A fall is a definite event in which a person loses balance and ends up on the floor or some firm ground. It is more prevalent in the population aged ≥65 years, in which the risk of falls has been estimated at 27.1%–34.8% of the total reported security incidents, making it a public health problem with a significant impact on hospitalized patients. This common adverse event reported in hospitals can result in multiple complications, ranging from minor blows or trauma to severe injuries and death (Healey et al., 2008; Anderson et al., 2016).

Different factors in older adults in the outpatient setting can increase the risk of falls, such as living alone, a decrease in instrumental capacity, previous falls, low blood pressure, weight deficit, cognitive impairment, and the use of drugs with anticholinergic load or others that increase the risk of falls [fall risk-increasing drugs (FRIDs)] (Machado-Alba et al., 2016; Machado-Duque et al., 2017; Zia et al., 2017; Andersen et al., 2020; Seppala et al., 2021). Male sex, history of hip replacement, previous falls, frailty, confusion, alterations in balance, and use of psychotropic medications or FRIDs are risk factors for falls in hospitalized patients (Krauss et al., 2005; Aryee et al., 2017; Castaldi et al., 2022; Ramos et al., 2023). There are few studies on medication use and falls in older adults in the hospital context or on the differences in the estimated risk of falls caused by different medications with psychotropic and sedative capacities nor are there many assessments of the association between falls and the anticholinergic risk scale (ARS) and FRIDs. Studies of hospital falls have focused mainly on the risks associated with the caregiver and nursing staff, as well as the comorbidities themselves or external factors, without clearly identifying the risk associated with medications and how these can be predictors of worse outcomes (Anderson et al., 2016).

The ARS was devised in 2008 by Rudolph et al. (2008). They hypothesized that drugs with anticholinergic potential are associated with falls or delirium, and they constructed a categorical scale in which the highest scores are associated with a greater probability of adverse events. The FRID scale proposes that the use of some psychotropic medications, such as opioids, antipsychotics, anxiolytics (benzodiazepines), hypnotics/sedatives, and antidepressants, increases the risk of falls, with greatest risk arising when a patient takes two or more drugs from this group (HR: 1.67; 95% CI: 1.58–1.77; Laflamme et al., 2015).

This association between the use of potentially inappropriate medications and falls in older adults is stronger in the presence of more comorbidities, which results in the prescription of more medications (Johnell and Fastbom, 2008) and increases the probability of pharmacologic interactions and other problems related to medications. A correct pharmacotherapeutic evaluation of a hospitalized patient, especially a patient aged ≥65 years and with polypharmacy, can identify inappropriate or risky prescriptions and can reveal the probability of suffering a fall and the potential need to avoid particular higher-risk medications (Maly et al., 2019).

Falls in hospital patients have multiple causes, potentially including risky medications. Additionally, available literature on risk factors is sparse, and no study on this topic has been carried out in Colombia. We aimed to establish the associations between the use of drugs according to the FRID classification and their anticholinergic load and the occurrence of falls in adults aged ≥65 years during their hospitalization at four hospital centers in Colombia from 2017 to 2020.

A total of 250 older hospitalized patients with one or more falls were analyzed from the included hospitals in Colombia during the study period. Four controls for each patient were included, for a total of 1,000 controls. The mean age of all patients (cases and controls) was 77.4 ± 7.4 years. Men predominated overall (n = 800, 64.0%) and in both groups.

The main reason for admission of patients and controls was community-acquired pneumonia (12.4% and 12.5%, respectively), followed by decompensated heart failure (10.8% and 10.5%), with no significant differences. No differences were found in the frequencies by sex or admission diagnoses between cases and controls, nor in the mean age. The most frequent comorbidities found were arterial hypertension (64.8% of patients and 64.3% of controls, p = 0.134) and a history of coronary disease (16.4% of patients and 30.2% of controls, p < 0.001). For most of the concomitant diseases, there were no significant differences between the groups, except for delirium, vertigo, coronary disease, and cerebrovascular events. Table 1 shows the sociodemographic variables, admission diagnoses, and comorbidities of the patients and controls, as well as the bivariate comparisons for each category.

Most of the falls occurred during hospitalization (n = 192 patients, 76.8%). Most often, the falls were related to the process of standing alone on the way to the bathroom without the help of a family member or nursing staff (n = 113, 45.2%).

The most prescribed drug groups during hospitalization were antihypertensives (88.0% in cases and 75.1% in controls), statins (62.0% in cases and 51.1% in controls), acetylsalicylic acid and other antiplatelet agents, proton pump inhibitors, antipyretic analgesics, NSAIDs, opioids, and psychotropics. For almost half of the groups of drugs identified, statistically significant differences were found in the frequency between cases and controls, including some antihypertensive, lipid-lowering, first-generation antihistamines, antidiabetics, antiparkinsonians, anticonvulsants, typical and atypical antipsychotics, benzodiazepines, antidepressants, opioid and nonopioid analgesics, and even disease-modifying antirheumatic drugs. Table 2 shows these groups and their differences in bivariate comparisons.

The patients used more antihypertensive and psychotropic drugs in the FRID classification as well as other drugs than the controls (p < 0.001). When evaluating the ARS scores, a higher mean score was identified in the cases (mean: 2.3 ± 1.9) than in the controls (mean: 1.4 ± 2.2) (p < 0.001). The most commonly used anticholinergic medications were trazodone (16.8% vs. 10.8%), quetiapine (12.4% vs. 3.1%), and hydroxyzine (8.0% vs. 3.9%) among cases and controls. The proportions of individuals with moderate (13.6% vs. 7.3%) and high anticholinergic risk (37.2% vs. 24.7%) were greater among the patient group. Polypharmacy, including both the mean number of drugs and the proportion of patients treated with five or more drugs, was greater among the patients than the controls (p < 0.001). Table 3 shows the frequencies and proportions of FRID use, polypharmacy, and ARS score.

This case‒control study in older hospitalized adults found that the risk of falling during the hospital stay was positively associated with the use of FRIDs and drugs with anticholinergic load. This provides evidence for the first time in Colombia of the use of potentially inappropriate drugs in this age group in the context of care within hospitals. These findings should elicit new strategies for the safer use of drugs.

The mean age of the patients and controls included was 77 years. In such a population, with their multiple comorbidities and high risk of frailty (Mudge and Hubbard, 2019), a fall can bring adverse consequences. In a study in Australia, Decalf et al. reported that falls were more frequent in men and in patients older than 70 years, and they occurred mainly in the bathroom and during the night (Decalf et al., 2019). Interestingly, suffering from delirium was also associated with a greater probability of suffering a fall during hospitalization, in line with earlier findings (Hoffman et al., 2019; Wedmann et al., 2019). Therefore, strategies to reduce the risk of delirium include nonpharmacologic measures like early mobilization, adequate hydration, and sleep enhancement.

Moreover, orientation to time and place, and hearing and vision optimization (Oh et al., 2017), as well as trying not to use drugs with an anticholinergic load and others that facilitate its appearance, can also have a positive impact on preventing hospitalized patients from falling and suffering related adverse effects (Tune and Egeli, 1999; Han et al., 2001; Cox et al., 2009; Salahudeen et al., 2016).

Another relevant finding was the association between polypharmacy (≥5 drugs) and an increased risk of falls (Gnjidic et al., 2012), similar to the findings in older hospitalized adults in Brazil reported by Alves–Ramos et al., who established that the association was attenuated after adjusting for comorbidities and increased by the use of FRIDs (Ramos et al., 2023). In Malaysia, Zia et al. recognized that polypharmacy was a predictor of the risk of falling before adjustment for FRIDs (Zia et al., 2017), and polypharmacy lost its significance after the inclusion of FRIDs, so these types of drugs are more significant than the number of active ingredients received by the patient (Gnjidic et al., 2012).

This analysis reflects the multifactorial nature of falls. The causative factors range from advanced age to the presence of comorbidities to the use of certain drugs, especially those with an anticholinergic load, as has been described in outpatients for almost 20 years. This can be attributed to the lack of accompaniment of relatives and nursing personnel, architectural barriers, and safety and patient protection elements (Rudolph et al., 2008). We identified an increased risk of falling (1.6 and 2.3 times greater for patients with low and high anticholinergic load, respectively) in hospitalized adults aged 65 years and older than in those not taking drugs with anticholinergic load. This finding is consistent with several previous studies, such as that of Akgün et al. in the Netherlands, who reported that each one-point increase in the ARS score increased the risk of suffering a fall, with an OR of 1.49 in older hospitalized adults (Akgün et al., 2022), and that reported by Dauphinot et al. in France, who found that higher anticholinergic load increased hospital falls (Dauphinot et al., 2014). This risk is also similar to that found in a study carried out in Colombia in older adults in an outpatient setting who suffered a fall with a hip fracture, in which a higher ARS score increased the probability of falling, which shows an important effect of acetylcholine-mediated peripheral and central nervous system blockage, particularly in older people in both outpatient (Collamati et al., 2016; Bishara et al., 2017) and inpatient settings (Wawruch et al., 2012).

FRIDs (antihypertensive, psychotropic, and others) were associated with the probability of suffering a fall within the hospital environment after multivariate adjustment, as has been identified in ambulatory patients (Milos et al., 2014; Zia et al., 2017; Castaldi et al., 2022; Ramos et al., 2023). It is important to know that the mechanisms of action and adverse reactions of different groups of FRIDs affect adults more than 65 years of age differently, but one thing they share in common is that they enhance the probability of falls, some of which are due to mechanisms such as decreased alertness, the induction of sedation or delirium, and other nonpsychotropic effects such as the generation of orthostatic hypotension or hypoglycemia (Park et al., 2015; Richardson et al., 2015; Nwadiugwu, 2020). However, it is clear that those drugs with an effect on the central nervous system had a stronger overall effect, as well as having individual associations with the risk of falls, as found for antiepileptic drugs and antipsychotics in our multivariate analysis.

The increased use of psychotropic medications, whether typical or atypical antipsychotics, increased the risk of falls in this population, as they were probably prescribed during an episode of psychomotor agitation or during a patient’s hyperactive delirium (Gareri et al., 2014) or in patients who suffered from a comorbidity such as schizophrenia or dementia. Importantly, the use of antipsychotics during an episode of delirium does not truly change the duration, severity, or mortality during the episode, but it does increase the risk of falling during the hospital stay (Neufeld et al., 2016). Antipsychotics can worsen cognitive decline in patients with dementia by altering dopamine levels in the central nervous system; these patients are included in the FRID group and have an anticholinergic load (Rudd et al., 2005; Rudolph et al., 2008; Siafis et al., 2018). The findings of this analysis provide evidence of the undesirable effects of this group of psychotropic drugs in hospitalized adults aged ≥65 and suggest the need to increase precautions for their use, avoid potentially inappropriate prescriptions, carry out greater surveillance and follow-up by healthcare teams, and improve patient safety systems (Prendergast et al., 2022).

The finding in the multivariate analysis of the association between the use of iSGLT2 and falls in adults aged ≥65 is quite interesting because, due to their relatively recent incorporation as antidiabetic drugs, it is striking that they may increase the risk of falling and suffering other adverse outcomes, perhaps related to their effect on glucose, diuresis, or hypotension. There have been reports of falls in older adults who were taking them (Taşdemir et al., 2020), and some authors have suggested deprescribing iSGLT2 and diuretics in older adults with heart failure due to the risk of falls (van Poelgeest et al., 2023). Additionally, iSGLT2 has been associated with altered bone metabolism and increased risk of fractures, so its use in older adults, who more frequently present osteoporosis, sarcopenia, and risk of fractures, should be considered (Ye et al., 2018). Further studies are warranted to explore this possible association.

With these results as a foundation, it should be possible to include older adults who take risky medications in the category of patients at high risk of falling, leading to greater focus on care by healthcare personnel and accompanying family members for these patients.

This study had some limitations, such as not establishing the date of drugs initiation, the dose used, the duration of the prescription during the hospital stay, the laboratory data, or the severity of the clinical conditions, all of which can influence the risk of falls. Some of its strengths are the union of information from four hospital centers in Colombia, a significant sample size of cases and controls, and the innovation of evaluating both the anticholinergic load and the prescription of FRIDs in hospitalized older adults in Colombia.

Adults aged ≥65 years are more likely to suffer a fall in the hospital if they are given a drug with anticholinergic load or a FRID, as well as if they have some comorbidities, especially delirium, vertigo, or heart disease (coronary ischemia and stroke). There is a clear need to prioritize the implementation of safe prescription criteria in older adults, to carry out a complete evaluation of the effectiveness and safety of the drugs used, and to recognize the potential risk of adverse reactions such as falls. It is important to disseminate these findings among those who care for patients and to carry out new investigations into the relationships that may exist between different medications and the risk of suffering a fall in the hospital.