Volume 56, Issue 1, Pages 52-60 (January 2003)

7 of 14

ABSTRACT

FULL TEXT

FULL-TEXT PDF (144 KB)

Comparative responsiveness of generic and specific quality-of-life instruments

Received 31 May 2001; received in revised form 2 May 2002; accepted 20 August 2002.

Abstract

We assessed the relative responsiveness of generic and specific quality of life instruments in 43 randomized controlled trials that compared head-to-head 31 generic and 84 specific instruments. Using weighted effect size as the metric of responsiveness, we assessed the impact of instrument type, disease category, and magnitude of underlying therapeutic effect on responsiveness, and assessed the responsiveness of specific instruments relative to the corresponding domains of generic measures. In studies with a nonzero therapeutic effect, specific instruments (mean = 0.57) were significantly more responsive than generic instruments (mean = 0.39, P = .01), and than the corresponding domains of generic instruments (mean = 0.40, P = .03). Studies with low, medium, and high therapeutic effects showed a corresponding gradation in responsiveness differences between specific and generic instruments. We conclude that, overall, specific instruments are more responsive than generic tools, and that investigators may come to misleading conclusions about relative instrument responsiveness if they include studies in which the magnitude of the underlying therapeutic effect is zero.

Keywords: Responsiveness, Health-related quality of life, Generic and specific instruments, Effect size, Randomized controlled trials

Article Outline

• Abstract

• 1. Introduction

• 2. Methods

• 2.1. Analysis

• 3. Results

• 3.1. Included and excluded studies

• 3.2. Instruments

• 3.3. Responsiveness

• 4. Discussion

• Acknowledgment

• Appendix 1.

• abbreviated MEDLINE search strategy (1992–1998)

• Appendix 2.

• effect size computations

• References

• Copyright

1. Introduction

Assessing change over time in health-related quality of life (HRQL) requires instruments capable of capturing any changes that, even if small, are important to patients. This instrument property, referred to as responsiveness, guides researchers' choice of HRQL measures for clinical trials.

Many commentators have suggested that targeted disease-specific or condition-specific HRQL instruments are likely to be more responsive than general or generic instruments, whose strengths include breadth and applicability across conditions and interventions. This intuitive view is based on the ability of specific instruments to focus on health aspects that are important to a specific patient group, as reflected by inclusion of multiple items in each relevant domain. Investigators cannot, however, apply specific instruments across conditions or diverse populations, making comparisons between these different populations impossible. As a result, researchers often have to use a combination of specific and generic instruments to achieve responsiveness and comparability.

Despite the prevailing wisdom, specific instruments have often proved no more responsive than generic instruments. For example, a specific instrument for patients undergoing knee replacement [1], an elder-specific instrument with individualized response items [2], an epilepsy-specific tool [3], and arthritis-specific instruments [4] proved no more responsive than generic tools or analogous generic domains. If, in general, specific instruments proved no more responsive than corresponding domains of generic instruments, the latter would suffice to assess impact of interventions, reducing respondent burden and simplifying outcome assessment. Thus, HRQL investigators are likely to find a comprehensive and unbiased assessment of the relative responsiveness of generic and specific instruments of use.

Murawski and Miederhoff synthesized published data on responsiveness of generic and specific instruments using effect size as a measure of responsiveness [5]. They found specific instruments more responsive than generic tools in studies in which both types of instruments were applied to the same patients but not when comparing studies that used only specific instruments to studies that used only generic instruments. Although providing useful data, this study has several limitations.

First, Murawski and Miederhoff included in the analysis nonrandomized and uncontrolled studies. Improvement in HRQL may result from a combination of natural history, placebo, the impact of the measurement process itself, and intervention effects. The psychologic dimensions of HRQL tools are particularly vulnerable to placebo effects, and effects of the measurement process itself [6]. It is possible that the relative responsiveness of generic and specific measures differs in observational studies and randomized trials; our interest is in the latter category of studies.

Second, Murawski and Miederhoff included studies of interventions without a clear effect on HRQL. Such studies contribute random error to the comparison of generic and specific measures, making differences in responsiveness more difficult to detect. Third, the relative responsiveness of generic and specific instruments may differ across types of interventions or disease categories, and these authors did not address this issue. Fourth, it is not clear how the investigators dealt with studies with multiple interventions. Finally, they did not include the Short Form-36 (SF-36), one of the most widely used generic measures.

Therefore, we undertook an analysis of the comparative responsiveness of generic and specific HRQL instruments used in randomized trials. Our systematic approach included clearly defined eligibility criteria for candidate studies, a comprehensive collection of randomized controlled trials (RCTs) that provided head-to-head comparisons of generic and specific instruments; and a methodology that allowed us to focus on trials with a nonzero underlying treatment effect. We hypothesized that: (1) specific HRQL instruments are more responsive than generic tools; (2) specific instruments are more responsive than corresponding generic-derived domains; (3) including studies in which there was no difference between treatment and control groups blunts the relative impact of instrument type on responsiveness.

2. Methods

We conceptualized HRQL as the patients' subjective experience of health status, and categorized instruments as generic if they were multiple- or single-item tools, applicable to a broad range of patient groups, diseases, and interventions and encompassing all relevant HRQL domains. We considered instruments to be specific if they targeted a particular patient group, disease, intervention or HRQL domain, regardless of the type of study in which the instruments were used. We excluded from the analysis functional scales not capturing the patients' subjective experience of health status. We included in the analysis all parallel group or crossover RCTs that measured HRQL with at least one generic and one specific instrument, applied head-to-head in at least two treatment groups or conditions.

We searched the literature from January1992 to December 1998. An initial exploration using the Cochrane database, the strategy recommended by the Cochrane collaboration to identify RCTs in Medline®, adding “Quality of life,” and “Outcome assessment (health care)” as MeSH terms, and a strategy for CINAHL® suggested by UK Clearing House on Health Outcomes proved highly nonspecific, yielding over 1,500 articles. We abandoned this strategy because it was impractical to review the full text of such a large number of articles. A simplified search strategy (Appendix 1) yielded 473 articles, of which 255 seemed potentially relevant after reviewing the abstract. To identify unpublished, nonindexed, and missed studies we reviewed references of relevant articles and book chapters, contacted 42 researchers in this field from North America and Europe (list available from authors), and posted requests for information in evidence-based health discussion lists (mailbase@mailbase.ac.uk). This disclosed 111 additional, potentially eligible studies. An independent search by a medical librarian found no other relevant studies. Two reviewers independently screened studies for eligibility, abstracted data, and resolved disagreements by consensus.

For each study, we gathered information on its design, disease category, number, and type of treatment groups, number of patients per treatment group, and number and type of HRQL instruments used. For each HRQL instrument, we collected data on its category (i.e., generic or specific), viewpoint (i.e., patient, clinician, other); mode of administration (i.e., self, interviewer, either, or both), and direction of change representing improvement and worsening. We also separated the individual domains of the generic instruments, and specified whether each comparison related to the entire instrument or a domain of that instrument. For each instrument and each treatment group we obtained, at minimum, the mean postintervention scores and their variance. We contacted authors of 22 eligible studies for missing information and obtained a satisfactory response from eight of them. When available, we also gathered preintervention and post- minus preintervention scores and their variance. In studies with multiple measurements over time we used the final measure for the analysis. In studies with more than two treatment groups, we selected the highest and the lowest intensity treatment groups for comparison.

To facilitate comparison of the responsiveness of specific instruments with that of corresponding subdomains of generic instruments, two reviewers (S.W. and S.M.) independently matched each specific instrument with a corresponding constituent subdomain of a generic instrument. Disagreement was resolved by consensus.

2.1. Analysis

We chose effect size as the metric of responsiveness (see Appendix 2). When the denominator of an effect size is the standard deviation of the difference between two observations, it is sometimes referred to as the “standardized response mean” (SRM). The SRM could be viewed as a variant of effect size, and for simplicity we will refer to the statistic as an effect size. We weighted effect sizes by the inverse of their variance before analysis in general linear models.

To test the hypothesis that apparent responsiveness will be attenuated if there are no differences between groups in HRQL, we assessed responsiveness in all studies and separately in those that fulfilled one of three intuitively chosen criteria for a true underlying therapeutic effect: (1) ⩾70% of the effect sizes generated by the HRQL instruments used in a particular study were in one direction and at least one had a P-value ⩽.05 (low threshold); (2) ⩾90% of the effect sizes were in the same direction and at least one had a P-value ⩽.05 (medium threshold); (3) ⩾90% of the effect sizes were in the same direction and at least one had a P-value ⩽.01 (high threshold). These were not mutually exclusive groups of studies, as each category included all the studies exceeding the specified threshold criterion.

Because the studies used more specific instruments than generic instruments, the criteria above could bias the analysis in favour of specific instruments. Therefore, we also used a fourth threshold: including any study in which a generic instrument was associated with a P-value of .05.

Ideally, we would have examined all variables in a single general linear model. However, we did not have enough data to consider all variables simultaneously, particularly in the data sets using the medium and high thresholds. Therefore, we used a series of general linear models to examine the impact of the following variables on effect size: type of instrument (generic versus specific), whole instrument vs. a domain or subdomain of an instrument, disease category, type of effect size calculation (see Appendix 2), and specific versus corresponding domain or subdomain of generic instrument (Table 3). Model 5 implemented an a priori contrast. In models 2–4, we also examined the interactions of these effects with type of instrument. Study was included as a predictor variable in all cases; and the dependent variable was effect size weighted by the inverse of its variance.

In cases where an article provided both domain and subdomain scores and whole instrument scores for the same instrument, we excluded the whole instrument scores from the analysis, because the latter are not independent of the domain or subdomain scores. We chose to retain domain or subdomain scores because, in theory, they are more likely to be responsive. We also conducted the analysis retaining the whole instrument scores rather than the domain or subdomain scores. Because the results did not change appreciably, we report only the analysis in which we retained the domain or subdomain scores.

3. Results

3.1. Included and excluded studies

Out of 87 potentially eligible randomized controlled trials, 40 satisfied eligibility criteria and the articles, or subsequent correspondence with the authors, provided the necessary information for our analysis 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46. Among the 40 articles included, two 18, 42 contained reports on two or more studies. Therefore, the number of studies included in the analysis is 43. We excluded 47 studies for the following reasons. Twenty-three studies gave HRQL data for one instrument category only (generic or specific) 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69. Sixteen studies had insufficient information 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85. Four studies reported QOL for patients in all treatment groups as a single group 86, 87, 88, 89. Three studies measured HRQL but did not provide these data 90, 91, 92. In one article HRQL was not measured at the same time and intervals in all patients [93]. Twenty-five of the 43 studies met our pre-established low threshold of therapeutic effect; of those 25, 16 also passed the medium threshold; and of those 16, 9 also passed the high threshold.

3.2. Instruments

Table 1 shows the number of instruments per disease category. A total of 115 HRQL instruments are included in the analysis, of which 31 (27.0%) are generic and 84 (73.0%) are specific (list available from authors). Investigators used the generic measures Sickness Impact Profile, Nottingham Health Profile, and SF-20 or SF-36, in 13, 6 and 8 articles, respectively.

Table 1.

Summary of instruments included in the analysis


Inclusion thresholda	Variable	Generic	Specific
All studies (N = 43)	Number of instruments per study: mean (SD)	1.5 (1.6)	2.9 (2.5)
	Number of instruments used: N (%)	31 (27.0)	84 (73.0)
	Number of instruments by disease category: N (%)
	Neurological (N = 6)	4 (36.4)	7 (63.6)
	Cardiovascular (N = 15)	12 (25.5)	35 (74.5)
	Pulmonary (N = 7)	5 (21.7)	18 (78.3)
	Musculoskeletal (N = 5)	14 (45.2)	17 (54.8)
	Other disease (N = 10)	6 (27.3)	16 (72.4)
	Number of instruments by completeness: N (%)
	Whole scale	27 (27.6)	71 (72.4)
	Domain or subdomain	7 (26.9)	19 (73.1)
Low threshold (N = 25)	Number of instruments per study: mean (SD)	1.7 (2.0)	2.8 (2.5)
	Number of instruments used: N (%)	24 (29.3)	58 (70.7)
	Number of instruments by disease category: N (%)
	Neurological (N = 5)	3 (33.3)	6 (66.7)
	Cardiovascular (N = 7)	9 (30.0)	21 (70.0)
	Pulmonary (N = 4)	3 (20.0)	12 (80.0)
	Musculoskeletal (N = 3)	13 (46.4)	15 (53.6)
	Other disease (N = 6)	4 (30.8)	9 (69.2)
	Number of instruments by completeness: N (%)
	Whole scale	21 (30.0)	49 (70.0)
	Domain or subdomain	6 (31.6)	13 (68.4)
Medium threshold (N = 16)	Number of instruments per study: mean (SD)	1.4 (0.5)	3.0 (2.7)
	Number of instruments used: N (%)	12 (22.6)	41 (77.4)
	Number of instruments by disease category: N (%)
	Neurological (N = 5)	3 (33.3)	6 (66.7)
	Cardiovascular (N = 5)	6 (27.3)	16 (72.7)
	Pulmonary (N = 1)	2 (22.2)	7 (77.8)
	Musculoskeletal (N = 1)	1 (14.3)	6 (85.7)
	Other disease (N = 4)	4 (36.4)	7 (63.6)
	Number of instruments by completeness: N (%)
	Whole scale	9 (20.5)	35 (79.5)
	Domain or subdomain	6 (42.9)	8 (57.1)
High threshold (N = 9)	Number of instruments per study: mean (SD)	1.4 (0.5)	4.4 (2.7)
	Number of instruments used: N (%)	11 (22.0)	39 (78.0)
	Number of instruments by disease category: N (%)
	Neurological (N = 1)	1 (16.7)	5 (83.3)
	Cardiovascular (N = 4)	6 (27.3)	16 (72.7)
	Pulmonary (N = 1)	2 (22.2)	7 (77.8)
	Musculoskeletal (N = 1)	1 (14.3)	6 (85.7)
	Other disease (N = 2)	3 (33.3)	6 (66.7)
	Number of instruments by completeness: N (%)
	Whole scale	8 (19.0)	34 (81.0)
	Domain or subdomain	5 (45.5)	6 (54.5)

Inclusion thresholds are as described in the analysis section. The number of studies meeting each inclusion threshold is shown in parentheses.

3.3. Responsiveness

Both overall responsiveness, and the difference in responsiveness between generic and specific tools are blunted in studies with low or zero therapeutic effect, as shown by a corresponding gradation of the mean effect size in studies with low, medium, and high therapeutic effect (Table 2). Table 3 shows results from general linear models constructed using studies that met the high therapeutic effect threshold. In studies with a nonzero underlying therapeutic effect, specific instruments (mean = 0.57, SE = 0.06) are significantly more responsive than generic tools (mean = 0.39, SE = 0.07). Moreover, specific instruments were significantly more responsive (mean = 0.56, SE = 0.06) than the analogous domains/subdomains of generic tools (mean = 0.40, SE = 0.08). As shown in Figure 1, the superior responsiveness of specific instruments was consistent across disease categories. The difference between specific and generic instruments was somewhat smaller for cardiovascular studies than for those in other disease categories. However, neither the main effect of disease category nor its interaction with type of instrument was significant (P = .41 and P = .48, respectively). The instrument's score used in the analysis (i.e., the global score, the domain, or the subdomain score), the type of effect size calculation (Appendix 2), and their interactions with the type of instrument, had no detectable impact on responsiveness.

Table 2.

Responsiveness of generic versus specific instruments at various thresholds of therapeutic effect


Inclusion threshold	Number of studies	Number of observations	Mean weighted effect sizea			P
			Overall	Generic	Specific
All studies	43	324	0.21 (0.18)	0.20 (0.18)	0.23 (0.18)	0.20
Low threshold	25	222	0.33 (0.25)	0.32 (0.25)	0.34 (0.25)	0.53
Medium threshold	16	141	0.43 (0.37)	0.38 (0.37)	0.47 (0.37)	0.02
High threshold	9	79	0.48 (0.05)	0.39 (0.07)	0.57 (0.06)	0.01

Cell entries are estimated marginal means (and standard errors) from general linear models with study and type of instrument as predictors of effect size (weighted by the inverse of variance). For studies that reported both Whole-scale and Domain scores, the former were excluded. The P-values are for the main effect of generic versus specific in the same analyses.

Table 3.

Impact of instrument type and other variables on responsiveness in nine studies with a high therapeutic effect (N = 79).


Modela	Effect	F	df	P
1	A. Generic vs. specific	7.14	1, 69	.01
2	A. Generic vs. specific	4.05	1, 67	.05
	B. Whole scale vs. domain or subdomain	0.85	1, 67	.36
	A × B interaction	0.71	1, 67	.40
3	A. Generic vs. specific	6.21	1, 65	.02
	B. Disease category	1.24	4, 4.29	.41
	A × B interaction	0.88	4, 65	.48
4	A. Generic vs. specific	4.94	1, 65	.03
	B. Type of effect size calculation	0.81	3, 9.55	.52
	A × B interaction	0.45	3, 65	.72
5	A. Specific vs. corresponding domain or subdomain of generic	5.85	1, 63	.02

Study was included as an independent variable in all models

View Large Image
Download to PowerPoint

Fig. 1. Mean weighted effect size by type of instrument and disease category.

An analysis restricted to studies in which at least one generic instrument was associated with a P-value of .05 demonstrated that the results favouring specific instruments are not due to a biased selection of studies. We identified six studies in which a generic instrument was associated with a P-value of .05 or less. These studies included 13 generic instruments and 38 specific instruments. The overall effect size in these studies was 0.46, with a standard error 0.06. The corresponding effect sizes and standard errors were 0.37 (0.09) for the generic instruments and 0.52 (0.06) for the specific instruments (difference between generic and specific P = .07).

4. Discussion

Our data demonstrate that, in randomized trials with a true underlying therapeutic effect, specific instruments are more responsive to change than generic instruments and than analogous domains of generic instruments. The results support the notion that researchers need to use a combination of specific and generic instruments to ensure responsiveness, breadth, and comparability across populations. The results, however, may or may not be generalizable to observational studies.

A key aspect of the methodology of this study is our restriction to within-study comparisons. In other words, we have not compared responsiveness across studies (one set of studies for generic instruments, and another set for specific studies) but rather restricted the analysis to within-study comparisons. Thus, in each study, the true underlying magnitude of the treatment effect is fixed, and any differences in apparent magnitude between instruments are attributable to the properties of the instruments being compared.

The main reason for superior responsiveness of specific instruments is likely to be their content, which focuses on aspects of health that are relevant to a particular group of patients. Instrument structure may also play a role. For example, instruments in which items have multiple response options are likely to be more responsive than those with dichotomous responses (e.g., NHP and SIP).

The superior responsiveness of specific instruments, compared with corresponding domains of generic instruments, may be explained by at least two factors. First, specific instruments generally explore a single domain in greater depth than the corresponding domain of a generic tool. Second, subdomains of generic instruments often have a much broader focus than that of corresponding specific instruments. For example, the focus of the subdomain “physical function” in the generic sickness impact profile is much broader than that of a specific instrument assessing dyspnea.

Although they both target aspects of HRQL, the different content of generic and specific instruments suggests that, to an appreciable extent, the two types of instruments are measuring different constructs. Thus, one way of conceptualizing our results is that investigators should choose instruments that treatments are most likely to impact. When treatment is designed primarily to modify symptoms of disease, and the resulting functional and emotional problems, specific instruments that target these aspects of patient experience are most likely to detect treatment effects.

Investigators have not resolved the optimal approach to measuring responsiveness. Indeed, investigators have suggested at least six methods. These include the effect size (the mean change divided by the standard deviation at baseline [94]), the standardized response mean (mean change divided by the standard deviation of change [95]), the responsiveness statistic (mean change over the standard deviation of change in stable subjects [96]), the t-test (mean change divided by the standard error of change), the efficiency statistic (t-ratio squared [97]), and the ROC curve (with an external anchor [98]). The impact of choice of method on results is uncertain. Stucki et al. found the same rank ordering of instruments using several responsiveness indices [99], whereas Wright et al. did not [100].

We chose two variants of the effect size as the measure of responsiveness for our analysis because of their wide use and familiarity. Furthermore, these effect sizes can be applied to a wide variety of presentations of data. This was a crucial issue for our study, in which we combined results across many studies with variable approaches to reporting their data. We were concerned that the results may vary according to the type of effect size measure available for a particular study. One of our general linear models addressed this issue. We found no systematic effect of the type of effect size on the magnitude of the estimate, nor any interaction between the type of effect size and the relative impact of specific and generic instruments.

An important question is how to apply estimates of responsiveness. More specifically, Murawski and Miederhoff [5] suggest that researchers cannot safely conceptualize responsiveness as an instrument's inherent attribute, which can be extrapolated from one clinical setting to another. Instead, they suggest that the population, disease, and intervention under assessment importantly influence responsiveness. Our findings confirm the notion that variables other than instrument category (e.g., underlying therapeutic effect and disease category) affect the assessment of responsiveness. However, the results suggest that considering an instrument as having inherent responsiveness remains a useful concept. If one tries to assess relative responsiveness in situations in which the treatment has no effect, one will find no difference in responsiveness between instruments. Thus, statements about instruments' responsiveness in studies without an underlying therapeutic effect or in cohort studies where most participants are unchanged 101, 102, should be interpreted with caution. Our findings challenge the notion that responsiveness is not an instrument's inherent, portable property because in studies with a true underlying therapeutic effect, head-to-head comparison of generic and specific instruments demonstrates a consistent trend towards superior responsiveness of specific tools across several disease categories.

At the same time, readers should not interpret our results to indicate that all specific instruments are more responsive than all generic instruments or even that a particular specific instrument will prove more responsive than a generic instrument in all contexts when they are used together. Individual specific and generic instruments are more or less responsive. Further, it may well be that relative responsiveness varies across patient populations, or across different interventions. In addition, even when generic instruments are less responsive, they may still provide very useful information beyond that provided by specific instruments. Because they are designed to capture all aspects of HRQL, generic instruments provide a broader context in which to interpret the information about change in HRQL. Our results show only that, in general, specific instruments are more responsive than are generic instruments, and should be interpreted accordingly.

In summary we have demonstrated that specific instruments tend to be more responsive than generic instruments, although the magnitude of the average difference is not large. We have also shown that investigators may come to misleading conclusions about relative instrument responsiveness if they include studies in which the true underlying treatment effect is zero. The results support a policy of including a specific instrument with demonstrated responsiveness in randomized trials in which detection of a treatment effect on health-related quality of life, if one exists, is a critical goal.

Acknowledgements

Drs. Geoff Norman and Stephen Walter provided invaluable assistance with statistical analyses. We thank the following authors who kindly responded to our request for additional information: Charles Goldstein, Ann Jacoby, Andreas Laupacis, J. Lonnqvist, George Torrance, Peter Tugwell, Ingela Wiklund, and Cindy Wong. Lisa Buckingham's skillful data management is appreciated.

Appendix 1.

abbreviated MEDLINE search strategy (1992–1998)

#1 13,327 QUALITY OF LIFE [MESH]

#2 10,853 #1 and [LA = “ENGLISH”]

#3 971 #2 and [PT = “RANDOMIZED-CONTROLLED TRIAL”]

#4 213,854 INSTRUMENT* or SCALE* or QUESTIONNAIRE* or TOOL*

#5 185,370 #4 and [LA = “ENGLISH”]

#6 473 #5 and #3

Appendix 2.

effect size computations

Parallel groups designs

When post-treatment minus pretreatment difference scores, and the standard deviations of those difference scores were available, we computed effect size as follows:

where D̄T = mean post–pre difference for treatment group

D̄C = mean post–pre difference for control group

Pooled SDD = pooled SD of the difference scores

If the pooled SD of the difference scores could not be computed, we substituted the pooled SD of the baseline scores. If only post-treatment scores were available, we used the following:

where X̄T = mean postintervention score for treatment group

X̄C = mean postintervention score for control group

Pooled SDD = pooled SD of the postintervention scores

Crossover designs

When possible, we computed effect size as follows:

If the pooled standard deviation for control and treatment conditions was not available, we substituted the standard deviation of either the control condition or a baseline measure, depending on which was available.

References

1. 1 Kantz ME, Harris WJ, Levitsky K, Ware JE, Davies AR. Methods for assessing condition-specific and generic functional status outcomes after total knee replacement. Med Care. 1992;30:MS240–MS252. MEDLINE | CrossRef

2. 2 Guyatt GH, Eagle DJ, Sackett B, Willan A, Griffith L, McIlroy W, et al. Measuring quality of life in the frail elderly. J Clin Epidemiol. 1993;46:1433–1444. MEDLINE | CrossRef

3. 3 Wiebe S, Rose K, Derry PA, McLachlan RS. Outcome assessment in epilepsy (comparative responsiveness of quality of life and psychosocial instruments). Epilepsia. 1997;38:430–438. MEDLINE | CrossRef

4. 4 Buchbinder R, Bombardier C, Yeung M, Tugwell P. Which outcome measures should be used in rheumatoid arthritis clinical trials? Clinical and quality-of-life measures' responsiveness to treatment in a randomized controlled trial. Arthritis Rheum. 1995;38:1568–1580. MEDLINE | CrossRef

5. 5 Murawski MM, Miederhoff PA. On the generalizability of statistical expressions of health related quality of life instrument responsiveness (a data synthesis). Qual Life Res. 1998;7:11–22. MEDLINE | CrossRef

6. 6 Bouchet C, Guillemin F, Briancon S. Nonspecific effects in longitudinal studies (impact on quality of life measures). J Clin Epidemiol. 1996;49:15–20. Abstract | Full-Text PDF (706 KB) | CrossRef

7. 7 Ahlmen M, Sullivan M, Bjelle A. Team versus non-team outpatient care in rheumatoid arthritis (a comprehensive outcome evaluation including an overall health measure). Arthritis Rheum. 1988;31:471–479. MEDLINE | CrossRef

8. 8 Applegate WB, Phillips HL, Schnaper H, Shepherd AMM, Schocken D, Luhr JC, et al. A randomized controlled trial of the effects of three antihypertensive agents on blood pressure control and quality of life in older women. Arch Intern Med. 1991;151:1817–1823. MEDLINE

9. 9 Boline PD, Kassak K, Bronfort G, Nelson C, Anderson AV. Spinal manipulation vs. amitriptyline for the treatment of chronic tension-type headaches (a randomized clinical trial). J Manipulative Physiol Ther. 1995;18:148–154. MEDLINE

10. 10 Bombardier C, Ware J, Russell IJ, Larson M, Chalmers A, Leighton Read J. Auranofin therapy and quality of life in patients with rheumatoid arthritis (results of a multicentre trial). Am J Med. 1986;81(4):565–578. Abstract | Full-Text PDF (1634 KB) | CrossRef

11. 11 Borson S, Mcdonald GJ, Gayle T, Deffebach M, Lakshminarayan S, VanTuinen C. Improvement in mood, physical symptoms, and function with nortriptyline for depression in patients with chronic obstructive pulmonary disease. Psychosomatics. 1992;33:190–201. MEDLINE

12. 12 Burman P, Broman JE, Hetta J, Wiklund I, Erfurth EM, Hagg E, et al. Quality of life in adults with growth hormone (GH) deficiency (response to treatment with recombinant human GH in a placebo-controlled 21-month trial). J Clin Endocrinol Metab. 1995;80:3585–3590.

13. 13 Chang SW, Fine R, Siegel D, Chesney M, Black D, Hulley SB. The impact of diuretic therapy on reported sexual function. Arch Intern Med. 1991;151:2402–2408. MEDLINE

14. 14 Churchill DN, Bird DR, Taylor DW, Beecroft ML, Gorman J, Wallace JE. Effect on high-flux hemodialysis on quality of life and neuropsychological function in chronic hemodialysis patients. Am J Nephrol. 1992;12:412–418. MEDLINE

15. 15 Coates A, Gebski V, Stat M, Bishop JF, Jeal PN, Woods RL, et al. Improving the quality of life during chemotherapy for advanced breast cancer (a comparison of intermittent and continuous treatment strategies). N Engl J Med. 1987;317:1490–1495. MEDLINE

16. 16 Coats AJS, Adamopoulos S, Meyer TE, Conway J, Sleight P. Effects of physical training in chronic heart failure. Lancet. 1990;335:63–66. Abstract | CrossRef

17. 17 Croog SH, Levine S, Testa MA, Brown B, Bulpitt CJ, Jenkins CD, et al. The effects of antihypertensive therapy on the quality of life. N Engl J Med. 1986;314:1657–1664. MEDLINE

18. 18 Croog SH, Kong BW, Levine S, Weir MR, Baume RM, Saunders E. The Black Hypertension Quality of Life Multicenter Trial Group. Hypertensive black men and women (quality of life and effects of antihypertensive medications). Arch Intern Med. 1990;150:1733–1741. MEDLINE

19. 19 Fletcher A, McLoone P, Bulpitt C. Quality of life on angina therapy (a randomized controlled trial of transdermal glyceryl trinitrate against placebo). Lancet. 1988;2:4–8. MEDLINE

20. 20 Goldstein RS, Gort EH, Stubbing D, Avendano MA, Guyatt GH. Randomised controlled trial of respiratory rehabilitation. Lancet. 1994;344:1394–1397. Abstract | CrossRef

21. 21 Grunfeld E, Mant D, Yudkin P, Adewuyi-Dalton R, Cole D, Stewart J, et al. Routine follow up of breast cancer in primary care (randomised trial). BMJ. 1996;313:665–669.

22. 22 Hilbrands LB, Hoitsma AJ, Koene RA. The effect of immunosuppressive drugs on quality of life after renal transplantation. Transplantation. 1995;59:1263–1270. MEDLINE

23. 23 Jacoby A, Johnson A, Chadwick D. Psychosocial outcomes of antiepileptic drug discontinuation. The Medical Research Council Antiepileptic Drug Withdrawal Study Group. Epilepsia. 1992;33:1123–1131. MEDLINE | CrossRef

24. 24 Jaeschke R, Guyatt GH, Willan A, Cook D, Harper S, Morris J, et al. Effect of increasing doses of beta agonists on spirometric parameters, exercise capacity, and quality of life in patients with chronic airflow limitation. Thorax. 1994;49:479–484. MEDLINE | CrossRef

25. 25 Jones BM, Taylor FJ, Wright OM, Harvison A, McBride M, Spratt PM, et al. Quality of life after heart transplantation in patients assigned to double- or triple-drug therapy. J Heart Transplant. 1990;9:392–396. MEDLINE

26. 26 Jones PW, Bosh TK. Quality of life changes in COPD patients treated with salmeterol. Am J Respir Crit Care Med. 1997;155:1283–1289.

27. 27 Keoghane SR, Lawrence KC, Jenkinson CP, Doll HA, Chappel DB, Cranston DW. The Oxford Laser Prostate Trial (sensitivity to change of three measures of outcome). Urology. 1996;47:43–47. Abstract | Full-Text PDF (478 KB) | CrossRef

28. 28 Laupacis A, Wong C, Churchill DN. The Canadian Erythropoietin Study Group. The use of generic and specific quality-of-life measures in hemodialysis in patients treated with erythropoietin. Control Clin Trials. 1991;12:168S–179S. MEDLINE

29. 29 Levenstein JH. Quality of life, symptoms and pulmonary function in asthma. A year-long multicentre double-blind trial of nedocromil sodium versus placebo. Fam Pract. 1991;8:402–403. MEDLINE

30. 30 Lonnqvist J, Sintonen H, Syvalahti E, Appelberg B, Koskinen T, Mannikko T, et al. Antidepressant efficacy and quality of life in depression (a double-blind study with moclobemide and fluoxetine). Acta Psychiatr Scand. 1994;89:363–369. CrossRef

31. 31 Omvik P, Thaulow E, Herland OB, Eide I, Midha R, Turner RR. Double-blind, parallel, comparative study on quality of life during treatment with amlodipine or enalapril in mild or moderate hypertensive patients (a multicentre study). J Hypertens. 1993;11:103–113. MEDLINE | CrossRef

32. 32 Petajan JH, Gappmaier E, White AT, Spencer MK, Mino L, Hicks RW. Impact of aerobic training on fitness and quality of life in multiple sclerosis. Ann Neurol. 1996;39:432–441. MEDLINE | CrossRef

33. 33 Regensteiner JG, Steiner JF, Hiatt WR. Exercise training improves functional status in patients with peripheral arterial disease. J Vasc Surg. 1996;23:104–115. Abstract | Full-Text PDF (990 KB) | CrossRef

34. 34 Regensteiner JG, Meyer TJ, Krupski WC, Cranford LS, Hiatt WR. Hospital vs home-based exercise rehabilitation for patients with peripheral arterial occlusive disease. Angiology. 1997;48:291–300. MEDLINE | CrossRef

35. 35 Ries AL, Kaplan RM, Limberg TM, Prewitt LM. Effects of pulmonary rehabilitation on physiologic and psychosocial outcomes in patients with chronic obstructive pulmonary disease. Ann Intern Med. 1995;122:823–832. MEDLINE

36. 36 Rutten van Molken MP, Custers F, van Doorslaer EK, Jansen CC, Heurman L, Maesen FP, et al. Comparison of performance of four instruments in evaluating the effects of salmeterol on asthma quality of life. Eur Respir J. 1995;8:888–898. MEDLINE

37. 37 Salek MS, Finlay AY, Luscombe DK, Allen BR, Berth Jones J, Camp RD, et al. Cyclosporin greatly improves the quality of life of adults with severe atopic dermatitis. A randomized, double-blind, placebo-controlled trial. Br J Dermatol. 1993;129:422–430. MEDLINE | CrossRef

38. 38 Singh NA, Clements KM, Fiatarone MA. A randomized controlled trial of the effect of exercise on sleep. Sleep. 1997;20:95–101. MEDLINE

39. 39 Smith DF, Baker GA, Davies G, Dewey M, Chadwick DW. Outcomes of add-on treatment with lamotrigine in partial epilepsy. Epilepsia. 1993;34(2):312–322. MEDLINE | CrossRef

40. 40 Stenstrom CH, Arge B, Sundbom A. Dynamic training versus relaxation training as home exercise for patients with inflammatory rheumatic diseases. A randomized controlled study. Scand J Rheumatol. 1996;25:28–33. MEDLINE | CrossRef

41. 41 Testa MA, Hollenberg NK, Anderson RB, Williams GH. Assessment of quality of life by patient with spouse during antihypertensive therapy with atenolol and nifedipine gastrointestinal therapeutic system. Am J Hypertens. 1991;4:363–373. MEDLINE

42. 42 Testa MA, Anderson RB, Nackley JF, Hollenberg NK. The Quality-of-Life Hypertension Study Group. Quality of life and antihypertensive therapy in men (a comparison of captopril with enalapril). N Engl J Med. 1993;328:907–913. MEDLINE | CrossRef

43. 43 Tugwell P, Bombardier C, Buchanan WW, Goldsmith C, Grace E, Bennett KJ, et al. Methotrexate in rheumatoid arthritis. Impact on quality of life assessed by traditional standard-item and individualized patient preference health status questionnaires. Arch Intern Med. 1990;150:59–62. MEDLINE

44. 44 Watson RG, Tham TC, Johnston BT, McDougall NI. Double blind cross-over placebo controlled study of omeprazole in the treatment of patients with reflux symptoms and physiological levels of acid reflux—the “sensitive oesophagus.”. Gut. 1997;40:587–590. MEDLINE

45. 45 Wiklund I, Karlberg J, Mattsson LA. Quality of life of postmenopausal women on a regimen of transdermal estradiol therapy: a double-blind placebo-controlled study. Am J Obstet Gynecol 1993 168(3pt1):824–830.

46. 46 Worsfold C, Langridge J, Spalding A, Mullee MA. Comparison between primary care physiotherapy education/advice clinics and traditional hospital based physiotherapy treatment (a randomized trial). Br J Gen Pract. 1996;46:165–168. MEDLINE

47. 47 Block G, Liss C, Reines S, Irr J, Nibbelink D. Comparison of immediate-release and controlled release carbidopa/levodopa in Parkinson's disease. A multicenter 5-year study. The CR First Study Group. Eur Neurol. 1997;37:23–27. MEDLINE | CrossRef

48. 48 Coleman EA, Buchner DM, Cress ME, Chan BK, de Lateur BJ. The relationship of joint symptoms with exercise performance in older adults. J Am Geriatr Soc. 1996;44:14–21. MEDLINE

49. 49 Heiligenstein JH, Ware JE, Beusterien KM, Roback PJ, Andrejasich C, Tollefson GD. Acute effects of fluoxetine versus placebo on functional health and well-being in late-life depression. Int Psychogeriatr. 1995;7(Suppl):125–137. CrossRef

50. 50 Hilding MB, Backbro B, Ryd L. Quality of life after knee arthroplasty. A randomized study of 3 designs in 42 patients, compared after 4 years. Acta Orthop Scand. 1997;68:156–160. MEDLINE

51. 51 Leanderson J, Wredmark T. Treatment of acute ankle sprain. Comparison of a semi-rigid ankle brace and compression bandage in 73 patients. Acta Orthop Scand. 1995;66:529–531. MEDLINE

52. 52 Bousquet J, Duchateau J, Pignat JC, Fayol C, Marquis P, Mariz S, et al. Improvement of quality of life by treatment with cetirizine in patients with perennial allergic rhinitis as determined by a French version of the SF-36 questionnaire. J Allergy Clin Immunol. 1996;98:309–316. Abstract | Full Text | Full-Text PDF (743 KB) | CrossRef

53. 53 Greenfield S, Kaplan S, Ware JE. Expanding patient involvement in care (effects on patient outcomes). Ann Intern Med. 1985;102:520–528. MEDLINE

54. 54 Greenfield S, Kaplan SH, Ware JE, Yano EM, Frank HJL. Patients' participation in medical care (effects on blood sugar control and quality of life in diabetes). J Gen Intern Med. 1988;3:448–457. MEDLINE | CrossRef

55. 55 Hallbook O, Hass U, Wanstrom A, Sjodahl R. Quality of life measurement after rectal excision for cancer. Comparison between straight and colonic J-pouch anastomosis. Scand J Gastroenterol. 1997;32:490–493. MEDLINE | CrossRef

56. 56 Hurny C, Bernhard J, Coates AS, Castiglione Gertsch M, Peterson HF, Gelber RD, et al. Impact of adjuvant therapy on quality of life in women with node-positive operable breast cancer. International Breast Cancer Study Group impact of adjuvant therapy on quality of life in women with node-positive operable breast cancer. International Breast Cancer Study Group. Lancet. 1996;347:1279–1284.

57. 57 Jenkinson C, Lawrence K, McWhinnie D, Gordon J. Sensitivity to change of health status measures in a randomized controlled trial (comparison of the COOP charts and the SF-36). Qual Life Res. 1995;4:47–52. MEDLINE | CrossRef

58. 58 Mansson A, Colleen S, Hermeren G, Johnson G. Which patients will benefit from psychosocial intervention after cystectomy for bladder cancer?. Br J Urol. 1997;80:50–57. MEDLINE

59. 59 O'Neill C, Normand C, Cupples M, McKnight A. A comparison of three measures of perceived distress (results from a study of angina patients in general practice in Northern Ireland). J Epidemiol Commun Health. 1996;50:202–206.

60. 60 Okamoto LJ, Noonan M, DeBoisblanc BP, Kellerman DJ. Fluticasone propionate improves quality of life in patients with asthma requiring oral corticosteroids. Ann Allergy Asthma Immunol. 1996;76:455–461. MEDLINE

61. 61 Patterson RB, Pinto B, Marcus B, Colucci A, Braun T, Roberts M. Value of a supervised exercise program for the therapy of arterial claudication. J Vasc Surg. 1997;25:312–318. Abstract | Full Text | Full-Text PDF (789 KB) | CrossRef

62. 62 Pocock SJ, Henderson RA, Seed P, Treasure T, Hampton JR. Quality of life, employment status, and anginal symptoms after coronary angioplasty or bypass surgery. 3-year follow-up in the Randomized Intervention Treatment of Angina (RITA) Trial. Circulation. 1996;94:135–142. MEDLINE

63. 63 Sand PK, Richardson DA, Staskin DR, Swift SE, Appell RA, Whitmore KE, et al. Pelvic floor electrical stimulation in the treatment of genuine stress incontinence (a multicenter, placebo-controlled trial). Am J Obstet Gynecol. 1995;173:72–79. Abstract | Full-Text PDF (1956 KB) | CrossRef

64. 64 Simon GE, VonKorff M, Heiligenstein JH, Revicki DA, Grothaus L, Katon W, et al. Initial antidepressant choice in primary care. Effectiveness and cost of fluoxetine vs tricyclic antidepressants. JAMA. 1996;275:1897–1902. MEDLINE

65. 65 Sullivan M, Katon W, Russo J, Dobie R, Sakai C. A randomized trial of nortriptyline for severe chronic tinnitus (effects on depression, disability, and tinnitus symptoms). Arch Intern Med. 1993;153:2251–2259. MEDLINE

66. 66 The German and Austrian Xamoterol Study Group . Double-blind placebo-controlled comparison of digoxin and xamoterol in chronic heart failure. Lancet. 1988;March:489–493.

67. 67 Weinberger M, Kirkman MS, Samsa GP, Shortliffe EA, Landsman PB, Cowper PA, et al. A nurse-coordinated intervention for primary care patients with non-insulin-dependent diabetes mellitus (impact on glycemic control and health-related quality of life). J Gen Intern Med. 1995;10:59–66. MEDLINE | CrossRef

68. 68 Werner RA, Kessler S. Effectiveness of an intensive outpatient rehabilitation program for postacute stroke patients. Am J Phys Med Rehabil. 1996;75:114–120. MEDLINE | CrossRef

69. 69 Williams HJ, Willkens RF, Samuelson CO, Alarcon GS, Guttadauria M, Yarboro C, et al. Comparison of low-dose oral pulse methotrexate and placebo in the treatment of rheumatoid arthritis (a controlled clinical trial). Arthritis Rheum. 1985;28:721–730. MEDLINE | CrossRef

70. 70 Bakker C, Rutten M, van Santen Hoeufft M, Bolwijn P, van Doorslaer E, Bennett K, et al. Patient utilities in fibromyalgia and the association with other outcome measures. J Rheumatol. 1995;22:1536–1543.

71. 71 Coulehan JL, Schulberg HC, Block MR, Madonia MJ, Rodriguez E. Treating depressed primary care patients improves their physical, mental, and social functioning. Arch Intern Med. 1997;157:1113–1120. MEDLINE

72. 72 Earlam S, Glover C, Davies M, Fordy C, Allen Mersh TG. Effect of regional and systemic fluorinated pyrimidine chemotherapy on quality of life in colorectal liver metastasis patients. J Clin Oncol. 1997;15:2022–2029.

73. 73 Engblom E, Korpilahti K, Hamalainen H, Ronnemaa T, Puukka P. Quality of life and return to work 5 years after coronary artery bypass surgery. Long-term results of cardiac rehabilitation. J Cardpulm Rehabil. 1997;17:29–36.

74. 74 Filipi CJ, Gaston Johansson F, McBride PJ, Murayama K, Gerhardt J, Cornet DA, et al. An assessment of pain and return to normal activity. Laparoscopic herniorrhaphy vs open tension-free Lichtenstein repair. Surg Endosc. 1996;10:983–986. MEDLINE | CrossRef

75. 75 Fletcher AE, Bulpitt CJ, Chase DM, Collins WCJ, Furberg CD, Goggin TK, et al. Quality of life with three antihypertensive treatments (cilazapril, atenolol, nifedipine). Hypertension. 1992;19:499–507.

76. 76 Lavin RA, Pappagallo M, Kuhlemeier KV. Cervical pain (a comparison of three pillows). Arch Phys Med Rehabil. 1997;78:193–198. Abstract | Full-Text PDF (761 KB) | CrossRef

77. 77 Mahajan P, Okamoto LJ, Schaberg A, Kellerman D, Schoenwetter WF. Impact of fluticasone propionate powder on health-related quality of life in patients with moderate asthma. J Asthma. 1997;34:227–234. MEDLINE | CrossRef

78. 78 Manzetti JD, Hoffman LA, Sereika SM, Sciurba FC, Griffith BP. Exercise, education, and quality of life in lung transplant candidates. J Heart Lung Transplant. 1994;13:297–305. MEDLINE

79. 79 Money SR, Herd JA, Isaacsohn JL, Davidson M, Cutler B, Heckman J, et al. Effect of cilostazol on walking distances in patients with intermittent claudication caused by peripheral vascular disease. J Vasc Surg. 1998;27:267–274. Abstract | Full Text | Full-Text PDF (81 KB) | CrossRef

80. 80 Persson LC, Carlsson CA, Carlsson JY. Long-lasting cervical radicular pain managed with surgery, physiotherapy, or a cervical collar. A prospective, randomized study. Spine. 1997;22:751–758. MEDLINE | CrossRef

81. 81 Rush DR, Stelmach WJ, Young TL, Kirchdoerfer LJ, Scott Lennox J, Holverson HE, et al. Clinical effectiveness and quality of life with ranitidine vs placebo in gastroesophageal reflux disease patients (a clinical experience network (CEN) study). J Fam Pract. 1995;41:126–136. MEDLINE

82. 82 Small GW, Birkett M, Meyers BS, Koran LM, Bystritsky A, Nemeroff CB. Impact of physical illness on quality of life and antidepressant response in geriatric major depression. Fluoxetine Collaborative Study Group. J Am Geriatr Soc. 1996;44:1220–1225. MEDLINE

83. 83 Tandon PK, Stander H, Schwarz RP. Analysis of quality of life data from a randomized, placebo-controlled heart-failure trial. J Clin Epidemiol. 1989;42(10):955–962. MEDLINE | CrossRef

84. 84 Weir MR, Berger ML, Weeks ML, Liss CL, Santanello NC. Comparison of the effects on quality of life and of the efficacy and tolerability of lovastatin versus pravastatin. The Quality of Life Multicenter Group comparison of the effects on quality of life and of the efficacy and tolerability of lovastatin versus pravastatin. The Quality of Life Multicenter Group. Am J Cardiol. 1996;77:475–479. Abstract | Full-Text PDF (670 KB) | CrossRef

85. 85 Whyman MR, Fowkes FG, Kerracher EM, Gillespie IN, Lee AJ, Housley E, et al. Randomised controlled trial of percutaneous transluminal angioplasty for intermittent claudication. Eur J Vasc Endovasc Surg. 1996;12:167–172. Abstract | Full-Text PDF (627 KB) | CrossRef

86. 86 Dougherty CM, Dewhurst T, Nichol WP, Spertus J. 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;51:569–575. Abstract | Full Text | Full-Text PDF (184 KB) | CrossRef

87. 87 Ganiats TG, Palinkas LA, Kaplan RM. Communication (comparison of quality of well-being scale and functional status index in patients with atrial fibrillation). Med Care. 1992;30:958–964. MEDLINE | CrossRef

88. 88 Laupacis A, Bourne R, Rorabeck C, Feeny D, Wong C, Tugwell P, et al. The effect of elective total hip replacement on health-related quality of life. J Bone Joint Surg. 1993;75-A:1619–1626.

89. 89 Wiklund I, Holst J, Karlberg J, Mattsson LA, Samsioe G, Sandin K, et al. A new methodological approach to the evaluation of quality of life in postmenopausal women. Maturitas. 1992;14:211–224. Abstract | Full-Text PDF (967 KB) | CrossRef

90. 90 Alemi F, Stephens RC, Javalghi RG, Dyches H, Butts J, Ghadiri A. A randomized trial of a telecommunications network for pregnant women who use cocaine. Med Care. 1996;34:OS10–OS20. MEDLINE

91. 91 Booth DC, Deupree RH, Hultgren HN, DeMaria AN, Scott SM, Luchi RJ. The Investigators of Veterans Affairs Cooperative Study No.28. Quality of life after bypass surgery for unstable angina (5-year follow-up results of a veterans affairs cooperative study). Circulation. 1991;83:87–95. MEDLINE

92. 92 Trial Participants RITA. Coronary angioplasty versus coronary artery bypass surgery (the Randomised Intervention Treatment of Angina (RITA) trial). Lancet. 1993;341:573–580. Abstract | CrossRef

93. 93 Lawrence K, McWhinnie D, Goodwin A, Doll H, Gordon A, Gray A, et al. Randomised controlled trial of laparoscopic versus open repair of inguinal hernia (early results). BMJ. 1995;311:981–985.

94. 94 Kazis LE, Anderson JJ, Meenan RF. Effect sizes for interpreting changes in health status. Med Care. 1989;27:S178–S189. MEDLINE | CrossRef

95. 95 Liang MH, Fossel AH, Larson MG. Comparisons of five health status instruments for orthopedic evaluation. Med Care. 1990;28:632–642. MEDLINE | CrossRef

96. 96 Guyatt G, Walter S, Norman G. Measuring change over time (assessing the usefulness of evaluative instruments. Measuring change over time: assessing the usefulness of evaluative instruments). J Chronic Dis. 1987;40:171–178. MEDLINE | CrossRef

97. 97 Liang MH, Larson MG, Cullen KE, Schwartz JA. Comparative measurement efficiency and sensitivity of five health status instruments for arthritis research. Arthritis Rheum. 1985;28:542–547. MEDLINE | CrossRef

98. 98 Deyo RA, Centor RM. Assessing the responsiveness of functional scales to clinical change (an analogy to diagnostic test performance). J Chronic Dis. 1986;39:897–906. MEDLINE | CrossRef

99. 99 Stucki G, Liang MH, Fossel AH, Katz JN. Relative responsiveness of condition-specific and generic health status measures in degenerative lumbar spinal stenosis. J Clin Epidemiol. 1995;48:1369–1378. Abstract | Full-Text PDF (936 KB) | CrossRef

100. 100 Wright JG, Young NL. A comparison of different indices of responsiveness. J Clin Epidemiol. 1997;50:239–246. Abstract | Full-Text PDF (900 KB) | CrossRef

101. 101 Hagen KB, Smedstad LM, Uhlig T, Kvien TK. The responsiveness of health status measures in patients with rheumatoid arthritis (comparison of disease-specific and generic instruments). J Rheumatol. 1999;26:1474–1480.

102. 102 Harper R, Brazier JE, Waterhouse JC, Walters SJ, Jones NM, Howard P. Comparison of outcome measures for patients with chronic obstructive pulmonary disease (COPD) in an outpatient setting. Thorax. 1997;52:879–887. MEDLINE | CrossRef

a Department of Clinical Neurological Sciences and Epidemiology and Biostatistics, University of Western Ontario, University Campus, London Health Sciences Centre, 339 Windermere Road, London, Ontario, Canada N6A 5A5

b Department of Clinical Epidemiology and Biostatistics, McMaster University, 1200 Main Street West, L8N 3Z5, Hamilton, Ontario, Canada

c The London Health Sciences Centre, 339 Windermere Road, London, Ontario, N6A 5A5, Canada

d Department of Medicine, McMaster University, 1200 Main Street West, L8N 3Z5, Hamilton, Ontario, Canada

Corresponding author. Tel.: 519- 663-3984; fax: 519-663-3164.

PII: S0895-4356(02)00537-1

7 of 14