Uncertainty with Long-term Predictions of Lower-Limb Amputation Prevalence and What This Means for Prosthetic and Orthotic Research

Being able to accurately estimate the number of people living with limb loss and predict how this might change over time provides valuable information to help institutions plan for the future needs of people living with limb loss. Similarly, if the prevalence of limb loss is expected to increase over time, institutions may invest in research and health care initiatives to curb the incidence rate and therefore reduce prevalence into the future. Unfortunately, it is difficult to be certain how the number of people living with limb loss may change into the future given that a recent systematic review1 found just one study that estimated the prevalence of lower-limb amputation.2 In this one study by Ziegler-Graham et al.,2 it was estimated that, in 2005, there were 1.6 million Americans living with lower-limb amputation and that this number would more than double by 2050. It is important to recognize that these estimates were based on incidence rate data collected from national hospital discharges from 1988 to 1999.2 The prediction assumed that the incidence rate would continue to increase over time and historical trends in population growth and mortality would also hold true into the future.2 Despite these assumptions, we can have confidence in the predicted prevalence for the year 2005 given the proximity to the data collection period, where uncertainty in these assumptions is unlikely to affect the prevalence estimate. However, when prevalence is predicted further into the future, well beyond the data collection period, there is less certainty that these assumptions will hold true. In the decade since Ziegler-Graham et al.2 published this influential work, a number of studies with more recent time series suggest that the incidence rate of lower-limb amputation has not increased over time as was assumed.3–9 Notwithstanding the various ways these incidence rate data were measured, standardized, and stratified, the incidence rate of lower-limb amputation seems to have remained stable or declined.3–9 Given this understanding, we are concerned that the long-term predictions of limb loss prevalence reported by Ziegler-Graham et al.2 are often used to justify research without regard for the uncertainty associated with such long-term predictions.10–13 For example, Rosenblatt and Ehrhardt10 state, “The number of persons with lower-limb amputation is expected to double over the coming decades…”. We contend there are many similar examples given that a forward citation search conducted on September 23, 2017, using Scopus, shows that this work by Ziegler-Graham et al.2 has been cited 583 times, and that publications in the Journal of Prosthetics and Orthotics are the most common sources of the citation. While the long-term projected increase in amputation prevalence reported by Ziegler-Graham et al.2 makes a compelling argument to justify research, we need to be cautious and thoughtful in our characterization of amputation prevalence to be confident that the research investment and policy decisions we make today are targeted appropriately to the needs of tomorrow. Despite uncertainty in the long-term prevalence estimates, there is a compelling need for research into the prevention of limb loss and how we can improve the outcomes for those living with amputation. Amputation is a life-changing event. Outcomes are often poor. Experiences of depression and anxiety are common. Quality of life is often diminished as is the ability to participate fully in activities that bring meaning to life. These are all compelling reasons to continue to pursue high-quality research, regardless of any uncertainty in the long-term prevalence of limb loss. We hope we have raised awareness of the uncertainty inherent in estimates of long-term amputation prevalence and encourage thoughtful use of these data in the professional literature.