The Frenchay Activities Index

Transcription

1 1173 The Frenchay Activities Index Assessment of Functional Status in Stroke Patients J. Schuling, MD; R. de Haan, MS; M. Limburg, MD, PhD; and K.H. Groenier, MS Downloaded from by on January 1, 2019 Background and Purpose: Assessment of functional status in stroke patients is of major importance in both clinical practice and outcome studies. The Frenchay Activities Index has been developed specifically for measuring disability and handicap in stroke patients. The purpose of the study was to evaluate the metric properties of this instrument and to obtain normal values in a group of unselected elderly subjects. Methods: The Frenchay Activities Index was tested in a group of stroke patients and a group of unselected subjects aged 65 or older. The functional status of the stroke patients was measured 26 weeks after stroke. Their prestroke status was registered retrospectively. Reliability and validity of the instrument were assessed. Results: The mean scores in the prestroke, poststroke, and control group demonstrated differences in functional status. The reliability of unweighted scores (range of Cronbach's et-coefficients, 0.78 to 0.87) was sufficient. The construct validity was supported by meaningful correlations between the Frenchay Activities Index and scores on the Barthel Index and Sickness Impact Profile. Principal-components analysis indicated that the Frenchay Activities Index showed two traits: instrumental disability and some aspects of handicap. The reliability of the instrument could be improved by deleting two items and by creating two subscale scores: domestic and outdoors activities. Conclusions: The Frenchay Activities Index is a useful stroke-specific instrument to assess functional status. Completion of the questionnaire is easy and takes only a few minutes. Currently, the instrument is suitable for use in patient care and cross-sectional, descriptive studies. (Stroke 1993;24: ) KEY WORDS * activities of daily living * cerebrovascular disorders * stroke assessment N o doubt the most important outcome of illness is survival. However, a wide range of possibilities lies between death and complete recovery. To assess these outcomes, the concept of functional status has been introduced. Functional status refers to the ability of people to look after themselves and to perform certain roles and tasks.'2 The number of instruments that have been developed to answer questions such as "Which activities of daily living can a patient perform?" and "To what degree is the patient dependent on the help of others?" is impressive. Some of these instruments are not developed for a specific target population and may be suitable for use in many patient populations (generic measures); other instruments are more sensitive to functional issues particularly relevant to a specific population of patients (disease-specific measures).3-5 Assessing the patients' disabilities and handicaps is of great importance in cerebrovascular research and patient care. Disability refers to the consequence of neurological impairments in terms of the patient's functional performance. Handicaps are concerned with the Received September 30, 1992; final revision received February 19, 1993; accepted March 18, From the Department of General Practice, University of Groningen (J.S., K.H.G.), and the Departments of Neurology (R. de H., M.L.) and Clinical Epidemiology and Biostatistics (R. de H.), Academic Medical Center, University of Amsterdam, the Netherlands. Correspondence to Department of General Practice, University of Groningen, A Deusinglaan 4, 9713 AW Groningen, the Netherlands (Dr Schuling). social disadvantages resulting from impairments and disabilities.6 Assessment of these outcomes allows one to adapt treatment and circumstances to the patient's needs, to enhance supportive care, and to evaluate the efficacy of therapeutic interventions.7-9 Most activities of daily living (ADL) scales, however, do not refer to the patient's ability to perform complex activities such as housekeeping, recreation, hobbies, and social interaction Because these so-called instrumental disabilities (IADL) may affect the quality of life considerably, they should be given due attention. Before setting goals for rehabilitation, one should obtain accurate information on the premorbid life-style of stroke patients; furthermore, to evaluate treatment programs, poststroke changes in activities should be recorded at specific time intervals so that a therapeutic strategy can be chosen and adjusted when necessary. However, in a busy office practice instruments to measure the patient's functional activities routinely should be concise and easy to understand. Holbrook and Skilbeck13 constructed an instrument that appears to meet these demands: the Frenchay Activities Index (FAI). This instrument measures activities that reflect a higher level of independence and social survival.13-1' Considering the simplicity of this instrument and its potential value in stroke rehabilitation studies, the FAI may prove to be appropriate to measure functional outcome in stroke patients.16 Studies concerning its reliability and validity, however, are rare. The purpose of this study is (1) to examine the metric properties of the FAI in prestroke, poststroke,

2 1174 Stroke Vol 24, No 8 August 1993 and nonstroke patient samples; (2) to compare the functional activities of stroke patients with those of an unselected group of elderly subjects in the community; and (3) to provide normal values of the FAI. Subjects and Methods The FAI comprises 15 items, each concerning an activity that requires some decision making and organizing on the part of the patient at home as well as outside the home (see also "Appendix"). The instrument depends on the patient's own report or on the report of relatives. Data can be collected either by means of an interview or a mailed questionnaire. The FAI consists of a single summary score (with a range of 15 to 60 points) as well as three subscale scores: domestic, leisure/work, and outdoors. We translated the FAI and tested it in a stroke group in a direct interview and in a control group of elderly people as a mailed questionnaire. In the stroke group the prestroke functional status was measured retrospectively (A); the present status was measured at 26 weeks poststroke (B). To measure prestroke functional status retrospectively, during the first 9 months of 1990, general practitioners in the northern region of the Netherlands reported all new patients with a stroke (first ever as well as recurring) to the research team.17 Patients were interviewed at fixed intervals up to 6 months after stroke. At the end of the first week patients were visited by a member of the research team to confirm the diagnosis. The patients' prestroke functional status was assessed by means of the FAI. No intervention took place; with the exception of patients living in a nursing home or having a stroke during a hospital admission, all stroke patients were included. To measure present functional status, at 26 weeks after stroke the interviewer completed the Barthel ADL Index, the FAI, and the Sickness Impact Profile. For the control group, all patients (aged 65 years or older) registered on the practice list of four general practitioners in the city of Groningen who were living at home received the FAI by mail; since virtually every Dutch citizen is registered in the practice of a general practitioner, these patients form a representative sample of elderly people living independently in an urban area in the Netherlands. We evaluated the homogeneity and validity of the FAI. Homogeneity (or internal consistency) was assessed by the Cronbach's a-coefficient.18 An a-coefficient greater than.80 is considered sufficient. If a scale or subscale merely intends to measure on group level, a more liberal standard (coefficient greater than.60) will be applied. To assess convergence and discriminant validity, the FAI was correlated with the Barthel Index and the Sickness Impact Profile.19 We assumed that, for the FAI to be valid, the scale scores had to correlate significantly with the disability scores of the Barthel Index and the subscales of the Sickness Impact Profile measuring ambulation, mobility, body care/movement, and home management. We also expected a substantial correlation between the FAI and the subscales of the Sickness Impact Profile measuring the amount of rest needed, social interaction, and time spent on recreation. Furthermore, we supposed that patients having dysfunctional scores on the Sickness Impact Profile's subscale of communication (and thus generally having larger TABLE 1. Score Distributions of Frenchay Activities Index in Percentages and Mean Sum Scores and Variances of Absolute Values Group FAI scores Control (n=216) Prestroke (n=92) Poststroke (n=96) Mean % CI SD FAI, Frenchay Activities Index; CI, confidence interval. strokes) would have a low FAI score. In addition, to demonstrate discriminant validity, we would expect the FAI scores to be unrelated to Sickness Impact Profile items focusing on emotional behavior (eg, "I laugh or cry suddenly," "I act irritably and impatiently with myself'), alertness (eg, "I react slowly to things that are said or done," "I do not finish things I start"), and eating. Finally, the construct validity of the FAI was evaluated by way of principal-components analysis. This procedure, based on the relations between the scale items, identifies a limited set of underlying dimensions (or factors) of a scale. The differences between mean sum scores of the three groups were analyzed with 95% confidence intervals (CIs). The reported correlations were calculated with Pearson's correlation coefficients. Results At the end of the study period, data on 185 stroke patients were reported. At the end of the 6 months' follow-up, 63 (34%) patients had died; a complete data set could not be obtained in all of the remaining 122 patients because of the severity of their clinical condition. An FAI score at 26 weeks after stroke was obtained for 96 patients, and 92 patients also completed the prestroke FAI. The mean age of the group was 74 years (median+sd, ; range, 41 to 92 years); 41% were male. Time needed to complete the FAI questionnaire was less than 5 minutes. Communication difficulties did not impede the completion of the FAI in any case. The FAI was mailed to 332 nonstroke control patients. We received 216 (65%) questionnaires that were correctly answered. The mean age of the respondents was 74 years (median+sd, ; range, 65 to 91 years); 36% were male. Table 1 reports the distribution of the scores and the means and SDs of the total scores. The mean total score in the control group made clear that in an unselected group of elderly subjects, impairment of functional status was quite common. The impact of the stroke

3 Reliability Coefficients Indicating Homogeneity of TABLE 2. Frenchay Activities Index Group Control (n=216) Prestroke (n=92) Poststroke (n=96) Mean interitem correlation a Total scale Domestic domain (items 1-5) Mean interitem correlation a Subscale Leisure/work domain (items 7, 9, 11, 13, 15) Mean interitem correlation a Subscale a If item 15 deleted Outdoors domain (items 6, 8, 10, 12, 14) Mean interitem correlation a Subscale a If item 14 deleted appeared to be substantial. In comparison with both prestroke functioning and reference scores, there was a substantial shift of the scores to the lower end of the scale (95% CI of the difference of the mean scores between poststroke and prestroke functioning, 4.03 to 7.41; 95% CI between poststroke group and reference group, 8.29 to 13.1). The prestroke FAI scores were lower than the scores of the reference group (95% CI of the difference, 2.68 to 7.24). Table 2 presents the reliability coefficients for both the total FAI and its three subscales: domestic, leisure/ work, and outdoors. The data indicated that the FAI was a homogeneous scale. The Cronbach's a-coefficients met the standards set previously. The data of the retrospective prestroke measurement, however, were less convincing. The leisure/work and outdoors domains included two weak items: 14 (reading books) and 15 (gainful work). Removal of these items had a positive effect on the reliability of the subscales. In the above analyses the unweighted scale scores were used. The same analyses, based on the weighted scores as suggested by Wade et al,14,15 did not improve the scale's reliability. Moreover, it had a rather negative effect (Cronbach's a of the weighted scores; poststroke group, a=.83; prestroke group, a=.70; control group, a=.77). The suggested weights were based on expected differences in sex. Separate analyses in our samples, however, did not show differences of mean total scores between men and women (95% CI of difference of the mean scores in the poststroke group, to.80; prestroke group, to 4.93; and control group, to.49). Neither did men and women in the three samples differ on the mean subscale scores (the nine 95% CIs are available on request). Furthermore, a-coefficients did not indicate that men or women scored with more reliability. Support for the construct validity of the instrument is shown in Table 3. As expected, there was a substantial convergent relation between the total scores of the FAI and the disability scores of the Barthel Index. This was also true for the subscales of the Sickness Impact Profile measuring household activities and physical functioning Schuling et al Frenchay Activities Index 1175 TABLE 3. Convergent-Discriminant Validity as Shown by Pearson's Correlation Coefficients (26 Weeks After Stroke) FAI n* Barthel Index Subscales of the Sickness Impact Profilet Home management Body care and movement Mobility Ambulation Recreation/pastimes Communication Eating Rest/sleep Social interaction Emotional behavior Alertness behavior FAI, Frenchay Activities Index. *The number of subjects varies because only those were involved who did not have missing values on the respective subscales. tbecause higher scores on the Sickness Impact Profile indicate more dysfunction, the correlations are negative. The subscale "work" and the item "sexual function" on the subscale "social interaction" were not included in the analysis because of too many missing values. (body care, mobility, and ambulation). Discriminant validity was supported by low correlations between the FAI and the subscales of emotional and alertness behavior. Table 4 reports the results of a principal-components analysis that showed that most of the score variance could be attributed to two factors, which explained 52% of the total variance in the stroke group. The first factor was closely related to the domestic activities as measured by items 1, 2, 3, and 4. Much weight was assigned TABLE 4. Principal-Components Analysis: Rotated* Loadings for Two-Factor Solution (Poststroke Group, n=96) Loadingst Loadings Item factor 1 factor 2 Communalityt 1. Preparing meals Washing up Washing clothes Light housework Driving/bus travel Outings/car rides Gardening House/car maintenance Social outings Pursuing hobby Heavy housework Local shopping Walking outdoors Reading books Gainful work *Rotation is a statistical procedure to facilitate the identification of the underlying dimensions. tloadings represent the correlation coefficients between factors and items. 4:Communality indicates the proportion of variance of each item explained by both factors.

4 1176 Stroke Vol 24, No 8 August 1993 Downloaded from by on January 1, 2019 to the second factor by items 10, 11, 12, and 13. The items that reflect a social component (7 and 9) were also moderately related to this factor. Items 5, 6, and, to a lesser extent, 8 were loading onto both factors. The score variances of items 14 and 15 could not be explained by the two factors. In the unselected group this same analysis produced even more definite results: items 1 to 4 loaded onto factor 1; items 6 to 13 onto factor 2; item 5 loaded onto both factors; and an absence of communality of items 14 and 15. According to the above results, the reliability of the FAI could be further improved by introducing two subscales: (1) the original domestic subscale as presented in Table 2 and (2) an outdoor subscale containing items 6 to 13 (stroke group, a=.82). Discussion The FAI proved to be a homogeneous scale that showed substantial validity. The instrument had the ability to distinguish between stroke patients' present and prestroke functioning and the functional status of unselected elderly. It appeared to have no ceiling effect. The FAI could be completed within a few minutes by means of a direct interview and as a mailed questionnaire, which enhances its value for researchers. The FAI measured, for the greater part, two concepts: (1) instrumental disabilities: indoor (items 1 to 5) and outdoor (items 6, 8, and 10 to 13) physical activities and (2) some elements of the handicap concept (items 7 and 9). We recommend deleting two clinimetrically weak items: "gainful work" and "reading books." Although, in general, the ability to work is an important indicator of handicap, this item is of little informative value in stroke research. Because stroke is a disease of the elderly, most patients are already retired at stroke onset. Because the scores on the item "reading books" were low in the three samples and hardly changed between prestroke and poststroke measurement, this item has little discriminative value as well. The weighting of scores is not necessary. Preferably a sum score is calculated by simply adding the ordinal values of the 13 items. If one wants to distinguish between domestic and outdoor activities, one can also calculate two summated subscale scores. Although the 65% response rate in our control group is not perfect, it is quite acceptable for a mailed questionnaire survey.20 The nonresponders did not differ from the total group with regard to the distribution of the characteristics sex and age, but they probably were more disabled than the responders. The missing data in the prestroke group were due to the fact that some patients were not able or preferred not to be interviewed completely because of the severity of their clinical condition. However, because prestroke and poststroke groups comprise the same patients, our conclusion, that the FAI can discriminate between prestroke and poststroke functional status, is not affected by the missing data. Of the great number of measuring instruments, we would like to discuss three instruments that could be considered as alternatives to the FAI. The Rankin scale, a five-grade modified handicap scale and by far the oldest scale, is suitable for epidemiological purposes. period of time (responsiveness) and would be less useful for monitoring progress or measuring detailed outcome The Functional Independence Measure is a comprehensive scale that includes ADL items, including a wide range of items concerning social behavior, communication, and cognitive impairment.24 This aggregation of neurological impairments and aspects of disabilities is conceptually confusing and masks the clinical meaning of total scores. Reliability is established for use with four-level responses but not for seven-level responses. The score weights are arbitrary, so that different disabilities cannot be compared.25 The Lawton scale is a Guttman scale of items concerning physical selfmaintenance and IADL. Its metric properties are well documented.2627 While the former section resembles the Barthel ADL Index, the content of the latter shows great similarity to the FAI. In our opinion, only this instrument provides a real alternative for the FAI. We chose the FAI because initially this instrument was developed specifically for stroke patients and because, compared with ADL scales, there is a need for supplementary data concerning the validity and reliability of IADL instruments. Like most disability scales, the FAI refers to a typical middle-class, western way of life, which limits its applicability to this kind of society. The usefulness of the instrument in patients with cognitive impairment or with aphasia needs further research, specifically with regard to the possibility of completion by relatives or friends. Although our data did support relevant clinimetric qualities of the FAI, future studies to assess the stability of the ratings (eg, interobserver reliability) and its ability to detect important health changes over a period of time (responsiveness) are necessary. Some support for responsiveness to within-patient changes over a period of time has been demonstrated in long-term follow-up.16 Currently, the FAI is suitable for use in both patient care and cross-sectional, descriptive outcome studies. In these latter studies, age-matched control groups should be viewed as an essential component of outcome research. Such a control group allows one to distinguish between disability and handicap effects related to the disease and those attributable to the aging process per se. As we have demonstrated, these normal values were not exchangeable with retrospective measurements of prestroke functioning. This was probably the result of increased preexisting morbidity in these patients and lower reliability of retrospective measuring. Appendix The Frenchay Activities Index15 Item Code In the last 3 months 1. Preparing main meals 1 =never 2. Washing up 2= < 1 time per week 3=1-2 times per week 4=most days 3. Washing clothes 1=never 4. Light housework 2=1-2 times in 3 months 5. Heavy housework 3=3-12 times in 3 months However, it probably lacks sensitivity to change over a 6. Local shopping 4 =..t 1 time per week

5 Item 7. Social outings 8. Walking outside > 15 minutes 9. Actively pursuing hobby 10. Driving car/bus travel In the last 6 months 11. Outings/car rides 12. Gardening 13. Household/car maintenance 14. Reading books 15. Gainful work Code 1 =never 2=1-2 times in 6 months 3=3-12 times in 6 months 4=.1 time per week 1 =never 2=light 3 = moderate 4=all necessary 1 =none 2=1 in 6 months 3=<1 in 2 weeks 4= > 1 in 2 weeks 1 = none 2=<10 h/wk 3=10-30 h/wk 4= >30 h/wk Acknowledgment This study was supported by De Nederlandse Hartstichting grants NHS and NHS References 1. Hall J, Masters G, Tarlo K, Andrews G. Measuring Outcomes of Health Services. Westmead, UK: Dept of Community Medicine, Westmead Centre; Nelson EC, Wasson J, Kirk J, Keller A, Clark D, Dietrich A, Stewart A, Zubkoff M. Assessment of function in routine clinical practice: description of the COOP Chart method and preliminary findings. J Chron Dis. 1987;40(suppl 1):55S-64S. 3. Bergner M. Measurement of health status. Med Care. 1985;23: Advances in health status assessment: conference proceedings. Med Care. 1989;27(suppl):1. 5. Bergner M, Kaplan RM, Ware JE. Evaluating health measures commentary: measuring overall health: an evaluation of three important approaches. J Chron Dis. 1987;40:23S-26S. Schuling et al Frenchay Activities Index World Health Organization. International Classification of Impairments, Disabilities and Handicaps. Geneva, Switzerland: World Health Organization; Gresham GE. Stroke outcome research. Stroke. 1986;17: Dombovy ML, Sandok BA, Basford JR. Rehabilitation for stroke: a review. Stroke. 1986;17: Wood-Dauphinee SL, Ivan Williams J, Shapiro SH. Examining outcome measures in a clinical study of stroke. Stroke. 1990;21: Wade DT, Collin C. The Barthel ADL Index: a standard measure of physical disability. Int Disability Studies 1988;10: Moskowitz E, McCann CB. Classification of disability in the chronically ill and ageing. J Chron Dis. 1957;5: Katz S, Ford AB, Moskowitz RW, Jackson BA, Jaffee MW. Studies of illness in the aged: the index of ADL: a standardized measure of biological and psychosocial function. JAMA. 1963;185: Holbrook M, Skilbeck CE. An activities index for use with stroke patients. Age Ageing. 1983;12: Wade DT, Legh-Smith J, Langton Hewer R. Social activities after stroke: measurement and natural history using the Frenchay Activities Index. Int Rehabil Med. 1985;7: Wade DT, Langton Hewer R, Skilbeck CE, David RM. Stroke: A Critical Approach to Diagnosis, Treatment and Management. London, England: Chapman and Hall; 1985: Greveson GC, Gray CS, French JM, James OFW. Long-term outcome for patients and carers following hospital admission for stroke. Age Ageing. 1991;20: Schuling J, Greidanus J. Stroke patients in the acute phase: characteristics and management. Ned Tijdschr Geneeskd. 1992;136:424. English Abstract. 18. Cronbach ILJ. Coefficient alpha and the internal structure of tests. Psychometrika. 1951;16: Bergner M, Bobbitt RA, Carter WB, Gilson BS. The Sickness Impact Profile: development and final revision of a health status measure. Med Care. 1981;19: Dillman DA. Mail and Telephone Surveys: The Total Design Method. New York, NY: John Wiley & Sons, Inc; 1978: Rankin J. Cerebral vascular accidents in patients over the age of 60, II: prognosis. Scott Med J. 1957;2: Adams GF. Grading of functional recovery from strokes. Age Ageing. 1975;4: Ebrahim S. Clinical Epidemiology of Stroke. Oxford, England: Oxford Medical Publications; 1990: Keith RA, Granger CV, Hamilton BB, Sherwin FS. The Functional Independence Measure: a new tool for rehabilitation. In: Eisenberg MG, Grzesiak RC, eds. Advances in Clinical Rehabilitation. New York, NY: Springer Publishing Co, Inc; 1987;1: Wade DT. Measurement in Neurological Rehabilitation. Oxford, England: Oxford University Press; 1992:76,84, Lawton MP, Brody EM. Assessment of older people: selfmaintaining and instrumental activities of daily living. Gerontologist. 1969;9: Lawton MP. Assessing the competence of older people. In: Kent DP, et al, eds. Research Planning and Action for the Elderly: The Power and Potential of Social Science. New York, NY: Behavioral Publications; 1972: