With the increasing number of total ankle arthroplasties (TAAs) [21, 22], physician guidelines on how to advise patients on returning to driving postoperatively are valuable. Evaluating safe return to driving after TAA is particularly important compared with safe return to driving after other lower-extremity procedures because the ankle was determined to be the most important mover in braking motions over the knee and hip . Studies have reported slight but remarkable impairment in braking motion after successful ankle arthrodesis, with a minimum 6-month postoperative evaluation [16, 23]. Although important in terms of ankle braking motion, these studies examined long-term postoperative effects; they did not evaluate the time at which patients can initially return to driving after ankle surgery. Additionally, there are differences in ROM after ankle arthrodesis and TAA [9, 12, 20, 24, 25] that make a separate evaluation of safe return to driving after TAA necessary.
The ability to accurately and efficiently determine when a patient can return to driving is important from both a patient safety and a medicolegal perspective [3, 10, 27]. Foot and ankle orthopaedic studies have suggested that most patients can safely return to driving at 9 weeks after open reduction and internal fixation of the right ankle and 8 weeks after right metatarsal osteotomy to correct hallux valgus [8, 18]. Although postoperative timeline recommendations are clinically helpful, an understanding of factors indicating that a patient cannot safely return to driving at the generally recommended time is also useful. But to our knowledge, no study has evaluated the timing of safe return to driving after TAA.
Therefore, we asked: (1) Does brake-reaction time return to a safe value within 6 weeks of TAA? (2) Are there factors associated with a delay of return of brake-reaction time to safe values after TAA?
Patients and Methods
Between July 2015 and November 2017, we performed 74 TAAs. Patients were treated with a Salto Talaris (Integra LifeSciences, Plainsboro, NJ, USA) fixed-bearing TAA by a fellowship-trained foot and ankle surgeon. Inclusion criteria for this study were that the adult patient drove regularly (at least 4 days per week) and had an active driver’s license. We excluded patients with pre-existing medical conditions that could impair their operation of a motor vehicle (for example, peripheral neuropathy, seizure disorders, and neuromuscular disorders). Additionally, patients consuming postoperative opioids at the time of testing and those with issues such as wound healing or deep vein thrombosis were ineligible. Further, 60 patients were enrolled, but one (2%) was excluded because of missing data, leaving 59 for analysis here (Fig. 1).
The study was conducted at a large multispecialty academic orthopaedic practice, and patients were cared for by five fellowship-trained foot and ankle orthopaedic surgeons (DIP, RJS, BSW, JND, SMR). Patient age ranged from 43 to 83 years (median 63 years); 59% of patients (35 of 59) were men, and 41% (24 of 59) were women.
A control group of 20 volunteer participants matched for age and sex who did not have right lower-extremity pathology or pain were was used to establish a passing brake-reaction time of 0.850 seconds. Participants in the control group performed the driving-simulation test under the same conditions as the TAA participants. The brake-reaction time results were evaluated in conjunction with brake-reaction time data supplied by the testing-equipment company using the Reaction Time Tester, Model RT-2S (Advanced Therapy Products, Glen Allen, VA, USA) . The median (range) brake-reaction time for the control group was 0.547 seconds (0.452 to 0.850). Therefore, a passing brake-reaction time was established as less than 0.850 seconds because this was the slowest brake-reaction time in our healthy control group and within the safe range of reported brake-reaction times (1.25 to 0.50) [7, 11].
In an investigation of safe driving, the brake-reaction time has been reported to be the most reproducible skill to test and is used by the United States National Highway Traffic Safety Administration when setting safe driving standards . In a previous study, the safe brake-reaction time ranged from 0.7 to 1.25 seconds . Using a validated instrument , we chose to set the threshold for a safe brake-reaction time at 0.850 seconds because this was the longest response time in our healthy control group. The 0.850-second threshold was used to determine if a patient had a passing or failed brake-reaction time at 6 weeks postoperatively.
At the 6-week postoperative visit, before brake-reaction time testing, patients in the study group completed a driver-readiness survey (Fig. 2), VAS for pain, and an American Orthopaedic Foot and Ankle Society-Hindfoot assessment. Furthermore, a fellowship-trained foot and ankle orthopaedic surgeon (DIP, RJS, BSW, JND, SMR) assessed patients for ankle plantarflexion and dorsiflexion ROM using dedicated weightbearing lateral radiographs made with the ankle in maximum plantar flexion and dorsiflexion.
Brake-reaction time was tested at 6 weeks postoperatively and repeated weekly until patients achieved a passing brake-reaction time (Fig. 3). The Reaction Time Tester is a commercial device that is accurate to 0.005 seconds; it requires ankle flexion and pressure applied to a pedal that simulates braking . A previous study of braking performance after hip arthroplasty defined sufficient force for brake application as at least 98 N . In a recent study that used the same braking simulator as in the current study, all participants who successfully depressed the brake pedal applied at least 93 N, suggesting that patients with a passing brake-reaction time in our study likely demonstrated both sufficient strength and reaction time to safely brake in an emergency situation.
The brake pedal from the Reaction Time Tester was placed at approximately the location of the pedals in a standard vehicle. Furthermore, red and green lights were placed at the patient’s eye level. A single trained investigator (ELM) performed all testing. Patients were either barefoot or allowed to wear comfortable shoes during testing; however, few patients underwent the brake-reaction time barefoot. Patients were instructed to depress the pedal to illuminate the green light. To prevent the participant from reacting before the light changed colors, the investigator stood behind the participant (Fig. 3). Then, the investigator randomly pushed “test” on a handheld device and the light immediately turned from green to red. The red light indicated to the patient that he or she should depress the brake pedal. The time between when the light turned red and the time at which the patient pressed down with adequate force on the brake pedal was defined as the brake-reaction time. The patient performed one practice trial and three recorded test trials. From an average of the three test trials, the patient’s brake-reaction time was calculated.
The four-question driver-readiness survey included the following questions: (1) “I think my brake reaction time is slower than most drivers my age”; (2) “I think my brake reaction time is faster than most drivers my age”; (3) “I think my brake reaction time is about the same as most drivers my age”; and (4) “Based on what I think my brake reaction time is, I think I am ready to drive.” Patient responses were Strongly Disagree, Disagree, Neither Agree nor Disagree, Agree, or Strongly Agree. For Question 1 (“slower”), responses were scored from 1 (“Strongly Agree”) to 5 (“Strongly Disagree”). Meanwhile, for Questions 2 through 4 (“faster, “same,” and “ready”), the responses were inversely scored from 1 (“Strongly Disagree”) to 5 (“Strongly Agree”). When validating the questionnaire, we removed Question 3 because of low external validity and total score correlation. As a result, the maximum total score was 15 points. Based on a large, prospective study validating a driver-readiness survey at our institution, a score of 10 to 15 points was determined as a passing score.
The VAS pain scale reported pain at rest, with 0 indicating no pain and 10 indicating worst possible pain. The patients completed a survey to self-report this information.
The American Orthopaedic Foot and Ankle Society-Hindfoot assessment ranges from a score of 0 to 100 as worst and best possible score, respectively . The MCID has not been established for TAA. However, the MCID for hallux valgus surgery ranges from 7.9 to 30.2 .
If the patient had a failed 6-week postoperative brake-reaction time, he or she took the test again each week until they reached a passing brake-reaction time. Additionally, at these retests, the patient was given the driver-readiness survey to complete.
After TAA, the patient’s leg was placed into a short-leg plaster splint and he or she was told not to bear weight. At the 2-week postoperative clinic visit, sutures were removed if incisions were adequately healed, and the patient began using a fracture boot. Furthermore, the patient was allowed to start active and passive ankle motion in the sagittal plane at home but he or she was instructed to avoid bearing weight. At the 6-week postoperative clinic visit, in the protected environment of physical therapy, patients were allowed to walk in a normal shoe. This allowance was made so that the physical therapist could work on strength and ROM. Meanwhile in a social environment, patients were instructed to remain in a fracture boot and bear weight as tolerated.
Primary and Secondary Study Outcomes
The primary study endpoint of interest was the percentage of patients who achieved a brake-reaction time of < 0.850 seconds at the assessment performed 6 weeks after surgery.
The secondary study endpoint examined the factors associated with a delay of return of brake-reaction time to safe values after TAA. We assessed this by comparing patients who did or did not pass the brake-reaction time and VAS scores for pain, American Orthopaedic Foot and Ankle Society-hindfoot scores, ankle plantarflexion, ankle dorsiflexion, and driver-readiness results.
A qualified statistician (CF) performed the statistical analysis. For the study and control groups, a descriptive univariate analysis was used. For the study group, the threshold of the time to recovery of the brake-reaction time was expressed using the mean, SD, and 95% confidence interval. Comparisons between patients who did or did not pass the brake-reaction time and VAS scores for pain, American Orthopaedic Foot and Ankle Society-hindfoot scores, ankle plantarflexion, ankle dorsiflexion, and driver-readiness results were assessed via Student’s t-test. Pain scores were log transformed for statistical analysis due to non-normal distribution of the data. In a separate analysis, driver-readiness results and brake-reaction time were evaluated using Spearman’s correlations for two variables. Finally, data were summarized by mean and SD; pain scores were summarized by median and interquartile range. Significance was determined at p < 0.05.
Most patients can safely return to driving within 6 weeks after TAA, based on brake reaction time and patient responses to a questionnaire. Ninety-two percent of patients (54 of 59) achieved a passing brake-reaction time and were considered safe to drive, and five patients had a failed brake-reaction time (Table 1). The mean brake-reaction time of the passing group was 0.626 seconds (± 0.111). In comparison, the average brake-reaction time for those with a failed brake-reaction time was 1.120 seconds (± 0.117; p < 0.001). At 9 weeks postoperatively, all patients who completed the study achieved a passing brake-reaction time. Among the five patients who initially had a failing brake-reaction time and returned for follow-up brake-reaction time testing, the patients had a passing brake-reaction time at an average (range) of 8.9 weeks after TAA (8.1 to 9.7 weeks).
Patients with a failed brake-reaction time at 6 weeks postoperatively had greater median VAS pain scores than those with a passing brake-reaction time (3 [IQR 2 to 7] versus 1 [IQR 0 to 3]; p = 0.022). However, the American Orthopaedic Foot and Ankle Society-hindfoot score did not differ, with an average American Orthopaedic Foot and Ankle Society-hindfoot score of 51.4 (± 7.9) for patients with a failed brake-reaction time and 62.9 (± 14.2) for those with a passing brake-reaction time (p = 0.082). Those who passed the brake-reaction time had greater ankle plantarflexion ROM than those who failed the brake-reaction time (24° [± 10°] versus 14° [± 5°]; p = 0.029). However, there were no differences in ankle dorsiflexion between groups; the average dorsiflexion angle for patients with a passing brake-reaction time was 8° (± 4°) compared with 7° (± 7°) for those with a failed brake-reaction time (p = 0.729). All five patients with a failed brake-reaction time at 6 weeks postoperatively also had failing driver-readiness survey scores. Better brake-reaction times were well correlated with better driver-readiness survey scores (R = -0.402, [95% CI -0.597 to -0.163]; p = 0.002).
With the increasing usage of TAA [4, 6, 15, 21, 22] and the related patient-safety and medicolegal pressures [3, 5, 10, 19], recommendations for when patients can safely return to driving and the clinical factors that indicate a lack of readiness are valuable. This is particularly relevant in view of the importance of the ankle in lower-extremity braking motion . Our goal was to determine if brake-reaction time returned to a safe value within 6 weeks following right TAA. Further, we hoped to discern if there were factors associated with a delay of return of brake-reaction time to safe values after TAA. In our investigation, although almost all patients were fit to drive at 6 weeks postoperatively, factors that were associated with a failing brake-reaction time at this time were a higher VAS score for pain, limited plantarflexion ROM, and a failed score on the driver-readiness survey.
There were several limitations. A limitation of the study is that the braking test was simulated, and many factors of cognitive coordination are involved when braking in a real automobile . Brake-reaction time is just one factor in safety for driving; some people are never safe to drive regardless of their ankle status. Second, although the ability to press the brake in a timely manner in response to something in the road is vital to avoiding accidents, it is not the only perioperative variable that must be considered. Additionally, preoperative testing of the braking time was not performed. Preoperative testing would have provided information on the time at which the patients’ brake-reaction time returned to baseline. However, we felt that comparing these patients with healthy participants provided a more rigorous standard because patients with ankle arthritis may have an abnormal preoperative brake-reaction time because of their ankle’s pathology. We did not examine preoperative VAS for pain, AOFAS score, and ROM in this study. We focused on postoperative measures because we felt this would better indicate patient recovery. Lastly, there may be a subset of patients whose brake-reaction time returns to within the normal range earlier than 6 weeks postoperatively; however, our study did not evaluate timepoints before 6 weeks.
Proportion of Patients Achieving Normal Brake-Reaction Time 6 Weeks After Surgery
We found that most patients who underwent TAA achieved brake-reaction times no different from those of the control group by 6 weeks after surgery. Although there is a precedent of comparing brake-reaction time after ankle arthrodesis to a control cohort, these patients were assessed over 6 months postoperatively [16, 23]. This is the first study to determine brake-reaction time after TAA as well the expediency of returning to driving following end-stage ankle arthritis surgery [8, 16, 23].
Factors Associated with Failing the Brake-Reaction Time Test at 6 Weeks
We found that patients with higher VAS pain scores and worse ROM were less likely to be able to drive within 6 weeks of surgery than those with lower scores and better ROM. No prior study has evaluated factors associated with brake-reaction time after TAA. However, Egol et al.  investigated safe brake reaction time and functional outcome scores following ankle fracture, and no relationship was found. Further, Jeng et al.  evaluated brake reaction time and functional scores in the ankle arthrodesis population. Similar to Egol et al.  in the ankle arthrodesis population, no relationship was found between reaction time and functional scores . Outside of the scope of orthopaedic surgery of the ankle, pain in TKA  and hallux valgus  and ROM in hallux valgus  were found to be predictors of brake-reaction time. Although future studies can further investigate thresholds for driving recommendations, in our study a VAS of 3 or greater and a minimum plantarflexion of 14° were indicators that the patient was unsafe to drive.
More than 90% of patients in this series achieved a safe brake-reaction time within 6 weeks of TAA, and those who did not were more likely to have had more pain and a stiffer ankle. Surgeons might counsel patients with persistent pain and stiffness at 6 weeks to delay driving for an additional 3 weeks, since by 9 weeks after TAA, all patients in this series had a brake-reaction time comparable with patients who had not undergone surgery. Future studies might elucidate what key gaps in knowledge remain and determine a practical way to answer these questions.
We thank Carol Foltz PhD, for her assistance in the preparation of this manuscript.
1. Brake A, Hernandez L, Mann L, Dickerson A, Reistetter T. The RT-2S simple reaction time tester: test/retest reliability. Research for the Older Adult Driving Initiative (ROADI), East Carolina University. https://www.atpwork.com/images/pdfs/test_retest.pdf
. Accessed July 26, 2018.
2. Chan HK, Chen JY, Zainul-Abidin SZ, Ying H, Koo K, Rikhraj IS. Minimal clinically important differences for American Orthopaedic Foot & Ankle Society score in hallux valgus surgery. Foot and Ankle Int. 2017;38:551-557.
3. Chen V, Chacko AT, Costello FV, Desrosiers N, Appleton P, Rodriguez EK. Driving after musculoskeletal injury. Addressing patient and surgeon concerns in an urban orthopaedic practice. J Bone Joint Surg Am. 2008;90:2791-2797.
4. Chou LB, Coughlin MT, Hansen S Jr, Haskell A, Lundeen G, Saltzman CL, Mann RA. Osteoarthritis of the ankle: the role of arthroplasty. J Am Acad Orthop Surg. 2008;16:249-259.
5. Coombes v Florio. 450 Mass 182 (2007). Available at: http://masscases.com/cases/sjc/450/450mass182.html
. Accessed July 26, 2018.
6. Cracchiolo A 3rd, Deorio JK. Design features of current total ankle replacements: implants and instrumentation. J Am Acad Orthop Surg. 2008;16:530-540.
7. Dobbs BM. Medical Conditions and Driving: A Review of the Scientific Literature (1960 – 2000). Association for the Advancement of Automotive Medicine; 2005. Available at: https://icsw.nhtsa.gov/people/injury/research/medical_condition_driving/pages/Sec1-Intro.htm
. Accessed July 26, 2018.
8. Egol KA, Sheikhazadeh A, Mogatederi S, Barnett A, Koval KJ. Lower-extremity function for driving an automobile after operative treatment of ankle fracture. J Bone Joint Surg Am. 2003;85:1185-1189.
9. Flavin R, Coleman SC, Tenenbaum S, Brodsky JW. Comparison of gait after total ankle arthroplasty and ankle arthrodesis. Foot Ankle Int.
10. Giddins GE, Hammerton A. "Doctor, when can I drive?': a medical and legal view of the implications of advice on driving after injury or operation. Injury. 1996;27:495-497.
11. Green M. “How long does it take to stop?” Methodological analysis of driver perception-brake times. Transport Hum Factors.
12. Hahn ME, Wright ES, Segal AD, Orendurff MS, Ledoux WR, Sangeorzan BJ. Comparative gait analysis of ankle arthrodesis and arthroplasty: initial findings of a prospective study. Foot Ankle Int.
13. Haslegrave CM. Force exertion. In: Delleman NJ, Chaffin DB, Haslegrave CM, eds. Working Postures and Movement: Tools for Evaluation and Engineering. 1st ed. Boca Raton, FL: CRC Press; 2004:367-397.
14. Hernandez VH, Ong A, Orozco F, Madden AM, Post Z. When is it safe for patients to drive after right total knee arthroplasty? J Arthroplasty. 2016;31:2495-2498.
15. Heuval AV, Bouwel SV, Dereymaeker G. Total ankle replacement. Design evolution and results. Acta Orthop Belg.
16. Jeng CL, Lin JS, Amoyal K, Campbell J, Myerson MS. Driving brake reaction time following right ankle arthrodesis. Foot Ankle Int.
17. MacDonald W, Own JW. The effect of total hip replacement on driving reactions. J Bone Joint Surg Br. 1988;70:202-205.
18. McDonald E, Shakked R, Daniel J, Pedowitz DI, Winters BS, Reb C, Lynch MK, Raikin SM. Driving after hallux valgus surgery. Foot Ankle Int.
19. Phillips JJ. Products liability for personal injury to minors. Virginia Law Review. 1970;56:1223-1241.
20. Piriou P, Culpan P, Mullins M, Cardon JN, Pozzi D, Judet T. Ankle replacement versus arthrodesis: a comparative gait analysis study. Foot Ankle Int.
21. Pugely AJ, Lu X, Amendola A, Callaghan JJ, Martin CT, Cram P. Trends in the use of total ankle replacement and ankle arthrodesis in the United States Medicare population. Foot Ankle Int.
22. Raikin SM, Rasouli MR, Espandar R, Maltenfort MG. Trends in treatment of advanced ankle arthropathy by total ankle replacement or ankle fusion. Foot Ankle Int.
23. Schwienbacher S, Aghayev E, Hofmann UK, Jordan M, Marmotti A, Röder C, Ipach I. Driving and emergency braking may be impaired after tibiotalar joint arthrodesis: conclusions after a case series. Int Orthop.
24. Seo SG, Kim EJ, Lee DJ, Bae KJ, Lee KM, Lee DY. Comparison of multisegmental foot and ankle motion between total ankle replacement and ankle arthrodesis in adults. Foot Ankle Int.
25. Singer S, Klejman S, Pinsker E, Houck J, Daniels T. Ankle arthroplasty and ankle arthrodesis: gait analysis compared with normal controls. J Bone Joint Surg Am.
26. Spalding TJ, Kiss J, Kyberd P, Turner-Smith A, Simpson AH. Driver reaction times after total knee replacement. J Bone Joint Surg Br.
27. Yousri T, Jackson M. Ankle fractures: When can I drive, doctor? A simulation study. Injury.