RESEARCH ARTICLE
Effect of Nocturnal Hemodialysis versus
Conventional Hemodialysis on End-Stage
Renal Disease: A Meta-Analysis and
Systematic Review
Fangjie Liu
1☯
, Yiting Sun
2☯
, Tianhua Xu
1
, Li Sun
1
, Linlin Liu
1
, Wei Sun
3
, Xin Feng
4
,
Jianfei Ma
1
, Lining Wang
1
, Li Yao
1
*
1 Department of Nephrology, The First Hospital of China Medical University, Shenyang, Liaoning, China,
2 Department of Clinical Medicine, China Medical University, Shenyang, Liaoning, China, 3 Department of
General Surgery, Shengjing Hospital of China Medical University, Shenyang, Liaoning, China, 4 Blood
Purification Center, Liaoning Electric Power Center Hospital, Shenyang, Liaoning, China
☯ These authors contributed equally to this work.
*
liyao_cmu@163.com
Abstract
Objectives
The purpose of this study is to assess the efficacy and safety of nocturnal hemodialysis on
end-stage renal disease (ESRD) patients.
Methods
We searched Medline, EmBase, and the Cochrance Central Register of Controlled Trials for
studies up to January 2016. Analysis was done to compare variant outcomes of different
hemodialysis schedules, including mortality, cardiovascular-associated variables, uremia-
associated variables, quality of life (QOL), side-effects, and drug usage.
Results
We collected and analyzed the results of 28 studies involving 22,508 patients in our meta-
analysis. The mortality results in this meta-analysis indicated that the nocturnal hemodialy-
sis (NHD) group was not significantly different from conventional hemodialysis (CHD) group
(Mortality: OR: 0.75; 95% confidence intervals (CIs): 0.52 to 1.10; p = 0.145), but the CHD
group had significantly fewer number of hospitalizations than the NHD group (OR: 1.54;
95%CI: 1.32 to 1.79; p<0.001). NHD was superior to CHD for cardiovascular-associated
(left ventricular hypertrophy [LVH]: SMD: -0.39; 95%CI: -0.68 to -0.10; p = 0.009, left ventric-
ular hypertrophy index [LVHI]: SMD: -0.64; 95%CI: -0.83 to -0.46; p<0.001) and uremia-
associated intervention results (Serum albumin: SMD: 0.89; 95%CI: 0.41 to 1.36; p<0.001).
For the assessment of quality of life, NHD treatment significantly improved the patients’
QOL only for SF36-Physical Components Summary (SMD: 0.43; 95%CI: 0.26 to 0.60;
p<0.001). NHD intervention was relatively better than CHD for anti-hypertensive drug usage
PLOS ONE | DOI:10.1371/journal.pone.0169203 January 20, 2017 1 / 15
a1111111111
a1111111111
a1111111111
a1111111111
a1111111111
OPEN ACCESS
Citation: Liu F, Sun Y, Xu T, Sun L, Liu L, Sun W, et
al. (2017) Effect of Nocturnal Hemodialysis versus
Conventional Hemodialysis on End-Stage Renal
Disease: A Meta-Analysis and Systematic Review.
PLoS ONE 12(1): e0169203. doi:10.1371/journal.
pone.0169203
Editor: Ping-Hsun Wu, Kaohsiung Medical
University Hospital, TAIWAN
Received: June 22, 2016
Accepted: December 12, 2016
Published: January 20, 2017
Copyright: © 2017 Liu et al. This is an open access
article distributed under the terms of the
Creative
Commons Attribution License
, which permits
unrestricted use, distribution, and reproduction in
any medium, provided the original author and
source are credited.
Data Availability Statement: All relevant data are
within the paper and its Supporting Information
files.
Funding: This work was supported by a grant from
the scientific research project for the Universities of
Educational Commission of Liaoning Province of
China (Grant No. L2011134), the Key Social
Development Program of Science and Technology
Commission of Liaoning Province of China (Grant
No. 2013225303), the Key Social Development
Program of Science and Technology Commission
(SMD: -0.48; 95%CI: -0.91 to -0.05; p = 0.005), and there was no difference between groups
in our side-effects assessment.
Conclusion
NHD and CHD performed similarly in terms of ESRD patients’ mortality and side-effects.
NHD was superior to CHD for cardiovascular-associated and uremia-associated results,
QOL, and drug usage; for number of hospitalizations, CHD was relatively better than NHD.
Introduction
End-stage renal disease (ESRD) is a chronic and progressive decline in kidney function, which
will eventually lead to uremia and death if it is not treated properly [
1]. However, with a prog-
ress of technology in past decades, the mortality have not improved significantly and exceeding
20% in chronic hemodialysis patients [
2, 3]. Cardiovascular events are the main driving force
for this high mortality. Therefore, there is a need for new methods to improve ESRD patients’
cardiovascular and mortality risk.
There are currently two main methods for treatment of ESRD patients. The first, renal
transplantation, is a permanent method to cure ESRD patients, however, that means ESRD
patients have an issue of having a proper kidney source, thus it has limited application [
4, 5].
The second, hemodialysis, is applied worldwide but has a high risk of cardiovascular complica-
tions and significantly reduces the quality of life of patients [6, 7]. Nocturnal hemodialysis
(NHD) is an important branch of hemodialysis [
8, 9]. The schedule for nocturnal dialysis is
3–7 times per week, 7–8 hours every time. This approach extends the effective duration of dial-
ysis without affecting the patient’s daytime activities making it more convenient as a method
of treatment. This approach has been widely used in Canada; however, the clinical results still
require further examination. Dialysis-related disease is defined as the complications caused by
long-term dialysis on ESRD patients; cardiovascular disease is the leading cause of death for
ESRD patients [
10–13].
Previously, several systemic reviews analyzed the mortality, blood pressure, and urinary-
related indexes of NHD for ESRD patients [
14–16]. However, the qualities of included studies
were relatively low and not comprehensive evaluated all relevant clinical outcomes. Our
research is up to date with recently published research and analyzes the effects of NHD by
mortality, cardiovascular-related variables, uremia-related variables, quality of life, side-effects,
and drug usage to provide better insight in clinical choices for dialysis methods.
Methods
Search strategy and selection criteria
This review was conducted and reported according to the Preferred Reporting Items for Sys-
tematic Reviews and Meta-Analysis Statement [
17] issued in 2009. Any studies that examined
NHD versus conventional hemodialysis (CHD) on ESRD patients were eligible for inclusion
in our study with no restrictions placed on language or publication status (published, or in
press). We searched the Medline, EmBase, and Cochrane Library electronic databases for arti-
cles published through January 2016 and used “nocturnal”, “dialysis”, “hemodialysis”, and
“controlled trials” as the search keywords. We also conducted manual searches of reference
lists from all relevant original articles and reviews to identify additional eligible studies.
Hemodialysis on End-Stage Renal Disease
PLOS ONE | DOI:10.1371/journal.pone.0169203 January 20, 2017 2 / 15
of Shenyang of China (Grant No. F16-206-9-04),
the Key Social Development Program of Science
and Technology Commission of Liaoning Province
of China (Grant No. 201404046). The funders had
no role in study design, data collection and
analysis, decision to publish, or preparation of the
manuscript.
Competing Interests: The authors have declared
that no competing interests exist.
A literature search was undertaken independently by 2 authors and any inconsistencies
were settled via group discussion. A study was eligible for inclusion if the following criteria
were met: (1) the trial investigated nocturnal hemodialysis NHD versus conventional hemodi-
alysis CHD; (2) all of patients included with ESRD; and (3) the outcomes variable included
one of the following: mortality, cardiovascular-associated variables, uremia-associated vari-
ables, quality of life, side-effect, and drug usage. Case series, reviews, and editorials were
excluded.
Data collection and quality assessment
Two reviewers independently extracted all data with disagreements resolved in consultation
with third-party investigators. The following items were extracted from the included articles:
first author, publication year, country, location or data source, study design, sample size, dis-
ease status, mean age, gender proportion, mean duration of dialysis, Dialysis session, and
reported outcomes. The outcome assessments included: mortality, cardiovascular-associated
variables, uremia-associated variables, quality of life, side-effects, and drug usage. In analysis,
the numerical changes between, before, and after dialysis of statistical indicators had priority
to be adopted, if not, the dialysis numerical indicators after dialysis was adopted. In addition,
the numerical units were adjusted for consistency, such as g/L and g/dL. Two reviewers inde-
pendently assessed the quality of included studies according to the Cochrane risk of bias tool
in the following six domains: selection, performance, detection, attrition, reporting and other
bias [
18].
Statistical analysis
For our meta-analysis, we used the inverse variance method to pool continuous data and the
Mantel-Haenszel method for dichotomous data; the results are presented as standardized
mean difference (SMD) with 95% confidence intervals (CIs) and odds ratio (OR) with 95%
CIs. The I
2
statistic was calculated to evaluate the extent of variability attributable to statistical
heterogeneity between trials. In the absence of statistical heterogeneity (I
2
50%), we used a
fixed-effect model, otherwise we used a random-effect model for traditional meta-analysis
[
19]. To investigate the sources of heterogeneity, predefined subgroup analysis were per-
formed: dialysis schedule and design bias. We assessed for publication bias using the Begg-
Mazumdar [20] and Egger’s test [21]. A non-parametric “Trim and Fill” method of assessing
publication bias was applied if needed [
22]. All tests were two tailed, and a p value of less than
0.05 was deemed statistically significant. We analyzed the data using Review Manager (Version
5.3) and STATA (Version 12.0).
Results
Our research returned 201 results after removing duplicates, from which we collected 28 trials
in our meta-analysis (
Fig 1). After a full text review, the reasons for exclusion of literature
included non-controlled trials, other intervention interference, other similar diseases, and lack
of desired outcome assessments. The general characteristics of the included studies are pre-
sented in
Table 1. In this research, included studies were mainly published in Canada, China,
the United States, the United Kingdom, Australia, and Turkey. The study design included
eight randomized controlled trials (RCTs) [
23–30], seven quasi-RCT [31–37], and thirteen
observational studies [
38–50].
A total number of 22,508 ESRD patients were examined. The average reported age of
patients was between 40–60 years while two studies did not mention the patients’ ages [
31, 32].
The number of men was slightly greater than the number of women. The follow-up time
Hemodialysis on End-Stage Renal Disease
PLOS ONE | DOI:10.1371/journal.pone.0169203 January 20, 2017 3 / 15
duration was 6 months to 36 months. The schedule for NHD was 3 nights/week or 6–7 nights/
week, and 3 times/week for CHD. The summary graph of risk of bias for each study is shown
in
Fig 2.
In our meta-analysis, mortality results were not significantly different between the NHD
group and the CHD group (OR: 0.75; 95%CI: 0.52 to 1.10; p = 0.145). For number of hospitali-
zations, the CHD group had significantly fewer than NHD group (OR: 1.54; 95%CI: 1.32 to
1.79; p<0.001); in addition, there was no significant difference between the two groups in the
number of infection hospitalizations (OR: 1.60; 95%CI: 0.48 to 5.35; p = 0.445).
Within cardiovascular-related variables, left ventricular hypertrophy (LVH, unit: g) and its
index (LVHI, unit: g/m
2
) results both indicate the NHD group has significantly fewer occur-
rences than the CHD group (LVH: SMD: -0.39; 95%CI: -0.68 to -0.10; p = 0.009, LVHI: SMD:
-0.64; 95%CI: -0.83 to -0.46; p<0.001). For the control of blood pressure, systolic blood pres-
sure results also show the NHD group is significantly better than the CHD group (Random
model: SMD: -0.33; 95%CI: -0.49 to -0.18; p<0.001, Fixed model: SMD: -0.17; 95%CI: -0.24 to
-0.1; p<0.001). The Diastolic blood pressure index also shows the NHD group is significantly
better than the CHD group (Diastolic blood pressure: SMD: -.032; 95%CI: -0.48 to -0.15;
p<0.001, Mean arterial pressure: SMD: -0.69; 95%CI: -1.19 to -0.19; p = 0.007, Pulse pressure:
SMD: -0.43; 95%CI: -0.75 to -0.12; p = 0.007).
For uremia-related variables, the concentration of serum albumin of the NHD group was sig-
nificantly greater than the CHD group (SMD: 0.89; 95%CI: 0.41 to 1.36; p<0.001); the concentra-
tion of serum hemoglobin of the NHD group was also significantly greater than the CHD group
(SMD: 0.42; 95%CI: 0.05 to 0.78; p = 0.025). The urea clearance index in the NHD group was sig-
nificantly higher than the CHD group (SMD: 2.61; 95%CI: 1.76 to 3.46; p<0.001), and urea
reduction ratio was also better in the NHD group (SMD: 1.39; 95%CI: 0.49 to 2.30; p = 0.003).
Fig 1. Flow diagram. PRISMA flow diagram.
doi:10.1371/journal.pone.0169203.g001
Hemodialysis on End-Stage Renal Disease
PLOS ONE | DOI:10.1371/journal.pone.0169203 January 20, 2017 4 / 15
Table 1. Characters of included studies.
Author Year Country Location or
data source
Study design Sample
size
(NHD)
Disease
status
Mean
age
(year)
Male
(%)
Mean
duration
of
dialysis
(mo)*
Dialysis session Reported
outcomes
NHD CHD
Chan[
38] 2002 Canada Toronto
General
Hospital
Observation
cohort study
41(28) ESRD
(end-
stage
renal
disease)
47
(11)
N/A NHD:
3.4Y;
CHD:
2.8Y
8–
10hours,
every
night
4hours,
3 times/
week
LVHI, BP, Hb.
Friedman
[
39]
2002 Canada Humber River
Regional
Hospital
Cross-sectional
cohort study
54(23) ESRD 44
(20–
65)
63.0% NHD:
100(83)
M; CHD:
29(17)M
6–
7nights/
week
3 times/
week
Albumin
Heidenheim
[
31]
2003 Canada London
(Canada)
Health
Sciences
Centre
Prospective
nonrandomized
(controlled)
study
45(12) ESRD N/A N/A 18M 6 nights/
week
3 times/
week
QOL;
Nesrallah
[
32]
2003 Canada London
(Canada)
Health
Sciences
Centre
Prospective
nonrandomized
(controlled)
study
43(12) ESRD N/A N/A 18M 6 nights/
week
3 times/
week
BP; Drug usage
Pierratos
[
40]
2004 Canada Humber River
Regional
Hospital
Retrospective
study
88 ESRD 49
(11)
65.0% 30(27)M 3–
4nights/
week
- QOL; LVH;
Lindsay[
33] 2004 Canada London
(Canada)
Health
Sciences
Centre
Prospective
controlled study
45(12) ESRD 46.7
(10.5)
(28–
76)
67.0% 5–36M 5–6
nights/
week
3 times/
week
BP; Mortality;
Schwartz
[
41]
2005 Canada Humber River
Regional
Hospital
Retrospective
cohort study
95(63) ESRD 49.7
(5.7)
68.0% 12M 5–6
nights/
week
3 times/
week
Hb; Drug usage
Culleton
[
23]
2007 Canada University of
Calgary and
University of
Alberta
Randomized
Controlled study
52(26) ESRD 54.1
(12.8)
62.7% 6M 6 nights/
week
3 times/
week
LVH; QOL; BP;
Drug usage
Johansen
[
42]
2009 U.S United States
Renal Data
System
database
Observation
cohort study
1034
(94)
ESRD 46.7
(17.4)
65.9% 36M 5–6
nights/
week
3 times/
week
Mortality;
Hospitalization
Manns[
24] 2009 Canada University of
Calgary and
University of
Alberta
Randomized
Controlled study
51(26) ESRD 54.1
(12.8)
62.7% 6M 5–6
nights/
week
3 times/
week
QOL
Powell[
43] 2009 U.K Western
Infirmary
renal unit
Case-Controlled
study
106(53) ESRD 51.2
(15.5)
74.5% >12M 3 times/
week
3 times/
week
URR; HB; BP;
Drug usage
van Eps[
44] 2010 Australia Princess
Alexandra
Hospital
Observation
cohort study
235(63) ESRD 56.5
(15.1)
63.8% 12M 3.5–4
times/
week
3 times/
week
Side-effects;
Mortality
Lacson[
45] 2010 U.S Fresenius
Medical Care,
North
America
Case-Controlled
study
15989
(655)
ESRD 61.9
(15)
53.6% 12M 3 times/
week
3 times/
week
Mortality;
Hospitalization;
QOL; BP
(Continued)
Hemodialysis on End-Stage Renal Disease
PLOS ONE | DOI:10.1371/journal.pone.0169203 January 20, 2017 5 / 15