Study design:
We conducted a parallel-group, non-blinded, randomised control trial at the haemato-oncology unit of University Malaya Medical Centre (UMMC), a tertiary university hospital with 1,617 beds in Kuala Lumpur, capital of Malaysia, from 1st October 2019 to 31st May 2020. Patients included were aged 18 years and above, had a histopathological diagnosis of haematological cancer according to World Health Organisation classification, and a fatigue score of ≥ 4 based on the fatigue subscale of Edmonton Symptom Assessment System (ESAS). Patients were excluded if they had impaired conscious level, cognitive impairment or any psychiatric illness that would prevent them from giving informed consent or participate fully in the study; active or past history of cancer of other system, or a haemoglobin level of < 8g/dl.(29)
Procedure:
Patients with haematological cancer attending the haematology clinic or admitted to the haematology ward of UMMC were consecutively screened for eligibility. The demographic and clinical data of the eligible patients, which include age, gender, ethnicity, religion, occupation, education level, marital status, type of haematological cancer, current status of cancer, types of cancer treatment, duration of cancer, blood parameters and other co-morbidities were obtained from the hospital Electronic Medical Record System. Any missing information was obtained by face-to-face interview with patients or relatives.
Patients who satisfied the inclusion criteria and agreed to participate in the study were randomly assigned into 2 groups based on computer-generated random numbers, in blocks of 10, with a one-to-one allocation ratio. Allocation sequence was concealed with sealed envelopes to prevent selection bias. Patients allocated to the intervention group received standard care plus a guided 30-minute mindful breathing session which consisted of four breathing exercises done consecutively. The four exercises included identifying the in-and out-breath, following the entire length of the breath, bringing the mind back to the body and relaxing the whole body.(30) Each exercise lasted 7.5 minutes. Guidance was given by one of the two research assistants, who were medical doctors. They were trained by one of the co-investigators, who was a palliative care physician, certified in mindfulness training.
The training included a brief introduction to the basic concepts of mindfulness, followed by a 30-minute mindful breathing session guided by the trainer. Guidance on delivering the intervention with attention to paralanguage (intonation, rate and rhythm of speech, pitch, articulation, use of silence, etc.) and body language (eye contact, facial expression, posture and bodily movement), followed by supervision of the actual delivery of the 30-minute mindful breathing session by each research assistant were performed. The instructions for the 30-minute mindful breathing are presented in Table 1. Patients in the control group received standard care alone. They were allowed to resume their usual activities 30 minutes prior to further assessment.
The study outcomes were assessed at minute 0 (before intervention – T0) and minute 30 (after intervention – T30). The outcomes at T0 and T30 include fatigue severity according to the fatigue subscale of ESAS, a unidimensional visual analogue scale (VAS) of 0 – 10, and the score of Functional Assessment of Chronic Illness Therapy (FACIT) Fatigue Scale Version 4, a multidimensional fatigue scale. At the end of the study, patients in the intervention group were asked about their feedback, any harm, and asked if they were satisfied and willing to practise 30-minute mindful breathing in their daily life.
The ESAS is a valid and reliable tool to assess nine common symptoms experienced by cancer patients.(31) The nine symptoms assessed include: pain, tiredness, nausea, depression, anxiety, drowsiness, loss of appetite, wellbeing and shortness of breath. An additional blank scale is given to assess each patient’s ‘other problems’ as needed. The severity for each symptom upon assessment has a rating from 0 to 10 on a numerical scale; with 0 indicating absence of the symptom and 10 indicating the worst symptom severity. For this study, the tiredness subscale was chosen to assess participants’ fatigue severity.
The FACIT Fatigue Scale is a 13-item multidimensional assessment tool to measure individual’s fatigue level during their normal daily activities over the past 7 days.(32) It has high internal validity (Cronbach’s alpha = 0.96) and high test-retest reliability (ICC = 0.95). Each participant’s level of fatigue is rated on a five point Likert scale (0 = not at all fatigued to 4 = very much fatigued). The total FACIT-fatigue score ranges from 0 to 52, with a higher score reflecting more fatigue.
The medical ethics approval was obtained from the Medical Ethic Committee of the UMMC (Ethics no: 201971-7588). Written informed consent were obtained from all the participants. The study was conducted according to the Declaration of Helsinki.
Statistical analysis:
The sample size was calculated based on the formula for a randomised control trial for continuous variables with statistical superiority design, n = [2 SD2(Z α/2 + Zβ)2] divided by d2,(33) in which n was sample size in each group; SD was standard deviation from previous study; Z α/2 was desired level of statistical significance, typically 1.96 for type-1 error of 0.05; Zβ was desired power, typically 0.842 for 80% power; d was effect size, the difference in means. In this study, we took the SD of 1.58 and d of 1.(21) Therefore, the minimum sample size was 78 (39 for each arms).
Results for continuous variables were expressed as mean ± standard deviation (SD), median or inter-quartile range depending on normality of the variable distributions; while results for categorical variables were expressed as percentages. Between-groups differences for continuous data was compared by using independent t-test or Mann-Whitney U test, as applicable; while between-group differences for categorical data was compared by using Chi-Square test or Fisher-Exact test, as applicable. A 2-sided p-value of less than 0.05 was considered as significant in this study. Statistical analyses were performed using the software package, Statistical Package for the Social Sciences (SPSS for windows version 25.0, SPSS Inc, Chicago, IL, USA).