Predictors of response to pulmonary rehabilitation in stable chronic obstructive pulmonary disease patients: A prospective cohort study.

Context: Pulmonary rehabilitation (PR) has become a standard of care in the management of chronic obstructive pulmonary disease (COPD). However, a significant proportion of the patients do not show benefit after the PR program. Aims: The study was planned to find different patient- and/or disease-related factors that may predict response to PR in stable COPD. Subjects and
Methods: A total of 102 stable COPD patients were prospectively enrolled. Baseline evaluation and investigations, including spirometry, arterial blood gas analysis, and bone mineral density assessment, were done. Thereafter, all patients underwent an 8-week comprehensive outpatient PR program that consisted of exercise training, education, nutritional, and psychological counseling. The response to PR was dichotomously (yes/no) defined by the combined improvement in exercise tolerance (6-min walk distance [6MWD] +54 m) and quality of life (St. George's Respiratory Questionnaire [SGRQ] score - 4 points) measured before and after the program. Thereafter, different predefined factors were analyzed for their possible association with the PR response.
Results: A total of 80 patients (78.4%) completed the PR program and were subjected for analysis. Out of them, 42 (52.5%) showed improvement in both 6MWD and SGRQ score (46 in 6MWD and 54 in SGRQ score). After application of multivariate logistic regression analysis, forced expiratory volume in 1 s <50% predicted (odds ratio [OR]: 2.9; 95% confidence interval [CI]: 1.1-8.3; P = 0.04) and osteoporosis (OR: 0.26; 95% CI: 0.13-0.53; P < 0.001) were found as independent factors predicting PR efficacy. Conclusions: Poor baseline lung function predicts a short-term improvement in exercise capacity and quality of life in COPD patients, whereas osteoporosis is a negative predictor of PR response. Active search for these factors may help in better patient selection, thus leading to improved outcome after PR.

Introduction

Chronic obstructive pulmonary disease (COPD) is a progressive airway disease affecting 8.4%–15% of the adult population worldwide.[1] India too carries a high burden of the disease with prevalence estimates ranging from 2% to 22% in men and 1.2%–19% in women.[2] Being an irreversible disease, the primary goal of COPD treatment is the improvement in symptoms and prevent/manage exacerbations.

Pulmonary rehabilitation (PR) is a nonpharmacological intervention that has been shown to improve dyspnea, exercise capacity, and quality of life in COPD patients.[34] It involves patient assessment-guided therapies that include exercise training, education, and behavior change. With a strong favorable evidence, PR is now the standard of care in COPD patients.[5] However, in spite of a well-organized PR program, a significant proportion of patients who complete the program do not attain expected benefits in terms of health-related quality of life (HRQoL) and/or exercise capacity.[678] Hence, it becomes crucial to identify the patients who are likely to achieve maximum benefit out of the PR program.

Previous studies have tried to evaluate different clinical-, physiological-, and disease-related factors that could predict PR efficacy. However, these have yielded inconsistent results with a variety of factors such as age,[6] dyspnea grade,[9] partial pressure of oxygen (PaO2) in arterial blood,[7] forced expiratory volume in 1 s (FEV1),[8] and body mass index (BMI);[7] comorbidities such as cardiac disease,[10] osteoporosis,[11] and psychiatric disturbances[6] found associated in different studies. Differences in study designs and/or PR programs used (hospital vs. home based) in these studies might have resulted in the discordant results.

Most of the evidence on PR use in COPD has been generated from Western studies with scarce data from India.[12] The results from these studies may not be extrapolated to Indian patients due to differences in the genetic makeup and the environmental influences that affect the disease course. Hence, the study was planned to generate data on the performance of PR in stable COPD patients and to evaluate factors that can predict short-term improvement after the intervention.

Subjects and Methods

This was a prospective observational study conducted in Government Medical College and Hospital, Chandigarh, between November 2015 and September 2017. Stable COPD patients on outpatient treatment who had not yet received PR were consecutively enrolled. COPD was diagnosed and staged as per the recent Global Initiative for Chronic Obstructive Lung Disease (GOLD) guidelines.[13] Based on the results from previous studies[78] and anticipating a minimal clinically important differences (MCID) size improvement in exercise capacity in 50% of patients, a sample size of 78 patients was required to attain 80% power at 5% level of significance. Patients with acute exacerbation in the preceding 6 weeks, unstable cardiac disease, and impaired mobility due to severe musculoskeletal/joint disorders were excluded. Informed written consent was obtained from all participants. The study was approved by the Institutional Ethics Committee.

Methodology

Demographic and clinical data were recorded, including the modified medical research council (mMRC) dyspnea grading.[14] Thereafter, each patient underwent a baseline spirometry as per standard American Thoracic Society guidelines,[15] using a spirometer make-Spiro Analyzer, RMS HELIOS-401. Staging of disease was done using postbronchodilator FEV1% predicted value, as per GOLD guidelines.[13] Arterial blood gas analysis was done on room air to measure the PaO2. Bone mineral density (BMD) was measured using dual energy X-ray absorptiometry scan. The results were interpreted based on the WHO guidelines according to which T score ≤−2.5 is diagnostic of osteoporosis.[16] The psychological component of ill health was also screened in all patients using a self-administered screening tool and General Health Questionnaire-12.[17] This tool is used to assess somatic symptoms, anxiety, insomnia, social dysfunction, and depression. According to it, patients having score >3 are likely to have a psychiatric problem and hence require an expert psychiatrist referral.

Pulmonary rehabilitation

After the initial assessment, patients underwent 8 weeks of comprehensive PR program that included exercise training, education, nutritional, and psychological counseling, if needed. The program consisted of two sessions of 1–1.5 h each, per week at the hospital-based PR center along with home-based exercise on the remaining days, that was in accordance with the previous studies.[67] Each supervised session consisted of exercise training which included four basic components, namely, (i) lower extremity aerobic exercise, (ii) strength and endurance training, (iii) upper extremity aerobic exercise, and (iv) ventilatory muscle training. The exercise was followed by education sessions covering issues such as medication use and adherence, nutrition, and coping strategies. The intensity of exercise was gradually increased as per response judged by the supervising investigator. Supplemental oxygen was given to the patients during PR whenever required. Patients were considered to have accomplished the PR program if they completed at least 75% or more of the sessions.[6]

Outcome parameters

Outcome after PR was measured in terms of functional exercise capacity and HRQoL using 6-min walk distance (6MWD) and St. George's Respiratory Questionnaire (SGRQ), respectively, performed before and after the PR program. A patient was said to have PR response if he/she had improvement in both outcome parameters above their MCID.

Six-min walk test was performed along a 30 m long flat, straight, enclosed corridor with a hard surface following standard guidelines.[18] The distance covered in meters at the end of 6 min was recorded as 6MWD. An increase of 54 meters or more was considered as MCID for 6MWD.[19] Patients also completed SGRQ that consisted of two parts, with eight questions each. Scores were expressed as a percentage of overall impairment where 100 represented the worst possible health status and 0 indicated the best. A change of four units or more in the overall score was considered as MCID for SGRQ score.[20]

Results were analyzed as number (%) of patients showing PR response. On the basis of the existing literature[467891021] and author's experience, certain patient- and disease-related factors, i.e., age, gender, BMI, mMRC dyspnea score, FEV1, PaO2, baseline 6MWD, baseline SGRQ, and osteoporosis were analyzed for their possible association with PR response.

Statistical analysis

Quantitative variables were summarized as mean ± standard deviation and qualitative variables as percentages. Paired t-test was used to compare change in 6MWD and SGRQ scores measured before and after PR sessions. We used binary logistic regression analysis with forward inclusion approach to find association between potential predictors and PR response. Multivariate models were used to measure the odds ratios with their 95% confidence intervals, for different factors predicting response. P value was considered significant at < 0.05. All statistical calculations were done using computer program SPSS (IBM SPSS Statistics 21.0; Armonk, NY, USA).

Results

Out of the 102 patients enrolled at baseline, 80 (78.4%) completed the PR program and were included for final analysis. Elderly (above 60 years; n = 41) and male gender (M: F = 12:1) comprised the majority of the patients. Moderate (50% ≤ FEV1 <80% predicted) or severe stage (30% ≤ FEV1 < 50% predicted) of the disease was seen in 78.7% of patients. Majority of them (94%) were ever smokers with 38 mean pack years. Sixteen patients had a low BMI (<18.5 kg/m2) with majority of them also having a very severe disease (FEV1 <30% predicted). Among the comorbidities evaluated, bone mineral derangement (osteoporosis and osteopenia) was the most common, seen in 62.5% of patients [Table 1].

Table 1

Baseline characteristics of the chronic obstructive pulmonary disease patients

Variable	Value*
Age (years)	60 60.9±9.1
Sex (male/female)	74/6
Ever smokers	75 (93.7)
Nutritional status: Underweight	16 (20)
Normal	53 (66.25)
Obese	11 (13.75)
COPD
Stage I	2 (2.5)
Stage II	36 (45)
Stage III	27 (33.75)
Stage IV	15 (18.75)
Osteoporosis	23 (28.7)
PaO2 (mmHg)	63.7±8.4
GHQ score
Score ≤3	69 (86.2)
Score >3	11 (13.8)
Baseline SGRQ	44.1±19.3
Baseline 6MWD (m)	318.4±119.5

*Values are written as mean±SD or n (%).SD: Standard deviation, GHQ: General health questionnaire, PaO2: partial pressure of oxygen in arterial blood, SGRQ: St. George’s Respiratory Questionnaire, 6MWD: 6-min walk distance, COPD: Chronic obstructive pulmonary disease

Change in outcome parameters after pulmonary rehabilitation

A statistically significant improvement was seen in both primary outcome parameters, i.e., 6-MWD and SGRQ score (P < 0.001) [Table 2]. A total of 46 (57.5%) and 54 (67.5%) patients demonstrated MCID in 6MWD and SGRQ score, respectively. Forty-two patients (52.5%) showed a positive response to PR as defined by improvement in both outcome parameters. Change in 6MWD strongly correlated with change in SGRQ score (Pearson correlation −0.52; P < 0.001).

Table 2

Change in the primary outcome parameters after pulmonary rehabilitation

Outcome parameter	Baseline value	Postrehabilitation value	Mean change	P
6MWD (n=80)	318+119 m	378+124 m	60+42m	<0.001
SGRQ score (n=80)	44.1+19.3	38.7+19.5	−5.3+3.2	<0.001

SGRQ: St. George’s Respiratory Questionnaire, 6MWD: 6-min walk distance

Factors predicting improvement in outcome parameters

On univariate analysis, age, BMI, FEV1 <50% predicted, osteoporosis, baseline SGRQ score, and gender showed an association with the PR response at 20% level of significance (P < 0.2). However, on multivariate stepwise logistic regression analysis using the above variables, osteoporosis and FEV1 <50% predicted were found as the independent factors predicting PR response [Table 3].

Table 3

Association of independent variables with improvement after pulmonary rehabilitation using logistic regression analysis

Variables	Univariate analysis			Multivariate analysis
	OR	95% CI	P	OR	95% CI	P
Age	0.96	0.92-1.01	0.15
Female	0.16	0.02-1.4	0.10
Current smoker	1.18	0.48-2.9	0.71
BMI	1.08	0.96-1.22	0.16
FEV1 <50% predicted	2.2	1.02-5.4	0.05	2.9	1.1-8.3	0.04
PaO2	1.01	0.96-1.07	0.59
mMRC dyspnea grade	1.2	0.55-2.7	0.61
GHQ score >3	1.1	0.31-3.9	0.88
Baseline 6MWD	1.002	0.99-1.005	0.38
Baseline SGRQ	0.97	0.95-1.01	0.06
Osteoporosis	0.29	0.15-0.56	<0.001	0.26	0.13-0.53	<0.001

95% CI: 95% confidence interval, OR: Odds ratio, BMI: Body mass index, PaO2: Arterial partial pressure of oxygen, GHQ: General health questionnaire, SGRQ: St. George’s Respiratory Questionnaire, 6MWD: 6-minute walk distance, FEV1: Forced expiratory volume in 1 s

Discussion

The present study was sought to find factors that could predict immediate outcome after PR in stable COPD patients. The results showed that severe and very severe COPD stage (FEV1 <50% predicted) was an independent predictor of improvement whereas osteoporosis was associated with no response to PR.

The study reinforced the existing scientific evidence regarding the benefit of PR in improving the exercise capacity and HRQoL in COPD patients.[4522] Fifty-two percent of patients in the study demonstrated improvement in both the outcome variables. The response rates are comparable to the previous studies that reported improvements of 50%–68%[68] and 50%–75%[723] in 6MWD and SGRQ score, respectively. It is pertinent to note that a significant proportion of patients in these studies, including ours (47.5%), did not show the expected results. Hence, knowledge of the factors predicting response is crucial for ensuring better PR efficacy.

The role of baseline lung functions in predicting benefit after PR in COPD has been investigated previously. However, the results have been discordant with few studies[7232425] showing its negligible value while others demonstrating a positive association between a worse baseline lung function and improvement in exercise capacity.[826] The present study demonstrated a significant improvement in both exercise capacity and HRQoL in patients with FEV1 <50% as compared to those with preserved lung function [Table 3]. Usually, patients with severely compromised lung function have a restricted mobility, that is, progressive in nature. This restriction further leads to muscle wasting as well as increases demotivation. It is plausible that this resultant deterioration might have increased the scope of improvement (to PR) in patients with poor lung function in the study. Nevertheless, patients with mild disease (FEV1 >50%) also showed improvement of 52 m (14.5% over baseline) in 6MWD though it was significantly less than patients with FEV1 <50% (66 m; 23.2%). Hence, it is pertinent to note that the presence of poor lung function should not serve to discourage PR use in patients with mild disease.

Osteoporosis is a disabling comorbidity that may be seen in up to 50% of the COPD patients.[27] A large prospective and multicenter study demonstrated a negative association of osteoporosis with 6MWD improvement (P = 0.01) but not with the quality of life[21] [Table 4]. However, the present study demonstrated a statistically significant association of osteoporosis with both outcome parameters. Pathogenesis of osteoporosis in COPD is multifactorial in nature with low daily physical activity, nutritional deficiency, and corticosteroid-induced side effects contributing to its development.[3031] Diminished response to PR in osteoporotic patients might be attributed to the increased bone fragility and muscle weakness/myopathy commonly seen in COPD patients.[21] The results highlight the need to screen all patients for osteoporosis before enrolling them for PR. However, the presence of osteoporosis should not be an absolute contraindication for PR but should guide the physician to treat the deficiency at the outset before initiating the intervention. It is likely that these patients with poor bone health may benefit from longer (12 weeks) sessions of comprehensive PR.

Table 4

Comparison of recent studies about predictors of outcome to pulmonary rehabilitation in chronic obstructive pulmonary disease patients

Study, year	Sample size	Type and duration of PR program	Outcome variables	Significant predictors of PR outcome
Spielmanns et al., 2016[28]	1492	Inpatient; 4 weeks	6MWD	Baseline 6MWD BODE index
Boutou et al., 2014[6]	787	Outpatient and home, 8-12 weeks	6MWD/ISWT	Baseline 6MWD/ISWT
			QoL (CAT, CRDQ)	Age, baseline health status, MRC dyspnea score
			HAD scale	Baseline HAD score
Zanini et al., 2013[8]	75	Inpatient; 3 weeks	6MWD	FEV1<50% TLCO<50%
Selzler et al., 2012[25]	676	Outpatient; 6-8 weeks	12MWD	Nil
			SF-36	Nil
			SGRQ	Nil
Altenburg et al., 2012[26]	102	Hospital based; 7 weeks	ESWT	FEV1% pred, Wpeak, change in lactate concentration, and ISWT distance
Van Ranst et al., 2011[23]	437	Inpatient and outpatient; 12 weeks	6MWD	Baseline 6MWD
			QoL (SGRQ)	Baseline SGRQ
Crisafulli et al., 2010[21]	316	Outpatient; 21 sessions (3 sessions per week)	6MWD	Osteoporosis
			MRC dyspnea score	PaCO2
			SGRQ	Nil
Vagaggini et al., 2009[7]	60	Outpatient and home; 8 weeks	6MWD	BMI>25 PaO2<60 mmHg
			SGRQ	Nil
Garrod et al., 2006[24]	74	Outpatient and home; 7 weeks	6MWD	Quadriceps torque, lower self-efficacy
			Quadriceps strength	Low FEV1
			QoL (SGRQ)	Higher 6MWD
Troosters et al., 2001[29]	49	Outpatient; 12 weeks	6MWD/incremental cycle ergometry	VE/MVV, PImax, Peripheral muscle strength
			CRDQ	Baseline CRDQ score
Present study	80	Outpatient and home; 8 weeks	6MWD	FEV1<50%, osteoporosis
			SGRQ	FEV1<50%, osteoporosis

6MWD: 6-minute walk distance, SGRQ: St. George’s Respiratory Questionnaire, ISWT: incremental shuttle walk test, QoL: Quality of life, CAT: COPD assessment test, CDRQ: Chronic respiratory disease Questionnaire, HAD: Hospital anxiety and depression, MRC: medical research council, SF-36: Short form health survey, ESWT: Endurance shuttle walk test, Wpeak: peak power output, PaCO2: Arterial partial pressure of carbon dioxide, BMI: Body mass index, VE/MVV: Ventilator reserve, PI max: Inspiratory muscle strength, PR: Pulmonary rehabilitation, FEV1: Forced expiratory volume in 1 s

A previous study demonstrated that patients with BMI >25 kg/m2 had a significant improvement in 6MWD[7] [Table 4]. However, the association was negated in another recent retrospective study that ruled out any effect of muscle depletion or obesity in achieving MCID in exercise tolerance or quality of life.[32] The present study also did not demonstrate any association between BMI and PR response. Certain comorbidities such as metabolic and heart diseases have also been evaluated as predictors of response to PR with varying results.[1011] However, these could not be evaluated in the study due to less number of these patients. Other variables such as PaO2,[7] dyspnea score, baseline 6MWD,[28] and SGRQ score[23] have also been evaluated previously; however, these did not yield any association in the present study [Table 4]. The difference in study design, PR programs, and outcome parameters used might be the reasons for these differences.

The present study evaluated a comprehensive list of different patient- and disease-related factors that could predict improvement after a PR program. Unlike previous studies, PR response was judged by the combination of 6MWD and SGRQ score that increased the authenticity and clinical relevance of the results. This is probably the first Indian study on the topic that also validated the importance of this nonpharmacological treatment modality in COPD. Despite being a standard of care, it is still underutilized or not available at many centers in India. The present study also had a few limitations. An absence of control arm might have falsely appreciated the perceived improvements after PR. However, all patients were already on pharmacological treatment for >6 months before enrollment, thus negating its confounding effect. Moreover, PR being a standard treatment strategy in COPD, creating a control arm in the study could have raised ethical issues. Few potential factors such ass peripheral and respiratory muscle strength[29] and social factors were not considered for evaluation that might have affected the results. Finally, increasing the sample size could have uncovered a few other predictors that might have been missed otherwise.

Conclusion

The results of the present study advocate the assessment of baseline lung function and BMD before enrolling COPD patients for the PR program. Incorporating these in prerehabilitation evaluation may help to achieve a better outcome after PR. Further studies focusing on individual factors in a large cohort of patients can further help to refine our approach to the effective use of PR.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.