Abstract
Background: Because there is no muscle disease specific measure of quality of life (QoL), we wanted to develop and validate an individualized muscle disease specific measure of QoL for adults suitable for both clinical and research use.
Methods: A literature review exploring QoL and its measurement resulted in the development of a theoretical model of QoL. This was used alongside qualitative interviews (n = 41) and a postal survey (n = 252) to design a questionnaire. The psychometric properties, validity (n = 95), reliability (n = 40), and responsiveness (n = 25) of the scale were assessed.
Results: The Individualized Neuromuscular Quality of Life questionnaire (INQoL) consists of 45 questions within 10 sections. Four of these focus on the impact of key muscle disease symptoms (weakness, locking [i.e., myotonia], pain, and fatigue), five look at the impact (degree and importance of impact) muscle disease has on particular areas of life, and one section asks about the positive and negative effects of treatment. The questionnaire is structured to allow for variations in the individual characteristics that influence quality of life. Psychometric evaluation established construct validity and test–retest reliability. A preliminary assessment of responsiveness was obtained.
Conclusions: The Individualized Neuromuscular Quality of Life is a validated muscle disease specific measure of quality of life developed from the experiences of patients with muscle disease and can be used for individuals or large samples.
Generic health-related quality of life (QoL) questionnaires have been used in a number of muscle diseases.1–5 However, these generic scales do not capture all the issues that are relevant to muscle disease and may incorporate issues that are superfluous to patients' experiences of muscle disease. Many scales also do not allow patients to indicate how much value an activity and its impact have on their own quality of life, especially because this may change as patients adapt to their condition.6,7 We therefore wanted to develop the Individualized Neuromuscular Quality of Life questionnaire (INQoL), which was specific to muscle disease and which allowed for individual variation.
The content of INQoL was derived from semistructured interviews with muscle disease patients. The items identified from these interviews were verified using a postal survey in a larger population. The structure of INQoL was based on the ICIDH-2 model of disease incorporating the concepts of Impairment, Activities, and Participation. We separated scores relating to symptom severity and their impact from the QoL domains so that it would be possible to attributed change on QoL either to an actual change in symptoms or to a change in the perceived QoL.6 The fact that QoL can be altered in either of these two ways explains the disability paradox whereby QoL seems to contradict more objective measures of health.7–9 For example, unexpectedly high levels of QoL have been reported in studies of chronically ill and disabled patients.10–12 If QoL is conceptualized as the discrepancy between the patients' ideal and actual states, this gap can be narrowed and QoL can be improved, either by bottom-up change in disease severity, or by top-down change or “response shift” caused by patients' reevaluation of their goals and their importance.13
Methods.
Subjects.
Eligible patients were defined as those having weakness or stiffness, i.e., myotonia, due to muscle disease, as confirmed by a combination of expert opinion, genetic testing, raised creatine kinase levels, neurophysiology, or muscle biopsy, of more than 6 months in duration. Subjects had to be older than 16 years and were excluded if they had major active comorbidities unrelated to muscle disease such as arthritis, respiratory disease (other than neuromuscular respiratory weakness), cardiovascular disease (other than muscle disease related cardiomyopathy), cognitive impairment that prevented comprehension of the questionnaires, or the inability to read English. Subjects were recruited from the muscle clinics of King's College Hospital, The National Hospitals for Neurology, and the Atkinson Morley's Hospital, London, and from the UK Myositis Support Group. The study received ethics approval from the institutions involved.
Interviews.
We conducted in-depth semistructured interviews to explore patients' perceptions of the effects of neuromuscular disease (NMD) on their lives (the interview schedule is given in appendix E-1 on the Neurology Web site at www.neurology.org). Three subgroups were chosen as being representative of the range of natural histories seen in these diseases. These were congenital progressive NMD such as Becker muscular dystrophy and the congenital myopathies, acquired relapsing–remitting NMD such as polymyositis or dermatomyositis, and late-onset progressive NMD such as inclusion body myositis, some limb-girdle muscular dystrophies, or dystrophic myotonias. The interviews were tape recorded and transcribed. A phenomenologic approach was used to guide the analysis and facilitate the exploration of patients' perceptions and experiences.20,21 Themes were extracted and clustered together into categories representing life domains influenced by NMD. A coding scheme was devised to represent the individual domain and subdomain categories and applied to the data. Finally, the validity of the coding scheme was verified through an external inspection by a second categorizer who applied the scheme to a sample of interviews.
Postal survey.
A postal survey was used to confirm that the life domains that emerged from the interviews were representative and to uncover any issues that may have been missed. The postal survey also obtained patients' rating of the degree to which the life domains were influenced by their muscle disease. The postal questionnaire was resent to nonresponders 2 weeks later, and a reminder letter was sent to remaining nonresponders after a further 2 weeks.
INQoL design.
Wording and layout was determined by reference to questionnaire design guidelines. INQoL was then piloted for appropriateness of layout, time frame for completion, response scaling, wording, and content validity.
Psychometric evaluation.
Reliability.
To assess test–retest reliability, participants completed INQoL twice, 1 week apart. A 1-week period was considered to be short enough to minimize the likelihood that change would occur in any of the dimensions and long enough to reduce the chance that patients would recollect their previous responses. Participants either returned to the clinic 1 week later to complete INQoL a second time or else elected to complete the questionnaire at home and mail the completed INQoL. Participants were given a choice of venue for completion of their second questionnaire because of practical considerations and the physical constraints of their condition (many traveled long distances to attend the clinic or were in discomfort).
Validity.
Construct validity was established by getting participants to complete several validated questionnaires and functional assessments. These were compared with responses to appropriate matched elements of INQoL. Table 1 summarizes the comparisons made and the scales and measures used. To reduce the burden of completing large numbers of questionnaires, patients were allocated to receive one of two different subsets of questionnaires designated as List A or List B (table 1). The order in which patients completed the questionnaires was randomized to diminish the influence of previous questionnaires on subsequent answers.
Table 1 Questionnaires and scales used in the validation study
Responsiveness.
We conducted a preliminary study of the responsiveness of INQoL with assessments at two time points: baseline and between 3 and 6 months later. At both times, they completed the INQoL, Short Form-36 (SF-36), Patient Generated Index (PGI), Arthritis Body Experience Scale, Barthel Index, Chalder Fatigue Questionnaire (CFQ), visual analog symptom scales, and the functional tests. On the second visit, patients also completed a five-point subjective change question that asked how much change if any they had experienced in their condition since the first appointment. Twenty-two subjects returned to the clinic for their second testing, but three elected to complete the questionnaire at home and mail it back.
Analysis.
One-week test–retest reliability was assessed using Bland and Altman's method for measuring agreement.23 In this method, limits of agreement are calculated to assess the size of differences between the first and second administrations of the questionnaire. Narrow limits of agreement indicate good test–retest reliability. The way that INQoL is constructed with one question for each dimension meant that it was neither possible nor appropriate to examine internal consistency within each of the subscales and philosophically inappropriate to evaluate it across the whole questionnaire because each domain is a distinct entity.
A priori hypotheses for the relationships between validated measures and INQoL were tested using Spearman Rho correlations. This ensured a robust validation process,14 diminishing the likelihood of an inflated type I error rate arising from multiple comparisons.
The responsiveness of the questionnaire to change over time was ascertained using effect sizes, with 0.8 or more reflecting a large change, 0.5 to 0.8 reflecting a moderate change, and 0.2 to 0.5 reflecting a small change. The effect sizes of all the scales used in the study were calculated to determine whether changes in scores could be related across the various measures and particularly to changes in INQoL.
Data were analyzed using SPSS version 10.0.
Results.
Interviews.
We interviewed 41 patients (15 men and 26 women) ranging from age 20 to 80 years. These included 21 patients with a congenital, slowly progressive NMD; 10 patients with an acquired, relapsing–remitting NMD; and 10 patients with an acquired, slowly progressive NMD. A number of broad life domains (table 2) emerged from the analysis of the qualitative interviews. Most patients reported an impact of NMD in each of these life domains. Domains associated with negative feelings included physical symptoms, degree of control over physical functioning, participation, relationships, negative social reactions, and perceived consequences of the condition. Positive effects were recorded, e.g., with improved family relationships resulting from the NMD.
Table 2 Life domains reported in interviews with NMD patients
Postal survey.
We received 252 completed questionnaires from the 537 sent out giving a response rate of 47%. Of the responders, 183 were women. Their ages ranged from 16 to 96 years (mean age 52.6 years, SD 16). The mean duration of their muscle disease was 11.8 years, and muscle diagnoses included congenital myopathies, limb-girdle muscular dystrophies, facioscapulohumeral muscular dystrophy, dystrophic and nondystrophic myotonias, and inflammatory myopathies. More than 60% of all respondents reported an impact on the life areas included in the questionnaire. Indeed, in all of the domains except Relationships, more than 80% of the respondents reported being at least “slightly affected.” In those receiving treatment, most respondents reported experiencing both beneficial and negative effects, which were rated as being “extremely important.” More than 90% of all those who reported an impact in the life domains also reported this impact to be important. Working life and Independence were given the highest importance ratings by more than 90% of the respondents. The Perception of the future and Partner relationship items also received very high ratings, with more than 80% selecting the categories of “very” and “extremely important.” Respondents' comments were concordant with, and clarified their responses to, the postal survey questions. New themes that emerged included difficulties in meeting potential partners and specific coping strategies used by respondents.
Piloting of INQoL.
The questionnaire was piloted, modified, and repiloted, three times using a total of 25 patients. Among the issues that emerged and were addressed were that Likert scales were preferred to graphical rating scales and that a seven-point Likert was preferred to a five-point Likert scale. The time frame of the questions was also changed from “over the last two weeks” to “at the moment.” It was believed that this would overcome difficulties in recall of experiences over a longer space of time but would avoid the problems associated with asking patients how they feel “today.” This is because the days on which patients attend to fill in a questionnaire are unlikely to be a typical day (i.e., hospital visit) and responses may also be more influenced by mood. The final revised questionnaire was assessed for reliability, validity, and responsiveness.
Psychometric evaluation.
Ninety-five people completed the questionnaire on at least one occasion during the validation studies. The male:female ratio of the entire sample was 1:2.2, and the mean age was 55 years (SD 17.08, min–max 19 to 97). See table 3 for characteristics of patients participating in each of the validation studies.
Table 3 Sample characteristics of patients participating in each study
Reliability.
The test–retest reliability of the scale is demonstrated by the mean difference between INQoL domain scores at first and second administration (table E-1). All but 2 of the 10 subscales demonstrated a mean change of less than two points, demonstrating very good stability. The other two subscales, Muscle weakness and Body image, had greater mean differences in score over the test period, but this still represented a small change over the whole score and an acceptable level of agreement/stability.
Construct validity.
Thirty-two patients completed the “timed walk” test, and 29 completed the “timed stands” task. The Muscle weakness scores demonstrated a strong correlation with the functional tests (table E-2). There were no significant relationships between other symptom scores (Pain, Fatigue, or Locking) and these two functional tests. Patients with more functional disability did seem to report NMD as having a greater impact upon their activities and their independence as evidenced by a strong correlation between both the Activities and the Independence scores of INQoL and the scores on these functional tests (table E-2). There was concordance between the visual analog scales for pain, weakness, and muscle “locking” and the corresponding subscales of INQoL. The relationship between INQoL Fatigue score and CFQ scores was weaker than the relationships demonstrated between the other symptoms and their reference measures (table E-2). High Activities impact scores on INQoL showed a reasonably strong relationship (Spearman Rho = −0.59; p > 0.001) with low Physical Functioning scores on the SF-36, suggesting that patients who report difficulties with mobility and self-care tasks will report more difficulty in performing physical activities and greater dissatisfaction with their ability to perform their activities. There was a strong relationship (Spearman Rho = 0.67; p > 0.001) between Independence scores of INQoL and scores on the Barthel Index, supporting the hypothesis that patients reporting difficulties with mobility and self-care tasks will report lower levels of independence and greater dissatisfaction with their degree of independence. Exploration of the Social relationships dimension of INQoL showed a modest but significant correlation between the SSQ6 and the Social relationships dimension of INQoL. The social impact domains of the SF-36, the Functional Limitations Profile (FLP), and the Nottingham Health Profile (NHP) demonstrated much stronger relationships. The Emotions subscale score of INQoL demonstrated a strong relationship with the Anxiety and Depression dimensions of the Hospital Anxiety and Depression scale and with the corresponding scales of the SF-36, the FLP, and the NHP (table E-2). Patients with a more negative body image reported more of an impact of NMD on satisfaction with their physical appearance (Spearman Rho −0.68, p > 0.001). As predicted, the QoL scores derived from FLP (Spearman Rho 0.76, p = 0.000) and from PGI (Spearman Rho −0.57, p > 0.001) demonstrated strong concordance with the overall INQoL score. There was also a considerable overlap between the domains generated by patients for PGI and those domains incorporated in INQoL.
Responsiveness to change.
Little change occurred on any of the INQoL dimensions over the follow-up period. The Pain (mean change 6.32, effect size 0.21) and Emotions dimensions (mean change −6.89, effect size −0.24) showed a small effect size over the 3- to 6-month period, but changes in the remaining scales were very small, with negligible effect sizes. The mean changes and effect sizes for the other questionnaires measured over the 3- to 6-month period are shown in table E-3. Patients who completed the questionnaire twice in a 3- to 6-month interval were divided into three groups based on whether they reported an improvement (n = 4), a deterioration (n = 7), or no change (n = 14) in their condition since their first administration of the questionnaires. Subgroup analysis of effect sizes in these groups of patients demonstrated changes in a number of dimensions in the expected direction (table E-4).
Discussion.
The qualitative analysis revealed that the experience of muscle disease influences many areas of life. There were similarities across patients, but no set pattern of disease impact. Factors including demographic factors, life circumstances, disposition, and type of muscle disease influenced individual patients' experience and perception of NMD effects. The life domains that emerged correspond to the broad domains believed to make up the QoL spectrum (physical, psychological, and social functioning).15,16 The issues associated with negative feelings have been documented in studies of patients with illness and disability.17,18). However, the life domains emerging from these interviews go beyond those captured by generic measures that have been used in studies of muscle disease patients. These included physical symptoms specific to muscle disease, relationships, perceptions of the future, employment, and self-perception.1–5,19–21 These findings affirmed the need to develop a new QoL measure, specific to patients with NMD.
Postal survey responses and comments supported and verified the exploratory interview data, with respondents reporting the influence of NMD across a broad spectrum of life areas. The postal survey data supported the diversity of the impact experienced by patients with NMD and showed that the effects of NMD are important to patients. These effects, already implied in the interviews, were explicitly represented in quantitative data from the survey. The results of the postal survey were used to select items on the basis of their relevance and importance to patients, thereby maximizing the likelihood that the measure would accurately capture QoL. This method was chosen because it was considered more appropriate to include domains of importance to patients rather than domains derived from a statistical model that may or may not accurately reflect the concerns of patients with NMD. To ensure that INQoL is representative, items were selected that covered a wide range of issues relevant to respondents at both the severe and mild ends of the spectrum. All of the chosen items also received high impact and importance ratings in the mail survey. One domain, Perception of the future, was left out because it was considered to be particularly dependent on personality and emotional state rather than being specifically related to muscle disease.
INQoL consists of 45 questions within 10 sections. Four of these focus on the impact of key muscle disease symptoms (Weakness, Locking (i.e., myotonia), Pain, and Fatigue), five look at the impact (degree and importance of impact) of NMD on particular areas of life, and one section asks about the effects of treatment. Participants respond using a seven-point Likert scale. Appendix E-2 gives an item example. Responses to items looking at symptoms and the impact of NMD on life domains are weighted by the perceived importance of the item, assigned by the participant. The final score is presented as a percentage of the maximum detrimental impact. Therefore, a higher percentage indicates greater impact. Participants' perception of their treatment is represented by two scores, reflecting the trade-off between the positive and negative effects of treatment now and in the future. Scores for each section can be presented as a profile. In addition, a composite score representing overall QoL, based on scores from the five sections assessing the impact of NMD on particular areas of life, can be calculated.
The definition of QoL as the discrepancy between experience and the patients' expected or ideal state was incorporated in questions by asking patients to rate their current state in relation to their ideal and worst state they can imagine across the life domains.22,23 To acknowledge the impact of individual circumstances on the experience of NMD, INQoL asks patients to rate the difficulties caused by specific symptoms and then the importance of these difficulties in terms of overall satisfaction with each lifestyle domain. To reflect the dynamic nature of QoL, INQoL was designed to separate out stages of NMD impact.24 For example, the effects of symptoms are separated from questions about life domains, thus allowing one to determine whether changes in symptoms relate to changes in life domain scores. Responses to the life domain questions are maintained as separate scores to enable particular problems to be identified and monitored over time. This separation also allows “shifts” in patients' internal standards to be identified if satisfaction with life domains has altered independently from a change in perceived symptoms.
Construct validation showed that there was a strong correlation between Muscle weakness and the functional tests supporting the validity of this symptom score given the considerable impact that weakness has on physical tasks. The lack of relationship between the Locking and Fatigue scores and the same timed tasks is a reflection of the lesser importance that each of these symptoms play in these particular short-term physical activities. The strength of the relationship between Symptom severity as measured by INQoL and established measures demonstrates the validity of INQoL. Indeed, the relationship may have been stronger if the established instruments measured symptom impact, as in INQoL. The weaker relationship between Fatigue and the CFQ may have been due to a difference in time frame. The CFQ asks about fatigue in the past 6 weeks compared with “at the moment” as measured by INQoL. In general, a difference in emphasis between INQoL and the established measures may have attenuated the strength of the relationships. For example, the correlation between the Physical Functioning scale (SF-36) and the Activities dimension of INQoL may have been moderated by the focus of the SF-36 on activities of daily living and not specifically asking about work or leisure activities as in INQoL. In addition, the small correlation between the social relationship subscale of INQoL and the SSQ6 may also be explained by the questionnaires measuring different constructs. INQoL measures the impact of NMD on relationships, whereas the SSQ6 measures the number of people available to provide social support in various circumstances. However, the moderate correlation coefficients demonstrated between the social support scores and QoL as measured by INQoL supports the hypothesis that that social support plays a role in maintaining QoL.
The composite QoL score of INQoL is generated by combining the scores that represent patients' satisfaction with each of the five dimensions with the importance attached to NMD impact on these dimensions. The PGI measures the impact of disease on the five most important life areas affected by the patient's condition, “all other health-related aspects of life,” and “all non-health aspects of life.” The significant relationship between these measures supports the validity of the INQoL composite score. The considerable overlap between the dimension of INQoL and domains generated by patients on the PGI further supports the content validity of the scale. The FLP covers a broad range of physical and psychological areas of life important to QoL in individual patients with physically disabling conditions. The strong relationship between the FLP and the QoL score on INQoL further supports the validity of the new scale. The associations of INQoL dimensions with scores on related measures supported the hypothesized underlying constructs of the scale. Correlation coefficients ranged between 0.29 and 0.78, and many were greater than 0.5, indicating that the reference measures are related to the subscales of INQoL but measure a different entity. Correlations of this magnitude were to be expected considering the differences between INQoL and established scales in their scaling methods, underlying constructs and their development for use in different circumstances. It might be thought that even these associations between INQoL and the FLP, the PGI, and many domains of the SF-36 suggest that these existing QoL instruments adequately measure QoL for NMD. However, the FLP is long, with a complicated scoring system, and a large number of FLP items are not relevant to a considerable number of individual patients, dampening respondents' motivation to respond accurately and to complete the assessment. The PGI involves the self-generation of individually relevant items, which can be difficult. The generic nature of the SF-36 means that many items, e.g., “have you been a happy person,” are not specifically relevant to the impact of NMD and could be influenced by many other factors.
INQoL therefore provides a more relevant and practical measure of QoL in NMD than these major generic measures of QoL. It allows respondents to rate the impact of specific symptoms, the influence of NMD on areas of their life, incorporates ratings of individual importance, and can be completed without the assistance of an interviewer. Although the construct validity of INQoL has been demonstrated, the validation process should continue as the questionnaire goes on to be used in different areas of research. INQoL has been validated for UK adults with a variety of muscle diseases that included congenital myopathies, limb-girdle muscular dystrophies, facioscapulohumeral muscular dystrophy, dystrophic and nondystrophic myotonias, and inflammatory myopathies. Although these are heterogeneous conditions, they share the property of physical disability usually due to progressive muscle weakness. Future studies can further explore the validity of INQoL using larger more homogenous diagnostic groups. Specific studies would be needed to extend the validity to other countries and to other neuromuscular diseases not included in our patient population, such as myasthenias and periodic paralysis.
Test–retest reliability was demonstrated in the INQoL dimensions. Pain, Fatigue, and Muscle weakness scores demonstrated slightly lower levels of agreement, but this may have been due to a real change in these symptoms between the two time points, particularly because they are likely to vary considerably over short periods of time. We chose a 1- week test–retest period as being a compromise between a period short enough to minimize actual change in condition while at the same time reducing carryover of previously recalled responses. The variance in scores from initial test to retest is likely to be due in part to the small sample size. Practical considerations allowed completion of INQoL at the second time point to be done at home, where the precise timing and conditions were less controlled and, consequently, variations may have reduced the consistency.
Evaluation of the responsiveness of INQoL must be regarded as being of a preliminary nature because the sample size was small and we were reliant on the ability of natural history changes in muscle disease in a 3- to 6-month period to demonstrate changes in INQoL domains. For most chronic muscle disease patients, little clinical change would be expected in such a short time scale. Not surprisingly, most of the changes reflected by the scales used in this study were small, and even in the parameters that demonstrated a substantial change, the sample size was too small for the effect size statistics to be conclusive. When the changes in the group as a whole were analyzed, the Pain and Emotions scales demonstrated small effect sizes. When patients were divided into groups representing their perception of whether they had improved, deteriorated, or stayed the same, changes in the expected direction indicated INQoL's potential to demonstrate responsiveness. The incorporation of INQoL into intervention studies with larger numbers of subjects should allow a better appreciation of its responsiveness.
A description of INQoL and the full INQoL questionnaire has been deposited with the MAPI Research Institute Web site, and inquiries regarding its use can be obtained from www.mapi-trust.org.
Acknowledgment
The authors thank Professor M. Hanna and Dr. R. McKeron for access to their patients, and the Myositis Support Group UK for their help with the mail survey.
Footnotes
Additional material related to this article can be found on the Neurology Web site. Go to www.neurology.org and scroll down the Table of Contents for the March 27 issue to find the title link for this article.
Funded by the King's College Hospital Joint Research Fund.
Disclosure: The authors report no conflicts of interest.
Received March 28, 2006. Accepted in final form November 27, 2006.
References
Letters: Rapid online correspondence
REQUIREMENTS
You must ensure that your Disclosures have been updated within the previous six months. Please go to our Submission Site to add or update your Disclosure information.
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
You May Also be Interested in
Hastening the Diagnosis of Amyotrophic Lateral Sclerosis
Dr. Brian Callaghan and Dr. Kellen Quigg