When Lightning Strikes Twice: Comorbid Amputation and Hemiplegia
December 2013 Issue
In the introduction to their recent systematic review of patients with both hemiplegia and lower-limb amputation, Hebert et al. identify the shared risk factors between the two events.1 The population of individuals with lower-limb amputations is largely characterized by older adults with diabetes and peripheral neuropathy. The underlying risk factors for amputation, including high cholesterol, high blood pressure, poorly controlled diabetes, and smoking, are also common risk factors for other vascular diseases, such as cardiac disease, peripheral neuropathy, and cerebrovascular disease. As a common manifestation of a cerebrovascular accident (CVA), hemiplegia is strongly associated with these same risk factors. While either event may precede the other, their shared risk factors create the potential for the comorbid existence of lower-limb amputation and hemiplegia.
Given that both events can be quite devastating to an individual's functional abilities, their comorbid existence would appear to present a substantial rehabilitative challenge. This article investigates the prevalence of hemiplegia with lower-limb amputation, its effect on prosthetic outcomes, and any predictive characteristics that might be used to moderate goals and expectations during rehabilitation.
The prevalence for this comorbid presentation has been cited at between 8.5 and 18 percent (Table 1). Interestingly, the highest prevalence rate was identified in the largest study cohort in a recent paper investigating the prognosis of U.S. veterans with some sort of preexisting neurological disorder and subsequent lower-limb amputation.2 Extracting their data from the Veterans Health Administration databases, Prvu-Bettger et al. identified 4,720 veterans who underwent transtibial, transfemoral, or hip disarticulation amputation between 2002 and 2004. They then extracted the International Classification of Diseases (ICD)-9 codes from their outpatient files three months prior to the hospital admission associated with their lower-limb amputation. Diagnoses were then divided into either etiologies or comorbidities. The former represents those likely to have directly contributed to the amputation, such as congenital deformity, chronic osteomyelitis, type I and II diabetes mellitus, peripheral circulation problems, and skin breakdown. Comorbidities were those diagnoses perceived as less likely to be related to lower-limb amputation. These were then broken down into neurological and non-neurological factors such as cancer, congestive heart failure, and hypertension. This first classification was the focus of the research paper and included such categories as stroke or hemiparesis, cerebral degenerative disease (such as Alzheimer's disease), movement disorders (such as Parkinson's disease), spinal cord injury (SCI) or paralysis, and peripheral nervous system disorders (such as peripheral neuropathy).
Forty-three percent of the study cohort were found to have at least one comorbid neurological diagnosis. Peripheral nervous system disorders were the most common, identified in 20 percent of the total cohort, followed closely by stroke or hemiplegia, occurring in 18 percent of the total cohort. The prevalence rates of the other neurologic comorbidities were much lower, including SCI/paralysis at 4 percent, and both cerebral degenerative disease and movement disorders at 2 percent.2 While peripheral nervous system disorders had the highest prevalence, they also had the second lowest hazard ratio of the neurological comorbidities, meaning these subjects generally had higher one-year survival rates. By contrast, the hazard ratio associated with stroke and hemiplegia was one of the highest, with a 30 percent mortality rate one-year post lower-limb amputation.2
In addition to prevalence, several studies have reported observations on the demographics of the laterality of the comorbidities, that is, whether the hemiplegia and amputation occurred on the same side and the sequence in which the two occurred. With respect to laterality, Herbert et al.'s systematic review observes that amputation and hemiplegia occur most often in the same leg, with a frequency ranging from 56-95 percent (Table 2).1 Of the seven studies reporting on laterality, only one reported a majority of cases in which the lower-limb amputation was contralateral to the hemiplegia.
The laterality of these comorbidities appears to be important because the four studies from Table 2 that examined the effects of laterality on ambulatory function found improved functional abilities among those with ipsilateral comorbidities compared to those with contralateral comorbidities. For example, in their review of the 23 patients with comorbid hemiplegia and lower-limb amputation seen between 1984-1994 at the rehabilitation department in which they worked, Chiu et al. observed that eight of the 16 patients with ipsilateral comorbidities returned to community ambulation.3 By contrast, only two of the seven patients (29 percent) with contralateral comorbidities reached this level of function.3 The apparent impact of contralateral comorbidities on the number of comorbid subjects able to return to community ambulation was similar to the apparent disadvantages experienced with a transfemoral amputation (25 percent), severe hemiplegia (27 percent), and being 60 years old or older (33 percent).3
In a similar study, Brunelli et al. reviewed the charts of the 56 patients with both hemiplegia and transfemoral amputation seen for inpatient service, over a five year period, at the hospital in which they worked.4 These subjects' ability to accomplish their activities of daily living and their ambulatory skills were assessed using the Barthel Index (BI) and Locomotor Capabilities Index (LCI) respectively. Upon discharge, the BI scores of the patients with ipsilateral comorbidities were 17 percent higher than they were for the patients with contralateral comorbidities. Similarly, LCI scores for those with ipsilateral comorbidities (18/42) were 50 percent higher than the LCI scores seen among those with contralateral comorbidities (12/42).4
The data on sequencing was less consistent. Three of the six studies reporting upon this demographic observed that hemiplegia preceded lower-limb amputation the majority of the time. Two additional studies found a balanced likelihood of either comorbidity preceding the other. A final study reported lower-limb amputation occurring first in the majority of their subjects. Therefore, no consistent sequencing trends have been observed (Table 3).1
At this time there is little data to suggest that the sequencing of the comorbidities matters. In their review, Chiu et al. observed a greater likelihood of returning to community ambulation among subjects whose lower-limb amputation preceded hemiplegia (67 percent) compared to those whose hemiplegia preceded lower-limb amputation (33 percent). However, these trends must be viewed cautiously as the former group only consisted of six subjects.3 By contrast, Brunelli et al. failed to observe any substantive differences in either BI or LCI scores between the amputation-first (n=26) and hemiplegia-first (n=19) cohorts.4
While the sequencing of the two comorbidities does not appear to predict function, there are several indications suggesting improved outcomes when the two comorbidities occur sequentially rather than simultaneously. Citing observations from some of the older trials on these populations, Herbert et al. summarize:
- "A longer time period between each subsequent disability may predict better outcomes."
- "Amputees who had been fitted with their prosthesis prior to having a stroke had better ambulatory function."
- "Patients who were ambulating prior to the second disability were more likely to achieve independent mobility than those who failed to ambulate after the first disability."
Other Predictive Factors
In addition to laterality, there were several other presentation characteristics that appeared to influence rehabilitation outcomes. Four separate studies have observed improved outcomes among individuals with transtibial amputations and hemiplegia compared to their peers with transfemoral amputations.1 For example, Chiu et al. reported that while 53 percent of the 15 subjects with transtibial amputation regained community ambulation, only 25 percent of their peers with transfemoral amputation regained similar mobility.3
Additionally, several studies observed that patients with less severe hemiplegia also experienced greater gains in their rehabilitation than those with more severe hemiplegia.1 From the Chiu et al. study, 58 percent of the 12 subjects with mild hemiplegia regained community ambulation compared to only 27 percent of the 11 subjects with more severe hemiplegia (Table 4).3 Similarly, among their cohort of subjects with transfemoral amputation, the average discharge BI scores of patients with mild hemiparesis was 26 percent higher than those with moderate hemiparesis. Similarly, the average LCI scores for those with mild hemiparesis (17/42) were higher than for those with more severe hemiplegia (13/42).4
There have also been preliminary indications that mental status and age may act as predictive factors. Within their cohort, Chiu et al. divided their subjects into two groups according to mental status. Those with an "intact" mental status were defined as "alert with normal judgment, orientation, memory, abstract thinking, and calculation with the ability to comprehend and learn." Subjects with "impaired" mental status were described as those who "showed decreased alertness, disorientation, or decreased ability of learning." Of the five subjects with impaired mental status, none achieved community ambulation, and only one was capable of indoor ambulation. The remaining subjects were classified as nonambulators. By contrast, 56 percent of the 18 subjects with intact mental status were capable of community ambulation, and with an additional 22 percent were capable of indoor ambulation.3
The effect of age as a predictor of functional outcomes was less readily apparent. Chiu et al. observed that six of the eight subjects classified as nonambulators were over the age of 65.3 Of their 13 older subjects, only four obtained community ambulation.3 However, while Brunelli et al. reported a 10 percent increase in the average discharge BI score of subjects younger than 65, there was no significant difference in the average LCI scores of the two cohorts.4
Any discussion about the likelihood of these comorbid patients returning to some level of functional ambulation must begin with an appreciation of the underlying sampling bias. The majority of the studies on the topic are derived from the data sets of inpatient rehabilitation departments, implying that a medical professional examined each of the included individuals and determined that they had a possibility of some return to function. There was certainly another subset of patients with amputation and comorbid hemiplegia whose rehabilitation potential was so poor that they would not have been referred to an inpatient rehabilitation setting.
In summarizing the data on the likelihood of patients being successfully fit with a prosthesis, Herbert et al. clarified that most studies did not directly report upon this metric. Instead, it was often derived by considering the number of patients with amputation and hemiplegia reported in the baselines of each study compared to the total number of patients who were reported to have reached an ambulatory level.1 Based on this analysis, the authors observed that five of the seven studies for which fit rate could be inferred had a rate greater than 58 percent.1
For those with amputation and comorbid hemiplegia who were successfully fit with a prosthesis, the ambulatory outcomes were generally quite good. Nonambulatory rates were generally reported at about 20-30 percent with the remainder of the subjects attaining either indoor or community ambulation. This allowed for the following evidence statement: "[O]nce individuals with hemiplegia and lower-limb amputation are selected for prosthetic rehabilitation, the fit rates and potential for achieving some level of functional ambulation are good, with most studies showing greater than a 58 percent success rate."1
Because of the shared risk factors involved, there is a surprisingly high prevalence of patients who experience both hemiplegia and lower-limb amputation. Ipsilateral presentations of these comorbidities are described more frequently than contralateral presentations. This is fortunate, as patients with ipsilateral presentations appear to have improved outcomes. There does not appear to be a dominant sequence of the two events among individuals with both conditions. However, outcomes are improved when the events occur sequentially and allow subjects to experience some level of rehabilitation with one event prior to experiencing the other. Advanced age and having a transfemoral amputation, moderate to severe hemiplegia, and impaired mental status are all predictive of worse ambulatory abilities. However, most comorbid patients with a lower-limb amputation and hemiplegia who are referred for rehabilitation appear to benefit from prosthetic rehabilitation.
Phil Stevens, MEd, CPO, FAAOP, is in clinical practice with Hanger Clinic, Salt Lake City, Utah. He can be reached at
- Herbert, S. J., M. W. Payne, D. L. Wolfe, A. B. Deathe, and M. Devlin. 2012. Comorbidities in amputation: A systematic review of hemiplegia and lower limb amputation. Disability and Rehabilitation 34 (23):1943-9.
- Prvu-Bettger, J.A., B. E. Bates, D. E. Bidelspach, and M. G. Stineman. 2009. Short- and long-term prognosis among veterans with neurological disorders and subsequent lower-extremity amputation. Neuroepidemiology 32 (1):4-10.
- Chiu, C. C., C. E. Chen, T. G. Wang, M. C. Lin, and I. N. Lien. 2000. Influencing factors and ambulation outcomes in patients with dual disabilities of hemiplegia and amputation. Archives of Physical Medicine and Rehabilitation 81 (1):14-17.
- Brunelli, S., T. Averna, P. Porcacchia, S. Paolucci, F. Di Meo, and M. Traballesi. 2006. Functional status and factors influencing the rehabilitation outcome of people affected by above-knee amputation and hemiparesis. Archives of Physical Medicine and Rehabilitation 87 (7):995-1000.