Original Article
Minimal Clinically Important Differences and Feasibility of Dyspnea-12 and the Multidimensional Dyspnea Profile in Cardiorespiratory Disease
Magnus P. Ekstr€om, MD, PhD, Hans Bornefalk, PhD, C. Magnus Sk€old, MD, PhD, Christer Janson, MD, PhD, Anders Blomberg, MD, PhD, Anna Bornefalk-Hermansson, PhD, Helena Igelstr€om, PhD, Jacob Sandberg, MD, and Josefin Sundh, MD, PhD
Faculty of Medicine (M.P.E., J.S.), Department of Clinical Sciences, Respiratory Medicine and Allergology, Lund University, Lund; Hans Bornefalk AB (H.B.), Vallentuna; Respiratory Medicine Unit (C.M.S.), Department of Medicine Solna and Center for Molecular Medicine, Karolinska Institutet, Stockholm; Department of Respiratory Medicine and Allergy (C.M.S.), Karolinska University Hospital Solna, Stockholm; Department of Medical Sciences, Respiratory, Allergy and Sleep Research (C.J.), Uppsala University, Uppsala; Department of Public Health and Clinical Medicine (A.B.), Section of Medicine, Ume a University, Umea; Uppsala Clinical Research Center (A.B.-H.), Uppsala University, Uppsala; Department of Neuroscience (H.I.), Uppsala University, Uppsala; and Faculty of Medicine and Health, Department of Respiratory Medicine (J.S.), € Orebro University, € Orebro, Sweden
Abstract
Context. Breathlessness is a cardinal symptom in cardiorespiratory disease and consists of multiple dimensions that can be measured using the instruments Dyspnea-12 (D12) and the Multidimensional Dyspnea Profile (MDP).
Objectives. The objective of the study is to determine the minimal clinically important differences (MCIDs) of all D12 and MDP summary and subdomain scores as well as the instruments’ feasibility in patients with cardiorespiratory disease.
Methods. Prospective multicenter cohort study of outpatients with diagnosed cardiorespiratory disease and breathlessness in daily life. D12 and MDP were assessed at baseline, after 30e90 minutes and two weeks. MCIDs were calculated using anchor- based and distributional methods for summary and subdomain scores. Feasibility was assessed as rate of missing data, help required, self-reported difficulty, and completion time.
Results. A total 182 outpatients (53.3% women) were included; main diagnoses were chronic obstructive pulmonary disease (COPD; 25%), asthma (21%), heart failure (19%), and idiopathic pulmonary fibrosis (19%). Anchor-based MCIDs were for D12 total score 2.83 (95% CI 1.99e3.66); D12 physical 1.81 (1.29e2.34); D12 affective 1.07 (0.64e1.49); MDP A1 unpleasantness 0.82 (0.56e1.08); MDP perception 4.63 (3.21e6.05), and MDP emotional score 2.37 (1.10e3.64). The estimates were consistent with small-to-moderate effect sizes using distributional analysis, and MCIDs were similar between COPD and non-COPD patients. The instruments were generally feasible and quick to use.
Conclusion. D12 and MDP are responsive to change and feasible for use for assessing multidimensional breathlessness in outpatients with cardiorespiratory disease. MCIDs were determined for use as endpoints in clinical trials. J Pain Symptom Manage 2020;60:968e975. Ó 2020 The Authors. Published by Elsevier Inc. on behalf of American Academy of Hospice and Palliative Medicine. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Key Words
Dyspnea, breathlessness, multidimensional, respiratory disease, heart disease, measurement
Address correspondence to: Magnus P. Ekstr€om, MD, PhD, Fac- ulty of Medicine, Department of Clinical Sciences, Respira- tory Medicine and Allergology, Lund University, SE-37185 Lund, Sweden. E-mail: pmekstrom@gmail.com
Accepted for publication: May 19, 2020.
Ó 2020 The Authors. Published by Elsevier Inc. on behalf of American Academy of Hospice and Palliative Medicine. This is an open access article under the CC BY license (http://
creativecommons.org/licenses/by/4.0/).
0885-3924/$ - see front matter
https://doi.org/10.1016/j.jpainsymman.2020.05.028Key Message
This study reports minimal clinically important dif- ferences for the instruments Dyspnea-12 and the Multidimensional Dyspnea Profile for measuring different dimensions of breathlessness in 182 outpa- tients with cardiorespiratory disease. The instruments were generally quick and feasible to use, and minimal clinically important differences are available for use in clinical trials.
Introduction
Chronic breathlessness is a disabling and distressing symptom in people with cardiorespiratory disease and is strongly associated with adverse health outcomes.
1,2Breathlessness is linked to reduced physical activity, worsening deconditioning, increased anxiety and depression, impaired quality of life, increased risk of hospitalization, and death.
1,3e5Breathlessness consists of different sensations that can vary in intensity.
1Several dimensions of breathless- ness can be differentiated by the individual: the expe- rienced intensity and unpleasantness, associated emotional response, and functional impact on the per- son’s life.
1Standardized multidimensional measure- ment is essential to quantify relevant aspects and to capture treatment effects in clinical trials as treat- ments may specifically affect certain breathlessness dimensions.
6,7Dyspnea-12 (D12)
8e12and the Multidimensional Dyspnea Profile (MDP)
13e15are recently deve- loped instruments to measure different dimen- sions of breathlessness. The instruments have been translated to and validated in a number of languages and across different cardiorespiratory diseases.
8e21However, knowledge on the minimal clinically important differences (MCIDs) and feasibility in patients in clinical practice is limited. Data on responsiveness and MCID are of fundamental importance to use the instruments as endpoints and to accurately capture effects of different breathlessness dimensions in therapeutic trials.
22For D12, an MCID of three points for the total score was suggested for patients with lung can- cer.
23MCIDs for MDP have not been published.
Similarly, data on feasibility of the instruments in clinical outpatients, including on ease and comple- tion time, are limited for D12
8,23and lacking for MDP.
15The aim of this article was to determine MCIDs for the D12 and MDP summary and subdomain scores and evaluate the instruments’ feasibility for measuring breathlessness in outpatients with cardiorespiratory disease.
Methods
Design and Population
This was a prospective, multicenter, cohort study of Swedish outpatients with breathlessness, and cardiore- spiratory disease diagnosed in accordance with cur- rent guidelines.
24e27Patients were recruited at five outpatient clinics (Blekinge Hospital, Karlskrona; Kar- olinska University Hospital in Solna, Stockholm; Uni- versity Hospital, Ume a; Uppsala University Hospital, Uppsala; and € Orebro University Hospital, € Orebro).
The present database was used to validate the factor structure, reliability, and concurrent validity of the Swedish D12 and MDP,
20,21but the present analyses have not been previously reported. This study was con- ducted in accordance with the amended Declaration of Helsinki. Written informed consent was obtained from all patients, and the protocol was approved by the regional ethical review board at Lund University (DNr: 2016/16).
Inclusion criteria were as follows: aged 18 years or older; a recorded chronic respiratory and/or cardiac disease; self-reported breathlessness during daily life defined as an answer yes to the question ‘‘Did you experience any breathlessness during the last two weeks?’’; and ability to provide written informed con- sent to participate in the study. Exclusion criteria were inability to write or understand Swedish adequately to participate; cognitive or other inability to participate in the study; and expected survival less than three months as judged by the responsible clinician.
Assessments
The D12
28and MDP
18were completed by the pa- tient at baseline, after 30e90 minutes, and after two weeks, as described elsewhere.
20,21To facilitate unbi- ased analysis, the order of D12 and MDP (used in all questionnaires) was randomized for each patient, us- ing questionnaires in opaque envelopes. All other as- sessments were similar for the patients. The period for all self-reported measures used in the analysis was during the previous two weeks.
At the baseline visit, demographics and physiolog- ical data including spirometry and ultrasound cardiog- raphy were measured or obtained from the medical records by the responsible physician.
20,21Patients self-reported the modified Medical Research Council Breathlessness Scale.
The patient completed a questionnaire after 30e
90 minutes (at the clinical visit) and a postal question-
naire after two weeks, which included the D12 and
MDP (in the patient’s randomized order) and the
global impression of change (GIC), which assessed
the change in breathlessness from the first assessment
on a seven-point ordinal scale (1 ¼ very much better;
2 ¼ much better; 3 ¼ minimally better; 4 ¼ no change;
5 ¼ minimally worse; 6 ¼ much worse; and 7 ¼ very much worse).
29Feasibility of D12 and MDP was assessed directly after each instrument at all assessments using the questions:
How difficult was it to understand and complete? (not at all; a little; moderate; and very difficult); Did you need help to understand and complete? (yes/no);
and How long time did it take to complete? (less than five minutes; five to 10 minutes; or more than 10 mi- nutes). The answer to the second question (need of help) was changed to yes by the study nurse or physician for patients who had responded no but according to the opinion of the staff actually required help during completion (clinical visit questionnaires only).
Statistical Analyses
MCIDs were determined using the anchor-based method in accordance with guidelines.
22,30Anchor- based methods determine the mean change in the score of interest (breathlessness) over time for people who experienced a change in another relevant and meaningful variable (anchor).
22,30We calculated the MCID as the mean change for each breathlessness score associated with a one unit’s change in the breathlessness GIC from baseline, using linear regres- sion. The estimates were presented with 95% CIs.
30The assumption of a similar change in the breathless- ness scores for each change along the GIC scale was found to be valid; F-tests of the linear regression models were all statistically significant at P < 0.001, and residuals were independent and normally distrib- uted as assumed in the model. Each GIC step was asso- ciated with a similar change in the breathlessness scores (Appendix Fig. 1) with mean values very close to the fitted mean change in the model with no statis- tically significant deviations.
MCIDs were additionally quantified using the distribution-based method with the proposed cutoffs for a small change (SD .25 of the score at baseline) and moderate change (SD .5).
22The methodology was similar to that used to calculate MCIDs for unidi- mensional scores of intensity and unpleasantness of chronic breathlessness
31,32and for the MCID of D12 in patients with lung cancer.
23Analyses were conducted for the D12 total score (range 0e36), physical subdomain (range 0e21) and affective subdomain scores (range 0e15), and the MDP A1 overall unpleasantness (range 0e10), percep- tion (sum of the A1 score and five intensities for the sensory qualities [SQs]; range 0e60), and emotional response scores (sum of intensities for the five emotional responses; range 0e50).
13,15,18For all these items, higher scores indicate worse breathlessness.
The D12 total score was imputed for 14 patients in accordance with Yorke et al.
33No other data were
imputed. D12 and MDP scores were plotted as recom- mended by the developers.
8,34,35MCIDs were esti- mated for all patients and separately for patients with chronic obstructive pulmonary disease (COPD) and with other main causes of breathlessness (non- COPD).
Feasibility was evaluated by tabulating the responses to the three feasibility questions for D12 and MDP (as- sessed after completing the instrument) at each time point and by the proportion of missing data in the in- strument scores.
Sample size was set to 180 enrolled patients based on previous validation studies.
8,13e15,34,36Statistical significance was defined as a two-tailed P < 0.05. Ana- lyses were conducted by biostatisticians with Matlab R2018b (Mathworks, Inc., Natick, MA).
Results
At five outpatient clinics, 182 patients were enrolled between August 29, 2016 and December 23, 2017. Pa- tient characteristics were as follows: mean age 68.6 (SD 13.8; range 19e91) years; 53.3% were women; main di- agnoses were COPD (25%), asthma (21%), heart fail- ure (19%), and idiopathic pulmonary fibrosis (19%);
and 70% had a modified Medical Research Council breathlessness score $2. Further patient characteris- tics have been detailed elsewhere.
21D12 and MDP scores at baseline are shown in Fig. 1.
Follow-up data were available for 179 patients (98%) after 30e90 minutes, for 162 patients (89%) after two weeks, and baseline characteristics were similar for people with and without follow-up data. The actual time between the first visit and the subsequent follow-up (projected at two weeks) was a mean 17.2 (SD 7.7; range 3e58) days. D12 and MDP scores at baseline and follow-up as well as mean change is shown in Table 1.
MCIDs for D12 and MDP summary and subdomain scores are shown in Table 2. At two weeks, data on the instruments and GIC were available for 149 patients;
30% reported improved breathlessness (GIC <4),
22% worse breathlessness (GIC >4), and 48% re-
ported unchanged breathlessness (GIC ¼ 4). One
step’s change on the GIC scale was associated with a
similar mean change in the D12 and MDP scores,
across the range of the GIC (Appendix Fig. 1). Using
the anchor-based method, the MCID was 2.83 (95% CI
1.99e3.66) points for the D12 total score and 0.82
(95% CI 0.56e1.08) for the MDP A1 unpleasantness
score. The MCIDs were consistent with small-to-
moderate effect sizes in the distribution-based analysis
(Table 2). When comparing patients with COPD and
other conditions, the CIs were wider because of
reduced power in the subgroups, but the MCID point
estimates were similar between patients with COPD and non-COPD (Table 2).
Feasibility for the instruments is shown in Table 3.
Difficulty to understand or complete the question- naires was rated as none or little by 78% for D12 and 69% for MDP at the first assessment (McNemar’s test, P ¼ 0.063 for patients who rated both). More pa- tients needed help to complete the whole MDP than D12 (36% vs. 19%; McNemar’s test, P < 0.001). For
MDP, most questions from patients pertained to how
to score their breathlessness during the last two weeks
(as breathlessness varied and was related to different
situations or activities) and how to rate of the presence
and best match for SQs; and some patients had to be
reminded to rate the best matching SQ (personal
communication from staff). D12 was completed in
less than five minutes by most of them (74.7%),
whereas completing the entire MDP took longer, but
Fig. 1. Dyspnea-12 and MDP scores in 182 outpatients with cardiorespiratory disease. a) Dyspnea-12 total score (maximum 36)
and physical subdomain score (maximum 21) and affective subdomain score (maximum 15); b) MDP overall A1 unpleasant-
ness score, SQ intensities, choices of best matching SQ, and A2 emotional response intensities. Mean scores SD. Best match-
ing SQ was missing for 28 (15.4%) patients. MDP ¼ Multidimensional Dyspnea Profile; SQ ¼ sensory quality.
80% completed the MDP within 10 minutes (Table 3).
Feasibility increased with repeated assessments for all three questions. The rate of missing data was higher for MDP than D12 (Table 3) and higher for the MDP perception (SQ intensities) than for the emotional responses.
Discussion
The present study reports MCIDs for all D12 and MDP summary and subdomain scores and shows that the instruments are feasible for multidimensional measurement of breathlessness in patients with cardio- respiratory disease.
The MCIDs were consistent between the anchor- based and distribution-based analyses. The present MCID for MDP A1 unpleasantness score of 0.82 is consistent with the published MCID for chronic breathlessness intensity of 0.89 (both measured on a 0e10 Numerical Rating Scale)
31and is similar to recently published MCIDs for unidimensional scores of breathlessness intensity and unpleasantness.
32The present MCID for the D12 total score is consistent
with the previously reported MCID of about three points in lung cancer,
23which supports the validity of the current analysis. The consistency of the MCIDs across patients with COPD and other conditions sug- gests that the MCIDs for measuring dimensions of breathlessness in outpatients may be similar across cardiorespiratory diseases.
These findings confirm previous limited data that the D12 and MDP instruments are feasible for use in patients in clinical care.
23,33The D12, a short and one-page instrument of 12 items, was perceived as easy to use by the patients and generally took only a few minutes to complete. Standardized procedures of measurement and the written instructions included in the instruments are important as a substantial pro- portion of patients required help to complete the questionnairedabout 20% for D12 and about one- third of patients for MDP. However, the need of help compares favorably to that reported for frequently used questionnaires such as 36.0% for the Clinical COPD Questionnaire and 53.9% for COPD Assess- ment Test.
37Most help needed for MDP related to the two-step process of selecting the presence and best match for SQs.
Table 1
Breathlessness Scores at Baseline and Follow-Up
Item Mean at Baseline Mean After Two Weeks Change From Baseline
D12
Total score 15.83 (9.1) 13.66 (9.7) 2.12 (6.9)
Physical subdomain 9.79 (5.3) 8.68 (5.7) 1.12 (4.4)
Affective subdomain 6.09 (4.4) 5.01 (4.3) 1.12 (3.4)
MDP
A1 unpleasantness 4.94 (2.5) 4.25 (2.4) 0.72 (2.1)
Perception 24.30 (14.9) 20.59 (15.4) 3.97 (11.2)
Emotional response 15.99 (13.2) 13.20 (12.9) 2.23 (9.6)
D12¼ Dyspnea-12; MDP ¼ Multidimensional Dyspnea Profile.
Data presented as means (SD). Change from baseline is the value at two weeks minus the baseline value for each participant.
Table 2
MCIDs for D12 and MDP
Item
All Patients COPD Patients Non-COPD Patients
Distribution-Based Method Anchor-Based Method
Small Change (SD .25)
Moderate Change
(SD .5) MCID (95% CI)
D12
Total score 2.25; n ¼ 167 4.5; n ¼ 167 2.83 (1.99e3.66); n ¼ 133 2.68 (1.16e4.20); n ¼ 27 2.97 (1.94e4.00); n ¼ 106 Physical subdomain 1.3; n ¼ 174 2.6; n ¼ 174 1.81 (1.29e2.34); n ¼ 142 1.36 (0.43e2.28); n ¼ 30 1.96 (1.30e2.61); n ¼ 112 Affective subdomain 1.1; n ¼ 174 2.2; n ¼ 174 1.07 (0.64e1.49); n ¼ 144 1.24 (0.40e2.09); n ¼ 32 1.06 (0.54e1.57); n ¼ 112 MDP
A1 unpleasantness 0.65; n ¼ 157 1.3; n ¼ 157 0.82 (0.56e1.08); n ¼ 149 0.58 (0.06e1.10); n ¼ 35 0.88 (0.56e1.19); n ¼ 114 Perception 3.7; n ¼ 146 7.4; n ¼ 146 4.63 (3.21e6.05); n ¼ 123 3.02 (0.06e5.99); n ¼ 28 5.09 (3.39e6.80); n ¼ 95 Emotional response 3.3; n ¼ 158 6.6; n ¼ 158 2.37 (1.10e3.64); n ¼ 142 2.07 (0.71 to 4.86); n ¼ 37 2.31 (0.83e3.79); n ¼ 105
MCIDs¼ minimal clinically important differences; D12 ¼ Dyspnea-12; MDP ¼ Multidimensional Dyspnea Profile; COPD ¼ chronic obstructive pulmonary dis- ease.MCID using the anchor-based method was analyzed using linear regression as the mean change in the D12 or MDP score associated with one unit’s change from baseline in the global impression of change of breathlessness at two weeks. MCIDs were analyzed for all patients and separately in patients with COPD and non- COPD as main cause of breathlessness. The distribution-based estimates used commonly used cutoffs for a small and moderate effect size.
Importantly, the D12 and MDP were developed with different intentions. D12 was developed to provide global ratings, whereas MDP was designed to provide separate ratings for three different dimensions (A1 overall unpleasantness, perception, and emotional response subdomain) as well as separate subscores, de- pending on the aim of the measurement. The required time to complete individual subitems of MDP is likely short compared with in the present study where feasibility and time pertain to completing the whole instrument.
A strength of the present study is that it included a relatively large sample of outpatients in clinical prac- tice across a range of relevant cardiorespiratory diag- noses. Completeness of data was high both for baseline and follow-up at two weeks. Recommended standard methods, similar to those used in previous studies, were used to facilitate comparison.
17,19,22,34Limitations include that subgroups were too small for detailed evaluation of any differences across spe- cific cardiorespiratory conditions. The consistency be- tween patients with COPD and other conditions in the present analysis and with previous reports across advanced life-limiting diseases
31,32and lung cancer supports the validity of the present MCIDs. The pre- sent findings pertain to the patient’s experience dur- ing the previous two weeks. Interestingly, MCIDs were similar to those previously reported for worst breathlessness during the last two weeks
23as well as for the intensity and unpleasantness of current breath- lessness and the recalled best, worst, and average breathlessness during the last 24 hours.
31,32Because patients were stable outpatients with diagnosed cardio- respiratory disease and breathlessness for at least two weeks, the present findings pertain mainly to chronic breathlessness.
The present findings have important implications for clinical care and research. For the clinician, D12 and MDP are validated and feasible tools to evaluate and follow-up relevant aspects of the cardinal symp- tom of breathlessness in cardiorespiratory disease.
They are validated for self-report by patients and should be used using the validated forms and instruc- tions. Research should evaluate the usefulness of the multidimensional instruments in further patient groups and settings, including advanced illness, pallia- tive care, as treatment outcome and in specific patient groups, as well as the instruments’ MCIDs for acute and acute-on-chronic breathlessness.
In conclusion, D12 and MDP are responsive and feasible for multidimensional measurement of breath- lessness in cardiorespiratory disease. MCIDs were es- tablished for the instruments’ summary and subdomain scores for use as endpoints in clinical trials.
Disclosures and Acknowledgments
The authors thank all research nurses who dedi- cated their time and work to conduct this study: nurses at the respiratory outpatient clinic in Karlskrona; Lisa Carlson, Karolinska University Hospital Solna, Stock- holm; Annika Johansson and Frida Holmstr€om, Uni- versity Hospital, Ume a; Karin Johansson, € Orebro University Hospital; and Jonatan Blomqvist in Lund for help with data input and quality checking. They thank Prof. Janelle York and Prof. Robert B. Banzett for valuable input on the project and interpretation of the instruments. The authors extend their warm thanks to all patients who made this research possible.
The authors declare no conflicts of interest.
Table 3
Feasibility of D12 and the MDP in 182 Outpatients With Cardiorespiratory Disease
Questions and Options
Baseline After 30e90 Minutes After Two Weeks
D12 MDP D12 MDP D12 MDP
Difficulty to understand and complete?
Not at all 45 32 49 40 55 41
A little 32 36 27 29 21 32
Moderate 16 19 12 14 15 13
Very difficult 4 7 3 2 3 4
Missing 3 6 8 16 6 9
Needed help to complete 19 36 14 14 6 4
Missing 6 7 8 16 7 14
Completion time
Less than five minutes 75 39 72 58 67 46
Five to 10 minutes 18 42 16 22 20 32
>10 minutes 4 13 3 3 5 10
Missing 3 7 9 17 7 12
D12¼ Dyspnea-12; MDP ¼ Multidimensional Dyspnea Profile.
Data presented as percentages. Feasibility was rated by the patient directly after the instrument and pertain to completing the whole D12 (one page) and MDP (four pages). Baseline and reassessment after 30e90 minutes was during a clinical visit, and follow-up at two weeks was by postal questionnaire completed by the patient in the home. The order of the D12 and MDP in the questionnaire was randomized but similar at the assessments for each patient.