Surgical correction of hallux valgus deformity by osteotomy of the first metatarsal is a common elective orthopaedic procedure, and early recovery is often characterized by pain and functional limitation. Patients undergoing lower limb surgery frequently ask when it is safe to return to driving since the ability to drive is important both in social and economic contexts. The advice given to patients is often anecdotal and based on the surgeon's experience as little scientific data are available on this subject.
A number of studies have investigated brake response time with regard to recovery after total knee and hip arthroplasty1-5. Those investigators reported that the ability to transfer the right foot from the accelerator to the brake pedal was not recovered for between four and eight weeks after surgery. Egol et al.6 assessed driving function after surgical treatment of ankle fractures and reported that braking times did not recover to control values until nine weeks after surgery.
Although first metatarsal osteotomy is considered a more minor procedure compared with lower limb arthroplasty, the effects of operating on the right foot are potentially more disabling when it comes to applying substantial amounts of pressure to the forefoot during emergency braking. At our institution, it is routine practice to advise patients to avoid driving for a minimum of six weeks after a first metatarsal osteotomy. This period of time is considered necessary to permit sufficient osseous union at the osteotomy site in order to safely allow forceful emergency braking, which is an essential function for safe driving. There is no current evidence, however, to indicate what the effects of first metatarsal surgery are with regard to emergency brake response time. We therefore conducted a prospective, observational study involving patients undergoing unilateral first metatarsal osteotomy on the right side to analyze the effect of this procedure on emergency brake response time and to determine whether six weeks was sufficient time for the emergency braking function to improve to the preoperative levels to enable a safe return to driving.
Individuals recruited into the study were identified at preoperative assessment clinics and had a diagnosis of symptomatic hallux valgus defined by a hallux valgus angle of >15° with no fixed deformities of the lesser toes. All operations were unilateral first metatarsal osteotomies on the right side performed by, or under the direct supervision of, the senior author (C.S.K.). Any patient who had a medical or physical condition that would affect the brake reaction time was excluded from the study. Testing was carried out for all patients at approximately two weeks (thirteen days) and six weeks (forty-one days) after surgery at prearranged outpatient clinics.
A chevron, Scarf, or basal first metatarsal osteotomy was performed, depending on the severity of the deformity. A soft-tissue release was carried out when it was considered necessary. The osteotomies were stabilized with use of screws and/or plates. Postoperative rehabilitation was standardized. Patients were advised to bear weight on the heel postoperatively for six weeks, after which forefoot weight-bearing was commenced. Simple dressings were applied, and cast immobilization was not used in any of the patients.
A brake reaction timer was used to calculate brake response times from all participants. This machine was based on a design described by Pierson et al.4 (automatic brake reaction timer, model 3548; American Automobile Association, Traffic Safety Department, Heathrow, Florida) and consists of a steering wheel, brake pedals, a liquid crystal display, a microprocessor, and a hand control unit. The screen was positioned at the subject's eye level, and the steering wheel was placed at a height to represent his or her normal driving position. A brake-accelerator pedal unit comparable with those in a vehicle with an automatic transmission was used and was fixed to the floor. Each subject was tested in the seated position, and the distance between the chair and the pedal was adjusted to simulate the individual's normal driving position. During the test protocol, the subject was asked initially to fully depress the accelerator pedal. The subject was then instructed to use the steering wheel to follow a randomly moving image displayed on the liquid crystal display. The purpose of this was to distract the subject so that they were not "primed" to react to the stimulus to brake. The test was initiated by the examiner who randomly depressed a switch on the hand control unit, which was kept out of sight of the patient. The unit then instructed the subject to depress the brake pedal by displaying the words "BRAKE NOW" across the screen. The subject would then release the accelerator pedal and immediately fully depress the brake pedal with the right foot.
In order to examine the different components of emergency braking, we used terminology based on that described by Green7. Total brake response time is defined as the amount of time an individual takes to perceive a potential danger and then initiate a motor response to apply pressure to the brake pedal. Total brake response time can be subdivided into two components that involve different psychomotor processes. Reaction time is the time it takes for the responder to perceive a danger and to decide on a motor response. Brake time is the time taken to perform the required muscle movements to lift the foot off of the accelerator pedal, move it to the brake pedal, and then depress the pedal. Surgery may prolong the reaction time by causing hesitation to react and similarly may affect the brake time by limiting movement because of pain or stiffness.
The machine recorded the time taken from the instruction to brake until the moment the patient began to lift the foot from the accelerator pedal; this was recorded as the reaction time. The machine also calculated the time taken from the moment the patient began to release the accelerator pedal to the moment the patient fully depressed the brake pedal; this was recorded as the brake time. Both of these values combined to give the total brake response time.
A minimum of three practice trials were performed at each visit, followed by ten test trials. The highest and lowest times were excluded. As such, we were able to assess the ability of a patient to perform a so-called emergency stop procedure, which one must be able to perform adequately in order to stop a vehicle quickly should the need arise.
The validity and reproducibility of testing equipment is fundamental in studies such as this. As such, the testing device was calibrated by our medical physics department to be accurate to within ±5 msec. In order to determine the reproducibility of the test, we calculated the difference between the highest and lowest total brake response values recorded for each patient (after excluding the two initial extreme values as described) and used this to calculate a mean value for each study group. The mean difference in brake response time was 42 msec for the control group and 54 msec for the study group.
During the testing process, each patient was asked to wear footwear of his or her own choice that would most closely resemble the footwear the patient would normally wear when driving. We did not think it was appropriate to perform the test without footwear as this would not mimic the normal driving scenario. Similarly, a surgical shoe was not worn during testing as driving in a surgical shoe would not be permitted by United Kingdom driving regulations.
A simple clinical test, termed the step test, was also performed as it was relevant to the function of driving a motor vehicle. This test is performed with the patient in the seated position with the right foot flat on the floor lateral to a 2-in (5-cm)-square strip of wood. The patient must lift the foot, transfer it over the wood block, and place it flat on the floor to the left of the block. The score relates to how many times the patient is able to perform this maneuver in a 10-sec period. The step test was used as a means of assessing the ability of an individual to raise, transfer, and then place the foot over an obstacle.
A control cohort, which consisted of twenty-eight healthy volunteers who were matched for age, sex, and driving frequency, was also tested (Table I). The control group consisted of individuals with no foot pathology and/or deformity or any other medical and/or physical condition that would affect the brake reaction time. This cohort was included in order to determine whether having symptomatic hallux valgus had an effect on brake response time.
Data analysis was performed with use of the chi-square (categorical data) and analysis of variance tests (continuous data). P values of <0.05 were considered significant. A post hoc power analysis was carried out to determine whether the sample size was adequate to show a significant difference between study groups. The study had a power of 0.8 to detect a difference of 98 msec with use of an alpha value of 0.05 and a standard deviation of 244 msec as observed in the study group.
Twenty-eight patients were enrolled in the study. The study and control groups were matched for age, sex, and frequency of driving (Table I). Table II and Figure 1 document total brake response time and its component elements at each assessment period.
At two weeks after surgery, only seven patients (25%) were able to complete the test. The total brake response time, the reaction time, and the brake time in this patient cohort were found to have increased at two weeks after surgery compared with preoperative values; however, the difference was not significant. The remaining patients reported that they were unable to perform the required tests because of pain in the affected foot. This would indicate that for the majority of patients, driving at two weeks after surgery is not possible as a consequence of postoperative pain.
At the six-week review, a significant reduction was seen in total brake response time, reaction time, and brake time compared with preoperative values (Table II, Fig. 1). Total brake response time remained significantly higher for the study group at six weeks compared with the control group (p = 0.047). Both brake time and reaction time were also higher in the study group at six weeks compared with the control cohort, although these differences were not significant (p = 0.114 and p = 0.445, respectively).
In the study group, the mean brake response time was lower for men (673 msec) compared with women (822 msec); however, this difference was not significant because of the small number of male participants in the study. This finding also applied to the control group.
Driving status was found to have an important effect on brake response times, with regular drivers recording significantly better times (Table III).
The type of osteotomy and whether a soft-tissue release was performed did not have a significant effect on total brake response time or its components at six weeks after surgery; however, because of the small sample size, there is insufficient statistical power to make meaningful comparisons.
No patient experienced a clinically important postoperative complication by the six-week follow-up examination, and all patients had returned to the use of their own footwear by this time. At six weeks, all osteotomies were judged to be healed on the basis of the radiographic appearance.
The question of when it is safe to resume driving is frequently posed by patients undergoing surgery to the lower limb. Excluding a patient from driving for an unnecessarily long period is unrealistic and may have substantial social and economic consequences. Equally important, advising a patient to return to driving prematurely is potentially dangerous and may make the surgeon vulnerable to litigation should the patient be involved in an accident while driving. Relatively small changes in emergency braking time can be important. If a car is traveling at 40 mph (64 km/h), it covers approximately 60 ft/sec (18 m/sec). A 100-msec increase in emergency braking time would mean that the car must travel an additional 6 ft (1.8 m) before coming to a complete stop. Such small changes in braking time may represent the difference between a collision and the ability to stop in time. Giddins and Hammerton discussed the medicolegal issues that arise when a patient who has recently suffered an injury or had a surgical procedure wishes to drive8. The legal position concerning people who drive after surgery is reasonably clear. An individual should not drive if, by reason of injury and/or surgery, the individual is likely to present a danger to himself/herself or to others. Additionally, under the terms of the patient's insurance policy, the insurer is usually entitled to refuse insurance coverage in cases when the driver had an accident while recovering from recent surgery that may have impaired his or her driving ability. However, when questioned, insurers indicated that coverage would not normally be withdrawn when the patient has reasonably relied on appropriate medical advice.
Studies of brake response time have described a wide variety of results that vary under specific conditions. Typical brake response times to anticipated stimuli have been reported in the range of =0.75 sec7. Such values are in keeping with those observed in our control cohort. One of the most important variables is driver expectation, which may affect brake response times by a factor of two7. We attempted to minimize the effects of expectation by asking the patient to concentrate and follow a moving target using the steering wheel and by the observer initiating the test randomly and out of sight. However, such precautions do not recreate the element of surprise that occurs when an unexpected danger presents itself during the actual driving experience, and this factor cannot be fully addressed in such artificial experiments. Brake reaction times are also subject to influence by other factors such as age, sex, and driving experience. As such, we used patient groups matched for these variables. Preoperatively, the study group had significantly increased total brake response time, reaction time, and brake time compared with the control group, which may be a consequence of symptomatic hallux valgus.
The results of this study show that the total brake response time after a first metatarsal osteotomy improves significantly by six weeks after surgery. These patients also had improvement in both reaction time and brake time compared with preoperative values. Although the total brake response time remained significantly higher for the study group at six weeks compared with the control group (p = 0.047), both the brake time and the reaction time were not significantly different and approached those of the healthy control group. However, both reaction time and brake time may improve with practice, and some of the improvement observed in the study group may not be entirely due to postoperative recovery and may represent a learning response7.
As we previously discussed, surgery to the foot or ankle may have a relatively greater effect specifically on brake function because of the inhibitive effects of pain when a brake pedal is forcibly depressed. In support of this, Egol et al.6 reported that brake times did not recover to control values until nine weeks after surgery in a cohort of patients undergoing operative fixation of ankle fractures. In that study, weight-bearing was not commenced until six weeks after surgery, when testing began, in order to allow time for the fracture to unite. The longer time to return to control values may reflect the location of the injury or indeed the inability to bear weight prior to testing. The patients in our study were advised to bear weight on the heel for the duration of the study. These two studies do not provide enough information to determine the effects of non-weight-bearing for a period of time prior to performing such brake response analysis.
In view of the improvement in the total brake response time by six weeks after surgery, the question arises as to whether one may return to driving sooner than six weeks. While we cannot specifically answer that question, it is clear that performing an emergency stop maneuver involves placing a substantial amount of force through the forefoot in order to rapidly depress the brake pedal. It is our opinion that waiting six weeks is sensible in that it allows a sufficient period of time for the osteotomy to heal and sustain that type of force without displacement.
This is the first study, as far as we know, to investigate emergency braking time after corrective osteotomy for hallux valgus. Assuming that other factors that are required to safely operate a motor vehicle are acceptable, these data suggest that patients may safely resume driving at six weeks. The ability to safely operate a vehicle is, however, multifactorial, and brake response time is only one of several important factors that should be considered.
Surgeons should take care when giving advice regarding driving after surgery and should record their advice in the clinic notes. In particular, they should bear in mind that the patient might intend or be required to drive a vehicle other than an automobile, which may entail greater or different demands. 