How Accurate Is Partial Weightbearing?

Abstract

The accuracy of partial weightbearing was assessed in six healthy volunteers and 23 patients who had sustained either a fracture of a lower limb or surgery. They were trained to weightbear partially using the conventional bathroom scale method and were assessed in a gait laboratory using force platforms. The amount of weight exerted on the involved limb during three-point crutch walking was determined. Four of six volunteers exerted a mean of 27% of body weight more than required. The remaining two volunteers exerted a mean of 8.5% of body weight less than required. Of the 23 patients, 21 exerted a mean of 35.3% of body weight more than that prescribed and two patients exerted a mean of 11.97% of body weight less than that prescribed. In both groups there was little relationship between the weightbearing prescribed and actual weightbearing. None of the patients or volunteers was able to reproduce the extent of partial weightbearing for which they were trained using the bathroom scale method, confirming that this technique of instructing patients in partial weightbearing is inaccurate.

Partial weightbearing is commonly advised after fracture, arthroplasty, amputation, or in patients with dysvascular limbs. It restricts weightbearing to safe limits therefore avoiding excessive or premature loading that may inhibit or stop healing.^{3,5,6,12–14,20}

Numerous clinical and animal studies have assessed the relationship between mechanical forces and the healing response, in particular fracture healing.^3,8–10,16 Early loading causes axial micromotion which has been shown to stimulate fracture healing^8–10,16,20 and is to be encouraged so long as it does not lead to loss of reduction or hinder healing of associated soft tissue injuries. In experimental fractures, axial amplitudes of 2 mm were shown to delay callus formation, although all fractures healed eventually.⁹ It has been reported that fractures treated in a cast show axial movements of 1–4 mm, whereas movements of 1–2 mm axially occur even in fractures rigidly fixed with external fixators.^2,9 Early loading of experimental fractures stabilized by flexible fixation systems was shown to generate an enormous amount of callus but produced significant delays in healing and the quality of newly formed tissue was reduced compared to fractures with reduced loading.¹ From these results, it seems that reduction of load transfer by delaying full weightbearing is advantageous for healing of fractures stabilized by flexible fixation systems.¹ It has been recommended that fracture mechanics should be controlled rigorously in the first 4–6 weeks to provide physiologic amplitudes of movement to avoid tissue damage that may result from high stresses and strains.^3,8,9,20 The extent of limitation of weightbearing depends on the stability and type of fracture fixation, its inherent fatigue life, and the systemic condition of the patient.^16,20

Although the beneficial role of functional weightbearing in fracture healing is well substantiated, its efficacy in uncemented arthroplasty is questionable.^1,15,21 Partial weightbearing is recommended in the initial period after uncemented arthroplasty because it is thought that early weightbearing would inhibit osseous ingrowth attributable to excessive micromotion at the implant-bone interface.^1,13,14 Animal studies reported that bone ingrowth occurs even with small movement (up to 28 μm) but excess movement (150 μm or more) resulted in attachment by mature connective tissue.^13,14 It is not known whether the latter is clinically significant. Immediate weightbearing after bilateral uncemented arthroplasty was shown to cause more initial subsidence of the femoral prosthesis but did not affect clinical outcome at the end of 2 years.¹⁵ In the absence of deleterious clinical consequences full weightbearing has been recommended after uncomplicated uncemented arthroplasty.^1,15,21 However partial weightbearing has been advised if the components are undersized, in the event of an unexpected intraoperative complication such as a fracture, trochanteric osteotomy, or removal of hardware that might compromise the femur.²¹

Although it seems important for certain patients to weightbear partially to avoid deleterious clinical consequences,^{3,5,13,14,20,21} it is not known if such patients would actually experience harmful consequences. The current study did not address the clinical effect of partial weightbearing but assessed the ability of patients and volunteers to weightbear partially after being trained using the bathroom scale technique which is a commonly used method.^6,7,19 In this method a patient places the involved limb on a bathroom scale and presses down on it until the desired weight limit is reached.^6,7,19 It is thought that on repeating this process the patient would form a mental image of the extent of pressure that needs to be exerted on the limb to achieve the required limit of weightbearing.^7,19 Also, it is thought that patients could reproduce the skill of partial weightbearing during three-point crutch walking.^7,19

MATERIALS AND METHODS

Twenty-three patients and six volunteers were included in this study, which was done during a period of 1 year. Ethical approval was obtained from the authors’ institutions.

The mean age of the volunteers was 30 years (range, 15–45 years). There were four males and two females and none had any visual, sensory, or motor impairment. Of the 23 patients, there were 12 males and 11 females, with a mean age of 41 years (range, 18–65 years). None of the patients had any significant medical or surgical condition apart from the orthopaedic condition for which they were being treated. The indications for weightbearing partially were uncemented hip replacement in 10 patients, nailing for a comminuted fracture of the tibia in seven patients, fixation of comminuted ankle fracture in three patients, comminuted hip fracture (fixed with a dynamic hip screw) in two patients, and revision of an ACL reconstruction in one patient.

A below-knee plaster walking cast was applied to the right leg of the healthy volunteers and they were trained to weightbear partially to an arbitrary weight of 30 kg, which represented 20–40% of the body weight, a range that is prescribed commonly in clinical practice.⁷ In the patient group, the magnitude of weightbearing allowed was prescribed in terms of percentage of body weight and varied in each case depending on the orthopaedic condition for which the patient was being treated.

All volunteers and patients were trained to weightbear partially by a physiotherapist using the bathroom scale technique as described earlier. After training, and the patients had formed a mental image of the prescribed partial weightbearing, they were taught to mobilize with crutches using three-point gait. A few days later, with the help of their crutches the subjects walked on a walkway incorporating Bertec force platforms (Bertec Corporation, Columbus, OH). The force platform calculated the ground reaction force produced between the plantar surface of the foot and the floor. All three components of the ground reaction force were used to calculate the amount of weight transmitted to the platform during partial weightbearing (Fig 1). Readings were taken during the entire stance phase in six separate walks and the mean of all six values was taken into account for each subject.

Fig 1.: A graph shows the ground reaction forces obtained during crutch walking.

Statistical analysis was done using SPSS software (SPSS Inc, Chicago, IL). Descriptive statistics were used to confirm that the data were normally distributed, and the technique of Bland and Altman⁴ was used to describe the distribution of differences between actual partial weightbearing and prescribed partial weightbearing. Spearman’s rank correlation was used to examine the correlation between prescribed weightbearing and actual weightbearing because based on simple plots a monotonic (rather than linear) relationship was assumed between prescribed weightbearing and actual weightbearing.

RESULTS

The extent of weightbearing was expressed as a percentage of the subject’s body weight. Four of six volunteers were weightbearing a mean of 27% of body weight more than required (range, 17.5%–40.5%). The remaining two volunteers exerted a mean of 8.5% of body weight less than required (range, 7.3%–9.6%). Of the 23 patients, 21 exerted a mean of 35.3% of body weight more than that prescribed (range, 3.14%–163.2%). Two patients exerted a mean of 11.97% of body weight less than that prescribed (range, 4.17%–19.8%).

A summary of the differences between actual partial weightbearing and prescribed partial weightbearing is shown in Table 1. The average difference was 37.2% of body weight with a standard deviation of 23.55% of body weight. A standard sample t test was used to assess any bias in the differences in both groups. The mean was statistically significantly different from 0 (t = 7.4, p < 0.001).

Table 1: Summary of Differences Between Prescribed and Actual Weightbearing

The correlation between prescribed weightbearing and actual weightbearing also was examined (Fig 2). Spearman’s rank correlation was not significant (r_s = 0.189; p = 0.399; 95% confidence interval, −0.261–0.572), suggesting that there is little relationship between the weightbearing prescribed and actual weightbearing. However the large confidence intervals do not preclude the possibility of a relationship between actual and prescribed weightbearing.

Fig 2.: A scatter diagram shows actual partial weightbearing (PWB) and prescribed partial weightbearing (both expressed in percentage of body weight). There is considerable difference between weightbearing prescribed and actual weightbearing.

When actual weightbearing was expressed as a percentage of prescribed weightbearing, it was apparent that the volunteers (controls) did better than the patients (Table 2). On comparing both groups there was no statistical difference in their response to weightbearing partially (p = 0.179). None of the patients or volunteers was able to reproduce the extent of partial weightbearing that they had been trained for (Fig 2).

Table 2: Summary of Actual Partial Weightbearing Expressed as a Percentage of Prescribed Partial Weightbearing

DISCUSSION

Although there have been no prospective randomized clinical trials on the consequences of increased or decreased weightbearing, data from clinical and animal studies have shown that magnitude and frequency of loading influences bone healing.^{3,8–10,13,14,16,21} For example, physiologic loading that occurs during axial micromotion encourages fracture healing^8–10,16 but excessive or premature loading may inhibit or stop healing.^3,5,8,14,16 Partial weightbearing acts by allowing physiologic strains while avoiding excessive or premature loading. Based on current evidence it seems that magnitude of weightbearing does have clinical consequences,^{3,5,8–10,16,20,21} therefore it seems important for certain patients to weightbear partially.

The current study did not address clinical effects of partial weightbearing but assessed the accuracy of patients and healthy volunteers in weightbearing partially after being trained using the bathroom scale technique.^6,7,19 Force platforms were used to assess accuracy of weightbearing because they provide exact values of the ground reaction force produced on weightbearing from which the extent of weightbearing was calculated. It has been shown that on comparing various methods used to train patients to weightbear partially, force platforms were 2.7 times more accurate than traditional methods.⁷

On comparing various methods of training patients to partial weightbear, Gray et al⁷ found that weightbearing on the therapist’s hand was crude and subjective (23.7% accuracy), bathroom scales are marginally better (26.2% accuracy), whereas force platforms give the best results (66.6% accuracy). Perren and Matter¹² reported overloading of the limb in all patients they tested although excess weightbearing was reduced in those who used a device that provided feedback. In the current study neither patients nor volunteers was able to weightbear partially to the prescribed limit after being trained using the bathroom scale technique (Figs 2, 3). Only two volunteers and two patients exerted less weight than that required.

Fig 3.: A bar diagram was used to compare actual weightbearing with prescribed weightbearing. Actual weightbearing was less in Subjects 4, 12, 17, and 24 but all other subjects were weightbearing more than prescribed. Subject 18 exerted 212% body weight more than required.

Other volunteers and patients exerted greater weight than that required, with two patients weightbearing almost fully (Fig 3). There was little relationship between the weightbearing prescribed and actual weightbearing (p = 0.399), and the response of both groups in weightbearing partially was not statistically significant (p = 0.179). These results indicate that in patients and volunteers, partial weightbearing is a difficult skill to master.

One patient exerted 212% of body weight more than that prescribed (Patient 18, Fig 3). This phenomenon can be explained by calculating the vertical ground reaction force (F) from the formula F = mg + ma where m = mass, g = acceleration caused by gravity, and a = acceleration of the body’s center of mass. During crutch walking a large downward vertical shift in the body’s center of mass is produced, which increases the acceleration of the body’s center of mass (a). Therefore a large shift in the center of mass increases the value of ma, producing a force greater than the body weight.

Overall, the ability of volunteers to weightbear partially was better than that of the patients (Table 2). This may be because of the added factor of pain that the patients’ experienced which may have compromised their ability to weightbear partially. Chow et al⁶ reported that among factors influencing the ability to weightbear partially, muscle power and mental state are the most significant. Other factors such as age, body weight, and type of surgery do not seem to have any effect.⁶ Trauma or surgery may temporarily decrease muscle power, which can compromise a patient’s ability to weightbear partially.⁶

Difficulty in partial weightbearing is related to the fact that training with bathroom scales uses a static mode whereas walking is a dynamic activity.⁷ In clinical practice, patients may not be able to reciprocate the ability to weightbear partially because the sensation of static loading while standing still may be different than while walking, patients may not retain the mental image for long, and although some patients may exceed the prescribed load other patients may not obtain that desired load.^{1,7,11,17,18,19}

Winstein et al¹⁹ showed that some form of feedback is essential to improve the patient’s ability to weightbear partially. Concurrent feedback (feedback given during the course of the activity) improved immediate performance but was detrimental for long-term learning, whereas postresponse feedback (feedback that is given after completion of activity) was relatively detrimental for immediate performance but improved long-term learning. They concluded that concurrent feedback and postresponse feedback are required to improve the patient’s ability to weightbear partially.¹⁹ In an experimental setting, several authors have used self-contained electronic devices that fit inside the patient’s shoe or walking cane, but these devices cannot provide postresponse feedback.^2,11,17,18 The ability of force platforms to provide concurrent and postresponse feedback could explain the better results that were obtained when using them for training subjects to weightbear partially.^7,19 Bathroom scales are unreliable because they do not give concurrent or postresponse feedback and the pointer on the scale sways considerably making any reading difficult.⁷

In the current study, neither patients nor volunteers were able to reciprocate partial weightbearing after being trained with the bathroom scale method. The bathroom scale method of teaching patients to weightbear partially is crude, inaccurate, and involves a static situation. An inexpensive, easy to use, dynamic device that provides concurrent and postresponse feedback could improve the patient’s ability to weightbear partially.

Acknowledgments

We thank Mr. Lee Shepstone of University of East Anglia and Mr. Rodney Bennett of West Wales General Hospital for statistical analysis, and all volunteers, patients, and physiotherapists for participating in this project.