Introduction
Dialysis initiation is a stressful experience for patients, with the initial months of treatment being critical in terms of both adaptation and mortality [
1‐
4]. Thus, it is not surprising that health-related quality of life in patients with kidney dysfunction requiring dialysis (KDRD) is substantially lower than the general population, a finding consistent across three continents in one analysis from the Dialysis Outcomes and Practice Patterns Study (DOPPS) [
5]. Symptoms of anxiety and/or depression are present in 25–50% of patients initiating hemodialysis (HD) [
6]. In a cross-sectional study involving patients with prevalent KDRD on maintenance HD, Weisbord et al. used the Dialysis Symptom Index (DSI) instrument and showed that 79% of the patients had symptoms of dry skin, 69% had fatigue, 50% had bone or joint pain, 49% had dry mouth, 45% had insomnia, 20% had feelings of sadness, and 31% had feelings of anxiety [
7]. However, few studies examined longitudinal patient-reported psychosomatic symptomatology after HD initiation and little is known about processes of care that impact psychological burden of HD [
8]. A few recent studies in patients with KDRD showed that patient-reported symptom burden, emotional well-being, and spiritual well-being vary widely from month to month regardless of whether overall trajectories improve or worsen over time [
9‐
11]. Therefore, serial collection of patient-reported health-related quality of life is a requisite for an accurate depiction of a full spectrum and dynamic of physical and emotional symptoms.
Dialysis initiation provokes substantial changes in lifestyle that require significant adjustments due to the intrusiveness of disease treatment into multiple life domains, which may underlie the greater than expected rates of anxiety [
12]. Consequently, it has been hypothesized that personalized HD therapy consisting of less frequent HD treatments at dialysis initiation in those who have suitable levels of residual kidney function, with subsequent adaptations in dialysis frequency to thrice-weekly or more often HD as residual kidney function declines—i.e., incremental HD—could be conducive to better adjustment to life style changes and better health-related quality of life. In spite of the alluring attributes of incremental HD rendered by other potential benefits such as better or non-inferior patient survival for significant economic savings, high-quality research focused on clinical outcomes and patient-reported outcomes with incremental HD is scarce [
13,
14]. Registry-based studies showed that about 30% of patients with incident KDRD could be treated with an initial schedule of less frequent HD based on their residual kidney function and morbidity profile [
15], yet incremental HD remains grossly underused in the United States [
16], possibly due to lack of clinical trials to validate the safety of this treatment approach as well as other potential barriers reviewed in former publications [
17].
Current data on incremental HD and patient outcomes is based on observational, retrospective studies and the results are mixed. In one study done in China, quality of life scores did not differ between the twice-weekly and thrice-weekly HD groups [
18]. Other studies, however, surmised that incremental-start HD, by virtue of better preservation of residual kidney function, confers better quality of life [
19,
20]. Thus, comparative effectiveness research in randomized clinical trials needs to be performed to elucidate, among other outcomes, longitudinal differences in psychosomatic symptomatology between incremental HD and conventional HD.
We undertook the first pilot clinical trial of incremental-start HD in patients with incident KDRD in the US (NCT 03740048). The primary objective of the TWOPLUS pilot clinical trial was to test the feasibility of implementing a schedule of incremental HD at outpatient dialysis facilities using a protocol of blood tests and timed urine collection that was embedded in usual clinical workflow, the results of which were published elsewhere [
21]. Exploratory analyses compared patient-reported outcomes using three instruments, the results of which are presented here. Using semi-structured interviews, we also evaluated participant’s perceptions on the intervention and provider perceived barriers to incremental HD.
Discussion
This pilot study showed dynamic, longitudinal changes in patient-reported outcomes with a trend for differences noted between domains assessed and between treatment groups. Specifically, the DSI improved (i.e., scores decreased) over time in both treatment groups, with higher improvements seen with thrice-weekly HD. In contrast, symptoms of anxiety (GAD-7 score) and depression (PHQ-9 score) were better (i.e., lower scores) while on schedules of twice-weekly HD, with results slightly moderated after adjustment for baseline covariates.
Longitudinal changes in patient-reported symptomatology and satisfaction with their care in people afflicted with kidney disease has received little attention [
27]. However, the importance of including patient-reported outcome measures that illustrate patient priorities for their healthcare has garnered growing recognition. Clinical trials are increasingly analyzing patient’s perspectives on treatment-related physical, functional and psychological impact [
28]. Outcomes important for the patients, and characterized by the patients, have the potential to influence healthcare policy and thus change practice paradigms [
29‐
31]. Thus, patient translation of treatment effects is becoming an essential tool to guide better allocation of funds and maximize the impact of the results for patients and society [
32]. With regards to KDRD, the period of dialysis initiation is demarcated by an abrupt decline in individual’s independence and a surge in morbidity, with excess mortality befalling during the first 6–12 months of HD therapy [
33]. One potential approach to mitigate some of the initial psychological burdens at dialysis initiation is to adopt the incremental approach, adding HD time or HD sessions as kidney residuals wane. To date, the effects of incrementally-prescribed twice-weekly HD and residual kidney function on health-related quality of life remain poorly delineated. Based on cross-sectional data from the China Dialysis Outcomes and Practice Patterns Study, patients dialyzing two times per week (26% of the patients in China) had longer treatment times and lower standardized Kt/V, but similar quality of life scores [
18]. In a prospective cohort study of 734 patients with incident KDRD in the US, those with urine output at baseline (defined as producing at least 250 mL/day) reported overall better QOL (
P = 0.05) and less dietary restriction (
P = 0.05) [
19]. Thus, longitudinal comparisons of patient-reported outcomes based on HD treatment schedule remains of particular interest. Thus, longitudinal comparisons of patient-reported outcomes based on HD treatment schedule remains of particular interest.
Notwithstanding the growing recognition of patient-reported symptoms and experiences in clinical research, the approach to interpreting patient-reported outcomes in chronically ill patients is challenging as different measures may assess varied aspects of patients’ experiences. This pilot study generated a few important observations. First, there could be a dichotomy in patient-reported experiences based on domains assessed. While items assessed with the DSI instrument indicated treatment-group independent improvement in the aggregate of physical and emotional symptoms, physiological well-being followed a treatment-dependent trend with lower burden of anxiety-related psychological symptoms being reported during the time period of twice-weekly HD. Based on these findings, we surmise that the timing of questionnaire administration should be an important consideration in the design of future, larger clinical trials. Prior studies pointed to multidimensional features of patient-reported quality of life measures that vary over time, emphasizing the need for serial assessments [
9,
34,
35]. For a clinical trial of twice-weekly HD start vs thrice-weekly HD, eliciting patient-reported symptoms
after conversion from twice-weekly to thrice-weekly HD will miss the opportunity to properly characterize patient-reported outcomes by different schedules of HD treatment. Given that transition from twice- to thrice-weekly HD is patient dependent and may occur at any time point during KDRD trajectory, frequent rather than distant time points of questionnaire administration ought to be part of the study design. As such, sensible selection of short, validated questionnaires for frequent instrument administration will be necessary to avoid respondent burden and minimize the risk of missing data. Second, results interpretation may prove to be problematic. In this pilot study, changes in DSI score were larger (i.e., better) with thrice-weekly HD, possibly because straightway introduction of thrice-weekly HD on a background of long-standing kidney dysfunction afforded swifter improvement in symptoms associated with uremia (e.g., appetite) than twice-weekly HD. In contrast, changes in psychological symptoms were better with twice-weekly HD. Moreover, a response shift phenomenon may occur among patients treated with thrice-weekly HD wherein patients experience a change in the meaning of their quality of life as a form of coping with illness. In such instance, patient-reported scores may improve not necessarily because quality of life per se has improved, diminishing the chance of finding differences between treatment groups even when differences may exist [
36]. Future studies should encompass the
then-test to assess whether, how and to what extent response shift occurred [
37]. Statistical analysis aside, what symptom domain matters more to patients in general and, more importantly, to each individual and at different stages in a patient’s disease trajectory should remain at the core of deciding dialysis prescription. Therefore, while patient-reported instruments are useful tools for clinical investigators, providers ought to continue to adapt clinical care based on each individual’s needs. Last, feedback obtained from participants helped the investigators understand what study-related processes of care require refinement. For example, the pilot study entailed inter-dialysis urine collection during follow-up, with timeframes of urine collection ranging from 45 to 68 hours. Employing equations of calculating residual kidney function based on 24-hour urine collection will likely decrease procedure burden on the patients [
38]. Frequent interaction with patients and reminder to perform urine collection were reported to help with task completion, making this strategy an important and simple tool to increase participant adherence to study protocol.
Results ought to be discussed through the lens of what constitutes a minimal important change, which is the smallest change in score in the construct measured that patients perceive as important [
39]. Importantly, the minimal important change is not a fixed characteristic of any patient-reported outcome measure; instead, it can vary across populations, disease severity, settings, study designs and analyses used to estimate this metric. Prior work suggested a minimum change in DSI score of 7 points [
40], in GAD-7 score of 4 points [
41], and in PHQ-9 score of 5 points [
42] as the cut-offs for minimal important change. Nevertheless, this metric has not been expressively elicited in patients treated with incremental HD. Therefore, with future studies, it will be important to establish a minimal important change anchored in the experiences and perceptions of patients treated with incremental HD.
Our study has limitations. Our study was not powered to detect statistical differences in patient-reported outcomes, but rather to explore the feasibility of implementing our trial procedures and to inform a power calculation for a future randomized controlled trial [
22]; thus, the results ought to be interpreted with caution. Due to the nature of the intervention, patients and providers were not blinded to allocated treatments, thereby weakening construct validity. The study was conducted at a single health system organization, thus limiting its generalizability. Receipt of HD according to the conventional thrice-weekly HD prior to study enrollment may have influenced patient-reported outcomes, albeit the extent to which this occurred remains speculative. In our pilot study, transition from twice- to thrice-weekly HD was defined a priori while in real life, the timing of transition cannot be precisely foretold. Future and larger clinical trials of incremental HD will need to emulate real-life prescription of incremental HD. Future studies should also analyze longitudinal effects on psychological burdens related to kidney function-tailored HD frequency. It is conceivable that while an initial HD schedule of twice-weekly HD may subtract from the initial stress and burden associated with thrice-weekly HD, the trade-off may consist in a later episode of anxiety related to increasing HD frequency. A recent pilot study done in the UK reported no signal of benefit of incremental HD, compared to conventional HD, in terms of patient-reported outcomes measured with EuroQol 5D-5L, PHQ-9, Illness Intrusiveness Rating Scale and Montreal Cognitive Assessment measured at baseline and months 6 and 12 [
43]. However, in the study done in the UK, patient-reported outcomes were collected at fixed time intervals and many of the patients originally treated with twice-weekly HD may had already experienced conversion to thrice-weekly HD at the time of questionnaire administration. Overall, the current studies cannot be interpreted as presence—or lack thereof—of a monotonic association between incremental HD and psychological symptoms until future larger clinical trials will include frequent measurements of patient-reported outcomes to detect if dynamic effects on psychological burdens exist.
In conclusion, this study suggests the prescribed HD treatment frequency may impact patient-reported psychological symptoms, with more HD dependency causing higher symptom burden of depression and anxiety. This approach, however, needs to be subjected to rigorous randomized clinical trials.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.