JBJS | Measuring Function of the Shoulder. A Cross-Sectional Comparison of Five Questionnaires*

The Journal of Bone & Joint Surgery, Volume 78, Issue 6

Articles | June 01, 1996

Measuring Function of the Shoulder. A Cross-Sectional Comparison of Five Questionnaires*

DORCAS E. BEATON, B.SC., O.T., M.SC.†; ROBIN R. RICHARDS, M.D., F.R.C.S.(C)†, TORONTO, ONTARIO, CANADA

Investigation performed at the Upper Extremity Reconstructive Service, St. Michael's Hospital and the University of Toronto, Toronto

The Journal of Bone & Joint Surgery. 1996; 78:882-90

5 Recommendations (Recommend) | 3 Comments | Saved by 3 Users Save Case

text A A A

Abstract

Measures of both generic and disease-specific health status are being developed and used with increasing frequency for the appraisal of musculoskeletal conditions. The purpose of this study was to compare prospectively the validity of five questionnaires in the assessment of function of the shoulder. Ninety subjects who had various problems related to the shoulder agreed to enter the study. All of the subjects completed a questionnaire package that included the Shoulder Pain and Disability Index, the Simple Shoulder Test, the Subjective Shoulder Rating Scale, the Modified American Shoulder and Elbow Surgeons Shoulder Patient Self-Evaluation Form, and the Shoulder Severity Index as well as a measure of generic health status (the acute version of the Short Form 36 [SF-36]) and two questions that asked the patient to rate the severity of the problem and his or her over-all health. Frequency distributions were created and compared among questionnaires. Spearman rank correlations were calculated to compare the questionnaires with each other and with other assessments. One-way analysis of variance was used to determine the ability of the questionnaires to discriminate between self-rated severity of the problem and over-all health.The frequency distributions were similar among the five shoulder questionnaires, but those of the five shoulder questionnaires differed from that of the SF-36. The correlations were good (0.73 = r = 0.80) among all of the five shoulder questionnaires except the Subjective Shoulder Rating Scale; they were lower with the Subjective Shoulder Rating Scale and the physical function dimension of the SF-36 (0.12 = r = 0.60). The shoulder questionnaires discriminated between levels of severity (p < 0.0001) but not between levels of over-all health (0.10 = p = 0.86).In this concurrent comparison of measures of shoulder-specific outcome in the same subjects, the shoulder questionnaires performed similarly, both in describing function of the shoulder and in discriminating between levels of severity. The shoulder questionnaires performed differently than the SF-36, which confirms the need to use both disease-specific and generic health-status measures to evaluate patients who have a problem related to the shoulder.

Figures in this Article

Introduction

Introduction | Materials and Methods | Results | Discussion | References

Outcomes research has encouraged the measurement of all of the possible effects of a disease or intervention^11,20. As a result, generic and disease-specific health-status measures are being used with increasing frequency for the appraisal of musculoskeletal conditions. This interest in comprehensive outcome measurement has also made it necessary to assess the measurement properties of different instruments to be certain that the outcome is being appraised accurately^{6,8,17,21,22,25}. Practicality, previous use in the same population, and evidence of the properties needed for the intended purpose are requirements for a valid outcome measure^{9,10,13,25,32,33}. Instruments may have varying strengths depending on the population (for example, very sick compared with relatively well patients) or the reason for their use (for example, to evaluate change over time or to describe a sample at one point in time). The selection of an instrument for an outcome measure must be context-specific and should be based on evidence that the instrument has the necessary measurement properties in the population being sampled for study or assessment^{8,15,18,24,25,33}. Describing a sample at one point in time is useful for the evaluation of gradients of health in a population or the assessment of relationships between different factors and the severity of the disease^10,18. The properties required for this discriminative role (describing a sample at one point in time) include a varied distribution of responses, correlations with other measures that try to capture a similar concept (construct validity), and the ability to discriminate between different levels of health or disability¹⁸.

Function of the shoulder has traditionally been assessed with measures—such as range of motion, strength, and pain—that reflect the local impact of a disorder. More recently, attention has been focused on the impact of a disorder on the ability to function in daily life, as evidenced by the move toward measures of health-related quality of life, both generic and disease-specific, often with use of questionnaires completed by the patient. Measurement properties (validity and reliability) have been assessed for outcome measures such as range of motion^12,29 or strength^5,30,34, and comparisons are often made among assessment techniques^3,29,31. Unfortunately, less is known about these same features with regard to measures of different patient-oriented outcomes such as the health-related quality-of-life questionnaires. As the use of such questionnaires increases, direct comparisons between instruments for the measurement of health-related quality of life are needed^8,17,21,25.

Most investigators support the use of disease or domain-specific measures of health-related quality of life together with generic measures^4,8,24,25. The former include items relevant and sensitive to the disorder being studied, and the latter allow for comparisons between conditions and may be sensitive to unexpected effects of a disorder^4,8. Previous research has shown that, compared with other generic instruments, the acute version of the Short Form 36 (SF-36) is a reliable and valid generic measure³⁵ of the health of patients who have a musculoskeletal condition^1,2,17. The purposes of the present study were, first, to compare the ability of five disease-specific measures of function of the shoulder to describe a sample of subjects who had disorders related to the shoulder (construct validity) and, second, to compare the responses to these disease-specific instruments with those to a generic health-status measure (the acute version of the SF-36). The underlying hypothesis was that the different shoulder questionnaires would perform in a similar manner (be highly correlated with each other and similarly discriminate across levels of severity) as they are conceptually similar in their goals and design. It was also hypothesized that the shoulder questionnaires would perform differently than the generic measure of health (the SF-36). Confirmation of the latter hypothesis would support the conceptual model, which suggests the need for both generic and disease-specific measures of health in outcomes research. Comparison of the shoulder questionnaires and the SF-36 also allowed examination of how well patients who have a disorder related to the shoulder will be described by the SF-36 in comparisons across conditions.

*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. Funds were received in total or partial support of the research or clinical study presented in this article. The funding source was a grant from the Health Sciences Research Centre at St. Michael's Hospital, Toronto, Ontario.

†St. Michael's Orthopaedic Associates, 55 Queen Street East, Suite 800, Toronto, Ontario M5C 1R6, Canada. Please address requests for reprints to Dr. Richards.

View Larger

TABLE I SELF-RATED SEVERITY OF PROBLEM RELATED TO THE SHOULDER AND OVER-ALL HEALTH

	No. of Subjects
Question	(N = 90)
Severity of problem
Very severe	21 (23%)
Moderate	39 (43%)
Mild	22 (24%)
Not severe	5 (6%)
No problem	1 (1%)
Not recorded	2 (2%)
Over-all health
Poor	2 (2%)
Fair	14 (16%)
Good	33 (37%)
Very good	31 (34%)
Excellent	9 (10%)
Not recorded	1 (1%)

Add to My JBJS

TABLE I SELF-RATED SEVERITY OF PROBLEM RELATED TO THE SHOULDER AND OVER-ALL HEALTH

	No. of Subjects
Question	(N = 90)
Severity of problem
Very severe	21 (23%)
Moderate	39 (43%)
Mild	22 (24%)
Not severe	5 (6%)
No problem	1 (1%)
Not recorded	2 (2%)
Over-all health
Poor	2 (2%)
Fair	14 (16%)
Good	33 (37%)
Very good	31 (34%)
Excellent	9 (10%)
Not recorded	1 (1%)

View Larger

TABLE II CONTENT OF INSTRUMENTS USED TO MEASURE FUNCTION OF THE SHOULDER


Instrument	Dimension/No. of Items	Scaling of Responses	Time to Complete	Ease of Scoring	No. of Usable Questionnaires (N = 90)
			(Mins.)
Shoulder Pain and Disability Index²⁸	Pain/5, function/8	10-cm visual-analog scales	3—5	Moderate	87 (97%)

Simple Shoulder Test²³	Pain/2, function/7, motion/3	Yes-or-no responses; some hierarchical scaling	<3	Easy	90 (100%)

Modified American Shoulder and Elbow Surgeons Form²⁷	Pain/2, function/13	4-point scale for each arm; visual-analog scale for pain and function; pain-drawing	3—5	Easy	88 (98%)

Subjective Shoulder Rating Scale¹⁹	Pain/1, motion/1, stability/1, activity/2	Multiple-choice answers; weighted by response	<3	Easy	87 (97%)

Shoulder Severity Index²⁶	Pain/7, function/20, strength/1, handicap/1, satisfaction/1	Variety of scales, all self-completed; weighted by response	7	Complex	79 (88%)

Acute version of Short Form 36 (SF-36)³⁵	Physical function, social function, emotional role function, physical role function, mental health, energy, pain, general health perceptions (over-all, unweighted mean across dimensions)	Variety of scales, used as a generic measure of health status	7—10	Moderate	87—89 (97—99%)

Add to My JBJS

TABLE II CONTENT OF INSTRUMENTS USED TO MEASURE FUNCTION OF THE SHOULDER


Instrument	Dimension/No. of Items	Scaling of Responses	Time to Complete	Ease of Scoring	No. of Usable Questionnaires (N = 90)
			(Mins.)
Shoulder Pain and Disability Index²⁸	Pain/5, function/8	10-cm visual-analog scales	3—5	Moderate	87 (97%)

Simple Shoulder Test²³	Pain/2, function/7, motion/3	Yes-or-no responses; some hierarchical scaling	<3	Easy	90 (100%)

Modified American Shoulder and Elbow Surgeons Form²⁷	Pain/2, function/13	4-point scale for each arm; visual-analog scale for pain and function; pain-drawing	3—5	Easy	88 (98%)

Subjective Shoulder Rating Scale¹⁹	Pain/1, motion/1, stability/1, activity/2	Multiple-choice answers; weighted by response	<3	Easy	87 (97%)

Shoulder Severity Index²⁶	Pain/7, function/20, strength/1, handicap/1, satisfaction/1	Variety of scales, all self-completed; weighted by response	7	Complex	79 (88%)

Acute version of Short Form 36 (SF-36)³⁵	Physical function, social function, emotional role function, physical role function, mental health, energy, pain, general health perceptions (over-all, unweighted mean across dimensions)	Variety of scales, used as a generic measure of health status	7—10	Moderate	87—89 (97—99%)

View Larger

TABLE IV RESULTS OF ANALYSIS OF VARIANCE ACROSS LEVELS OF SELF-RATED OVER-ALL HEALTH AND SEVERITY OF PROBLEM RELATED TO THE SHOULDER

*The F test statistic from one-way random-effects analysis of variance and the associated p value for each instrument. Significant results (p = 0.05) are in bold.

	Over-All Health*		Severity of Problem Related to Shoulder*
Instrument	F Test Statistic	P Value	F Test Statistic	P Value
Acute version of Short Form 36 (SF-36)³⁵
Physical function	4.94	0.001	3.20	0.02
Pain	1.64	0.17	15.79	0.0001
Mean across dimensions	7.02	0.0001	7.57	0.0001
Shoulder Pain and Disability Index²⁸	1.56	0.20	19.69	0.0001
Shoulder Severity Index²⁶	1.21	0.32	20.95	0.0001
Subjective Shoulder Rating Scale¹⁹	0.40	0.86	8.73	0.0001
Simple Shoulder Test²³	1.73	0.15	10.31	0.0001
Modified American Shoulder and Elbow Surgeons Form²⁷	1.60	0.10	12.11	0.0001

Add to My JBJS

TABLE IV RESULTS OF ANALYSIS OF VARIANCE ACROSS LEVELS OF SELF-RATED OVER-ALL HEALTH AND SEVERITY OF PROBLEM RELATED TO THE SHOULDER

*The F test statistic from one-way random-effects analysis of variance and the associated p value for each instrument. Significant results (p = 0.05) are in bold.

	Over-All Health*		Severity of Problem Related to Shoulder*
Instrument	F Test Statistic	P Value	F Test Statistic	P Value
Acute version of Short Form 36 (SF-36)³⁵
Physical function	4.94	0.001	3.20	0.02
Pain	1.64	0.17	15.79	0.0001
Mean across dimensions	7.02	0.0001	7.57	0.0001
Shoulder Pain and Disability Index²⁸	1.56	0.20	19.69	0.0001
Shoulder Severity Index²⁶	1.21	0.32	20.95	0.0001
Subjective Shoulder Rating Scale¹⁹	0.40	0.86	8.73	0.0001
Simple Shoulder Test²³	1.73	0.15	10.31	0.0001
Modified American Shoulder and Elbow Surgeons Form²⁷	1.60	0.10	12.11	0.0001

View Large | Download Slide(.ppt)
Add to My JBJS

Graph comparing the distributions of responses for the five shoulder questionnaires completed by the same sample of subjects who were seen for pain or dysfunction of the shoulder. The bars represent the percentage of the sample in each score range. SPADI = Shoulder Pain and Disability Index, SSRS = Subjective Shoulder Rating Scale, M-ASES = Modified American Shoulder and Elbow Surgeons Patient Self-Evaluation Form, SSI = Shoulder Severity Index, and SST = Simple Shoulder Test.

View Large | Download Slide(.ppt)
Add to My JBJS

Graph comparing the responses obtained from a disease-specific health-status instrument (Shoulder Pain and Disability Index [SPADI]) and a generic health-status instrument (the physical function dimension of the acute version of Short Form 36 [SF-36]). The SF-36 depicts a healthier state than the disease-specific instrument. Both questionnaires were administered at the same time to the same sample of subjects who had pain in the shoulder. The bars represent the percentage of the sample in each score range.

View Large | Download Slide(.ppt)
Add to My JBJS

Graph of the mean scores on the acute version of Short Form 36 (SF-36) for normal subjects³⁵ and the mean scores for the subjects in the present study, who were seen for problems related to the shoulder. Pain, physical role function, and emotional role function appear to be sensitive to the impact of the problem related to the shoulder. Physical function is not as low as expected for this problem. The mean score reflects an unweighted mean across dimensions, a proxy for an aggregate score. The developers of the SF-36 do not recommend aggregation of the scores.

View Large | Download Slide(.ppt)
Add to My JBJS

TABLE III

Materials and Methods

Introduction | Materials and Methods | Results | Discussion | References

Sample

Ninety subjects with a variety of problems related to the shoulder who had been referred to an upper-extremity clinic agreed to participate in the study. The inclusion criteria were a disorder related to the shoulder, the ability to read and write in English (in order to complete the questionnaires), and a willingness to participate in the study. The protocol was approved by the hospital ethics board, and all subjects signed an informed-consent form.

The mean age of the subjects (and standard deviation) was 48 ± 15.5 years (range, eighteen to seventy-seven years; median, forty-eight years). There were fifty men and forty women. The diagnoses included impingement syndrome in fifty-five subjects (61 per cent), osteoarthrosis in seventeen (19 per cent), instability in five (6 per cent), humeral malunion in four (4 per cent), and various other abnormalities in nine (10 per cent). The mean duration of the symptoms before the assessment was 50 ± 57.6 months (range, two to 420 months; median, thirty months). Sixty subjects (67 per cent) rated the severity of the problem related to the shoulder as moderate or very severe. However, most of these subjects (sixty-four; 71 per cent) rated their over-all health as good or very good (Table I).

All assessments were done by the study therapist, independent of the attending surgeon. The subjects agreed to complete a questionnaire package that included five questionnaires related to function of the shoulder (the Shoulder Pain and Disability Index, the Simple Shoulder Test, the Subjective Shoulder Rating Scale, the Modified American Shoulder and Elbow Surgeons Shoulder Patient Self-Evaluation Form, and the Shoulder Severity Index), the acute version of the SF-36, and questions related to self-rated severity of the disability of the shoulder and over-all health. The study therapist then assessed range of motion. The subjects were allowed to return the package by mail and were provided with a stamped, addressed envelope. The subjects were reminded by telephone if they did not return the package.

Questionnaire Package

A review of the literature revealed a group of instruments that could be considered disease-specific measures of the functional status of patients who have a problem related to the shoulder^{14,19,23,26-28} (Table II). As far as we know, none of these instruments have been previously tested in a concurrent comparison with other questionnaires. Few have been formally tested for reliability and validity^23,28. All of the instruments include questions regarding activities of daily living and pain. The Subjective Shoulder Rating Scale¹⁹ also includes questions related to stability and motion. In addition, the Shoulder Severity Index²⁶ includes questions regarding strength, handicap, and satisfaction with the treatment. The subjects were encouraged by the study therapist to write in the booklet any comments regarding the content of the questionnaires.

Management and Analysis of Data

All data were entered and checked for data-entry errors with use of Epi Info (version 5.01b)⁷. Analyses were done on SAS software (version 6.08, Statistical Analysis System; SAS Institute, Cary, North Carolina) for personal computers. The scores were transformed, if necessary, such that the most positive state of health on each instrument was equal to 100 points. Although methods are being developed for the aggregation of scores³⁵ on the SF-36, they were not available for this analysis and the unweighted mean across dimensions was used as a proxy for an aggregate score. This is not recommended by the developers of the SF-36, but it has been used previously¹⁷. Frequency distributions of the responses to the five shoulder questionnaires and to the SF-36 were constructed with use of 10-point intervals for the distribution ranges.

Construct validity was assessed in three ways. First, Spearman rank correlations were calculated between various measures of function of the shoulder, including other questionnaires, measures of over-all health, and active range of motion (the maximum angle between the arm and the trunk as viewed from any direction and as measured with a goniometer). High correlations suggest good convergent validity of the instruments. Second, the ability of the instruments to discriminate between self-rated levels of disability of the shoulder was assessed with analysis of variance (one-way random effects). Self-rated severity of the problem related to the shoulder was selected as a reasonable comparison to test the construct validity of a shoulder-function questionnaire. If a problem is rated by the patient as being more severe, the patient is probably having more difficulty with daily activities. A strong instrument is expected to differentiate between at least two of these levels, with higher scores for less severe problems related to the shoulder. Third, the shoulder questionnaires were compared with generic measures of health: the acute version of the SF-36 and self-rated over-all health. This comparison was performed to examine the need for both a generic and a disease-specific measure of health status. This final assessment of construct validity was analyzed with Spearman rank correlation between the SF-36 and the shoulder questionnaires as well as with analysis of variance to determine the ability to discriminate between different levels of over-all health (F-test significance across different levels).

Results

Introduction | Materials and Methods | Results | Discussion | References

All ninety subjects completed the questionnaire package. After correction for missing values or duplicate responses, we were able to use most of the questionnaires; the rate of correct completion was 97 to 100 per cent for all of the instruments except for the Shoulder Severity Index, for which the rate was 88 per cent (Table II).

Subjects' Comments Regarding Content of Questionnaires

The subjects were encouraged to comment on any difficulties that they had encountered while completing the questionnaires. Several subjects had difficulty responding to the question on the Shoulder Severity Index regarding satisfaction with the treatment because they had not yet had treatment or had received multiple treatments for the shoulder. Several subjects could not respond to items on the Shoulder Severity Index regarding driving, particularly with regard to changing gears, because they did not drive. Other comments included ways that the subject had adapted in order to avoid difficulty performing an activity. One example is "I have to stand up" when "pouring from a jug," with the subject rating minimum difficulty with the task. Other examples include using a long-handled brush to comb the hair or asking for specifications regarding the height of the "high shelf" or the weight of the item that is being reached on the high shelf. The subjects also indicated that the difficulty encountered may depend on the time of day or on other demands on the shoulder.

Distribution of Responses

The scores for the Subjective Shoulder Rating Scale tended to be higher (mode, 61 to 70 points), possibly reflecting the more impairment-oriented nature of its questions (for example, instability or motion) (Fig. 1). The other questionnaires displayed similar distributions with relatively normal distributions (the modes had a central tendency between 31 and 40 points or 41 and 50 points, and the distributions were not skewed). The distributions for the physical function dimension of the acute version of the SF-36 differed from those for the shoulder questionnaires (Fig. 2). The distributions for the SF-36 reflect a sample that appears to be less disabled than indicated by the shoulder questionnaires. The mean score for physical function was 60 points on the SF-36, compared with 42 points on the Shoulder Pain and Disability Index. The mean results for the sample on each of the dimensions of the SF-36 were compared with the normative values (unadjusted for age or gender)³⁵ (Fig. 3). Physical role function, emotional role function, and pain were farthest from the normative values, followed by physical function. Mental health, general health, and energy were close to the normative values.

Construct Validity

The shoulder instruments correlated only moderately well with each other (0.47 = r = 0.80), although this improved significantly when the Subjective Shoulder Rating Scale was excluded (0.73 = r = 0.80) (Table III). All correlations between instruments were significant (p = 0.05) except for that between the Subjective Shoulder Rating Scale and the physical function dimension of the SF-36 (p > 0.05). Correlations of the shoulder instruments with either physical function or the over-all score on the SF-36 were lower than those between the shoulder questionnaires (excluding the Subjective Shoulder Rating Scale) (0.58 = r = 0.72, compared with 0.73 = r = 0.80) (Table III). Correlations between active elevation of the shoulder and the questionnaires were low (r = 0.45 for the Modified American Shoulder and Elbow Surgeons Form) or not significant (p > 0.05) (all other instruments).

Analysis of variance confirmed the ability of the shoulder instruments to discriminate between levels of self-rated severity of the problem related to the shoulder (the F tests were significant [p = 0.0001]) (Table IV). The acute version of the SF-36 also had significant F tests across the severity rating (range, 3.20 to 15.79; 0.0001 = p = 0.02), although the gradient across these levels was less than that for the shoulder instruments. Analysis of variance also demonstrated that the dimensions of the SF-36 were able to differentiate between over-all health ratings better than the shoulder instruments (the F tests for the SF-36 were significant at p = 0.05; the F tests for the shoulder questionnaires were not significant, with 0.10 = p = 0.86) (Table IV). The performance of the pain dimension of the SF-36 was similar to that of the shoulder instruments, with better discrimination across the severity rating for the shoulder than across the over-all health rating.

Discussion

Introduction | Materials and Methods | Results | Discussion | References

Outcomes research has been the focus of a great deal of attention in the literature. By definition, outcomes should reflect all possible effects of a disease or intervention¹⁹. Conceptually, this has been described as the measurement of effects at the level of specific tissues or joint restrictions (impairments—for example, range of motion or strength), the level of ability to perform daily activities (disability), and the level of role function (handicap—for example, return to work). A complete outcome battery includes measures of each level, each offering a different but important insight into the impact of the disease or the treatment on an individual. The questionnaires used in this study were for the most part disability outcomes that are employed to examine the impact of abnormalities of the shoulder on the ability to function in daily activities. Some of the questionnaires also include questions regarding impairment (pain or motion) and handicap (the ability to work). Although a great deal of effort has gone into increasing the reliability and validity of impairment measures, ongoing work in this area is still needed for the disability and handicap measures, including health-related quality-of-life questionnaires.

The present study confirmed the similarity of five disease-specific measures of the health-related quality of life for individuals with conditions related to the shoulder; the measures with a disability focus were especially similar. There were similarities in distributions as well as good correlations among the disability-oriented questionnaires, and it was possible to discriminate between different levels of self-rated disability of the shoulder. Although the correlations were good, some variation in the responses was found in both the correlation coefficients and the distributions. If the shoulder instruments were identical in terms of what they assessed, these distributions would have been identical because the same sample of subjects completed each of the questionnaires. Measures that were more impairment-oriented (whether measured by the observer or self-rated), such as range of motion or the Subjective Shoulder Rating Scale, did not correlate as well with measures that assessed the impact of the disorder on the capacity to perform or on the performance of daily activities (disability).

It was hypothesized that the shoulder instruments would have lower correlations with the generic measure of health (the SF-36) than with each other and would also be less powerful in their ability to discriminate between levels of over-all health (a role for a generic measure) than the SF-36. This was confirmed and supports the need for both generic and disease-specific measures of health status when assessing function of the shoulder.

The physical function dimension of the SF-36 was not as sensitive to the disability experienced by people who have problems related to the shoulder. In this study, in which patients were being seen in a tertiary-care center for the problem related to the shoulder, the scores on the physical function dimension of the SF-36 were relatively high. The same patients had lower scores (more disability) on the disease-specific measures (and therefore lower correlations between the physical function dimension of the SF-36 and the shoulder instruments). Even though the physical function dimension of the SF-36 was able to discriminate between different levels of severity of the problem related to the shoulder, the breadth of the scores for these levels was less than that for the disease-specific instruments. Only the pain dimension of the SF-36 was sensitive to the impact of the problem related to the shoulder in this sample, possibly because pain was the main symptom in the shoulder for most patients at the time of presentation. Pain is often considered to be at the level of impairment, which supports this finding that it is a unique dimension in the disability and handicap-oriented SF-36. Physical role function had a low mean value, but this may have been due to a large number of people who were retired and thus not in their work roles, possibly accounting for the dimension's lack of discrimination across disease severity. It is important to include both generic and disease-specific health-status instruments in a battery of outcome measures^4,8,24.

There is concern that the instrument designed to be used for comparison across conditions (the SF-36 in this study) was not sensitive to the disability experienced by these subjects. It is therefore possible that, in comparisons across conditions, our subjects will appear healthier or less disabled than subjects who have symptoms in other anatomical areas (such as low-back pain). Until studies are performed to evaluate what is actually being measured with these generic instruments, it will not be known if the potential lack of sensitivity to patients who have a problem related to the upper extremity is due to the content of the instrument (physical function is dominated by items related to the lower extremity) or whether these subjects truly are less disabled than those who have musculoskeletal problems in different anatomical regions.

A modified version of the American Shoulder and Elbow Surgeons Patient Self-Evaluation Form was included in this study²⁷. The modifications were made so that the instrument would reflect dysfunction of the entire extremity rather than just that of the shoulder. The supposition was that the upper extremity functions as a kinematic chain. According to this paradigm, the shoulder (as well as the elbow, forearm, and wrist) positions the hand for grasp and manipulation of the environment¹⁶. In the present study, the Modified American Shoulder and Elbow Surgeons Patient Self-Evaluation Form performed as well as the other shoulder instruments did. Therefore, there may be a role for a questionnaire that deals with the entire upper extremity (domain-specific) rather than each joint or condition within the upper extremity. For example, a condition involving the shoulder, elbow, wrist, or hand affects, to a greater or lesser extent, the ability to use a telephone. A domain-specific instrument could be used when the diagnosis is less certain or when more than one part of the extremity is affected.

The measurement of outcomes has become increasingly important. This study demonstrated the measurement properties of five disease-specific measures of function of the shoulder in the same sample of subjects. The focus was on how the instruments performed in a discriminative role: measurement of disability at one point in time. Four of the five questionnaires performed well when compared with each other and appeared to be sensitive to the disability experienced by these subjects. The choice of which questionnaire to use to assess patients at one point in time might therefore be based on practical considerations, such as the ease of scoring or the applicability across a variety of conditions affecting the upper extremity. These instruments allowed assessment of a different aspect of the impact of a problem related to the shoulder than either range of motion or that measured with the acute version of the SF-36, a generic measure of health. Outcome assessments require measures of impairment as well as both generic and disease-specific measures of health-related quality of life in order for the full impact of a problem related to the shoulder on the patient to be understood.

Generic measures may not be as sensitive to the disability experienced by patients who have problems related to the shoulder. Additional research is needed to investigate whether this is due to the instrument (for example, if there is less content dealing with the upper extremity) or if dysfunction of the shoulder is less disabling than other musculoskeletal conditions, such as low-back pain. Additional longitudinal research is required to understand how sensitive these same instruments are to clinical change, as is needed for instruments that are to be used to measure change in function over time, such as in clinical trials.

References

Introduction | Materials and Methods | Results | Discussion | References

Beaton, D. E.; Bombardier, C.; and |and |Hogg-Johnson, S.: Health status measurement: can your tool colour your results. A comparison of five different measures of health status used in the same sample [abstract]. Arthrit. and Rheumat.,36: 193, 1993.36193 1993

Beaton, D. E.; Bombardier, C.; and |and |Hogg-Johnson, S.: Reliability and responsiveness of five generic health-related quality of life instruments [abstract]. Arthrit. and Rheumat.,36: 193, 1993.36193 1993

Beaton, D. E.; O'Driscoll, S. W.; and |and |Richards, R. R.: Grip strength testing using the BTE work simulator and the Jamar dynamometer: a comparative study. Baltimore therapeutic equipment. J. Hand Surg.,20A: 293-298, 1995.20A293 1995

Bergner, M., and |and |Rothman, M. L.: Health status measures: an overview and guide for selection. Ann. Rev. Pub. Health,8: 191-210, 1987.8191 1987 [CrossRef]

Bohannon, R. W.: Test-retest reliability of hand-held dynamometry during a single session of strength assessment. Phys. Ther.,66: 206-209, 1986.66206 1986 [PubMed]

Bombardier, C., and |and |Tugwell, P.: Methodological considerations in functional assessment. J. Rheumat.,14 (Supplement 15): 6-10, 1987.14 (Supplement 15)6 1987

Dean, A. G.; Barton, J. A.; Barton, A. H.; and Dicer, R. C.: Epi Info, Version 5. A Word Processing, Database, and Statistics Program for Epidemiology on Microcomputers. Stone Mountain, Georgia, USD, 1990.

Deyo, R. A.; Andersson, G.; Bombardier, C.; Cherkin, D. C.; Keller, R. B.; Lee, C. K.; Liang, M. H.; Lipscomb, B.; Shekelle, P.; Spratt, K. F.; and |and |Weinstein, J. N.: Outcome measures for studying patients with low back pain. Spine,19(18S): 2032S-2036S, 1994.19(18S)2032 1994 [CrossRef]

Feinstein, A. R.: Clinical biostatistics. XLI. Hard science, soft data, and the challenges of choosing clinical variables in research. Clin. Pharm. and Ther.,22: 485-498, 1977.22485 1977

Feinstein, A. R.: Clinimetrics. New York, Yale University Press, 1987.

Fletcher, R. H.; Fletcher, S. W.; and Wagner, E. H.: Clinical Epidemiology: the Essentials. Ed. 2. Baltimore, Williams and Wilkins, 1988.

Gajdosik, R. L., and |and |Bohannon, R. W.: Clinical measurement of range of motion. Review of goniometry emphasizing reliability and validity. Phys. Ther.,67: 1867-1872, 1987.671867 1987 [PubMed]

Gartland, J. J.: Orthopaedic clinical research. Deficiencies in experimental design and determinations of outcome. J. Bone and Joint Surg.,70-A: 1357-1364, Oct. 1988.70-A1357 1988

Gerber, C.: Integrated scoring systems for the functional assessment of the shoulder. In The Shoulder: a Balance of Mobility and Stability, pp. 531-550. Edited by F. A. Matsen, III, F. H. Fu, and R. J. Hawkins. Rosemont, Illinois, The American Academy of Orthopaedic Surgeons, 1993.

Guyatt, G. H.; Bombardier, C.; and |and |Tugwell, P. X.: Measuring disease-specific quality of life in clinical trials. Canadian Med. Assn. J.,134: 889-895, 1986.134889 1986

Kapandji, I. A.: The Physiology of the Joints, translated by L. H. Honore. Ed. 5, vol. 1, The Upper Limb. New York, Churchill Livingstone, 1982.

Katz, J. N.; Larson, M. G.; Phillips, C. B.; Fossel, A. H.; and |and |Liang, M. H.: Comparative measurement sensitivity of short and longer health status instruments. Med. Care,30: 917-925, 1992.30917 1992 [PubMed][CrossRef]

Kirshner, B., and |and |Guyatt, G.: A methodological framework for assessing health indices. J. Chronic Dis.,38: 27-36, 1985.3827 1985 [PubMed][CrossRef]

Kohn, D.; Geyer, M.; and Wülker, N.: The Subjective Shoulder Rating Scale (SSRS)—an examiner-independent scoring system. Read at the International Congress on Surgery of the Shoulder, Paris, July 12-15, 1992.

Last, J. M.: A Dictionary of Epidemiology. Ed. 2, p. 92. New York, Oxford University Press, 1988.

Liang, M. H.; Fossel, A. H.; and |and |Larson, M. G.: Comparisons of five health status instruments for orthopedic evaluation. Med. Care,28: 632-642, 1990.28632 1990 [PubMed][CrossRef]

Liang, M. H.; Larson, M. G.; Cullen, K. E.; and |and |Schwartz, J. A.: Comparative measurement efficiency and sensitivity of five health status instruments for arthritis research. Arthrit. and Rheumat.,28: 542-547, 1985.28542 1985 [CrossRef]

Lippitt, S. B.; Harryman, D. T., II; and Matsen, F. A., III: A practical tool for evaluation of function: the Simple Shoulder Test. In The Shoulder: a Balance of Mobility and Stability, pp. 501-518. Edited by F. A. Matsen, III, F. H. Fu, and R. J. Hawkins. Rosemont, Illinois, The American Academy of Orthopaedic Surgeons, 1993.

Patrick, D. L., and |and |Deyo, R. A.: Generic and disease-specific measures in assessing health status and quality of life. Med. Care,27 (Supplement 3): 217-S232, 1989.27 (Supplement 3)217 1989 [CrossRef]

Patrick, D. L., and Erickson, P.: Health Status and Health Policy. Quality of Life in Health Care Evaluation and Resource Allocation. New York, Oxford University Press, 1993.

Patte, D.: Directions for the Use of the Index Severity for Painful and/or Chronically Disabled Shoulders, pp. 36-41. Paris, The First Open Congress of the European Society of Surgery of the Shoulder and Elbow, 1987.

Richards, R. R.; An, K.-N.; Bigliani, L. U.; Friedman, R. J.; Gartsman, G. M.; Gristina, A. G.; Iannotti, J. P.; Mow, V. C.; Sidles, J. A.; and |and |Zuckerman, J. D.: A standardized method for the assessment of shoulder function. J. Shoulder and Elbow Surg.,3: 347-352, 1994.3347 1994 [CrossRef]

Roach, K. E.; Budiman-Mak, E.; Songsiridej, N.; and |and |Lertratanakul, Y.: Development of a shoulder pain and disability index. Arthrit. Care and Res.,4: 143-149, 1991.4143 1991 [CrossRef]

Rothstein, J. M.; Miller, P. J.; and |and |Roettiger, R. F.: Goniometric reliability in a clinical setting. Elbow and knee measurements. Phys. Ther.,63: 1611-1615, 1983.631611 1983 [PubMed]

Sapega, A. A.: Current concepts review. Muscle performance evaluation in orthopaedic practice. J. Bone and Joint Surg.,72-A: 1562-1574, Dec. 1990.72-A1562 1990

Stratford, P. W.: Summarizing the results of multiple strength trials: truth or consequence. Physiother. Canada,44: 14-18, 1992.4414 1992

Streiner, D. L., and Norman, G. R.: Health Measurement Scales: a Practical Guide to Their Development and Use. New York, Oxford University Press, 1989.

Tugwell, P., and |and |Bombardier, C.: A methodologic framework for developing and selecting endpoints in clinical trials. J. Rheumat.,9: 758-762, 1982.9758 1982

Wadsworth, C. T.; Krishnan, R.; Sear, M.; Harrold, J.; and |and |Nielsen, D. H.: Intrarater reliability of manual muscle testing and hand-held dynametric muscle testing. Phys. Ther.,67: 1342-1347, 1987.671342 1987 [PubMed]

Ware, J. E., Jr.; Snow, K. K.; Kosinski, M.; and Gandek, B.: SF-36 Health Survey: Manual and Interpretation Guide. Boston, The Health Institute, 1993.

Topics

View Large | Download Slide(.ppt)
Add to My JBJS

TABLE III

View Large | Download Slide(.ppt)
Add to My JBJS

View Larger

TABLE I SELF-RATED SEVERITY OF PROBLEM RELATED TO THE SHOULDER AND OVER-ALL HEALTH

	No. of Subjects
Question	(N = 90)
Severity of problem
Very severe	21 (23%)
Moderate	39 (43%)
Mild	22 (24%)
Not severe	5 (6%)
No problem	1 (1%)
Not recorded	2 (2%)
Over-all health
Poor	2 (2%)
Fair	14 (16%)
Good	33 (37%)
Very good	31 (34%)
Excellent	9 (10%)
Not recorded	1 (1%)

Add to My JBJS

TABLE I SELF-RATED SEVERITY OF PROBLEM RELATED TO THE SHOULDER AND OVER-ALL HEALTH

	No. of Subjects
Question	(N = 90)
Severity of problem
Very severe	21 (23%)
Moderate	39 (43%)
Mild	22 (24%)
Not severe	5 (6%)
No problem	1 (1%)
Not recorded	2 (2%)
Over-all health
Poor	2 (2%)
Fair	14 (16%)
Good	33 (37%)
Very good	31 (34%)
Excellent	9 (10%)
Not recorded	1 (1%)

View Larger

TABLE II CONTENT OF INSTRUMENTS USED TO MEASURE FUNCTION OF THE SHOULDER


Instrument	Dimension/No. of Items	Scaling of Responses	Time to Complete	Ease of Scoring	No. of Usable Questionnaires (N = 90)
			(Mins.)
Shoulder Pain and Disability Index²⁸	Pain/5, function/8	10-cm visual-analog scales	3—5	Moderate	87 (97%)

Simple Shoulder Test²³	Pain/2, function/7, motion/3	Yes-or-no responses; some hierarchical scaling	<3	Easy	90 (100%)

Modified American Shoulder and Elbow Surgeons Form²⁷	Pain/2, function/13	4-point scale for each arm; visual-analog scale for pain and function; pain-drawing	3—5	Easy	88 (98%)

Subjective Shoulder Rating Scale¹⁹	Pain/1, motion/1, stability/1, activity/2	Multiple-choice answers; weighted by response	<3	Easy	87 (97%)

Shoulder Severity Index²⁶	Pain/7, function/20, strength/1, handicap/1, satisfaction/1	Variety of scales, all self-completed; weighted by response	7	Complex	79 (88%)

Acute version of Short Form 36 (SF-36)³⁵	Physical function, social function, emotional role function, physical role function, mental health, energy, pain, general health perceptions (over-all, unweighted mean across dimensions)	Variety of scales, used as a generic measure of health status	7—10	Moderate	87—89 (97—99%)

Add to My JBJS

TABLE II CONTENT OF INSTRUMENTS USED TO MEASURE FUNCTION OF THE SHOULDER


Instrument	Dimension/No. of Items	Scaling of Responses	Time to Complete	Ease of Scoring	No. of Usable Questionnaires (N = 90)
			(Mins.)
Shoulder Pain and Disability Index²⁸	Pain/5, function/8	10-cm visual-analog scales	3—5	Moderate	87 (97%)

Simple Shoulder Test²³	Pain/2, function/7, motion/3	Yes-or-no responses; some hierarchical scaling	<3	Easy	90 (100%)

Modified American Shoulder and Elbow Surgeons Form²⁷	Pain/2, function/13	4-point scale for each arm; visual-analog scale for pain and function; pain-drawing	3—5	Easy	88 (98%)

Subjective Shoulder Rating Scale¹⁹	Pain/1, motion/1, stability/1, activity/2	Multiple-choice answers; weighted by response	<3	Easy	87 (97%)

Shoulder Severity Index²⁶	Pain/7, function/20, strength/1, handicap/1, satisfaction/1	Variety of scales, all self-completed; weighted by response	7	Complex	79 (88%)

Acute version of Short Form 36 (SF-36)³⁵	Physical function, social function, emotional role function, physical role function, mental health, energy, pain, general health perceptions (over-all, unweighted mean across dimensions)	Variety of scales, used as a generic measure of health status	7—10	Moderate	87—89 (97—99%)

View Larger

TABLE IV RESULTS OF ANALYSIS OF VARIANCE ACROSS LEVELS OF SELF-RATED OVER-ALL HEALTH AND SEVERITY OF PROBLEM RELATED TO THE SHOULDER

*The F test statistic from one-way random-effects analysis of variance and the associated p value for each instrument. Significant results (p = 0.05) are in bold.

	Over-All Health*		Severity of Problem Related to Shoulder*
Instrument	F Test Statistic	P Value	F Test Statistic	P Value
Acute version of Short Form 36 (SF-36)³⁵
Physical function	4.94	0.001	3.20	0.02
Pain	1.64	0.17	15.79	0.0001
Mean across dimensions	7.02	0.0001	7.57	0.0001
Shoulder Pain and Disability Index²⁸	1.56	0.20	19.69	0.0001
Shoulder Severity Index²⁶	1.21	0.32	20.95	0.0001
Subjective Shoulder Rating Scale¹⁹	0.40	0.86	8.73	0.0001
Simple Shoulder Test²³	1.73	0.15	10.31	0.0001
Modified American Shoulder and Elbow Surgeons Form²⁷	1.60	0.10	12.11	0.0001

Add to My JBJS

TABLE IV RESULTS OF ANALYSIS OF VARIANCE ACROSS LEVELS OF SELF-RATED OVER-ALL HEALTH AND SEVERITY OF PROBLEM RELATED TO THE SHOULDER

*The F test statistic from one-way random-effects analysis of variance and the associated p value for each instrument. Significant results (p = 0.05) are in bold.

	Over-All Health*		Severity of Problem Related to Shoulder*
Instrument	F Test Statistic	P Value	F Test Statistic	P Value
Acute version of Short Form 36 (SF-36)³⁵
Physical function	4.94	0.001	3.20	0.02
Pain	1.64	0.17	15.79	0.0001
Mean across dimensions	7.02	0.0001	7.57	0.0001
Shoulder Pain and Disability Index²⁸	1.56	0.20	19.69	0.0001
Shoulder Severity Index²⁶	1.21	0.32	20.95	0.0001
Subjective Shoulder Rating Scale¹⁹	0.40	0.86	8.73	0.0001
Simple Shoulder Test²³	1.73	0.15	10.31	0.0001
Modified American Shoulder and Elbow Surgeons Form²⁷	1.60	0.10	12.11	0.0001