Answers are grounded in this paper and cite their sources.
To the Editor: Participants of clinical trials are often asked to recall the number of falls in the previous 12 months, yet the accuracy of this 12-month recall in older people is often questioned. There is little information from studies performing a head-to-head comparison of different assessment strategies to determine the likely accuracy of the data collected. The current analysis is a head-to-head comparison of falls ascertainment comparing monthly calendars with a 12-month recall question in 2,096 community-dwelling women at high risk of falls and fracture. All participants in the Vital D study were aged 70 and older at recruitment (2003–2005) and prospectively completed a monthly ascertainment of falls for the next 3 to 5 years. In 2007, all participants completed a 12-month falls recall question that categorized falls into no falls, a few times, several, and regular falls. During the 12-month interval, 42.8% (898/2,096) of participants experienced a fall according to their monthly calendars, of whom 77.1% (692/898) correctly recalled the fall(s). The 206 participants who failed to recall their fall were older (P=.03). Of participants who selected no fall, 84.6% were correct. Few women selected the higher fall categories of several-times or regular faller (4.1% and 0.4%, respectively), and the sensitivity of these categories was low (30.4–33.3%). Simply categorizing participants into fallers or nonfallers had 77.1% sensitivity and 94.2% specificity. The sensitivity (77.1%) was similar to the 6-month recall (74%) but lower than the 12-month recall (87%) reported previously.1 The specificity of both studies was similar (current study 94.2%; previous study1 93%). The head-to-head comparison of ascertainment methods shows that, over the four categories of recall, 82.4% (1,727/2,096) of participants correctly categorized their fall history on 12-month recall, but nonfallers and those who fell only a few times were more likely to recall correctly (94.2% correct for nonfallers and 68.0% correct for a few times fallers). Few women fell more than four times in 1 year (1.5%; 32/2,096), and of those, half did not report having several or regular falls (16/32) (Table 1). Although there were 70 nonfallers (5.8%) who incorrectly reported falling on the 12-month recall, the general classification of fallers or nonfallers has a high sensitivity (77.1%) and specificity (94.2%). Because participants in the Vital D study have been completing the monthly falls calendar for several years, it was possible to ascertain that only 35.7% of the 70 women who incorrectly recalled a fall had sustained a fall in the 3-months before the specific 12-month recall interval. The head-to-head comparison of the falls data poses an unavoidable limitation of the study, because the participants recalling their falls over the previous 12 months had also been posting a monthly record of their falls over the same period. The 22.9% of participants who did not recall their fall(s) is likely to be an underestimate, because the monthly calendar ascertainment would improve their recall of falls. Even with the most rigorous reporting methodology, it is likely that falls are underreported.2 It has been noted that denial can be a factor in underreporting, because some older people take pride in being nonfallers and “don't want to blot their copybook” by reporting a fall.3 Older people can blame external factors for their fall and not count it as a “true” fall. Few studies would have sufficient power to detect differences between several-time fallers and regular fallers. The proportion of nonfaller, one-time faller, or more than one fall was 57.2%, 25.7%, and 17.1%. Asking older women to select from fewer categories provides a reasonable alternative for studies in which intensive ascertainment of falls is not feasible. With the aging of most Western populations, the consequences of injurious falls and their effect on quality of life and the economic burden to the health system continues to grow. It is hoped that, by reporting this head-to-head comparison of prospective ascertainment and 12-month recall, researchers can make informed choices in designing studies that incorporate some falls risk data but in which a more-intensive ascertainment of fall events is not feasible. Conflict of Interest: The editor in chief has reviewed the conflict of interest checklist provided by the authors and has determined that the authors have no financial or any other kind of personal conflicts with this letter. Author Contributions: KMS, MAK, and GCN are Chief Investigators of the Vital D study (ISRCTN83409867), and all three contributed to writing the letter. KMS is also the project manager, and ALH is the study coordinator, performed some of the statistical analyses, and reviewed the draft manuscripts. Sponsor's Role: This research is wholly funded through government funding agencies of medical research, and the manuscript is without bias in the presentation of results.
No takes yet. Share an insight, caveat, or question.
Sanders et al. (2009) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: