Falls after stroke are common. Previous research has shown that almost every other, 48%, of patients with a first-ever stroke fell at least once during the first year after stroke (1). Moreover, most of the patients fell during the first three months after stroke onset. However, the exact time-point for the falls was not analyzed (1). In addition, many patients are falling already during hospital stay (2-5). Reducing the number of falls during inpatient rehabilitation on a stroke unit is a high-priority mission. Greater knowledge concerning the factors that best predict falls is eligible.

The aim of the study is to analyze the significance of different physiological, psychological and medical factors for predicting the risk of falling in patients with stroke per day of care at a stroke unit.

Sample size assessment: According statistical consultation it is required to include at least 150 patients. The target figure is to include 500 patients. That number enables sub analyzes.

Inclusion criterion: All patients with clinical stroke (infarction as bleeding, first-time or recurrent) at the stroke unit 354 will be asked for study participation.

Exclusion criterion: Patients not willing to participate in the study.

The response variable is fall. A fall is defined as an event which results in a person coming to rest inadvertently on the ground or floor or other lower level. The definition of a fall refers to the World Health Organization (http://www.who.int/mediacentre/factsheets/fs344/en/ 2014-09-29). Data regarding falls will be collected from MedControl Pro (without any official possible expansions). MedControl Pro is a web based program for deviations. For validation of data concerning falls, assessment of medical journals will also be implemented.

The significance of the following covariates will be used in the analyses:

1. The patient's confidence in their own balance (postural control).

   - Do you feel that your balance (postural control) is impaired?

2. Fear of falling

   - Are you afraid to fall? (Fear of falling)

3. Postural control in everyday activities

   • The modified version of the Postural Assessment Scale for Stroke Patients (SwePASS)

4. Postural control; tandem stance (A decisive moment?)

   - Do the patient manages tandem stance 30 s or not?/use of scale 0-4

5. Risk taking behaviour

   • Is the patient inclined to take a risk in tandem stance?

6. Self reported previous physical activity (PA) level.

   - Grimby & Saltin scale and/or 150 minutes of PA/week or not

7. Cognitive ability

   - Montreal Cognitive Assessment (MoCa)

8. Medication

   - Registration of drugs and their Anatomical Therapeutic Chemical (ATC) classification codes

9. Blood pressure

   - Blood pressure and pulse lying and after 1 and 3 minutes standing

10. The use of walking aids

11. The use of wheelchair

12. Age

13. Sex

14. Length of stay (in days) on the stroke unit.

In addition, the National Institutes of Health Stroke Scale (NIHSS) result will be registered and used to describe the study population.

Data collection:

Data collection will take place once, as fast as possible but at the latest 4 days after admission to the stroke unit. There are two exceptions:

First, data collection regarding drugs will take place day 4 after admission. However, in case of a fall before day 4 after admission, the data collection will be based on the drug list the current fall day.

Second, data concerning cognition using MoCA will be performed once sometime during the stroke-unit stay (due to organisation issues). That is, not at any particular time, but the greater the suspicion of cognitive impairment, the later during the stroke unit stay.

Statistical analysis:

The poisson regression analysis will be used in order to address number of falls (fall risk) per day and night at the stroke unit. Poisson regression takes into account the wide variation in length of stay at the stroke unit.

Inclusion Criteria:

• Patients with clinical stroke, whether first-ever or recurrent stroke and whether caused by cerebral infarction or cerebral hemorrhage
• Patients 18 years old or older

Exclusion Criterion:

• Patients unwilling to participate in the study.

The group will consist of consecutively included patients with clinical stroke at the stroke unit at Sahlgrenska University Hospital/Östra, Gothenburg.