predictors of hand function in patients with rheumatoid arthritis

Predictors of Hand Function in Patients with Rheumatoid Arthritis

Berit Dellhag and Carol S. Burckhardt

Objectives. The aims of this study were to determine: (IJ what factors predict patient self-estimated hand function; [2J what factors predict actual hand function; and (3) the relationship among actual hand function, patient estimates of hand function, and self-assessed activities of daily living [ADLJ. Methods. Fifty-two patients with rheumatoid arthritis completed wrist and hand mobility measures, grip strength, pain, stiffness, and estimated hand function tests, along with the Sollerman Grip Function Test [actual hand function), Health Assessment Ques- tionnaire (HAQ, and subscales of the Arthritis Impact Measurement Scales [AIMS]. Results. Grip strength and stiffness were the strongest predictors of self-estimated hand function. Flexion and extension deficits in digits II through V were the strongest predictors of actual hand function. Actual hand function and self-estimated hand function were significantly correlated with each other and with the HAQ and AIMS subscales. Conclusions. Measures of deficit are the most useful in predicting actual hand function, whereas measures of strength and flexibility are most useful for estimated hand function.

Key Words: Hand function; Activities of daily living; Rheumatoid arthritis; Outcome.

Berit Dellhag, OT, is at the Department of Occupational Therapy, Sahlgren University Hospital, Gothenburg, Sweden; and Carol S. Burckhardt, RN, PhD, is Professor, Department of Mental Health Nursing, and Assistant Professor of Medicine (Research), Division of Arthritis and Rheumatic Diseases, School of Medicine, Oregon Health Sciences University, Portland, Oregon.

Address correspondence to Berit Dellhag, OT, Department of Oc- cupational Therapy, Sahlgren University Hospital, 41345 Gothen- berg, Sweden.

Submitted for publication May 7, 1993; accepted June 7, 1994. 0 1995 by the American College of Rheumatology.

Over 90% of rheumatoid arthritis (RA] patients are believed to have some involvement of their hand joints [l]. By far the most commonly involved joints in RA are the metacarpophalangeal (MCP) joints, proximal interphalangeal (PIP] joints, and wrists. Decreased PIP and MCP flexion or extension are thought to make a major contribution to functional disability because grasp and fine manipulation are compromised [Z].

Patients often complain of difficulty in performing everyday tasks. These problems with activities of daily living (ADL) may be the result of hand pain, deformity, and loss of range of motion, as well as decreased mus- cle strength or fatigue in the proximal upper extremities [3]. Decreased grip strength and impaired hand function are well documented in the literature and seem to be almost universal among RA patients [4-81. The impact of decreased grip strength on the RA patient has also been documented. Jonsson and Larsson, for example, found a highly significant correlation between grip strength and total locomotion scores [9], whereas Pincus et al. noted that decreased grip strength was among the predictors of increased mortality in RA patients [lo].

We have noticed in our clinical work that patients often underestimate their hand functioning in ADLs when compared to actual testing of those functions, whereas Mathiesen et al. [11] have reported the op- posite finding. About 23% of their sample regarded their hand function as unaffected and yet they were unable to perform a hand function test developed by Recht et al. [12]. Thus, a patient's estimate of hand function may not be the best predictor of actual hand function or actual ADL ability.

Because one of the goals of occupational therapy with RA patients is to maintain or increase the strength, joint motion, and endurance of the upper extremities in order to enhance functional independence, it would be useful for clinicians to better understand the re-

16 0893-7524/95/$5.00

Arthritis Care and Research Hand Function in Rheumatoid Arthritis 17

TABLE 1

Means, Standard Deviations, and Ranges of All Variables in the Study

Variable Mean SD Range

ROM deficit (flex) digits 11-V (mm] 54.7 45.8 ROM deficit [ext) digits 11-V [mm) 32.1 50.5 Deficit opposition digit I [degrees) 8.4 13.2 Extension (abd] digit I [degrees) 26.8 8.1 Dorsal flexion (ext] wrist (degrees] 46.1 14.7 Volar flexion [flex) wrist [degrees] 45.1 18.3 Grip strength [maximum] (Newtons] 101.3 80.9 Grip strength (average) [Newtons) 79.9 67.5 Pain on resisted motion (0-9) 2.6 2.3 Pain on nonresisted motion (0-100) 26.1 24.0 Stiffness (0-100) 28.2 25.1 Acutal hand function (0-801 74.5 6.0

Estimated hand function [0-100) 63.9 21.9 Health Assessment Questionnaire (0-3) 1.1 0.6 AIMS Dexterity (0-10) 4.7 3.1 AIMS ADL (0-10) 0.8 1.3

(Sollerman Grip Function Test)

AIMS Household Activities (0-10) 1.2 1.0

0-156 0-232 0-45

10-46 0-70 0-80 8-344 1-300 0-8.2 0-93 0-90

40-80

19-100 0-2.4 0-10 0-6.3 0-3.1

SD, standard deviation; ROM. range of motion; flex, flexion; ext, extension; abd. abduction; AIMS, Arthritis Impact Measurement Scales; ADL, Activities of Daily Living.

lationships among actual hand function, patients’ per- ceptions of their hand function, and ADL activities when planning treatment. The aims of this study were to determine: (1) what factors predict self-estimated hand function, (2) what factors predict actual hand function, and (3) whether actual hand function and patients’ estimates of hand function are related to self- assessed ADL function.

PATIENTS AND METHODS

Patients Fifty-two patients (33 women and 19 men) with a

diagnosis of seropositive RA participated in the study. Their mean age was 53 years (range, 29-69) and the mean duration of disease was 7.7 years (range 6-10). They were all in American Rheumatism Association (ARA, now the American College of Rheumatology, ACR) functional classes I and I1 [13]. The study sample was described more completely in an earlier com- munication [14].

Measurements The measures consisted of the following:

1. Range of motion flexion and extension deficits of digits I1 through V and opposition deficit of digit I was measured in the dominant hand and expressed in millimeters (mm) [15]. Flexion of digits I through

IV was defined as the distance between the distal point of the digits and the distal palmar crease. Extension in digits I through IV was measured with the hand resting supinated on a table with the fore- arm outside the table. The distance between the nail-bed of the extended fingers and the table was recorded. Opposition of digit I was measured as the distance between the distal point of the thumb top and the base of digit V.

2. Range of motion in abduction/extension of digit I was measured as the angle between the metacarpal bones of the thumb and digit I1 parallel to the plane. Volar and dorsal flexion of the wrist in the dominant hand were measured in degrees using a go- niometer [15].

3. Self-estimated hand function was measured with a 100-mm visual analogue scale. Before performing any grip strength or hand function tests, patients were asked to put a mark on a line somewhere between the endpoints of “no hand function” and “full hand function.” Hand function was defined as the ability to use the hand to perform some activity. Test-retest reliability for a 1-week interval was 0.91 in a group of 28 patients with RA (Dellhag, unpublished data).

4. Grip strength in the dominant hand was measured with the electronic Grippit instrument. This instrument measures the actual force produced by the hand when squeezing a handle with maximum in- tensity during 10 seconds. Maximum power and average power are both recorded in Newtons. Norms for healthy women and women with RA as well as a complete description of this new instrument have been published [16,17].

5. Actual hand function was measured by the Soll- erman grip function test [18]. This is a criterion- referenced test that defines grip function as the ability of the hands to perform a given task with a permitted grip within a set period of time. It was developed specifically to measure hand movements needed in ADL while minimizing elbow-shoulder movement. The device consists of a board that re- quires the performance of 20 hand tasks, 16 one- handed and four two-handed, such as unscrewing a lid, putting a key in a lock and turning, and writing with a pencil. The subject’s performance is graded from 4 (best) to 0 (worst) using as criteria the com- bination of actual performance of the activity and the use of a correct grip. Each of the 20 hand tasks has a list of permitted grips. The test has high inter- rater reliability and reproducibility (r = 0.98). Work by Wilton [19] reported a high correlation between the Sollerman test and the norm-referenced Jebsen Taylor Hand Function Test [ Z O ] .

18 Dellhag and Burckhardt

TABLE 2

Correlations among the Potential Predictor Variables, Actual Hand Function, and Estimated Hand Function

Vol. 8, No. 1, March 1995

TABLE 3

Variable

Actual Estimated hand hand

function function

ROM deficit (flex) digits 11-V ROM deficit [ext] digits 11-V Deficit opposition digit I Extension [abd] digit I Dorsal flexion [ext) wrist Volar flexion [flex) wrist Grip strength (maximum] Grip strength [mean) Pain on resisted motion Pain on nonresisted motion Stiffness

-0.51' -0.54" -0.38b

0.31" 0.21 0.09 0.40b 0.40b

-0.43b -0.45" -0.34"

-0.23 -0.38b -0.3gb

0.38b 0.20 0.09 0.59" 0.58"

-0.41b -0.43b -0.45"

Results of the Regression Analyses: Relationship of Five Predictors to Actual Hand Function and Estimated Hand Function

Actual hand Estimated hand function function

Predictors F P F P

1.93 0.17 ROM deficit [flex) digits 11-V 3.91 0.05 ROM deficit [ext] digits 11-V 9.85 0.003 2.76 0.10 Deficit opposition digit I 0.95 0.33 2.72 0.11 Grip strength [maximum) 0.52 0.47 14.12 0.0005 Stiffness 0.75 0.39 3.96 0.05

Full model RZ 0.46 0.49

Numbers in boldface type indicate significant predictors. ROM, range of motion; flex, flexion; ext, extension.

" P < 0.05. b P < 0.01. = P < 0.001. ROM, range of motion; flex. flexion; ext, extension; abd, abduction.

6. Pain on resisted motion during performance of the grip function test was measured on a 10-point scale

7. Pain on nonresisted motion and stiffness was measured on a separate 100-mm visual analogue scale.

8. A self-administered Swedish version of the Health Assessment Questionnaire (HAQ) disability index [22,23] and three subscales of the Swedish version of the Arthritis Impact Measurement Scales (AIMS] measured ADL ability [24]. The results reported here are from measurements made at the beginning of a study of hand treatments for RA patients [14].

Statistical Analysis Means and standard deviations were calculated for

all variables in the study. Bivariate correlations between all the variables were determined. Based on the outcome of the correlational analysis, predictor variables were selected for inclusion in two multiple regression analyses where the predictors were re- gressed simultaneously first on actual hand function and then on self-estimated hand function.

RESULTS

Means and standard deviations of the variables are shown in Table 1. A summary of the Pearson correlation coefficients between the predictor variables of flexion/extension, grip strength, pain, and stiffness and the outcome variables of actual hand function and estimated hand function are shown in Table 2. All variables with the exception of the volar and dorsal

wrist flexion tests were significantly correlated with the outcome measurements. Five variables were selected for inclusion in the multiple regression analyses. These variables were chosen not only because of their significant zero-order correlations with the outcome variables but also because of their independence from each other. No predictor variable correlated higher than 0.50 with any other predictor.

The results of the regression analyses are summa- rized in Table 3. The five variables explained nearly 50% of the variance in estimated hand function. Grip strength and stiffness were the strongest predictors. The same five variables explained 46% of the variance in actual hand function. Flexion and extension deficits in digits I1 through V were the strongest predictors of actual function.

The relationships among actual hand function, estimated hand function, and self-assessed ADL are shown in Table 4. All correlations except one were significant at P < 0.05 and ranged in strength from

TABLE 4

Pearson Correlation CoeBcients between Actual Hand Function, Estimated Hand Function, and Activities of Daily Living Variables

Actual hand Estimated hand function function ADL

HAQ -0.50" -0.43b AIMS Dexterity -0.33" -0.49" AIMS ADL -0.57= -0.24 AIMS Household -0.33" -0.3gb

P < 0.05. P < 0.01.

= P < 0.001. ADL, activities of daily living; HAQ, Health Assessment Questionnaire;

AIMS, Arthritis Impact Measurement Scales.

Arthritis Care and Research Hand Function in Rheumatoid Arthritis 19

-0.24 between AIMS ADL and estimated hand function to -0.57 for AIMS ADL and actual hand function. The correlation between actual hand function and estimated hand function was r = 0.54.

DISCUSSION

This study showed that almost one-half of the variance in actual hand function and estimated hand function could be predicted. In each analysis, two variables were shown to be the strongest predictors-flexion and extension of the fingers for actual hand function, and grip strength and stiffness for estimated hand function.

As has been pointed out earlier [2], actual impairment in flexion and extension of the PIP and MCP joints are the most predictive of grasping ability and fine manipulation. However, unlike the findings of Ryu et al. [25] that showed adequate wrist mobility (40 degrees each of flexion and extension] to be necessary for ADL activities, wrist mobility had no correlation with grip function in our study. Most of our subjects had adequate range of motion. So this lack of correlation may have been largely a matter of restriction in range of the wrist motion variables rather than a true lack of correlation.

Whereas actual hand function was best predicted by impairment measures, estimated hand function was best predicted by stiffness, which is entirely self-estimated, and grip strength, which for its reliability is partially dependent on subject motivation to perform maximally [17,26]. Given this finding, it is important for clinicians to understand that estimated hand function is partially a personal perception contingent on the way in which the patient sees himself/herself at a particular time. McPhee [27] has pointed out that the ability to use the hands effectively in everyday tasks depends not only on anatomical integrity but also mental status and motivation.

Most of the patients in this study had little difficulty performing each task in the Sollerman grip function test [18] with a permitted grip and within the 20-second time limit for each task. Thus, their scores on the Soll- erman test were high. Yet self-estimates of hand function were widely dispersed and only moderately correlated with actual hand function. Perhaps some patients underestimated their hand function because they believed that they should be able to do tasks more quickly and easily. They may also have been com- paring themselves to earlier abilities or other persons.

Our finding of underestimation of hand function contrasts with Mathiesen et al. [l l] , who found that about a quarter of their sample thought they had no

hand dysfunction when, according to testing, they did. The Recht hand test, used in the study by Mathiesen et al. [ll], required that the patient report no pain and have a normal hand mobility in order to get a negative score; in other words, it measures hand function using lack of impairment as the criterion. The patients in that study [ l l ] who rated their hand function as un- impaired may have been thinking in terms of ability to accomplish a task, not whether they had no pain and normal mobility. This contrast highlights the dif- ference between impairment (functional and struc- tural departures from normality) and disability (restriction or lack of functional ability and activity] as defined by the World Health Organization [28].

The Sollerman grip function test was developed to measure grips necessary for functional ADL such as eating, personal hygiene, dressing, driving a car, writing, and using a telephone. It is therefore not surprising that it correlated most highly with the HAQ and the AIMS ADL subscale, both of which contain many of these activities.

Estimated hand function correlated most highly with the AIMS dexterity and household activities subscales; both of these subscales are composed of activities di- rectly contingent upon hand function. The correlation between estimated hand function and the AIMS ADL scale was not significant. We speculate that this may have been the case because some patients used assis- tive devices; some had developed their own methods for accomplishing these activities; or the level of disease severity was not sufficient to have caused deficits requiring help, which is what the AIMS ADL is de- signed to determine.

We conclude that measures of grip strength and range of motion deficits as well as measures of ADL ability are useful in predicting a patient’s actual hand function and self-estimated hand function. It is important, however, to be clear as to how hand function is measured because measures of impairment could yield far different scores than measures of inability to perform the task. The Sollerman grip function test attempts to deal with this problem by giving patients enough time to complete a task and some leeway in what type of grip is used. As such it seems capable of giving the most true picture of actual hand function in ADL situations.

Further work using patients in ACR functional class I11 could extend the relevance of the predictors for explaining actual and estimated hand function by in- creasing the variance in the Sollerman grip function test scores, as well as that of some of the other tests, most notably wrist movement. Also, because only about one-half of the variance was explained by the predictors, it could be useful to search for other clinical

20 Dellhag and Burckhardt Vol. 8, No. 1, March 1995

indicators that would help clinicians more accurately predict their patients’ functioning.

Self-estimations of ability to function are receiving more attention in the rheumatology literature as re- searchers and clinicians recognize the role that perception plays in determining patient behavior [29,30]. Because actual hand function and estimated hand function are only moderately correlated, we suggest that attending to both may give the occupational ther- apist more comprehensive information when assisting patients to realize their potential for self-care.

REFERENCES

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

Harris ED Jr: The clinical features of rheumatoid arthritis. In Kelly WN, Harris ED Jr, S Ruddy, Sledge CB (eds): Textbook of Rheumatology, 3rd ed. Philadelphia, WB Saunders, 1989, pp 947-948. Melvin JL: Rheumatoid Disease, Occupational Therapy and Rehabilitation, 2nd ed. Philadelphia, FA Davis, 1982, p 32. Phillips C: Rehabilitation of the patient with rheumatoid hand involvement. Phys Ther 69:1091-1098, 1989. Myers DB, Grennan DM, Palmer DG: Hand grip function in patients with rheumatoid arthritis. Arch Phys Med Rehabil61:369-373, 1980. MacBain KP: Assessment of function in the rheumatoid hand. Can J Occup Ther 37:95-103,1970. An K-N, Chao EYS, Askew LJ: Hand strength measurement instruments. Arch Phys Med Rehabil 61:366-368, 1980. Spiegel JS, Paulus HE, Ward NB, et al: What are we measuring? An examination of walk time and grip strength. J Rheumatol 14:80-86, 1987. Jones E, Hanly JG, Mooney R, et al: Strength and function in the normal and rheumatoid hand. J Rheumatol

Jonsson B, Larsson S-E: Hand function and total locomotion status in rheumatoid arthritis-an epidemiologic study. Acta Orthop Scand 61:339-343, 1990. Pincus T, Callahan LF, Vaughn WK: Questionnaire, walking time and button test measures of functional ca- pacity as predictive markers for mortality in rheumatoid arthritis. J Rheumatol 14240-251, 1987. Mathiesen FK, Rasmussen JO, Recht L, Lithman T: Im- pairment of grip function in rheumatoid arthritis-stud- ies with a simple hand test. Scand J Rheumatol 20:209- 212,1991. Recht L, Rasmussen JO, Mathiesen FK, Lithman T: Con- struction of a simple test for assessment of hand function in primary care. Scand J Prim Health Care 8:225-231, 1990. Steinbrocker 0, Traeger CH, Batterman RC: Therapeu- tic criteria in rheumatoid arthritis. JAMA 140:659-662, 1949.

181313-318, 1991.

14. Dellhag B, Wollersjo I, Bjelle A: Effect of active hand exercise and wax bath treatment in rheumatoid arthritis patients. Arthritis Care Res 5:87-92, 1992.

15. American Academy of Orthopaedic Surgeons: Joint Mo- tion. Method of Measuring and Recording. Reprinted by The British Orthopaedic Association. London, En- gland, Churchill Livingston, 1966, pp 16-29.

16. Nordenskiold U: Elastic wrist orthoses. Reduction of pain and increase in grip force for women with rheumatoid arthritis. Arthritis Care Res 3:158-162, 1990.

17. Nordenskiold U, Grimby G: Grip force in patients with rheumatoid arthritis and fibromyalgia and in healthy subjects. Scand J Rheumatol 22:14-19, 1993.

18. Sollerman C: Assessment of grip function: evaluation of a new test method. In Sollerman C, Handens greppf- unktion, Analys och utvardering samt en ny testmetod [Hand grip function: analysis and evaluation and a new test method]. Thesis, University of Gothenburg, Goth- enburg, Sweden, 1980, pp 1-21.

19. Wilton JC: Hand function evaluation: a comparative study between an American and a Swedish instrument. In Focus 90, loth International Congress, World Federation of Occupational Therapists, Melbourne, Australia, 1990,

20. Jebsen RH, Taylor N, Trieschmann RB, Trotter MJ, Howard LA: An objective and standardized test of hand function. Arch Phys Med Rehabil 50:311-319, 1969.

21. Borg G: Subjective effort and physical abilities. Scand J Rehabil Med 10 (Suppl6):108-113, 1978.

22. Fries JF, Spitz P, Kraines RG, Holman HR: Measurement of patient outcome in arthritis. Arthritis Rheum 23:137- 145, 1980.

23. Ekdahl C, Eberhart K, Anderson SI, Svensson B: As- sessing disability in patients with rheumatoid arthritis: use of a Swedish version of the Stanford health assessment questionnaire. Scand J Rheumatol17:263-271,1988.

24. Archenholtz B, Bjelle A: Reliability and validity of a Swedish version of the arthritis impact measurement scales (AIMS). Scand J Rheumatol20:219, 1991.

25. Ryu J, Cooney WP 111, Askew LJ, An K-N, Chao EY: Functional ranges of motion of the wrist joint. J Hand Surg 16A:409-419, 1991.

26. Smith GA, Nelson RC, Sadoff SJ, Sadoff AM: Assessing sincerity of effort in maximal grip strength tests. Am J Phys Med Rehabil 68:73-80, 1989.

27. McPhee SD: Functional hand evaluations: A review. Am J Occup Ther 41:158-163, 1987.

28. World Health Organization: International Classification of Impairments, Disabilities and Handicaps. Geneva, Switzerland, World Health Organization, 1980.

29. Lorig K, Chastain RL, Ung E, Shoor S, Holman HR: Development and evaluation of a scale to measure per- ceived self-efficacy in people with arthritis. Arthritis Rheum 3237-44.1989.

30. Taal E, Rasker JJ, Seydel ER, Wiegman 0: Health status, adherence with health recommendations, self-efficacy and social support in patients with rheumatoid arthritis. Patient Educ Counsel 20:63-76, 1993.

pp 365-368.

predictors of hand function in patients with rheumatoid arthritis

Documents