Symptom burden according to dialysis day of the week in three times a week haemodialysis patients.

BACKGROUND: Haemodialysis patients experience significant symptom burden and effects on health-related quality of life. Studies have shown increases in fluid overload, hospitalization and mortality immediately after the long interdialytic interval in thrice weekly in-centre haemodialysis patients, however the relationship between the dialytic interval and patient reported outcome measures (PROMs) has not been quantified and the extent to which dialysis day of PROM completion needs to be standardised is unknown.
METHODS: Three times a week haemodialysis patients participating in a stepped wedge trial to increase patient participation in haemodialysis tasks completed PROMs (POS-S Renal symptom score and EQ-5D-5L) at recruitment, six, 12 and 18 months. Time from the long interdialytic interval, HD day of the week, and HD days vs non-HD days were included in mixed effects Linear Regression, estimating severity (none to overwhelming treated as 0 to 4) of 17 symptoms and EQ-5D-5L, adjusting for age, sex, time on HD, control versus intervention and Charlson Comorbidity Score.
RESULTS: 517 patients completed 1659 YHS questionnaires that could be assigned HD day (510 on Mon/Tue/Sun, 549 on Wed/Thu/Tue, 308 on Fri/Sat/Thu and 269 on non-HD days). With the exception of restless legs and skin changes, there was no statistically significant change in symptom severity or EQ-5D-5L with increasing time from the long interdialytic interval. Patients who responded on non-HD days had higher severity of poor appetite, constipation, difficulty sleeping, poor mobility and depression (approximately 0.2 severity level), and lower EQ-5D-5L (-0.06, CI -0.09 to -0.03) compared to HD days.
CONCLUSIONS: Measuring symptom severity and EQ-5D-5L in haemodialysis populations does not need to account for dialysis schedule, but completion either on HD or non-HD days could introduce bias that may impact evaluation of interventions. Researchers should ensure completion of these instruments are standardized on either dialysis or non-dialysis days.

Introduction

People receiving haemodialysis (HD) for end stage renal failure (ESRF) experience significantly impaired health-related quality of life (HRQoL) and high symptom burden [1]. When looking to improve outcomes on haemodialysis, patients, clinicians and policymakers are increasingly focusing on HRQoL and symptoms in addition to the traditional endpoints including cardiovascular disease and survival [2]. Patients state HRQoL is a basic aspect of health [3], and studies have shown these patient-reported outcomes measures (PROMs) are more strongly associated with the risk of death and hospitalization than clinical parameters [4]. Although use of PROMs in haemodialysis studies are increasing, standardisation in terms of how measures are implemented is lacking. In addition to the unmet need is the undetected need: healthcare providers fail to commonly recognise and treat the physical and emotional symptoms experienced by HD patients [5, 6]. Significant changes in the severity of symptoms occur at a median of 3 months, justifying regular surveillance of symptoms in the dialysis population [7], interventions based on monitoring symptoms to improve outcomes in haemodialysis patients [8], and the robust measurement of symptoms in these settings.

In addition to the response of symptom severity to interventions in the clinical or experimental setting over time, the HD schedule may introduce further variation. It is recognised that the varying intervals between haemodialysis sessions have an association with volume status, uraemia and electrolyte imbalance, measures of cardiac function, hospitalisation and death [9, 10]. When Standardized Outcome in Nephrology (SONG-HD), a consensus exercise to establish core outcomes to be measured and reported in haemodialysis trials, spoke to patients about their experience of fatigue, they reported “extreme fatigue on Sunday night because it was 2.5 days without treatment” [2]. Furthermore patients’ perceptions of their symptoms or quality of life may be more important than objective clinical assessments using validated instruments [11]. There is expanding literature that PROMs are not only effected by psychosocial issues, stress, emotions, patient characteristics and co-morbidities [12, 13], the environment in which the instrument is completed may influence the result [11].

HRQoL has been shown to be a predictor of morbidity and mortality in haemodialysis patients [14, 15] and HRQoL measures play an important role in evaluating cost effectiveness of treatment. Failure to account for any underlying differences in symptom severity due to the day of instrument completion in relation to the dialysis schedule could bias the impact and effectiveness of interventions for symptoms and HRQoL, which have been prioritised by the patients and clinicians [2] and could lead to failure of new treatment or interventions to be approved. To quantify this, we used data from a large stepped wedge randomised controlled trial with aim to determine the association between symptom burden and the haemodialysis schedule in three times a week haemodialysis patients and to explore the effect of PROMs completion on dialysis and non-dialysis days, accounting for patient characteristics.

Material and methods

Study design

Data for this study was obtained from SHAREHD Stepped Wedge Cluster Randomised Trial (SWCRT) [16, 17]. which evaluated a quality improvement collaborative designed to create an environment to support in-center HD patients to dialyse more independently. The evaluation ran for 18 months with an additional six months sustainability, and was conducted in 12 renal centres in England. It ran from October 2016 to October 2018: following a control period of six months (baseline), six centres participated in the intervention immediately (step one) with six centres joining after a further six months (step two).

Setting

Trained, delegated research nurses gained written informed consent to participate from prevalent HD patients established on centre-based haemodialysis. The study adhered to the declaration of Helsinki, ethical approval was obtained from West London & GTAC Research Ethics Committee (IRAS project ID 212395) and the trial was registered (ISRCTN Number 93999549). This presented analysis was specified in the research protocol [16].

The recruitment exclusion criteria were patients too unwell to engage in the study, as judged by the clinical team, or patients unable to understand written and verbal communication in English. For this specific analysis, we excluded patients who had missing data for the adjustment covariates, did not dialyse three times a week, or we could not assign a HD schedule for (Fig 1). Failure to link comorbidity using National Health Service (NHS) number by NHS digital resulted in missing comorbidity data and was assumed at random. Centres participating in the trial were: Sheffield Teaching Hospital NHS Foundation Trust, Central Manchester Healthcare Trust, City Hospitals Sunderland NHS Foundation Trust, East & North Hertfordshire NHS Trust, Guy’s & St Thomas NHS Foundation Trust, Heart of England Foundation Trust, Leeds teaching Hospitals NHS Trust, The Royal Wolverhampton NHS Trust, North Bristol NHS Trust, University Hospital of North Midlands NHS Trust, Nottingham University Hospitals NHS Trust.

Fig 1

Flow diagram of the participants and completed questionnaires in the study.

Participants and data collection

Patients participating in SHAREHD were asked to complete instruments at baseline, six, 12, and 18 months at either dialysis unit or home or clinic. They were allowed to complete these instruments freely at the time and day they choose. A delegated member of the local research team collected research nurse-completed and self-completed paper instruments, which included patient demography and dialysis schedules. The Think Kidneys Your Health Survey (YHS) includes the POS-S Renal [18] and EQ-5D-5L [19]: POS-S renal consists of 17 symptoms commonly experienced by HD patients: weakness, poor mobility, pain, difficulty sleeping, breathlessness, drowsiness, feeling anxious, itching, dry mouth, restless legs, feeling depressed, poor appetite, changes in skin, constipation, nausea, diarrhoea and vomiting. Each symptom is scored on a 5-level ordinal scale ranging from not at all to overwhelmingly. The EQ-5D-5L is a five-item standardized instrument developed as a measure of generic health-related quality of life and it consists of a descriptive system which comprises five dimensions: mobility, self-care, usual activities, pain/discomfort and anxiety/depression, for which respondents select one of five levels from no problems to unbearable or being unable to perform the domain. From these responses a utility scale can be calculated from zero to one: zero being death and one being in perfect health. This generic measure and associated utility value are used to compare health between groups of individuals with chronic conditions, particularly for health economic evaluations.

Sociodemographic variables including age, gender, ethnicity and education were also collected. The Modified End Stage Renal Disease (ESRD) Charlson comorbidity index (CCI) score [14, 20] was calculated using established algorithms and weights from hospitalisation data obtained through linkage to hospital episode statistics by the NHS Digital Data Access Request Service.

Study size

The informing SWCRT sample size was determined using a recommended ICC value of 0.05 [21], to have a 90% power to detect an increase in the proportion undertaking five or more haemodialysis-related tasks from 15% to 30% with statistically significance at the 5% two-sided level: 12 clusters of 25 patients, with six clusters randomised at each step of SWCRT was arrived at. In recognition of the background mortality and renal transplantation rate and to mitigate the risk of incomplete data collection, the target recruitment per participating site was doubled to 50 [17].

Statistical analysis

Dialysis day of the week (the day of the week in relation to HD schedule) was determined from the date that the instrument was completed, combined with the dialysis schedule (routine attendance on Mon/Wed/Fri, Tue/Thu/Sat or Sun/Tue/Thu) reported at that study timepoint. Descriptive statistics were used to summarize the frequency and distribution of the questionnaires and other baseline characteristics in the study cohort, with symptoms reported recoded as none vs mild to overwhelming based on the distribution of responses. As visually a number of symptoms showed a progressive improvement from the long interdialytic interval, in the primary analysis, dialysis day of the week was treated as a continuous variable. The change in symptom severity was assessed for each additional dialysis session after the two-day break: Mon/Wed/Fri HD patients completing their questionnaires on Monday were assigned to HD day 1 (HD1), with Wednesday responses to HD day 2 (HD2) and Friday responses to HD day 3 (HD3) and therefore, HD1 was assigned to those who dialyzed and completed the instruments on Mon/Tue/Sun, HD2 to Wed/Thu/Tue and HD3 to Fri/Sat/Thu respectively. Patients who completed instruments on non-dialysis days were assigned to non-HD day.

Mixed effects linear regression was employed, with the patient specified as a random intercept, therefore allowing changes in symptom burden according to the dialysis schedule to be estimated within an individual. For the purposes of statistical analysis, the five-point responses to the symptom questions were coded 0 = not at all; 1 = slightly; 2 = moderately; 3 = severely and 4 = overwhelmingly. With the symptom recoded 0 to 4, the linear mixed effects model reports a coefficient representing the size of the change in severity level associated with a variable. e.g., a value of 1.0 would represent a change from none to mild or severe to overwhelming. Models predicting severity of each symptom were adjusted for age (<40, 40–65, >65), sex, years receiving dialysis (less than 1 year, 1–5 years, more than 5 years), comorbidity (Modified ESRD Charlson Index Score 0, 1–5 or more than 5) and steps (control arm versus intervention arm). Years on dialysis was defined as the time between the first day of dialysis treatment and the study entry date.

Sensitivity analyses exploring the relationship between symptoms and dialysis schedule were performed: HD2 and HD3 were compared individually to HD1 as a categorical variable, and HD days were compared to non-HD days. Additional analyses exploring the symptom burden affected by patient baseline characteristics and co morbidities was performed adjusted for dialysis day of the week treating HD days as continuous variable. P value < 0.05 was considered the threshold for statistical significance. All analyses were carried out using SPSS 26.

Results

Of the 586 haemodialysis patients recruited to the SHAREHD trial, 552 in-centre HD patients from the 12 participating renal centres provided data during the baseline phase. Thirty-five participants who were not dialysing 3 times a week (3xW) were excluded (6.3%). 1659 YHS questionnaire instruments were completed by 517 3xW participants at baseline, 6 and 12 months, 18 months. HD schedule at the beginning of the study was Mon/Wed/Fri (n = 266, 48.2%), Tue/Thu/Sat (n = 228, 41.3%) and Sun/Tue/Thu (n = 23, 4.3%) (Fig 1). 32 patients (6%) changed their HD schedule during the study. There were 475 observations at baseline control period (step 0), 459 observations at step 1 (212 observations in control group and 247 in intervention group) and 725 observations at step 2 (intervention group).

Out of the YHS questionnaires completed by 517 participants, 510 were assigned to HD1 (Mon/Tue/Sun), 549 to HD2 (Wed/Thu/Tue), 308 to HD3 (Fri/Sat/Thu) and 269 to non-HD day. 23 YHS questionnaires were excluded due to missing adjustment covariates.

Patient’s characteristics

The demography of patients stratified with haemodialysis schedules are outlined in Table 1 and patients who dialysed on Mon/Wed/Fri and Tue/Thu/Sat are younger but have higher comorbidities scores compared to patients who dialysed on Sun/Tue/Thu. The majority of patients were male (61%) with a mean age of 63 years, with 23.2% having been on haemodialysis for less than a year. Caucasian patients accounted for 81.2% of the patients. The Mean Modified Charlson score index was 2.85 (SD 2.77).

Table 1

Demography of patients stratified by haemodialysis pattern at baseline.

		Mon/Wed/Fri	Tue/Thu/Sat	Sun/Tue/Thu	Total
Number of haemodialysis patients		N = 266	N = 288	N = 23	N = 517
Total number of observations		885	697	77	1659
Mean Age		61.5 (SD 15.51)	64.6 (SD 15.40)	66 (SD 16.29)	63 (SD 15.56)
Sex	Male	61% (158/259)	62% (134/216)	57.1% (12/21)	61.3% (304/496)
Ethnicity	White	81.7% (210/257)	80.7% (171/212)	81% (17/21)	81.2% (398/490)
Education	No formal education	29.2% (74/253)	38.5% (82/213)	57.1% (12/21)	34.5% (168/487)
	High education (1–3) *	49.4% (125/253)	42.3% (90/213)	23.8% (5/21)	45.2% (220/487)
	Higher education (4–6) **	21.3% (54/253)	19.2% (41/213)	19.0% (4/21)	20.3% (99/487)
Myocardial infarction		17.2% (42)	23.1% (48)	9.5% (2)	19.5% (92)
Heart Failure		19.3% (47)	18.8% (39)	23.8% (5)	19.2% (91)
Cancer		7.8% (19)	9.1% (19)	4.8% (1)	8.2% (39)
Connective Tissue Disease		7.4% (18)	2.4% (5)	4.8% (1)	5.1% (24)
Cerebral vascular accident		7.4% (18)	9.6% (20)	0.0% (0)	8% (38)
DM without complication		37.3% (91)	37.0% (77)	33.3% (7)	37% (175)
DM with complication		23.8% (58)	26.0% (54)	9.5% (2)	24.1.% (114)
Pulmonary Disease		20.1% (49)	23.6% (49)	9.5% (2)	21.2% (100)
Peripheral vascular disease		27.0% (66)	25.0% (52)	14.3% (3)	25.5% (121)
Severe Liver disease		0.8% (2)	1.0% (2)	0.0% (0)	0.8% (4)
Lymphoproliferative disease		1.2% (3)	1.0% (2)	0.0% (0)	1.1% (5)
Metastatic cancer		0.8% (2)	0.8% (2)	0.0% (0)	0.8% (4)
Paraplegia		0.8% (2)	2.4% (5)	0.0% (0)	1.5% (7)
Modified Charlson score index (score 0–16) ***	Score 0	23.8% (58)	22.1% (46)	38.1% (8)	23.7% (112)
	Score 1–5	63.1% (154)	60.6% (126)	61.9% (13)	61.9% (293)
	Score >5	13.1% (32)	17.3% (36)	0% (0)	14.4% (68)
Years on dialysis	<1yr on RRT	17.9% (40)	28.3% (51)	36.8% (7)	23.2% (98)
	1–5 year	52% (116)	43.3%(78)	57.9% (11)	48.6% (205)
	>5 years	30% (67)	28.3% (51)	5.3% (1)	28.2% (119)

*High education (1 = professional qualification, 2 = ‘O’ level/GSCE equivalent,3 = Apprenticeship).

**Higher education (4 = ‘A’level/higher equivalent,5 = Degree or higher, 6 = Diploma).

*** Higher Modified Charlson score indicates high comorbidities.

Prevalence of symptoms and association with patient characteristics

Table 2 presents symptom prevalence at baseline defined as the proportion of patients with presence of symptoms (mild, moderate, severe or overwhelming symptoms), stratified by haemodialysis schedules that patients are receiving. Fig 2 shows the severity of each of the 17 symptoms, stratified by HD Days (Dialysis Day of the week: HD1, HD2, HD3) (S1 Table).

Table 2

Symptoms prevalence at baseline (mild or worse) stratified according to haemodialysis schedule.

	Mon/Wed/Fri	Tue/Thu/Sat	Sun/Tue/Thu	Total
Number of Patients	n = 266	n = 228	n = 23	n = 517
eq5d5l	0.68	0.69	0.74	0.70
Pain	(160) 62.0%	(138) 63.0%	(14) 60.9%	(312) 62.4%
Breathlessness	(146) 57.3%	(126) 57.0%	(14) 60.9%	(286) 57.3%
Weakness	(219) 84.9%	(179) 80.3%	(18) 78.3%	(416) 82.5%
Nausea	(99) 38.2%	(87) 39.5%	(9) 39.1%	(195) 38.8%
Vomiting	(54) 20.8%	(52) 23.5%	(4) 17.4%	(110) 21.9%
Poor appetite	(129) 50.0%	(110) 49.5%	(12) 52.2%	(251) 49.9%
Constipation	(97) 37.5%	(80) 36.5%	(6) 27.3%	(183) 36.6%
Sore mouth	(119) 45.8%	(117) 52.9%	(9) 39.1%	(245) 48.6%
Drowsiness	(165) 64.2%	(157) 70.7%	(15) 65.2%	(337) 67.1%
Poor mobility	(177) 68.6%	(159) 71.0%	(14) 60.9%	(350) 69.3%
Itching	(169) 65.0%	(142) 64.8%	(18) 78.3%	(329) 65.5%
Difficulty sleeping	(170) 65.6%	(146) 65.5%	(18) 78.3%	(334) 66.1%
Restless legs	(138) 53.5%	(114) 51.8%	(16) 69.6%	(268) 53.5%
Changes in skin	(132) 51.4%	(106) 48.2%	(11) 50.0%	(249) 49.9%
Diarrhoea	(82) 31.7%	(53) 24.2%	(9) 39.1%	(144) 28.7%
Anxious	(132) 51.0%	(105) 47.3%	(10) 43.5%	(247) 49.0%
Depressed	(125) 48.4%	(108) 48.6%	(9) 39.1%	(242) 48.1%

Fig 2

Symptom severity stratified by HD day (dialysis day of the week).

The observations from 4 time points (baseline, six, 12 and 18 months) were used to inform this figure (S1 Table).

In descending frequency, the five most prevalent symptoms were: weakness (82.5%), poor mobility (69.3%), drowsiness (67.1%), difficulty in sleeping (66.1%) and itching (65.5%) (Table 2). Although symptom prevalence varies by HD schedule patients were receiving, patients who dialysed on Sun/Tue/Thu have less symptom prevalence in 11 out of 17 symptoms compared to other two HD schedules (Table 2). Higher symptom burden was found only for breathlessness, poor appetite, itching, difficulty sleeping, diarrhoea and restless legs in Sun/Tue/Thu HD schedule. The multivariable adjusted association between patient characteristics and each of the symptoms in the POS-S renal and additionally adjusted for dialysis day of the week (HD1, HD2, HD3) is reported separately (S2 Table). Age, sex, years on dialysis and comorbidities independently predict symptom severity. With the exception of skin changes which were significantly higher in female, there were no significant differences in other symptom burden between males and females (S2 Table). Compared to age group 40–65 years old, older patients (>65 years) have statistically significant less symptoms burden in a range of symptoms: pain, weakness, nausea, vomiting, poor appetite, constipation, drowsiness, difficulty sleeping, restless legs, anxiety, and depression. Symptom severity was higher in a range of symptoms in respondents who had received dialysis for more than five years and had greater comorbidity defined by a Charlson comorbidity score of 5 or more. Importantly, no patient characteristic resulted in an increase in symptom severity that exceeded 1.0, representing a change in level of severity (e.g., from moderate to severe). There were 62 participants who completed only one instrument throughout the study period. Comparing this cohort with patients who completed the instruments more than once in follow up period, patient characteristics and severity of symptom burden at baseline were similar (S3 and S4 Tables).

Effect of symptom burden and change in symptom score according to dialysis day of the week

Fig 3 and S5 Table report the changes in symptom severity associated with increasing time from the long interdialytic interval, HD days compared to non-HD days, and individual HD days (HD1 vs HD2 and HD1 vs HD3). Estimating symptom changes over time after 2-day break (time from HD1) using multivariable mixed effects linear regression analysis adjusted for age, sex, comorbidity, control versus intervention period, and time on dialysis revealed that among 17 symptoms of POS-S renal, only restless legs (effect size 0.1, P = 0.014, 95% CI 0.02 to 0.18) and changes in skin (effect size 0.08, P = 0.03, 95% CI 0.01 to0.16) significantly worsened with increasing time from 2-day break (S5 Table). Increasing time from the long interdialytic interval (e.g., 0, 2 and 4 days) was not significantly associated with a change in severity for remaining 15 symptoms. Control compared to intervention period did not significantly influence the severity of any of the 15 symptoms.

Fig 3

Multivariable mixed effects linear regression comparing each hemodialysis day after long break, each HD day from HD1, HD day vs non-HD day (S5 Table).

Compared to HD days, the severity of responses on non-HD days were significantly higher for poor appetite (effect size 0.2, P = 0.004, 95% CI 0.06 to 0.34), constipation (effect size 0.19, P = 0.005, 95% CI 0.06 to 0.31), difficulty sleeping (effect size 0.2, P = 0.012, 95% CI 0.04 to 0.36) and depression (effect size 0.17, P = 0.016, 95% CI 0.03 to 0.31) (Fig 3) (S5 Table).

Comparing the HD days individuals to each other, patient’s symptoms worsen significantly on the second HD day (HD2) compared to the first (HD1) for breathlessness (effect size 0.17, P = 0.004, 95% CI 0.05 to 0.28), weakness (effect size 0.2, P = 0.002, 95% CI 0.07 to 0.33), difficulty sleeping (effect size 0.23, P = 0.001, 95% CI 0.09 to 0.36), restless legs (effect size 0.18, P = 0.006, 95% CI 0.05 to 0.32) and changes in skin (effect size 0.13, P = 0.04, 95% CI 0.01 to 0.26) (Fig 3) (S5 Table). Comparing HD1 and HD3, we found that restless legs and changes in skin worsen significantly on HD3. None of the symptoms improved significantly in both HD2 and HD3 comparing categorically to HD1.

Overall, the mean EQ-5D-5L score was 0.70 (1 being perfectly healthy and 0 being dead) (Table 2). There was no significant change in EQ-5D-5L score with increasing time from the long interdialytic interval and comparing HD2 and HD3 to HD1 (Fig 3) (S5 Table). However, EQ-5D-5L was significantly lower in non-HD days compared to HD Days (effect size -0.06, 95% CI -0.09 to -0.03) (Fig 3) (S5 Table).

Discussion

This observational study evaluated the symptom burden associated with dialysis day of the week in three times a week haemodialysis patients using POS-S Renal and EQ-5D-5L [19] at baseline, six months, 12 months and 18 months. The baseline characteristics are comparable to national registry data describing prevalent patients [22] and high levels of comorbidity were reported. This analysis showed that with the exception of restless legs and changes in skin, symptom severity was not significantly different with increasing time from two-day break although milder symptom severity on the first HD day compared to other HD days was observed in breathlessness, weakness, difficulty sleeping, restless legs and changes in skin. Symptom burden was significantly higher and EQ-5D-5L was significantly lower in non-HD days compared to HD days. No patient characteristic resulted in an increase in symptom severity that exceeded 1.0, representing a change in level of severity (e.g., from moderate to severe) although older patients (>65yrs) reported less symptoms and respondents receiving HD more than five years and more comorbid respondents reported more symptoms which reached statistical significance.

Symptoms prevalence in this study is comparable to other studies [23] including a systematic review on symptoms prevalence in ESRF patients [24], for prevalence, type of symptoms reported and cohort mean HRQoL. Variation in severity of symptoms in relation to patient characteristics and comorbidities was also comparable [13]. Despite associations between the interdialytic interval and a range of outcomes, although we observed trends, we did not find statistically significant variation in most symptoms over the dialysis day of the week. Explanations for this include changes in volume, cardiac and uraemic markers being of insufficient size to manifest in PROMs, and inadequate sample size. Change in effect size for symptoms were less than 0.5 points on the scale of 0 (none) to 4 (overwhelming): Using distribution-based method ½ SD [25], the observed changes are less than minimum importance difference (MID) although they statistically significant. Although EQ5D-5L was lower in non-HD days, it is less than mean minimum important difference (MID) estimate of 0.074 (range -0.011 to 0.140) for the three-level version of the EQ-5D from eight non-renal longitudinal studies in eleven patient groups [26].

The strengths of this study are that it tackles a previously unexplored question, which could have altered practice in the range of ongoing clinical trials aiming to improve symptoms, and that this study repeatedly sampled individuals over an 18 months period for a range of symptoms. Weaknesses include the assumption of linearity across the levels of severity, which we used visual inspection of the underlying trends (Fig 2) and the application of exploratory non-linear methods during the development of the study to justify. For instruments completed on a dialysis day, we were unable to determine the timing of the completion of the instruments in relation to the dialysis session which can be either before or during or after dialysis session. This may be relevant as patients may feel more unwell immediately after dialysis due to dialysis process itself affecting their symptom burden. However, it is our experience that most patients completed the instruments once they established on the haemodialysis machine. As the participants were allowed to complete the instruments on the day of their choice either on non-dialysis day or dialysis day and they may choose to complete these questionnaires on the day they feel better, this has potential impact on the outcome of symptom burden in relation to dialysis days. We performed multiple testing with different variables with statistical significance considered as a p value less than 0.05, and concluded adjustment for dialysis day of the week is not required as most of the results were non-significant. Correction of p values for multiple testing would further reduce significance, not altering our conclusions, and is therefore is not required.

We identified significant associations between symptom severity in different age groups, comorbidity and time on dialysis, mandating their adjustment in observational studies and potentially in clinical trials as studies have shown this can result in a reduced sample size [27, 28]. However, providing these assessments are consistently performed on a standardized day, either dialysis or non-dialysis day, and comparisons across providers are adjusted for demography, the timing of the assessment in relation to dialysis day of the week does not need to be standardised.

Symptom severity stratified by HD day (dialysis day of the week: HD1, HD2, HD3).

(DOCX)

Click here for additional data file.

Symptom score determined by baseline characteristics adjusted for dialysis day of the week (HD1, HD2, HD3).

(DOCX)

Click here for additional data file.

Baseline patient characteristics of participants who completed 1 instrument and >1 instruments throughout the study.

(DOCX)

Click here for additional data file.

5 common symptoms comparing their severity at baseline for participants who completed only 1 instrument and >1 instruments throughout the study.

(DOCX)

Click here for additional data file.

Multivariable mixed effects linear regression comparing each hemodialysis day after long break, each HD day from HD1, HD day vs non-HD day.

(DOCX)

Click here for additional data file.

STROBE check list for observational study.

(DOCX)

Click here for additional data file.

7 Feb 2022

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Gianpaolo Reboldi, MD, MSc, PhD

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf  and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please include the names of each of the renal centres in the study

3. Thank you for stating the following financial disclosure:

“The Health Foundation (Scaling Up Round 2) funded the SHAREHD study and had no role in this study design, data collection, analysis, interpretation, or writing of the report.”

Please state what role the funders took in the study.  If the funders had no role, please state: "The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript."

If this statement is not correct you must amend it as needed.

Please include this amended Role of Funder statement in your cover letter; we will change the online submission form on your behalf.

4. Thank you for stating the following in the Competing Interests section:

“PH conducts research funded by Vifor Pharma. JF has received speaker honoraria from Fresenius medical care, and conducts research funded by the National Institute of Health Research (NIHR), Vifor Pharma and Novartis. MEW has received speaker honoraria Fresenius and Baxter, has acted on an advisory board for Baxter and has conducted research funded by NIHR. SJW has received book royalties from Wiley and has received funds from NIHR, the Department of Health and Medical Research Council.”

Please confirm that this does not alter your adherence to all PLOS ONE policies on sharing data and materials, by including the following statement: "This does not alter our adherence to  PLOS ONE policies on sharing data and materials.” (as detailed online in our guide for authors http://journals.plos.org/plosone/s/competing-interests).  If there are restrictions on sharing of data and/or materials, please state these. Please note that we cannot proceed with consideration of your article until this information has been declared.

Please include your updated Competing Interests statement in your cover letter; we will change the online submission form on your behalf.

5. We note that you have stated that you will provide repository information for your data at acceptance. Should your manuscript be accepted for publication, we will hold it until you provide the relevant accession numbers or DOIs necessary to access your data. If you wish to make changes to your Data Availability statement, please describe these changes in your cover letter and we will update your Data Availability statement to reflect the information you provide

6. In your Data Availability statement, you have not specified where the minimal data set underlying the results described in your manuscript can be found. PLOS defines a study's minimal data set as the underlying data used to reach the conclusions drawn in the manuscript and any additional data required to replicate the reported study findings in their entirety. All PLOS journals require that the minimal data set be made fully available. For more information about our data policy, please see http://journals.plos.org/plosone/s/data-availability.

Upon re-submitting your revised manuscript, please upload your study’s minimal underlying data set as either Supporting Information files or to a stable, public repository and include the relevant URLs, DOIs, or accession numbers within your revised cover letter. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories. Any potentially identifying patient information must be fully anonymized.

Important: If there are ethical or legal restrictions to sharing your data publicly, please explain these restrictions in detail. Please see our guidelines for more information on what we consider unacceptable restrictions to publicly sharing data: http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions. Note that it is not acceptable for the authors to be the sole named individuals responsible for ensuring data access.

We will update your Data Availability statement to reflect the information you provide in your cover letter.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: No

Reviewer #2: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

Reviewer #2: No

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The manuscript presents secondary analysis of longitudinal data obtained from the SHAREHD stepped wedge CRT. While the study objectives sound interesting, I have the following questions.

1. It is not clear where the actual study trial was published (please mention a Reference!). If not, this manuscript needs a thorough rewrite, describing the trial (using detailed CONSORT items), and the analysis conducted in that context.

2. Please state the desired sample size/power under which the original CRT was powered, and data generated, to have some context. Better to write a full paragraph on what sample size/power was considered, in light of the primary response variable.

3. Mixed-effects linear regression was used to model severity, considering it as a "continuous" variable, whereas, in reality, it should have been an ordinal (response) variable. Ordinal regresson under a generalized linear mixed modeling framework can be readily implemented in SAS, or R. Hence, the analysis require a through redo.

4. Was there no need of considering any interaction terms in the list of covariables controlled for? In particular, understanding the changing patterns with time, given the longitudinal nature of the study.

5. Statements such as effect size 0.1, p 0.014, 95% CI ....require proper punctuations throughout the manuscript. Please edit thoroughly; it doesn't read well in the current form.

Reviewer #2: GENERAL COMMENTS

- The way the study has been designed, it should be considered an observational cohort study – therefore ensure your manuscript adheres to STROBE reporting guidelines

- Please use the current international nomenclature for kidney disease published in the journal Kidney International (Levey et al 2020), such as kidney failure (https://www.kidney-international.org/article/S0085-2538(20)30233-7/fulltext)

- Avoid use of uncommon abbreviations (e.g. DoW)

- Write numbers 0-9 in full (e.g. line 119 ‘6 months’)

ABSTRACT

- Background: further clarify that the rationale for the study relates to the need to know to what extent time and place of PROM completion should be standardised for research and clinical assessment.

- Methods: explain what the trial aimed to investigate because it provides relevant background context to your data collection

- Methods: Readers may not know what POS-S Renal refers to – either write in full or use ‘symptom burden score’

- Results: provide some reference to facilitate interpretation of the reported EQ-5D utility score

- Conclusions: significance of ‘location of completion’ not clear from the findings reported in the Results section

INTRODUCTION

- Explain what SONG-HD is (line 100)

- The statement ‘There is expanding literature that PROMS are not only effected by psychosocial issues, stress, emotions, and co-morbidities, the environment in which the instrument is completed may influence the result’ (lines 104-106) forms a key part of the rationale for this study – it needs more convincing support, rather than a single reference to an article that’s nearly 20 years old.

- ‘Failure to account for any underlying differences in symptom severity due to the haemodialysis schedule or location of completion at data collection,symptom analysis could underestimate the impact of interventions in this patient group’ (lines 108-9): impact could either be under- or overestimated, so better to state that this ‘failure to account….’ may distort evaluations of effectiveness of interventions

- No mention of HRQoL in the study’s aim as formulated in lines 111-3.

METHODS

- ‘Patients participating in SHAREHD were asked to complete instruments at baseline, 6, 12, and 18 months at either at dialysis unit or at home’ (lines 134-5): provide more detail on the extent to which the time and place of PROMs completion was protocolised. If this aspect was not protocolised –i.e. patients were free to complete it when and where they chose—you need to acknowledge the potential impact of this on the relationship between dialysis day and PROMs scores in the Discussion. This is relevant because the time/place participants chose for completion may be related to your outcome measures of interest: people’s likelihood of completing a PROM on a dialysis or non-dialysis day may depend on their symptom burden or QoL at that time (e.g. if their symptoms are worse on non-dialysis days, they may choose not to complete a PROM on those days because they are too poorly and wait for a dialysis day when they tend to feel better).

- ‘We excluded patients who had missing data’ (line 132) – were people excluded if they had any missing data or only when they had a certain level of missingness? Please provide more detail.

- Related to the previous point: provide information on how characteristics of people compared between those in- and excluded in the analysis. Also, if it appears you excluded a substantial number of participants because of missing outcome/PROMs measurements, please reflect in the Discussion on the fact that if and when people complete a PROM may be associated to their symptom burden/QoL, which in turn would impact on your findings (also see my first comment under METHODS).

- In the conclusion section of the abstract, the ‘location of completion’ seems central to the study’s key findings. Yet, it is unclear from the Methods how this aspect was defined, recorded and analysed, and there are no findings presented in the Results that are clearly and explicitly linked to and supporting this conclusion.

- Statistical analysis: State what level at which you considered a finding statistically significant. This level should be corrected for multiple testing (e.g. using Bonferroni correction), especially considering the large number of tests presented in Tables 3 and 4.

- Statistical analysis: related to the previous point, please try to reduce the number of tests, in particular those presented in Table 4. For example, you could consider running an ANOVA to explore whether there was any difference between HD1-3 for each symptom, and only compare specific days if the ANOVA test suggested this was the case. This would enable you to reduce the number of tests for most symptoms in Table 4 from 4 to 2.

- Statistical analysis: the intervention evaluated in the SHAREHD trial was likely to have had an impact on PROMs scores. Therefore, the time in the trial when PROMs measurement were taken (i.e. during control, intervention or sustainability period) needs to be accounted for in the analysis.

RESULTS

- Figure 1: it seems that no participants were excluded due to missing data – is this correct? Please clarify in the text and in the figure.

- Provide footnotes to Table 1 to explain all abbreviations in the table, as well as what the higher education levels refer to, and to clarify that higher Charlson scores represent more comorbidities

- I’m unclear what the relevance is of the analysis presented in Table 3 in relation to the study’s aim. The way it’s been presented seems to suggest the study is interested in exploring the relationship between patient characteristics and symptom scores, rather than between time/place of PROMs completion and PROMs scores – please provide a rationale for this analysis in the Methods and more explicitly relate the findings presented in the Results to the study aim.

DISCUSSION

- ‘Like other studies, we demonstrate that symptom severity is also susceptible tochange, in this study over a period of 18 months’ (line 297-8). This was what you presumably were trying to demonstrate in your SWT but not in the current study. Please clarify this.

- ‘We assume that completion of instruments on non-dialysis days occurred at home, and are unable to determine the timing of the completion of the instruments in relation to the dialysis session. However, it is our experience that most patients completed the instruments once established on the dialysis machine’ (lines 312-5). Clarify this assumption earlier on (in Methods and Results) and reflect in the Discussion how uncertainty within this assumption may have affected your findings and conclusions.

- Related to the previous point: specify what you mean with ‘timing of the completion of the instruments in relation to the dialysis session’ – it seems to refer to the exact timing of completion on a HD day. If this is this case, please clarify why this is relevant to mention as a limitation.

- ‘The high proportion of patients who change the severity of their response…’ (lines 318-9). Apart from it being better to talk about ‘report more severe symptoms’ instead of ‘change the severity of their response’, I’m unsure of the point you’re trying to make. Please consider to rephrase and more clearly link it to the study’s aim and the findings you presented.

- Please include some reflections on how your study and findings relate to the ‘expanding literature’ mentioned in the Introduction (lines 104-6)

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: Yes: Sabine N van der Veer

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

6 May 2022

I've submitted the separate word document for responses to the reviewers. Some of the tables aren't included in the responses below due to formatting issues but those said tables were included in the word document attached with this submission.

Comments from Editor

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming.

We have ensured the revised manuscript adhered to PLOS One’s journal requirement and used PACE digital diagnostic tool for the figures included in this revised manuscript.

2. Please include the names of each of the renal centres in the study

We have included the names of 12 renal centres in the method section and study design.

3. Thank you for stating the following financial disclosure:

“The Health Foundation (Scaling Up Round 2) funded the SHAREHD study and had no role in this study design, data collection, analysis, interpretation, or writing of the report.”

We have stated this disclosure in the manuscript as well as in cover letter. "The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript."

4. Thank you for stating the following in the Competing Interests section:

“PH conducts research funded by Vifor Pharma. JF has received speaker honoraria from Fresenius medical care, and conducts research funded by the National Institute of Health Research (NIHR), Vifor Pharma and Novartis. MEW has received speaker honoraria Fresenius and Baxter, has acted on an advisory board for Baxter and has conducted research funded by NIHR. SJW has received book royalties from Wiley and has received funds from NIHR, the Department of Health and Medical Research Council.”

Please confirm that this does not alter your adherence to all PLOS ONE policies on sharing data and materials, by including the following statement: "This does not alter our adherence to PLOS ONE policies on sharing data and materials.” (as detailed online in our guide for authors http://journals.plos.org/plosone/s/competing-interests). If there are restrictions on sharing of data and/or materials, please state these. Please note that we cannot proceed with consideration of your article until this information has been declared.

We have added this statement to the cover letter.

5. We note that you have stated that you will provide repository information for your data at acceptance. Should your manuscript be accepted for publication, we will hold it until you provide the relevant accession numbers or DOIs necessary to access your data. If you wish to make changes to your Data Availability statement, please describe these changes in your cover letter and we will update your Data Availability statement to reflect the information you provide.

Please see the response to question number 6.

6. In your Data Availability statement, you have not specified where the minimal data set underlying the results described in your manuscript can be found. PLOS defines a study's minimal data set as the underlying data used to reach the conclusions drawn in the manuscript and any additional data required to replicate the reported study findings in their entirety. All PLOS journals require that the minimal data set be made fully available. For more information about our data policy, please see http://journals.plos.org/plosone/s/data-availability.

A minimal dataset required to reach the conclusions drawn from this manuscript required the linkage of identifiable patient information collected during the trial to Hospital Episode Statistics data, which at the time of writing is provided by the NHS Digital Data Access Request Service (NHS DARS, https://digital.nhs.uk/services/data-access-request-service-dars), and then appropriate processing. An application to NHS DARS can be submitted detailing lawful processing of the combined dataset and the period which HES data is required for. NHS DARS would verify appropriate permissions were in place as a result of this process. A data sharing agreement between the relevant parties would allow data to be transferred from the University of Sheffield to NHS DARS and on to those wishing to perform the enclosed analyses. Please contact ctru@sheffield.ac.uk for further information about the unlinked dataset which has the personal information required for linkage.

In response to Reviewer 1:

1. It is not clear where the actual study trial was published (please mention a Reference!). If not, this manuscript needs a thorough rewrite, describing the trial (using detailed CONSORT items), and the analysis conducted in that context.

It was published during the submission period of this manuscript to PLOS ONE. We have updated the references to this trial publication: A breakthrough series collaborative to increase patient participation with hemodialysis tasks: A stepped wedge cluster randomised controlled trial (plos.org) Reference number 14 in Manuscript.

2. Please state the desired sample size/power under which the original CRT was powered, and data generated, to have some context. Better to write a full paragraph on what sample size/power was considered, in light of the primary response variable.

We have included an abridged sample size/power calculation from original CRT (Line number 143-149). A full link to the trial was added in a reference. (Reference number 14)

“The informing clinical trial sample size to determined using a recommended ICC value of 0.05 (20), to have a 90% power to detect an increase in the proportion undertaking five or more haemodialysis-related tasks from 15% to 30% as statistically significant at the 5% two-sided level: 12 clusters of 25 patients, with six clusters randomised at each step of SWCRT was arrived at. In recognition of the background mortality and renal transplantation rate and to mitigate the risk of incomplete data collection, the target recruitment per participating site was doubled to 50 (14).”

3. Mixed-effects linear regression was used to model severity, considering it as a "continuous" variable, whereas, in reality, it should have been an ordinal (response) variable. Ordinal regression under a generalized linear mixed modelling framework can be readily implemented in SAS, or R. Hence, the analysis require a through redo.

Thank you for highlighting this, which we acknowledge. This analysis was informed by the skills, software and time available within the research group and used as a vehicle to train the first author. Prior to the use of mixed effects linear regression, we recognised that an ordered mixed effects logistic regression could be applied, and trialled this in STATA. This was primarily to consider if linearity could be assumed, and reviewed the cut-points of the underlying latent variable (in the analysis presented below, pain). These largely incremented by constant values leading us to believe linearity could be assumed.

Integration method: mvaghermite Integration pts. = 7

Wald chi2(1) = 2.06

Log likelihood = -1608.2202 Prob > chi2 = 0.1517

------------------------------------------------------------------------------

YSQ1 | Odds ratio Std. err. z P>|z| [95% conf. interval]

-------------+----------------------------------------------------------------

HD Day | .9387223 .0414065 -1.43 0.152 .8609761 1.023489

-------------+----------------------------------------------------------------

/cut1 | -1.09009 .1953785 -1.473024 -.7071547

/cut2 | .3218006 .1913726 -.0532829 .6968841

/cut3 | 1.980382 .2056775 1.577262 2.383503

/cut4 | 4.218393 .2698197 3.689556 4.74723

-------------+----------------------------------------------------------------

SHAREHD_ID |

var(_cons)| 2.756961 .4098724 2.06008 3.689583

------------------------------------------------------------------------------

Note: Estimates are transformed only in the first equation to odds ratios.

LR test vs. ologit model: chibar2(01) = 174.86 Prob >= chibar2 = 0.0000

In addition, the output does not lend itself to an assessment of what is clinically meaningful for the patient or clinical audience for the following reasons:

a) The odds ratio of moving from one level to another does not allow the reader to readily link this to the scale of the PROM

b) The estimate (odds ratio) does not allow linkage to a measure of clinically meaningful difference (either a whole level movement or an estimate based on the half-SD rule).

Finally, the second reviewer did propose other linear methods to screen for differences.

4. Was there no need of considering any interaction terms in the list of covariables controlled for? In particular, understanding the changing patterns with time, given the longitudinal nature of the study.

Thank you – this is an important topic for the renal community. To acknowledge the original nature of the SWCRT and therefore factor time, we have re-run the model adjusted for the steps (control and intervention). Symptom score did not change statistically significantly comparing the control period and intervention period. Analysis in relation to dialysis day of the week adjusted for control and intervention period did not result in change in symptom score. By including steps, effectively we also have between 6-12 months differentiation in time.

Because of the number of symptoms and attendant models informing this manuscript we did not feel it was appropriate to try and incorporate any inference around how symptoms change over time – this is a large topic and we hope to report something specific around this in the future. We were concerned that adding interactions could potentially add more type 1 errors.

5. Statements such as effect size 0.1, p=0.014, 95% CI -0.02 to 0.2 etc. ....require proper punctuations throughout the manuscript. Please edit thoroughly; it doesn't read well in the current form.

We have edited the results in the format that was consistent throughout manuscript.

Reviewer 2:

GENERAL COMMENTS

1. The way the study has been designed; it should be considered an observational cohort study – therefore ensure your manuscript adheres to STROBE reporting guidelines

We have submitted a STROBE reporting guideline checklist in supplementary material (S4 table)

2. Please use the current international nomenclature for kidney disease published in the journal Kidney International (Levey et al 2020), such as kidney failure (https://www.kidney-international.org/article/S0085-2538(20)30233-7/fulltext)

Avoid use of uncommon abbreviations (e.g. DoW)

Write numbers 0-9 in full (e.g. line 119 ‘6 months’)

We have edited and changed the abbreviations and numbers as reviewer 2 suggested.

ABSTRACT

3. Background: further clarify that the rationale for the study relates to the need to know to what extent time and place of PROM completion should be standardised for research and clinical assessment.

Due to word limitation for abstract, some of this detail was previously removed but we have edited to clarify the aim of this study with the following: Line number 30-32

“The relationship between dialytic interval and patient reported outcome measures (PROM) has not been quantified and the extent to which dialysis day of PROM completion need to be standardised for is unknown.”

4. Methods: explain what the trial aimed to investigate because it provides relevant background context to your data collection

We have added some background info of SHAREHD in the abstract.

“Three times a week haemodialysis patients participating in a stepped wedge trial to increase patient participation in haemodialysis tasks completed PROMs (POS-S Renal symptom score and EQ5D-5L) at recruitment, six, 12 and 18 months.” Line number 33-35

5. Methods: Readers may not know what POS-S Renal refers to – either write in full or use ‘symptom burden score’

We have addressed these instruments as PROMS symptom score in the abstract.

“PROMs (POS-S Renal symptom score and EQ5D-5L)” Line number 34-35

6. Results: provide some reference to facilitate interpretation of the reported EQ-5D utility score

We have removed the EQ5D utility score from abstract due to the 300 word limitation and result section focused mainly on the main topic of HD schedule and symptoms but it remained in main manuscript.

7. Conclusions: significance of ‘location of completion’ not clear from the findings reported in the Results section

Location of completion is not the same as non-dialysis day of the week (this is an assumption). 70% of patients assigned to non- HD group completed questionnaires at home. We have changed the language to non-dialysis day instead of location.

INTRODUCTION

8. Explain what SONG-HD is (line 100)

I’ve expanded SONG_HD and its purpose.

“When Standardized Outcome in Nephrology (SONG-HD) , consensus group to establish core outcome to be measured and reported in haemodialysis trial,” Line number 74-75

9. The statement ‘There is expanding literature that PROMS are not only affected by psychosocial issues, stress, emotions, and co-morbidities, the environment in which the instrument is completed may influence the result’ (lines 104-106) forms a key part of the rationale for this study – it needs more convincing support, rather than a single reference to an article that’s nearly 20 years old.

There are more contemporary evidence-based literature on effects of psychosocial issues, stress, emotions, patients characteristics influencing PROMS measures which we have referenced. We were only able to identify the very old reference of the effect of environment. We have moved away from the issue of environment to make the focus of the manuscript the schedule.

10. ‘Failure to account for any underlying differences in symptom severity due to the haemodialysis schedule or location of completion at data collection, symptom analysis could underestimate the impact of interventions in this patient group’ (lines 108-9): impact could either be under- or overestimated, so better to state that this ‘failure to account….’ may distort evaluations of effectiveness of intervention

We completely agree that bias can be both ways either over or underestimated and amended this statement in revised manuscript.

“Failure to account for any underlying differences in symptom severity due to the day of instrument completion in relation to the dialysis schedule could bias the impact and effectiveness of interventions in this patient group”. Line number 84-86

11. No mention of HRQoL in the study’s aim as formulated in lines 111-113.

Although we have led the introduction with symptom burden and effect on HRQoL in HD patients, this study aim is to explore the effect of symptom burden as a mean of HRQoL in relation to HD day of the week.

“To quantify this, we used data from a large stepped wedge randomised controlled trial with aim to determine the association between symptom burden and the haemodialysis schedule in three times a week haemodialysis patients and to explore the effect of PROMS completion in dialysis and non dialysis days, accounting for patient characteristics.” Line number 86-89

Methods

12. ‘Patients participating in SHAREHD were asked to complete instruments at baseline, 6, 12, and 18 months at either at dialysis unit or at home’ (lines 134-5): provide more detail on the extent to which the time and place of PROMs completion was protocolised. If this aspect was not protocolised –i.e. patients were free to complete it when and where they chose—you need to acknowledge the potential impact of this on the relationship between dialysis day and PROMs scores in the Discussion. This is relevant because the time/place participants chose for completion may be related to your outcome measures of interest: people’s likelihood of completing a PROM on a dialysis or non-dialysis day may depend on their symptom burden or QoL at that time (e.g. if their symptoms are worse on non-dialysis days, they may choose not to complete a PROM on those days because they are too poorly and wait for a dialysis day when they tend to feel better).

Research protocol did not specify where to compete the questionnaires. But participants were asked to tick where they completed the instruments: either at home, dialysis unit or clinic. It is possible that frailer people might take instruments home to complete. But our analyses showed that only 16.7% of older people (>65yr) completed the instruments on non-dialysis days and 12.8% of patients with high CCI score (score >5) completed on non-dialysis days. In addition, our recommendation mitigates against this (older patients >65 have less symptoms in our analysis). I've included the relevant analyses and tables in the word documents attached with this submission (document labeled as responses to reviewers)

13. ‘We excluded patients who had missing data’ (line 132) – were people excluded if they had any missing data or only when they had a certain level of missingness? Please provide more detail.

We excluded missing data for the adjustment covariates and mechanism of missingness for comorbidity is failure to link via NHS Digital – assumed at random.

14. Related to the previous point: provide information on how characteristics of people compared between those in- and excluded in the analysis. Also, if it appears you excluded a substantial number of participants because of missing outcome/PROMs measurements, please reflect in the Discussion on the fact that if and when people complete a PROM may be associated to their symptom burden/QoL, which in turn would impact on your findings (also see my first comment under METHODS).

We have 62 participants who completed only 1 instrument throughout the study period. Comparing this cohort with patients who completed the instruments more than once in follow up period, patient characteristics are similar (perhaps slightly younger and less co morbid) and severity of symptom burden at baseline is less (we have provided 5 most prevalence symptoms as an example). Therefore, we did not think this will impact the outcome of the analysis we performed to demonstrate the relationship between symptom burden and dialysis day of the week. I've included the relevant analyses and tables in the word documents attached with this submission (document labeled as responses to reviewers)

15. In the conclusion section of the abstract, the ‘location of completion’ seems central to the study’s key findings. Yet, it is unclear from the Methods how this aspect was defined, recorded and analysed, and there are no findings presented in the Results that are clearly and explicitly linked to and supporting this conclusion.

Majority of patients who were assigned to non HD days completed instruments at home and HD days at renal units. We changed the language to non HD day in line with wider question of dialysis schedule.

16. Statistical analysis: State what level at which you considered a finding statistically significant. This level should be corrected for multiple testing (e.g. using Bonferroni correction), especially considering the large number of tests presented in Tables 3 and 4.

We performed multiple comparison testing with different variable and therefore, we agreed that P value should be lowered. But given that most results were negative, assumption of P value might probably caveat. Based on our analysis, we don’t need to adjust for Dialysis day of the week. Therefore, we believe a Bonferroni correction would not change that and if anything, its significant will be less significant.

17. Statistical analysis: related to the previous point, please try to reduce the number of tests, in particular those presented in Table 4. For example, you could consider running an ANOVA to explore whether there was any difference between HD1-3 for each symptom, and only compare specific days if the ANOVA test suggested this was the case. This would enable you to reduce the number of tests for most symptoms in Table 4 from 4 to 2.

We agree that there are a lot of tests and that this introduces a range of problems. Although an ANOVA can be performed to compare symptoms change between HD1 vs HD2/HD3 (e.g. categorical vs continuous), it does not fully capitalise on the statistical efficiency of treating HD days as continuous variable as we do in one subset of our models.

Although doing many tests does introduce the risk of a type I error, we have generally said that the timing of assessment in relation to the dialysis schedule does not need to be accounted for (which is more likely to be a type II error)

The other threat of so many models is reader fatigue. In order to reduce the number of tables, we have reported these results in figure form in the main manuscript and I’ve submitted the table as supplementary material. Figure 3 and S3 Table

18. Statistical analysis: the intervention evaluated in the SHAREHD trial was likely to have had an impact on PROMs scores. Therefore, the time in the trial when PROMs measurement were taken (i.e. during control, intervention or sustainability period) needs to be accounted for in the analysis.

We have re-run the model adjusted for steps (control and intervention). Symptom score did not change statistically significantly comparing the control period and intervention period and these are described in the manuscript (Line number 248-250) . Analysis in relation to dialysis day of the week adjusted for control and intervention period did not result in change in symptom score.

“Symptom score comparing the control period and intervention period predicting the symptom severity independently in this model did not reach statistical significance”.

RESULTS

19. Figure 1: it seems that no participants were excluded due to missing data – is this correct? Please clarify in the text and in the figure.

We excluded missing data for the adjustment covariates and but no missing data for PROMS instrument itself. We have edited the flow diagram figure for more clarification.

20. Provide footnotes to Table 1 to explain all abbreviations in the table, as well as what the higher education levels refer to, and to clarify that higher Charlson scores represent more comorbidities

We have expanded the abbreviations in table and foot notes added to clarify education levels and comorbidities.

21. I’m unclear what the relevance is of the analysis presented in Table 3 in relation to the study’s aim. The way it’s been presented seems to suggest the study is interested in exploring the relationship between patient characteristics and symptom scores, rather than between time/place of PROMs completion and PROMs scores – please provide a rationale for this analysis in the Methods and more explicitly relate the findings presented in the Results to the study aim.

We included this relationship between patient characteristics and symptoms so that others designing the clinical trials can refer to it in order to assist with power calculations. This will mandate their adjustment in observational studies and potentially in clinical trials as studies have shown this can result in a reduced sample size. We appreciate the reviewer’s point and have moved this table to supplementary material to reduce the number of tables as we agreed that these finding are not directly linked to the study aim.

DISCUSSION

22. ‘Like other studies, we demonstrate that symptom severity is also susceptible to change, in this study over a period of 18 months’ (line 297-8). This was what you presumably were trying to demonstrate in your SWT but not in the current study. Please clarify this.

We agreed that the main purpose of this study is not to demonstrate the change in symptom burden over 18 months. We demonstrated symptoms change over a week according to dialysis schedules although it was not statistically significant. We apologise that this statement introduces confusion. Therefore, we decided not to include this statement in our revised manuscript.

23. ‘We assume that completion of instruments on non-dialysis days occurred at home, and are unable to determine the timing of the completion of the instruments in relation to the dialysis session. However, it is our experience that most patients completed the instruments once established on the dialysis machine’ (lines 312-5). Clarify this assumption earlier on (in Methods and Results) and reflect in the Discussion how uncertainty within this assumption may have affected your findings and conclusions.

We have changed the language of location of completion to non-dialysis days and reflected this in discussion. Line number 308-311

“As the participants were allowed to complete the instruments on the day of their choice either on non-dialysis day or dialysis day and they may choose to complete these questionnaires on the day they feel better, this has potential impact on the outcome of symptom burden in relation to dialysis days.”

24. Related to the previous point: specify what you mean with ‘timing of the completion of the instruments in relation to the dialysis session’ – it seems to refer to the exact timing of completion on a HD day. If this is this case, please clarify why this is relevant to mention as a limitation.

We have expanded this in discussion session. (line number 303-308)

“For instruments completed on a dialysis day we are unable to determine the timing of the completion of the instruments in relation to the dialysis session which can be either before or during or after dialysis session. This may be relevant as patients may feel more unwell immediately after dialysis due to dialysis process itself affecting their symptom burden. However, it is our experience that most patients completed the instruments once established on the dialysis machine.”

25. ‘The high proportion of patients who change the severity of their response…’ (lines 318-9). Apart from it being better to talk about ‘report more severe symptoms’ instead of ‘change the severity of their response’, I’m unsure of the point you’re trying to make. Please consider to rephrase and more clearly link it to the study’s aim and the findings you presented.

As we answered in question number 22, we agreed that this study aim was not to demonstrate change in symptom burden, therefore, we decided not to include this statement.

26. Please include some reflections on how your study and findings relate to the ‘expanding literature’ mentioned in the Introduction (lines 104-6)

We added additional reflection and conclusion related to this statement.

“Symptoms prevalence in this study is comparable to other studies (22) including a systematic review on symptoms prevalence in ESRF patients (23), both in prevalence, type of symptoms reported. How severity of symptom varies in relation to patient characteristics and comorbidities was also comparable (13) . Line number 285-288

“We identified significant associations between symptom severity in different age groups, comorbidity and time on dialysis, mandating their adjustment in observational studies and potentially in clinical trials as studies have shown this can result in a reduced sample size(26)(27).” Line number 312-315

Submitted filename: Responses to reviewers.docx

Click here for additional data file.

30 Jun 2022

13 Aug 2022

REBUTTAL LETTER FOR SUBMISSION OF REVISED MANUSCRIPT

To The Editor in Chief,

PLOS ONE Journal Date: 11/08/2022

We are grateful for your and the reviewers’ comment on the manuscript “Symptom Burden According To Dialysis Day Of The Week In Three Times A Week Haemodialysis Patients “. We’ve revised and modified the manuscript according to reviewers’ critiques. In the following, we address each of their comment.

Journal Requirements:

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

I’ve reviewed the reference list and they are complete. There was no citation of retracted papers.

Response to reviewer 2

1. Your response to my point 11 ("No mention of HRQoL in the study’s aim") is unclear: what does 'effect of symptom burden as a mean of HRQoL' mean? I strongly recommend you say something about the role/importance of HRQoL at the end of the introduction, so that this doesn't come as a surprise to readers when they get to the Methods and Results (e.g. that it was a secondary outcome of interest?)

We’ve added the role of HRQoL in this study. (line number 83-88)

“HRQoL has been shown to be a predictor of morbidity and mortality in haemodialysis patients (14)(15) and HRQoL measures play an important role in evaluating cost effectiveness of treatment. Failure to account for any underlying differences in symptom severity due to the day of instrument completion in relation to the dialysis schedule could bias the impact and effectiveness of interventions for symptoms and HRQoL, which have been prioritised by the patients and clinicians(2) and could lead to failure of new treatment or interventions to be approved.”

2. Your response to my point 14 (“provide information on how characteristics of people compared between those in- and excluded in the analysis”) is adequate. However, I suggest you include the relevant analyses/tables as supplementary materials, so that interested readers also have access to this information.

We’ve added this in result section and provided supplementary tables. (line number 234-237 and Supplementary table 3 and 4).

“There were 62 participants who completed only one instrument throughout the study period. Comparing this cohort with patients who completed the instruments more than once in follow up period, patient characteristics and severity of symptom burden at baseline were similar (S3 Table, S4 Table).”

3. Your response to my point 16 (“State the level at which you considered a finding statistically significant”) is unclear: it’s hard to follow your argument for why statistical significance is (or isn’t?) relevant. But regardless of what you decide to use as a threshold, please clarify for readers in the manuscript (and not just for me in the response letter) what your definition of ‘statistical significance’ is and how you accounted for multiple testing (or why this is not required).

We’ve clarified this in method as well as in discussion section. (line number 178, 309-313)

“P value < 0.05 was considered the threshold for statistical significance.”

“We performed multiple testing with different variables with statistical significance considered as a p value less than 0.05, and concluded adjustment for dialysis day of the week is not required as most of the results were non-significant. Correction of p values for multiple testing would further reduce significance, not altering our conclusions, and is therefore is not required.”

4. There are several grammatical/style errors throughout the newly added sections in the abstract and main text - the manuscript therefore requires one, last thorough round of text editing by a native English speaker prior to submission.

We hope we have identified and corrected the errors.

Submitted filename: Response to reviewer(s).docx

Click here for additional data file.

1 Sep 2022

Symptom burden according to dialysis day of the week in three times a week Haemodialysis patients

PONE-D-21-25372R2

Dear Dr. Hnynn Si,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Gianpaolo Reboldi, MD, MSc, PhD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

6 Sep 2022

PONE-D-21-25372R2

Symptom Burden According To Dialysis Day Of The Week In Three Times A Week Haemodialysis Patients

Dear Dr. Hnynn Si:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Prof Gianpaolo Reboldi

Academic Editor

PLOS ONE