Factors associated with falls in hemodialysis patients: a case-control study

Objective: to identify possible associations between a higher probability of falls among hemodialysis patients and laboratory values, comorbidities, pharmacological treatment, hemodynamic changes, dialysis results and stabilometric alterations. Method: this was a retrospective case-control study with hemodialysis patients. Patients in a hemodialysis unit who had suffered one or more falls were included in the case group. Patients from the same unit who had not suffered falls were the controls. Data were gathered from the patients’ clinical history and also from the results of a balance test conducted six months before the study. Results: thirty-one patients were included (10 cases and 21 controls). Intradialytic body weight change was significantly greater among cases (p <0.05). Patients in the case group also presented greater lateral instability after dialysis (p <0.05). Other factors such as high blood pressure, antihypertensives, beta-blockers, and lower heart rates were also associated with falls. Conclusion: a greater intradialytic weight change was associated with an increase in risk of falls. Nursing staff can control these factors to prevent the incidence of falls in dialysis patients.


Introduction
Moderate to severe Chronic Kidney Disease (CKD) (stages 3-5) affects 6.8%-9.5% of the population (1) and involves the accumulation of waste substances such as uremic toxins, which cannot be eliminated due to impaired renal function. In Latin America the prevalence of patients treated with hemodialysis (HD) is 451 per million inhabitants (2) . In these circumstances, patients must undergo dialysis several times a week, with the aim of eliminating uremic toxins and excess fluids, as well as rebalancing the concentrations of ions and other substances, which affect the body's homeostasis. These biochemical alterations affect the functioning of organs and systems related to balance; in fact, hyponatremia, which affects 6%-29% of patients receiving HD (3) , has been associated with an increased risk of falls.
The excess liquid that must be extracted from patients varies depending on the weight gain that they experience in the interdialytic period and the difference between their calculated optimum weight or dry weight, which is defined as weight when there is no fluid excess or deficiency, without the presence of detectable peripheral edemas, with normal blood pressure, and with no postural hypotension.
The excess volume is removed during dialysis, which lasts about 4 hours. The higher the weight gain, the higher the ultrafiltration rate required, resulting in an increased risk of hypotension during dialysis or postHD orthostatic hypotension (4)(5) , both situations associated with greater morbidity and mortality in patients receiving HD (6)(7) . Therefore, HD produces hemodynamic changes and acute homeostasis, affecting postural control. Previous studies have observed that, after HD sessions, patients suffer changes in postural control (8)(9) . Likewise, severe CKD, even when treated by HD, results in the progressive deterioration of structures involved in balance. One example is HD-related amyloidosis (10) , which affects joints such as the hip, which plays an important role in postural control in older adults (11) . In addition, patients treated with HD usually have other comorbidities, which require treatment and in many cases lead to the polymedication of HD patients, which poses a greater risk of falls (12) .
Consequently, numerous factors may place HD patients' postural control at risk. Preventing falls in HD patients is essential, because the consequences in terms of quality of life, associated morbidity and reduction in life expectancy are very important (13)(14) . The nursing staff in charge of our dialysis unit is responsible for connecting, supervising and disconnecting the dialyzed patient. In these processes, clinical situations can occur, which are already contemplated by protocols, given that greater postural instability is observed after the sessions. In turn, other subclinical situations possibly related to the factors already mentioned continue to pose a risk, which keeps the incidence of falls high among our patients, at levels similar to those observed in prevalence studies, in which the incidence was between 1-1.6 falls per patient-year (15)(16) .
The aim of this study was to identify possible associations between a higher probability of falls among hemodialysis patients and the laboratory values, comorbidities, pharmacological treatment, hemodynamic changes, dialysis results and stabilometric alterations.

Study design
This was a retrospective case-control study with a ratio of 1 case/2 controls, with hemodialysis patients. The STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guide for observational studies was followed, as recommended by the EQUATOR network.

Study location and period
The study was conducted in the hemodialysis unit of the Infanta Leonor University Hospital (HUIL), in Madrid (Spain), between January and October 2019.

Participants
The study was conducted with 31 patients: 10 cases Patients in the hemodialysis unit who had suffered one or more falls in the last 6 months and who reported this in the questionnaires administered monthly by the nursing staff were considered cases (n=10). Thus, the participants were unaware that the event "fall" was what categorized them as cases in the study. The controls (n=21) were patients from the same unit who did not report falls during the study period and they were also blind to this criterion. The nurses who had collected the data relative to falls were also unaware of this study.

Data collection
After the cases were recruited, the researchers obtained by the dialyzer were analyzed. We also evaluated a balance study conducted 6 months earlier in these same patients, using an AMTI AccuGait force platform, previously validated by other study (17) . In this study, each patient underwent a stabilometry test before (preHD) and immediately after (postHD) the same dialysis session.

Variables
The researchers gathered the following information regarding the general characteristics of the patients: age, sex, body mass index (BMI) and years in renal replacement therapy. Kt/v is Kt divided by the volume (v) of urea distribution.
All these factors were continuous quantitative variables.
Hemodynamic variables collected before and after HD were also analyzed. The following were recorded: 1) systolic blood pressure (SBP), diastolic blood pressure (DBP) and pulse pressure; 2) heart rate; 3) dry weight, preHD weight and postHD weight. Changes and differences between preHD and postHD variables were calculated. All these factors were continuous quantitative variables.
The stabilometry variables studied as factors were extracted from a stabilometry study conducted with hemodialized patients 6 months prior to this study and all the patients that composed this sample were included.
In this stabilometry study, the balance of patients was tested before and after HD, following the same protocol for all patients, which had been used previously in a similar study (18) . The tests were carried out by two people trained to handle the platform in an office prepared for this purpose, where the lighting conditions were the same throughout the day. The variables analyzed were: 1) average displacement range of the center of pressure (CoP) on the Y axis (Y range) and X axis (X range) measured in cm (centimeter); 2) maximum and average velocity (Vymax, Vxmax and Vavg) of these movements measured in cm/s (centimeter/second); and 3) the area that included the displacement of the CoP with 95% confidence (Area95) measured cm 2 (square centimeters).
All these factors were continuous quantitative variables.

Data processing and analysis
After the database was built, it was cleansed. The

Ethical aspects
This study was approved by the Ethics Committee

Results
A total of 31 patients, 19 (61.3%) men and 12 (38.7%) women, participated in the study. No significant gender differences were found among the cases (6 men and 4 Rev. Latino-Am. Enfermagem 2021;29:e3505. women) and the controls (13 men and 8 women). Table 1 shows the differences between cases and controls in terms of age, BMI, years treated with HD, laboratory variables, comorbidities and polymedication status. Participants in the case group were 10 times less likely to be hypertensive than the control group (OR= 0.105, 95%CI=0.02-0.71). A significant difference was observed between mean levels of beta-2 microglobulin between cases and controls, with a 95% confidence level, the mean levels of cases were 0.09-9.39 higher than controls (95%CI=0.09-9.39).  Perez-Gurbindo I, Alvarez-Mendez AM, Perez-Garcia R, Arribas-Cobo P, Angulo-Carrere MT. weight recorded by the nursing team before and after dialysis. Table 3 presents the results.
We analyzed the distribution of HD variables as recorded by the dialysis machine, in addition to the hemodynamic values of blood pressure, heart rate and  The cut-off point was set at 1.1 kg, 1.9 kg and 2.7 kg, which resulted in a sensitivity of 100%, 70% and 40%, respectively and a specificity of 28.6%, 66.7% and 95.2%, respectively. The Youden index (J=0.367) indicated that the point that determined the highest sensitivity and specificity together was 1.9 kg. Table 4 presents the means of cases and controls in terms of the variables obtained in the balance test, performed before and after the same dialysis session. This group of variables did not present a normal distribution (K-S p<0.05), so the Mann-Whitney U test was used to assess whether there were significant differences between cases and controls.
Based on the other results of this study, patients receiving dialysis who fall also had significantly lower heart rates after HD than the controls. Physiologically, heart rate is supposed to increase when volemia decreases.
If this does not occur, brain perfusion may be affected.
The decrease in cardiac stimulus may be due to patients' medication and in fact, this study found that patients receiving HD and who fall are more often treated with beta-blockers, which is consistent with the fact that these drugs are included into the group of those that increase the risk of falls (24) . However, the association between postHD heart rate and beta-blocker treatment was not statistically significant, so this hypothesis was ruled out.
It seems likely that patients with a limited chronotropic response due to intrinsic or extrinsic causes are at greater risk of falls when undergoing HD.
As for the other drugs analyzed here, all belonging to the group of medications that increase the risk of falls, it was observed that patients who fell took antihypertensives more frequently than those in the control group. These findings are based on the association between lower preHD blood pressure values and increased risk of hypotension (25) and falls (26) .
Regarding laboratory values, dialysis patients with a history of falls had higher levels of beta-2 microglobulin.
Increased levels of this protein are due to the passing of time in renal replacement therapy: after 15 years receiving HD, about 80% of patients present with dialysis-related amyloidosis (10) , which can affect structures related to The controls presented significant differences (p<0.05) in the PreHD and PostHD data for all the stabilometric variables, except for V x max. In contrast, the differences related to HD among the cases were not significant (p>0.05) for any of the variables.

Discussion
This case-control study describes is the first of its kind to describe intradialytic weight change as a factor associated with falls in dialysis patients. The study showed that patients in HD who suffer from falls have greater intradialytic weight changes. When patients arrive at the HD session, they are weighed by the nursing staff, who determine if there is excess weight relative to their reference weight or dry weight.
The greater the weight difference, the greater the volume of liquid that should be removed. This usually occurs in patients who do not have good adherence to treatment or dietary guidelines (20) . The session lasts an average of 4 hours, so to remove a greater volume, a higher ultrafiltration rate must be used, which can generate a greater risk of intradialytic hypotension (21) and orthostatic hypotension (22) . Previous studies have related these events with falls, and this study directly demonstrated that intradialytic weight change is a factor related to the risk of falls. Falls are factors of poor prognosis in patients who receive dialysis (14) and after observing an incidence of falls of 32%, similar to the 37% described in a recent study (23) , it would be interesting to analyze the ability to prevent falls in Perez-Gurbindo I, Alvarez-Mendez AM, Perez-Garcia R, Arribas-Cobo P, Angulo-Carrere MT.
motor ability such as joints or the central nervous system.  (29) .
After dialysis, both cases and controls experienced an increase in the ranges, speeds and area of displacement of the CP, indicating the acute effect of HD on postural balance. This effect of HD was consistent with the results of other study (9) and it used a similar methodology. Unlike another study (18) which observed an increase in postHD lateral range associated with a higher risk of falls, in our study, we found no significant differences between Limitations of this study include mainly its retrospective methodology. The time between assessments and the fall that defined the cases varied within the 6-month period considered in this study. Furthermore, the fall that defined a case was always outside the hospital and in the period between sessions, but the circumstances and the exact time of the falls were not taken into account.

Conclusion
This case-control study identified some factors change. Furthermore, the cases also presented lateral instability and a lower heart rate at the end of dialysis than the controls.