Galvez-Barron C.,Fundacio Privada Sant Antoni Abat Research Unit |
Sanguino M.J.,Hospital Sant Antoni Abat |
Narvaiza L.,Hospital Sant Antoni Abat |
Cavestany F.,Hospital Sant Antoni Abat |
And 4 more authors.
Aging Clinical and Experimental Research | Year: 2013
Objective: Given the impact of recurrent falls in older people, risk evaluation for falling is an important part of geriatric assessment. Available clinical tools usually do not include patients' self-perceived risk of falling. The objective of this study was to evaluate association with and predictive capacity of self-perceived risk of falling in recurrent falls. Methods: Prospective cohort study. Patients attending a geriatric outpatients' clinic were recruited (Pfeiffer score <5). A baseline assessment and follow-up over 14 months was scheduled for each patient. Self-perceived risk of falling was assessed through four questions. Association with falls was evaluated through relative risk, survival curves (Kaplan-Meier), and Cox regression (recurrent falls as outcome variables). Predictive capacity was evaluated through sensitivity, specificity, and predictive values. Results: 52 participants answered all 4 questions, and 15 participants (27.2 %) presented recurrent falls. Question 1 (Do you think you may fall in the next few months?) was associated with the occurrence of recurrent falls according to relative risk [3.88 (CI95 %:1.48-10.09)] and survival curves (log Rank, p 0.004). Such relationship is maintained over time. Cox-regression also showed significant difference in relation to the answer for question 1 and recurrent falls [hazard ratio 4.044 (CI: 1.410-11.597); p 0.009]. Sensitivity, specificity, positive and negative predictive values (question 1) were 53.3 % (CI95 %:28.1-78.6), 84.2 % (CI95 %:72.6-95.8), 57.1 % (CI95 %:31.2-83.1), and 82.1 % (CI95 %:70.0-94.1), respectively. Conclusions: Patients' self-perceived risk of falling is related to recurrent falls amongst people with a high risk of falling and this parameter might be useful in falling risk evaluation. © 2013 Springer International Publishing Switzerland.