Partial-Foot Amputation and Its Unexpected Evidence

Home > Articles > Partial-Foot Amputation and Its Unexpected Evidence
By Phil Stevens, MEd, CPO, FAAOP

Research takes a long time. The process requires a research team to design an experiment, secure funding, recruit study subjects, analyze the data, and push its findings through the time-consuming process of peer review before the published study finally becomes available. Often, the arrival of this final product occurs well after those in clinical practice have already reached similar conclusions based on collective, anecdotal observations. This reality can cause clinicians to discount the value of the published literature as something that merely confirms what we already know. However, a profession historically based on assumptions is bound to encounter a few surprises. Nowhere has this been more evident than in the management of partial-foot amputation.

This article presents several recent publications that may challenge some of the conventional assumptions regarding the treatment of this population.

The Big Picture

Two of our profession's more experienced researchers, Michael Dillon, PhD, BPO (Hons), and Stefania Fatone, PhD, BPO (Hons), recently collaborated on a special communication article that serves as a good starting point for this conversation.1 The authors opened their discussion by reviewing some rather striking statistics on the complication rates reported with the various levels of partial-foot amputation that may surprise many clinicians (Table 1). The authors observed that as many as 50 percent of all individuals who undergo a partial-foot amputation will experience complications such as skin breakdown, ulceration, or wound failures. Further, while the healing rates following transtibial amputation are consistently over 75 percent, only about 50 percent of partial-foot amputations will ultimately heal. When that healing doesn't occur, re-amputation is often indicated. These revision amputations are performed in about one-third of all patients who undergo an initial partial-foot amputation, an incidence nearly twice that reported for an initial transtibial amputation.1 Yet despite these less than encouraging observations from the literature, at a time when amputation proximal to the ankle is becoming less common in many parts of the world, the incidence of partial-foot amputation is rapidly increasing.1

Table 1: Complication rates observed with partial-foot amputation, as summarized by Dillon and Fatone.1

Given that transtibial amputation is an alternate surgical option in all partial-foot amputation cases, the disparities between these rather concerning outcomes and the increasing incidence of partial-foot amputation appears to be based on faulty assumptions among the medical community. Dillon and Fatone addressed three of these apparent assumptions with published evidence.1

Assumption 1:

By preserving the anatomic ankle joint and vertical limb length, partial-foot amputation permits push-off and a more normal post-amputation gait. To the extent that it has been studied, the evidence suggests that the ability to generate a concentric plantarflexion contraction is only maintained when the partial-foot amputation is distal to the metatarsophalangeal (MTP) joints. Partial-foot amputations proximal to that level, whether transmetatarsal, Lisfranc, or Chopart, result in negligible power generation at the ankle, irrespective of what orthotic or prosthetic device may be worn.1 Given that the power generation associated with partial-foot amputation is comparable to that observed following transtibial amputation, it is hardly surprising to note that both groups appear to adopt similar compensatory strategies at the hip joint.1 In addition, the decreases in self-selected walking speeds appear to be very similar in both groups, irrespective of the length of the partial-foot amputation, when such groups are matched to control groups for the influence of confounding systemic disease considerations.1 Given the similarities observed following partial-foot amputations and transtibial amputations with respect to the absent push-off compensatory actions at the hip and the reduced walking velocity, the assumption that partial-foot amputation preserves a more normal gait post-amputation can be reasonably called into question.

Assumption 2:

Gait following partial-foot amputation is less energy expensive than that observed following transtibial amputation. This assumption appears to be an unfounded extension of the more thoroughly investigated phenomenon at more proximal amputation levels.1 A progressive decrease in the energy costs of ambulation has been consistently observed with the progressively distal transfemoral, transtibial, and Symes amputation levels. However, the extrapolation that the extended residual-limb length of partial-foot amputation will yield still greater energy savings does not appear to be warranted. To the limited extent that such studies have been undertaken, they have failed to show a reduction in energy costs.1 These observations are bolstered by the theoretical arguments that, unlike their peers with transfemoral, transtibial, or Symes amputations, where distinctly different biomechanic gait strategies are observed, there are no such differences between subjects with a partial-foot amputation and a transtibial amputation.1 Also, as observed before, the relative similarities in self-selected walking speeds, which are often decreased post-amputation to accommodate new increases in energy costs, appear to be similar following partial-foot amputation and transtibial amputation.1

Assumption 3:

Partial-foot amputation is associated with an improved quality of life (QOL) compared to transtibial amputation. Unfortunately, to the extent that QOL following partial-foot amputation has been investigated, these studies have failed to discriminate between the various levels of partial-foot amputation. Thus, it is difficult to confirm whether the QOL associated with the relatively distal MTP amputation level is any different from that experienced following Chopart amputation. To the extent that the aggregate observations of QOL following partial-foot amputation have been reported, they have been found to be comparable to those observed following transtibial amputation.1 The compromised QOL reported following partial-foot amputation may be more easily accepted and understood by considering the examples suggested by Dillon and Fatone in their article. While there is a certain advantage to being able to get up in the middle of the night and go to the restroom without donning a prosthesis, this must be counterbalanced by the compromised ability of such patients to participate in strenuous recreational activities like running and other sports.1

Given the discouraging outcomes collectively observed following partial-foot amputation with respect to such considerations as complications, healing, and secondary amputation, it is important for the decision makers in such cases to understand the viability of the assumptions upon which they will ultimately base their choice. To the extent that these assumptions appear to be flawed, the medical community will need to adjust its rationales and expectations regarding partial-foot amputation.

What about Balance?

The shortcomings of partial-foot amputation appear to extend beyond those cited previously. It's intuitive to suppose that if the residual-limb length is long enough to preserve contact with the ground, patients with a partial-foot amputation might demonstrate improved balance compared to those with a transtibial amputation. However, the observations of Kanade et al. appear to suggest that if the amputation is secondary to diabetes, that difference in limb length may not improve balance after all.2

In their investigation, the authors set out to compare the standing balance of four different patient cohorts, all with diabetes as their primary diagnosis. This included controls with no lower-limb compromise (n=23), patients with an existing foot ulceration (EFU) (n=23), patients with a partial-foot amputation (n=16), and patients with a transtibial amputation (n=22). All tested subjects were living independently in the community and walked independently to perform their activities of daily living. Patients with obvious confounders such as stroke, recent fractures, acute cardiovascular symptoms, or drug and alcohol dependence were excluded. All subjects were reasonably matched with regard to their average age, height, body mass, and body mass index. The average duration of diabetes for those in the EFU, partial-foot amputation, and transtibial amputation cohorts was twice that of the control group.2

Computerized dynamic posturography. Photograph courtesy of Natus Medical.

Standing balance was measured using posturography, in which the subjects stood on a force plate allowing for a measurement of the relative movements of the center of pressure (COP) during quiet standing. The relative excursions of the COP in the sagittal and coronal planes could then be reported, along with a total excursion value, reflecting the total distance traveled by the COP in any direction during the 30-second standing trial. Increases in COP migration are generally associated with decreased postural stability.

The authors found that all four groups with diabetes produced total COP excursion values that were higher than those observed in healthy controls. While legacy data suggested a mean total excursion value among healthy controls of 0.55m, among the diabetic cohort with no current lower-limb compromise, this mean value rose just over 50 percent to 0.86m. Among the three remaining cohorts, it rose another 40 percent to about 1.2m. While the differences between the three diabetic cohorts with existing lower-limb compromise were minimal, the highest average COP excursion values were observed among the partial-foot amputation cohort, followed by the transtibial amputation and EFU groups.2

The term partial-foot amputation encompasses a range of limb deficiencies. In this last study, the group was composed of five subjects with transmetatarsal amputations, four subjects with ray amputations, five subjects with hallux amputations, and two subjects with amputations of two or more digits. The message is clear. Even with more distal levels of partial-foot amputation, subjects with diabetes experience as much or more balance compromise as that seen among patients with diabetes who have undergone a transtibial amputation.

Can Balance Be Improved?

In the Kanade et al. study, patients were tested with an "appropriate shoe-filler." Would their balance measures have improved with a more proximal device? Preliminary data from a small case series conducted in the Pacific Northwest suggest that such improvements might be unlikely.3 In this investigation, a smaller cohort of six subjects with transmetatarsal amputation secondary to diabetic neuropathy were evaluated. Each subject experienced two conditions. In the first, subjects were fitted with a custom, total-contact insole and a shoe modified with a long, rigid shank. In the second, an Allard BlueRocker™ AFO was placed in the shoe, with the custom insole placed over the plantar plate of the AFO.3

The subjects ambulated across a ten-meter walkway, variously stepping over obstacles that were 2.5 percent or 10 percent of their body height, simulating a door threshold or curb. In this study, COP data was collected and constantly compared against the center of mass (COM) positions to create instantaneous COM-COP inclination angles (defined as the angle formed between a vertical line passing through the COP and the line connecting the COM to the COP). Balance compromise has been associated with larger medial COM-COP inclination angles.

The authors observed that while the addition of the BlueRocker led to a nonsignificant increase in the cohorts' mean gait velocity, for five of the six subjects, the addition of the Blue Rocker failed to alter the medial COM-COP angles appreciably. Interestingly, the only subject who experienced a substantial improvement in this balance parameter with the more restrictive intervention was also the most debilitated subject, with the longest standing duration of his diabetes and highest Timed Up and Go (TUG) score.3

Taken collectively, these studies highlight the substantial balance compromise associated with diabetes-related and comorbid partial-foot amputation. Further, they suggest that for many patients, variations in the orthotic intervention may not affect these parameters.

Affecting the Center of Pressure

In addition to informing upon balance considerations, monitoring the COP during partial-foot amputation ambulation can provide other information with respect to such considerations as weight shifting during forward progression and the loads placed on the often at-risk distal end of the residual limb following partial-foot amputation.

In what should now be seen as a classic study in the growing body of partial-foot amputation research literature, Dillon and Barker reported on the inability of traditional partial-foot amputation solutions (toe-fillers and slipper socks) to restore functional foot length during ambulation. In their investigation, the sagittal progressions of the COPs of patients with MTP-, transmetatarsal-, Lisfranc-, and Chopart-level partial-foot amputations were monitored relative to the residual-limb lengths. The authors observed that in all of those subjects with longer residual limbs and filler-style prostheses, the COP stayed posterior to the end of the residual limb until the contralateral limb had accepted axial loads. In other words, there was no functional restoration of foot length.

By contrast, in the patients with a Chopart-level partial-foot amputation whose devices were characterized by a stiff forefoot, restricted dorsiflexion, and a proximal build height that allowed the forces across the toe lever to be transferred to the anterior aspect of the shank, the COP was observed to travel well anterior to the distal end of the residual limb prior to the contralateral heel strike.4 Thus, if the objective of the prosthesis is to maintain a smooth advancement of the COP following partial-foot amputation, the same design criteria identified in the Chopart-level devices would appear to be indicated.

In their analysis of why those subjects with more distal partial-foot amputations and minimalist prostheses adopted their characteristic gait pattern, Dillon and Barker suggested that it may have been due, in part, to an attempt to "spare the sensitive distal stump from the extreme forces typically observed during terminal stance by not allowing the COP to progress across the distal end until after contralateral heel contact when the magnitude of the ground reaction force was rapidly diminishing."4 This concept finds further support in data from the insole/BlueRocker study described earlier.3 In addition to monitoring the COP of the entire body as measured through the force plates that made up the flooring of the research lab, the authors used an in-shoe Tekscan® F-Scan® sensor system to monitor the plantar pressures between the residual foot and the custom insole in three of their subjects. When these patients walked in the insole/shoe combination, the plantar pressure was observed to travel an average of about 7cm, or most of the length of the residual foot. When the same subjects walked with the addition of the BlueRocker AFO, the travel distance of the COP was reduced by about 2.5cm, which spared the distal end of the residual limb from load bearing pressures. This reduction of the axial pressure against the distal plantar surface of the foot may have contributed to the modest increases in both gait velocity and step length observed in the BlueRocker condition.3


The prosthetic management of partial-foot amputation represents one area in the O&P literature where many of our profession's collective assumptions appear to be wrong. At the broadest level, considerations regarding the surgical viability and functional benefits of the partial-foot amputation may need to be more carefully considered. With respect to the prosthetic management of partial-foot amputation, the ability of certain design types to affect such considerations as balance, efficiency in gait, and limb loading patterns are beginning to be better understood. Viewed collectively, the recent literature regarding partial-foot amputation continues to expand the profession's understanding of this unique patient population.

Phil Stevens, MEd, CPO, FAAOP, is in clinical practice with Hanger Clinic, Salt Lake City, Utah. He can be reached at


  1. Dillon, M. P., and S. Fatone. 2013. Deliberations about the functional benefits and complications of partial foot amputation: Do we pay heed to the purported benefits at the expense of minimizing complications? Archives of Physical Medicine and Rehabilitation 94 (8):1429-35.
  2. Kanade, R. V., R. W. Van Deursen, K. G. Harding, and P. E. Price. 2008. Investigation of standing balance in patients with diabetic neuropathy at different stages of foot complications. Clinical Biomechanics 23 (9):1183-91.
  3. Spaulding, S. E., T. Chen, and L. S. Chou. 2012. Selection of an above or below-ankle orthosis for individuals with neuropathic partial foot amputation: A pilot study. Prosthetics and Orthotics International 36 (2):217-24.
  4. Dillon, M. P., and T. M. Barker. 2006. Can partial foot prostheses effectively restore foot length? Prosthetics and Orthotics International 30 (1):17-23.