We conducted a parallel-group randomised controlled pilot trial; an equal randomisation of 1:1 was chosen. Inclusion criteria were (1) age ≥18 years, (2) evidence on a High-Resolution Chest CT (HRCT) of fibrosing interstitial disease19 (with at least one of the following: traction bronchiectasis and honeycombing by CT scan) (3) and evidence of advanced disease, defined as the partial pressure of oxygen in arterial blood (PaO2) ≤ 60 mmHg at room air, a decline in forced vital capacity (FVC) ≥ 10% in the previous 6 months, or a GAP index (a multidimensional index consisting of Gender [G], Age [A], and two lung physiology variables [P], forced vital capacity [FVC] and diffusion lung carbon monoxide [DLCO]) at least 3.20,21 High-Resolution Chest CT was discussed during a weekly multidisciplinary meeting by a team formed by radiologists, pulmonologists, surgeons, rheumatologists and pathologists. Exclusion criteria were: diagnosis of active cancer, treatment with anti-fibrotic therapy, and ongoing palliative pharmacological treatments. Although anti-fibrotic therapy is considered the standard of care for patients with fibrosing ILD and only a fraction of patients cannot or choose not to receive these medications, we decided to recruit only patients not taking anti-fibrotic drugs, to avoid potential bias of slowed progression (See Supplementary Material for additional data). Fifty consecutive outpatients at our ILD center were enrolled between October 2016 and September 2019. Subjects were randomly assigned to two groups, based on the activation of the palliative care intervention (intervention group) or being in the usual care group. Randomisation was performed using a generated computer sequence with block of 10. An independent statistical consultant set up the web-based randomisation process to assign eligible participants to intervention or control groups by remote allocation; no one directly involved in the project had access to allocation codes. All patients were followed-up for 12 months by their regular pulmonologist who was not blinded on group allocation.

Patients in the intervention group were scheduled to meet, at least every 6 weeks, for the whole duration of the study, a physiotherapist, a psychologist, a dedicated nurse, a palliative care doctor, and a pulmonologist experienced in ILD care. In the intervention group, the visit addressed several topics which included evaluation of patient's understanding of their illness and prognosis and establishing goals of care. Medical therapy (anxiolytic and anti-depressive medications, low-dose opioids, cough depressant, oxygen titration, etc.) was also initiated or changed to ameliorate the symptoms.

Patients allocated to the intervention group also participated in a rehabilitation program consisting of breathing and motor exercises. The same physiotherapist supervised all therapeutic sessions (60 min each) during the study period; respiratory exercises consisted of breathing control techniques, and were executed in a sitting, relaxed position. Motor exercises mainly consisted of active movements of the lower limbs directed at flexibility exercise and short active stretching of quadriceps, calves, and gluteus.22,23 Those subjects who were able to maintain a standing position without incurring severe dyspnoea, fatigue, hemodynamic instability exercised with bilateral semi-squat, standing up on tiptoe. (See Supplementary Material for additional data.). Pulmonary rehabilitation programs provided to patients with interstitial lung diseases typically include a period of 3-4 weeks of daily incremental exercise training sessions.24 Indeed, our programs lasted 12 months and it can be more aptly defined as rehabilitative counselling. To this end, patients were asked to repeat the exercises they had practised at home and they were also specifically consulted about and instructed on home-based physical activity during their visits; patients were encouraged to engage in physical activity, such as walking at their own pace for > 30 minutes most days of the week. Patients attending rehabilitative sessions were under oxygen therapy. During the treatment sessions each patient was monitored via a pulse oximeter; for those subjects who manifested O2 desaturation (i.e. peripheral oxygen saturation (SpO2)< 85%),25 the O2 supply was adjusted to guarantee a peripheral oxygen saturation (SpO2) >88%.

Patients in the usual care group followed the regular center schedule for medical visits (at least 3 times in a year) with a pulmonologist experienced in ILD care; medical therapy was also initiated or changed to ameliorate the symptoms. Management trails for both intervention group and usual care group during follow-up is illustrated in Table 1.

In both groups, at every visit, the pulmonologist recorded the intensity of perceived dyspnoea – using the Borg scale15 (score 0-10, where 10 is the worst) - and cough intensity - using the Visual Analog Scale16 (VAS) (score 0-100, where 100 is the worst). Spirometry and DLCO were performed at the baseline, using standardized references.26,27 We also obtained the Maugeri Respiratory Questionnaire (MRQr) reduced form18 for QoL impairment for chronic respiratory failure patients (score 0-28, where 28 is the worst), and the Center for Epidemiologic Studies Depression Score (CES-D)17 (score 0-60, where 60 is the worst). A brief explanation of these tools is reported in the Supplementary Material.

To explore the impact of the intervention on physical functioning over time, we examined lower body flexibility (chair sit-and-reach test)28 and strength (30-s Chair-Stand Test)29 at baseline in the intervention group. Thirty-s (30s) chair stand,28,29 consisting of counting the number of times the patient comes to a full standing position in 30 seconds. During execution of the test, the subjects were asked to place their hands on the opposite shoulder keeping their arms close to the chest. The normal range of scores in men and women was considered as predicted values.28,29 The chair sit-and-reach test was performed starting from a sitting position with a leg extended and hands reaching towards the toes.28 Measurement of the distance between extended fingers and the tip of the toe gives the result. The normal range of scores in men and women were considered the predicted ones.28 To avoid excessive fatigue and to allow a certain degree of progression, the chair sit-and-reach30 was executed before the 30-s chair stand test.

To define the sample sizes of the two groups included in the experiment, we assessed power using the Borg scale: from data collected in the 4 previous years in our clinic in this subset of patients, a mean score of 5 with a standard deviation 2.5 was estimated. In order to distinguish an average effect, change of 2 points in the scale with α=0.05 and a power of 0.8, a sample size of 25 for each treatment group was selected. Data were evaluated with an intention-to-treat analysis. Summary data are reported as mean or percentage unless otherwise stated. Homogeneity between groups was assessed to identify potential systematic differences at the baseline: chi-square independence tests were used for categorical variables, two samples independent t-tests were adopted for numerical variables after an assessment of normality and homoscedasticity assumptions. In the case of assumption violation, the Wilcoxon signed-rank test was used.

To assess the effect of the different treatments on the symptom and QoL scales, two-samples t-tests were performed to compare the mean score differences between the baseline and the follow-up (12 months) in the treatment groups. To investigate the evolution of the monitored indicators in time, we used a two-way ANOVA model with treatment group and visit time (baseline, after 6 months, and after 12 months) as factors. Adjusted p-values were produced to correct the level for multiple testing: the multivariate t distribution accounts for the correlations among the multiple hypothesis.31 The Wilcoxon test for paired samples was used to investigate if differences in the distribution of SpO2 during chair stand test, could be detected after 6 and 12 months in the intervention group. To compute the sample size for future randomised controlled trials (RCTs) aimed at assessing the statistical mean difference between the two groups for CES-D and Borg scale at 6 and 12 months (with confidence level α=0.05), the approach by Vierron and Girauderau32 was used. In the study, a p-value less than 0.05 was considered statistically relevant. The statistical analyses were carried out in the R environment.33

Focusing on the differences between the values registered for the studied scales between the baseline and after 12 months of treatment, changes in the means of the treatment groups were identified (Table 3). The Borg scale at 12 months showed a worsening in dyspnoea in both groups, such worsening being less evident in the intervention group (+0.93 ± 3.34 vs +3.42 ± 2.5, p-value: 0.03). A similar trend was also observed for the CES-D scale (+1.57 ± 7.95 vs +6.58 ± 8.55, p-value: 0.01). There were no differences for the other scales.

Results concerning the scales differences at 6 month and 12 months are shown in Table 4 and Table 5. Moreover, in Fig. 2, the absolute variations over time for the different scales are shown. We did see an improvement in dyspnoea and depression: specifically, Borg scale's and CES-D scale's values remained stable in patients from the intervention group, while they both deteriorated in patients from the control group. Particularly, changes > 30% in the CES-D represent a minimal clinically important difference.34 In contrast, we noted a constant decline in VAS scale values for cough in both groups: based on such observation, we suppose that palliative care is not sufficient to alleviate the burden of debilitating cough, which most ILD patients share.

Figure 2.

Fitted values and 95% confidence intervals for the studied scale in time. The triangles highlight a statistically significant difference with respect to the baseline (see Tables 4-5).

(0.31MB).

Regarding the Borg scale, no differences in time were observed in the intervention group (p-value between baseline and 6 months 0.31, p-value between baseline and 12° month 0.46), while Borg scale values worsened in usual care group (p-value between baseline and 6 months 0.03, p-value between baseline and 12° month < 0.001). Similarly, CES-D scales’ values remained stable in the intervention group's patients (p-value between baseline and 6 months 0.88; p-value between baseline and 12° month 0.67); on the contrary, there was a worsening in CES-D values in usual care group (p-value between baseline and 6 months 0.02; p-value between baseline and 12° month <0.001). The VAS scale values for cough worsened both in intervention and usual care group (p-value group effect: 0.88), especially in the second 6-month period (p-value between baseline and 12° month: 0.04 for the intervention group and p-value 0.002 for the usual care group). A similar pattern can be observed for SpO2/FiO2 (p-value between baseline and 12° month: 0.01 for the intervention group and 0.007 for the usual care group). Regarding the MRQr scale, only a slight decrease in usual care group was detected after 12 months (p-value: 0.04), although no decrease in the intervention group was detected (p-value: 0.94). To test for differences in survival of subjects included in the two groups, Kaplan-Meier curves were estimated: differences were not significant (p-value=0.11).

The mean value of the chair sit-and-reach test at the baseline was 11.1 cm ± 6.6, 10.4 cm ± 10.4 after 6 months, and 10.4 cm ± 8.0 after 12 months. Similarly, lower body strength in the intervention group at baseline, using the 30-s chair stand, was evaluated in all patients, and the mean value was 5.9 ± 3.6 repetitions; 5.2 ± 3.4 repetitions after 6 months and 3.6± 2.8 repetitions after 12 months. Based on a one-way ANOVA model, we did not observe changes in these measures during the study period for either the lower body flexibility or strength.

Intervention group did better than usual care group for at least one of the primary outcomes (reducing the increase of dyspnoea during the time course). We may suggest therefore that this variable should be considered in designing a phase III study. A 6 months study will allow us to detect some benefits quicker, and eventually reduce the loss of number of patients due to their deaths. Indeed, when we consider feasibility for an outpatients based-program, we also have to take into consideration other items such as costs of transportations, loss of working-days by caregivers, and distance from the hospitals. We estimate that, considering CES-D and Borg scale as primary outcomes at 6 months and assuming an equal enrolment rate by centers, a sample size of 146 patients (total of intervention + control), is required to obtain a power of 0.80, fixing the intra-center correlation equal to 0.1.

Our pilot trial has several limitations. First, this is a single-center study and may not generalize to other areas; indeed, very few (if any) of the Respiratory Units in our Country have a structured palliative care program for ILDs patients, so that even psychological support is not usually provided, if not as inpatients. This may be different in other geographical locations.12,40 Second, we had a relatively small sample size: a larger randomised trial is necessary to confirm our findings. Third, the chair stand and chair sit and reach tests were not recorded in the usual care group, and it was not possible to record pulmonary function tests (PFTs) in most of patients due to severe impairment and subsequent difficulties in performing them, especially concerning DLCO, most of them being on oxygen therapy. In keeping with the last point, some patients could not perform the six minute walking test (6MWT) and, most important, a portion of those who could perform the test (14/25) were not able to repeat it at 12 months, due to the pandemic, since the dedicated corridor was closed to outpatients. Lastly, the exclusion of patients who were not taking anti-fibrotic medication, limit the generalization of our results in patients receiving these drugs.