J Nephrol DOI 10.1007/s40620-014-0153-1
ORIGINAL ARTICLE
Hemodialysis in satellite dialysis units: incidence of patient fallback to the in-center dialysis unit Aurore Barthelemy • Thierry Lobbedez • Clemence Bechade • Patrick Henri Jean-Marie Batho • Eric Cardineau • Alain Jeanson • Isabelle Landru • Franc¸oise Lefort • Jacky Potier • Elie Zagdoun • Pascal Thibon
•
Received: 2 June 2014 / Accepted: 9 October 2014 Ó Italian Society of Nephrology 2014
Abstract Background Patient fallbacks occur when a patient treated in a satellite dialysis unit (SDU) must be transferred to an in-center dialysis unit. Fallbacks have an impact on the in-center dialysis unit organization. This study was carried out to estimate the incidence and risk factors of patient fallback. Methods Between 01/01/2006 and 31/12/2010 there were 193 patients starting dialysis in SDUs in one French region. The primary outcome was the incidence of temporary and permanent fallbacks with or without hospitalization. A survival analysis and binomial negative regression were used for the statistical analysis. Results Among the 193 patients, 117 (60.6 %; 95 % confidence interval [CI]: 53.3–67.6) had at least one fallback, which occurred within a median of 249 days (interquartile range [IQR]: 71–469) after the first session in the
SDU. The median number of fallbacks by subject was 1 (IQR: 0–4). The median duration of the fallback period was 4 days (IQR: 1–8) and median number of dialysis sessions during the fallback time was 1 (IQR: 1–3). Of the 494 temporary fallbacks, 210 were due to patient hospitalization, the main cause of which was cardiovascular disease. At univariate analysis, patients permanently transferred to in-center hemodialysis units were older, had more peripheral arteriopathy, cancer and arrhythmia. At multivariate analysis, peripheral arteriopathy (relative risk [RR] 2.06, 95 % CI 1.05–4.09) and the center (center 2: RR 0.42, 95 % CI 0.21–0.84; center 3: RR 2.88, 95 % CI 1.20–6.91) were significantly associated with the number of fallbacks. Conclusion Fallback is a common event in hemodialysis patients treated in SDUs. Yet, the SDU system operates well since a third of patients treated in these units are still in SDUs at 2 years of follow-up. Factors associated with patient fallback are the center and cardiovascular disease.
A. Barthelemy C. Bechade J.-M. Batho P. Thibon Renal Epidemiology and Information Network, CHU de Caen, 14000 Caen, France
I. Landru CH Robert Bisson, Nephrology, 14100 Lisieux, France
A. Barthelemy T. Lobbedez C. Bechade P. Henri CHU de Caen, Nephrology, 14000 Caen, France T. Lobbedez (&) Nephrology, CHU de Caen, 14033 Caen CEDEX 9, France e-mail: [email protected] J.-M. Batho CHP Saint-Martin, Nephrology, 14000 Caen, France
F. Lefort CHU de Caen Division of Medical Informatics, 14000 Caen, France J. Potier CH Pasteur, Nephrology, 50100 Cherbourg, France E. Zagdoun CH Me´morial France Etats-Unis, Nephrology, 50009 Saint-Loˆ, France
E. Cardineau CH Inter-communal Alenc¸on-Mamers, Nephrology, 61000 Alenc¸on, France A. Jeanson CH Jacques Monod, Nephrology, 61100 Flers, France
123
J Nephrol
Keywords Transfer
Hemodialysis Satellite hemodialysis
Introduction In France, hemodialysis patients (HD) are treated either in an in-center hemodialysis unit (ICHDU), medical dialysis unit (MDU) or satellite dialysis unit (SDU). ICHDUs must be localized in academic or community hospitals. With the exception of some regions, patients treated in an MDU or SDU are under the responsibility of the nephrologists of the ICHDU. MDUs are not necessarily located in hospital facilities. Patients treated in an MDU are seen by a nephrologist once a week whereas patients treated in an SDU are seen by a nephrologist once a month. However, patients in the MDU and SDU can be seen by the nephrologist on call of the ICHDU whenever necessary [1]. Patients treated in an SDU are younger and less comorbid than those treated in the ICHDU [2]. In France, of 36,109 patients treated by hemodialysis in 2012, 8,001 were treated in SDUs [2]. Of the prevalent patients treated in an SDU in 2012, 658 patients were transferred to ICHDUs. Patients treated in an SDU have a higher quality of life compared with those treated in ICHDUs [3–6]. In addition, the SDU is associated with a lower transport-related time and cost than in-center hemodialysis [3–9]. Fallbacks can be defined as any situation where a patient treated in an SDU must be transferred to the ICDHU. Patient fallback can occur for a limited period of time or permanently. As patient fallbacks are frequently unplanned, transfers to the ICHDU have a major impact on the incenter dialysis unit organization. In order to deal with patient fallbacks, it was decided arbitrarily by the French health care authorities that each ICHDU must have one hemodialysis machine available for every 40 SDU or MDU patients under their responsibility [1]. Although patient fallback is associated with in-center organization issues, to our knowledge there is only one study from Canada which has focused on this issue [10]. This study was conducted to estimate the fallback incidence in a cohort of patients starting renal replacement therapy in SDUs in one French region between 2006 and 2010. The study was also carried out to identify risk factors for patient fallback.
31/12/2010. Patients were followed for at least two years after initiation of dialysis in the SDU. Only patients older than 18 years were included in the study. These hemodialysis patients were treated in ten SDUs, each under the responsibility of one of the seven ICHDU facilities in our region, which is characterized by a greater number of patients living in a rural area compared to the rest of France. Definition of the events The event of interest was the fallback during the follow-up period. Fallback was defined by at least one dialysis session performed in the ICHDU whatever the cause. In addition, temporary fallbacks, defined by more than one dialysis session in the ICHDU within a period of less than 2 months (with or without hospitalization), were registered. Patients switched to the ICHDU for more than 2 months were considered permanently transferred to the in-center dialysis unit. Consequently there were three types of fallback: single fallback (one dialysis session), temporary fallback (more than one dialysis session within less than 2 months) and permanent fallback. The time between SDU initiation and the first fallback was calculated. The number of fallbacks and the number of dialysis sessions performed during both single fallback and temporary fallback were collected. After inclusion, patients were followed until one of the following events occurred: renal transplantation, death, or loss to follow-up. If none of these events happened, the end of the observation period was 31st December 2012, so ensuring that each patient could be followed for at least 2 years after inclusion. Patients were therefore followed for a maximum of 7 years. Data collection A regional subset of patients was extracted from the national registry of the Renal Epidemiology and Information Network (REIN) to create a cohort of incident patients who started treatment in SDUs during the study period. Information regarding patient fallback was extracted from the French Medical Information System Program (PMSI). The PMSI routinely collects information about patient hospitalization, with regular cross-checks for consistency and completeness. In each ICHDU, information about the main cause of hospitalization, the length of stay and the number of in-center dialysis sessions was collected.
Materials and methods Statistical analysis Study population This was a retrospective open cohort study of 193 patients who began treatment in SDUs between 01/01/2006 and
123
Based on the study of Lindsay et al. [10], in which 58 % of satellite dialysis patients had at least one fallback during 1 year in SDU, our power calculation, with an error of
J Nephrol Table 1 Baseline characteristics of patients SDU patients 193 N (%)
ICHDU patients 775 N (%)
Home patients 4 N (%)
pd
Gender: male Age at start of dialysisa
132 (68.4) 59.3 (47.7–68.3)
470 (61.2) 70.2 (57.5–78.8)
3 (75.0) 65.4 (53.6–73.0)
0.09 \0.0001
BMIa in kg/m2 (missing: 259)
25.1 (22.0–29.0)
25.1 (21.5–29.2)
27.4 (24.2–31.2)
0.45
Full-time
52 (31.9)
74 (11.2)
2 (50.0)
Retired
80 (49.1)
486 (73.7)
1 (25.0)
Unemployed or disabled
31 (19.0)
100 (15.1)
1 (25.0)
Active
24 (13.6)
87 (12.5)
0 (0)
Past smokers
37 (21.0)
153 (22.0)
1 (33.3)
Non-smokers
115 (65.4)
455 (65.5)
2 (66.7) 0
Number of patients
Covariate Socio-demographics
Employment (missing: 145)
Smoking status (missing: 98)
0.9
Transportation (missing: 424) Ambulance
80 (75.5)
383 (86.8)
Self-transportation
25 (23.6)
55 (12.5)
0
Other
1 (0.9)
3 (0.7)
1 (100.0)
26 (13.9)
186 (24.7)
0 (0.0)
Diabetic
21 (10.9)
123 (17.9)
0 (0)
Vascular
20 (10.4)
173 (25.1)
0 (0)
Glomerulonephritis
62 (32.1)
126 (18.3)
1 (50.0)
Tubulo-interstitial
17 (8.8)
70 (10.2)
1 (0)
Hereditary
31 (16.1)
40 (5.8)
1 (0)
Malformative
5 (2.6)
10 (1.4)
0 (0)
Other
1 (50.0)
Disability (missing: 30) Underlying nephropathy
37 (19.2)
146 (21.2)
Awaiting renal transplantation (missing: 53)
32 (16.6)
49 (6.3)
0 (0)
Reasons for non-registration
97 (50.2)
589 (76.0)
2 (50.0)
Medical contraindication
48 (24.9)
82 (10.6)
2 (50.0)
Pre-transplant check-up
4 (2.0)
6 (0.8)
0
Patient decline
2 (1.1)
6 (0.8)
0
0.006 0.0001
\0.0001
Other Duration of in-center HD (years)a (missing: 5)
0.7 (0.3–3.1)
–
2.9 (0.5–14.2)
5
–
0
Biological parameters Serum albumina (g/l) (missing: 283)
36 (32.0–38.9)
33.0 (29.0–33.3)
32.5 (29.5–36.2)
\0.0001
Hemoglobina (g/dl) (missing: 106)
11.2 (10.2–12.4)
10.9 (9.7–12.1)
12.7 (9.4–13.0)
0.01
0 Kt/v (missing:372)
1.4 (1.2–1.6)
1.2 (1.0–1.5)
0
–
First dialysis in emergency
22 (11.4)
146 (18.9)
1(25.0)
0.01
First dialysis with catheter
61 (31.6)
366 (47.2)
2(50.0)
\0.0001 \0.0001
Condition of HD initiation
Dialysis technique Hemodialysis
185 (95.9)
638 (83.2)
4 (100)
Hemodiafiltration
8 (4.2)
128 (16.7)
0 (0)
Biofiltration
0(0)
1 (0.1)
0 (0)
123
J Nephrol Table 1 continued Number of patients
SDU patients 193 N (%)
ICHDU patients 775 N (%)
Home patients 4 N (%)
pd
\0.0001
Type of vascular access (missing: 27) Arteriovenous fistula
184 (95.3)
374 (48.8)
4 (100)
Graft
6 (3.1)
13 (1.7)
0 (0)
Catheter
0 (0)
364 (47.4)
0 (0)
Comorbidities Hypertension
170 (88.1)
656 (84.6)
4( 100)
0.2
Diabetes
38 (19.7)
234 (30.2)
0 (0)
0.004
Type 1
8 (4.2)
40 (17.1)
0 (0)
\0.001
Type 2
30 (15.5)
194 (82.9)
0 (0)
0.001
27 (14.0) 23 (11.9)
217 (28.0) 193 (24.9)
0 (0) 1 (25.0)
0.01 \0.001
Chronic heart failureb Coronaropathyc Myocardial infarction
11 (5.7)
91 (11.7)
0 (0)
0.09
Peripheral arteriopathy
17 (8.8)
152 (19.6)
0 (0)
0.002
Cancer
16 (8.3)
98 (12.6)
0 (0)
\0.001
Chronic respiratory failure
4 (2.1)
65 (8.4)
0 (0)
0.17
Arrhythmia
16 (8.3)
159 (20.5)
0 (0)
0.59
Liver cirrhosis
1 (0.5)
15 (1.9)
0 (0)
0.77
Positive HCV serology
3 (1.6)
14 (1.9)
1 (33.3)
0.47
9
58
1
2(1.1)
5 (0.7)
0 (0)
9
58
1
0 (0)
2 (0.3)
0 (0)
9
58
1
Missing Positive HBV serology Missing Positive HIV serology Missing Biological values are at start of satellite dialysis
SDU satellite hemodialysis unit, ICHDU in-center hemodialysis unit, BMI body mass index, HD hemodialysis, HCV hepatitis C virus, HBV hepatitis B virus, HIV human immunodeficiency virus a
Median (interquartile range)
b
Symptoms of chronic heart failure according to New York Heart Association classification
c
Defined by history of percutaneous transluminal coronary angioplasty or coronary artery bypass surgery or coronaropathy documented by stress test or a myocardial scintigraphy d Comparison between ICHDU patients and SDU patient by Chi2 test. Home patients were not included for comparison because the number of effective cases was too little
10 % (two tailed), showed that 94 patients had to be included in the study to measure fallback frequency with a precision of 90 %. We looked for the outliers and checked the quality of the data. Continuous variables were expressed as median value, 1st and 3rd quartile, and mean and standard deviation (SD). Categorical variables were expressed as frequency and percentage. The Chi2 test, Fischer’s test, variance analysis or Kruskal–Wallis test were used for the group comparisons. The associations between each categorical covariate and the first fallback within 2 years were estimated with Kaplan–Meier survival curves and log-rank test whereas a Cox model was used for continuous covariates. The link between each covariate and the number of fallbacks within 2 years was evaluated by both univariate and multivariate analysis with a binomial
123
negative regression model. A p value of less than 0.05 was considered to be statistically significant. Statistical analyses were performed with the SAS software (SAS Institute Inc., Cary, NC, USA). This study was approved by the ethical committee: Comite´ de Protection des Personnes Nord-Ouest (September 19th 2013, reference: A13-D41-VOL.18).
Results Patient characteristics There were 193 incident patients who started dialysis in one of the ten satellite units of our region during the study
J Nephrol 45,0 40,0 39.4 35,0
Percent of subjects
30,0 25,0 20,0 15,0
12.9 9.4 9.4 9.9
10,0
4.1
5,0 0,0
0
1
2
3
4
5
2.6 6
4.1 1.6 1.6 7
8
9
0.5 0.5 1.0
0.5 0.5
1.0
0.5
10 11 12 13 14 15 16 17 18 19 20 21 22 23
0.5 24 25 26 27
28
frequency of fallbacks
Fig. 1 Frequency of fallbacks
period. The mean follow-up duration was 3.2 ± 2.2 years. No patient was lost to follow-up during the study. Mean age was 57.5 ± 14.8 years, the sex ratio (M/F) was 2.2, and the median dialysis vintage before SDU was 0.7 years (interquartile range [IQR]: 0.3–3.1). Patients’ sociodemographic characteristics, underlying nephropathy, comorbidities and biological parameters at satellite dialysis initiation are provided in Table 1. No patients treated in SDU were dialyzed through a hemodialysis catheter. Characteristics of patients treated in ICHDUs and by home-hemodialysis are also presented Table 1. Compared with patients treated in SDUs, incident in-center dialysis patients were older, less frequently registered on the waiting list for renal transplantation, had more disabilities, and more comorbidities. Incidence and characteristics of the fallback Of these 193 patients, 117 (60.6 %; 95 % confidence interval [CI] 53.3–67.6) experienced at least one permanent or temporary fallback. The median time between satellite dialysis start and the first fallback was 249 days (IQR: 71–469). There were 527 fallbacks (either permanent or for a limited period of time) which corresponded to a median of 1 (IQR: 0–4) fallback per patient. During the study period there were 2418 dialysis sessions linked to patient fallbacks, corresponding to four dialysis sessions per patient-year. The fallback incidence was 0.5 per patient-
year. Among the 117 patients, 25 (12.9 %) had only one fallback episode, and 92 patients (47.7 %) had multiple fallbacks (Fig. 1). Of the 527 fallbacks, 44.4 % and 85.5 % occurred, respectively, within 6 and 24 months after satellite dialysis initiation. Median duration of the fallback period was 4 days (IQR: 1–8) and the median number of dialysis sessions during the fallback period was 1 (IQR: 1–3). Of the 494 temporary fallbacks, 210 were due to patient hospitalization. Medical non-dialysis problems represented the main cause of hospitalization (cardiovascular disease: 18.1 %, gastrointestinal disease: 15.4 %). Hospitalizations were, otherwise, related to surgical procedures (15.9 %), vascular access issues (11.1 %) and infections (9.1 %). Patients who underwent permanent fallback were older (mean age 64.0 ± 15.0 years vs. 56.2 ± 14.5, and had more comorbidities than the other patients (peripheral arteriopathy 18.2 vs. 6.9 %, p = 0.04; cancer 21.2 vs. 5.6 %, p = 0.003; arrhythmia 18.2 vs. 6.2 %, p = 0.02). Detailed information about fallbacks is presented in Table 2. There were significant differences between centers regarding the patients and fallbacks characteristics (Table 3). Outcome of dialysis At the end of the study, 61 subjects (31.6 %) were still being treated in SDU, 33 (17.1 %) had been transferred
123
J Nephrol Table 2 Fallback characteristics Types of fallback
Single
Temporary without hospitalization
Temporary with hospitalization
Permanent
N (%)
185 (35.1)
99 (18.8)
210 (39.8)
33 (6.3)
Time before first fallback event (days) Mean (SD)
470.9 (558.9)
478.7 (480.8)
498.5 (451.8)
699.0 (535.3)
Median (IQR)
246.5 (85–729)
340.0 (64–780)
413 (127–735)
522 (363–1062)
630.7 (574.4)
819.2 (598.7)
699.0 (535.3)
466 (133–1027)
701.5 (353-1,210)
522 (363–1062)
Time before fallback event (days) Mean (SD)
631.8 (598.1)
Median (IQR) 474.0 (176–919) Dialysis sessions during fallback period Mean (SD)
–
5.7 (5.4)
2.8 (3.1)
31.9 (20.3)
Median (IQR)
–
3 (2–7)
2 (1–3)
28 (27–38)
Duration of the fallback period (days) Mean (SD)
–
13.6 (13.2)
7.8 (10)
70.7 (48.5)
Median (IQR)
–
8 (4–19)
5 (3–8)
62 (50–76)
SD standard deviation, IQR interquartile range
Table 3 Particularities of patients and fallbacks by center Satellite units
1
2
3
4
5
6
7
p
Number of patients
15
30
9
23
57
8
51
– \0.001
Age (years) Median
62
63
51
66
61
51
49
IQR
40–67
57–70
46–56
60–73
55–69
36–65
38–61
First dialysis in emergency N (%)
4 (26.7)
2 (6.7)
0 (0.0)
0 (0.0)
2 (3.5)
1 (12.5)
13 (25.5)
0.01
Awaiting renal transplantation N (%)
1 (6.7)
3 (10.0)
3 (33.3)
0 (0.0)
8 (14.0)
2 (25.0)
15 (29.4)
0.02 \0.001
Duration of ICHD before SDU Median
0.8
0.3
1.7
1.0
0.6
0.9
0.7
IQR
0.6–1.6
0.2–0.9
1.1–15.6
0.6–2.9
0.2–1.3
0.3–1.1
0.3–5.8
14 (93.3)
16 (53.3)
8 (88.9)
20 (87.0)
26 (45.6)
6 (75.0)
27 (52.9)
\0.001 \0.001
Fallbacks N (%) Number of fallbacks per patient Median IQR
3.0
1.0
7.0
3.0
0.0
1.5
1.0
2.0–4.0
0.0–2.0
3.0–12.0
1.0–5.0
0.0–4.0
0.5–3.0
0.0–3.0
Number of dialysis during fallbacks Median
1
2
1
2
2
1.5
1
IQR
1–3
1–4
1–1
1–3
1–4
1–2
1–4
\0.001
N number, ICHD in-center hemodialysis, SDU satellite hemodialysis unit; IQR interquartile range
to an ICHDU, and among these 8 had returned to the SDU. Of the 193 patients included, there were 70 (36.3 %) transplantations, 20 (10.4 %) deaths, 8 (4.1 %) transfers outside the region and 1 (0.5 %) renal function recovery. Factors associated with patient fallbacks The global Kaplan–Meier curve of 2-year survival free of fallback is represented in Fig. 2. The cumulative
123
probabilities of fallbacks at 6, 12, 18 and 24 months were respectively: 27.8, 40.9, 53.7 and 62.2 %. In the survival analysis, there was no association between patient characteristics and the occurrence of fallbacks. Binomial negative regression analysis showed that underlying nephropathy (hereditary nephropathy vs. other types: relative risk [RR] 0.43, 95 % CI 0.23–0.83), peripheral arteriopathy (RR 2.08, 95 % CI 1.06–4.06), and the center (center 2: RR 0.44, 95 % CI 0.21–0.97; center 3: RR 3.30, 95 % CI 1.22–8.80) were associated with the number of
J Nephrol
Number of subjects at risk during follow-up 193
6 months
12 months
18 months
24 months
141
120
102
93
Fig. 2 Fallback-free survival curve in the two first years of follow-up
patient fallbacks (Table 4). In the multivariate analysis, peripheral arteriopathy (RR 2.06, 95 % CI 1.05–4.09) and center (center 2: RR 0.42, 95 % CI 0.21–0.84; center 3: RR 2.88, 95 % CI 1.20–6.91) were independently associated with the number of fallbacks.
Discussion To our knowledge, this is the first report on the outcome of SDU patient fallbacks in France. Our findings are in line with the results of a study from Canada [10]. However, in the study of Lindsay et al., the main cause of fallback was access-related, whereas medical problems were the principal causes of fallback in our study. This could be explained by the fact that hemodialysis catheters are not authorized in SDUs in France. Data from the Dialysis Outcomes and Practices Patterns Study (DOPPS) showed that, for French hemodialysis patients, the mean number of annual hospitalizations was 1.43 per patient/year with a mean length of stay of 9.2 days [11]. The principal causes of hospitalization were cardiovascular and gastrointestinal diseases, vascular access issues, or infections. Similar results were observed in the United States [12–18]. Even though the causes of hospitalization in our study were similar to those reported in the literature, the number of infection-related hospitalizations was relatively lower. Again, this could be explained by the type of vascular access used in the SDU in France. In our region, 21 % of patients on hemodialysis were treated in SDUs, which is close to the rate of 20.4 % observed in France as a whole. Patients treated in the SDU
had similar features to those treated in SDUs in France [2]. However, there were more SDU patients who had at least one comorbidity (99 %) compared to patients treated in SDUs (58.4 %). In 2012 there were important discrepancies in France regarding the rate of patients treated in satellite dialysis units. Since there were less MDU facilities in our region compared to the rest of France, this could have influenced the type of patients treated in SDU. Indeed in France, there are no formal criteria for selecting patients eligible for SDU. In some regions, only patients without comorbidity are treated in SDUs while comorbid patients are treated in an MDU or ICHDU. In our region, there are only a few MDUs. Thus patients were more likely to have dialysis sessions in an SDU. Furthermore, in our region, many patients live in rural territories; having the dialysis session in the SDU can reduce transportation time for these patients. These elements could explain the proportion of comorbid patients treated in SDU in our region. Therefore, comorbidities could, in part, have affected fallback incidence. Differences between centers could reflect differences in medical practice between centers, as there was no significant difference regarding patient comorbidities between the centers. At the same time, one may argue that characteristics of the centers, which were not analyzed in our study, could have influenced patient fallback. Our study has some limitations. First, only the main diagnosis was collected for each fallback. In addition, for temporary fallbacks without hospitalization, causes of fallback were not identified. Patients could have been hospitalized outside the region, leading to missing fallbacks in our data collection. These could have led to a lack of power in the statistical analysis. There were some missing data in our study. In particular, the Kt/V value was missing for 53.4 % of the included patients, which is similar to the missing rate observed in the 2012 annual report of the REIN registry. A better characterization of the patient fallback, especially when it becomes permanent, could help nephrologists deal with allocation of the population between incenter units and satellite units. In our study, among comorbidities, only arteriopathy was associated with the number of fallbacks in the multivariate analysis. The center effect observed in our study suggests that differences in medical practice and factors related to dialysis centers, such as SDU distance from the ICHDU, numbers of nephrologists and nurses per patient treated in SDU, number of dialysis generators, and transportation time between the patient’s home and the SDU could influence fallback occurrence. A complementary study from our team including center characteristics will be conducted to study the center effect observed in our first results. Our study provides information regarding the number of
123
J Nephrol Table 4 Univariate analysis: association between covariate and occurrence of fallbacks and number of fallbacks in the first 2 years of follow-up Factor
Fallbacks (n, %)
Gender
p*
Number mean ± SD
RR (95 % CI)
p**
0.24
Women
28 (45.9)
1.5 ± 2.2
Ref
–
Men
72 (54.5)
1.6 ± 2.7
1.09 (0.68–1.73)
0.72
Age at dialysis initiation
0.58
18–35 years
10 (52.6)
1.5 ± 2.0
Ref
–
36–60 years
37 (44.6)
1.5 ± 2.8
1.01 (0.46–2.22)
0.97
More than 60 years
53 (58.24)
1.7 ± 2.2
1.19 (0.55–2.59)
0.65
0.4 ± 0.8
0.35 (0.08–1.60)
0.18
BMI (kg/m2) (missing: 38) \18
0.96 2 (28.6)
18–25
25 (52.1)
1.2 ± 1.7
Ref
–
25–30 C30 kg/m2
37 (56.7) 17 (56.7)
1.8 ± 2.8 2.0 ± 2.7
1.48 (0.85–2.6) 1.68 (0.86–3.29)
0.16 0.13
1.3 ± 2.0
Ref
–
Employment (missing: 30) Full-time
0.48 23 (44.2)
Unemployed or disabled
17 (54.8)
1.9 ± 3.5
1.47 (0.76–2.85)
0.25
Retired
47 (58.7)
1.8 ± 2.2
1.41 (0.83–2.38)
0.20
Smoking status (missing: 17)
0.52
Active smokers
15 (62.5)
1.9 ± 2.7
1.33 (0.70–2.50)
0.38
Past smokers
18 (48.6)
2.0 ± 3.4
1.42 (0.84–2.41)
0.19
Non smokers
60 (52.2)
1.4 ± 2.0
Ref
–
Transportation (missing: 87)
0.68
Ambulance
43 (53.7)
1.8 ± 2.3
0.94 (0.47–1.87)
0.85
Self-transportation
14 (56.0)
1.9 ± 2.3
ref
–
Disability (missing: 6)
16 (61.5)
1.8 ± 2.4
1.11 (0.59–2.07)
0.74
Underlying nephropathy
0.76 0.09
Other than hereditary
90 (55.5)
1.8 ± 2.6
Ref
–
Hereditary Awaiting transplantation
10 (32.3) 14 (43.7)
0.95
0.8 ± 1.5 1.4 ± 2.2
0.43 (0.23–0.83) 0.82 (0.46–1.48)
0.01 0.52
First dialysis in emergency
12 (54.5)
0.50
1.4 ± 1.9
0.89 (0.45–1.75)
0.73
First dialysis with catheter
35 (57.4)
0.57
1.7 ± 2.0
1.05 (0.66–1.67)
0.83
Hemoglobin (g/dl) (missing:15) \11 11–13 [13 Albumin (g/l) median (IQR) (missing: 47)
0.30 43 (54.4)
1.7 ± 2.7
1.71 (0.85–3.46)
0.13
9 (34.6)
1.0 ± 1.8
Ref
–
1.7 ± 2.3
1.71 (0.84–3.48)
0.14
0.97 (0.93–1.02)
0.29 0.99
48 (54.5) 36.0 (32–39)
Dialysis technique
0.82 0.81
HD
95 (51.3)
1.6 ± 2.4
Ref
HDF
5 (62.5)
1.6 ± 2.7
(0.34–2.94)
Duration of in-center HD (months) (missing: 5)
0.52
\6
33 (46.5)
1.3 ± 2.5
Ref
–
6–12
26 (60.5)
2.1 ± 2.7
1.23 (0.88-2.82)
0.12
12–24 [24
9 (45.0) 32 (59.3)
1.6 ± 2.6 1.7 ± 2.1
1.23 (0.57–2.66) 1.29 (0.74–2.23)
0.59 0.37
Comorbidities Hypertension Stroke Transient ischemic attack Coronaropathy
123
88 (51.8)
0.95
1.6 ± 2.5
1.27 (0.64–2.51)
0.49
6 (54.5)
0.18
2.8 ± 5.1
1.82 (0.79–4.20)
0.16
2 (50.0)
0.45
0.8 ± 1.0
0.46 (0.08–2.50)
0.37
13 (56.5)
0.90
1.7 ± 2.5
1.09 (0.56–2.10)
0.80
J Nephrol Table 4 continued Factor Diabetes
Fallbacks (n, %)
p*
Number mean ± SD
RR (95 % CI)
p**
21 (55.3)
0.68
1.9 ± 2.7
1.24 (0.73–2.12)
0.41
1 (25.0)
0.77
4.2 ± 8.5
2.72 (0.75–9.9)
0.13
Arrhythmia Arteriopathy
10 (62.5) 11 (64.7)
0.79 0.59
2.4 ± 2.9 3.1 ± 4.5
1.53 (0.72–3.2) 2.07 (1.06–4.06)
0.26 0.03
Cancer
Chronic respiratory failure
12 (75.0)
0.13
3.0 ± 4.4
2.01 (0.99–4.06)
0.05
Myocardial infarction
7 (63.6)
0.90
1.3 ± 1.3
0.78 (0.30–2.00)
0.60
Chronic heart failure
14 (51.9)
0.90
1.6 ± 2.7
0.98 (0.53–1.82)
0.96
HD hemodialysis, HDF hemodiafiltration, BMI body mass index, n number, SD standard deviation, RR relative risk, CI confidence interval * p value of the log rank test ** p value of the negative binomial regression
dialysis machines required for the patient fallback. Indeed, we found that fallbacks corresponded to four dialysis sessions per patient-year.
Conclusion In satellite dialysis units, patient fallbacks are a frequent event within the first 2 years of SDU and are mainly related to patients’ hospitalization. Nevertheless, the system of SDU operates well, since most of the fallbacks end with the patient’s return to the SDU. Only arteriopathy and the center were associated with the number of fallbacks. Complementary studies are needed to study the center effects. Conflict of interest The authors have do not have any conflict of interest.
References 1. Ministe`re de la sante´, de la famille et des personnes handicape´es. De´cret n° 2002-1198 du 23 septembre 2002 relatif aux conditions techniques de fonctionnement des e´tablissements de sante´ qui exercent l’activite´ de traitement de l’insuffisance re´nale chronique par la pratique de l’e´puration extrare´nale et modifiant le code de la sante´ publique :troisie`me partie : De´crets. JO n° 224 du 25 septembre 2002, p. 15813 2. REIN- annual report, 2012 3. Diamant MJ, Young A, Gallo K et al (2011) Hemodialysis in a satellite unit: clinical performance target attainment and healthrelated quality of life. Clin J Am Soc Nephrol 6:1692–1699 4. Malmstro¨m RK, Roine RP, Heikkila¨ A et al (2008) Cost analysis and health-related quality of life of home and self-care satellite haemodialysis. Nephrol Dial Transplant 23:1990–1996 5. Mazairac AHA, Grooteman MPC, Blankestijn PJ et al (2012) Differences in quality of life of hemodialysis patients between dialysis centers. Qual Life Res 21:299–307
6. Diamant MJ, Harwood L, Movva S et al (2010) A comparison of quality of life and travel-related factors between in-center and satellite-based hemodialysis patients. Clin J Am Soc Nephrol 5:268–274 7. Prakash S, Austin PC, Oliver MJ, Garg AX, Blake PG, Hux JE (2007) Regional effects of satellite haemodialysis units on renal replacement therapy in non-urban Ontario Canada. Nephrol Dial Transplant 22:2297–2303 8. Soroka SD, Kiberd BA, Jacobs P (2005) The marginal cost of satellite versus in-center hemodialysis. Hemodial Int 9:196–201 9. Roderick P, Armitage A, Nicholson T et al (2004) A clinical and cost evaluation of hemodialysis in renal satellite units in England and Wales. Am J Kidney Dis 44:121–131 10. Lindsay RM, Hux J, Holland D et al (2009) An investigation of satellite hemodialysis fallbacks in the province of Ontario. Clin J Am Soc Nephrol 4:603–608 11. Rayner HC, Pisoni RL, Bommer J et al (2004) Mortality and hospitalization in haemodialysis patients in five European countries: results from the Dialysis Outcomes and Practice Patterns Study (DOPPS). Nephrol Dial Transplant 19:108–208 12. Arneson TJ, Liu J, Qiu Y, Gilbertson DT, Foley RN, Collins AJ (2010) Hospital treatment for fluid overload in the Medicare hemodialysis population. Clin J Am Soc Nephrol 5:1054–1063 13. Schoonover KL, Hickson LJ, Norby SM et al (2013) Risk factors for hospitalization among older, incident hemodialysis patients. Nephrology (Carlton) 18:712–717 14. Ng LJ, Chen F, Pisoni RL et al (2011) Hospitalization risks related to vascular access type among incident US hemodialysis patients. Nephrol Dial Transplant 26:3659–3666 15. Lacson E Jr, Wang W, Lazarus JM, Hakim RM (2010) Change in vascular access and hospitalization risk in long term hemodialysis patients. Clin J Am Soc Nephrol 5:1996–2003 16. Dalrymple LS, Johansen KL, Chertow GM et al (2010) Infectionrelated hospitalizations in older patients with ESRD. Am J Kidney Dis 56:522–530 17. US Renal Data System, USRDS (2013) Annual data report: atlas of chronic kidney disease and end-stage renal disease in the united states, National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda 18. Usvyat LA, Kooman JP, van der Sande FM et al (2014) Dynamics of hospitalizations in hemodialysis patients: results from a large US provider. Nephrol Dial Transplant 29:442–448
123
ORIGINAL ARTICLE
Hemodialysis in satellite dialysis units: incidence of patient fallback to the in-center dialysis unit Aurore Barthelemy • Thierry Lobbedez • Clemence Bechade • Patrick Henri Jean-Marie Batho • Eric Cardineau • Alain Jeanson • Isabelle Landru • Franc¸oise Lefort • Jacky Potier • Elie Zagdoun • Pascal Thibon
•
Received: 2 June 2014 / Accepted: 9 October 2014 Ó Italian Society of Nephrology 2014
Abstract Background Patient fallbacks occur when a patient treated in a satellite dialysis unit (SDU) must be transferred to an in-center dialysis unit. Fallbacks have an impact on the in-center dialysis unit organization. This study was carried out to estimate the incidence and risk factors of patient fallback. Methods Between 01/01/2006 and 31/12/2010 there were 193 patients starting dialysis in SDUs in one French region. The primary outcome was the incidence of temporary and permanent fallbacks with or without hospitalization. A survival analysis and binomial negative regression were used for the statistical analysis. Results Among the 193 patients, 117 (60.6 %; 95 % confidence interval [CI]: 53.3–67.6) had at least one fallback, which occurred within a median of 249 days (interquartile range [IQR]: 71–469) after the first session in the
SDU. The median number of fallbacks by subject was 1 (IQR: 0–4). The median duration of the fallback period was 4 days (IQR: 1–8) and median number of dialysis sessions during the fallback time was 1 (IQR: 1–3). Of the 494 temporary fallbacks, 210 were due to patient hospitalization, the main cause of which was cardiovascular disease. At univariate analysis, patients permanently transferred to in-center hemodialysis units were older, had more peripheral arteriopathy, cancer and arrhythmia. At multivariate analysis, peripheral arteriopathy (relative risk [RR] 2.06, 95 % CI 1.05–4.09) and the center (center 2: RR 0.42, 95 % CI 0.21–0.84; center 3: RR 2.88, 95 % CI 1.20–6.91) were significantly associated with the number of fallbacks. Conclusion Fallback is a common event in hemodialysis patients treated in SDUs. Yet, the SDU system operates well since a third of patients treated in these units are still in SDUs at 2 years of follow-up. Factors associated with patient fallback are the center and cardiovascular disease.
A. Barthelemy C. Bechade J.-M. Batho P. Thibon Renal Epidemiology and Information Network, CHU de Caen, 14000 Caen, France
I. Landru CH Robert Bisson, Nephrology, 14100 Lisieux, France
A. Barthelemy T. Lobbedez C. Bechade P. Henri CHU de Caen, Nephrology, 14000 Caen, France T. Lobbedez (&) Nephrology, CHU de Caen, 14033 Caen CEDEX 9, France e-mail: [email protected] J.-M. Batho CHP Saint-Martin, Nephrology, 14000 Caen, France
F. Lefort CHU de Caen Division of Medical Informatics, 14000 Caen, France J. Potier CH Pasteur, Nephrology, 50100 Cherbourg, France E. Zagdoun CH Me´morial France Etats-Unis, Nephrology, 50009 Saint-Loˆ, France
E. Cardineau CH Inter-communal Alenc¸on-Mamers, Nephrology, 61000 Alenc¸on, France A. Jeanson CH Jacques Monod, Nephrology, 61100 Flers, France
123
J Nephrol
Keywords Transfer
Hemodialysis Satellite hemodialysis
Introduction In France, hemodialysis patients (HD) are treated either in an in-center hemodialysis unit (ICHDU), medical dialysis unit (MDU) or satellite dialysis unit (SDU). ICHDUs must be localized in academic or community hospitals. With the exception of some regions, patients treated in an MDU or SDU are under the responsibility of the nephrologists of the ICHDU. MDUs are not necessarily located in hospital facilities. Patients treated in an MDU are seen by a nephrologist once a week whereas patients treated in an SDU are seen by a nephrologist once a month. However, patients in the MDU and SDU can be seen by the nephrologist on call of the ICHDU whenever necessary [1]. Patients treated in an SDU are younger and less comorbid than those treated in the ICHDU [2]. In France, of 36,109 patients treated by hemodialysis in 2012, 8,001 were treated in SDUs [2]. Of the prevalent patients treated in an SDU in 2012, 658 patients were transferred to ICHDUs. Patients treated in an SDU have a higher quality of life compared with those treated in ICHDUs [3–6]. In addition, the SDU is associated with a lower transport-related time and cost than in-center hemodialysis [3–9]. Fallbacks can be defined as any situation where a patient treated in an SDU must be transferred to the ICDHU. Patient fallback can occur for a limited period of time or permanently. As patient fallbacks are frequently unplanned, transfers to the ICHDU have a major impact on the incenter dialysis unit organization. In order to deal with patient fallbacks, it was decided arbitrarily by the French health care authorities that each ICHDU must have one hemodialysis machine available for every 40 SDU or MDU patients under their responsibility [1]. Although patient fallback is associated with in-center organization issues, to our knowledge there is only one study from Canada which has focused on this issue [10]. This study was conducted to estimate the fallback incidence in a cohort of patients starting renal replacement therapy in SDUs in one French region between 2006 and 2010. The study was also carried out to identify risk factors for patient fallback.
31/12/2010. Patients were followed for at least two years after initiation of dialysis in the SDU. Only patients older than 18 years were included in the study. These hemodialysis patients were treated in ten SDUs, each under the responsibility of one of the seven ICHDU facilities in our region, which is characterized by a greater number of patients living in a rural area compared to the rest of France. Definition of the events The event of interest was the fallback during the follow-up period. Fallback was defined by at least one dialysis session performed in the ICHDU whatever the cause. In addition, temporary fallbacks, defined by more than one dialysis session in the ICHDU within a period of less than 2 months (with or without hospitalization), were registered. Patients switched to the ICHDU for more than 2 months were considered permanently transferred to the in-center dialysis unit. Consequently there were three types of fallback: single fallback (one dialysis session), temporary fallback (more than one dialysis session within less than 2 months) and permanent fallback. The time between SDU initiation and the first fallback was calculated. The number of fallbacks and the number of dialysis sessions performed during both single fallback and temporary fallback were collected. After inclusion, patients were followed until one of the following events occurred: renal transplantation, death, or loss to follow-up. If none of these events happened, the end of the observation period was 31st December 2012, so ensuring that each patient could be followed for at least 2 years after inclusion. Patients were therefore followed for a maximum of 7 years. Data collection A regional subset of patients was extracted from the national registry of the Renal Epidemiology and Information Network (REIN) to create a cohort of incident patients who started treatment in SDUs during the study period. Information regarding patient fallback was extracted from the French Medical Information System Program (PMSI). The PMSI routinely collects information about patient hospitalization, with regular cross-checks for consistency and completeness. In each ICHDU, information about the main cause of hospitalization, the length of stay and the number of in-center dialysis sessions was collected.
Materials and methods Statistical analysis Study population This was a retrospective open cohort study of 193 patients who began treatment in SDUs between 01/01/2006 and
123
Based on the study of Lindsay et al. [10], in which 58 % of satellite dialysis patients had at least one fallback during 1 year in SDU, our power calculation, with an error of
J Nephrol Table 1 Baseline characteristics of patients SDU patients 193 N (%)
ICHDU patients 775 N (%)
Home patients 4 N (%)
pd
Gender: male Age at start of dialysisa
132 (68.4) 59.3 (47.7–68.3)
470 (61.2) 70.2 (57.5–78.8)
3 (75.0) 65.4 (53.6–73.0)
0.09 \0.0001
BMIa in kg/m2 (missing: 259)
25.1 (22.0–29.0)
25.1 (21.5–29.2)
27.4 (24.2–31.2)
0.45
Full-time
52 (31.9)
74 (11.2)
2 (50.0)
Retired
80 (49.1)
486 (73.7)
1 (25.0)
Unemployed or disabled
31 (19.0)
100 (15.1)
1 (25.0)
Active
24 (13.6)
87 (12.5)
0 (0)
Past smokers
37 (21.0)
153 (22.0)
1 (33.3)
Non-smokers
115 (65.4)
455 (65.5)
2 (66.7) 0
Number of patients
Covariate Socio-demographics
Employment (missing: 145)
Smoking status (missing: 98)
0.9
Transportation (missing: 424) Ambulance
80 (75.5)
383 (86.8)
Self-transportation
25 (23.6)
55 (12.5)
0
Other
1 (0.9)
3 (0.7)
1 (100.0)
26 (13.9)
186 (24.7)
0 (0.0)
Diabetic
21 (10.9)
123 (17.9)
0 (0)
Vascular
20 (10.4)
173 (25.1)
0 (0)
Glomerulonephritis
62 (32.1)
126 (18.3)
1 (50.0)
Tubulo-interstitial
17 (8.8)
70 (10.2)
1 (0)
Hereditary
31 (16.1)
40 (5.8)
1 (0)
Malformative
5 (2.6)
10 (1.4)
0 (0)
Other
1 (50.0)
Disability (missing: 30) Underlying nephropathy
37 (19.2)
146 (21.2)
Awaiting renal transplantation (missing: 53)
32 (16.6)
49 (6.3)
0 (0)
Reasons for non-registration
97 (50.2)
589 (76.0)
2 (50.0)
Medical contraindication
48 (24.9)
82 (10.6)
2 (50.0)
Pre-transplant check-up
4 (2.0)
6 (0.8)
0
Patient decline
2 (1.1)
6 (0.8)
0
0.006 0.0001
\0.0001
Other Duration of in-center HD (years)a (missing: 5)
0.7 (0.3–3.1)
–
2.9 (0.5–14.2)
5
–
0
Biological parameters Serum albumina (g/l) (missing: 283)
36 (32.0–38.9)
33.0 (29.0–33.3)
32.5 (29.5–36.2)
\0.0001
Hemoglobina (g/dl) (missing: 106)
11.2 (10.2–12.4)
10.9 (9.7–12.1)
12.7 (9.4–13.0)
0.01
0 Kt/v (missing:372)
1.4 (1.2–1.6)
1.2 (1.0–1.5)
0
–
First dialysis in emergency
22 (11.4)
146 (18.9)
1(25.0)
0.01
First dialysis with catheter
61 (31.6)
366 (47.2)
2(50.0)
\0.0001 \0.0001
Condition of HD initiation
Dialysis technique Hemodialysis
185 (95.9)
638 (83.2)
4 (100)
Hemodiafiltration
8 (4.2)
128 (16.7)
0 (0)
Biofiltration
0(0)
1 (0.1)
0 (0)
123
J Nephrol Table 1 continued Number of patients
SDU patients 193 N (%)
ICHDU patients 775 N (%)
Home patients 4 N (%)
pd
\0.0001
Type of vascular access (missing: 27) Arteriovenous fistula
184 (95.3)
374 (48.8)
4 (100)
Graft
6 (3.1)
13 (1.7)
0 (0)
Catheter
0 (0)
364 (47.4)
0 (0)
Comorbidities Hypertension
170 (88.1)
656 (84.6)
4( 100)
0.2
Diabetes
38 (19.7)
234 (30.2)
0 (0)
0.004
Type 1
8 (4.2)
40 (17.1)
0 (0)
\0.001
Type 2
30 (15.5)
194 (82.9)
0 (0)
0.001
27 (14.0) 23 (11.9)
217 (28.0) 193 (24.9)
0 (0) 1 (25.0)
0.01 \0.001
Chronic heart failureb Coronaropathyc Myocardial infarction
11 (5.7)
91 (11.7)
0 (0)
0.09
Peripheral arteriopathy
17 (8.8)
152 (19.6)
0 (0)
0.002
Cancer
16 (8.3)
98 (12.6)
0 (0)
\0.001
Chronic respiratory failure
4 (2.1)
65 (8.4)
0 (0)
0.17
Arrhythmia
16 (8.3)
159 (20.5)
0 (0)
0.59
Liver cirrhosis
1 (0.5)
15 (1.9)
0 (0)
0.77
Positive HCV serology
3 (1.6)
14 (1.9)
1 (33.3)
0.47
9
58
1
2(1.1)
5 (0.7)
0 (0)
9
58
1
0 (0)
2 (0.3)
0 (0)
9
58
1
Missing Positive HBV serology Missing Positive HIV serology Missing Biological values are at start of satellite dialysis
SDU satellite hemodialysis unit, ICHDU in-center hemodialysis unit, BMI body mass index, HD hemodialysis, HCV hepatitis C virus, HBV hepatitis B virus, HIV human immunodeficiency virus a
Median (interquartile range)
b
Symptoms of chronic heart failure according to New York Heart Association classification
c
Defined by history of percutaneous transluminal coronary angioplasty or coronary artery bypass surgery or coronaropathy documented by stress test or a myocardial scintigraphy d Comparison between ICHDU patients and SDU patient by Chi2 test. Home patients were not included for comparison because the number of effective cases was too little
10 % (two tailed), showed that 94 patients had to be included in the study to measure fallback frequency with a precision of 90 %. We looked for the outliers and checked the quality of the data. Continuous variables were expressed as median value, 1st and 3rd quartile, and mean and standard deviation (SD). Categorical variables were expressed as frequency and percentage. The Chi2 test, Fischer’s test, variance analysis or Kruskal–Wallis test were used for the group comparisons. The associations between each categorical covariate and the first fallback within 2 years were estimated with Kaplan–Meier survival curves and log-rank test whereas a Cox model was used for continuous covariates. The link between each covariate and the number of fallbacks within 2 years was evaluated by both univariate and multivariate analysis with a binomial
123
negative regression model. A p value of less than 0.05 was considered to be statistically significant. Statistical analyses were performed with the SAS software (SAS Institute Inc., Cary, NC, USA). This study was approved by the ethical committee: Comite´ de Protection des Personnes Nord-Ouest (September 19th 2013, reference: A13-D41-VOL.18).
Results Patient characteristics There were 193 incident patients who started dialysis in one of the ten satellite units of our region during the study
J Nephrol 45,0 40,0 39.4 35,0
Percent of subjects
30,0 25,0 20,0 15,0
12.9 9.4 9.4 9.9
10,0
4.1
5,0 0,0
0
1
2
3
4
5
2.6 6
4.1 1.6 1.6 7
8
9
0.5 0.5 1.0
0.5 0.5
1.0
0.5
10 11 12 13 14 15 16 17 18 19 20 21 22 23
0.5 24 25 26 27
28
frequency of fallbacks
Fig. 1 Frequency of fallbacks
period. The mean follow-up duration was 3.2 ± 2.2 years. No patient was lost to follow-up during the study. Mean age was 57.5 ± 14.8 years, the sex ratio (M/F) was 2.2, and the median dialysis vintage before SDU was 0.7 years (interquartile range [IQR]: 0.3–3.1). Patients’ sociodemographic characteristics, underlying nephropathy, comorbidities and biological parameters at satellite dialysis initiation are provided in Table 1. No patients treated in SDU were dialyzed through a hemodialysis catheter. Characteristics of patients treated in ICHDUs and by home-hemodialysis are also presented Table 1. Compared with patients treated in SDUs, incident in-center dialysis patients were older, less frequently registered on the waiting list for renal transplantation, had more disabilities, and more comorbidities. Incidence and characteristics of the fallback Of these 193 patients, 117 (60.6 %; 95 % confidence interval [CI] 53.3–67.6) experienced at least one permanent or temporary fallback. The median time between satellite dialysis start and the first fallback was 249 days (IQR: 71–469). There were 527 fallbacks (either permanent or for a limited period of time) which corresponded to a median of 1 (IQR: 0–4) fallback per patient. During the study period there were 2418 dialysis sessions linked to patient fallbacks, corresponding to four dialysis sessions per patient-year. The fallback incidence was 0.5 per patient-
year. Among the 117 patients, 25 (12.9 %) had only one fallback episode, and 92 patients (47.7 %) had multiple fallbacks (Fig. 1). Of the 527 fallbacks, 44.4 % and 85.5 % occurred, respectively, within 6 and 24 months after satellite dialysis initiation. Median duration of the fallback period was 4 days (IQR: 1–8) and the median number of dialysis sessions during the fallback period was 1 (IQR: 1–3). Of the 494 temporary fallbacks, 210 were due to patient hospitalization. Medical non-dialysis problems represented the main cause of hospitalization (cardiovascular disease: 18.1 %, gastrointestinal disease: 15.4 %). Hospitalizations were, otherwise, related to surgical procedures (15.9 %), vascular access issues (11.1 %) and infections (9.1 %). Patients who underwent permanent fallback were older (mean age 64.0 ± 15.0 years vs. 56.2 ± 14.5, and had more comorbidities than the other patients (peripheral arteriopathy 18.2 vs. 6.9 %, p = 0.04; cancer 21.2 vs. 5.6 %, p = 0.003; arrhythmia 18.2 vs. 6.2 %, p = 0.02). Detailed information about fallbacks is presented in Table 2. There were significant differences between centers regarding the patients and fallbacks characteristics (Table 3). Outcome of dialysis At the end of the study, 61 subjects (31.6 %) were still being treated in SDU, 33 (17.1 %) had been transferred
123
J Nephrol Table 2 Fallback characteristics Types of fallback
Single
Temporary without hospitalization
Temporary with hospitalization
Permanent
N (%)
185 (35.1)
99 (18.8)
210 (39.8)
33 (6.3)
Time before first fallback event (days) Mean (SD)
470.9 (558.9)
478.7 (480.8)
498.5 (451.8)
699.0 (535.3)
Median (IQR)
246.5 (85–729)
340.0 (64–780)
413 (127–735)
522 (363–1062)
630.7 (574.4)
819.2 (598.7)
699.0 (535.3)
466 (133–1027)
701.5 (353-1,210)
522 (363–1062)
Time before fallback event (days) Mean (SD)
631.8 (598.1)
Median (IQR) 474.0 (176–919) Dialysis sessions during fallback period Mean (SD)
–
5.7 (5.4)
2.8 (3.1)
31.9 (20.3)
Median (IQR)
–
3 (2–7)
2 (1–3)
28 (27–38)
Duration of the fallback period (days) Mean (SD)
–
13.6 (13.2)
7.8 (10)
70.7 (48.5)
Median (IQR)
–
8 (4–19)
5 (3–8)
62 (50–76)
SD standard deviation, IQR interquartile range
Table 3 Particularities of patients and fallbacks by center Satellite units
1
2
3
4
5
6
7
p
Number of patients
15
30
9
23
57
8
51
– \0.001
Age (years) Median
62
63
51
66
61
51
49
IQR
40–67
57–70
46–56
60–73
55–69
36–65
38–61
First dialysis in emergency N (%)
4 (26.7)
2 (6.7)
0 (0.0)
0 (0.0)
2 (3.5)
1 (12.5)
13 (25.5)
0.01
Awaiting renal transplantation N (%)
1 (6.7)
3 (10.0)
3 (33.3)
0 (0.0)
8 (14.0)
2 (25.0)
15 (29.4)
0.02 \0.001
Duration of ICHD before SDU Median
0.8
0.3
1.7
1.0
0.6
0.9
0.7
IQR
0.6–1.6
0.2–0.9
1.1–15.6
0.6–2.9
0.2–1.3
0.3–1.1
0.3–5.8
14 (93.3)
16 (53.3)
8 (88.9)
20 (87.0)
26 (45.6)
6 (75.0)
27 (52.9)
\0.001 \0.001
Fallbacks N (%) Number of fallbacks per patient Median IQR
3.0
1.0
7.0
3.0
0.0
1.5
1.0
2.0–4.0
0.0–2.0
3.0–12.0
1.0–5.0
0.0–4.0
0.5–3.0
0.0–3.0
Number of dialysis during fallbacks Median
1
2
1
2
2
1.5
1
IQR
1–3
1–4
1–1
1–3
1–4
1–2
1–4
\0.001
N number, ICHD in-center hemodialysis, SDU satellite hemodialysis unit; IQR interquartile range
to an ICHDU, and among these 8 had returned to the SDU. Of the 193 patients included, there were 70 (36.3 %) transplantations, 20 (10.4 %) deaths, 8 (4.1 %) transfers outside the region and 1 (0.5 %) renal function recovery. Factors associated with patient fallbacks The global Kaplan–Meier curve of 2-year survival free of fallback is represented in Fig. 2. The cumulative
123
probabilities of fallbacks at 6, 12, 18 and 24 months were respectively: 27.8, 40.9, 53.7 and 62.2 %. In the survival analysis, there was no association between patient characteristics and the occurrence of fallbacks. Binomial negative regression analysis showed that underlying nephropathy (hereditary nephropathy vs. other types: relative risk [RR] 0.43, 95 % CI 0.23–0.83), peripheral arteriopathy (RR 2.08, 95 % CI 1.06–4.06), and the center (center 2: RR 0.44, 95 % CI 0.21–0.97; center 3: RR 3.30, 95 % CI 1.22–8.80) were associated with the number of
J Nephrol
Number of subjects at risk during follow-up 193
6 months
12 months
18 months
24 months
141
120
102
93
Fig. 2 Fallback-free survival curve in the two first years of follow-up
patient fallbacks (Table 4). In the multivariate analysis, peripheral arteriopathy (RR 2.06, 95 % CI 1.05–4.09) and center (center 2: RR 0.42, 95 % CI 0.21–0.84; center 3: RR 2.88, 95 % CI 1.20–6.91) were independently associated with the number of fallbacks.
Discussion To our knowledge, this is the first report on the outcome of SDU patient fallbacks in France. Our findings are in line with the results of a study from Canada [10]. However, in the study of Lindsay et al., the main cause of fallback was access-related, whereas medical problems were the principal causes of fallback in our study. This could be explained by the fact that hemodialysis catheters are not authorized in SDUs in France. Data from the Dialysis Outcomes and Practices Patterns Study (DOPPS) showed that, for French hemodialysis patients, the mean number of annual hospitalizations was 1.43 per patient/year with a mean length of stay of 9.2 days [11]. The principal causes of hospitalization were cardiovascular and gastrointestinal diseases, vascular access issues, or infections. Similar results were observed in the United States [12–18]. Even though the causes of hospitalization in our study were similar to those reported in the literature, the number of infection-related hospitalizations was relatively lower. Again, this could be explained by the type of vascular access used in the SDU in France. In our region, 21 % of patients on hemodialysis were treated in SDUs, which is close to the rate of 20.4 % observed in France as a whole. Patients treated in the SDU
had similar features to those treated in SDUs in France [2]. However, there were more SDU patients who had at least one comorbidity (99 %) compared to patients treated in SDUs (58.4 %). In 2012 there were important discrepancies in France regarding the rate of patients treated in satellite dialysis units. Since there were less MDU facilities in our region compared to the rest of France, this could have influenced the type of patients treated in SDU. Indeed in France, there are no formal criteria for selecting patients eligible for SDU. In some regions, only patients without comorbidity are treated in SDUs while comorbid patients are treated in an MDU or ICHDU. In our region, there are only a few MDUs. Thus patients were more likely to have dialysis sessions in an SDU. Furthermore, in our region, many patients live in rural territories; having the dialysis session in the SDU can reduce transportation time for these patients. These elements could explain the proportion of comorbid patients treated in SDU in our region. Therefore, comorbidities could, in part, have affected fallback incidence. Differences between centers could reflect differences in medical practice between centers, as there was no significant difference regarding patient comorbidities between the centers. At the same time, one may argue that characteristics of the centers, which were not analyzed in our study, could have influenced patient fallback. Our study has some limitations. First, only the main diagnosis was collected for each fallback. In addition, for temporary fallbacks without hospitalization, causes of fallback were not identified. Patients could have been hospitalized outside the region, leading to missing fallbacks in our data collection. These could have led to a lack of power in the statistical analysis. There were some missing data in our study. In particular, the Kt/V value was missing for 53.4 % of the included patients, which is similar to the missing rate observed in the 2012 annual report of the REIN registry. A better characterization of the patient fallback, especially when it becomes permanent, could help nephrologists deal with allocation of the population between incenter units and satellite units. In our study, among comorbidities, only arteriopathy was associated with the number of fallbacks in the multivariate analysis. The center effect observed in our study suggests that differences in medical practice and factors related to dialysis centers, such as SDU distance from the ICHDU, numbers of nephrologists and nurses per patient treated in SDU, number of dialysis generators, and transportation time between the patient’s home and the SDU could influence fallback occurrence. A complementary study from our team including center characteristics will be conducted to study the center effect observed in our first results. Our study provides information regarding the number of
123
J Nephrol Table 4 Univariate analysis: association between covariate and occurrence of fallbacks and number of fallbacks in the first 2 years of follow-up Factor
Fallbacks (n, %)
Gender
p*
Number mean ± SD
RR (95 % CI)
p**
0.24
Women
28 (45.9)
1.5 ± 2.2
Ref
–
Men
72 (54.5)
1.6 ± 2.7
1.09 (0.68–1.73)
0.72
Age at dialysis initiation
0.58
18–35 years
10 (52.6)
1.5 ± 2.0
Ref
–
36–60 years
37 (44.6)
1.5 ± 2.8
1.01 (0.46–2.22)
0.97
More than 60 years
53 (58.24)
1.7 ± 2.2
1.19 (0.55–2.59)
0.65
0.4 ± 0.8
0.35 (0.08–1.60)
0.18
BMI (kg/m2) (missing: 38) \18
0.96 2 (28.6)
18–25
25 (52.1)
1.2 ± 1.7
Ref
–
25–30 C30 kg/m2
37 (56.7) 17 (56.7)
1.8 ± 2.8 2.0 ± 2.7
1.48 (0.85–2.6) 1.68 (0.86–3.29)
0.16 0.13
1.3 ± 2.0
Ref
–
Employment (missing: 30) Full-time
0.48 23 (44.2)
Unemployed or disabled
17 (54.8)
1.9 ± 3.5
1.47 (0.76–2.85)
0.25
Retired
47 (58.7)
1.8 ± 2.2
1.41 (0.83–2.38)
0.20
Smoking status (missing: 17)
0.52
Active smokers
15 (62.5)
1.9 ± 2.7
1.33 (0.70–2.50)
0.38
Past smokers
18 (48.6)
2.0 ± 3.4
1.42 (0.84–2.41)
0.19
Non smokers
60 (52.2)
1.4 ± 2.0
Ref
–
Transportation (missing: 87)
0.68
Ambulance
43 (53.7)
1.8 ± 2.3
0.94 (0.47–1.87)
0.85
Self-transportation
14 (56.0)
1.9 ± 2.3
ref
–
Disability (missing: 6)
16 (61.5)
1.8 ± 2.4
1.11 (0.59–2.07)
0.74
Underlying nephropathy
0.76 0.09
Other than hereditary
90 (55.5)
1.8 ± 2.6
Ref
–
Hereditary Awaiting transplantation
10 (32.3) 14 (43.7)
0.95
0.8 ± 1.5 1.4 ± 2.2
0.43 (0.23–0.83) 0.82 (0.46–1.48)
0.01 0.52
First dialysis in emergency
12 (54.5)
0.50
1.4 ± 1.9
0.89 (0.45–1.75)
0.73
First dialysis with catheter
35 (57.4)
0.57
1.7 ± 2.0
1.05 (0.66–1.67)
0.83
Hemoglobin (g/dl) (missing:15) \11 11–13 [13 Albumin (g/l) median (IQR) (missing: 47)
0.30 43 (54.4)
1.7 ± 2.7
1.71 (0.85–3.46)
0.13
9 (34.6)
1.0 ± 1.8
Ref
–
1.7 ± 2.3
1.71 (0.84–3.48)
0.14
0.97 (0.93–1.02)
0.29 0.99
48 (54.5) 36.0 (32–39)
Dialysis technique
0.82 0.81
HD
95 (51.3)
1.6 ± 2.4
Ref
HDF
5 (62.5)
1.6 ± 2.7
(0.34–2.94)
Duration of in-center HD (months) (missing: 5)
0.52
\6
33 (46.5)
1.3 ± 2.5
Ref
–
6–12
26 (60.5)
2.1 ± 2.7
1.23 (0.88-2.82)
0.12
12–24 [24
9 (45.0) 32 (59.3)
1.6 ± 2.6 1.7 ± 2.1
1.23 (0.57–2.66) 1.29 (0.74–2.23)
0.59 0.37
Comorbidities Hypertension Stroke Transient ischemic attack Coronaropathy
123
88 (51.8)
0.95
1.6 ± 2.5
1.27 (0.64–2.51)
0.49
6 (54.5)
0.18
2.8 ± 5.1
1.82 (0.79–4.20)
0.16
2 (50.0)
0.45
0.8 ± 1.0
0.46 (0.08–2.50)
0.37
13 (56.5)
0.90
1.7 ± 2.5
1.09 (0.56–2.10)
0.80
J Nephrol Table 4 continued Factor Diabetes
Fallbacks (n, %)
p*
Number mean ± SD
RR (95 % CI)
p**
21 (55.3)
0.68
1.9 ± 2.7
1.24 (0.73–2.12)
0.41
1 (25.0)
0.77
4.2 ± 8.5
2.72 (0.75–9.9)
0.13
Arrhythmia Arteriopathy
10 (62.5) 11 (64.7)
0.79 0.59
2.4 ± 2.9 3.1 ± 4.5
1.53 (0.72–3.2) 2.07 (1.06–4.06)
0.26 0.03
Cancer
Chronic respiratory failure
12 (75.0)
0.13
3.0 ± 4.4
2.01 (0.99–4.06)
0.05
Myocardial infarction
7 (63.6)
0.90
1.3 ± 1.3
0.78 (0.30–2.00)
0.60
Chronic heart failure
14 (51.9)
0.90
1.6 ± 2.7
0.98 (0.53–1.82)
0.96
HD hemodialysis, HDF hemodiafiltration, BMI body mass index, n number, SD standard deviation, RR relative risk, CI confidence interval * p value of the log rank test ** p value of the negative binomial regression
dialysis machines required for the patient fallback. Indeed, we found that fallbacks corresponded to four dialysis sessions per patient-year.
Conclusion In satellite dialysis units, patient fallbacks are a frequent event within the first 2 years of SDU and are mainly related to patients’ hospitalization. Nevertheless, the system of SDU operates well, since most of the fallbacks end with the patient’s return to the SDU. Only arteriopathy and the center were associated with the number of fallbacks. Complementary studies are needed to study the center effects. Conflict of interest The authors have do not have any conflict of interest.
References 1. Ministe`re de la sante´, de la famille et des personnes handicape´es. De´cret n° 2002-1198 du 23 septembre 2002 relatif aux conditions techniques de fonctionnement des e´tablissements de sante´ qui exercent l’activite´ de traitement de l’insuffisance re´nale chronique par la pratique de l’e´puration extrare´nale et modifiant le code de la sante´ publique :troisie`me partie : De´crets. JO n° 224 du 25 septembre 2002, p. 15813 2. REIN- annual report, 2012 3. Diamant MJ, Young A, Gallo K et al (2011) Hemodialysis in a satellite unit: clinical performance target attainment and healthrelated quality of life. Clin J Am Soc Nephrol 6:1692–1699 4. Malmstro¨m RK, Roine RP, Heikkila¨ A et al (2008) Cost analysis and health-related quality of life of home and self-care satellite haemodialysis. Nephrol Dial Transplant 23:1990–1996 5. Mazairac AHA, Grooteman MPC, Blankestijn PJ et al (2012) Differences in quality of life of hemodialysis patients between dialysis centers. Qual Life Res 21:299–307
6. Diamant MJ, Harwood L, Movva S et al (2010) A comparison of quality of life and travel-related factors between in-center and satellite-based hemodialysis patients. Clin J Am Soc Nephrol 5:268–274 7. Prakash S, Austin PC, Oliver MJ, Garg AX, Blake PG, Hux JE (2007) Regional effects of satellite haemodialysis units on renal replacement therapy in non-urban Ontario Canada. Nephrol Dial Transplant 22:2297–2303 8. Soroka SD, Kiberd BA, Jacobs P (2005) The marginal cost of satellite versus in-center hemodialysis. Hemodial Int 9:196–201 9. Roderick P, Armitage A, Nicholson T et al (2004) A clinical and cost evaluation of hemodialysis in renal satellite units in England and Wales. Am J Kidney Dis 44:121–131 10. Lindsay RM, Hux J, Holland D et al (2009) An investigation of satellite hemodialysis fallbacks in the province of Ontario. Clin J Am Soc Nephrol 4:603–608 11. Rayner HC, Pisoni RL, Bommer J et al (2004) Mortality and hospitalization in haemodialysis patients in five European countries: results from the Dialysis Outcomes and Practice Patterns Study (DOPPS). Nephrol Dial Transplant 19:108–208 12. Arneson TJ, Liu J, Qiu Y, Gilbertson DT, Foley RN, Collins AJ (2010) Hospital treatment for fluid overload in the Medicare hemodialysis population. Clin J Am Soc Nephrol 5:1054–1063 13. Schoonover KL, Hickson LJ, Norby SM et al (2013) Risk factors for hospitalization among older, incident hemodialysis patients. Nephrology (Carlton) 18:712–717 14. Ng LJ, Chen F, Pisoni RL et al (2011) Hospitalization risks related to vascular access type among incident US hemodialysis patients. Nephrol Dial Transplant 26:3659–3666 15. Lacson E Jr, Wang W, Lazarus JM, Hakim RM (2010) Change in vascular access and hospitalization risk in long term hemodialysis patients. Clin J Am Soc Nephrol 5:1996–2003 16. Dalrymple LS, Johansen KL, Chertow GM et al (2010) Infectionrelated hospitalizations in older patients with ESRD. Am J Kidney Dis 56:522–530 17. US Renal Data System, USRDS (2013) Annual data report: atlas of chronic kidney disease and end-stage renal disease in the united states, National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda 18. Usvyat LA, Kooman JP, van der Sande FM et al (2014) Dynamics of hospitalizations in hemodialysis patients: results from a large US provider. Nephrol Dial Transplant 29:442–448
123
