Abstract
Objective:
To construct a needs assessment instrument for older people using a standardized terminology (International classification of functioning, disability, and health [ICF]) and assess its psychometrical properties.
Method:
An instrument was developed comprising questions to older people regarding their perceived care needs. The instrument’s reliability, validity, and utility were tested. Forty-one social workers and 251 older people participated.
Results:
The questions were sufficiently unambiguous (inter-rater reliability, intraclass correlation = .60–.80); measured a person’s care and service needs to a satisfactory extent (criteria validity, agreement between social workers’ and older people’s assessments = 72–94%); both social workers and older people considered the questions useful; and the needs of older people were documented in social records to a greater extent when the instrument was used.
Conclusion:
The psychometric properties of the instrument support its use by social workers to gain relevant information on elder care needs.
Keywords
Introduction
The Swedish society’s care for older people is regulated by the Socials Services Act. The Act’s principal goal is that the social services should promote economic and social security, equal living conditions, and an active participation in society. The law also states that the 290 municipalities in Sweden each have a responsibility to ensure that help and support are offered to those who need it. The municipalities have the authority to decide if public care is necessary in order to ensure a reasonable standard of living for the individual.
A basic value in Swedish welfare is that care and services should be offered to those in need. When considering elder care and service needs, social workers must try to answer two basic questions: Do a person’s problems create a nonreasonable standard of living? How can care needs best be met while adhering to relevant regulations and best known practice? The responsibility to decide on these matters places high demands on social workers. It is, therefore, important that social workers get both correct and sufficient information in order to be able to make the best possible decision. In the case of older people, the decision typically relates to whether they are in need of some kind of care or services at home or should instead be offered a place in a nursing home (or another form of residential care).
The concept of need is complex and definitions may vary depending on the context. One definition that may be of relevance for elder care concerns health and disability: the present condition should be established, a goal for health should be formulated, and a need of care exists if there is a difference between the present condition and the goal (Liss, 1990). But the significance of disabilities, physical or cognitive, may differ depending on the individual situation (Kaufman, 1994). For example, the need for help or care is often dependent on whether or not the older person has relatives who are able to provide support.
It has been argued that the needs of older people are not always given sufficient weight in decisions regarding public elder care. In particular, social, psychological, and existential needs have been shown to be considered to a lesser extent (Janlöv, 2006; Lindelöf & Rönnbäck, 2004; Olaison, 2009, 2010). Even health problems are not given sufficient emphasis in some cases (Karlsson, 2008). One explanation for these observed results may be that social service decisions must be made within set cost limitations. These restrictions may pose a dilemma for social workers since it is not always obvious which needs entitle a person to public care (Dunér & Nordström, 2006; Thorslund & Larsson, 2002).
Instruments used within the Swedish municipal elder care (Socialstyrelsen, 2002) measure, for example, cognitive ability (Folstein, Folstein, & McHugh, 1975), ability to handle everyday life (Katz, Ford, Moskowitz, Jackson, & Jaffe, 1963), well-being (e.g., depression, Yesavage & Brink, 1983), and quality of life (Hillerås, Jorm, Herliz, & Winblad, 2001). Other instruments are more comprehensive and aim to capture all information necessary for health and social care (e.g., Hawes et al., 1997). There are also simple questionnaires for self-rated health (e.g., Jylhä, 2002) and instruments for drug-related symptoms (Hedström, Lidström, & Hulter Åsberg, 2009). The majority of these instruments have been developed internationally and later translated to Swedish. To our knowledge, no instrument aiming to assess one or several elder care needs is based on a standardized terminology.
The International Classification of Functioning, Disability, and Health (ICF) is developed by the World Health Organization (WHO) as a means for international communication (WHO, 2001). ICF offers a standardized framework for terms related to health and disabilities and is based on a biopsychosocial model which includes components such as activities and participation, body functions, and body structures. It also considers contextual components such as environmental and personal factors. The structure of ICF is hierarchical and a generic scale (0 = no problem to 4 = total problem) can be used when assessing functioning and the influence of contextual factors. This classification system makes comparisons over time possible, as well as comparisons between both caregivers and countries. The ICF has been translated to Swedish and included in the national interdisciplinary terminology resources (Socialstyrelsen, 2010).
Aim
The aim of this study was to develop a relevant, reliable, and valid instrument that can be used by social workers to assess older people’s needs of care and services. Although based on ICF terminology, the instrument must be adapted to the Swedish context.
Method
An instrument was first constructed and then tested for reliability, validity, and utility in three successive phases of data collection. See Table 1 for methodological overview.
Methodological Overview.
Note. ICF = International classification of functioning, disability, and health.
Seven municipalities were selected to represent major cities of >200,000 inhabitants (one city), medium-sized cities of 50,000–200,000 inhabitants (two cities), and rural areas of <8 inhabitants per km2 (four areas; Sveriges Kommuner och Landsting [SKL], 2010). This was made in order to ensure variation in the nature of the cases, in so far as cases in elder care vary by region and type of community. Ethical approval of this study was granted by the regional ethical board in Stockholm in April 2011 (reg. No. 2011/396-31/5).
Construction of an Instrument
Thirty-five terms describing everyday activities and two terms describing safety and security were initially identified in a pilot study as particularly relevant for elder care. The first version of the instrument therefore consisted of 37 terms, formulated as questions about difficulties (ICF response scale from 0 to 4) and help or support needs (response scale = yes/no). The terms were subcategories belonging to nine different ICF domains of activities:
Learning and applying knowledge—for example, to mend something that is broken or to throw away bad food.
General tasks and demand—for example, to eat regularly, follow a medical ordination, or keep an appointment.
Communication—for example, to hear and understand speech, see and understand text, be able to speak Swedish and to use a phone, and an alarm or a computer.
Mobility—for example, to be able to get up from a chair or a bed, pick up a pen, move between different floors, go out on a balcony, or take a walk in the neighborhood.
Self-care—for example, to wash oneself, cut one’s nails, comb one’s hair, brush one’s teeth, eat healthy, or follow health advices.
Domestic life—for example, to cook and serve food, gather and throw garbage, repair clothes, maintain means of aid, and take care of plants or animals.
Interpersonal interactions and relationship—for example, to have contacts with family/relatives, friends/neighbors, or organizations such as public authorities, health centers, or the hairdresser.
Major life areas—for example, to pay bills or handle money when shopping.
Community, social and civic life—for example, to participate in a club or association, practice a hobby, visit church (synagogue, mosque) or a graveyard, or vote in general elections.
The instrument should be used by social workers (also called needs assessors) in a semi-structured interview. Preferably, the older person would answer the questions for each ICF domain and report if he or she has any difficulties and is therefore in need of help or support. A preliminary manual was formulated according to the explanations of each term given in the ICF. Some of the terms were later explained by presenting explicit and illustrative examples in a plain language.
Reliability Test—First Data Collection Phase
Social workers’ inter-rater reliability was tested during assessments of elder care needs using the first version of the instrument. The purpose was to measure the agreement among social workers when interpreting an older person’s answers to each question. Forty-one female social workers participated, with an average of 7.60 years (standard deviation (SD) = 6.39, range = 1 month–30 years) of work experience in the field.
The older people were recruited prospectively during September to December 2011. Those who only applied for security alarms or meal distribution as well as those who were only partook in discharge care planning at a hospital were excluded. The older people were first informed verbally about the study and then asked if they wanted to participate. Those who agreed to participate were given information on a consent form that they had to sign before participating.
One hundred and fifty one older people participated (99 women and 52 men), with an average age of 83.57 years (SD = 7.51, range = 58–97). All applied for some kind of elder care or were subject to a follow-up. The objective was to collect data from about 150 cases for double assessments in order to achieve sufficient data for statistical calculations (Donner & Eliasziw, 1987). Four older people declined to participate due to fatigue, illness, or poor hearing during the initial phone contacts. Two persons declined as they were afraid to let two social workers enter their home due to recent media coverage reporting people who had entered homes with false social worker identification.
The social workers worked in pairs. Both social workers in each pair conducted independent assessments of each older person using the instrument. Although one social worker interviewed the older person, the other strictly observed the dialogue.
Validity Test—Second Data Collection Phase
Criteria validity was tested with help from social workers, using a revised, second version of the instrument (see Results section). The social workers now worked individually as they usually do. The social workers also rated if they themselves found that the older person needed help based on each of the ICF activity domains. The agreement between older people’s and social workers’ ratings of elder care needs was calculated. The social worker’s assessment was regarded as a criterion. This is because it was made with reference to all available information such as information from relatives, medical staff, and documents, in addition to the social worker’s own observations during the meeting with the older person. The social worker’s assessment is of particular relevance as it is the social worker who makes the formal decision about public elder care, a decision which may be appealed in court. Twenty-three social workers participated, with an average of 8.61 years (SD = 6.11, range = 1 month–22 years) of work experience in the field.
The older people were recruited prospectively among those who applied for elder care or were subject to a follow-up during January to April 2012. Individuals subject to discharge care planning in hospitals and those who had a dementia diagnosis were excluded. As in the first data collection, the older people were informed verbally and asked if they wanted to participate. Those who agreed to participate were given information on a consent form that they had to sign before participating.
One hundred older people participated (70 men and 30 women), with an average age of 82.47 years (SD = 7.12, range = 56–99). About 100 cases were estimated to be sufficient for this particular test (Gardner & Altman, 1989).
In addition, ratings from the first data collection were analyzed using factor analysis in order to obtain a measure of construct validity (i.e., the extent to which the ICF domains are confirmed empirically by the present data).
Test of Utility—Third Data Collection Phase
The test of utility consisted of two parts. First, information was collected during the first two data collection phases described previously. All social workers answered a questionnaire (eight questions, Appendix C) about the usefulness of the instrument. Furthermore, the social workers also asked if the older person had any comments on the instrument. A total of 201 older people gave such comments.
Second, social records from assessments of 40 cases where the instrument was used were examined and compared to 40 cases where the instrument was not used. These comparisons were made in order to determine whether or not use of the instrument contributed to an increased identification of older people’s needs. Each record chosen for comparison was matched according to the complexity of a case where the instrument was used in order to ensure that the matched case pairs represented as similar cases as possible. These 80 case records were selected by the social workers, copied, and edited in order to ensure anonymity.
Data Analysis
Measures of inter-rater reliability were obtained by calculating the agreement between social workers on the 5-degree ICF scale (estimates based on pairwise comparisons for each question and each older person), using intraclass correlations (ICC).
Measures of criteria validity were obtained by calculating the agreement between the older person’s answer (Do you need help? Yes/No) and the social worker’s assessment (Public care? Yes/No). The calculations were based on individual pairwise comparisons and reported as percentages. In addition, construct validity (if overall domains of activity could be identified) was explored using exploratory factor analysis on the data used for the reliability test. The factor analysis method applied was oblique with maximum likelihood extraction and varimax rotation.
Both older people’s and social workers’ reported opinions about usability were analyzed descriptively. Two people independently categorized the responses as negative, neutral, or positive. The social records were reviewed and the number of ICF activity domains mentioned in each was recorded. The review was conducted independently by three reviewers using an audit form. One of the reviewers was blinded to the coding (with or without instrument). The initial agreement between reviewers was in total 86% (113 of 800 variables [10 variables × 80 records] were rated differently). The variables that were coded differently by the reviewers were discussed until consensus was reached.
Results
In total, 41 social workers and 251 older people participated in the 3 data-collection phases. Twenty-three of the social workers participated in both the reliability and the validity studies.
Reliability
The degree of pairwise agreement between social workers regarding their interpretation of the older people’s answers to the 37 questions in the first version of the instrument varied between .60 and .88 per question (Table 2). Although the degree of agreement differed between questions, the results can be considered to be generally good (Fleiss, 1986).
Summary of Results From the Three Data Collection Phases.
Note. ICC = intraclass correlation; ICF = International classification of functioning, disability, and health. Type of test, analysis, type and number of participants, and number of ICF variables.
However, the two terms with the lowest correlation coefficients were perceived as difficult to distinguish from other terms by the majority of social workers. These two terms were therefore excluded in the second version of the instrument (the terms are Sense of security and To maintain basic body position). The social workers also suggested some additional areas of importance for older people. As a result, seven questions were added in the second version: (i) solving problems; (ii) complex economic transactions; (iii) voting and change the rest of the sentence to (iv) feeling sad; (v) feeling loss of appetite; (vi) housing; and (vii) personal support.
Validity
The agreement between the older person’s opinion of whether she or he felt a need for help or support and the social worker’s opinion of the older person’s need for public care was tested using the 42 questions in the second version of the instrument. The degree of agreement varied between 72% and 94% per ICF domain (Table 2). It can, therefore, be concluded that the older people and the social workers were in agreement as regard the need for help in the majority of cases.
The cases where the social workers and older people did not agree were examined more closely. The most common point of disagreement (4–18%, depending on the ICF domain) was when a social worker assessed that there was a need for help despite the fact that the older person did not want any help.
A further measure of validity was obtained using a factor analysis with data from the reliability study. Eight factors with eigenvalues above 1.0 were initially extracted. Most of the 37 variables loaded positively onto several factors (Appendix A). Only three ICF domains were identified as coherent factors: (i) Domestic life; (ii) Interactions and relationships; and (iii) Community, social, and civic life. All eight factors correlated to varying degrees. These results suggest that, when applied in the context of elder care, the different activity domains in the ICF are not distinct but overlapping as a specific activity may belong to several domains. For example, if a person has mobility problems, he or she is also likely to have difficulties managing domestic life.
Overall, the results suggest that the questions based on ICF terms sufficiently captured the older person’s perception of his or her need for help or support. Furthermore, the ICF activity domains were not independent but instead correlated to one another.
Utility—Perceived Usefulness
The majority (81%, 163 of 201 people) of the older people who gave their opinion of the questions were generally neutral or positive. They described the questions as good, easy to respond to, and/or relevant. Of these 163 responses, 76 were neutral, 4 people saw both advantages and disadvantages, and 83 gave explicitly positive comments. They reported, for example, that standardized questions can make it easier for them to describe and to reflect about their life situation in general. The remaining 38 of the 201 people’s responses reflected more negative views, for example, that there were too many questions, that some questions felt intrusive, and that it was hard to grade the difficulties they had.
The majority (88%, 36 of 41) of the social workers reported that the instrument was both useful and contained relevant questions, but that some of the questions were difficult to ask an older person. Six social workers also pointed out that it took longer at first to complete the interview based on the instrument as they were unaccustomed to it. Overall, the social workers expressed that the instrument provided both advantages and disadvantages when compared to assessing the need of older people without any instrument. Examples of disadvantages included difficulties using a response scale to assess the degree of difficulty; that certain questions were not included (mainly about health); that some questions were difficult to ask (mostly on religion, relationships, and decision making); that the conversation became rigid when using the instrument; and that the terms were not adapted to plain language. Advantages of using the instrument included the experience that reticent people can be induced to express themselves; that the questions can help the older person to reconsider his or her situation; that the ICF domains cover an older person’s daily life to a large extent; that the instrument gives a structure to the assessment process; and that it may clarify an older person’s difficulties in ambiguous cases.
The time to work through the instrument for an individual older person varied between 10 minutes and 2 hours, depending on the complexity of the case.
Utility—Documented Aspects of Everyday Life
A greater number of ICF activity domains were mentioned in social records when the instrument was used than when no instrument was used. In particular, activities such as Major life areas (here only financial transactions) and Social community, social, and civic life were documented to a greater extent when the instrument was used (Table 2).
Final Version
A final version of the instrument was developed after the three data collection phases were completed. The instrument can be described as a questionnaire for a semi-structured interview, where questions with fixed response alternatives are complemented with the possibility to give more details in free text. The instrument consists of nine general questions regarding difficulties in everyday life. If the older person reports difficulties on any of these general domains, more detailed questions can then be asked based on the 42 ICF codes (see Table 3). The social worker is supposed to ask if the older person wants help or support every time he or she reports a difficulty on a general domain. Additional questions address home and personal support respectively (two questions) and how the person has been feeling lately (four questions).
ICF Domains With Codes for Detailed Questions in the Final Version of BAS.
Note. BAS = Behov Av Stöd; ICF = International classification of functioning, disability, and health.
The instrument should be used by social workers when interviewing people applying for elder care or when current elder care is followed-up. Both the instrument and the manual are available in Swedish at the website of the National Board of Health and Welfare, www.socialstyrelsen.se (in Swedish BAS = Behov Av Stöd, corresponds to Need Of Care). For an English version of the instrument, please see Appendix B.
Discussion
The psychometric testing showed that the questions in the instrument were sufficiently unambiguous to be useful in the context they were developed for; that the questions measure different aspects of an older person’s perceived need of social care to a satisfactory extent; that the majority of both older people and social workers considered the questions to be useful; and that the needs of older people were documented to a greater extent in social records when the instrument was used compared to when no instrument was used.
Although the instrument mainly comprises questions about everyday activities, an interview can touch on many different aspects of a person’s situation. Physical, psychological, social, and existential aspects may be discussed if the older person so wishes. The various ICF domains are overlapping. This is not surprising as, for example, physical ability may affect responses on the ICF domains Mobility, Self-care, Domestic life, and Communication. Mental functions in terms of cognitive functions could affect responses on the domains Learning and applying knowledge, General tasks and demand, and Major life areas. Mental functions in terms of emotional aspects can also be expected to affect responses to the questions exploring safety, sadness, and loss of appetite. Furthermore, social circumstances can reflect not only responses on the domain Interpersonal interactions and relationships but also existential issues (e.g., to participate in ceremonies).
The responses to a single instrument cannot provide all possible relevant information when making elder care decisions (Chernesky & Gutheil, 2008). An instrument can, however, contribute to more systematic assessments, which can promote social justice and older people’s rights to individualized care and services. Unlike other instruments used in elder care (e.g., Activities of Daily Living [ADL] or Resident Assessment Instrument [RAI] [Hawes et al., 1997; Katz et al., 1963]), the instrument developed and presented in this study (called BAS) is designed to enable a comprehensive assessment of how older people perceive their need of care. Furthermore, to our knowledge, BAS is the first instrument that aims to capture the need of elder care based on a standardized international terminology.
Still, BAS has several limitations. For example: (a) it is often necessary to collect additional information about, for example, medical diagnoses, housing, and/or family relationships; (b) as elder care in Sweden should be adapted to each individual’s unique situation, no cutoff level was established; (c) the instrument is yet to be tested on people with dementia or when cognitive impairment is suspected, since needs assessment in these cases often requires specific information from relatives; and (d) reliability testing was restricted to inter-rater reliability. A test–retest was judged problematic due to systematic error as a result of the potential change in older people’s needs over time; (e) there was variation in the social workers’ level of work experience and the most inexperienced social workers’ ratings may be questioned. Nonetheless, the variation in work experience captured here reflects the population of social workers; (f) the social records used for the test of utility were chosen by the social workers themselves. This procedure may have biased the results and the measure of utility may, therefore, benefit from further testing; (g) as the national terminology for health and social care is underdevelopment, the instrument may need to be revised; (h) the instrument’s development was constrained by its adaption to the ICF terms, which generally were experienced as difficult to interpret. An introduction to the instrument is therefore considered necessary before use that clarifies the meaning of the terms and how to use the instrument.
In conclusion, the psychometric properties of the BAS instrument support its use by social workers to systematically gain relevant information on elder care needs. BAS can therefore provide a valuable tool when making assessments and decisions regarding the need of elder care.
Footnotes
Appendix A
Summary of Communalities (h 2) and Main Loadings (F 1–4) for the Factor Analysis for the First Version of the Instrument BAS (Need of Care).
| ICF domains | ICF items | Factor Loadings | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| h 2 | F 1 | F 2 | F 3 | F 4 | F 5 | F 6 | F 7 | F 8 | ||
| Learning and applying knowledge | Solving problems/making decisions | .521 | .014 | −.101 | .216 | .078 | −.087 |
|
.204 | .206 |
| General tasks and demand | Carrying out daily routines | .511 | .224 | .142 | .257 | .006 | .088 |
|
.130 | .170 |
| Communication | Speaking | .082 | −.052 | −.052 | −.050 | .022 | −.039 | .044 | .073 | −.019 |
| Receiving spoken messages | .417 | .062 | −.074 | .074 | .224 | .023 | .224 | .299 | .430 | |
| Receiving written messages | .697 | .110 | −.150 | .183 |
|
.126 | .179 | .087 | .167 | |
| Writing messages | .983 | .206 | .196 | .193 |
|
−.043 | .025 | .110 | .151 | |
| Using communication devices | .704 | .074 | .183 | .125 | .415 | −.128 |
|
.285 | .055 | |
| Mobility | Changing basic body position | .694 | .170 |
|
.010 | .000 | .266 | −.075 | .076 | .097 |
| Keeping basic body position | .623 | .152 |
|
.096 | −.232 | .172 | −.100 | −.020 | .183 | |
| Lifting and carrying objects | .579 | .499 | .393 | .027 | .121 | .307 | −.051 | .172 | −.050 | |
| Fine hand use | .592 | .186 |
|
.068 | .123 | .051 | −.016 | .187 | −.049 | |
| Walking | .549 | .235 | .474 | −.019 | −.036 | .488 | −.107 | .125 | .024 | |
| Climbing (stairs) | .629 | .235 | .352 | .047 | −.030 |
|
−.054 | .108 | .153 | |
| Moving around within the home | .701 | .156 |
|
.117 | −.052 |
|
.045 | −.133 | .06 | |
| Moving around outside home | .879 | .275 | .271 | .191 | .179 |
|
.081 | .036 | .145 | |
| Self-care | Washing one self | .626 | .452 | .469 | .006 | .037 | .173 | .194 | −.123 | .334 |
| Caring for body parts | .593 | .427 | .283 | .212 | .211 | .230 | .355 | −.081 | .103 | |
| Toileting | .712 | .219 |
|
.032 | .084 | −.011 | .275 | −.027 | −.024 | |
| Dressing | .687 | .380 |
|
−.024 | .101 | .190 | .136 | −.179 | −.035 | |
| Eating and drinking | .593 | .307 |
|
−.010 | .078 | .048 | .315 | −.183 | −.044 | |
| Looking after one’s health | .537 | .491 | .242 | .250 | .067 | .016 | .340 | −.061 | .062 | |
| Domestic life | Shopping | .631 |
|
.170 | .068 | .198 | .277 | .046 | .154 | .078 |
| Preparing meals | .677 |
|
.184 | .091 | .087 | .256 | .203 | −.009 | .080 | |
| Cleaning cooking area and utensils | .730 |
|
.196 | .066 | −.075 | .115 | .058 | .136 | −.022 | |
| Disposing of garbage | .711 |
|
.235 | .282 | .190 | .080 | .042 | −.070 | .188 | |
| Washing/drying clothes/garments | .555 |
|
.148 | .103 | .188 | .005 | .094 | .078 | .176 | |
| Cleaning living area | .532 |
|
.130 | −.115 | −.005 | .138 | −.082 | −.091 | .270 | |
| Caring for household objects | .968 |
|
.376 | .191 | .114 | .038 | .127 | .130 | .033 | |
| Interpersonal interactions/relationship | Formal relationship | .548 | .101 | .012 | .182 | .134 | .037 | .351 | .159 | .486 |
| Informal social relationship | .576 | .126 | .174 | .310 | .031 | .105 | .040 | .032 |
|
|
| Family relationship | .324 | .174 | −.006 | −.032 | .129 | .081 | .105 | .050 |
|
|
| Major life areas | Basic economic transactions | .531 | .249 | .060 | .297 | .428 | .138 | .352 | .142 | .169 |
| Community, social and civic life | Community life | .755 | .165 | .068 |
|
.133 | −.027 | .124 | .013 | .108 |
| Recreation and leisure | .922 | .047 | .061 |
|
.196 | .135 | .224 | .055 | −.048 | |
| Religion and spirituality | .696 | .108 | −.011 |
|
.086 | .088 | .118 | .080 | .173 | |
Note. Loadings >.5 are marked bold.
Appendix B
Appendix C
Acknowledgments
The authors would like to express their thanks to the social workers and older people who participated in the data collection, Bo Ekehammar and Knut Sundell for valuable comments on the analyses, and AnnaLena Hållner, Svend Erik Mathiassen, and Sandra South for valuable comments on the article.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The study has been financed by the Ministry of Health and Social Affairs.