Thorough assessment of pelvic alignment includes measurement of positioning in the sagittal and frontal planes. The angle of orientation in the sagittal plane is commonly referred to as pelvic tilt.1 Anterior pelvic tilt has been linked to contributing to the etiology of common musculoskeletal pathologies such as patellofemoral pain syndrome,2–4 lumbopelvic dysfunction,5 and anterior cruciate ligament deficiencies.6,7 The frontal plane is measured relative to the iliac crest height and is often referred to as leg length discrepancy8,9 or pelvic height difference.10 Traditionally, the sagittal and frontal plane alignment of the pelvis has been evaluated by measures via radiography10–12 or by simple visual palpation and estimation.13,14 Although advanced imaging procedures may be the most accurate measure, clinicians are often limited in their time and resources, making imaging a less practical form of assessment.10
In the clinical setting, visual observations and palpation are commonly used as an assessment for pelvic alignment. A caliper–inclinometer, the palpation meter (PALM; Performance Attainment Associates, Lind-strom, Minnesota), has been used to obtain objective numeric measurements of both pelvic tilt and pelvic height difference.8–10,15 Previous research has shown the assessment of static frontal and sagittal plane pelvic alignment to be reliable using the PALM.8,10,16 In addition to assessing the reliability of the PALM, Petrone et al10 compared measurements of the PALM to standing anteroposterior radiographs to determine their validity. In the assessment of pelvic tilt, the PALM was determined to have excellent validity through intraclass correlation coefficients (ICC = 0.90, 0.92), whereas less accurate validity was shown in the assessment of pelvic height difference (ICC = 0.76, 0.78).10 Although the PALM has been shown to be a reliable and valid measure of pelvic tilt and pelvic height difference, it still has associated clinical limitations.
Clinicians may not have access to the PALM and therefore must still be able to adequately examine pelvic position without the aid of a measuring device or radiography. In these situations, clinicians may need to rely solely on a visual judgment of alignment for their assessment. Previous research on the efficacy of visual ratings and estimations in place of objective measurements for range of motion and static alignment of the pelvis, hip, and knee have shown mixed results and warrant further investigation.14,17–19 Accuracy of visual ratings are said to be affected by clinician experience, the use of preestablished range of motion estimates, and body composition.14,17,18 We aimed to investigate the reliability of the PALM device, the reliability of visual observations by a novice clinician, and the relationship between the 2 methods.
Twenty healthy individuals (6 men, 14 women; age = 22.75 ± 1.59 years; height = 171.44 ± 11.59 cm; mass = 73.49 ± 16.37 kg) were recruited to participate. Individuals were excluded if they reported a history of prior lower extremity or back surgery or systemic or neurological disease or if they possessed abnormal neurological status. This study was approved by the university’s institutional review board, and all participants gave their informed consent prior to data collection.
Three athletic trainers (2 examiners, 1 recorder) with less than 2 years of clinical athletic training experience performed all clinical measures and observations. The recorder was used to reduce examiner bias. All of the athletic trainers completed two 2-hour training sessions conducted by an athletic training clinician researcher (D.A.H.) with 8 years of clinical experience prior to testing. A manual of written instructions and photographic demonstrations of each measure was provided to each examiner.
Pelvic tilt and height difference were measured with the PALM, which consisted of a bubble inclinometer and 2 depth caliper arms. The central display provides the readings of the distance between the caliper arms in centimeters and the bubble inclinometer in degrees. The PALM was suspended from the examiner’s neck with the adjustable cord and held level in front of the examiner. The caliper arms were placed on 2 bony landmarks of the participant, and the distance between the landmarks and the degree of inclination was measured. In addition to the raw values, the PALM is equipped with a slide rule calculator that uses the 2 values to provide a composite measure of height difference between the landmarks.
On arrival, the participants read and signed the informed consent form, and demographic information consisting of age, height (cm), mass (self-reported in kg), true and functional leg length (cm), and activity level were recorded. For continuity, all measures were taken on the dominant limb, which was determined by asking the participants which leg is used to kick a ball for greatest distance. Participants were randomly assigned to 1 of the 2 examiners through a randomization software program.
A foot template was created for each participant to standardize foot placement between examiners. While standing on a laminate tile floor, participants took 10 marching steps in place and then assumed a shoulder-width stance with feet facing forward on the template sheet made of laminated paper. The examiner instructed participants to stand as close to their normal standing posture as possible. Once in place, the examiner traced the outline of their feet onto the template with a dry-erase marker. Participants were asked to remain standing on their template with their arms crossed over the chest as the measurements were taken. Participants stood on the same surface for all testing, and the foot template was used for both examiners throughout the testing session.
Anterior-Posterior Tilt. The examiner stood to the side of the participant’s dominant leg and identified the anterior and posterior superior iliac spines (ASIS and PSIS, respectively) through palpation and marked them with a marking pen. The examiner was instructed to use the most prominent aspects of the landmarks as the marking point. The examiner placed the index finger of each hand on the respective landmark and assumed a sagittal view of the participant’s pelvis (Figure 1). The examiner moved to be at eye level with the landmarks and used the dominant eye to initially make a visual judgment on the direction of pelvic tilt by comparing the levels of the ASIS and PSIS and decided whether the participant’s pelvic position was in 1 of 3 categories: posterior tilt, neutral, or excessive anterior tilt. This visual judgment was reported to the recorder.
Figure 1. Visual observation of anterior–posterior tilt.
The examiner then obtained an objective measure of the direction and magnitude of pelvic tilt by placing the palpation tips of the PALM on the marked points of the ASIS and PSIS (Figure 2). After the examiner stated that the measurement was complete, the recorder recorded the measurement. Each examiner performed 3 trials with the PALM, and data were recorded to the nearest degree. Negative values were assigned to a posteriorly directed inclination to distinguish the data.
Figure 2. Measurement of anterior–posterior tilt with the palpation meter.
Pelvic Height Difference. The examiner stood behind the participant and identified the most central superior portions of the left and right iliac crests through palpation. The examiner then rested her hands firmly on the iliac crests and assumed a frontal plane view of the participant’s pelvis. The examiner noted the position of her hands on the iliac crests in relation to one another and categorized the participants into 1 of 3 categories: left tilt (left side lower), right tilt (right side lower), or neutral (Figure 3). This judgment was reported to the recorder.
Figure 3. Visual observation of pelvic height difference.
The examiner then obtained an objective measure of the pelvic obliquity and magnitude of pelvic height difference by placing the palpation tips of the PALM on the respective portions of the iliac crests (Figure 4). After the examiner stated that the measurement was complete, the recorder recorded the value. The examiner performed 3 trials with the PALM, and obliquity was measured to the nearest degree, whereas pelvic height difference was measured to the nearest 10th of a centimeter. Negative values were assigned to a left tilt to distinguish the data.
Figure 4. Measurement of pelvic height difference with the palpation meter.
After the 1 examiner completed both observations and PALM measurements for anterior-posterior tilt and pelvic height difference, the marks were removed from the skin using an alcohol pad. Once clean, participants took their foot templates and moved to the second examiner. The test session was complete after both examiners rated and measured each participant. Participants returned for a second identical test session with a minimum of 24 hours between the 2 sessions, and the order of examiners was reversed for randomization. Participants were instructed not to change their daily activities in between the 2 testing sessions, and they were encouraged to refrain from additional therapy or stretching.
All data were analyzed using the SPSS version 16.0 statistical package (IBM Corp, Armonk, New York). Two models of ICCs were used to evaluate intra-examiner, inter-examiner, and intersession reliability. An ICC (3,1) was calculated using 3 trials performed by each examiner during the first session for intra-examiner reliability and the means of the 3 trials across each test session for intersession reliability. This model allows for a focused examination of the 2 novice examiners used in this study by viewing them as a fixed effect, while participants are the only random factor of variance.20 An ICC (2,k) was calculated using the mean of 3 trials performed by the 2 examiners in the first test session for inter-examiner reliability. This model allows the results to be generalized to examiners outside of those used in this study by viewing both the participants and examiners as random factors of variance.20 Reliability of the categorical data was analyzed using percent agreement, defined as the number of agreements divided by the total number of possible agreements.21 Objective data were categorized based on normalized values and were compared with the observational categories. Each observed category was assigned a range of values to group the raw data for comparison based on normative values established by others that used the PALM or comparable methods.2,8,10,22–26 Normalized categories for the objective data of pelvic tilt consisted of posterior tilt (< 0°), neutral (0° to 10°), and anterior tilt (> 10°). Normalized categories for the objective data of pelvic height difference consisted of: left tilt (less than −11 mm), neutral (−11 to 11 mm), right tilt (greater than 11 mm).
Portney and Watkins27 suggested that ICC values from 0.25 to 0.50 indicate a fair relationship, with values from 0.50 to 0.75 indicating a moderate to good relationship and values of 0.75 or higher indicating a good to excellent relationship. The ICC (3,1) for the intra-examiner reliability objective measure of pelvic tilt was excellent, and the measure for pelvic height difference was good, with examiner 1 demonstrating slightly higher but not statistically significant different reliability when compared with examiner 2 (Table 1). Inter-examiner reliability (ICC 2,k) for pelvic height difference (0.777) and pelvic tilt (0.823) was good.
Table 1: Reliability of Examiners
Intersession reliability was more variable as evidenced by the wider range of ICC (3,1). The measure of pelvic tilt again demonstrated greater consistency in measurement with good reliability coefficients for examiner 1 (0.779) and examiner 2 (0.801), respectively. The intersession reliability measure of pelvic height difference produced fair (0.370) and moderate (0.530) reliability for examiner 1 and examiner 2, respectively.
Agreement of the observations within examiners between the test sessions ranged from 55% to 75% for pelvic tilt. Observations of pelvic height difference demonstrated a greater number of agreements ranging from 70% to 80% (Table 2).
Table 2: Intersession Reliability of Observations
Reliability Between Observations and Normative Value Categories
For the first test session, the visual observation category of each examiner was compared with normative categories from the raw data of the PALM measurement. Examiner 1 had a lower percent agreement (65%) than examiner 2 (80%) for the measure of pelvic tilt, indicating that the visual observation did not match the normative categories as often. For the measure of pelvic height difference, there was less discrepancy between examiners, as they ranged from 90% to 95% agreement, indicating that the normative categories and visual observations were similar between both examiners (Table 3). The frequency counts for the visual observation and normative category placement, as well as the means for the raw data, are reported in Table 4.
Table 3: Reliability Between Observation and Normative Categories
Table 4: Frequencies and Means of Observations and Normative Category Placement
Our examiner reliability for the measurement of pelvic tilt using the PALM met or exceeded those previously reported by others.8,16 Shultz et al16 reported maximum intraclass correlations of 0.98 for intra-examiner and 0.68 inter-examiner reliability for previously trained examiners when assessing pelvic angle using the PALM, whereas Hagins et al8 reported ICC values of 0.98 and 0.89 for intra-examiner and inter-examiner reliability. Differences in examiner training and experience could provide explanation for the higher values from those studies when compared with novice clinicians in the current study. The measures of pelvic height difference using the PALM were in the lower range of values previously reported by others.10,16 Our good intra-examiner reliability coefficients (0.82, 0.83) closely match those reported by Hagins et al8 (0.84) but fall below the high coefficients reported by Petrone et al10 (0.94 to 0.99). Inversely, the inter-examiner reliability coefficients for pelvic height difference met or exceeded previously reported values of 0.6515 and 0.70 to 0.97.10 Our results show that relatively novice clinicians can use the PALM to assess pelvic alignment in a consistent manner within a test session.
When the measurements were compared across 2 test sessions, a sizeable discrepancy was found between the ICC for both pelvic tilt and pelvic height difference. Intersession reliability for the measure of pelvic tilt was good, but it was only fair to moderate for the measure of pelvic height difference. The ASIS and PSIS are easily palpable landmarks with a small, focused area, thus leading to less variability in measurement of pelvic tilt. In contrast, when assessing pelvic height difference, the iliac crests are more difficult to palpate due the amount of adipose and soft tissue. Our participants had variable body composition, which may have resulted in interference in the most accurate palpation of the iliac crests and led to inconsistency of the measurement. However, Mann et al14 found that body composition had no effect on the reliability of this palpation method. It is possible that the examiners were better at the visual estimation of pelvic height difference and that objective measure between sessions is not as accurate.
Although we demonstrated that the PALM was a reliable patient assessment device, clinicians may not have one available for their use and will most likely perform a visual assessment of their patient. Therefore, our second objective was to assess the reliability of a visual estimation technique for pelvic alignment. When the observations for pelvic alignment were compared across 2 test sessions, percent agreements were only fair to moderate, with those of examiner 2 rating slightly better. Our results do not improve on the fair to moderate reliability of this method reported by Mann et al.14 In addition, Mann et al14 analyzed the correlation between number of agreements and experience level of the tester and found none to be present. Their findings support our ability to generalize the results of the novice examiners used in our study.
Our third objective was to compare the observational rating with the raw data obtained from the PALM to determine whether the observation could correctly categorize each participant. Previous studies comparing clinical measurements to visual observations or estimates for range of motion have shown mixed results in the reliability and accuracy in those judgments.17,18,28 Reliability of observational ratings seems to improve when a numerical estimate is involved as opposed to placing the participant into a category.17,28 Our reliability between the observational rating and normative value category was moderate for pelvic tilt and very high for pelvic height difference. These findings support the hypothesis that the observation made by a clinician will likely match a range of objectively measured values and allow for interchangeable use.
The frequency counts (Table 4) demonstrate the dispersal of our participants across the categories. Participants demonstrated a greater degree of variability for pelvic tilt than pelvic height difference, but most participants were in the neutral category for both conditions. Several hypotheses can explain these observed findings. The measurement technique may have been a factor, given that the palpation landmarks for the measure of pelvic tilt are easily located and may have allowed for more precise measurements and judgments, accurately placing those participants in the categories outside of neutral. Although Mann et al14 did not find a statistical correlation between body composition and examiner agreement, they do acknowledge that participants reported that not all examiners palpated the same area, and thus could have affected the measurements in this study as well. We used a healthy, asymptomatic group of participants; therefore, it is expected to see the majority of our participants fall into the defined neutral categories. It is plausible that examiners were unable to detect a marked difference for pelvic height either due to soft tissue interference or the normalcy of our participants. The lack of variation among the participants could account for the greater percent agreements seen with pelvic height difference than pelvic tilt.
Conclusion and Limitations
Our study had some limitations that could have affected the study results. Although percent agreement is easily interpreted at face value, the kappa statistic is a stronger reliability model for categorical data because it is corrected for chance and allows for stronger interpretation of results.21 Given that our observational data were not distributed in all 3 categories for each examiner, we could not use the kappa statistic, which presented a statistical limitation in our analysis. In addition, examiners in our study were not instructed to make an estimation of the degree of pelvic alignment as part of their observation, and it is unknown whether that procedure would have affected our results. Because our observations consistently matched the proposed normative value ranges for each category, it is unlikely that the addition of estimation is necessary. Given the fact that our participants were asymptomatic and most were placed in the neutral category, it is possible that the marked difference the examiner looked for to rate outside of the neutral category may exist in a symptomatic population and may not require the distinct measurement to be identified. Future research in the assessment of pelvic alignment should include a symptomatic population to determine whether malalignment is present and how objective visual measures compare to normative values.
Implications for Clinical Practice
Although the PALM is relatively inexpensive and a moderately to highly reliable tool to assess pelvic tilt and pelvic height difference, clinicians may not have access to this device. Currently, clinicians often rely on a quick visual estimate rather than an objective measure such as the PALM to determine pelvic alignment when evaluating the lower extremity. On the basis of our results of comparing visual observations to measurements obtained by the PALM, clinicians can adequately assess patients visually and may not need to take the distinct PALM measurement. Patient treatment decisions are not typically based on the degree of pelvic tilt or objective difference in pelvic height but instead are determined by a directional category. If clinicians do not have the resources to obtain a PALM device, they can continue to assess their patients visually and be confident that they are obtaining adequate information to treat them.
- Preece SJ, Willan P, Nester CJ, Graham-Smith P, Herrington L, Bowker P. Variation in pelvic morphology may prevent the identification of anterior pelvic tilt. J Man Manip Ther. 2008;16(2):113–117 doi:10.1179/106698108790818459 [CrossRef] .
- Nguyen A-D, Shultz SJ. Identifying relationships among lower extremity alignment characteristics. J Athl Train. 2009;44(5):511–518 doi:10.4085/1062-6050-44.5.511 [CrossRef] .
- Hruska R. Pelvic stability influences lower-extremity kinematics. Biomechanics. June1998;5:23–29.
- Boling MC, Padua DA, Creighton RA. Concentric and eccentric torque of the hip musculature in individuals with and without patellofemoral pain. J Athl Train. 2009;44(1):7–13 doi:10.4085/1062-6050-44.1.7 [CrossRef] .
- Schache AG, Blanch PD, Murphy AT. Relation of anterior pelvic tilt during running to clinical and kinematic measures of hip extension. Br J Sports Med. 2000;34(4):279–283 doi:10.1136/bjsm.34.4.279 [CrossRef] .
- Hertel J, Dorfman JH, Brahman RA. Lower extremity malalignments and anterior cruciate ligament injury history. J Sport Sci Med. 2004;3(4):220–225.
- Louden JK, Jenkins W, Louden KL. The relationship between static posture and ACL injury in female athletes. J Orthop Sports Phys Ther. 1996;24(2):91–97.
- Hagins M, Brown M, Cook C, et al. Intratester and intertester reliability of the palpation meter (PALM) in measuring pelvic position. J Man Manip Ther. 1998;6(3):130–136.
- Krawiec CJ, Denegar CR, Hertel J, Salvaterra GF, Buckley WE. Static innominate asymmetry and leg length discrepancy in asymptomatic collegiate athletes. Man Ther. 2003;8(4):207–213 doi:10.1016/S1356-689X(03)00012-2 [CrossRef] .
- Petrone MR, Guinn J, Reddin A, Sutlive TG, Flynn TW, Garber MP. The accuracy of the palpation meter (PALM) for measuring pelvic crest height difference and leg length discrepancy. J Orthop Sports Phys Ther. 2003;33(6):319–325.
- Mannello DM. Leg length inequality. J Manipulative Physiol Ther. 1992;15(9):576–590.
- Devereaux MD, Lachmann SM. Patello-femoral arthralgia in athletes attending a sports injury clinic. Br J Sports Med. 1984;18(1):18–21 doi:10.1136/bjsm.18.1.18 [CrossRef] .
- Alviso DJ, Dong GT, Lentell GL. Intertester reliability for measuring pelvic tilt in standing. Phys Ther. 1988;68(9):1347–1351.
- Mann M, Glasheen-Wray M, Nyberg R. Therapist agreement for palpation and observation of iliac crest heights. Phys Ther. 1984;64(3):334–338.
- Herrington L. Assessment of the degree of pelvic tilt within a normal asymptomatic population. Man Ther. 2011;16(6):646–648 doi:10.1016/j.math.2011.04.006 [CrossRef] .
- Shultz SJ, Nguyen AD, Windley TC, Kulas AS, Botic TL, Beynnon BD. Intratester and intertester reliability of clinical measures of lower extremity anatomic characteristics: implications for multicenter studies. Clin J Sports Med. 2006;16(2):155–161 doi:10.1097/00042752-200603000-00012 [CrossRef] .
- Watkins MA, Riddle DL, Lamb RL, Personius WJ. Reliability of goniometric measurements and visual estimates of knee range of motion obtained in a clinical setting. Phys Ther. 1991;71(2):90–97.
- Peeler J, Anderson JE. Reliability of the Ely’s test for assessing rectus femoris muscle flexibility and joint range of motion. J Orthop Res. 2008;26(6):793–799 doi:10.1002/jor.20556 [CrossRef] .
- Hanada E, Kirby RL, Mitchell M, Swuste JM. Measuring leg-length discrepancy by the “iliac crest palpation and book correction” method: reliability and validity. Arch Phys Med Rehabil. 2001;82(7):938–942 doi:10.1053/apmr.2001.22622 [CrossRef] .
- Weir JP. Quantifying test-retest reliability using the intraclass correlation coefficient and the SEM. J Strength Cond Res. 2005;19(1):231–240.
- Hunt RJ. Percent agreement, Pearson’s correlation, and kappa as measures of inter-examiner reliability. J Dent Res. 1986;65(2):128–130 doi:10.1177/00220345860650020701 [CrossRef] .
- Blustein SM, D’Amico JC. Limb length discrepancy. Identification, clinical significance, and management. J Am Podiatr Med Assoc. 1985;75(4):200–206.
- Gross MT, Burns CB, Chapman SW, et al. Reliability and validity of rigid lift and pelvic leveling device method in assessing functional leg length inequality. J Orthop Sports Phys Ther. 1998;27(4):285–294.
- Jonson SR, Gross MT. Intraexaminer reliability, interexaminer reliability, and mean values for nine lower extremity skeletal measures in healthy naval midshipmen. J Orthop Sports Phys Ther. 1997;25(4):253–263.
- Medina McKeon JM, Hertel J. Sex differences and representative values for 6 lower extremity alignment measures. J Athl Train. 2009;44(3):249–255 doi:10.4085/1062-6050-44.3.249 [CrossRef] .
- Woerman AL, Binder-Macleod SA. Leg length discrepancy assessment: accuracy and precision in five clinical methods of evaluation. J Orthop Sports Phys Ther. 1984;5(5):230–239.
- Portney LG, Watkins MP. Foundations of Clinical Research: Applications to Practice. 3rd ed. Upper Saddle River, NJ: Pearson Prentice Hall; 2009.
- Holm I, Bolstad B, Lütken T, Ervik A, Røkkum M, Steen H. Reliability of goniometric measurements and visual estimates of hip ROM in patients with osteoarthritis. Physiother Res Int. 2000;5(4):241–248 doi:10.1002/pri.204 [CrossRef] .
Reliability of Examiners
|PELVIC TILT||PELVIC HEIGHT DIFFERENCE|
|Intra-examiner (ICC 3,1)|
| Examiner 1||0.935||0.832|
| Examiner 2||0.923||0.824|
|Inter-examiner (ICC 2,k)||0.823||0.777|
|Intersession (ICC 3,1)|
| Examiner 1||0.779||0.370|
| Examiner 2||0.801||0.530|
Intersession Reliability of Observations
| Examiner 1||55%|
| Examiner 2||75%|
|Pelvic height difference|
| Examiner 1||70%|
| Examiner 2||85%|
Reliability Between Observation and Normative Categories
| Examiner 1||65%|
| Examiner 2||80%|
|Pelvic height difference|
| Examiner 1||90%|
| Examiner 2||95%|
Frequencies and Means of Observations and Normative Category Placement
|PELVIC TILT||PT, < 0°||N, 0° TO −10°||AT, >10°|
| Data category||0||17||3|
| Mean ± SD||4.08 ± 2.41||11.78 ± 1.26|
| Data category||0||15||5|
| Mean ± SD||4.38 ± 2.09||10.73 ± 0.43|
|PELVIC HEIGHT DIFFERENCE||LT, < −11 mm||N, −11 TO 11 mm||RT, >11 mm|
| Data category||1||17||2|
| Mean ± SD||−11.67 ± 0.0||4.68 ± 4.25||14.67 ± 2.83|
| Data category||0||19||1|
| Mean ± SD||1.64 ± 4.85||12.50 ± 0.0|