- Email: [email protected]
HD-16: a new quality of life instrument specifically designed for insomnia
Sleep Medicine 6 (2005) 191–198 www.elsevier.com/locate/sleep
Fast-track article
HD-16: a new quality of life instrument specifically designed for insomnia D. Leger*, K. Scheuermaier, T. Raffray, A. Metlaine, D. Choudat, C. Guilleminault Centre du Sommeil, Service de Physiologie, Hoˆtel-Dieu de Paris, 1 pl. du Parvis Notre Dame, 75181 Paris cedex 04, France Received 12 November 2004; received in revised form 4 March 2005; accepted 7 March 2005
Abstract Background and purpose: To design a new quality of life (QoL) instrument specifically for insomnia. Patients and methods: Based on severe insomniacs’ interviews, we have built a new quality of life scale that has been tested in one group of 240 severe insomniacs, in one group of 422 mild insomniacs and in one group of 391 good sleepers. Ten steps led to the construction of a specific QoL scale. Results: Five dimensions have been validated as both relevant and independent from each other. Sixteen items out of the 43 initially tested were retained and significantly different within the groups in each dimension. Based on the 16 items selected, we called the scale Hotel Dieu 16 (HD-16). We have therefore verified the score’s specificity (correlation score of C0.36) and the reliability of the scale (Cronbach coefficient aZ0.78). Conclusion: HD-16 may be used as a focused instrument to better assess an insomniac’s quality of life. q 2005 Elsevier B.V. All rights reserved. Keywords: Insomnia; Quality of life; Sleep; Severe insomnia
1. Introduction Epidemiological studies have shown that insomnia is a very common disease in the general population, affecting between 9 and 15% of adults [1–3]. In addition to disturbed sleep, insomnia is defined as having a negative impact on daytime functions. In the International Classification of Sleep Disorders (ICSD) [4], ‘insomnia is always associated with a complaint of decreased functioning during wakefulness’. In the Diagnostic and Statistical manual of mental Disorders IV (DSM-IV) [5] insomnia is ‘serious enough to induce severe fatigue, irritability or disability in daytime functioning’ and is ‘frequently accompanied by non-specific symptoms such as mood disorders, memory troubles or lack of concentration’. However, previous studies which have evaluated the daytime impact of insomnia with objective parameters, such as multiple sleep latency tests [6] failed to demonstrate objective impairment of insomniacs during the day, despite * Corresponding author. Tel.: C33 142 348 243; fax: C33 142 348 227. E-mail address: [email protected] (D. Leger).
1389-9457/$ - see front matter q 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.sleep.2005.03.013
the fact that, subjectively, insomniacs complained of impaired daytime functioning. Hence, recent studies have concentrated on evaluating the impact of insomnia while other instruments have targeted more subjective dimensions including the perceptions by insomniacs of both nighttime and daytime performance. This may be achieved with the quality of life (QoL) instruments. QoL is a complex and multidimensional term that has been defined as “a concept encompassing a broad range of physical and psychological characteristics and limitations which describe an individual’s ability to function and to derive satisfaction from doing so” [7]. It includes the following domains: physical (the ability to conduct activities of daily living), psychological (or emotional) and social (interactions with family, friends and community). Insomnia may affect all three of these domains. Several studies have now used the SF-36, a generic and well-validated QoL instrument, to assess an impaired QoL in insomniacs [8–12]. Other generic scales such as the Nottingham Health Profile (NHP) and the Sickness Impact Profile (SIP) have also been used [13]. Both have shown that insomnia impaired QoL similarly with chronic disease such as hypertension or chronic obstructive pulmonary disease [13]. However, there is, until now, a lack of specific instruments assessing QoL
192
D. Leger et al. / Sleep Medicine 6 (2005) 191–198
disturbances in insomniacs after controlling for comorbid illnesses and reflecting the degree of severity of insomnia. In many other chronic diseases, it has been shown that a disease-focused QoL instrument was useful to accurately evaluate its impact on the daily lives of patients and to show the efficacy of some treatments [14–15]. We also believe that in the field of insomnia there is a need for specific instruments to better encompass the QoL of patients. The objective of the present study was to build a new QoL instrument specifically designed for insomniacs.
2. Method 2.1. Subjects and controls The SOFRES, a French poll institute, selected the subjects for study. The SOFRES regularly surveys a group of 11,372 individuals representing the French general population. Initially, each individual received a first questionnaire (Q1) assessing insomnia based on the DSMIV and ICSD criteria [4–5]. The SOFRES methodology and the questionnaire used in epidemiological surveys to select groups of mild and severe insomniacs have been previously described and published [2,9]. Severe insomniacs (SI) were categorized as having at least two of the sleep complaints defined in the DSM-IV and ICSD definitions of insomnia (i.e. difficulty initiating sleep or maintaining sleep or nonrestorative sleep) that lasts for at least one month (DSM-IV 307.42), almost nightly with an impaired social or occupational functioning (ICSD). A group of subjects with mild insomnia (MI) and a group of good sleepers (GS) were statistically matched by age, sex, occupation, location and marital status to the SI group. The MI group comprised of people who occasionally complained of sleep difficulties but did not meet the ICSD–DSM-IV criteria for insomnia. The GS group did not complain of any sleep disturbance. The statistical matching is a software process of the SOFRES database, which found groups of MI and GS that were statistically similar to the group of SI with respect to age, sex, occupation, location and marital status. Subjects were not matched by pairs, but groups were statistically not different among these variables. The study was approved by the ethics committee of local authorities of the University Hospital, and the confidentiality of personal data was respected. 2.2. Questionnaire In order to assess quality of life in subjects without depression or anxiety disorders, we subsequently sent a second questionnaire (Q2) to Q1-responders. Q2 had a twofold orientation: (a) detecting psychiatric disorders (based on the DSM-IV definitions of anxiety and depression), and (b) assessing insomnia-specific QoL. General information such as sex, age, occupation, marital and matrimonial status,
location and habitat were also recorded. The questionnaire was mailed with a postage paid return envelope. The participants received neither fee nor reward for taking part in the study. In a first part, Q2 included 10 items based on the DSM-IV minimum criteria for anxiety and depression to eliminate those patients presenting such conditions [5]. An individual was classified as ‘presenting a depression profile’, and thus eliminated from the study, when he presented two of the three following symptoms for more than 15 days: (i) being sad or depressed most of the time almost every day during the day, (ii) having lost interest in activities that normally please him or her, and (iii) being tired all the time. An individual was classified as ‘presenting an anxiety profile’ when he had one of two main anxiety symptoms (being anxious or worried most of the time, or being apprehensive towards daily problems) and was still presenting this (these) symptom(s) and had this (these) symptom(s) for more than six months. After the reception of the questionnaire, persons with psychiatric disorders according to DSM-IV minimum criteria for anxiety and depression were eliminated from each group. The three groups (SI, MI and GS) were once more compared for their socio-economic status. A mathematical adjustment was necessary to make the groups similar vis a` vis sex ratio. This was done by weighting the data with a weighting process based on the deviation between the group studied and the SI group (i.e. if the SI group included 50% of men and 50% of women and the GS group contained 25% of men and 75% of women, the system would have given a !2 weighting coefficient to the data collected from GS men and a !2/3 weighting coefficient to the data collected from the GS women). This system is called ‘the method of quotas’. After this statistical adjustment, we again controlled the three groups for the professional status of the head of the family, for the marital status of the person interviewed, and for housing and living conditions, and found no significant differences. The second part displayed 43 items of the specific scale QoL questionnaire. This study on quality of life was built according to the rules and steps defined by the European regulatory issues on quality of life assessment [16]. 2.3. Nine steps led to the construction of the specific QoL scale 2.3.1. Step 1. Selection of specific items for the questionnaire To design this questionnaire, two psychosociologists conducted detailed and in-depth interviews of 20 severe insomniacs. These severe insomniacs were selected from the national SOFRES panel, based on the socio-demographic profile of severe insomniacs defined in a previous national epidemiological survey [2]. Interviews were tape recorded and transcribed, and psychosociologists performed a semantic analysis of recordings. A meeting with the two psychosociologists, two sleep specialists and a linguist was
D. Leger et al. / Sleep Medicine 6 (2005) 191–198
then organized to select the most frequently reported and most relevant items with regard to the clinical experience of the four members of the panel. We therefore proposed to the groups a global QoL questionnaire. 2.3.2. Step 2. Determination of the five dimensions specific to QoL Another goal of the meeting was to find a consensus that the impact of insomnia on QoL may be divided in several dimensions which reflect one aspect of the patients’ QoL. There was no strong support in the literature allowing us to define these dimensions. The international definitions of insomnia [4,5] include the terms of fatigue, irritability, distress, but they do not describe further the consequences of insomnia. There is no study in our knowledge describing nighttime and daytime impacts of insomnia in a large sample. However, we tried to validate the attribution of each of the 43 items to one of the five hypothetical dimensions using a similar process than the one used by the SF-36. 2.3.3. Step 3. Harmonization of the value of the items For each item selected, the subjects had to answer whether, in their case, this item was absolutely true, rather true, neither true nor false, rather false, or absolutely false. All of the items in the same dimension have to be positive or negative. We decided to transpose positive items into negative items to avoid requiring reverse ratings. 2.3.4. Step 4. Transposition of the semantic scale into a numerical scale The solution, which consisted of attributing points ranging, for example, from 1 for absolutely true to 5 for absolutely false, is purely arbitrary, as ‘absolutely false’ does not necessarily have the same weight as ‘absolutely true’. We thus derived numeric scores for each degree of the scale from a factorial analysis of correspondence [16–17], a technique well suited to the analysis of contingency tables opposing on one side the 43 items of the scale and on the other side the five possible answers. 2.3.5. Step 5. Statistical validation of the five dimensions and attribution of each item to one dimension In order to validate the relevancy and each dimension’s statistical independence when compared to the other presupposed dimensions and the attribution of each item to one dimension, we performed a principal components analysis (PCA). Two steps have been completed: (i) a calculation of the average score per item in the three groups in order to check item continuity and to find the most discriminate items, and (ii) a PCA with axis rotation [17] in each dimension in order to elucidate the correlation between items and to avoid redundancies. 2.3.6. Step 6. Calculation of the score per dimension This consisted in aggregating the scores of the various items in one dimension into a final score for the dimension
193
itself. In Step 5, we had determined the weight of each item in the sub-dimensions representative of each dimension. Each sub-dimension has an inertia rate, which can be equated to the weight of each axis in the dimension studied. The score for a dimension was hence determined as the average of the scores for corresponding items, weighted by the mean level of inertia of sub-dimensions to which the item significantly contributed. 2.3.7. Step 7. Convergent validity of the scale The goal of this step is to assess how the scale measures only the phenomenon on which it is supposed to focus. There must thus be a significant correlation between the score and the clinical state of the three groups studied. In our case, the phenomenon measured is quality of life. The clinical states are the levels of sleep disturbance (good sleepers, mild insomniacs and severe insomniacs). However, since clinical state does not equate quality of life, and since the scale is supposed to correlate well to quality of life and less to the clinical state itself, the correlation should not be too strong. 2.3.8. Step 8. Verification of the reliability of the scale The Cronbach coefficient a has been used to test the reliability. 2.3.9. Step 9. Global score calculation In order to calculate the global score of the HD-16, we used a time trade-off (TTO) test [17–18]. According to this method, subjects had to give a score between 0 and 10 on two assessments presented together, with a total score of 10. Nine hundred and fifteen subjects performed that step (346 GS, 372 MI, and 197 SI). The final scores for the five dimensions of the HD-16 were calculated in the three groups and compared within the groups using the t-test.
3. Results 3.1. Subjects and controls Of the 8625 individuals who responded to the poll (76%), 690 were classified as severe insomniacs (8.1%). A group of 700 persons with mild insomnia (MI) and a group of 600 good sleepers (GS) were statistically matched by age, sex, occupation, location and marital status. 3.2. Questionnaire Seventy-eight percent of severe insomniacs (538 individuals), 81% of others (1060 individuals, of which 614 mild insomniacs and 446 good sleepers) responded to Q2. Persons with psychiatric disorders according to DSM-IV minimum criteria for anxiety and depression were eliminated from each group. After the screening, the three groups
194
D. Leger et al. / Sleep Medicine 6 (2005) 191–198
1598 responders to Q2
Others 66% 1060 individuals (614 MI+446 GS)
Severe insomniacs 34% 538 individuals
After screening for depression and anxiety (D/A) profiles
Without D/A profiles 15% 240 individuals
Presenting with D/A 19% 298 individuals
Presenting with D/A 15% 247 individuals
Without D/A profiles 51% 813 individuals
eliminated from the study Severe Insomniacs (SI group) 240 individuals
Mild Insomniacs (MI group) 422 individuals
Good Sleepers (GS group) 391 individuals
Fig. 1. Selection of the three groups after Questionnaire Q2.
were thus reduced to 240 severe insomniacs, 422 mild insomniacs and 391 good sleepers (see Fig. 1). These three groups were once more compared for their socio-economic status. They were not statistically different in terms of age (52.3 years for severe insomniacs, 50.0 for mild insomniacs and 50.9 for good sleepers), occupation (43% active and 57% with no professional activity in the three groups), marital and matrimonial status, location and habitat. However, they were different regarding sex ratio: the SI group was composed of 26% men and 74% women, the MI group of 32% men and 68% women and the GS group of 35% men and 65% women. A mathematical adjustment was made to make the groups similar vis a` vis sex ratio (26% of men and 74% of women). 3.3. Design of the insomnia-specific QoL scale We described the results of the 10 steps, which led to the construction of the specific QoL scale. 3.3.1. Step 1. Selection of specific items for the questionnaire One hundred and twenty-seven items were retained, and the expert group finally selected the 43 most frequently reported and relevant items (one on three). Several items have been eliminated from the most reported list because they were not considered relevant enough: (i) those addressing group-specific (such as women-specific, working adults-specific items) vs. individual-specific concerns, and (ii) items which had no quality of life implications (assertions such as: “I have difficulties paying my bills on
time” were discarded from the questionnaire). A 43-item questionnaire was built and proposed to the three groups: SI, MI and GS. 3.3.2. Step 2. Determination of the five dimensions specific of QoL The meeting led to a consensus that the impact of insomnia on QoL, and the 43 items we selected may be gathered in five descriptive dimensions: (a) physical; (b) energy, will to do things; (c) cognitive (concentration, attention, memory); (d) social (relationship with others); (e) psychological. (Table 1). 3.3.3. Step 3. Harmonization of the items’ value Within the 42 items, 29 were negative and 14 were positive. The 14 positive items have been translated into negative items to avoid requiring reverse ratings (i.e. “I have a good memory” was transposed in “I do not have a good memory”). 3.3.4. Step 4. Transposition of the semantic scale into a numerical scale With a factorial analysis of correspondence we obtained the following scores: Absolutely true: C10.0, Rather true: C6.0, Neither true nor false: C1.0, Rather false: K2.5, Absolutely false: K4.5. These scores, respectively, corresponded to the following data on the axis one: Absolutely true: C0.99, Rather true: 0.59, Neither true nor false: 0.08, Rather false: K0.25, Absolutely false: K0.43.
D. Leger et al. / Sleep Medicine 6 (2005) 191–198 Table 1 Forty-three items initially included in the questionnaire Physical It takes me a long time to wake up in the morning When I wake up, I feel tired already If I go out during the day, I often come back feeling tired I feel slower during some parts of the day I am often sleepy Physical activities make me feel tired If I can, I avoid moving Energy, will to do things I need to be active I am full of energy Most of the time, I don’t have the energy to practice sports I often put off doing things Most of the time, I don’t feel like going out of my house I experience difficulties finishing something I’ve started Most of the time, I don’t have the energy to go out and shop People say I am full of energy Cognitive (concentration, attention, memorization) I am not well organized I have a good memory I cannot concentrate Most of the time, my mind is clear I can easily face new situations Sometimes, I feel my reflexes are slower I regularly break things because of clumsiness I regularly have little accidents Social (relationship with others) In the morning, I don’t even tolerate that someone talks to me Some days, I don’t even want to call my own friends Sometimes, I am aware that I am not kind with people I am comfortable acquainting with people I don’t know I easily lose patience with others I talk easily when I find myself in a group I get along easily with people I often need to be left alone People tell me I am always open to listen to others I am often rough when I talk to people I care about my appearance Psychological I resent vacations or week-ends People tell me I am calm I become irritable easily People tell me I am a cheerful person When I wake up, I feel calm I often think I only run on nervousness I am often anxious I am often worried I am very susceptible
3.3.5. Step 5. Statistical validation of the five dimensions and attribution of each item to one dimension At the end of this analysis, the five dimensions were validated as both relevant and independent of each other. During this step, 13 items were moved from one dimension to another, and the most relevant items for each dimension were enhanced. Each item retained was selected for its ability to represent the sub-dimensions. Hence, the dimension ‘Physical role’ had three sub-dimensions, best represented without redundancy and independence by the four following items: “In the morning, I don’t even
195
tolerate someone talking to me”, “I feel slower during certain parts of the day”, “Most of the time, I don’t have enough energy to practice sports” and “Physical activities make me feel tired”. Eventually, 16 of 43 items were retained because they were the most statistically informative for each dimension. The items eliminated correlated well to other items of the same dimension and were thus not informative. The remaining items were eventually used to calculate the scores for the scale displayed in Table 2. Based on the 16 items selected, and on the name of our hospital which has a history of 16 centuries, we called the scale ‘Hotel Dieu 16’, or HD-16. 3.3.6. Step 6. Calculation of the score per dimension The score per dimension has hence been determined as the average of corresponding item scores, weighted by the mean level of inertia of sub-dimensions to which the item significantly contributed (see Table 2). For example, the item ‘Physical activities make me feel tired’ in the ‘Physical role’ dimension contributed significantly to sub-dimensions 1 and 3 of this dimension. Sub-dimensions 1 and 3 had an inertia rate of 24.31 and 15.36, respectively, and their contributions are 57.01 and 46.61%, respectively. Thus, the coefficient calculated for this item was (24.31!0.5701C 15.36!0.4661)Z21.02. 3.3.7. Step 7. Verification of the convergent validity of the scale We found a correlation score of C0.36 (on a scale from K1 to C1) between the scores and the clinical states. 3.3.8. Step 8. Verification of the reliability of the scale The Cronbach coefficient a was 0.78; if it was over 0.65, it assessed reliability. 3.3.9. Step 9. Global score calculation Nine hundred and fifteen subjects performed this step (346 GS, 372 MI, and 197 SI). GS had a preference for physical fitness (22.039) rather than for concentration, attention, or memory (19.421). SI emphasized relationships with others (21.596) rather than concentration, attention, or memory (17.572). We used these results as level-headedness coefficients in the global score calculation. The sub-scores in each dimension and the total HD-16 score are represented in Table 3 and Fig. 2. The three groups were statistically different for each dimension sub-score and for the total score.
4. Discussion In this study, based on the testimonies of severe insomniacs, we investigated whether it was possible to propose a new QoL instrument focused on insomnia.
196
D. Leger et al. / Sleep Medicine 6 (2005) 191–198
Table 2 HD-16 scale I. Physical role 1. In the morning, I don’t even tolerate that someone talks to me 2. I feel slower during certain parts of the day 3. Most of the time, I don’t have the energy to practice sports 4. Physical activities make me feel tired II. Energy, will to do things 5. Most of the time, I don’t feel like going out of my house 6. I experience difficulties finishing something I’ve started 7. Some days, I don’t even want to call my own friends III. Concentration, attention, memorization 8. I cannot concentrate 9. Most of the time, my mind is not clear 10. I regularly break things because of clumsiness IV. Relationship with the others 11. People say I am not full of energy 12. People tell me I am not a calm person 13. I become irritable easily 14. I often need to be left alone V. Psychological role 15. I often think my nerves are my only drive 16. I am often worried
A A A A
B B B B
C C C C
D D D D
E E E E
A A A
B B B
C C C
D D D
E E E
A A A
B B B
C C C
D D D
E E E
A A A A
B B B B
C C C C
D D D D
E E E E
A A
B B
C C
D D
E E
For each item, please choose the adapted proposal between the following options and complete the column ‘letter’: A, absolutely true; B, rather true; C, neither true nor false; D, rather false; E, absolutely false. To calculate the final HD-16 score. Step 1: Calculate one sub-score for each of the five dimensions (I–V) using for each item answer: AZC10, BZC6, CZC1, DZK2.5, EZK4.5. IZitem 1 score!12.19Citem 2 score!18.31Citem 3 score!6.60Citem 4 score! 21.02. IIZitem 5 score!25.45Citem 6 score!24.90Citem 7 score!19.66. IIIZitem 8 score!23.17Citem 9 score!25.29Citem 10 score!18.73. IVZ item 11 score!10.62Citem 12 score!8.92Citem 13 score!14.08Citem 14 score!8.52.VZitem 15 score!K12.86Citem 16 score!32.17. Step 2: The HD-16 score is I!0.21815CII!0.19770CIII!0.17542CIV!0.21279CV!0.19594.
We carefully designed, following the recommendations of the European regulatory issues on quality of life assessment [16], a short and simple instrument, the HD-16, which differentiated severe insomniacs from mild insomniacs and good sleepers. We believe that this instrument is a new step toward understanding the consequences of insomnia in the
daily lives of patients and might be used, in a second step, in the evaluation of treatments. However, we understand that HD-16 has several limitations. The first one is that we cannot conclude, based on this one study, that HD-16 is an insomnia-specific scale. Many of the 16 items we selected are rather unspecific
Table 3 Results of the different dimensions in the specific HD-16 scale
Physical role: average GSD Range Energy, will to do things average GSD Range Concentration, attention, memorization: average GSD Range Relationship with the others: average GSD Range Psychological role: average AverageGSD Range HD-16 total score: average GSD Range
Good sleepers, GS
Mild insomnia, MI
Severe insomnia, SI
GS vs. MI, P (t-test)
MI vs. SI, P (t-test)
GS vs. SI, P (t-test)
67.41 G8.23 45–87 79.88 G10.23 39–84 83.04
59.02 G6.29 48–85 73.71 G9.4 42–74 78.84
52.90 G8.35 37–75 68.62 G7.56 38–79 75.54
0.000
0.000
0.000
0.002
0.001
0.000
G12.52 63–98 69.36
G10.14 54–89 64.06
G6.82 60–93 60.05
0.000
0.000
0.000
G7.68 51–81 67.76 G7.45 51–91 73.1 G9.24 54–93
G9.55 48–81 57.16 G8.68 31–71 66.0 G8.75 46–83
G6.52 42–74 49.73 G6.21 32–77 61.0 G7.43 39–85
0.000
0.000
0.002
0.000
0.007
0.000
D. Leger et al. / Sleep Medicine 6 (2005) 191–198
197
Physical role
67,41
Phychological role
Energy, will to do things
59,02 73,71
67,76 52,90
57,16
79,88 68,62
49,73 60,05 64,06 75,54 69,36
78,84
83,04
Social: Relationships with the others Legend:
= good sleepers,
= mild insomniacs
Cognitive: Concentration, attention, memorization = severe insomniacs.
Fig. 2. HD-16 among the three groups.
(i.e. “People say that I am not full of energy”, “Physical activities make me feel tired”) and might also be significant in other chronic diseases. Other items seem more specific and are part of the admitted definitions of insomnia (i.e. “I become irritable easily”, “I cannot concentrate”). All of these items came from the testimonies of severe insomniacs transcribed by psychologists and selected by an expert group. They were also significantly different within the three groups. However, the items certainly have to be tested in other sleep disorders and chronic diseases to ascertain to what extent they are really specific to insomnia. We therefore proposed to consider the HD-16 as an insomniafocussed QoL scale rather than as a specific QoL scale on insomnia. In order to prove the specificity of HD-16, it would be necessary to demonstrate that this scale does not vary in other chronic diseases. However, we also believe that sleep disorders are an important part of well being in the QoL of chronic diseases sufferers, and HD-16 would be therefore significantly impaired in other diseases [19,20]. Many studies have in fact shown that insomnia was, by itself, an important contributor in the QoL of chronic diseases, like cancers [21–22], Parkinson disease [23], haemodialysis [24], HIV [25] or psychiatric disorders [26]. At least one study has clearly demonstrated that there was a close relationship between insomnia and chronic illness, and it was therefore difficult to determine the key factor reducing the QoL of patients [27]. In a very comprehensive review on self-reported measures of insomnia in adults, the authors wholly recommend the development of instruments integrating the daytime and nighttime distress of insomniacs and the quality of life [28].
Another aspect is to clarify whether it was useful to design an insomnia-focussed QoL scale rather to use a generic scale like the SF-36. SF-36 has been used extensively in the general population and in many groups of patients [29]. It has been shown that insomnia had a similar impact on QoL that hypertension or chronic obstructive pulmonary disease [13]. SF-36 is easy to complete and seems to be very sensitive to insomnia [8–12]. In a very similar protocol to the one used in this study, we demonstrated that the six dimensions of the SF-36 varied with the severity of insomnia [9]. However, SF36 is also unspecific and we believe that there is a need for a more focussed scale based on the insomniac’s own feeling. Several studies have shown the discrepancy between the insomniac’s poor health perception and the absence of objective assessment by doctors and specialists [1,20, 28–30]. A focussed instrument might be the ideal mean to better reflect the impact of sleep dissatisfaction and insomnia on well being. However, we acknowledge that the HD-16 has to be used in more extensive groups to demonstrate whether it is useful in surveying the natural history of insomniacs and the impact of treatments in the daily lives of patients. In this study, we did not test–retest the scale in the same group of patients to assess the temporal stability of the measure. This would have required that about 10–20% of the initial sample completed the scale a second time, after an interval of a few weeks, and the computation of a simple correlation between test–retest scores. This was not planned in our original protocol, and we were not sure that the temporal stability of insomnia by itself in the three groups would be verified. However, we have scheduled this validation in a next step study.
198
D. Leger et al. / Sleep Medicine 6 (2005) 191–198
The original language of the scale also has to be discussed. The study was designed in France and the questionnaire was therefore in French. Items have been translated for the purpose of this manuscript and this translation, which was done by a native English speaker, has to be considered with circumspection. Even when translation is perfect, certain terms or phrases may have different meaning in different countries. Future studies would probably have to be tested in small samples of subjects in order to understand the meaning of each item. We recommend that for non-English studies, researchers obtain the original French version from the authors, in order to translate it more accurately in their own language with the help of the two versions. Despite these limitations, we believe that the HD-16 is a significant step in the understanding of insomnia and well being. The World Health Organization (WHO) consensus conference has recommended that QoL-specific instruments be developed in order to better understand the day-to-day functioning of patients and better prevent the consequences of insomnia [31]. Our findings suggest that HD-16 can be one of these future instruments and could be used extensively in future studies.
References [1] Ohayon MM. Epidemiology of insomnia: what we know and what we still need to learn. Sleep Med Rev 2002;6(2):97–111. [2] Leger D, Guilleminault C, Dreyfus JP, et al. Prevalence of insomnia in a survey of 12 778 adults in France. J Sleep Res 2000;9:35–42. [3] Hajak G. Epidemiology of severe insomnia and its consequences in Germany. Eur Arch Psychiatry Clin Neurosci 2001;251:49–56. [4] ICSD. International classification of sleep disorders: diagnostic and coding manual. Rochester, MN: American Sleep Disorders Association; 1990. [5] DSM-IV. Diagnostic and statistical manual of mental disorders. 4th ed. Washington, DC: American Psychiatric Association; 1994. [6] Stepanski E, Zorick F, Roehrs T, et al. Daytime alertness in patients with chronic insomnia compared with asymptomatic control subjects. Sleep 1998;11:54–60. [7] Walker SR, Rosser R. Quality of life: assessment and application. Lancaster, OH: MTP Press; 1998. [8] Zamitt GK, Weiner J, Damato N, et al. Quality of life in people with insomnia. Sleep 1999;22:S379–S85. [9] Leger D, Scheuermaier K, Philip P, et al. SF-36: evaluation of quality of life of severe and mild insomniacs compared with good sleepers. Psychosom Med 2001;63:49–55. [10] Morgan K, Dixon S, Mathers N, et al. Psychological treatment for insomnia in the regulation of long-term hypnotic drug use. Health Technol Assess 2004;8:1–68. [11] Manocchia M, Keller S, Ware JE. Sleep problems, health-related quality of life, work functioning and health care utilization among the chronically ill. Qual Life Res 2001;10:331–45.
[12] Schubert CR, Cruickshanks KJ, Dalton DS, et al. Prevalence of sleep problems and quality of life in an older population. Sleep 2002;25: 889–93. [13] Reimer MA, Flemons WW. Quality of life in sleep disorders. Sleep Med Rev 2003;7:335–49. [14] Frosch DL, Kaplan RM, Ganiats TG, et al. Validity of selfadministered quality of well-being scale in musculoskeletal disease. Arthritis Rheum 2004;15(51):28–33. [15] Spertus J, Jones P, Poler S, Rocha-Singh K. The peripheral artery questionnaire: a new disease-specific health status measure for patients with peripheral arterial disease. Am Heart J 2004;147:301–8. [16] Chassany O, Sagnier P, Marquis P, et al. Patient-reported outcomes: the example of health-related quality of life—a European guidance document for the improved integration of health-related quality of life assessment in the drug regulatory process. Drug Inf J 2002;36:209–38. [17] Dolan P, Jones-Lee M. The trade-off: a note on the effect of lifetime reallocation consumption and disounting. J Health Econ 1997;16: 731–9. [18] Green C, Brazier J, Deverill M. Valuing health-related quality of life. A review of health state valuation techniques. Pharmacoeconomics 2000;17:151–65. [19] Ford DE, Kamerow DB. Epidemiologic study of sleep disturbances and psychiatric disorders: an opportunity for prevention? J Am Med Assoc 1989;262:1479–84. [20] Kappler C, Hohagen F. Psychosocial aspects of insomnia. Results of a study in general practice. Eur Arch Psychiatry Clin Neurosci 2003; 253:49–52. [21] Redeker NS, Lev EL, Ruggiero J. Insomnia, fatigue, anxiety, depression, and quality of life of cancer patients undergoing chemotherapy. Sch Inq Nurs Pract 2000;14:275–90. [22] Stark D, Kiely M, Smith A, et al. Anxiety disorders in cancer patients: their nature, associations and quality of life. J Clin Oncol 2002;20: 3137–31748. [23] Caap-Ahlgren M, Dehlin O. Insomnia and depressive symptoms in patients with Parkinson disease. Relationship to health-related quality of life. An interview study of patients living at home. Arch Gerontol Geriatr 2001;32:23–33. [24] Iliescu EA, Coo H, McMurray MH, et al. Quality of sleep and healthrelated quality of life in haemodialysis patients. Nephrol Dial Transplant 2003;18:126–32. [25] Nokes KM, Kendrew J. Correlates of sleep quality in persons with HIV disease. J Assoc Nurses AIDS Care 2001;12:17–22. [26] Stein MB, Barrett-Connor E. Quality of life in older adults receiving medications for anxiety, depression or insomnia. Findings from a community based study. Am J Geriatr Psychiatry 2002;10:568–74. [27] Katz DA, McHorney CA. The relationship between insomnia and health-related quality of life in patients with chronic illness. J Fam Pract 2002;51:229–35. [28] Moul DE, Hall M, Pilkonis PA, Buysse DJ. Self-report measures of insomnia in adults: rationales, choices, and needs. Sleep Med Rev 2004;8:177–98. [29] Jenkinson C, Layte R, Coulter A, Wright L. Evidence for the sensitivity of the Sf-36 health status measure to inequalities in health: result from the Oxford healthy lifestyles survey. J Epidemiol Community Health 1996;50:377–80. [30] Riedel BW, Lichstein KL. Insomnia and daytime functioning. Sleep Med Rev 2000;4:277–98. [31] World Health Organization. Insomnia, an international consensus conference report—Versailles, 13–15 October 1998. Worldwide project on sleep and health. Division of mental health and prevention of substance abuse. World Health Organization report, Geneva; 1998.
Fast-track article
HD-16: a new quality of life instrument specifically designed for insomnia D. Leger*, K. Scheuermaier, T. Raffray, A. Metlaine, D. Choudat, C. Guilleminault Centre du Sommeil, Service de Physiologie, Hoˆtel-Dieu de Paris, 1 pl. du Parvis Notre Dame, 75181 Paris cedex 04, France Received 12 November 2004; received in revised form 4 March 2005; accepted 7 March 2005
Abstract Background and purpose: To design a new quality of life (QoL) instrument specifically for insomnia. Patients and methods: Based on severe insomniacs’ interviews, we have built a new quality of life scale that has been tested in one group of 240 severe insomniacs, in one group of 422 mild insomniacs and in one group of 391 good sleepers. Ten steps led to the construction of a specific QoL scale. Results: Five dimensions have been validated as both relevant and independent from each other. Sixteen items out of the 43 initially tested were retained and significantly different within the groups in each dimension. Based on the 16 items selected, we called the scale Hotel Dieu 16 (HD-16). We have therefore verified the score’s specificity (correlation score of C0.36) and the reliability of the scale (Cronbach coefficient aZ0.78). Conclusion: HD-16 may be used as a focused instrument to better assess an insomniac’s quality of life. q 2005 Elsevier B.V. All rights reserved. Keywords: Insomnia; Quality of life; Sleep; Severe insomnia
1. Introduction Epidemiological studies have shown that insomnia is a very common disease in the general population, affecting between 9 and 15% of adults [1–3]. In addition to disturbed sleep, insomnia is defined as having a negative impact on daytime functions. In the International Classification of Sleep Disorders (ICSD) [4], ‘insomnia is always associated with a complaint of decreased functioning during wakefulness’. In the Diagnostic and Statistical manual of mental Disorders IV (DSM-IV) [5] insomnia is ‘serious enough to induce severe fatigue, irritability or disability in daytime functioning’ and is ‘frequently accompanied by non-specific symptoms such as mood disorders, memory troubles or lack of concentration’. However, previous studies which have evaluated the daytime impact of insomnia with objective parameters, such as multiple sleep latency tests [6] failed to demonstrate objective impairment of insomniacs during the day, despite * Corresponding author. Tel.: C33 142 348 243; fax: C33 142 348 227. E-mail address: [email protected] (D. Leger).
1389-9457/$ - see front matter q 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.sleep.2005.03.013
the fact that, subjectively, insomniacs complained of impaired daytime functioning. Hence, recent studies have concentrated on evaluating the impact of insomnia while other instruments have targeted more subjective dimensions including the perceptions by insomniacs of both nighttime and daytime performance. This may be achieved with the quality of life (QoL) instruments. QoL is a complex and multidimensional term that has been defined as “a concept encompassing a broad range of physical and psychological characteristics and limitations which describe an individual’s ability to function and to derive satisfaction from doing so” [7]. It includes the following domains: physical (the ability to conduct activities of daily living), psychological (or emotional) and social (interactions with family, friends and community). Insomnia may affect all three of these domains. Several studies have now used the SF-36, a generic and well-validated QoL instrument, to assess an impaired QoL in insomniacs [8–12]. Other generic scales such as the Nottingham Health Profile (NHP) and the Sickness Impact Profile (SIP) have also been used [13]. Both have shown that insomnia impaired QoL similarly with chronic disease such as hypertension or chronic obstructive pulmonary disease [13]. However, there is, until now, a lack of specific instruments assessing QoL
192
D. Leger et al. / Sleep Medicine 6 (2005) 191–198
disturbances in insomniacs after controlling for comorbid illnesses and reflecting the degree of severity of insomnia. In many other chronic diseases, it has been shown that a disease-focused QoL instrument was useful to accurately evaluate its impact on the daily lives of patients and to show the efficacy of some treatments [14–15]. We also believe that in the field of insomnia there is a need for specific instruments to better encompass the QoL of patients. The objective of the present study was to build a new QoL instrument specifically designed for insomniacs.
2. Method 2.1. Subjects and controls The SOFRES, a French poll institute, selected the subjects for study. The SOFRES regularly surveys a group of 11,372 individuals representing the French general population. Initially, each individual received a first questionnaire (Q1) assessing insomnia based on the DSMIV and ICSD criteria [4–5]. The SOFRES methodology and the questionnaire used in epidemiological surveys to select groups of mild and severe insomniacs have been previously described and published [2,9]. Severe insomniacs (SI) were categorized as having at least two of the sleep complaints defined in the DSM-IV and ICSD definitions of insomnia (i.e. difficulty initiating sleep or maintaining sleep or nonrestorative sleep) that lasts for at least one month (DSM-IV 307.42), almost nightly with an impaired social or occupational functioning (ICSD). A group of subjects with mild insomnia (MI) and a group of good sleepers (GS) were statistically matched by age, sex, occupation, location and marital status to the SI group. The MI group comprised of people who occasionally complained of sleep difficulties but did not meet the ICSD–DSM-IV criteria for insomnia. The GS group did not complain of any sleep disturbance. The statistical matching is a software process of the SOFRES database, which found groups of MI and GS that were statistically similar to the group of SI with respect to age, sex, occupation, location and marital status. Subjects were not matched by pairs, but groups were statistically not different among these variables. The study was approved by the ethics committee of local authorities of the University Hospital, and the confidentiality of personal data was respected. 2.2. Questionnaire In order to assess quality of life in subjects without depression or anxiety disorders, we subsequently sent a second questionnaire (Q2) to Q1-responders. Q2 had a twofold orientation: (a) detecting psychiatric disorders (based on the DSM-IV definitions of anxiety and depression), and (b) assessing insomnia-specific QoL. General information such as sex, age, occupation, marital and matrimonial status,
location and habitat were also recorded. The questionnaire was mailed with a postage paid return envelope. The participants received neither fee nor reward for taking part in the study. In a first part, Q2 included 10 items based on the DSM-IV minimum criteria for anxiety and depression to eliminate those patients presenting such conditions [5]. An individual was classified as ‘presenting a depression profile’, and thus eliminated from the study, when he presented two of the three following symptoms for more than 15 days: (i) being sad or depressed most of the time almost every day during the day, (ii) having lost interest in activities that normally please him or her, and (iii) being tired all the time. An individual was classified as ‘presenting an anxiety profile’ when he had one of two main anxiety symptoms (being anxious or worried most of the time, or being apprehensive towards daily problems) and was still presenting this (these) symptom(s) and had this (these) symptom(s) for more than six months. After the reception of the questionnaire, persons with psychiatric disorders according to DSM-IV minimum criteria for anxiety and depression were eliminated from each group. The three groups (SI, MI and GS) were once more compared for their socio-economic status. A mathematical adjustment was necessary to make the groups similar vis a` vis sex ratio. This was done by weighting the data with a weighting process based on the deviation between the group studied and the SI group (i.e. if the SI group included 50% of men and 50% of women and the GS group contained 25% of men and 75% of women, the system would have given a !2 weighting coefficient to the data collected from GS men and a !2/3 weighting coefficient to the data collected from the GS women). This system is called ‘the method of quotas’. After this statistical adjustment, we again controlled the three groups for the professional status of the head of the family, for the marital status of the person interviewed, and for housing and living conditions, and found no significant differences. The second part displayed 43 items of the specific scale QoL questionnaire. This study on quality of life was built according to the rules and steps defined by the European regulatory issues on quality of life assessment [16]. 2.3. Nine steps led to the construction of the specific QoL scale 2.3.1. Step 1. Selection of specific items for the questionnaire To design this questionnaire, two psychosociologists conducted detailed and in-depth interviews of 20 severe insomniacs. These severe insomniacs were selected from the national SOFRES panel, based on the socio-demographic profile of severe insomniacs defined in a previous national epidemiological survey [2]. Interviews were tape recorded and transcribed, and psychosociologists performed a semantic analysis of recordings. A meeting with the two psychosociologists, two sleep specialists and a linguist was
D. Leger et al. / Sleep Medicine 6 (2005) 191–198
then organized to select the most frequently reported and most relevant items with regard to the clinical experience of the four members of the panel. We therefore proposed to the groups a global QoL questionnaire. 2.3.2. Step 2. Determination of the five dimensions specific to QoL Another goal of the meeting was to find a consensus that the impact of insomnia on QoL may be divided in several dimensions which reflect one aspect of the patients’ QoL. There was no strong support in the literature allowing us to define these dimensions. The international definitions of insomnia [4,5] include the terms of fatigue, irritability, distress, but they do not describe further the consequences of insomnia. There is no study in our knowledge describing nighttime and daytime impacts of insomnia in a large sample. However, we tried to validate the attribution of each of the 43 items to one of the five hypothetical dimensions using a similar process than the one used by the SF-36. 2.3.3. Step 3. Harmonization of the value of the items For each item selected, the subjects had to answer whether, in their case, this item was absolutely true, rather true, neither true nor false, rather false, or absolutely false. All of the items in the same dimension have to be positive or negative. We decided to transpose positive items into negative items to avoid requiring reverse ratings. 2.3.4. Step 4. Transposition of the semantic scale into a numerical scale The solution, which consisted of attributing points ranging, for example, from 1 for absolutely true to 5 for absolutely false, is purely arbitrary, as ‘absolutely false’ does not necessarily have the same weight as ‘absolutely true’. We thus derived numeric scores for each degree of the scale from a factorial analysis of correspondence [16–17], a technique well suited to the analysis of contingency tables opposing on one side the 43 items of the scale and on the other side the five possible answers. 2.3.5. Step 5. Statistical validation of the five dimensions and attribution of each item to one dimension In order to validate the relevancy and each dimension’s statistical independence when compared to the other presupposed dimensions and the attribution of each item to one dimension, we performed a principal components analysis (PCA). Two steps have been completed: (i) a calculation of the average score per item in the three groups in order to check item continuity and to find the most discriminate items, and (ii) a PCA with axis rotation [17] in each dimension in order to elucidate the correlation between items and to avoid redundancies. 2.3.6. Step 6. Calculation of the score per dimension This consisted in aggregating the scores of the various items in one dimension into a final score for the dimension
193
itself. In Step 5, we had determined the weight of each item in the sub-dimensions representative of each dimension. Each sub-dimension has an inertia rate, which can be equated to the weight of each axis in the dimension studied. The score for a dimension was hence determined as the average of the scores for corresponding items, weighted by the mean level of inertia of sub-dimensions to which the item significantly contributed. 2.3.7. Step 7. Convergent validity of the scale The goal of this step is to assess how the scale measures only the phenomenon on which it is supposed to focus. There must thus be a significant correlation between the score and the clinical state of the three groups studied. In our case, the phenomenon measured is quality of life. The clinical states are the levels of sleep disturbance (good sleepers, mild insomniacs and severe insomniacs). However, since clinical state does not equate quality of life, and since the scale is supposed to correlate well to quality of life and less to the clinical state itself, the correlation should not be too strong. 2.3.8. Step 8. Verification of the reliability of the scale The Cronbach coefficient a has been used to test the reliability. 2.3.9. Step 9. Global score calculation In order to calculate the global score of the HD-16, we used a time trade-off (TTO) test [17–18]. According to this method, subjects had to give a score between 0 and 10 on two assessments presented together, with a total score of 10. Nine hundred and fifteen subjects performed that step (346 GS, 372 MI, and 197 SI). The final scores for the five dimensions of the HD-16 were calculated in the three groups and compared within the groups using the t-test.
3. Results 3.1. Subjects and controls Of the 8625 individuals who responded to the poll (76%), 690 were classified as severe insomniacs (8.1%). A group of 700 persons with mild insomnia (MI) and a group of 600 good sleepers (GS) were statistically matched by age, sex, occupation, location and marital status. 3.2. Questionnaire Seventy-eight percent of severe insomniacs (538 individuals), 81% of others (1060 individuals, of which 614 mild insomniacs and 446 good sleepers) responded to Q2. Persons with psychiatric disorders according to DSM-IV minimum criteria for anxiety and depression were eliminated from each group. After the screening, the three groups
194
D. Leger et al. / Sleep Medicine 6 (2005) 191–198
1598 responders to Q2
Others 66% 1060 individuals (614 MI+446 GS)
Severe insomniacs 34% 538 individuals
After screening for depression and anxiety (D/A) profiles
Without D/A profiles 15% 240 individuals
Presenting with D/A 19% 298 individuals
Presenting with D/A 15% 247 individuals
Without D/A profiles 51% 813 individuals
eliminated from the study Severe Insomniacs (SI group) 240 individuals
Mild Insomniacs (MI group) 422 individuals
Good Sleepers (GS group) 391 individuals
Fig. 1. Selection of the three groups after Questionnaire Q2.
were thus reduced to 240 severe insomniacs, 422 mild insomniacs and 391 good sleepers (see Fig. 1). These three groups were once more compared for their socio-economic status. They were not statistically different in terms of age (52.3 years for severe insomniacs, 50.0 for mild insomniacs and 50.9 for good sleepers), occupation (43% active and 57% with no professional activity in the three groups), marital and matrimonial status, location and habitat. However, they were different regarding sex ratio: the SI group was composed of 26% men and 74% women, the MI group of 32% men and 68% women and the GS group of 35% men and 65% women. A mathematical adjustment was made to make the groups similar vis a` vis sex ratio (26% of men and 74% of women). 3.3. Design of the insomnia-specific QoL scale We described the results of the 10 steps, which led to the construction of the specific QoL scale. 3.3.1. Step 1. Selection of specific items for the questionnaire One hundred and twenty-seven items were retained, and the expert group finally selected the 43 most frequently reported and relevant items (one on three). Several items have been eliminated from the most reported list because they were not considered relevant enough: (i) those addressing group-specific (such as women-specific, working adults-specific items) vs. individual-specific concerns, and (ii) items which had no quality of life implications (assertions such as: “I have difficulties paying my bills on
time” were discarded from the questionnaire). A 43-item questionnaire was built and proposed to the three groups: SI, MI and GS. 3.3.2. Step 2. Determination of the five dimensions specific of QoL The meeting led to a consensus that the impact of insomnia on QoL, and the 43 items we selected may be gathered in five descriptive dimensions: (a) physical; (b) energy, will to do things; (c) cognitive (concentration, attention, memory); (d) social (relationship with others); (e) psychological. (Table 1). 3.3.3. Step 3. Harmonization of the items’ value Within the 42 items, 29 were negative and 14 were positive. The 14 positive items have been translated into negative items to avoid requiring reverse ratings (i.e. “I have a good memory” was transposed in “I do not have a good memory”). 3.3.4. Step 4. Transposition of the semantic scale into a numerical scale With a factorial analysis of correspondence we obtained the following scores: Absolutely true: C10.0, Rather true: C6.0, Neither true nor false: C1.0, Rather false: K2.5, Absolutely false: K4.5. These scores, respectively, corresponded to the following data on the axis one: Absolutely true: C0.99, Rather true: 0.59, Neither true nor false: 0.08, Rather false: K0.25, Absolutely false: K0.43.
D. Leger et al. / Sleep Medicine 6 (2005) 191–198 Table 1 Forty-three items initially included in the questionnaire Physical It takes me a long time to wake up in the morning When I wake up, I feel tired already If I go out during the day, I often come back feeling tired I feel slower during some parts of the day I am often sleepy Physical activities make me feel tired If I can, I avoid moving Energy, will to do things I need to be active I am full of energy Most of the time, I don’t have the energy to practice sports I often put off doing things Most of the time, I don’t feel like going out of my house I experience difficulties finishing something I’ve started Most of the time, I don’t have the energy to go out and shop People say I am full of energy Cognitive (concentration, attention, memorization) I am not well organized I have a good memory I cannot concentrate Most of the time, my mind is clear I can easily face new situations Sometimes, I feel my reflexes are slower I regularly break things because of clumsiness I regularly have little accidents Social (relationship with others) In the morning, I don’t even tolerate that someone talks to me Some days, I don’t even want to call my own friends Sometimes, I am aware that I am not kind with people I am comfortable acquainting with people I don’t know I easily lose patience with others I talk easily when I find myself in a group I get along easily with people I often need to be left alone People tell me I am always open to listen to others I am often rough when I talk to people I care about my appearance Psychological I resent vacations or week-ends People tell me I am calm I become irritable easily People tell me I am a cheerful person When I wake up, I feel calm I often think I only run on nervousness I am often anxious I am often worried I am very susceptible
3.3.5. Step 5. Statistical validation of the five dimensions and attribution of each item to one dimension At the end of this analysis, the five dimensions were validated as both relevant and independent of each other. During this step, 13 items were moved from one dimension to another, and the most relevant items for each dimension were enhanced. Each item retained was selected for its ability to represent the sub-dimensions. Hence, the dimension ‘Physical role’ had three sub-dimensions, best represented without redundancy and independence by the four following items: “In the morning, I don’t even
195
tolerate someone talking to me”, “I feel slower during certain parts of the day”, “Most of the time, I don’t have enough energy to practice sports” and “Physical activities make me feel tired”. Eventually, 16 of 43 items were retained because they were the most statistically informative for each dimension. The items eliminated correlated well to other items of the same dimension and were thus not informative. The remaining items were eventually used to calculate the scores for the scale displayed in Table 2. Based on the 16 items selected, and on the name of our hospital which has a history of 16 centuries, we called the scale ‘Hotel Dieu 16’, or HD-16. 3.3.6. Step 6. Calculation of the score per dimension The score per dimension has hence been determined as the average of corresponding item scores, weighted by the mean level of inertia of sub-dimensions to which the item significantly contributed (see Table 2). For example, the item ‘Physical activities make me feel tired’ in the ‘Physical role’ dimension contributed significantly to sub-dimensions 1 and 3 of this dimension. Sub-dimensions 1 and 3 had an inertia rate of 24.31 and 15.36, respectively, and their contributions are 57.01 and 46.61%, respectively. Thus, the coefficient calculated for this item was (24.31!0.5701C 15.36!0.4661)Z21.02. 3.3.7. Step 7. Verification of the convergent validity of the scale We found a correlation score of C0.36 (on a scale from K1 to C1) between the scores and the clinical states. 3.3.8. Step 8. Verification of the reliability of the scale The Cronbach coefficient a was 0.78; if it was over 0.65, it assessed reliability. 3.3.9. Step 9. Global score calculation Nine hundred and fifteen subjects performed this step (346 GS, 372 MI, and 197 SI). GS had a preference for physical fitness (22.039) rather than for concentration, attention, or memory (19.421). SI emphasized relationships with others (21.596) rather than concentration, attention, or memory (17.572). We used these results as level-headedness coefficients in the global score calculation. The sub-scores in each dimension and the total HD-16 score are represented in Table 3 and Fig. 2. The three groups were statistically different for each dimension sub-score and for the total score.
4. Discussion In this study, based on the testimonies of severe insomniacs, we investigated whether it was possible to propose a new QoL instrument focused on insomnia.
196
D. Leger et al. / Sleep Medicine 6 (2005) 191–198
Table 2 HD-16 scale I. Physical role 1. In the morning, I don’t even tolerate that someone talks to me 2. I feel slower during certain parts of the day 3. Most of the time, I don’t have the energy to practice sports 4. Physical activities make me feel tired II. Energy, will to do things 5. Most of the time, I don’t feel like going out of my house 6. I experience difficulties finishing something I’ve started 7. Some days, I don’t even want to call my own friends III. Concentration, attention, memorization 8. I cannot concentrate 9. Most of the time, my mind is not clear 10. I regularly break things because of clumsiness IV. Relationship with the others 11. People say I am not full of energy 12. People tell me I am not a calm person 13. I become irritable easily 14. I often need to be left alone V. Psychological role 15. I often think my nerves are my only drive 16. I am often worried
A A A A
B B B B
C C C C
D D D D
E E E E
A A A
B B B
C C C
D D D
E E E
A A A
B B B
C C C
D D D
E E E
A A A A
B B B B
C C C C
D D D D
E E E E
A A
B B
C C
D D
E E
For each item, please choose the adapted proposal between the following options and complete the column ‘letter’: A, absolutely true; B, rather true; C, neither true nor false; D, rather false; E, absolutely false. To calculate the final HD-16 score. Step 1: Calculate one sub-score for each of the five dimensions (I–V) using for each item answer: AZC10, BZC6, CZC1, DZK2.5, EZK4.5. IZitem 1 score!12.19Citem 2 score!18.31Citem 3 score!6.60Citem 4 score! 21.02. IIZitem 5 score!25.45Citem 6 score!24.90Citem 7 score!19.66. IIIZitem 8 score!23.17Citem 9 score!25.29Citem 10 score!18.73. IVZ item 11 score!10.62Citem 12 score!8.92Citem 13 score!14.08Citem 14 score!8.52.VZitem 15 score!K12.86Citem 16 score!32.17. Step 2: The HD-16 score is I!0.21815CII!0.19770CIII!0.17542CIV!0.21279CV!0.19594.
We carefully designed, following the recommendations of the European regulatory issues on quality of life assessment [16], a short and simple instrument, the HD-16, which differentiated severe insomniacs from mild insomniacs and good sleepers. We believe that this instrument is a new step toward understanding the consequences of insomnia in the
daily lives of patients and might be used, in a second step, in the evaluation of treatments. However, we understand that HD-16 has several limitations. The first one is that we cannot conclude, based on this one study, that HD-16 is an insomnia-specific scale. Many of the 16 items we selected are rather unspecific
Table 3 Results of the different dimensions in the specific HD-16 scale
Physical role: average GSD Range Energy, will to do things average GSD Range Concentration, attention, memorization: average GSD Range Relationship with the others: average GSD Range Psychological role: average AverageGSD Range HD-16 total score: average GSD Range
Good sleepers, GS
Mild insomnia, MI
Severe insomnia, SI
GS vs. MI, P (t-test)
MI vs. SI, P (t-test)
GS vs. SI, P (t-test)
67.41 G8.23 45–87 79.88 G10.23 39–84 83.04
59.02 G6.29 48–85 73.71 G9.4 42–74 78.84
52.90 G8.35 37–75 68.62 G7.56 38–79 75.54
0.000
0.000
0.000
0.002
0.001
0.000
G12.52 63–98 69.36
G10.14 54–89 64.06
G6.82 60–93 60.05
0.000
0.000
0.000
G7.68 51–81 67.76 G7.45 51–91 73.1 G9.24 54–93
G9.55 48–81 57.16 G8.68 31–71 66.0 G8.75 46–83
G6.52 42–74 49.73 G6.21 32–77 61.0 G7.43 39–85
0.000
0.000
0.002
0.000
0.007
0.000
D. Leger et al. / Sleep Medicine 6 (2005) 191–198
197
Physical role
67,41
Phychological role
Energy, will to do things
59,02 73,71
67,76 52,90
57,16
79,88 68,62
49,73 60,05 64,06 75,54 69,36
78,84
83,04
Social: Relationships with the others Legend:
= good sleepers,
= mild insomniacs
Cognitive: Concentration, attention, memorization = severe insomniacs.
Fig. 2. HD-16 among the three groups.
(i.e. “People say that I am not full of energy”, “Physical activities make me feel tired”) and might also be significant in other chronic diseases. Other items seem more specific and are part of the admitted definitions of insomnia (i.e. “I become irritable easily”, “I cannot concentrate”). All of these items came from the testimonies of severe insomniacs transcribed by psychologists and selected by an expert group. They were also significantly different within the three groups. However, the items certainly have to be tested in other sleep disorders and chronic diseases to ascertain to what extent they are really specific to insomnia. We therefore proposed to consider the HD-16 as an insomniafocussed QoL scale rather than as a specific QoL scale on insomnia. In order to prove the specificity of HD-16, it would be necessary to demonstrate that this scale does not vary in other chronic diseases. However, we also believe that sleep disorders are an important part of well being in the QoL of chronic diseases sufferers, and HD-16 would be therefore significantly impaired in other diseases [19,20]. Many studies have in fact shown that insomnia was, by itself, an important contributor in the QoL of chronic diseases, like cancers [21–22], Parkinson disease [23], haemodialysis [24], HIV [25] or psychiatric disorders [26]. At least one study has clearly demonstrated that there was a close relationship between insomnia and chronic illness, and it was therefore difficult to determine the key factor reducing the QoL of patients [27]. In a very comprehensive review on self-reported measures of insomnia in adults, the authors wholly recommend the development of instruments integrating the daytime and nighttime distress of insomniacs and the quality of life [28].
Another aspect is to clarify whether it was useful to design an insomnia-focussed QoL scale rather to use a generic scale like the SF-36. SF-36 has been used extensively in the general population and in many groups of patients [29]. It has been shown that insomnia had a similar impact on QoL that hypertension or chronic obstructive pulmonary disease [13]. SF-36 is easy to complete and seems to be very sensitive to insomnia [8–12]. In a very similar protocol to the one used in this study, we demonstrated that the six dimensions of the SF-36 varied with the severity of insomnia [9]. However, SF36 is also unspecific and we believe that there is a need for a more focussed scale based on the insomniac’s own feeling. Several studies have shown the discrepancy between the insomniac’s poor health perception and the absence of objective assessment by doctors and specialists [1,20, 28–30]. A focussed instrument might be the ideal mean to better reflect the impact of sleep dissatisfaction and insomnia on well being. However, we acknowledge that the HD-16 has to be used in more extensive groups to demonstrate whether it is useful in surveying the natural history of insomniacs and the impact of treatments in the daily lives of patients. In this study, we did not test–retest the scale in the same group of patients to assess the temporal stability of the measure. This would have required that about 10–20% of the initial sample completed the scale a second time, after an interval of a few weeks, and the computation of a simple correlation between test–retest scores. This was not planned in our original protocol, and we were not sure that the temporal stability of insomnia by itself in the three groups would be verified. However, we have scheduled this validation in a next step study.
198
D. Leger et al. / Sleep Medicine 6 (2005) 191–198
The original language of the scale also has to be discussed. The study was designed in France and the questionnaire was therefore in French. Items have been translated for the purpose of this manuscript and this translation, which was done by a native English speaker, has to be considered with circumspection. Even when translation is perfect, certain terms or phrases may have different meaning in different countries. Future studies would probably have to be tested in small samples of subjects in order to understand the meaning of each item. We recommend that for non-English studies, researchers obtain the original French version from the authors, in order to translate it more accurately in their own language with the help of the two versions. Despite these limitations, we believe that the HD-16 is a significant step in the understanding of insomnia and well being. The World Health Organization (WHO) consensus conference has recommended that QoL-specific instruments be developed in order to better understand the day-to-day functioning of patients and better prevent the consequences of insomnia [31]. Our findings suggest that HD-16 can be one of these future instruments and could be used extensively in future studies.
References [1] Ohayon MM. Epidemiology of insomnia: what we know and what we still need to learn. Sleep Med Rev 2002;6(2):97–111. [2] Leger D, Guilleminault C, Dreyfus JP, et al. Prevalence of insomnia in a survey of 12 778 adults in France. J Sleep Res 2000;9:35–42. [3] Hajak G. Epidemiology of severe insomnia and its consequences in Germany. Eur Arch Psychiatry Clin Neurosci 2001;251:49–56. [4] ICSD. International classification of sleep disorders: diagnostic and coding manual. Rochester, MN: American Sleep Disorders Association; 1990. [5] DSM-IV. Diagnostic and statistical manual of mental disorders. 4th ed. Washington, DC: American Psychiatric Association; 1994. [6] Stepanski E, Zorick F, Roehrs T, et al. Daytime alertness in patients with chronic insomnia compared with asymptomatic control subjects. Sleep 1998;11:54–60. [7] Walker SR, Rosser R. Quality of life: assessment and application. Lancaster, OH: MTP Press; 1998. [8] Zamitt GK, Weiner J, Damato N, et al. Quality of life in people with insomnia. Sleep 1999;22:S379–S85. [9] Leger D, Scheuermaier K, Philip P, et al. SF-36: evaluation of quality of life of severe and mild insomniacs compared with good sleepers. Psychosom Med 2001;63:49–55. [10] Morgan K, Dixon S, Mathers N, et al. Psychological treatment for insomnia in the regulation of long-term hypnotic drug use. Health Technol Assess 2004;8:1–68. [11] Manocchia M, Keller S, Ware JE. Sleep problems, health-related quality of life, work functioning and health care utilization among the chronically ill. Qual Life Res 2001;10:331–45.
[12] Schubert CR, Cruickshanks KJ, Dalton DS, et al. Prevalence of sleep problems and quality of life in an older population. Sleep 2002;25: 889–93. [13] Reimer MA, Flemons WW. Quality of life in sleep disorders. Sleep Med Rev 2003;7:335–49. [14] Frosch DL, Kaplan RM, Ganiats TG, et al. Validity of selfadministered quality of well-being scale in musculoskeletal disease. Arthritis Rheum 2004;15(51):28–33. [15] Spertus J, Jones P, Poler S, Rocha-Singh K. The peripheral artery questionnaire: a new disease-specific health status measure for patients with peripheral arterial disease. Am Heart J 2004;147:301–8. [16] Chassany O, Sagnier P, Marquis P, et al. Patient-reported outcomes: the example of health-related quality of life—a European guidance document for the improved integration of health-related quality of life assessment in the drug regulatory process. Drug Inf J 2002;36:209–38. [17] Dolan P, Jones-Lee M. The trade-off: a note on the effect of lifetime reallocation consumption and disounting. J Health Econ 1997;16: 731–9. [18] Green C, Brazier J, Deverill M. Valuing health-related quality of life. A review of health state valuation techniques. Pharmacoeconomics 2000;17:151–65. [19] Ford DE, Kamerow DB. Epidemiologic study of sleep disturbances and psychiatric disorders: an opportunity for prevention? J Am Med Assoc 1989;262:1479–84. [20] Kappler C, Hohagen F. Psychosocial aspects of insomnia. Results of a study in general practice. Eur Arch Psychiatry Clin Neurosci 2003; 253:49–52. [21] Redeker NS, Lev EL, Ruggiero J. Insomnia, fatigue, anxiety, depression, and quality of life of cancer patients undergoing chemotherapy. Sch Inq Nurs Pract 2000;14:275–90. [22] Stark D, Kiely M, Smith A, et al. Anxiety disorders in cancer patients: their nature, associations and quality of life. J Clin Oncol 2002;20: 3137–31748. [23] Caap-Ahlgren M, Dehlin O. Insomnia and depressive symptoms in patients with Parkinson disease. Relationship to health-related quality of life. An interview study of patients living at home. Arch Gerontol Geriatr 2001;32:23–33. [24] Iliescu EA, Coo H, McMurray MH, et al. Quality of sleep and healthrelated quality of life in haemodialysis patients. Nephrol Dial Transplant 2003;18:126–32. [25] Nokes KM, Kendrew J. Correlates of sleep quality in persons with HIV disease. J Assoc Nurses AIDS Care 2001;12:17–22. [26] Stein MB, Barrett-Connor E. Quality of life in older adults receiving medications for anxiety, depression or insomnia. Findings from a community based study. Am J Geriatr Psychiatry 2002;10:568–74. [27] Katz DA, McHorney CA. The relationship between insomnia and health-related quality of life in patients with chronic illness. J Fam Pract 2002;51:229–35. [28] Moul DE, Hall M, Pilkonis PA, Buysse DJ. Self-report measures of insomnia in adults: rationales, choices, and needs. Sleep Med Rev 2004;8:177–98. [29] Jenkinson C, Layte R, Coulter A, Wright L. Evidence for the sensitivity of the Sf-36 health status measure to inequalities in health: result from the Oxford healthy lifestyles survey. J Epidemiol Community Health 1996;50:377–80. [30] Riedel BW, Lichstein KL. Insomnia and daytime functioning. Sleep Med Rev 2000;4:277–98. [31] World Health Organization. Insomnia, an international consensus conference report—Versailles, 13–15 October 1998. Worldwide project on sleep and health. Division of mental health and prevention of substance abuse. World Health Organization report, Geneva; 1998.