Article Text
PDF
Abstract
OBJECTIVES To inform researchers and clinicians about the most appropriate generic and disease specific measures of health related quality of life for use among people with ischaemic heart disease.
METHODS MEDLINE and BIDS were searched for research papers which contained a report of at least one of the three most common generic instruments or at least one of the five disease specific instruments used with ischaemic heart disease patients. Evidence for the validity, reliability, and sensitivity of these instruments was critically appraised.
RESULTS Of the three generic measures—the Nottingham health profile, sickness impact profile, and short form 36 (SF-36)—the SF-36 appears to offer the most reliable, valid, and sensitive assessment of quality of life. However, a few of the SF-36 subscales lack a sufficient degree of sensitivity to detect change in a patient's clinical condition. According to the best available evidence, the quality of life after myocardial infarction questionnaire should be preferred to the Seattle angina questionnaire, the quality of life index cardiac version, the angina pectoris quality of life questionnaire, and the summary index. Overall, research on disease specific measures is sparse compared to the number of studies which have investigated generic measures.
CONCLUSIONS An assessment of the quality of life of people with ischaemic heart disease should comprise a disease specific measure in addition to a generic measure. The SF-36 and the quality of life after myocardial infarction questionnaire (version 2) are the most appropriate currently available generic and disease specific measures of health related quality of life, respectively. Further research into the measurement of health related quality of life of people with ischaemic heart disease is required in order to address the problems (such as lack of sensitivity to detect change) identified by the review.
- quality of life
- outcome measurement
- ischaemic heart disease
Statistics from Altmetric.com
An important aspect of service evaluation and development is the assessment of the nature and extent to which an intervention or treatment impacts on a patient's illness or condition and on their quality of life. Instruments designed to measure health related quality of life (HRQoL) can be divided into two categories—generic or disease specific. Generic measures provide a broad assessment of the health status of an individual and allow comparisons of HRQoL between groups of patients with different conditions. Often, these instruments are used to monitor progress and to assist in the distribution of resources. Disease specific instruments are designed to measure HRQoL by tapping those areas of life which may be affected by a specific condition or illness. Although these instruments are narrow in focus, they do have a couple of advantages over their generic counterparts. Firstly, disease specific instruments comprise domains or areas which are related closely to the areas of life explored by clinicians, and secondly, disease specific instruments tend to be more sensitive to detecting change in health status than generic instruments.1 As there are benefits with each type of instrument, it is recommended often that both should be used when evaluating HRQoL. However, a critical analysis of the properties of the growing range of generic and disease specific measures is necessary in order to guide and direct researchers and clinicians towards the most appropriate measures in terms of reliability, validity, and sensitivity to change.
Generic instruments
Several generic measures of HRQoL, or health status, have been used among people with ischaemic heart disease. The most commonly used generic instruments in heart disease are the Nottingham health profile (NHP)2 (used in approximately 40% of studies), the short form 36 (SF-36),3 and the sickness impact profile (SIP)4 (both used in approximately 24% of studies). Although there are other generic measures of HRQoL, this paper concentrates on reviewing the research evidence concerning the reliability, validity, and sensitivity of the three most commonly used instruments among this patient group.
The studies reviewed were obtained by a search of the MEDLINE and BIDS databases. Research which did not involve people with an ischaemic heart disease (for example, studies concerned solely with the impact on HRQoL of heart failure or heart transplantation) or which were devoted to the validation of a translated version of the instrument, were excluded from this review. The final list of studies reviewed included samples of patients who were described as having angina, myocardial infarction (MI), coronary artery disease, ischaemic heart disease, or patients who had undergone coronary artery bypass graft (CABG) or angioplasty. Thus, all of the studies reviewed in this paper have drawn their samples from the population of patients with ischaemic heart disease, as defined by the International Classification of Diseases, 10th revision (ICD-10).
Each article extracted from the literature search was examined for information about: reliability in the form of test-retest reliability coefficients and internal consistency coefficients; and sensitivity as indicated by reports of effect sizes (mean change between two points in time divided by baseline SD) or standardised response means (mean change between two points in time divided by SD of change scores). Evidence for validity was also sought in terms of whether an instrument appeared to be valid for use in discriminative or evaluative studies.5 For example, an instrument may be described as having good discriminative validity when patients with severe angina attain significantly higher scores than patients with mild angina, whereas an instrument with evaluative validity is likely to detect significant differences between pre- and post-CABG scores.
Only brief details of the content of each instrument are given, as several authors provide comprehensive reports of their content and development.6 ,7
NOTTINGHAM HEALTH PROFILE
Seventeen studies which used the NHP to measure HRQoL in ischaemic heart disease were reviewed. The NHP has been described as a useful survey tool in terms of assessing whether or not patients have a severe health problem but it does not provide a comprehensive measure of HRQoL.8 Nevertheless, it tends to be used for this purpose in practice. The NHP is divided into two parts. Part 1 requires a yes or no response to 38 statements, which are grouped into six scales: mobility, pain, energy, sleep, emotional reactions, and social isolation. Part 2 of the NHP asks about the effects of health on seven areas of daily life: work, looking after the home, social life, home life, sex life, interests and hobbies, and holidays. The NHP is short and can be administered very quickly because of the limited response choices.
Many studies use part 1 only of the NHP and so relatively little psychometric work has been done on part 2. Results are inconsistent regarding the ability of the NHP scales to discriminate between the clinical classes of angina. This inconsistency may be because the clinical classification systems are not highly regarded as criterion measures.9 Results are also inconsistent regarding the ability of the NHP to discriminate between people with heart disease and healthy people.
All scales have moderate to high test-retest reliability but preintervention ceiling effects and poor responsiveness indices suggest that the NHP may be an inappropriate choice of instrument for evaluative studies. In summary, the NHP may be used to track large changes in health status such as pre- to postcardiac surgery but does not appear to be sensitive to smaller changes within groups. In particular, some scales, such as emotion, appear to have weak validity.
SHORT FORM 36
Ten studies which used the SF-36 to assess HRQoL among people with heart disease were reviewed. The SF-36 is perhaps the most well known and widely used generic health status measure. The 36 items on the questionnaire are grouped into eight scales: physical functioning, social functioning, role limitations caused by physical problems, role limitations caused by emotional problems, mental health, energy/vitality, bodily pain, and general health, and another single item which solicits a self assessment of health change over the past year.
Overall, the SF-36 appears to have good psychometric properties, though further research is required to investigate the sensitivity to change of the SF-36. There is no evidence to indicate that it can discriminate between classes of angina. However, some researchers and clinicians have warned against using the angina classification system as a criterion measure, as it may be insensitive to change, as mentioned previously. As an evaluative tool in the field of ischaemic heart disease, the mental health and general health scales do not appear to be responsive to change, and the role emotional and role physical scales are prone to ceiling effects. These scales may not measure specific changes that a patient with ischaemic heart disease experiences as his/her condition improves or deteriorates. So, results obtained from these four scales should be interpreted with caution.
SICKNESS IMPACT PROFILE
A review of nine studies which used the SIP to measure HRQoL among ischaemic heart disease patients was conducted. The SIP consists of 136 items grouped into 12 categories: sleep and rest, eating, work, home management, recreation and pastimes, ambulation, mobility, body care and movement, social interaction, alertness behaviour, emotional behaviour, and communication. Through cluster and factor analytical techniques three of these categories were further aggregated into a physical dimension (ambulation, mobility, and body care and movement) and four others into a psychosocial dimension (social interaction, alertness behaviour, emotional behaviour, and communication). The other five dimensions cannot be grouped in a coherent manner but an overall score for the SIP can be obtained.10
Internal consistency of the total SIP is high, but there is little information about the internal consistency of the 12 separate scales or the two separate dimensions. There is some information to support the discriminative validity of the physical and psychosocial dimensions and the total score of the SIP, but there is little to suggest that the 12 separate scales of the SIP would be of any value in a discriminative study. The same is true for test-retest reliability. The total SIP score appears to be responsive to changes in patients' health status after surgery but is not responsive to changes over time post-MI. The available evidence suggests that the SIP should not be separated into 12 scales but should be used to obtain a total score or scores for the physical and psychosocial dimensions when used in either discriminative or evaluative studies among patients with ischaemic heart disease.
COMPARISON OF GENERIC INSTRUMENTS
All three generic instruments cover the key areas of physical, social, and emotional functioning. However, the SF-36 and the NHP appear to have better content validity in the field of heart disease because they both cover the areas of energy/vitality and bodily pain. A measure of these domains would be appropriate and useful, if not essential, when exploring the HRQoL of people with heart disease.
A head to head comparison of the SIP and the NHP has been conducted among patients with angina (73% male, average age 65 years)11 and among post-MI patients (in two studies: 40% male, average age 72.7 years12; 81% male, average age 62 years13). Little psychometric differences were found between the two instruments. It was suggested in all these studies that the SIP was slightly more sensitive, though it is much longer and so takes much more time to administer than the NHP.
The NHP has also been compared with the SF-36 for a sample of several types of patients (angina, post-MI and postcardiac surgery) undertaking a cardiac rehabilitation programme (57% male, mean age 57.1 years). The SF-36 had higher internal consistency coefficients and gave clearer evidence of discriminative validity than the NHP.14
Disease specific instruments
Research regarding disease specific HRQoL measures in heart disease was sparse before 1990.15 ,16 Currently, there are approximately 10 instruments. The main ones are the quality of life after myocardial infarction (QLMI),17 the Seattle angina questionnaire (SAQ),18 the quality of life index (QLI)-cardiac version,19 the angina pectoris quality of life questionnaire (APQLQ),20 and the summary index.21
The MEDLINE and BIDS databases were searched using the names of each of the disease specific instruments listed above and the same exclusion criteria were used as outlined for the generic instruments.
QUALITY OF LIFE AFTER MYOCARDIAL INFARCTION
The 26 items on the QLMI questionnaire are grouped into five domains: symptoms, restriction, confidence, self esteem, and emotion.22 Recently, assessments of the psychometric properties and a refinement of the content of the QLMI has been conducted.23 ,24 The refined questionnaire (comprising the QLMI with two original questions removed and three new questions added) is known as the QLMI-2 and groups 27 items into three domains: emotional, physical, and social.
The QLMI-2 appears to have better psychometric properties than the original QLMI, as the domains of the QLMI-2 have higher internal consistency estimates and are based on factor analysis. However, the evaluative properties of the QLMI-2 have yet to be investigated. Most of the QLMI domains have a moderate to strong evaluative dimension, submitting high estimates of test-retest reliability and moderate to high responsiveness indices. As there is little difference in the content of the QLMI and the QLMI-2, it is likely that the QLMI-2 would be at least equally as useful as the QLMI in evaluative studies.
SEATTLE ANGINA QUESTIONNAIRE
The SAQ consists of 19 items, which are grouped into five separate domains: physical limitation, anginal stability, anginal frequency, treatment satisfaction, and disease perception. It is described by the authors as a disease specific functional status measure but seven out of the 19 questions tap social and emotional issues, which makes the SAQ definable as a measure of disease specific HRQoL.
All SAQ domains appear to be psychometrically sound. However, the treatment satisfaction and anginal stability domains may not be suitable for evaluative purposes, as the responsiveness estimate of the former and the test-retest reliability estimates of the latter are very low. Perhaps we should not expect treatment satisfaction to improve along with an improvement in the other SAQ domains. It is possible that a patient's health can improve and their HRQoL can improve, but their satisfaction with treatment may not increase any further. It could also be that the anginal stability scale is very sensitive to change in clinical condition and so perhaps this domain will still identify minor changes in a patient who is defined as stable by other measures.
QUALITY OF LIFE INDEX-CARDIAC VERSION
The QLI was first designed for use with dialysis patients19 but has been adapted for use with other groups, including cardiac patients. The most recent form of the QLI to be used with cardiac patients is the QLI-cardiac version III, which contains 72 items. These items are divided into two equal parts: part 1 measures the satisfaction of patients with various life domains; and part 2 measures the importance of these domains to the patient. Scores on part 1 are weighted by the responses on part 2. Four domains are covered by the 72 items: health and functioning, socioeconomic, psychosocial/spiritual, and family.
Although the version of the QLI discussed here has been designed specifically for cardiac patients, the QLI was not developed originally for use with this population. In addition, it has not been used often with cardiac patients. Therefore, it is unsurprising that further work is required to investigate the discriminative and evaluative value of the QLI-cardiac version.
ANGINA PECTORIS QUALITY OF LIFE QUESTIONNAIRE
The APQLQ has 22 items, which are divided into four scales: physical activities, somatic symptoms, emotional distress, and life satisfaction. The APQLQ has good psychometric properties for discriminative purposes. However, further work, such as an examination of responsiveness and test-retest reliability, is required.
SUMMARY INDEX
The APQLQ has been combined with the anginal impact questionnaire21 and the psychological general well being scale25 to produce a summary index for health related quality of life assessment in angina. The anginal impact questionnaire and the psychological general well being scale both consist of 22 items. The summary index, which combines all three questionnaires, consists of 51 items and can be divided into six categories: impact of angina on daily life, physical exertion, vitality, alertness, self control, and emotional function.
The summary index has received relatively little research attention. However, the reliability and responsiveness properties of the instrument appear to be good. Further evidence to support the validity of the summary index is required. The length of the index and its lack of “user friendliness”, overall, may make it difficult to apply in clinical practice.
COMPARISON OF DISEASE SPECIFIC INSTRUMENTS
A head to head comparison of the SAQ and the QLI-cardiac version has been conducted among patients with angina (all male, average age 65 years).26 The SAQ was able to discriminate between patients in Canadian Cardiovascular Society (CCS) classes I–III, whereas the QLI could not. The SAQ was also found to be slightly more sensitive to clinical changes than the QLI.
No other head to head comparisons of these instruments have been undertaken. This is an area of need in the research if we are to be able to pick the most appropriate disease specific HRQoL instrument in a given situation.
Conclusions
Any review of HRQoL instruments faces the same problem—finding an agreed criterion or “gold standard” against which each instrument can be judged. In the absence of a criterion, HRQoL instruments are often validated against clinical measures of health. This is not wholly satisfactory but it does allow us to judge the validity of the HRQoL instruments in relative terms, as is the case in this review, and clinical measures are meaningful to the practitioner. The psychometric evidence suggests that the SF-36 is the best available generic measure of HRQoL among people with ischaemic heart disease, though some of the SF-36 scales may not be useful in evaluative studies. The SF-36 is the shortest of the three generic measures reviewed, which makes it more acceptable to patients, clinicians, and service providers.
The best disease specific measures appear to be the QLMI-2 or the SAQ. In light of current evidence, a more detailed psychometric analysis of the QLMI-2 has been conducted, when compared to the SAQ, and so it is the recommended choice of disease specific HRQoL measurement instrument for use with patients who have ischaemic heart disease.
This review makes it apparent that some aspects of generic instruments may not be sensitive or perhaps relevant to the specific problems faced by ischaemic heart disease patients. There are also aspects of the disease which impact on a person's life in ways which are not measured by the generic instruments. In order to measure change in HRQoL more accurately and to explore those domains which are affected particularly by ischaemic heart disease, a disease specific measure of HRQoL provides a useful, if not a necessary, addition to the generic instrument.
Acknowledgments
This work was supported by a grant from the Northern Ireland Chest Heart and Stroke Association.
References
Additional article outlining the measures of health related quality of life appears on the Heart website www.heartjnl.com
Supplementary materials
Evidence regarding the validity, reliability, and responsiveness of health related quality of life instruments among people with ischaemic heart disease
Generic instruments
Nottingham Health Profile (NHP)
Validity
Major floor (minimum possible score on the scale) and ceiling (maximum possible score) effects have been reported for the NHP.1 All scales on part one of the NHP were found to produce a ceiling effect (indicating no problems with health status) for at least 25% (50% in the case of the energy and social isolation scales) of a sample of mostly female patients six months after myocardial infarction (MI).2 Ceiling effects have also been reported for between 16.6% and 62.5% of a sample of mostly male patients beginning a cardiac rehabilitation programme on all scales of part 1.3
An examination of the discriminative validity of part 1 of the NHP has shown a significant difference on all six scales between the New York Heart Association (NYHA) classes I, II, and III for people with angina (79% male; mean age 60 years).4 Later studies (with 73% of samples male; mean age 64-65 years) found that the Emotion and Sleep scales did not discriminate between the NYHA classes, unlike the other four scales in part one.1 5 6 The same insignificant results were reported when the Canadian Cardiovascular Society (CCS) classification system was used with a slightly younger all male group.7 Two of these studies also reported that for part 2 of the NHP, only the Home Maintenance scale was able to discriminate between NYHA classes I to III.1 6 There is a poor consensus about the ability of the NHP to discriminate between people with heart disease and samples of healthy people. Part 1 of the NHP has been found to discriminate between healthy people and people with stable coronary artery disease, angioplasty, or coronary artery bypass graft surgery (CABG) (68% male; mean age 59.2 years).8 9 However, using a sample with a similar male:female ratio as the latter studies, only the Mobility scale of part one of the NHP was able to discriminate between healthy people and people who underwent CABG or heart valve operations, and even this relation did not hold for those younger than 55 years.10 Furthermore, only the Energy and Pain scales in part 1 and a total score for part 2 have indicated a significant difference between healthy people and those who had experienced MI within the previous six months (60% female; mean age 73 years).2All six scales of part 1 portray a significant improvement from before to after cardiac surgery in a predominantly male group11-13 and a significant correlation was found between the NHP and ratings of chest pain for a sample of patients two years after cardiac surgery.14 Moreover, when used with people who had experienced MI, the six scales of part 1 showed a significant difference between those able to complete more than 12 minutes on a treadmill test and those who could not; those who had a history of angina within the previous month and those who had no recent history of angina15; those with frequent angina and those with infrequent angina; and finally, those with severe dyspnoea and those with mild or no dyspnoea.16
Reliability
Some researchers have used Cronbach's alpha to provide an indication of the internal consistency of part one of the NHP.3 9 In these studies, the sample was a mixture of people with stable coronary artery disease and those who had angioplasty or CABG and in both studies the alpha values were similar, ranging from 0.58 to 0.84. The Emotion scale had the highest level of internal consistency in both studies, and the Pain, Energy, and Isolation scales submitted the lowest alpha values. Cronbach's alpha ranges from 0 to 1 - the closer to 1, the stronger the internal consistency. Usually, an alpha value of about 0.7 or above is taken as evidence of acceptable internal consistency.
The test�retest reliability of both parts of the NHP was examined among a group of people six months to two years after MI.2 The coefficients for the six scales in part 1 ranged from 0.65 to 0.86. A total score for part 2 had a coefficient of 0.88. Test�retest reliability is a correlation coefficient; usually values of about 0.6 or above are taken as evidence of an acceptable level of reliability.Responsiveness
The responsiveness to change of the NHP is an area that has received little attention. However, effect sizes were calculated after the NHP was administered to people at six weeks and again at six months after MI. The effect sizes of all scales on part 1 were found to be very small (0.01 to 0.17), indicating poor responsiveness to a clinical change which other generic instruments were able to detect.17
Table 1 Summary of the discriminative and evaluative properties of the NHP in ischaemic heart disease Evaluative properties
Discriminative properties
Positive findings All scales on part 1 detected a significant improvement in patients' condition due to cardiac surgery All scales on part 1 could discriminate between clinical classes of angina (NYHA) Test�retest reliability for all scales in part 1 ranged from 0.65 to 0.86 Part 1 could discriminate between patients with heart disease and healthy people Test�retest reliability for a total score in part 2 was 0.88 There was a significant correlation between the NHP score and patients' ratings of chest pain after cardiac surgery All scales in part 1 could discriminate between patients with frequent and infrequent angina, between patients with severe and mild dyspnoea, and between patients who could complete a 12 minute treadmill test and those who could not Internal consistency estimates for all scales on part 1 range from 0.58 to 0.84 Negative findings Major ceiling effects for all scales The Emotion and Sleep scales did not discriminate between clinical classes of angina (NYHA and CCS) All scales on part 1 displayed poor responsiveness to change 6 of the 7 scales on part 2 were unable to discriminate between the clinical classes of angina (NYHA) 5 of the 6 scales on part 1 were unable to discriminate between patients who underwent cardiac surgery and healthy people 4 of the 6 scales on part 1 were unable to discriminate between patients post-MI and healthy people Short Form 36 (SF-36)
Validity
The Role Physical and Role Emotional scales appear to demonstrate minor floor and ceiling effects. Follow up studies of predominantly male patients (mean age 57 to 67 years) with angina or after MI,18 19 undergoing cardiac surgery,20 or taking part in cardiac rehabilitation,3 have estimated that between 15.2% and 62.5% of samples scored 0 (minimum possible score) on these two scales and between 16.4% and 57% scored 100 (maximum possible score) on these scales before beginning the rehabilitation programme or before undergoing surgery. Additionally, the same studies found that the Social Functioning scale and, to a lesser extent, the Bodily Pain scale demonstrated a ceiling effect. It is not unusual for people with ischaemic heart disease to attain high scores, though the large proportion of extremely high scorers suggests that some of the SF-36 scales have weak sensitivity. Furthermore, the ceiling effect indicates that ischaemic heart disease patients have no problems in these areas. This is further evidence for the suggestion that these SF-36 scales are not sensitive to specific problems experienced by ischaemic heart disease patients.All of the SF-36 scales, except Mental Health in the case of women, indicated a significant improvement from before to after participating in cardiac rehabilitation (mean age of sample 61.4 years)21; and all of the SF-36 scales, except Mental Health and Role Physical, detected a significant deterioration in the clinical condition of people with coronary artery disease (84% male; mean age 60 years).22 The General Health scale failed to show a significant change between the preoperative and postoperative state of mostly male cardiac surgery patients (mean age 60 years).20
The Mental Health and Energy scales, respectively were the only ones unable to differentiate significantly between people with coronary artery disease who received a health promotion intervention and those who received usual services (58% male; mean age 66 years).23 In the case of people with angina, studies (predominantly male; mean age 65 to 67 years)19 24 have revealed that the SF-36 scales were unable to distinguish between clinical classes of angina (as determined by the CCS's classification system) and all scales have a low correlation with patients' (96% male; mean age 68 years) single item ratings of current health.25 However, all of the SF-36 scales, except Role Emotional, were found to differentiate significantly between ischaemic heart disease patients who were experiencing angina and those who had a history of angina but were not currently experiencing this symptom.19 Moreover, preoperative scores on the SF-36 scales for male cardiac surgery patients were reported to be lower than norms for the general population26 but approximated to the norms postoperatively.20
Reliability
Estimates of internal consistency (using Cronbach's alpha) for the eight scales of the SF-36 range from 0.76 to 0.93 for people with angina,19 24 from 0.73 to 0.9 for patients undergoing cardiac surgery,20 26 and from 0.72 to 0.85 for cardiac rehabilitation participants, the largest section of whom had an MI while others had an angina attack or had undergone cardiac surgery.27Test�retest reliability after a delay of two weeks ranged from 0.67 to 0.84 for all the SF-36 scales except Bodily Pain, Social Functioning, Role Physical, and Role Emotional for angina patients24 and ranged from 0.7 to 0.84 for all scales except Role Emotional for a group of cardiac rehabilitation participants.27
Responsiveness
Effect sizes and standardised response means (SRM) for all eight scales of the SF-36, based on preoperative and postoperative scores, are available for people who underwent angioplasty procedures.20 Six of the eight scales showed moderate to high responsiveness to change � values ranged from 0.52 to 0.86 for the SRM and from 0.53 to 1.12 for the effect sizes, with each scale submitting a similar value on both indices. However, the Mental Health scale indicated a low responsiveness to change (SRM = 0.33; effect size = 0.32) and the responsiveness of the General Health scale was very poor (SRM = 0.05; effect size = 0.06).SRMs have also been reported after the administration of the SF-36 at two points in time, three months apart, to a group of patients who were mostly diagnosed with stable coronary artery disease.22 The pattern of the resulting SRMs is similar to that reported for angioplasty patients.20 The SRM values ranged from 0.5 to 1.1 except for the Mental Health scale (SRM = 0.4) and the General Health scale (SRM = 0.1).
Table 2 Summary of the discriminative and evaluative properties of the SF-36 in ischaemic heart disease Evaluative properties Discriminative properties Positive findings Most of the scales showed no major floor or ceiling effects All scales could differentiate between preoperative cardiac surgery patients and the general population 6 of the 8 scales detected a deterioration in patients' clinical condition Scores for postoperative cardiac surgery patients on all scales approximate scores for the general population 7 of the 8 scales detected an improvement in the patients' condition, due to cardiac surgery 7 of the 8 scales could differentiate between symptomatic and asymptomatic angina patients Test�retest reliability for most scales ranges from 0.67 to 0.84 Internal consistency estimates for all scales range from 0.72 to 0.93 6 of the 8 scales show high responsiveness to change Negative findings Major ceiling effects on the Role Physical, Role Emotional, and Social Functioning scales None of the scales could differentiate between the clinical classes of angina (CCS) Deterioration in clinical condition not detected by Mental Health and Role Physical scales All scales have a low correlation with patients' ratings of current health Improvement in condition, due to cardiac surgery, not detected by General Health scale Role Emotional scale could not differentiate between symptomatic and asymptomatic angina patients The Bodily Pain, Social Functioning, Role Physical, and Role Emotional scales have shown poor test�retest reliability The Mental Health and General Health scales are not very responsive to change Sickness Impact Profile (SIP)
Validity
The SIP does not suffer from floor and ceiling effects to any large degree.1 The total SIP score has indicated a significant improvement in female patients over the six months after hospital discharge after cardiac surgery28 and for a mixed sex sample over six weeks after hospital discharge after cardiac surgery.29 The Ambulation, Body Care, Social Interaction, Communication, and Alertness Behaviour scales of the SIP, the physical and psychosocial dimensions, and the total score of the SIP have been shown to discriminate to a significant level between the NYHA classes I and III among men, but there is little evidence to suggest that the other scales can discriminate between these classes of angina.1 6 A significant difference of scores on the physical dimension between those who took part in a cardiac recovery programme and those who did not was found for an overwhelmingly male sample (mean age 64 years) after CABG.30 Finally, only the Ambulation, Recreation, and Emotion scales, and the total SIP score indicated a significant difference between a mainly female group of people after MI and a group of healthy people with similar age and sex characteristics.2Reliability
The internal consistencies of the three scales which make up the physical dimension for a mostly male sample (mean age 64 years) after CABG were examined, and estimates ranged between 0.55 and 0.84.30 The internal consistency of the total SIP score for a sample of women after CABG (mean age 62 years) was 0.94.28 The same estimate was reported for a sample of male patients after MI.31For post-MI patients, test�retest reliability estimates of the total SIP score have been reported at 0.92 for a male sample31 and 0.86 for a mainly female sample.2 This latter study also included an estimate of the test�retest reliability of the physical (0.80) and psychosocial (0.82) dimensions of the SIP.
Responsiveness
A high effect size (0.73) for the Mobility scale and moderate effect sizes for the Body Care (0.3), Home Management (0.28), Eating (0.27), Emotion (0.27), and Work (0.23) scales of the SIP have been reported when administered at six weeks and six months after MI to a group of mostly men (mean age 62 years).17 Effect sizes for the remaining scales ranged from 0.01 to 0.15. The effect size for the total score was 0.05. This contrasts with the magnitude of the effect size (1.02) for the total SIP score when used with a group of women at six weeks and six months after cardiac surgery.28
Table 3 Summary of the discriminative and evaluative properties of the SIP in ischaemic heart disease Evaluative properties Discriminative properties Positive findings No major floor or ceiling effects reported 5 of the 12 scales, the physical and psychosocial dimension and the total SIP score were able to discriminate between clinical classes of angina (NYHA) Total SIP score indicated improvements in patients' condition, due to cardiac surgery The physical dimension was able to discriminate between those who entered rehabilitation after cardiac surgery and those who did not Test�retest reliability of the total SIP score has been estimated at 0.86 and 0.92 The total SIP score was able to discriminate between post-MI patients and healthy people Test�retest reliability of the physical and psychosocial dimensions have been estimated at 0.80 and 0.82 respectively Internal consistencies for scales underlying the physical dimension were estimated to range from 0.55 to 0.84 6 of the 12 scales had moderate to high responsiveness indices when administered to post-MI patients Internal consistency estimate of the total SIP score was 0.94 The total SIP score had a high responsiveness to change in patients' condition due to cardiac surgery Negative findings 6 of the 12 scales and the total SIP score had low responsiveness indices when administered to post-MI patients 7 of the 12 scales were unable to discriminate between the clinical classes of angina (NYHA) Most of the SIP scales were not able to discriminate between post-MI patients and healthy people Disease specific instruments
Quality of Life after Myocardial Infarction (QLMI)
Validity
The QLMI-2 appears to have good construct validity in so far as different factor analytic studies produced confirmation of a three factor or domain structure, which explained 66.5% of the variance.32 33 A study of the discriminative validity of the QLMI-2 found a significant difference on all three domains between patients with or without a previous MI; and between those readmitted to hospital within six months of an MI and those who avoided readmission.33Reliability
Test�retest reliability between 8 and 12 months after MI ranged from 0.75 to 0.87 for the five domains of the QLMI. Test�retest reliability is a correlation coefficient; usually values of about 0.6 or above are taken as evidence of an acceptable level of reliability. Internal consistency estimates (using Cronbach's alpha) for these domains ranged from 0.5 to 0.78.34 However, when reduced to the three domains of the QLMI-2, internal consistency improved (alpha = 0.93 tp 0.95).33 Cronbach's alpha ranges from 0 to 1 � the closer to 1 the stronger the internal consistency. Usually, an alpha value of about 0.7 or above is taken as evidence of acceptable internal consistency.Responsiveness
Standardised response mean values for the QLMI ranged from 0.57 to 1.43 for four of the five domains.34 The SRM value for the symptoms domain was 0.27, indicating low sensitivity. The responsiveness of the QLMI-2 has not been studied (or at least reported).
Seattle Angina Questionnaire (SAQ)
Validity
Criterion related validity, by means of correlations with performance on a treadmill test, was found for the physical limitation domain of the SAQ (mean age 61 years; 95% male).35 Correlational designs, using samples with a similar average age and sex ratio to the latter study have also found high criterion related validity between the anginal stability domain and patients' assessment of change over three months; the treatment satisfaction domain and a patient satisfaction questionnaire; and the disease perception domain and the general health scale of the SF-36.35 The last result was repeated and a moderate correlation between the SAQ physical dimension and the SF-36 physical scale was also reported for an all male sample (mean age 65 years).24Scores on the SAQ anginal stability domain discriminated significantly between people with stable and unstable angina (mean age 59 years; 83% male).35 In addition, scores on all of the SAQ domains, except treatment satisfaction, differed significantly between patients in classes I to III of the CCS classification system.24 Furthermore, scores on all scales of the SAQ changed in a manner consistent with the direction of change reported by patients (that is, SAQ scores increased for patients who claimed to experience an improvement in their condition and SAQ scores decreased for patients who claimed to experience a deterioration in their condition) � this was reported not to be the case for the SF-36 scales.22
Reliability
An all male sample (mean age 65 years) with stable angina recorded alpha estimates of internal consistency ranging from 0.66 to 0.89; and two week test�retest reliability coefficients ranging from 0.58 to 0.80 for all SAQ domains except anginal stability (test�retest coefficient = 0.33).24 Three month test�retest reliability coefficients for a sample of people with stable coronary artery disease (mean age 61 years; 95% male) were estimated to range between 0.76 to 0.83 for all SAQ domains, except anginal stability (0.24).35Responsiveness
A responsiveness statistic (mean change in score among patients who improve divided by the standard deviation of scores in stable patients) was calculated for all scales of the SAQ and the results ranged from 1.2 to 2.3 except for the treatment satisfaction domain (0.1). These responsiveness indices were much higher than those found for the SF-36.22
Quality of Life Index � Cardiac Version (QLI)
Validity
The QLI total score and the health and functioning scale have indicated a significant improvement in health related quality of life (HRQoL) after angioplasty for a mostly male sample (83% male; mean age 59 years).36 Significant improvements on the total QLI score and the health and functioning and socioeconomic scales were recorded also for a group of male patients after taking part in a three month angina management programme (mean age 65 years).24 For the same sample, the QLI health and functioning scale correlated moderately with the general health and physical functioning scales of the SF-36; and the QLI psychosocial scale correlated moderately with the mental health, social functioning, and general health scales of the SF-36. However, no significant differences were found between the CCS classifications I to III on any scales of the QLI, including the total score.24Reliability
Internal consistency estimates of the total QLI score, when used with patients undergoing angioplasty, ranged from 0.86 to 0.96.36 37 Test�retest reliability for the total QLI when used with this group of patients has been estimated at 0.87.37 For people with angina, internal consistency estimates range from 0.70 to 0.93 and test�retest reliability estimates range from 0.68 to 0.79.24
Angina Pectoris Quality of Life Questionnaire (APQLQ)
Validity
All scales except life satisfaction were found to correlate highly and significantly with a rating of angina severity; and the somatic symptoms and physical activities scales showed moderate and significant negative correlations with time spent on a treadmill test (mean age of sample 59.5 years; 84% male).38 The APQLQ scales also displayed concurrent validity with relevant SF-36 scales for a mostly male sample (mean age 67 years). The physical activities scale correlated highly (0.76) with the SF-36 physical mobility scale; the emotional distress scale correlated highly (0.84) with the SF-36 mental health scale; and the somatic symptoms scale correlated highly (0.76) with the SF-36 vitality scale.39 In terms of discriminant validity, all of the APQLQ scales and the total APQLQ score indicated a significant difference between symptomatic and asymptomatic patients with coronary artery disease.19 39 Furthermore, the physical activities scale of the APQLQ detected a significant improvement after CABG surgery for a sample of patients with a mean age of 62.5 years.40 This improvement was maintained at three months, one year, and two years after the surgery.41 None of the other APQLQ scales were administered in the latter two studies.Reliability
Internal consistency estimates for the four scales of the APQLQ range from 0.82 to 0.90.39 The total APQLQ has a reported internal consistency estimate of 0.95.19
Summary Index
Validity
The Summary Index discriminated between men with stable angina (mean age 61.7 years) who had two or fewer angina related symptoms and those with more than two symptoms. However, there was no significant difference on the Summary Index between patients with either a high or a low angina attack rate.42 The Summary Index total score showed a significant improvement in HRQoL for patients with angina (all male; mean age 60-61 years) who received transdermal nitroglycerin or long acting oral nitrate medication.43Reliability
Test�retest (four week interval) reliability coefficients for the separate categories of the Summary Index ranged from 0.69 to 0.84. The reliability coefficient for the total Summary Index score was 0.84. Internal consistency estimates for the Summary Index categories were reported to range from 0.91 to 0.98.42Responsiveness
Standardised response means ranged from 0.18 (for the Alertness category) to 0.52 (for the Physical Exertion category). Responsiveness estimates for the total Summary Index score ranged from 0.25 to 0.44.421. Bulpitt CJ. Quality of life as an outcome measure. Postgrad Med J 1997;73:613-16.
2. Visser MC, Koudstaal PJ, Erdman RA, et al. Measuring quality of life in patients with myocardial infarction or stroke: a feasibility study of four questionnaires in the Netherlands. J Epidemiol Commun Health 1995;49:513-17.
3. Dempster M, Bradley J, Wallace E, et al. Measuring quality of life in cardiac rehabilitation: comparing the Short Form 36 and the Nottingham Health Profile. Coronary Health Care 1997;1:211-17.
4. O'Brien BJ, Buxton MJ, Patterson DL. Relationship between functional status and health-related quality-of-life after myocardial infarction. Med Care 1993;31:950-5.
5. Ekeberg O, Klemsdal TO, Kjeldsen SE. Quality of life on enalapril after acute myocardial infarction. Eur Heart J 1994;15:1135-9.
6. Visser MC, Fletcher AE, Parr G, et al. A comparison of three quality of life instruments in subjects with angina pectoris: the Sickness Impact Profile, the Nottingham Health Profile, and the Quality of Well Being Scale. J Clin Epidemiol 1994;47:157-63.
7. Permanyer-Miralda G, Alonso J, Antó JM, et al. Comparison of perceived health status and conventional functional evaluation in stable patients with coronary artery disease. J Clin Epidemiol 1991;44:779-86.
8. Lukkarinen H, Hentinen M. Assessment of quality of life with the Nottingham Health Profile among patients with coronary heart disease. J Adv Nurs 1997;26:73-84.
9. Lukkarinen H, Hentinen M. Assessment of quality of life with the Nottingham Health Profile among women with coronary artery disease. Heart Lung 1998;27:189-99.
10. Trouillet JL, Scheimberg A, Vuagnat A, et al. Long-term outcome and quality of life of patients requiring multidisciplinary intensive care unit admission after cardiac operations. J Thorac Cardiovasc Surg 1996;112:926-34.
11. Caine N, Harrison SCW, Sharples LD, et al. Prospective study of quality of life before and after coronary artery bypass grafting. BMJ 1991;302:511-16.
12. Klersy C, Collarini L, Morellini MC, et al. Heart surgery and quality of life: a prospective study on ischemic patients. Eur J Cardiothorac Surg 1997;12:602-9.
13. Chocron S, Etievent JP, Viel JF, et al. Prospective study of quality of life before and after open heart operations. Ann Thorac Surg 1996;61:153-7.
14. Sjöland H, Wiklund I, Caidahl K, et al. Relationship between quality of life and exercise test findings after coronary artery bypass surgery. Int J Cardiol 1995;51:221-32.
15. Skinner JS, Albers CJ, Hall RJ, et al. Comparison of Nottingham Health Profile (NHP) scores with exercise duration and measures of ischaemia during treadmill exercise testing in patients with coronary artery disease. Eur Heart J 1995;16:1561-5.
16. Wiklund I, Herlitz J, Hjalmarson A. Quality of life five years after myocardial infarction. Eur Heart J 1989;10:464-72.
17. Taylor R, Kirby B, Burdon D, et al. The assessment of recovery in patients after myocardial infarction using three generic quality-of-life measures. J Cardiopulm Rehabil 1998;18:139-44.
18. Ware JE, Snow KK, Kosinski MK, et al. SF-36 Health Survey Manual and Interpretation Guide. Boston: The Health Institute, New England Medical Center; 1993.
19. Marquis P, Fayol C, Joire JE. Clinical validation of a quality of life questionnaire in angina pectoris patients. Eur Heart J 1995;16:1554-60.
20. Krumholz HM, McHorney CA, Clark L, et al. Changes in health after elective percutaneous coronary revascularization - A comparison of generic and specific measures. Med Care 1996;34:754-9.
21. Lavie CJ, Milani RV. Effects of cardiac rehabilitation and exercise training on exercise capacity, coronary risk factors, behavioural characteristics and quality of life in women. Am J Cardiol 1995;75:340-3.
22. Spertus, JA, Winder JA, Dewhurst TA, et al. Monitoring the quality of life in patients with coronary artery disease. Am J Cardiol 1994;74:1240-4.
23. Campbell NC, Thain J, Deans HG, et al. Secondary prevention clinics for coronary heart disease: randomised trial of effect on health. BMJ 1998;316:1434-7.
24. Dougherty CM, Dewhurst T, Nichol WP, et al. Comparison of three quality of life instruments in stable angina pectoris: Seattle Angina Questionnaire, Short Form Health Survey (SF-36), and Quality of Life Index-Cardiac Version III. J Clin Epidemiol 1998;51:569-75.
25. Chen AY, Daley J, Thibault GE. Angina patients' ratings of current health and health without angina: associations with severity of angina and comorbidity. Med Decis Making 1996;16:169-77.
26. McCarthy MJJ, Shroyer AL, Sethi GK, et al. Self-report measures for assessing treatment outcomes in cardiac surgery patients. Med Care 1995;33:OS76-85
27. Jette DU, Downing J. Health status of individuals entering a cardiac rehabilitation program as measured by the Medical Outcomes Study 36 item short form survey. Phys Ther 1994;74:521-7.
28. Redeker NS, Mason DJ, Wykpisz E, et al. Women's patterns of activity over 6 months after coronary artery bypass surgery. Heart Lung 1995;24:502-11.
29. Artinian NT, Duggan C, Miller P. Age differences in patient recovery patterns following coronary artery bypass surgery. Am J Crit Care 1993;2:453-61.
30. Moore SM. The effects of a discharge information intervention on recovery outcomes following coronary artery bypass surgery. Int J Nurs Stud 1996;33:181-9.
31. Munro BH, Creamer AM, Haggerty MR, et al. Effect of relaxation therapy on post-myocardial infarction patients' rehabilitation. Nurs Res 1988;37:231-5.
32. Lim LL-Y, Valenti LA, Knapp JC, et al. A self administered quality of life questionnaire after acute myocardial infarction. J Clin Epidemiol 1993;46:1249-56.
33. Valenti L, Lim L, Heller RF, et al. An improved questionnaire for assessing quality of life after acute myocardial infarction. Qual Life Res 1996;5:151-61.
34. Hillers TK, Guyatt GH, Oldridge N, et al. Quality of life after myocardial infarction. J Clin Epidemiol 1994;47:1287-96.
35. Spertus JA, Winder JA, Dewhurst TA, et al. Development and evaluation of the Seattle Angina Questionnaire - a new functional status measure for coronary artery disease. J Am Coll Cardiol 1995;25:333-41.
36. Bliley AV, Ferrans CE. Quality of life after coronary angioplasty. Heart Lung 1993;22:193-9.
37. Faris J, Stotts N. The effect of percutaneous transluminal coronary angioplasty on quality of life. Progr Cardiovasc Nurs 1990;5:132-40.
38. Wiklund I, Comerford MB, Dimenäs E. The relationship between exercise tolerance and quality of life in angina pectoris. Clin Cardiol 1991;14:204-8.
39. Marquis P, Fayol C, Joire JE, et al. Psychometric properties of a specific quality of life questionnaire in angina pectoris patients. Qual Life Res 1995;4:540-6.
40. Sjoland H, Wiklund I, Caidahl K, et al. Improvement in quality of life and exercise capacity after coronary bypass surgery. Arch Intern Med 1996;156:265-71.
41. Sjoland H, Hartford M, Caidahl K, et al. Improvement in various estimates of quality of life after coronary artery bypass grafting in patients with and without a history of hypertension. J Hypertens 1997;15:1033-9.
42. Wilson A, Wiklund I, Lahti T, et al. A summary index for the assessment of quality of life in angina pectoris. J Clin Epidemiol 1991;44:981-8.
43. Nissinen A, Wiklund I, Lahti T, et al. Antianginal therapy and quality of life - a comparison of the effects of transdermal nitroglycerin and long-acting oral nitrates. J Clin Epidemiol 1991;44:989-97.