Health-related quality of life (HRQL) is an important outcome measure in patients with chronic respiratory failure. These patients suffer physical, psychological, and social limitations in daily life due to severely progressive underlying diseases, and cure is not possible in the vast majority of patients. The St. George’s Respiratory Questionnaire (SGRQ) [
Thus, for HRQL assessment of these patients, the Maugeri Respiratory Failure (MRF) Questionnaire [
We recruited 56 stable outpatients receiving NIV due to chronic hypercapnic respiratory failure as described in detail previously [
HRQL was evaluated using the Japanese versions of the SGRQ (respiratory-specific questionnaire) [
Dyspnea during activities of daily living was evaluated by the modified Medical Research Council (mMRC) dyspnea scale [
Results are expressed as mean ± standard deviation. Spearman rank correlation tests were performed to analyze the relationships between two sets of data. Stepwise multiple regression analyses were used to identify those variables that could best predict the HRQL scores using variables that were significant between two sets of data.
Baseline characteristics of 56 patients receiving long-term NIV due to chronic hypercapnic respiratory failure are shown in Table
Baseline characteristics of 56 patients using noninvasive ventilation.
Characteristics | Total ( |
COPD ( |
Tb ( |
---|---|---|---|
Age, years | 73 ± 8 | 70 ± 8 | 76 ± 7 |
BMI, kg/m2 | 20.5 ± 3.7 | 20.6 ± 3.8 | 20.4 ± 3.8 |
Cumulative smoking, pack-years | 29 ± 34 | 50 ± 31 | 12 ± 26 |
FEV1, % predicted | 29.9 ± 10.6 | 25.7 ± 10.3 | 33.0 ± 10.3 |
FVC, % predicted | 45.7 ± 18.7 | 56.1 ± 22.7 | 38.1 ± 11.0 |
PaO2, kPa | 10.7 ± 2.2 | 10.2 ± 2.2 | 11.1 ± 2.2 |
PaCO2, kPa | 7.7 ± 1.4 | 7.5 ± 1.5 | 7.9 ± 1.4 |
Base excess, mmol/L | 5.4 ± 3.4 | 4.8 ± 3.0 | 6.0 ± 3.6 |
mMRC dyspnea | 2.3 ± 1.1 | 2.3 ± 1.0 | 2.3 ± 1.1 |
HADS anxiety | 5.6 ± 3.7 | 5.5 ± 3.9 | 5.7 ± 3.6 |
HADS depression | 7.4 ± 3.8 | 7.3 ± 3.5 | 7.5 ± 4.0 |
SRI respiratory complaints | 62.3 ± 21.1 | 61.0 ± 21.2 | 62.5 ± 21.5 |
SRI physical functioning | 52.1 ± 19.2 | 55.3 ± 17.3 | 50.5 ± 20.7 |
SRI attendant symptoms and sleep | 61.3 ± 16.8 | 58.4 ± 18.1 | 63.0 ± 16.0 |
SRI social relationships | 53.5 ± 17.6 | 55.6 ± 15.4 | 51.8 ± 20.0 |
SRI anxiety | 54.2 ± 24.3 | 55.0 ± 24.0 | 52.7 ± 25.4 |
SRI psychological well-being | 54.3 ± 17.0 | 55.1 ± 18.0 | 53.8 ± 17.0 |
SRI social functioning | 54.5 ± 21.1 | 56.1 ± 18.9 | 54.0 ± 23.3 |
SRI summary | 56.0 ± 15.3 | 56.6 ± 14.7 | 55.5 ± 16.4 |
MRF-26 daily activities | 43.2 ± 30.5 | 39.8 ± 28.4 | 44.9 ± 31.9 |
MRF-26 perceived disability | 52.7 ± 28.3 | 48.4 ± 26.9 | 54.3 ± 29.5 |
MRF-26 total | 48.0 ± 26.7 | 43.8 ± 22.8 | 49.9 ± 29.1 |
SGRQ symptoms | 50.4 ± 22.1 | 50.9 ± 23.6 | 49.6 ± 22.1 |
SGRQ activities | 74.1 ± 17.4 | 75.7 ± 16.4 | 72.8 ± 18.5 |
SGRQ impacts | 44.6 ± 22.4 | 45.3 ± 23.4 | 43.0 ± 22.2 |
SGRQ total | 54.5 ± 18.5 | 55.4 ± 19.1 | 53.1 ± 18.5 |
Theoretical scores range from 0 to 4 in the mMRC, 0 to 21 in the HADS, and 0 to 100 in the SRI, MRF-26, and SGRQ. For the SRI, minimum scores indicate poor status, and for the mMRC, HADS, MRF-26, and SGRQ, maximum scores indicate poor status; BMI = body mass index; COPD = chronic obstructive pulmonary disease; Tb = pulmonary tuberculosis sequelae; FEV1 = forced expiratory volume in one second; FVC = forced vital capacity; PaO2 = arterial oxygen pressure; PaCO2 = arterial carbon dioxide pressure; mMRC = modified Medical Research Council; HADS = Hospital Anxiety and Depression Scale; SRI = Severe Respiratory Insufficiency; MRF = Maugeri Respiratory Failure; SGRQ = St. George’s Respiratory Questionnaire.
Spearman’s rank correlation coefficients (
Correlations between HRQL questionnaires.
SGRQ | ||||
---|---|---|---|---|
Symptoms | Activities | Impacts | Total | |
SRI | ||||
Respiratory complaints |
|
|
|
|
Physical functioning |
|
|
|
|
Attendant symptoms and sleep | — |
|
|
|
Social relationships |
|
|
|
|
Anxiety |
|
|
|
|
Psychological well-being |
|
|
|
|
Social functioning |
|
|
|
|
Summary score |
|
|
|
|
MRF-26 | ||||
Daily activities | 0.50 |
|
|
0.80 |
Perceived disability | 0.46 | 0.48 | 0.69 | 0.69 |
Total | 0.51 | 0.73 | 0.78 | 0.81 |
Missing values (—) indicate no significant relationships (
In all 3 instruments, SRI, MRF-26, and SGRQ, there was a significant relationship with mMRC dyspnea and HADS anxiety and depression (Table
Correlations between HRQL questionnaires and physiological measures.
SRI | MRF-26 | SGRQ | |||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
RC | PF | AS | SR | AX | WB | SF | SS | Dai | Per | Tot | Sym | Act | Imp | Tot | |
Age | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — |
BMI | — | 0.31 | — | 0.31 | — | — | — | 0.29 | −0.38 | — | −0.33 | — | — | — | — |
Smoking | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — |
FEV1 | 0.30 | — | — | — | — | — | — | — | −0.35 | — | — | −0.31 | −0.40 | — | −0.31 |
FVC | 0.39 | — | — | — | — | 0.28 | — | 0.30 | −0.49 | −0.38 | −0.47 | — | −0.42 | — | −0.33 |
PaO2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — |
PaCO2 | −0.41 | — | — | −0.28 | — | — | −0.38 | −0.38 | 0.38 | — | 0.33 | — | — | — | — |
BE | — | — | — | — | — | — | −0.30 | −0.28 | 0.34 | — | 0.27 | — | — | — | — |
Dyspnea | −0.60 | −0.45 | −0.29 | −0.46 | −0.31 | −0.42 | −0.45 | −0.56 | 0.72 | 0.46 | 0.51 | 0.59 | 0.72 | 0.62 | 0.69 |
Anxiety | −0.60 | −0.34 | −0.54 | −0.50 | −0.51 | −0.67 | −0.68 | −0.69 | 0.46 | 0.57 | 0.73 | 0.39 | 0.36 | 0.51 | 0.51 |
Depression | −0.60 | −0.48 | −0.50 | −0.44 | −0.60 | −0.56 | −0.53 | −0.70 | 0.58 | 0.62 | 0.78 | 0.37 | 0.45 | 0.59 | 0.59 |
Missing values (—) indicate no significant relationships (
BMI was partly weakly but significantly related to SRI-PF, SRI-SR, SRI-SS, and daily activities and the total of the MRF-26 (
Multiple regression analyses were performed to investigate factors explaining summary or total HRQL scores using 2 models (excluding or including patient-reported measurements of dyspnea and depression (Models I and II, resp.)) (Table
Results of multiple regression analyses.
SRI summary | MRF-26 total | SGRQ total | ||||
---|---|---|---|---|---|---|
|
|
|
|
|
| |
|
||||||
BMI, kg/m2 | — | — | −0.27 | 0.09 | ||
FEV1, % predicted | −0.31 | 0.09 | ||||
FVC, % predicted | — | — | −0.38 | 0.16 | — | — |
PaCO2, kPa | −0.31 | 0.15 | — | — | ||
BE, mmol/L | — | — | — | — | ||
|
||||||
Cumulative |
0.15 | 0.25 | 0.09 | |||
|
||||||
|
||||||
BMI, kg/m2 | — | — | −0.17 | 0.05 | ||
FEV1, % predicted | — | — | ||||
FVC, % predicted | — | — | — | — | — | — |
PaCO2, kPa | — | — | — | — | ||
BE, mmol/L | — | — | — | — | ||
mMRC dyspnea | −0.45 | 0.27 | 0.55 | 0.37 | 0.63 | 0.46 |
HADS depression | −0.49 | 0.34 | 0.37 | 0.24 | 0.29 | 0.17 |
|
||||||
Cumulative |
0.61 | 0.67 | 0.63 |
All values represent coefficient of determination (
We cross-sectionally evaluated and compared two respiratory failure-specific questionnaires, the SRI and MRF-26, in addition to the SGRQ. We found that both were valid in patients receiving NIV due to chronic hypercapnic respiratory failure. Dyspnea and psychological status were the major determinants in the multiple regression analyses. There were some differences in their relationships with physiological measurements.
In a recent clinical trial, disease-specific questionnaires were shown to be capable of detecting beneficial changes in HRQL following NIV treatments in patients with severe COPD apart from the generic questionnaire [
We also examined to what extent HRQL scores were correlated with relevant clinical measurements. The mMRC and HADS were significantly related to all domains of the SRI, MRF-26, and SGRQ. Thus, the SRI and MRF-26 reflected physical and psychological impairments in health similarly to the well-established SGRQ, although the MRF-26 had no specific psychological domains. In contrast, relationships between HRQL and physiological measurements were not strong, indicating that HRQL is somewhat independent of physiological parameters, and cannot be estimated from pulmonary function or blood gases alone. Regarding pulmonary function, FEV1 and FVC were significantly but weakly related to only some domains of the HRQL like SRI-RC, daily activities in the MRF-26, and activity in the SGRQ, but most relationships were nonsignificant. In addition, arterial blood gas measurements such as PaCO2 and BE were significantly weakly related to some domains of the SRI and MRF-26, but not the SGRQ. This may be connected to the notion that those instruments were originally developed specifically for severely affected patients with arterial blood gas abnormalities, unlike the SGRQ.
This is the first study comparing the determinants of the SRI, MRF-26, and SGRQ in patients with chronic respiratory failure receiving NIV. Multiple regression analyses showed that dyspnea and depression were their main determinants, explaining over 60% of the variance. In studies comparing the SRI and MRF-28 (the earlier version of the MRF-26) in patients with severe COPD [
In multiple regression analyses, where explanatory factors were limited to BMI, pulmonary function, and arterial blood gas, the MRF-26 was explained by 25% of the variance, which was more than in the SRI (15%) and SGRQ (9%). This indicates that the MRF-26 placed greater emphasis on daily restrictions due to physiological impairments than the SRI. In a longitudinal analysis in patients receiving NIV, although both the SRI and MRF-26 strongly predicted mortality in univariate analyses, the SRI was predictive independently of BMI, FVC, and PaCO2, but the MRF-26 was not [
The SRI contains 49 items divided into 7 subscales, and each is scored by a 5-point Likert scale while the MRF-26 contains 26 items divided into only 2 domains, and each is measured with yes/no. Thus, structurally, the SRI would measure diversified health impairments more discriminatively than the MRF-26, although it may take more time to be administered. While the MRF was developed for patients with chronic respiratory failure irrespective of NIV, the SRI was primarily developed for patients with home NIV, and it recently was validated in patients with COPD receiving long-term oxygen therapy [
As another limitation of the present study, the sample size was relatively small. Unfortunately, we could not collect a larger number of such severely ill patients who needed long-term NIV. However, we consider that an increase in the sample size would not greatly change the main results regarding the characteristics of the questionnaires assessed by their relationships with other clinical measures.
We demonstrated that both the SRI and MRF-26 were reliable questionnaires in patients with chronic hypercapnic respiratory failure receiving NIV. Dyspnea and psychological status were similarly their main determinants while their relationships with physiological measurements were weaker. The SRI covers more psychological health impairments than the MRF.
The authors declare that no funding sources influenced the preparation of any part of this manuscript including collection, interpretation, and presentation of the data.
The authors declare that they have no conflicts of interest.
This work was supported in part by a grant to the Respiratory Failure Research Group from the Ministry of Health, Labour and Welfare and in part by Health, Labour and Welfare Sciences Research Grants, Research on Region Medical. Department of Respiratory Care and Sleep Control Medicine, Graduate School of Medicine, Kyoto University, is funded by endowments from Philips-Respironics, Teijin Pharma, Fukuda Denshi, and Fukuda Lifetec Keiji to Kyoto University.