tools for the systematic assessment of individual patients’ needs for help, care or support, to be used for clinical purposes—not for research or other purposes. It sought to undertake a rapid appraisal sufficiently comprehensive to identify existing research, as well as gaps in the field; we also sought to identify methodological and clinical issues requiring consideration when developing tools in future.

The working definition of assessment that informed the review was that assessment is about collecting information on a person’s needs and circumstances and making sense of that information to identify needs and decide on what support or treatment to offer. It was conceived as (1) a clearly defined process done with or by the person with cancer (2) involving some form of consistent framework and (3) involving regular comprehensive assessment at clearly defined intervals (4) based on patients’ accounts of their needs and wishes, which they expect professional care to meet and finally (5) informing the decisions of a range of healthcare professionals involved in care. A patient needs assessment tool is defined in this paper as a collection of questions, scales and other means of obtaining information that, together, provide a consistent and comprehensive system through which patients’ range of needs for support and care can be explored [9]. A tool should enable health care professionals to understand the specific needs patients would like to be met through professional care after use at clearly defined points in the patient pathway.

Materials and methods

Searching the literature

A rapid literature search was undertaken following the general principles outlined in Glasziou’s Systematic Review Guide [10]. It was not intended to be an exhaustive review and was conducted in a limited timeframe, so it was important to quickly identify tools most pertinent to the subject of interest. A systematic search was performed using Medline, Embase, British Nursing Index, ERIC and CINHAL databases across the period 1984–2004 to identify patient assessment tools developed for use in routine cancer care. Search terms used either singularly or in combination were ‘cancer’, ‘patient participation’, patient centred’ and ‘assessment’ in the thesaurus and index lists of the relevant databases. Also ‘free text’ words were used to supplement the search terms [medical subject heading (MeSH) search terms in the case of Medline]. The review was designed to focus on tools for the systematic assessment of needs of individual patients with cancer. Our focus was solely on tools with an application in clinical care, in contrast to other reviews and texts that include tools for research on quality of life assessment and outcome measurement [1115]. To arrive at a working definition of tools to be included in the review inclusion and exclusion, criteria were developed (see Table 1). Manual searches of the bibliographies of searched articles and reviews in the field were also conducted, and contacts were made with experts in the field.
Table 1

Needs assessment tool inclusion and exclusion criteria

Included

Excluded

Tools developed with purpose of assessment of patients’ needs for clinical—not for research or other–purposes

Tools only about one type of need only, e.g. in relation to symptoms, emotions

Tools focused primarily on needs related to health status

Tools concerned with assessing health-related quality of life in the context of a group level application

Tools focused on the needs of adult patients with cancer

Tools solely assessing needs for, and satisfaction with, aspects of health care

Tools designed for the purpose of assessing experienced needs for care in the context of individual patient application

Tools solely assessing patients’ opinions on the quality of care received from professionals

 

Tools designed originally primarily for purposes other than individual needs assessment, e.g. audit

 

Tools for which there was insufficient information published to conduct an appraisal

Figure 1 illustrates our search and selection process. Titles and abstracts were screened for inclusion by one reviewer (JM), with decisions checked by a second (AR). Where the relevance of a paper was not clear, the full text was retrieved and the content examined by both reviewers to determine if it met the inclusion criteria. Papers that met the inclusion criteria were read and summarized by the research team using a structured data extraction form.
https://static-content.springer.com/image/art%3A10.1007%2Fs00520-006-0205-8/MediaObjects/520_2006_205_Fig1_HTML.gif
Fig. 1

Flow chart summarizing number of studies identified and ultimately included

Evaluation criteria

We appraised each tool against a range of properties [1, 1618] suggested in the literature as pertinent to whether a tool might be suitable for routine use (see Table 2). In addition to validity and reliability, these include responsiveness to change, feasibility (including evidence on acceptability as well as burden on both patients and professionals), intended purpose and administration method.
Table 2

Properties required of tools proposed for use in routine clinical practice

Properties

Definition

Validity

Involves judging whether a tool measures what it intends to measure

Reliability

Requires that a tool is reproducible and internally consistent

Appropriateness

Considers the match between specific purpose, population and setting for which a tool was developed and intended use

Responsiveness to change

Addresses whether a tool is sensitive to changes of importance to patients and detects clinically meaningful change

Feasibility

Addresses the extent of effort, burden and disruption to staff and clinical care arising from use of a tool

Acceptability

Considers how acceptable a tool is for respondents to complete

Each tool was assessed in terms of the following aspects.

Purpose

This referred to the tool’s original intended use(s) and application.

Question format

This referred to the tool’s response format.

Validity

The number of items and domain structure for each tool were summarized. Precise content was recorded by classifying items of a questionnaire along dimensions of need for care. This followed a similar approach to Osse et al. [19] who split these between needs related to health-related quality of life and problems in the quality of care. The current review mapped content on needs related to (a) health status (see Table 3) and (b) needs for and satisfaction with health care (see Table 4). Although needs for and satisfaction with health care was not the main focus of the review, it became obvious early on that some of the tools contained items related to this aspect, and so we chose to map this aspect. Needs related to health-related quality of life were adapted from Fitzsimmons and Ahmedzai [20] and Ahmedzai et al. [21], and problems in relation to quality of care were adapted from McCusker [22] and Wiggers et al. [23]. Also noted were the methods used to identify the problems and needs included in the tools (e.g. interviews, expert panels and literature reviews). The results of any empirical validation of the tools such as attempts to establish convergent or discriminant validity were noted.
Table 3

Domains used to compare tool content and number of items: needs related to health status

Domain

Examples

Disease symptoms and treatment related side effects

Pain

Fatigue

Dyspnoea

Appetite

Nausea and vomiting

Constipation

Physical functioning

Mobility

Self-care activities such as washing and dressing

Activities of daily living such as household chores, meal preparation

Physical activity

Disability

Psychological well being

Depression

Anxiety

Adjustment to illness

Coping

Fear

Self esteem

Body image and sexuality

Life satisfaction

Spiritual well being

Concerns with meaning/purpose of life

Concerns about death and afterlife

Conflicted or challenged belief system

Cultural concerns

Cognitive

Confusion

Memory

Concentration

Social

Partner relationships

Family relationships

Coping with children

Ability to carry out hobbies, interests, leisure activities

Occupational

Work

Financial status and sources of financial support and assistance

Global assessment

Global health

Global quality of life

Table 4

Domains used to compare tool content and number of items: needs for and satisfaction with health care

Domain

Examples

Participation in care, involvement in decision making and information provision

Preferred degree of involvement

Adequacy of information provision re: diagnosis, treatment and investigations

Opportunity to ask questions

Accessibility of care, help and support

Proximity of service to home

Transport to and from clinic

Home care provision

Provision of equipment, aids and appliances in the home

Flexibility of care delivery

Choice over clinic appointment time

Choice over location of treatment

Continuity of care

Healthcare professional aware of previous history and experience

Seeing same doctor at consultation

Co-ordination

Health care professionals technical competence, interpersonal skills and availability

Technical competence

Adequacy of explanation

Available at time of need

Waiting times

At clinic

For an appointment

Environment of care

Cleanliness

Space

Privacy/dignity

Diagnosis disclosed in private

Support and help for family members/carers

Information for family members

Emotional support

Support for caregivers at home

General satisfaction with care

 

Reliability

This was examined in terms of internal consistency and test–retest reproducibility.

Responsiveness

This was assessed through examining a tool’s ability to detect change over time.

Feasibility

This was gauged through examining a range of features, including number of items, time and ease of administration, reading level, whether the tool had been primarily designed for research purposes or for integration in clinical care, data on acceptability, availability in different formats and administration method.

Findings

Our search strategy of the electronic databases identified 1,803 possible relevant papers. Applying the inclusion criteria left only 91 papers directly related to the assessment of needs for care in cancer patients. There was disagreement in around one of ten cases between the two reviewers at the initial searching and inspection stage. Papers were excluded if the tools were designed for purposes other than assessment of individual patient needs or were concerned with one type of need or aspect of care only (for example, in relation to symptoms). A number of tools fell at the margins, but after careful examination, they were excluded as they were developed principally as outcome measures to assess the effectiveness of palliative care intervention [24, 25], specifically for audit purposes [26], or developed as a survey of patient need rather than a clinical tool [27, 28].

In this manner, the selection was reduced to 36 papers in which 15 tools had been reported. The tools for cancer patients that fulfilled our inclusion criteria were as follows:
  • Cancer rehabilitation evaluation system (CARES) [29, 30], previously Care Inventory of Problem Situations (CIPS) [31]

  • Cancer care monitor (CCM) [32]

  • Creating better health outcomes by improving communication about patients’ experiences assessment (CHOICEs) [3335]

  • Concerns checklist [3639] (originally developed by Devlen [40])

  • Distress management tool [41]

  • Initial health assessment (IHA) form [42]

  • Needs evaluation questionnaire (NEQ)[4345]

  • Needs at the end-of-life screening tool (NEST) [4648]

  • Oncology clinic patient checklist (OCPC) [49]

  • Patient needs assessment tool (PNAT) [50]

  • Problems and needs in palliative care instrument (PNPC) [51, 52]

  • Problems checklist [53, 54]

  • Supportive care needs survey (SCNS) [55, 56], previously cancer patients needs questionnaire (CPNQ) [57] or CNQ

  • Sheffield profile for assessment and referral to care (SPARC) [21, 58, 59]

  • Symptoms and concerns checklist [6062]

Table 5 summarizes the evaluation of the selected tools.
Table 5

Summary of tool properties

Instrument and country of origin

Purpose and population

Items and domains

Question format

Validity

Reliability

Responsiveness

Feasibility

Content validity

Construct validity

Internal consistency

Reproducibility

CARES [2931] (previously CIPS); USA

To measure rehabilitation needs and quality of life, and find how cancer affects psychosocial, physical and behavioural states—in patients with cancer

139 items; 6 domains:

•Physical (26)

•Psychosocial (44)

•Medical interaction (11)

•Sexual (8)

•Marital (18)

•Miscellaneous (32)

5-point scales (0– 4); “Does not apply” to “Applies very much”. Plus “do you want help?” (Yes/No) option

•Literature

•Interviews with patients and family members

•Expert review

High correlation with interview (trained interviewer). Construct validity with:

•The symptom checklist-90 (SCL-90)

•Dyadic adjustment scale (DAS)

•Karnofsky performance status (KPS)

•Visual analogue scales for QoL before (QoL-B) and after (QoL-A) cancer [tested when CARES was CIPS]

Domains: ranged from α = 0.87–0.94

Time: soon-after first test; r = 091 (domains: r = 0.84– 0.95, 87% agreement) Time: 1 week (results “excellent”)

Demonstrated ability to capture significant change over time in women recovering from surgery (In SF) at 1, 7 and 13 months post-surgery.

Time: 20 min

Reading level: unknown

Acceptability: translation available in Spanish (for Hispanic patients), short version available (CARES-SF), and version available for AIDS patients (HOPES)

Made for: care and research. Short form available suited to repeated application in clinical trials

CCM [32]; USA

To screen high frequency cancer- related symptoms and assess overall symptom severity and QoL—in patients with cancer

38 items; 6 domains:

•General physical symptoms (11)

•Treatment side effects (8)

•Acute distress (4)

•Despair (7)

•Impaired ambulation (4)

•Impaired performance (4) [plus one global QoL index]

11-point Likert scales (0–10); “Not a problem” to “As bad as possible”

•Literature

•Physician judgements

•Review by professionals and patients

Demonstrated convergent and divergent validity through comparison with:

•The brief symptom inventory (BSI)

•The medical outcomes study SF-36 Health survey (SF-36)

•The memorial symptom assessment scale (MSAS)

•The life satisfaction index–short form (LSI)

•The satisfaction with life scale (SWLS)

•Factor analysis

Domains: ranged from 0.80–0.89

Time: between 1 and 7 days apart (correlations ranged from r = 0.90 to 0.74) Time: between 8 and 14 days apart (correlations ranged from r = 0.87 to 0.74)

To be addressed in forthcoming studies

Time: 20 min to complete paper version, 12 min to complete electronic version

Reading level: 85% completed high school or greater education

Acceptability: patients expressed a strong preference for the electronic form (versus the paper form) Made for: screening in a clinical setting

CHOICEs assessment [3335]; Norway

To assess patients’ symptoms, functional problems and preferences— completed by outpatient cancer patients before to consultation

112 items (symptoms and problems); 6 domains:

•Cancer specific symptoms

•Functional problems

•Physical

•Psychosocial

•Emotional

•Spiritual

Plus 2 global ratings: health and QoL [instrument not made available by developer]

(Yes/No) option. Degree of severity and bother of symptoms. Analogue scales (0–10) rating “Importance” of problems as priorities for treatment/care

•Literature

•Review by an expert focus group specialists in cancer care

•Review by patients

To be addressed in forthcoming studies—ongoing study on convergent and discriminant validity and plans to carry out further work

“Ease of Use”: α = 0.98 “Satisfaction”: α = 0.86

Tool currently being evaluated

Time: 9 min (median; SD = 10.35; range = 0.5 to 49 min); 25% percent of the sample used ≤5 min

Reading level: unknown

Acceptability: (a) ease of use: 80% without assistance, 20% some assistance (weak, disability, convenience).

Positive overall ‘Ease of Use’ score = 5.06 (range −16 to +16) (b)

Satisfaction: scores positively skewed in both groups

Made for: care. To support clinicians during consultations

Concerns checklist [3640];UK

To elicit and register main concerns of patient—various groups and hospice patients

Refined version  = 12 items, original source  = 53 items; 3 domains

•Illness (7)

•Practical (2)

•Psychological (3)

Two forms (of refined version):

•Self completion: 5-point scale (Not a worry– Extremely worried)

•Interview schedule (open questions)

•Literature

• Retrospective study data

• Pilot work

• Review by patients

Factor analysis-3 factor solution

Time: unknown

Reading level: unknown

Acceptability: unknown

Made for: to alert clinicians to nature and number of concerns. Has been used to assess the effect of communication skills training and nurses’ skills in eliciting concerns

Distress management tool [41]; USA

Screening tool for rapid assessment—in cancer patients

36 items; 5 domains:

• Practical problems (5)

• Family problems (2)

• Emotional problems (6)

• Spiritual/religious concerns (1)

• Physical problems (21)

• Plus 1 general distress item

One rating scale (distress thermometre; 0–10) “Extreme distress” to “No distress”. Plus 33 statements (Yes/No)

• Literature

• Expert review—NCCN panel

Time: Unknown

Reading level: unknown

Acceptability: unknown

Made for: the clinical setting

IHA form [42]; Canada

To aid clinicians in recognition and documentation of supportive care needs—of patients with newly diagnosed cancer during their first visit to a comprehensive cancer centre

Unclear. Common supportive care needs in 7 domains:

• Physical

• Psychological

• Practical

• Financial concerns

• Informational needs

• Other special needs

• Important personal resources

Checklist, prompts (facilitating discussion), plus space for management plan

• Literature

• Patient survey of supportive care needs

• Pilot test

Time: unknown

Reading level: Unknown

Acceptability: focus groups with professionals determined effective form of instrument. Introduced in practice and evaluated

Made for: use as a multidisciplinary tool in the clinical setting. Used mostly by nurses

NEQ [4345]; Italy

Assessment of informative, psychological and social needs—of hospitalized cancer patients

23 items; 4 domains:

• Information regarding diagnosis/ prognosis (2)

• Information regarding examination/treatment (2)

• Communication (4)

• Relational (3)

• Plus 12 additional items

Statements (Yes/No)

• Interviews with patients

Factor analysis on the scale only partially confirms the hypothesized structure. Later study demonstrated good fit

Domains: ranged from α = 0.63–0.76

Time: 1 week (Cohen’s kappa ranged from 0.54–0.94). Being further developed and tested 2 items removed; originally 25—to improve reliability

Time: 5 min

Reading level: unknown

Acceptability: 63% of patients OK; 24% incomplete; 3% missing values. Only available in Italian

Made for: use as an aid in clinical encounters

NEST [4648]; USA

Developed as a framework that is based on patients’ experiences and perspectives regarding their care—to measure subjective experiences of patients at end-of-life

13 items; 10 dimensions:

• Financial burden (1)

• Access to care (1)

• Social connectedness (1)

• Caregiving needs (1)

• Psychological distress (2)

• Spirituality/religiousness (1)

• Personal acceptance (1)

• Sense of purpose (1)

• Patient–clinician relationship (1)

• Clinician communication (1)

Plus 2 additional items

5-point Likert scale (strong agreement–strong disagreement) or discrete responses

• Literature

• Interviews

• Symptom items from other scales

• Pilot work

• Expert review

Domains: baseline-ranged from α = 0.63–0.85; follow up-ranged from α = 0.64–0.89

Time: unknown

Reading level: unknown

Acceptability: 69.2% patients found the interview was helpful. Equally reliable short form has 13 items

Made for: use as screening tool for the clinical setting and possibly assess impact of interventions

OCPC [49]; USA

To systematically assess problems related to cancer and its treatment—in adult patients in outpatients clinics

86 Items; 15 domains:

• Information (12)

• Fatigue (3)

• Pain (3)

• Nutrition (7)

• Speech and language (4)

• Respiration (3)

• Bowel and bladder (9)

• Transportation (2)

• Mobility (5)

• Self and home care (8)

• Vocational and educational (5)

• Interests and activities (6)

• Family (5)

• Interpersonal relationships (4)

• Emotional (7)

• Plus 3 open-ended questions

Checklist for each item—to indicate prevalence of problem. Plus 3 open-ended questions

• Data from previous research

• Based on items from other tool

• Based on items from other tool

Time: unknown

Reading level: unknown

Acceptability: Checklist was accepted for practical use—process evaluation by staff (100% response rate) and patients (78%) (after 4 months); positive response from nursing staff. Usefulness, 82% (pilot work, n = 11 patients)

Made for: use in care

PNAT [50]; USA

A tool to screen for potential problems in physical and psychological functioning—in cancer patients

16 items; 3 domains:

• Physical (6)

• Psychological (5)

• Social (5)

Overall discomfort (symptom distress) also measured

Five-item scale (no impairment–severe impairment) for each item, within context of structured interview

• Literature

• Clinical experience

• Review by multi-professional group

Physical domain correlates with:

• KPS Psychological domain correlates with:

• Global Adjustment to Illness Scale (GAIS)

• Memorial Pain Assessment Scale (MPAC)

• Beck Depression Inventory (BDI)

• BSI Social domain correlates with:

• Interpersonal Support Evaluation List (ISEL)

Domains: Ranged from 0.85–0.94

Interrater reliability: concordance co-efficient: 0.87 for physical, 0.76 for psychological, and 0.73 for social; Spearman rank order correlation coefficient: 0.59–0.98

Time: 20–30 min

Reading level: unknown

Acceptability: item selection reviewed and rated by nurses, physicians, psychologists and social work members of service

Made for: clinical practice; by oncology nurses and physicians (because it is the clinician who initially triages a patient’s care) and social workers

PNPC [51, 52]; The Netherlands

A checklist of problems patients experience in palliative care and their needs for care. To support provision of care tailored to specific demands of individual patients with advanced cancer

138 items; 13 domains:

• Activities of daily living (7)

• Physical symptoms (18)

• Role activities (4)

• Financial and administrative issues (5)

• Social issues (15)

• Psychological issues (15)

• Spiritual issues (5)

• Autonomy (9)

• Informational needs (9)

• Problems in consultations (3)

• Overriding problems in quality of care (9)

• Concerning the GP (20)

• Concerning the specialist (19)

• Experienced problems: 3 options (Yes–Somewhat–No)

• Needs for care: 3 options(“Yes, more”-“As much as now”-“No”)

• Literature

• Interviews with patients and their partners

• Interviews with professionals

• Pilot work

• Item content analysis of existing tools

Convergent validity demonstrated by comparison with two HRQL measures:

• The European Organization for Research and Treatment of Cancer (EORTC)

• The COOP-WONCA quality-of-life measure

Domains: Problem aspect – ranged from α = 0.67–0.89 Need for care aspect ranged from α = 0.73–0.92

Acknowledged that further studies needed to develop and test

Time: unknown

Reading level: unknown

Acceptability: original tool in Dutch. Has been translated to English (further validation required). Appropriateness and comprehensiveness also assessed

Made for: use as a patient-centred tool in clinical practice. Can be used for research

Problems checklist [53, 54]; UK

Developed to assess the prevalence and severity of psychosocial problems experienced by the cancer patients

16 items; 4 domains:

• Daily living (4)

• Relationships (5)

• Economics (2)

• Emotions (3)

Plus 2 other

4-point scale (0 = no difficulty, 3 = severe difficulty). An additional category of ‘5 = does not apply to me’ was included in the rating

• Literature

• Audit data

• Research study (n = 505)

Factor analysis endorsed the 4 factor structure (accounting for 64% of variance) with the components on Economics and Emotions being particularly credible

Domains: ranged from α = 0.70–0.82

Time: unknown

Reading level: unknown

Acceptability: patients found it quick and easy to complete

Made for: use to identify/highlight problems not usually picked up in routine practice

SCNS—Long form (LF59) and short form (SF-34) [57] (previously CPNQ or CNQ) [5557]; Australia

Developed to assess perceived needs—of patients with cancer. Plus a supplementary modules

59 items; 6 domains:

• Psychological (22)

• Health system and information (15)

• Physical and daily living (7)

• Patient care and support (8)

• Sexuality (3)

• Additional (4)

Selection; “Applicable” or “Satisfied”. Then 3-point “Need” scale (3–5); “Low need” to “High need”

• Review of CNQ tool by oncology specialists and patients

• Pilot work

•Factor analysis revealed 5 factors were identified which together accounted for 64% of the total variance

Domains: Ranged from: α = 0.87–0.97

Longitudinal study underway

Time: 15–20 min

Reading level: 11–13 years

Acceptability: rated by patients; was easy to understand and complete; 7–12% missing data for each question. Trial of use by touch screen in clinic demonstrated acceptability

Made for: quality assessment, research and intervention

SPARC [21, 58, 59]; UK

Developed to assess the distress caused by advanced illness and to screen symptoms and problems to guide referrals to specialist and palliative care—for patients with advanced illness

45 items; 7 domains:

• Communication and information (1)

• Physical symptoms (21)

• Psychological issues (9)

• Religious and spiritual issues (2)

• Independence and activity (3)

• Family and social issues (4)

• Treatment issues (5)

• Items: help/information/contact with professionals (Yes/No)

• Remaining items: 4-point rating scale (0–3); “Not at all” to “Very much”

• Literature

• Interviews with patients and professionals

• Cognitive interviewing

• Consultation with experts

• Pilot work

Inter-item correlations examined, and item-total correlation

Time: 45 min (median), range of 15 to 105 min

Reading level: unknown

Acceptability: assessed during iterative process of development

Made for: professionals as a screening measure, to aid their referral decisions. Future version planned to allow patients and carers to use self-rated measure to support need for specialist palliative care

Symptoms and concerns checklist [6062]; UK

To determine prevalence and severity of symptoms and concerns in routine practice as adjuvant to clinical assessment—of patients with advanced cancer

32 items (29–32 items); 4 domains:

• Physical symptoms (11)

• Cognitive/psychological (4)

• Other concerns (14)

• Patient defined (3)

Rating scale—‘how much of a problem’ (0–3); “not at all” to “very much”

• Literature

• Expert panel

• Pilot work

• Patient interviews

Generally demonstrated convergent validity when compared with:

• Symptom Distress Scale (SDS)

• Palliative care Outcome Scale (POS)—Patient version Able to discriminate between different groups of patients (e.g. Outpatients vs. hospital inpatients)

Overall: α = 0.85

Time: over 2 consecutive days—weighted Kappa 0.35–0.77

Time: 5 min

Reading level: unknown

Acceptability: 97% felt comprehensive, 82% felt easy to complete, 79% good idea, 98% participated, 97% completed all items

Made for: clinical use. Also for audit and research

Purpose and population

Table 5 shows the many different purposes for which the tools were developed. These include focus on the health status of patients, providing information on particular symptoms or problems, personal resources and sources of support, care preferences and satisfaction with care. Often, the experience of a problem—or dissatisfaction with an element of care—is not clearly distinguished from a need for more or better care. A small number of instruments address needs for care (help or support), such as the CARES, the SCNS, the PNPC and the distress management tool. Whereas developers imply that many tools may be used for screening (identifying those who might benefit from referral for more in-depth evaluation or who might warrant intervention), only three tools were specifically developed for this purpose. The CCM screens for high frequency cancer-related symptoms, side effects and current concerns at each clinic visit; the SPARC and concerns checklist identify palliative care needs of patients to guide referral of patients and families to palliative care. The CCM, the concerns checklist, the symptoms and concerns checklist and the IHA are put forward as aids for clinicians to enable them to recognize and document supportive care needs as an adjunct to consultation.

Most tools address the needs of a general population of cancer patients and were developed with mixed groups. A few, however, were developed to address the needs of specific groups of patients, such as those with advanced cancer (PNPC, SPARC and symptoms and concerns checklist) or, even more specifically, at the end of life (NEST). Most were developed for use in an outpatient clinic; exceptions are the NEQ and the NEST (designed for hospitalized cancer patients) and the SPARC and symptoms and concerns checklist (both of which can be used in a primary care setting).

Content

Content of tools is drawn from a number of sources, including existing literature, review of item content of existing tools, reliance on clinical experience of tool developers, and in some cases, research by the tool developers. Content is often refined (and reduced) through an iterative process involving reference to the opinions of health care professionals and patients, pilot work and, in some cases, through recourse to statistical techniques such as factor analysis (for example, for the SCNS and the NEQ).

Table 5 provides an indication of the domains covered by the tools and their relative emphasis in terms of different areas of need. The content of tools was compared using the domains of needs related to health status and needs for, and satisfaction with, health care. Tables 6 and 7 provide a more detailed examination of the content of the individual tools with respect to these domains.
Table 6

Comparison of content of assessment instruments compared by reference to needs related to health status

Instruments→

CARESa

CCM

CHOICEsa

Concerns checklist (refined version)

Distress management toola

IHAa

NEQ

NEST

OCPC

PNAT

PNPC

Problems checklist

SCNS

SPARC

Symptoms and concerns checklist

Domains ↓

Symptoms and side effects

 

19

 

16

1

1

17

 

17

 

5

21

11

Physical functioning

8

7

2

1

1

8

4

7

2

2

3

1

Psychological well being

11

 

8

4

1

5

5

24

4

22

10

9

Spiritual well being

  

 

1

 

1

3

  

5

1

2

2

1

Cognitive

    

1

   

4

2

 

1

 

1

1

Social

 

3

5

2

1

10

2

18

5

1

5

2

Occupational

   

2

2

2

1

4

1

7

2

1

1

2

Global assessment

 

1

 

1

    

1

     

aUnable to map accurately, as instrument not made available by tool developer. Information taken from publications

√ Indicates possible coverage

Table 7

Comparison of content of assessment instruments compared by reference to needs for and satisfaction with health care

Instruments→

CARESa

CCM

CHOICEsa

Concerns checklist (refined version)

Distress management toola

IHAa

NEQ

NEST

OCPC

PNAT

PNPC

Problems checklist

SCNS

SPARC

Symptoms and concerns checklist

Domains↓

Participation in care and information provision

     

5

2

12

 

24

 

9

6

1

Accessibility of care

       

2

2

1

5

 

2

  

Flexibility of care delivery

          

3

 

4

  

Continuity of care

          

5

 

1

  

Healthcare professionals technical and interpersonal skills

     

5

1

  

25

1

4

5

1

Provision of equipment

               

Waiting times

            

1

  

Environment

      

1

     

1

  

Privacy and dignity

      

1

     

2

  

Support for carers

        

4

 

2

 

2

  

Global satisfaction

               

aUnable to map accurately, as instrument not made available by tool developer. Information taken from publications

√ Indicates possible coverage

Degree of coverage varies widely. The most comprehensive instruments with respect to health status (defined as covering to some degree the full range of needs related to health status) include the PNPC, OCPC, symptoms and concerns checklist, SCNS and SPARC. In particular, the distress management tool and SPARC covered all the dimensions in our classification related to health status. Coverage did not necessarily equate with length; some of the shortest instruments proved to have items across a number of domains. As noted above, only the CARES, SCNS and PNPC addressed the dual facets of how much problem a particular issue presents, along with the degree to which patients wish help or attention to respond to this need. Broadly, symptoms and side effects, together with physical and psychological functioning, were more fully covered than were other aspects of need such as spirituality (often purely confined to issues of religion), cognition and occupation.

Both the SCNS and the PNPC possess a broader coverage of issues related to satisfaction with health care than the other tools reviewed. It is not surprising that the majority of the tools did not cover this aspect in any depth, as their purpose was directed at needs with respect to health status rather than assessing quality of care received.

Question format and scoring

Question formats reflect purpose. These include adjectival, semantic differential and Likert-type scales that address particular aspects of need such as the degree to which a problem is experienced, the degree of bother or the degree of importance. Checklists tend to utilize dichotomous items (yes or no) to indicate wants for help or the presence or absence of a need. Many of the tools adopt a combination of formats to accommodate different types of questions.

The majority of the tools adopt some kind of scoring system especially at the individual item and domain level (CCM, PNAT, SCNS, problems checklist and IHA). Most of them do not pursue an overall score; however, the CARES, CCM and concerns checklist are exceptions to this. In a few cases, tool developers adopted more sophisticated scoring algorithms that would require a computer to generate scores. There was little reference to the interpretation of scores.

Some tool developers specifically discuss an intention to move to a system whereby the score might be used to distinguish those patients who warrant referral to different disciplines or specific services (distress management tool and SPARC) or to provide a basis for targeted interventions (PNAT). Other tools, however, focus solely on revealing the presence or absence of a need that could then be examined more closely within a clinical encounter (NEQ, symptoms and concerns checklist). However, on the whole, there was an absence of discussion on how the assessment results might be used to plan care or how to facilitate clinical interpretation of individual scores for case-finding and monitoring purposes.

Validity

A wide variety of approaches were taken to ensure and examine tool validity. The extent to which validity was demonstrated also varied greatly.

For content validity, tools tend to rely heavily on existing literature; indeed, this was the most common approach (used in ten cases). In one case (PNPC), the developer also selected items from existing tools. Panels of professionals and patients were frequently used to derive, select or review items. Pre-testing of items was undertaken for some tools (SCNS, SPARC, symptoms and concerns checklist and OCPC). The generation of items through interviews with patients was a feature of a number of tools, although the scale on which this was undertaken varied. Some authors used extensive procedures to derive items based on patients’ accounts: combined with literature reviews, review by patient and professional panels and pre and/or pilot testing procedures. The CARES and SPARC, symptoms and concerns checklist and PNPC are examples of the results of such careful procedures. Beyond descriptions of how item contents were derived, there was often little description of the process of tool development.

Construct validity was generally assessed by examining the relationship between the tool under development and a measure of another concept to which it was theoretically related. On the whole, where this was examined, construct validity was satisfactorily demonstrated [63]. Further empirical validation of the tool beyond its initial construction by the original developers was undertaken only for the CARES and SCNS. These tools also feature successive verification by increasing the numbers of researchers. A number of tool developers report plans to conduct further validation studies. Exploratory and confirmatory factor analysis was used at various stages of tool development in a small number of cases (SCNS, NEQ and NEST). Validity testing is absent for the distress management tool, IHA and OCPC. Two instruments had demonstrated they had the ability to discriminate between different conditions, for example, at different phases of illness (CARES) or different groups of patients (symptoms and concerns checklist).

Reliability

Data on the reliability of tools are shown in Table 5. Test–retest reliability was assessed in the CARES, CCM, NEQ and, symptom and concerns checklist. For some other tools, only internal consistency was demonstrated, mostly through computing Cronbach’s alpha. The majority fell well short of the recommended reliability standards for individual patient application, which ranges from a low of 0.90 to a high of 0.95, the desired standard [64]. Internal consistency was also examined through Cohen’s kappa, a coefficient of agreement.

For scales designed for use by health care professionals, some investigators assessed inter-rater reliability (consistency of a measure when administered by different interviewers) through employing a coefficient of concordance, as was the case for the PNAT. The PNAT achieved moderate to strong concordance.

Responsiveness

Responsiveness (or sensitivity to change) has rarely been addressed. One exception is the cancer rehabilitation evaluation system—short form (CARES-SF); a study following up patients 1, 7 and 13 months after surgery demonstrated its responsiveness to actual changes occurring over time [65]. Improvements in scores were demonstrated as time from surgery lengthened.

Feasibility

Practical considerations are important in the clinical setting, affecting both patients and professionals involved in the assessment process, especially as time is necessarily limited. Yet, only limited data were found on feasibility. Wide ranges of completion times were recorded ranging from 5 to 45 min. Electronic versions of assessment tools appear to have shorter completion times, and there is some evidence that patients, if given the choice, prefer an electronic format (for example the CCM). Tools employ varying methods of administration, including both paper and pencil format and electronic solutions; some offer a choice on this issue. Moreover, some require self-completion by the patient, whereas others require an interview with a professional.

The level of support required for people to complete a tool could be an important factor in determining the feasibility of a tool in practice. Only a few papers referred to this dimension. In those involving self-completion by patients, some mention were made of having help available for the CHOICEs tool (computer-based) and the touch screen version of the SCNS. Of the professionally administered tools, only the developers of the PNAT and IHA had considered training issues.

Acceptability for the target group is clearly a prime consideration, and there have been some attempts, albeit involving variable degrees of rigour, to collect information on the content, process and outcome of using a tool from the viewpoint of patients or health care professionals. Appraisals have been made, for example, of ease of use (CHOICEs), degree of missing data (NEQ) and degree of patient satisfaction with the tool (CHOICEs). The suitability of tools for diverse populations has not been well explored; only a few developers have examined the reading level of their tool, and still fewer have sought to develop versions in different languages.

Discussion

The searches identified 15 tools designed for routine clinical assessment of patients’ needs. Most had been carefully constructed but lacked generalizability across the cancer trajectory and focused on one particular context for care (such as outpatients) or point in the cancer pathway. A few had been developed without recourse to patient input and thus liable to assess needs seen to be important solely from a professional perspective. All had very different organizing structures, and few covered all the dimensions of need, failing to offer a comprehensive approach to assessment. This is probably due to the different viewpoints from which the tools were developed. Some topics generally regarded as important in cancer and palliative care, such as spirituality, were often missing. Our concern in this review has been to identify tools intended for clinical use to address the needs of individual patients. The tools identified in this paper were selected as they professed to have been developed to assess patients needs but, when examined more closely, were employed for differing purposes: to identify and prioritize actual problems; to identify patient preferences regarding treatment and care; to monitor responses to treatment and changes in symptoms, functioning or well-being; or to screen for potential problems. A few discriminated between the assessment of health problems and the desire to receive care, which can result in ambiguity about whether or not a patient wants help. Most are constructed to assess experienced problems in health status rather than perceived needs for care. This leaves clinicians needing to interpret the practical consequences of information gathered as a result of the assessment and does not address ‘desire’ for care even when a need is identified.

Most tools employed a self-report format, and a number of computer-assisted systems of assessment have been developed. Future research should be undertaken to help us understand the circumstances under which a particular administration format might be used to best effect. Computerized assessment solutions are gaining in popularity and have demonstrated that computer-assisted approaches to assessment are acceptable to patients undergoing cancer treatment [3335, 6668]. As a greater proportion of health records become computerized, the desire for an electronic solution to needs assessment is only likely to grow. One of the obvious advantages of this mode of assessment is that it can result in the immediate availability of information for all members of the multidisciplinary team to be used during the process of treatment and care decision-making and also that assessment data and clinical data are co-located, if not integrated.

Tools to be used in routine clinical care should depict a robust image of reality and possess reasonable psychometric properties. Validity and reliability have been addressed to varying degrees: in some tools thoroughly, in others not at all. The merit of qualitative methods to ensure that tools are aligned closely to cancer patients’ needs should not be overlooked during the development process.

The dynamic nature of need and the desirability of monitoring patients’ responses to care interventions over the course of their illness and treatment experience mean that tools should have the ability to capture change. In terms of timing of assessment, ideally, patients should be initially assessed on entry to the health care system and periodically throughout treatment. Other key points of assessment might be on commencing each new treatment, when finishing treatment and entering routine follow up or on referral to palliative care. It is disappointing that very few tools had been tested over time for their responsiveness to change. Without this information, it is not possible to understand how tools might perform over time and whether they are able to capture changing needs in different groups of patients. Cancer is not one disease but a general diagnosis for over 100 different diseases with different natural histories and treatments. It is questionable whether one tool might reasonably cover all stages of illness and environments for care, and there might be an argument for supplementing ‘core’ content of a generic tool with items specific to particular populations to achieve comprehensive coverage.

Feasibility and acceptability to both patients and professionals may be a critical factor in determining the clinical possibilities of these tools and the degree to which they might be integrated in daily care; yet, information on their feasibility for use in routine care was very scarce. Future evaluation of current or new tools should address issues such as patients’ responses to being asked to complete needs assessment questionnaires and time needed by professionals to administer and use the information generated and subsequent effect on the care process and service utilization. Evaluations to date lack a robust examination of circumstances surrounding non-response, an area that must be amplified in future work, if tools are to be sufficiently evaluated with regard to their potential clinical utility. The process of tool development is not well articulated for the majority, and steps taken to evolve tools would benefit from more elaboration. Dialogue with clinicians and patients while designing and testing new tools will contribute to the development of feasible and acceptable tools that are fit for purpose.

Tools need to be effective, that is, they should result in a direction of care that has measurable beneficial outcome in a broad spectrum of patients. Although not the focus of this review, it is worth noting that prospective trials as to the effectiveness of these tools are lacking. Whereas there is enthusiasm to deploy assessment tools in practice, attention should be directed at the measurable outcomes of care that might be improved. Assessment alone may not lead automatically to improved patient outcomes, as without subsequent care planning improved recognition of needs would not necessarily lead to improved patient management. Outcomes might include decreased levels of unmet need, enhanced patient–clinician communication, improved patient well being and satisfaction with care.

This review is not an exhaustive survey of all existing tools and has limitations. First, the search was limited to the most common databases. Second, our search strategies inevitably would have resulted in missing some publications. Its specificity was limited largely due to the many inconsistencies in the assignment of thesaurus and index terms to papers. Most of the tools were derived from hand searches, contacting experts and personal knowledge, indicating that MeSH terms are inadequate. However, probably, as has been acknowledged in other reviews of this type [19], the most important limitation was publication bias. In all likelihood, cancer care professionals have developed tools for needs assessment that they find useful without ever seeking to publish them. However, when we gathered tools currently in use from cancer care providers across England, we found that few were using tools for assessment of needs, and those that were in use were derived from tools included in this review or those with a specific focus on a single area of need [69]. The fact that healthcare professionals, where they choose to use a tool, are choosing to use custom designed ones, rather than those published in the literature, could suggest they fall short in a number of aspects.

Conclusion

During the process of conducting the review, we have come to identify the most desirable properties of tools designed to address the assessment of patient need. Similar to Osse et al. [19], we feel that tools of this nature should function as a bridge in communication between patients and healthcare professionals and promote a sense of partnership. They should address issues that patients would like professional help with and give them the opportunity to express their needs, which results in a feeling that their situation has been acknowledged. At the same time, they should provide healthcare professionals with information relevant to the planning of care. They should be easy to use from the dual perspective of patients and clinicians, easy to understand and complete and prove easy for clinicians to extract important issues at a glance.

From work to date, we conclude that the main focus for tool improvement should concentrate on aspects such as responsiveness, feasibility, appropriateness and acceptability rather than development exercises to map the range of needs to be assessed. There is, in our opinion, sufficient information to indicate the content of tools (except perhaps in relation to spiritual, cognition and occupational aspects) but a dearth of information exploring clinical utility. Tool developers should work in collaboration with patients and professionals to ensure that the product has a coherent fit with the care process.

Clearly, there is widespread recognition of the need for assessment in cancer care, as evidenced by the proliferation of tools around this issue. Assessment of patients’ needs is fundamental to understanding their experiences and developing an effective care response. The activity serves to direct professionals to who and how they might focus their care efforts. Although some of the tools in our review appear to have considerable merit, there is nothing to suggest that one is more fit for purpose than another. The lack of testing for use in practical care is a severe limitation of all the tools reviewed. This leaves many questions unanswered on how a particular tool might support patient management in the ‘real’ world of cancer care.

Acknowledgement

The work on which this paper is based was funded by the Cancer Action Team. The views expressed, findings and conclusions are those of the authors and not necessarily the views of the Cancer Action Team.

Copyright information

© Springer-Verlag 2007