Improvement in Driving Reaction Time After Total Hip Arthroplasty

Abstract

Decisions made regarding resumption of driving after total hip arthroplasty may be determined by a combination of factors including driving reaction time and when postsurgical precautions need no longer be adhered to. Ninety patients, ranging in age from 34 to 85 years old were recruited after total hip arthroplasty to measure driving reaction time preoperatively and from 1 to 52 weeks postoperatively. Driving reaction time worsened 1 week postoperatively for patients who had a right hip arthroplasty. The driving reaction time then improved up to 1 year postoperatively. Patients who had a left arthroplasty improved from 1 week postoperative. In general, patients reach their preoperative driving reaction time 4 to 6 weeks postoperatively and continue to improve.

Driving a car is an essential component of daily life for many older people. In 1985 there were 155 million American drivers, 9.5% of whom were older than 65 years and who regard driving as necessary to maintain their independence. ¹ The Transportation Research Board predicts that by the year 2020, 20% to 25% of the population with driving privileges will be older than 65 years. ^6,10,11

In 1999, 168,000 total hip arthroplasties were done in the United States. ¹⁰ The average age of patients having a total hip arthroplasty is 65 years. ⁵ The primary reason a total hip arthroplasty is done is to relieve pain. ⁵ A painful hip affects a patient’s ability to do basic activities of daily living including driving a car. ^5,7 During postoperative rehabilitation after total hip arthroplasty, inability to drive a car may have tremendous impact on a patient’s life. ^4,7 One of the questions most commonly asked by patients after total hip arthroplasty is when can they return to driving. ⁴ Healthcare professionals must be concerned with monitoring elderly individuals’ ability to drive, for their safety and the safety of others. These concerns need to be balanced with restoring and preserving the individual’s ability to participate in normal activities. ^4,7

Many physicians make recommendations for resumption of driving after total hip arthroplasty based on a 4- to 6-week soft tissue healing and patient recuperation time. These recommendations are made empirically without scientific data. Most patients no longer are required to adhere to total hip arthroplasty precautions after 4 to 6 weeks postoperatively. ⁹ After total hip arthroplasty using a posterior approach, patients must avoid hip flexion greater than 90°, internal rotation past neutral, and hip adduction past neutral. ⁴ These movements must be avoided singularly or in combination to avoid hip dislocation because during surgery the surgeons dislocate the anatomic hip when doing a total hip arthroplasty. Patients are cautioned not to duplicate these movements until the muscles and surrounding tissue around the new hip have sufficiently healed. ⁹ By telling patients who had a total hip arthroplasty to adhere to total hip arthroplasty precautions, their hip motions are being restricted in all planes to avoid hip dislocation. Some surgeons do not allow patients who have had a hip replacement to drive until their first postoperative office visit, which is usually 4 to 6 weeks after surgery. One objective measure of driving capabilities is driving reaction time. It was postulated that total hip arthroplasty would have an effect on driving reaction time. Driving reaction time has been studied for more than 2 decades using different driving simulators.

The purpose of the current study was twofold: to determine the rate of improvement in driving reaction time after total hip arthroplasty, and to determine when driving reaction time after total hip arthroplasty returns to preoperative values. A prospective observational trial was done that measured preoperative and serial postoperative driving reaction times in patients having unilateral total hip arthroplasty.

MATERIALS AND METHODS

The eligibility criteria for inclusion in the study included: adult patients, having a unilateral cemented or hybrid total hip arthroplasty secondary to osteoarthritis, with a valid drivers license and who primarily drove a car with an automatic gear shift and used their right foot to brake. Patients were excluded from the study if they did not speak English, were nondrivers, had not driven for more than 3 months before surgery, principally drove a car having a manual gear shift, or used their left foot to brake. Patients who used their left foot to brake were considered to be two-footed drivers. Patients were recruited in the study on admission to the hospital, and informed consent was obtained. The operating room list was checked daily to determine the number of patients having total hip arthroplasty from a pool of five surgeons participating in the study. One hundred fifty consecutive patients were approached prospectively to obtain a study cohort of 100 patients. There were 50 ineligible patients, 32 were nondrivers, eight had not driven for more than 3 months before surgery, six were non-English speaking, and four used their left foot to brake. All patients who were asked, agreed to participate. Vision was not assessed during this study. The minimum visual acuity requirement to obtain a driver’s license in New York State is corrected vision that measures 20/40. Subjects were asked whether they had a change in prescription during the course of the study and there were no reported changes in vision. All patients met the New York State requirements for vision. Ten patients were dropped from the study because they had no postoperative data or because they no longer met the inclusion criteria: three patients were discharged from the hospital before the second driving reaction time test and subsequently did not return for any postoperative visits; three patients became two-footed drivers postoperatively (used the left foot to brake); three patients admitted for unilateral total hip arthroplasty had staged bilateral total hip arthroplasties, and one patient’s surgery was canceled.

Demographics

The 90 subjects ranged in age from 34 to 85 years, and had a mean age of 68.9 years. There were 56 men and 34 women. There were 52 right and 38 left total hip arthroplasties. Height, weight, age, gender, and number of years driving were compared. There were no statistically significant differences between groups as defined by the side that had the total hip arthroplasty (Table 1). Medications were recorded during the course of the study and there were no significant changes in cardiac, psychotropic, or anticholinergic medications. There was a decrease in the amount of pain medication taken postoperatively when compared with medication taken preoperatively.

TABLE 1: Demographics

Driving Reaction Time Assessment

The automatic brake reaction timer (model 3548) was obtained from the American Automobile Association (Traffic Safety Department, American Automobile Association, Heathrow, FL) and is the same as their testing device. It is a driving simulator system that has a car brake and accelerator pedal. A digital display stop clock allows the examiner to measure reaction time to the nearest 1/100 of a second. To adhere to total hip arthroplasty precautions, a standard 24-inch high chair was used for testing to avoid hip dislocation. All subjects were given the same instructions regarding the use of the driving reaction time simulator. The driving simulator was placed in front of each subject. Each subject was instructed to place his or her right foot on the accelerator and keep it depressed to the floorboard sufficiently to keep a green light lit on the driving reaction time simulator. Then subjects were instructed that as soon as a red light appeared they were to place their right foot on the brake pedal as quickly as possible. The timing between the green and red lights was controlled randomly for 3, 4, or 5 seconds. The subjects were given three practice trials and then 10 test runs. The highest and lowest reaction times were eliminated and the remaining eight trials were averaged. This method of testing was recommended by the manufacturer. Subjects were tested 24 hours before surgery, and on four occasions after surgery: at 1 week, 4 to 6 weeks, 26 weeks, and 52 weeks postoperatively. The driving reaction time assessments done postoperatively coincided with scheduled visits with the surgeon.

Before commencing the current study, the reproducibility of the test was assessed. A random selection of 100 people was asked to use the machine on two occasions. The results of this test were compared. There was no significant difference between the two tests. The mean test scores on the two test dates were within 3/100 of a second.

This pilot study also served to test the conditions of the experiment, and the instructions to the participants.

Normative driving reaction time data were obtained from the American Automobile Association, Traffic Safety Department (Table 2). ¹ The driving reaction time was tested using the same driving reaction time simulator used in the current study. The norms for the American Automobile Association represent the distribution of scores obtained when testing 1262 persons. There were 688 women and 574 men ranging in age from 21 to 80 years. Separate reaction time score distributions were developed for men and women. Each chart contains the reaction time scores for the total population, plus distributions for four subgroups divided by age as follows: 21 to 35 years, 36 to 55 years, 56 to 65 years, and older than 66 years.

TABLE 2: American Automobile Association, Traffic Safety and Engineering Department Reaction Time Norms by Percentile Rank 0–100 With Five Subgroups A–E

There is no safe driving reaction time established or recommended.

Physical Therapy Protocol

All patients recruited in the study had a diagnosis of osteoarthritis of the hip and had a cemented or hybrid total hip arthroplasty under epidural anesthesia using a posterior approach. Physical therapy was initiated on the first postoperative day. All patients were instructed to adhere to the same total hip arthroplasty precautions. A therapeutic exercise program was initiated that included gluteal setting, quadriceps setting, supine hip and knee flexion to 45°, and supine hip internal rotation to neutral. Transfer training in and out of bed and ambulation training with a walker were initiated on the first postoperative day. The patients’ weightbearing status ranged from partial weightbearing (46%) to weightbearing as tolerated (54%). The average length of hospital stay was 7 days. More than 50% of the patients were discharged home using a cane, and by 4 weeks postoperatively, all patients were ambulating with a cane.

Statistical Analysis

Data were collected and entered into a computerized database. Data were checked and descriptive statistics and bivariate analysis were done using the Kwikstat statistical software package (TexaSoft, Cedar Hill, TX). Then the dataset was exported to SPSS (version 10, SPSS Inc, Chicago, IL) and the driving reaction time was analyzed using analysis of variance (ANOVA). The data were analyzed by repeated measures ANOVA which dropped cases with missing data and by random effects ANOVA which included all cases.

RESULTS

One hundred patients were recruited. Ninety were tested preoperatively. Patients did not always complete all five tests. Those who omitted the 1-week test tended to have a slower reaction time preoperatively (0.6 versus 0.56). The difference in the drop-out rate between the genders was not statistically significant (p = 0.09).

Fifty-five patients (35 men, 64%; and 20 women, 59%) completed the test 1 year later. All patients were scheduled to return for their 1-year followup. Of the 35 patients who did not complete the driving reaction time test at 52 weeks, all were contacted by telephone. Nine relocated out of state after their surgery; 11 lived out of state and did not wish to be followed up in New York; seven were seen by their physician at 52 weeks but were missed by the therapist for testing; four patients canceled a scheduled visit; one patient was hospitalized after abdominal surgery, and three were lost to followup.

Driving Patterns

Before surgery, 90% of the subjects drove four or more times a week. Seven percent drove once a week. Three percent drove once a month. Five percent of the subjects were city drivers. Forty-five percent of the subjects drove in the suburbs. Fifty percent drove in the city and the suburbs. Chi square analysis comparing subjects with right and left total hip arthroplasties with location of driving (city versus suburbs versus both) was done. The difference was not statistically significant.

Changes in Driving Reaction Time

The mean driving reaction time for all subjects increased (or worsened) 1 week postoperatively compared with before surgery. At 4 to 6 weeks (p < 0.001), 26 weeks (p < 0.001), and 1 year (p < 0.001) postoperatively, the mean driving reaction time for all subjects was decreased significantly (or improved) compared with their preoperative driving reaction time score (Table 3).

TABLE 3: Driving Reaction Time—Gender and Side Involved

The person with the slowest driving reaction time at 1 week postoperatively had an average of 2.29 seconds for 10 test runs, where the times ranged from 1.69 to 3.52 seconds. After dropping the highest and lowest scores, the average was 2.21 seconds. The person with the fastest driving reaction time at 1 week postoperatively had an average of .43 seconds for 10 test runs, where the times ranged from .31 to .67 seconds. After dropping the highest and lowest scores, the average was .42 seconds. These patients had a right total hip arthroplasty.

The driving reaction time mean scores were compared with the results obtained by the American Automobile Association using a percentile rank score. By percentile rank score, men improved from the fifty-fifth percentile (0.53 seconds) preoperatively to the sixty-seventh percentile (0.48 seconds) at 1 year postoperatively. Women improved from the forty-fifth percentile (0.59 seconds) to the eighty-fifth percentile (0.47 seconds). Although the average male patient who had a total hip arthroplasty had driving reaction time scores that were faster than their female counterparts preoperatively, at 1 year postoperatively women had scores that were slightly faster than the scores of the men.

The influence of gender or side on driving reaction time was estimated using repeated measures ANOVA. Neither gender (p = 0.6), side (p = 0.4), nor an interaction term (p = 0.4) had a statistically significant effect on the driving reaction time.

Preoperatively, the driving reaction time was slower for women than for men (p = 0.034). The difference was not statistically significant at any other time. The mean driving reaction time preoperatively for patients with right or left total hip arthroplasty was 0.56 seconds. The change in driving reaction time from preoperative driving reaction time to 1 week was not statistically significant. The change in driving reaction time between preoperative and each of the other times was statistically significant (p < 0.001). This also was true for each gender subgroup. The within subject change had a statistically significant linear trend, but the higher order components were not statistically significant. For all patients, the change between 6 weeks and 1 year was statistically significant (p = 0.006), but the change was not statistically significant between 26 weeks and 1 year. For women, the change between 6 weeks and 1 year postoperatively, and between 26 weeks and 1 year postoperatively, were statistically significant (p ≤ 0.001). The mean driving reaction time score 1 year after surgery was 0.08 seconds faster than before surgery for patients who had a left total hip arthroplasty, and 0.09 seconds faster for patients who had a right total hip arthroplasty. The effect of time on driving reaction time was statistically significant for all patients and for men and women separately (p < 0.001).

DISCUSSION

Reaction time is considered a complex task. ^8,10,13 Reaction time is defined as the amount of time an individual takes to respond and complete a movement after a stimulus has been presented. This task includes different psychomotor processes. ^8,10,13 The effect of aging on reaction time is controversial in that researchers do not agree on whether it is response time or the effector movement time which is impaired. ^11,12 For the current study, driving reaction time was defined as the amount of time it takes after the red light stimulus appears on the driving simulator for the subject to place his or her right foot on the brake pedal. This is inclusive of the response time and effector movement time.

The normative data published by the traffic safety division of the American Automobile Association represent driving reaction time by percentile rank 0 to 100. They are subdivided into five categories of the total population according to percentile rank score by gender (A–E), (Table 2). For example, for men the ninetieth to one-hundredth percentile represents 0 to 0.34 seconds, (A category); seventy-fifth to eighty-ninth percentile represents 0.341 to 0.36 seconds, (B category); twenty-fifth to seventy-fourth percentile represents 0.361 to 0.430 seconds, (C category); tenth to twenty-fourth percentile represents 0.431 to 0.58 seconds, (D category); and zero to ninth percentile represents 0.581 seconds or greater, (E category). The female patients who had total hip arthroplasty had a mean preoperative score of 0.59 seconds. The 1-week postoperative driving reaction time score was in the E category or the lowest American Automobile Association category which was the zero to ninth percentile. These same patients at 4 to 6 weeks, 26 weeks, and 52 weeks postoperatively were in the American Automobile Association category D which was the tenth to twenty-fifth percentile.

Davies and Watts ^2,3 and Sivak et al ¹² investigated the effects that ergonomic brake and accelerator pedal placement had on driving reaction time. The effect of prescription and illegal drugs on driving reaction time and the effect that numerous neurologic diseases had on driving reaction time also were studied. There has been only one study in the literature, done by MacDonald and Owen ⁷ in 1988, which addressed total hip arthroplasty and driving reaction time. There were several limitations in their study. Their sample size was 25 subjects. The subjects were tested twice postoperatively, at 8 weeks, and again at 32 weeks. There was no control for diagnosis, type of gearshift, type of anesthesia, type of prosthesis used, or postoperative physical therapy protocol.

There are no state or national regulations which address delayed driving reaction time. There are no accepted values reported by the American Automobile Association or the department of motor vehicles regarding what is a safe or optimal score on the driving reaction time test. There are no state or national guidelines to assist the healthcare provider in determining what driving reaction time score should be used to advise patients when they may or may not resume driving after a total hip arthroplasty. The mean preoperative driving reaction time for all subjects was 0.56 seconds. At 4 to 6 weeks postoperatively the mean driving reaction time for all subjects was 0.50. Because this score was significantly lower (faster) than the preoperative score, advice to patients based on driving reaction time would be to allow patients to resume driving 4 to 6 weeks postoperatively after total hip arthroplasty.

The current study differed from that of MacDonald and Owen ⁷ in that the study by MacDonald and Owen was done in England where there is a predominance of cars with clutches in which drivers use two feet to drive. In their study there was no control for the type of car driven whereas in the current study all subjects drove a car with an automatic transmission where the right foot is used for the gas and brake. The current subjects were tested at serial points postoperatively whereas in the study of MacDonald and Owen, the subjects were tested only twice postoperatively, the first driving reaction time test postoperatively was done at 8 weeks, and the patients were tested again at 32 weeks.

The current results indicate that at 1 week postoperatively patients who had a left total hip arthroplasty had a faster reaction time than they had preoperatively. From the standpoint of solely reaction time, patients who had a left total hip arthroplasty could resume driving 1 week postoperatively if medically allowed to do so. For patients who had a right total hip arthroplasty, there was an improvement at 4 to 6 weeks. The mean driving reaction time score at 4 to 6 weeks for patients who had a right total hip arthroplasty was 0.50 seconds, which is 0.06 seconds faster than their preoperative score. The 0.50 seconds score represents the sixty-third percentile rank score for national norms for men and the seventy-fifth percentile rank score for women. It seems prudent to recommend that patients who have a right total hip arthroplasty wait 4 to 6 weeks after surgery before resuming driving. It is not uncommon for physicians to allow patients who have a left total hip arthroplasty to resume driving sooner than patients who have a right total hip arthroplasty.

Initially it was not expected that there would be any change in the driving reaction time of patients who had a left total hip arthroplasty. All patients who had a left total hip arthroplasty in the current study asked why they were being tested when they drive with their right foot and were having a left total hip arthroplasty. When the movements of the left hip and pelvis were analyzed during subject testing, it was observed that when driving a car with an automatic gear shift, although one uses only their right lower extremity for the gas and brake pedals, when the right foot is moved on and off the gas and brake pedals weight almost always is shifted to the left hip and pelvis. Although a visual analog scale was not used to assess pain, 85% of the patients who had a left total hip arthroplasty reported increased pain in the left hip when driving before surgery.

It is necessary to discuss with the patient, the length of time the patient will be required to adhere to total hip arthroplasty precautions, the type of automobile the patient is driving, and whether it is possible to maintain total hip arthroplasty precautions while entering and exiting the car. The standard precautions for patients having total hip arthroplasty to avoid hip dislocation are to avoid extremes of hip flexion, adduction, and internal rotation. It may not be advisable to allow patients who drive sports cars or cars in which their knee is higher than their hip when seated in the drivers seat, and have a left total hip arthroplasty, to drive 1 week postoperatively, because of the inability to maintain total hip arthroplasty precautions while sitting in the drivers seat, and when entering and exiting the vehicle. If the same patient drove a truck, a van, or sport utility vehicle, and was able to adhere to total hip arthroplasty precautions while entering and exiting the vehicle, resumption of driving is possible. An open dialog with the surgeon can help the patient decide when to resume driving.

The mean preoperative driving reaction time score for all subjects was 0.56 seconds, which is in the fifty-fifth percentile for men and forty-fifth percentile for women. Before surgery most patients had an impairment in psychomotor performance with a delayed driving reaction time score. The driving simulator obtained from American Automobile Association is inexpensive, portable, and easy to operate and takes approximately 3 to 4 minutes to do this driving reaction time functional assessment. This is a simple model simulator, which solely measures reaction time. For patients having lower extremity procedures, one can determine the patient’s reaction time and compare it with the national average, and recommendations can be made regarding driving strategies.

What does this mean in practical terms? If a car is traveling 55 miles per hour, the car moves 80 feet per second. One tenth of a second in driving reaction time would mean that the car must travel 8 feet farther before coming to a complete stop. One tenth of a second may mean the difference between a fatal motor vehicle accident and the ability to stop in time. Functional assessment of driving skills using a driving simulator may assist the healthcare professional to determine the proper time when patients with a musculoskeletal impairment may resume driving. Driving strategies include decreasing driving speed and a longer distance between cars in front and behind. It is critical for physicians and other healthcare providers to be aware that a slower than normal response time will affect a patient’s ability to brake when driving. In making the determination as to when patients may resume driving after a total hip arthroplasty it is important to consider the total hip arthroplasty precaution time frame and the postoperative driving reaction time.

Limitations of the Study and Recommendations for Additional Studies

Hip pain and muscle strength were not addressed in this study. A visual analog scale was not used to measure hip pain, and no objective measurements of strength were used to assess hip strength. All patients were able to do the driving reaction time tests, and had enough strength and range of motion (ROM) in their right lower extremity to depress the accelerator and brake. Recommendations for additional studies would include objectively measuring strength, pain, and hip ROM, and determine how physical therapy intervention addressing impairments affect driving reaction time.

The apparatus is not necessarily the perfect reproduction for all automobile set-ups. The placement of gas and brake pedals in vehicles varies and some people may have found the situation less realistic than other participants. The lack of the appropriate noise, the seat, the lack of the usual steering column and controls, were differences. Some people also had more trouble simulating an emergency they did not see and were wary of breaking the apparatus. Therefore the simulation is not a perfect test. There also are legal implications not addressed in this study. A person who wishes to drive must consider what a court might think, if he or she were involved in an accident 4 weeks after a total hip arthroplasty.

The current data apply only to the population studied, the age groups, and unilateral total hip arthroplasty. Other populations may have different results.

All patients who had a total hip arthroplasty had a delayed reaction time preoperatively compared with healthy subjects. In the current study driving reaction time, when measured by right foot braking time, is affected to a large extent by osteoarthritis of the right hip. Most patients who have had a right total hip arthroplasty have recovered sufficient hip function to return to driving 4 to 6 weeks postoperatively. Some patients who have had a left total hip arthroplasty have recovered sufficient hip function to return to driving 1 week postoperatively, however adhering to total hip arthroplasty precautions is imperative. Advice regarding when to resume driving after total hip arthroplasty should be determined by side of surgery, time for adherence to total hip arthroplasty precautions, and comparison between preoperative and postoperative driving reaction times.

Acknowledgments

The authors thank Robin Benick, Emily Peterson, Linda Murray, and Jimmy Dempsey for invaluable assistance.