Background

Falls constitute a leading preventable cause of geriatric injuries and hospitalization, prolonged recovery times and deaths [1, 2]. More than a third of older adults fall each year, meaning worldwide a high number of patients [3]. Fall-related injuries can be serious, occuring approximately in 10% of falls, and consequencing an urgent hospitalization [3, 4]. Furthermore, falls decrease self-sufficiency and quality of life among older adults, because falls can result in fear of falling, loss of confidence, mobility and ability to live independently [5].

Fall risk is a multifactorial problem associated with numerous intrinsic and extrinsic factors [6, 7]. Intrinsic factors are related to ageing, including age-related changes to the pharmacokinetic and pharmacodynamic effects of pharmacotherapies. Age-related pharmacokinetic changes are represented by drug absorption, distribution, metabolism and elimination route. The drug absorption is usually decreased due to changes in the gastrointestinal tract, e.g., reduced gastric secretion, loss of mucosal intestinal surface and decreased blood flow in splanchnic area. The drug distribution is affected by many factors, e.g., increase in fat compartment, decrease in total body water, muscle mass and serum albumin level. The distribution volume of water soluble drugs is decreased leading to their potential toxicity, while the distribution volume of lipid soluble drugs is increased leading to prolongation of elimination half-life and accumulation of the drug in fatty tissues. The drug metabolism is influenced by decreased hepatic blood flow and low activity of liver enzymes. The elimination route is limited due to reduction in renal clearance [7, 8]. Pharmacodynamic changes involve altered sensitivity to many pharmacological agents, increased sensitivity in psychotropic or cardiovascular drugs, in contrast with decreased capacity to respond to physiological challenges and side effects of the drug therapy, e.g., orthostatic hypotension [1, 9].

Extrinsic factors include fall risks related to the environment where individuals live, e.g., how medicines are taken at home [6, 10]. Prescribed medication is an important and potentially underappreciated contributor to falls [11, 12]. The older adults are susceptible to polypharmacy and higher risk of falls [13]. Polypharmacy can increase the risk of medication-related falls, especially when at least one established fall risk-increasing drug is part of the patient’s daily regimen [14]. A leading mechanism for increased risk can be sedation which slows reaction time, orthostatic hypotension, syncope, dizziness, drowziness, and blood pressure change or impaired balance [15]. Normally, long-acting medications pose a higher risk, but the metabolism and half-life of otherwise short-acting drugs can be also prolonged in older adults [16, 17]. Previous studies have presented various methods reducing fall risks in older adults, such as supplementation of vitamin D and calcium, cataract surgery, hip protectors, modification of hazard environment and adjustment of medications [18, 19]. In addition, stability of the human body is a crucial factor for reducing falls [20]. Recommended actions to improve stability include exercise programs focusing on balance and muscle strength [21]. There is a good evidence for reducing medication-induced falls. In psychotropics, the fall risk can be reduced by prospective management of adverse effects such as drowsiness, dizziness, slow reaction time and orthostatic hypotension [22, 23]. Drugs reducing blood pressure are associated with fall risk because of their hypotension effect. Thus, adjusting anti-hypertensive medication may reduce syncope and falls [24, 25]. The fall risk prevention with anticoagulants should focus on bleeding events such as a cerebral hemorrhage and associated falls [26]. Many drugs disturb vision and vestibular system, which exacerbate gait disturbance in older adults [27, 28]. In case of falls, reduced bone mineral density and osteoporosis can increase risk of fracture and more severe consequences [29].

Evidence on medication-related fall risks is not always integrated into multifactorial risk assessments which contribute to preventive programmes for reducing falls [30, 31]. Incorporating these risks could improve accuracy of existing tools routinely used in geriatric care units as many of the current workflows do not contain medication as a risk factor [32]. This study aimed to provide a better screen on falls in institutions by developing a preliminary categorization of fall-risk-increasing drugs (FRIDs) that could be added as a risk factor to the existing fall risk assessment tool routinely used in geriatric care units.

Methods

Patient data

The patient data used in this study were retrospectively derived from two different types of health care institutions, i.e., a hospital and a nursing home in Brno, Czech Republic. Medication data on all patients who had fallen at least once during their stay at a health care institution (hospital or nursing home) within a 2-year observational period were derived from patient records. The fall is defined by World Health Organization as an event which results in a person coming to rest inadvertently on the ground or floor or other lower level [31]. The falls during a stay at a health care institution were documented as undesirable events in a protocol, which contained information about circumstances of the fall such as a description of fall, exact time and place, and consequences, e.g., severe injuries, bone fracture or head injury, bruises or bleeding risk. Descriptive and demographic information on patients (sex, age, length of stay, MMSE score and medication use) were collected from the patient records. Inclusion criteria of the study participants were 1) age ≥ 60 years old, 2) evidence of at least one fall during their stay in a geriatric care unit, 3) evidence of fall documented by health care professionals (nurses, physicians). Exclusion criteria were 1) age < 60 years old, 2) absence of falls during their stay in a geriatric care unit, 3) no evidence of falls documented by health care professionals. The nursing home had less descriptive records of their residents, e.g., they missed a cognitive impairment scale.

Developing categorization of medicines increasing fall risk (FRIDs)

The medication-related fall risk was determined by categorizing all medicines patients used according to each medicine’s fall risk. The criteria for this categorization was derived from the staturory summaries of product characteristics (SmPCs) approved by authorities as part of marketing authorization within European Union countries [33]. The SmPC information of each medicine was systematically reviewed by two researchers (JM, KV) to derive the fall risk. We applied information on a direct connection between the fall risk and adverse drug effects (ADEs) disturbing the patient’s balance as presented in SmPCs. This information was used to create an A, B, C medication-related fall risk categorization. Usually the mechanism leading to medication-related fall risk is one or more of the following adverse drug effects: sedation, orthostatic hypotension, syncope, dizziness, drowsiness, changes in blood pressure or impaired balance [34, 35]. All these adverse drug effects have a negative influence on a patient’s balance, which can then predispose the patient to falls [36].

Adverse drug effects posing fall risk are classified in the SmPC as very common (≥1/10), common (≥ 1/100), uncommon (≥ 1/1000), rare (≥ 1/10000) and very rare (< 1/10000) (Table 1). Medicines with at least one of the above-mentioned adverse drug effect frequencies of “very common” or “common” in the SmPC were categorized as high fall risk medicines (category A). Similarly, the medicines with an adverse drug effect frequency “uncommon” were categorized as moderate fall risk medicines (risk level B) and those with rare or very rare frequency as no fall risk medicines (risk level C). The high-risk category A medicines are considered as a predisposition factor to fall, because the occurrence of adverse drug effects connected to falls is documented by at least one patient per each 100 patients using the same active substance. The risk information provided in the SmPC of the brand product used by the patients involved in the study was validated by comparing the information with the SmPC information of another similar product on the market.

Table 1 Fall risk level criteria: categorization according to information derived from the summaries of product characteristics (SmPC) on the frequency of adverse drug effects (ADEs) contributing to fall risk (sedation, orthostatic hypotension, syncope, dizziness, drowsiness, changes in blood pressure or impaired balance)
Full size table

Frequency of use of high fall risk medicines in study participants with a fall history

The second phase of the study focused on counting the frequency of use of high fall risk medicines in study participants with a fall history in institutions. This phase made use of the existing routine practice at the time of the study of assessing and scoring all patients for fall risk by nurses in the routine admission procedure. Both institutions used the same set of 4 risk items other than medications to estimate fall risk (Table 2). The original Morse Fall Score [37] had been shortened to a time-saving procedure of 4 items suitable for clinical practice. Each fall risk item yielded 1 point and they assessed 1) history of falls, 2) mental condition, 3) physical condition, and 4) occurrence of dizziness or drowsiness in older patients. The estimation was based on patient interviews on admission or notes from patient records. Direct questions to the patient were dependent on the patient’s health condition. As this fall risk estimation did not consider patient’s medication as a contributor to the fall risk, the patient data on medication use were retrospectively collected from the patient records by the researcher (JM). All the medicines were classified into three risk categories (high, moderate and none) using the new categorization based on the fall risk information in the SmPCs. For the descriptive feasibility testing, the medication data of the patients in the hospital and nursing home were combined to provide the widest possible coverage of medicines used by older adults in institutions.

Table 2 The items included in the existing fall risk estimation routinely used in the two health care institutions participating in the study (a hospital and a nursing home)
Full size table

Statistical analysis

Descriptive statistics were used in both study phases to categorize active substances according to their fall risk, and to test the categorizations with a retrospective sample of patients from a secondary care hospital and a nursing home. All patients were grouped together for the analysis. Data were processed in Microsoft Excel, version 2016. A pivot table was created to analyze worksheet data. Descriptive statistics were used for patient data to count metric items such as frequencies, means, ranges, and standard deviations. The categorization of medicines according to their fall risk used ATC codes, specified in levels of anatomical main group, therapeutic subgroup, and pharmacological subgroup [38]. The commonly used active substances in the high fall risk category A were processed from the retrospective patient data as the frequency of use.

Results

Study participants

In total 188 fall-experienced older patients were retrospectively drawn from the patient registers of both institutions included in this study covering the 2-year period from January 2016 to December 2017 (Table 3). Of the patients, 128 (68%) were from the geriatric department of a secondary care hospital and 60 (32%) from a nursing home. Of them, 103 were females and 85 males with the mean age of 79 years (range: 60–97 years, standard deviation ±18.5). They used on average 9.3 medicaments (range: 2–17, standard deviation ±7.5) during a stay in a health care institution. The mean length of stay of the study patients in the secondary care hospital (n = 128) was 15.8 days (range: 1–56 days, standard deviation ±27.5). The most common causes of hospitalization were cardiovascular problems, and/or respiratory and urogenital tract infections. The half (52%) of the patients had at least a mild cognitive impairment (≤ 23 points), the mean score in the Mini Mental State Exam (MMSE) being 21.7 points (range: 5–30 points, standard deviation ±12.5). The stay for nursing home residents (n = 60) was permanent. Data about their cognitive function was not available. They suffered from many chronic diseases (cardiac, neurologic) and used chronic long-term medications.

Table 3 Characteristics of the participants from two health care institutions included in the study (n = 188, of which hospital inpatients, n = 128; nursing home residents, n = 60)
Full size table

The fall risk scores of study participants assessed on admission

Medications were not included in the routinely used assessment. Nurses had assessed and scored all patients (n = 188) for fall risk on admission using the original 4 risk items (score range 0–4). This fall risk assessment is a part of the regular procedure in each patient admission. More than half (53%, n = 100) of the patients were scored as having at least high fall risk (had 3 or 4 risk scores), the remaining patients (46%, n = 88) being scored with medium or lower risk (Table 4).

Table 4 Distribution of fall risk scores in fall-experienced patients (n = 188) using 4 fall risk items routinely assessed by nurses on admission (each item yielding 1 point, score range 0–4). Medications were not included in the assessement
Full size table

Categorization of medicines according to their fall risk by applying SmPCs and frequency of use

Fall-experienced patients (n = 188) used altogether 1748 medicaments. Of these medicaments, 710 (41%) belonged to the category of high fall risk (category A), 331 (19%) moderate risk (category B) and 707 (40%) no risk (category C). The mean number of high-risk medications (category A) in use per patient was 3.8 medications, minimum 0 (n = 2), maximum 8 (n = 4) and median being 3 (n = 46). The majority (93%) of high-risk category A medicines (n = 710) used by the patients were for the nervous system (53%, n = 375) and the cardiovascular system (40%, n = 281) (Fig. 1). The nervous system medicines in use were mainly psychotropic medications (73% of the medicines in use in this category), while analgesics (16%), antiepileptics (8%) and antiparkinson drugs (3%) represented a minority. Psychotropic medications (n = 274) consisted of 1) psycholeptics (n = 179, representing 65% of the medicines in this category) such as antipsychotics (n = 109, 40%), anxiolytics (n = 55, 20%), hypnotics and sedatives (n = 15, 5%), and 2) psychoanaleptics (n = 95, representing 35% of the medicines in this category) such as antidepressants (n = 65, 24%) and anti-dementia drugs (n = 30, 11%). The high fall risk cardiovacular medications (n = 281) included agents acting on the renin-angiotension system (RAS) (n = 92, 33% of the medicines in use in this category), beta blocking agents (n = 85, 30%), cardiac therapy (n = 54, 19%), calcium channel blockers (n = 27, 10%) and other therapeutic subgroups (n = 23, 8%).

Fig. 1
figure 1

Categorization of high fall-risk Category A medicines (n = 710) according to therapeutic group and number of users. The medicines are organized according to the ATC Classification [38].. Anatomical main group (1st level, dark grey), Therapeutic subgroup (2nd level, blue), Pharmacological subgroup (3rd level, grey)

Full size image

Of the 216 active substances used by the patients, 102 (47%) were classified in the high-risk category (fall risk level A), 29 (13%) in the moderate risk category (fall risk level B) and 85 (40%) in no risk category (fall risk level C). The leading mechanisms of ADEs consequencing in high fall risk of most commonly used active ingredients by study participants are presented in Table 5. The most commonly used active ingredients in the high fall risk category A (n = 710) are presented in Table 6, included metoprolol (n = 49, beta blocking agents), perindopril (n = 38, ACE inhibitors), tiapride (n = 36, antipsychotics), quetiapine (n = 24, antipsychotics), tramadol (n = 24, opioids), citalopram (n = 23, antipressants), amlodipine (n = 23, calcium channel blockers), melperone (n = 23, antipsychotics), bromazepam (n = 20, anxiolytics), digoxin (n = 18, cardiac glycosides), and ramipril (n = 18, ACE inhibitors).

Table 5 The most commonly used fall risk increasing drugs (Category A) having very common (≥1/10) or common (≥1/100) frequency of ADEs connected to fall risk according to statutory summary of product characteristics (SmPCs)
Full size table
Table 6 The preliminary categorization of high fall risk medicines (Category A) basing on their adverse drug effect (ADE) profiles in statutory summaries of product characteristics (SmPCs) and frequency of use in older patients who had experienced at least one documented fall in a geriatric care unit (n = 188). The medicines are organized according to the ATC Classification [38]
Full size table

Discussion

This study developed a preliminary categorization to identify medicines that may increase a fall risk in older adults in institutions, by using evidence derived from statutory medicines information and retrospective patient records. Our results indicate that numerous medicines have such an ADE profile that can lead to an increased fall risk and that the use of these high fall-risk medicines, even concomitant use of several active ingredients, is common in older adults during a hospital stay or while living in a nursing home. These findings suggest incorporating a medication-related fall risk indicator in a routine fall risk assessment on admission to geriatric care units. In long-term care units such as nursing homes, it may be useful to repeat the fall risk assessment regularly, e.g., at least once a year.

The high-fall risk categorization presented in this study considered also patient data which gives an idea of ​​the prevalence of clinical use of different medicines. By this way it was possible to identify a few widely used medicines with a high fall risk. The top three clearly most commonly used high fall risk active substances were metoprolol (beta blocking agent), perindopril (ACE inhibitor) and tiapride (antipsychotic agent). Of the top ten active ingredients, 6/10 were psychotropics (3 of them antipsychotics) and 4/10 cardiovascular agents. Optimizing the use of these few medicines in terms of prospectively managing fall risk could make a remarkable change in the incidence of actual falls in geriatric care units.

Our findings concerning the “ranking “of drugs with a high fall risk are in line with previous research. The association with falls has been consistently reported for psychotropic and cardiovascular medicines [39,40,41,42,43,44,45]. The recent systematic reviews and meta-analyses (Seppala [43] and de Vries [44]) confirmed the association between certain drug classes and fall risk from psychotropics, i.e., antidepressants (OR = 1.57), antipsychotics (OR = 1.54), and benzodiazepines (OR = 1.42), and cardiovascular drugs, i.e., cardiac glycosides (OR = 1.6), antiarrhythmics (OR = 1.27), vazodilatators (OR = 1.03), and ACE inhibitors (OR = 1.03). The increased fall risk is also evidenced in patients using opioids (OR = 1.61), antiepileptics (OR = 1.55), anti-Parkinson drugs (OR = 1.54), and NSAIDs (OR = 1.09) [46, 47]. Furthermore, the previous study from Czech Republic (Maly et al. [48]) classified drugs that affected to the nervous system (antipsychotics, antidepressants, analgesics) and to the cardiovascular system (diuretics, beta blocking agents, agents acting on the renin-angiotensin system) as the most frequently used fall risk-increasing drugs in hospitals. Thus, many studies have reached similar conclusions, although the fall-increasing medicines have been presented slightly in different order. The identification of increased risk of falls in different drug classes might be a crucial risk factor for falls. The European Geriatric Medicine Society and Finnish Expert Group on Fall-Risk-Increasing Drugs (FRIDs) have concluded that the knowledge about the risk of falls associated with therapeutic classes and individual medications can help in fall prevention [49].

According to our findings, the fall risk assessment tool presented in this study could be helpful to prevent medication-related falls and increase quality of geriatric care in health care institutions. This is supported by the findings of previous studies that have reported the medication use as a remarkable but modifiable fall risk factor [48, 50]. Therefore, it is reasonable to implement a strategy to avoid use of fall risk-increasing drugs in the routine practice in geriatric care units. Nevertheless, many of developed fall risk assessment tools do not consider medication as a risk factor included in the tool. A review of twenty fall risk assessment tools found that only seven of the tools consider medication use as a risk factor [32]. More recent tools include medication use in a fall risk assessment, e.g., The Johns Hopkins Fall Risk Assessment Tool: it ranks opiates, anticonvulsants, anti-hypertensives, diuretics, hypnotics, laxatives, sedatives, and psychotropics among high fall risk drugs [51].The Johns Hopkins Fall Risk Assessment Tool is well-structured and validated, but the total number of risk points to evaluate is high (maximum 28). This makes it more time consuming for health care providers to assess the fall risk than by using a tool with fewer risk points, such as the five items suggested to be used in our updated fall risk assessment tool. Our tool is also unique in the way that the list of fall risk medicines was created basing on statutory medicines information presented in the summaries of product characteristics approved by European Union (EU) authorities.

This study indicates that the current practice of fall risk estimation in geriatric care units without including medications underestimates the actual fall risk. More than half (53%, n = 100) of the patients included in our study scored at least a high fall risk, but the remaining 46% (n = 88) of these fall-experienced patients scored a medium risk (35%), a low risk (9%) or no risk (3%). The addition of a new item scoring the exposure to the high fall risk medication could refine the prospective risk assessments. This new risk item represented by high fall risk medicines (FRIDs) should be considered in each patient admission or medication change and periodically assessed in the existing fall risk assessment tool by health care experts.

Fine-tuning the fall risk scoring system requires further research and comparisons with the contents of other existing tools. Further research should compare how much adding high risk medicines to the fall risk assessment tool changes the risk score and makes the risk score more accurate in terms of predicting and preventing actual falls. It also might be useful to assess whether concomitant use of more than one FRID medicines will elevate the fall risk. Determining this in future studies may be necessary because the concomitant use of several drugs that increase the risk of falls was common in our data (on average 3.8 high risk medicaments per patient). In addition to this kind of risk verification research, more generalizable results are needed by testing the fall risk assessment tool with a larger number of geriatric patients from a larger number of hospital wards and nursing homes. These studies should be based on prospective patient data. Furthermore, the fall risk profiles and patterns of the high fall risk medicines should be further investigated in older adults in hospital and home care settings.

Study limitations

When interpreting the results of this study it is important to keep in mind that the study does not cover the complete spectrum of medicines used by older adults, only the medicines used by the 188 study patients. Furthermore, the retrospective patient and medication data from the patient records may perform as a source of bias. The data related to the nursing home residents were not documented in such detail compared to that of hospital patients. The differences in the detailedness of documentation concern e.g., information on cognitive impairment or changes made to medications of individual patients. Documentation may have missed minor falls which were not reported to the staff by the patients.

Conclusion

It was possible to develop a preliminary categorization of FRIDs basing on their adverse drug effect profile in SmPCs and frequency of use in older patients who had experienced at least one documented fall in a geriatric care unit. Even though more than half of the fall-experienced older adults had at least high fall risk scores on admission according to the fall risk items routinely assessed, their fall risk might have been underestimated as use of high fall risk medicaments was common, even concomitant use. Further studies are needed to develop the FRID categorization and assess its impact on fall risk.