TWO DIFFERENT DURATION OF OUTPATIENT PULMONARY REHABILITATION PROGRAMME

Comparing the effectiveness of two different duration of outpatient pulmonary rehabilitation programme

Comparing the effectiveness of two different duration of outpatient pulmonary rehabilitation programme.

Comparison between “goods” and “poor”-attenders

Table 1 compares the baseline characteristics of the “good” attenders and “poor” attenders.

Table 1.

Comparison of baseline variables between eventual good attenders (67% of total sessions attended) and poor attenders (<67% of total sessions attended)

Variable

Good attenders (n=180)

Poor attenders (n=62)

p-Value

Male (%)

56.1

69.4

0.15

Age (years)

67.0±9.0

67.6±8.3

0.64

BMI (kg/m2)

26.4±6.0

24.7±5.8

0.06

COPD diagnosis (%)

81.8

87.9

0.44

Current smokers (%)

17.7

56.5

<0.001

FEV1 (% pred)

40.5±14.8

36.6±14.2

0.07

MRC dyspnoea score*

4 (3-5)

5 (4-5)

<0.001

SGRQ total score

62.4±15.8

65.5±14.8

0.20

Hospital admissions in last year

0.82±1.34

1.42±1.43

0.004

Major co-morbidities (%)

48.3

61.3

0.08

Distance from PR centre (miles)

9.8±9.1

9.0±7.0

0.55

Length of journey (%)

 <5 min

2.2

0

 6-15 min

27.2

8.1

<0.001

 16-30 min

44.4

33.9

 >30 min

25.6

58.1

Short rehabilitation (%)

56.1

43.5

0.09

Univariate analysis

A positive relationship with attendance was found with higher BMI (p<0.05), fewer hospital admissions in the last year (p<0.01), current (self-reported) non-smokers (p<0.01), a lower MRC dyspnoea score (p<0.01) and a shorter journey time to reach the PRP (p<0.001).

Regression analysis

Table 2 shows the results of multiple regression analysis, including only those variables associated with poor attendance. The following were not significantly associated with attendance and were excluded from the linear regression model: age (p=0.53), gender (p=0.93), FEV1% (p=0.90), co-morbidity (p=0.84), respiratory diagnosis (p=0.68), baseline SGRQ (p=0.42) and BMI (p=0.27). Distance from PRP centre showed a trend but was not a statistically significant predictor in our model (p=0.08). Overall our regression model still only accounts for 18% of the variance in log ratio of attendance.

Table 2.

Results of multiple linear regression analysis for attendance using transformed data: Ln {(attendances+1)/(absences+1)}

Log attendance ratio

Change statistics

Variables added at each stage

Adjusted R2

R2 change

F change

d.f.

Sig. F

1. Journey duration

0.062

0.062

15.309

1

<0.001

2. Journey duration and MRC score

0.117

0.055

0.13

2

<0.001

3. Journey duration and MRC score and smoker

0.141

0.024

2.425

3

<0.001

4. Journey duration and MRC score and smoker and type of PRP

0.164

0.023

1.243

4

<0.001

5. Journey duration and MRC score and smoker and type of PRP and number of hospital admissions in the preceding year

0.176

0.012

1.343

5

<0.001

In this retrospective analysis using routine and easily accessible hospital data, we have demonstrated that patients are less likely to attend a PRP if they are current smokers, attend a long rehabilitation programme, previously had frequent exacerbations requiring hospital admission, have a high MRC dyspnoea score or endure a long journey.

Adherence is defined as “the extent to which a patient's behaviour coincides with medical advice” rather than concordance where the patient and doctor both agree on and negotiate conduct. Patients were strongly advised to attend all 18 sessions so adherence is the more appropriate term. Human behaviour is a complex phenomenon, which is influenced by personal ...