The Six-Minute Walk Test Predicts Ambulation After TKA
Prior to undertaking the definitive study, a pilot survey of patients awaiting knee and hip replacement was performed to determine a time parameter for an extended walk test which may be functionally relevant for this patient group. Sixty-four consecutive individuals [n = 47 awaiting TKA, mean age 68.3 (SD 10.7) years, 77% female; n = 17 awaiting total hip arthroplasty, mean age 61.5 (SD 10.9), 71% female] attending the pre-operative joint replacement education class at a public hospital completed a written survey comprising close-ended questions about current and expected 1-year walk times. At the time of the survey, 52% of participants reported a maximum walking duration of 10 minutes or less before they had to stop due to their leg symptoms. As anticipated, participants expected to be able to walk much further 1-year post surgery; 86% of participants expected to be able to walk 25 minutes non-stop, and over 50% expected to be able to walk 50 minutes non-stop one year post-surgery. From these data, we reasoned that a 30-minute walk (30MW) test was in accordance with patient expectations for functional mobility one year after TKA.
A convenience sample of a minimum of 60 subjects (30 TKA recipients and 30 age-matched healthy controls) was planned for this study. The sample size was a compromise between gathering sufficient data for repeatability and cross-sectional analyses based on what is known from previous timed mobility trials in TKA and elderly populations (n=22 to 51), and the need to test within the cooler, drier months (Autumn, Winter) to reduce environmental confounders.
For the TKA cohort, patients who underwent TKA 12 to 18 months prior were identified from an existing hospital database and a random sample was selected using a computer-generated sequence. The selected individuals were contacted and screened via a telephone interview. Eligible patients were invited to participate. To improve generalizability, patients with other co-morbidities and other musculoskeletal pain were not excluded, as patients who undergo TKA commonly present with multiple co-morbidities and other joint disease. However, the following exclusion criteria were used to ensure patient safety during walk test performance: unstable angina, myocardial infarction, or cardiopulmonary, spinal or lower limb surgery in the previous six months; a recent history of uncontrolled hypertension; and inability to comprehend English.
Healthy control volunteers were recruited from the community through flyers placed in hospital grounds and through word-of-mouth. To allow comparison with a genuinely healthy population of this age group, volunteers were excluded if they had acute or chronic respiratory, cardiac, neurological and musculoskeletal disorders involving significant mobility impairment. Eligibility was determined via telephone interview.
All participants who provided written, informed consent were enrolled into the trial. The protocol and consent forms were reviewed and approved by the South Western Sydney Local Health District Human Research Ethics Committee. The characteristics of the participants including age, gender. height, and body mass index (BMI) are summarized in Table 1.
Testing took place at three outdoor sites for the convenience of the participants. An outdoor setting was used to simulate normal functional walking conditions. Well demarcated, level, paved, concrete or bitumen footpaths devoid of obstacles were chosen. Participants were tested under dry conditions with ambient temperature between 13 to 24 degrees Celsius. Time of testing was constant for each participant. Participants were instructed to wear supportive footwear and comfortable clothing.
Resting heart rate (HR) was measured (Polar F4 or F6, Polar Electro Oy, Finland) at the commencement of the testing session. As with other studies evaluating 6MW tests, HR was monitored as an indicator of participant effort. The average and maximum HRs achieved, together with the time spent with HR greater than 65% of age-predicted HR maximum, were recorded. 65% of age-predicted HR maximum was used to indicate whether a moderate level of exercise intensity had been achieved.
Four assessors used standardized instructions to administer the tests. All subjects performed walk tests in the same order. The TUG and 6MW tests were conducted first. A seated rest of at least 10 minutes was given after the 6MW test. On restoration of resting heart rate, the participant performed the 30MW test.
TUG The procedures for the TUG test was based on published protocols. A standard height chair of 45cm with arm rests was placed on an outdoor level footpath, and a line 3 metres from the chair was drawn. The participant was instructed to stand up from the chair, walk to the line, turn around, walk back to the chair and sit down as quickly and safely as possible. The assessor commenced timing as the participant leaned forward to stand up, and ceased when the participant's hips made contact with the seat to sit down. A minimum of two tests was performed for each participant with the best time used in the analysis. High repeatability of the TUG has been established in patients awaiting TKA (ICC 0.75).
6MW Test The procedure for the 6MW test was based on published guidelines. A 25 metre section of the outdoor footpath was demarcated for this test. The participant was instructed to walk as far as possible for six minutes, up and down the demarcated footpath, pivoting to turn at the end of each lap. Timing commenced as the participant stepped over the start line. Standardized encouragement was given to the patient after each minute. The participant was instructed to stop at six minutes, or prior if they were unable to complete six minutes, and to maintain their position whilst the assessor measured the final partial lap with a trundle wheel. The use of a walking aid and standing rests were permitted. One test was performed for each participant as we did not want fatigue from completing two 6MW tests to undermine performance in the longer test. High repeatability of the 6MW test has been established in patients awaiting TKA (ICC 0.94).
30MW Test For the 30MW test, a 200 m section of the footpath was demarcated. A 200 m lap was chosen as it kept the frequency of turns low whilst keeping the participant in full view of the investigator at all times. The curvature of the path varied slightly between sites. Four subjects (three not included in the study) performed reliability tests of the 30MW test between the sites. Subjects varied within 5% at each site and this was considered an acceptable level of variation given the need to be flexible with the location of the testing site.
The participant was instructed to walk as far as possible and safely for 30 minutes along the 200 m track, pivoting to turn at the end of each lap. Participants were permitted to rest in standing at any time if required and to resume walking as soon as able. Encouragement was given by the assessor after every lap. The assessor commenced timing as the participant stepped over the starting line and ceased at 30 minutes. At this point, the participant was instructed to maintain their position on the track while the assessor measured the distance of the partial lap with a trundle wheel. Testing was terminated if the participant reported that they were unable to continue and the distance covered in the abbreviated test was recorded as the distance covered in 30 minutes. To evaluate gait characteristics, the assessor recorded the number of laps walked, the number of steps and time taken during one lap in the initial five minutes and again during the last five minutes of the test, and the number and duration of any rests taken, noting the reasons for resting. From this, average cadence, average step length and total 30MW distance were calculated. To measure test-retest reliability of the 30MW test, participants repeated the test within two weeks of the first test using the same track.
Western Ontario and McMaster Universities Osteoarthritis Index VA3.1 (WOMAC) The WOMAC index was used to measure patient-reported pain and function. The pain and function subscales of the WOMAC index were scored using a visual analogue scale on a 10cm line for each question and a total was calculated for each subscale (worst score 50 for pain, 170 for function).
Physical Activity Questionnaire The questions on physical activity from the National Physical Activity Survey, were used as the data provide a contemporary population-based comparison for our cohort. The total number of sessions and total duration of physical activity was calculated by summating reported sessions and duration of walking over 10 minutes, and moderate and vigorous activity.
Descriptive statistics were calculated for continuous variables, and distribution was examined. Independent t-tests and Mann–Whitney U tests were used to compare subject characteristics and walking test variables as appropriate. Chi square test was used to compare difference in participants requiring rests between the groups. Paired t-tests were used compare the difference in average walking speed between 6MW and 30MW tests. Repeatability of the 30MW test was measured by calculating the intra-class coefficient (ICC) (two way random, absolute agreement). Correlation coefficients (Pearson's or Spearman's rank as appropriate) were computed between pairs of walk tests, and multiple linear regressions were used to determine the predictors of the 30MW test in the TKA cohort, respectively. Multiple linear regression modeling was used to establish predictive models for WOMAC function score and total time of physical activity in the TKA cohort. To compare the predictive value of the three walk tests, three models were generated for each dependent variable using gender, body mass index (BMI), and one of the walk tests as independent variables. Each model was limited to three independent variables due to the sample size. These variables were determined a priori, and were selected based on current literature. Appropriate regression diagnostics were used in modeling. For all analyses, P < 0.05 was deemed as significant. Data from the first testing session of the 30MW test were used in the regression analyses.