Risk assessment and incidence of falls in adult hospitalized patients 1

Abstract Objectives: assess the risk of falls in adult hospitalized patients and verify the incidence of the event in this environment. Method: cohort study, with approval by the Research Ethics Committee, which monitored 831 patients hospitalized at a university hospital. The Morse Fall Scale (MFS) was used to assess the risk and patients with high risk (≥45 points) were considered exposed to falls. Results: the mean MFS score was 39.4 (±19.4) points. Between the first and the final assessment, the score increased by 4.6%. The first assessment score presented a strong and positive correlation with the final assessment score (r=0.810; p=0.000). Conclusion: the higher the risk score for falls when the patient is admitted, the higher the score at the end of the hospitalization period and vice-versa. The incidence rate corresponded to 1.68% with a higher percentage of patients classified at high risk of falls.

The incidence rate (IR) was calculated as the ratio between the number of new cases of falls and the total of person-time produced between on the total number of patients monitored, according to the equation (12) : PT, refers to the interval between the baseline t 0 and moment t; I represents the number of new cases that emerged between t 0 and t; and PT represents the quantity of person-time the population accumulated during the study.
The decrease in the muscle strength can be a factor predisposing to falls and is not included in the MFS. Thus, the test by Rossi and Mistrorigo (11) , scored from zero to five, was used to assess the muscle strength in each upper and lower limb. The higher the score, the greater the patient's muscle strength. For the analyses, the limb assessment was grouped in upper and lower limbs and the score was dichotomized into reduced (0 to 4 points) and preserved (5 points). 44 points are at moderate risk of falls; and patients with 45 points or more are at high risk of falls (10) . Patients classified as high risk were considered exposed to falls (MFS score of 45 or higher). Low and moderate-risk patients (MFS between 0 and 44) were considered not exposed to the event.

Results
Among the 864 patients hospitalized at the investigated services between March 11 th and July 11 th 2013 who complied with the inclusion criteria, 831 were monitored daily to assess the risk and occurrence of falls. The losses (N=33; 3.8%) were due to refusals to participate. The Variation Coefficient of the MFS is similar to its standard deviation, but related to the patient's average score. Hence, one may say that, on average, the same patient's score during the period assessed varied by 18.5%. It is highlighted that, in total, 337 patients presented zero variation in the MFS score during the period (one assessment day or MFS score equal to zero).
In Table 2, the patients' distribution according to the MFS items is described. Rev. Latino-Am. Enfermagem 2017;25:e2862.
In Table 3, the patients' risk classifications for falls on the first and final assessment day and the mean classification are described, according to the MFS score. Between the first and final assessments, the MFS score increased by 4.6%. The score on the first assessment revealed a strong and positive correlation with the score on the final assessment (r=0.810; p=0.000), that is, the higher the risk score for falls when the patient was admitted, the higher the score at the end of the hospitalization period and vice-versa.
During the 122 days of monitoring, among the 831 patients assessed, 19 dropped to the floor. That implies an average 4.7 falls per month. The fall incidence rate per person/day in the total group of 6400 patients/ day corresponded to 1.68% (95%CI; 1.51 -1.72%).
As regards the accumulated frequency, which directly estimates the probability/risk that an individual develops the outcome during a specific time period, was equal to 2.28 (95%CI: 1.66 -2.91).
In Table 4, the absolute and relative frequencies When comparing the MFS scores with the presence and absence of falls between the groups (with and without falls), a higher MFS score was detected across the assessment period in the group with falls ( Figure 1).
When the scores were compared intragroup, it was observed that, among the patients who did not present falls, the average scores ranged between 33.8 and 60.0 points on the MFS. In the group with falls, however, the mean scores ranged between 55.0 and 80.0 points, that is, a higher variation when compared to the group without falls.

Discussion
What the length of the assessment and, inherently, of the hospitalization was concerned, most patients were assessed between two and ten times, with an average length of hospitalization of 7.7 days (±9.2). In that sense, in another study, on average, the assessed patients were in hospital for 3.1 days (±2.57) and the length of hospitalization was longer in cases of falls (13) . Hence, the longer the length of hospitalization, the greater the patient's risk of falls (OR=3.2; p<0.01) (13) .
As regards the mean MFS scores, previous studies found averages that differed from this study (39.4 ±19.4 points). In a study that monitored patients similar to the persons assessed in this study, the mean MFS score was 31.7 (±16.9), corresponding to a moderate risk of falls (13) . In another study, a higher mean MFS score was found (57.2), corresponding to a high risk of falls (7) . The latter was developed at a rehabilitation service, where a higher percentage of patients experienced limitations and difficulties, mainly related to walking (7) . Hence, the mean MFS scores and, consequently, the profile of the hospitalized patients will depend on the service offered in hospital.   investigated hearing impairment as a factor predisposing to falls but found no significant result.
What the other findings are concerned, other studies did not evidence a significant difference either for falls related to sex (7,13) and age (7) . Regarding the variable musculoskeletal problem, these study results differed from the findings in other studies (14,18) that found a significant association between high risk for falls and the presence of musculoskeletal disorders. Thus, using this tool to classify the patients and, based on the risk identification, listing prevention strategies, turns into an ally in the nurse's work process and in the promotion of patient safety in the hospital context.
Based on the results, some strategies can be cited that can be included in the care plan: use specific instruments to predict the risk of falls, one of which is the MFS: train the team on the appropriate way to assess the patient and implement the strategies; advise patients/companions on the risk factors that can entail falls; and identify high-risk patients, using a signal at the headrest or a specific wristband, among other strategies (7)(8)22) .
The assessment period is appointed as a limitation, considering that the prevalence of the investigated outcome is low, demanding a larger number of

Conclusion
The largest group of hospitalized patients was classified as at high risk for falls according to the MFS.
The incidence rate of falls corresponded to 1.68% and it was verified a higher percentage of patients who fell were classified in the category high risk for falls. These data signal that the MFS can be used to assess the risk of falls, with a view to identifying factors that contribute to the occurrence of this incident in the hospital context, as the scale assesses different items.
Although low, the incidence rate of falls detected in this study appoints the need to sensitize the health professionals to the occurrence of these incidents in hospitals. Being closer to the patient, the nursing team is an important ally in the prevention of falls.
This proximity permits the early identification of risk situations and favors the nurse's planning of actions, in cooperation with the multidisciplinary team, with a view to reducing the falls rate, which interferes in the continuity of care and in patient safety.