In a cohort study, Lacson et al. [5] compared 746 patients who switched from standard HD to LNHD in a dialysis center and 2,062 patients who remained on standard HD. The hourly ultrafiltration rate (UF) decreased from 11 to 6 ml/h/kg in the LNHD group, remaining constant at 12 ml/h/kg in the standard HD group. This result is all the more interesting given that the speed of UF is currently recognized as a risk factor for mortality [6] and that the patients who had switched to LNHD showed a significant increase in interdialytic weight gain of 0.5 kg on average. Interdialytic weight gain is associated with the risk of arterial hypotension during the session [7]. The study of Lacson et al. does not mention this complication in their standard LNHD–HD comparative study. However, Ok et al. [8] reported that 247 patients switched from standard HD to LNHD in a dialysis unit compared to 247 patients who remained in standard HD, matched on 45 criteria and followed for 1 year. The incidence of in-session hypotension requiring saline infusion was comparable at the baseline—60.4 and 67.0 episodes per 1,000 sessions, respectively, decreasing in the LNHD group to 21.2 episodes (p < 0.0001) and increasing to 80.3 episodes/1,000 sessions in the standard HD group (p = 0.15).

While phosphatemia quickly reaches a plateau during the session, phosphate clearance is prolonged throughout the dialysis session, as shown by Gutzwiller et al. [9] with 5-hour sessions. In the “case control” study by Ok et al. [8], the mean phosphatemia over 1 year was significantly lower in the LNHD group (3.87 ± 1.20 mg/dl) than in the standard HD group (4.96 ± 1.14 mg/dl), whereas it was comparable at the beginning of the study (4.63 ± 1.32 and 4.82 ± 1.26). The prescription of phosphate binders, stable over 1 year in conventional HD (82.9% of patients treated at the end of 1 year), fell from 83.0% to 22.4% among patients in LNHD. These results confirmed the cohort study by Lacson et al. [10], which found a significant decrease in blood phosphorus levels in LNHD (5.73 to 5.02 mg/dl (p < 0.001)), whereas it increased in standard HD (5.75 to 5.85 mg/dl (p = 0.01)).

In the Turkish case control study [8], 91 and 85 patients, respectively, were analyzed by echocardiography at the beginning and end of the study. Cardiac parameters (atrial volume, ejection fraction, ventricular end-diastolic diameter, and ventricular mass index) were identical at the baseline between the two groups and did not change in standard HD. In LNHD, atrial and ventricular end-diastolic volume and ventricular mass index decreased, and ejection fraction increased significantly. After 1 year, atrial and ventricular end-diastolic volume and ventricular mass index were significantly lower in LNHD than in standard HD. However, the ejection fraction did not differ between the two groups at 1 year. Blood pressure did not vary significantly in both groups during the follow-up year, but the prescription of antihypertensive drugs, stable in the standard HD group, decreased from 22% to 8% of patients in the LNHD group. In their study of an equivalent duration (1 year), Karur et al. [11] covered right ventricular remodeling by switching 30 patients from standard HD to LNHD compared to 37 patients who remained in standard HD. The right ventricular telesystolic and telediastolic volume indices and the right ventricular mass index decreased in the LNHD patients with a stable ejection fraction, but only the right ventricular end-diastolic volume index significantly decreased in the LNHD group compared to the standard HD group after 1 year. In a non-randomized controlled study evaluating cardiac parameters 6 months after the transfer of 13 patients to LNHD compared to 12 patients who remained in standard HD, the data was in favor of a significant reduction in cardiac fibrosis [12].

In the HEMO study, the plasma β2M level above which an increased risk of mortality is associated is 27.5 mg/l [13]. In a crossover study, Chazot et al. [14] showed that long HD (average of 6.4 hours thrice per week) with low-permeability membranes did not result in a significant decrease in β2M (43.3 mg/l), whereas with a high-permeability membrane under the same dialysis conditions, β2M decreased to 27.5 mg/L. With the same high permeability dialyzer (polysulfone), Eloot et al. [15] showed that over sessions of 4, 6, and 8 hours, increasing the session length increases the mass of β2M extracted, the plasma volume cleared of β2M, and the β2M extraction ratio (ratio of dialyzer clearance to blood flow), confirming a beneficial effect of session length regarding the clearance of medium molecules. In hemodiafiltration (HDF), for equal duration in long dialysis (7–8 hours/session), Maduell et al. [16] showed that a high convective volume (35–50 vs. 20–30 liters/session) still optimizes the β2M reduction rate.

The use of erythropoiesis-stimulating agents (ESAs) was studied in the Turkish case control study [8]. After 1 year of follow-up, the percentage of patients on ESA and the weekly dose of ESA remained stable (53.7% and 2,819 IU/week), while both parameters decreased significantly in LNHD (patients on ESA: 55.5% to 24.7%; weekly EPO dose in LNHD decreased from 3,026 to 1,697 IU/week).

The first two aspects grouped in this paragraph are critical elements of the sustainability of an LNHD program, especially sleep, as highlighted by Laruelle et al. [17]. Ok et al. [8] studied the evolution in standard HD after the transfer to LNHD of memory capacities, cognitive functions and mental health, perceived pain, vital momentum, and anxiety and depression scores. In the standard HD group, an increase in perceived pain and a decrease in vitality and mental health were noted after 1 year. These parameters remained stable in the LNHD patients, who saw an improvement in memory abilities. More recently, Dumaine et al. [18] showed among 36 patients transferred to LNHD that patients whose baseline quality of life scores were below the group median showed significant improvement at 12 months in mental, physical, and chronic kidney disease–related symptoms and impact. It is notable that the duration of dialysis did not impact the quality of life and even appeared to improve it.

In our experience in Tassin with regard to sleep, in a cohort of 36 patients [4], 86% of the patients declared that they had slept at least 4 hours during the session, 30% of the patients finished their night at the center after disconnection, and 81% slept again after returning home (but not those who left for work, who were few in number). A quarter of the patients considered their sleep to be of poor quality and took hypnotics both on dialysis and at home. Recently, Hull et al. [19] evaluated sleep in 36 patients in an anonymized fashion: 24 continued the LNHD program, and 12 left the program. The ability to fall asleep was found to be impaired in 71% and 86% (LNHD continuation and LNHD discontinuation, respectively) of the cases. The ability to stay asleep was unchanged in 42% and 8% of the patients, respectively, whereas the deterioration of sleep and/or the feeling of “restless legs” increased during the sessions in 54% and 92% of the patients, respectively. The overall quality of sleep among patients who remained in LNHD was judged to be unchanged (37.5%), improved (4%), and slightly (33%) or significantly (21%) worse. Patients who left the LNHD program were judged to fare significantly worse overall in 83% of the cases, slightly affected and unchanged in 8% of the cases. In the Tassin’s experience, poorer sleep during the session is not a frequent cause of discontinuation of LNHD, which is also reported by Lacson et al. [10]. Hull et al. [19] emphasized the value of a “sleep kit” with a mask and earplugs to help patients adapt to LNHD. It is also important to emphasize the logistical setup with, where possible, individual rooms or cubicles that promote sleep, as illustrated in Figure 1.

One of the reasons for choosing LNHD is to increase daytime availability for working people, among others. There is little data on this subject. Li et al [20] compared incident patients on LNHD (20 patients) and peritoneal dialysis (80 patients). After one year, the percentage of employed LNHD patients remained stable (from 75% to 80%, mean age 47 years), whereas it dropped in peritoneal dialysis from 60% to 33% (mean age 52 years).

As mentioned above, no randomized controlled trials confirm reduced hospitalizations and mortality in patients treated with LNHD compared to standard dialysis—but cohort data exists. Lacson et al. [10] showed a 25% reduced risk of mortality at 2 years with in-center LNHD compared to standard in-center HD (746 vs. 2,062 propensity score matched patients; p = 0.004). In the study by Ok et al. (242 patients transferred to LNHD vs. 242 patients remaining in standard HD matched on 35 criteria [8]), the 1-year hospitalization rate was 5.4 versus 18.8 days per 100 patients/month (p = 0.002), and 1-year mortality was reduced by 72% (p = 0.02). Another U.S. study compared survival in 1,206 patients treated with LNHD and 111,707 patients treated with standard HD, with a 33% survival advantage in LNHD [21]. More recently, the ERA-EDTA registry confirmed this 27% survival advantage in favor of LNHD [22].

In conclusion, this data, even with scientific limitations, shows that LNHD has many clinical advantages on known criteria associated with mortality (UF rate, phosphatemia; cardiac conditions), and the mortality data is convergent. The patient’s quality of life is not impacted by session length. The main obstacle is sleep, which is not optimal. In spite of this, the surveys show a large adherence of the patients to the modality.

In this facility, LNHD started from its opening on August 4, 1969, at a time when patients were young and active. The LNHD program closed permanently on February 8, 2021. This closure represents a good summary of the practices to be avoided for the sustainability of the LNHD program (Table II).

Since its arrival in 2009, the operational management of the group that owns the CRAT has expressed and repeated its wish to see the LNHD program disappear for financial and organizational reasons. The financial difficulties are concrete with the absence of adequate pricing for the additional costs of the modality (night work, deficit patient/ID ratio, increased expenses). The additional cost of the modality has been calculated at 42€ per session [17]. The decrease in the in-center tariffs made it impossible to compensate for the LNHD deficit when the two modalities coexist. The Regional HealthCare Authorities in 2018 was sollicitated, without success, to expose the difficulties and the risk of disappearance of LNHD to find a solution. The messages remained “dead letter.”

In addition, there are recurrent human resource difficulties in recruiting night staff. These difficulties, combined with a lack of vision and managerial interest in the quality of treatment and a lack of knowledge of medical results, have maintained a deleterious environment for the future of LNHD in the facility for over 10 years. The medical, nursing, and management teams were destabilized by these recurrent assaults, which were a source of demotivation. The conclusion of this observation is that proactive leadership on the part of the management teams is essential for the sustainability of the LNHD program.

“À la carte” dialysis was gradually implemented on the CRAT, facilitated by a large number of stations and a culture of long dialysis with a dedicated area for non-standard dialysis, in contrast to the uniform industrial model of “one size fits all” prescribing denounced by one of the pioneers of dialysis, Lee W. Henderson [23]. This flexibility was also necessary to respond to numerous requests from patients to reduce their session length. The individualized prescription of the duration was also applied in nocturnal dialysis, resulting in durations varying from 5 to 8 hours per session with, as a consequence, permanent activity in the dialysis room, bringing light and noise incompatible with sleep. This was also favored by late afternoon connection times and the end of sessions after 11:00 p.m. In addition, some patients who had been dialyzed for many years in LNHD and who were attached to this modality progressively lost part of their autonomy, complicating the management with a need for stretchering that was incompatible with sleeping LNHD.

Another consequence of the variable duration of sessions was an extremely low patient/caregiver ratio at the end of the night, with inactivity that was deleterious to the organization of work. Finally, the premises used for LNHD were the same as those used during the day, with no particular arrangements to promote relative privacy, whereas the CRAT had moved to new premises in 2011. No thought had been given to the subject, and this was a mistake.

In practice, LNHD must be prescribed to selected patients in self-care or low-medicalized units modality, performed in premises that favor relative privacy, with a specific time frame that does not overlap with evening dialysis, and with uniform non-negotiable session lengths. It can be practiced at home, but scant data exists on this particular aspect. The expected clinical benefits are the same.

Patients are not easy to convince for LNHD despite its undeniable clinical advantages because of the length of dialysis, the wish for many patients to spend the night “in their own bed,” the difficulty to leave the spouse alone for the night, and the fear of not sleeping. It is important to ensure that in the preparation phase for dialysis (“CKD 5 pathway”), patients are informed about this modality if it is available and about its clinical and socio-professional advantages. Moreover, the “sleep kits” mentioned above represent a measure that can improve adherence if the fear of sleep disturba,nce is the main obstacle [19]. Moreover, access to transversal functions such as dietary, psychological, or social care is more complicated as patients often refuse to return to the dialysis center during the day and outside of the sessions. The development of telephone or video interviews can provide solutions that were well tested during the COVID-19 pandemic. Finally, patients’ associations or testimonies, such as those of Fabrice Huré [3], are crucial. Given the refusal of the ARS to discuss LNHD within the former CRAT, the mobilization of patient associations could have changed the situation.

Caregivers may not commit to LNHD for fear of the lack of vigilance with night work and the absence of a doctor on site in the event of a serious event linked to care, suc-h as accidents involving the disconnection of the vascular approach. The programs designed with only one caregiver (1 nurse for 6 patients in a selfcare-unit ) are not necessarily adapted, posing the problem of isolation in the event of an incident and of the “isolated worker” for the caregiver, which is subject to strict regulations.

Nephrologists are not always convinced of the relevance of LNHD in the absence of randomized controlled trials. They are the linchpin of patient education, which is crucial. It is the job of our learned societies to promote this modality and to convince them because the data and feedback are abundant and positive.

Management must be convinced that this care is optimal and beneficial for patients. Collaboration with nephrologists is essential to make or keep the project or program durable. They must explore all possibilities of financing projects proposed by regional agencies to finally make LNHD emerge as an innovative modality, even if it has existed almost since the beginning of dialysis.