Ultrasonography to Assess the Efficacy of Osteopathic Manipulative Treatment for Lumbar Spine Asymmetry

Participants

We recruited volunteers older than 18 years from the medical school and local community via fliers, emails, and personal invitation from other participants. All participants were asymptomatic at presentation. All participants received a diagnosis, measurements, and OMT. Individuals were excluded if they demonstrated contraindications to receiving OMT (including lack of somatic dysfunction) or were not able to provide written informed consent.

Palpatory Assessment

To assess lumbar spine somatic dysfunction, osteopathic physicians (J.P., J.C, C.E., K.B., A.O., who are noted in the Acknowledgments section), who were assisted by third-year osteopathic medical students (S.M., J.W., A.K.), palpated the participant's lumbar spine. Bony landmarks and the lumbar vertebrae with the most substantial somatic dysfunction were identified, and a diagnosis was determined and recorded by the assessing physician (J.P., J.C., C.E., K.B., who are noted in the Acknowledgments section) to be used in creating (J.P., J.C, C.E., K.B.) a management plan. The vertebrae was palpated and diagnosed in 3 planes of motion: flexion/extension, rotation, and side-bending.²⁰ This vertebrae was then marked by the physician (J.P., J.C, C.E., K.B.) with a felt-tip pen to be used as a landmark, not a definitive point of measurement. The observers who performed US (J.G. and A.K.) were blinded to the osteopathic palpatory examination results. Lumbar somatic dysfunction could be related to dysfunction in biomechanically related areas. Other biomechanically-related structures, including the sacrum, pelvis, thoracic diaphragm, pelvic diaphragm, psoas, and pubic symphysis, were also assessed and diagnosed as deemed necessary by the treating physician (J.P., J.C, C.E., K.B.).

Grayscale US Assessment

Immediately following diagnosis and immediately after a single OMT session, participants underwent US assessment of their lumbar spine with an Acuson S3000 Ultrasound Scanner (Siemens Medical Solutions). A curvilinear transducer was used to optimize musculoskeletal clarity of transverse processes. Gain and depth were adjusted as needed. All measurements were made with the participant in a prone position using the following settings: scanning frequency, 3.5 MHz; image depth, 7 cm; dynamic range, 60; single-image focus; tissue harmonic imaging. All US imaging of the lumbar spine was performed twice by 2 observers (J.G., who had 30 years of experience performing US, and A.K., who had 2 years of basic training). Each observer took 2 sets of measurements for testing intraobserver variability. The 2 observers performed US separately for testing interobserver variability. The transducer was placed directly over the landmark being measured with only enough applied pressure to achieve satisfactory imaging (Figure 1). The US images were saved, and measurements were performed using the Digital Imaging and Communications in Medicine viewer computer program by multiple observers (J.W., A.K., S.M.) with duplicate measurements made for inter- and intraobserver variability.

Figure 1.

Ultrasonography (US) of the lumbar spine was performed with the participant prone and with the back muscles relaxed. The transducer (white arrow) was directly placed on the skin of lower back over the transverse processes (L1-L5). Markings were made on the skin by the physician to ensure that US images were recorded at the same vertebral level before and after osteopathic manipulative treatment. These marks were guidelines for the US operators and not specific points of measurement.

To assess the rotational position of each vertebral level (L1-L5), the distance from skin to the most posterolateral surface of the transverse processes was measured (R_L and R_R; Figure 2). Distance between adjacent transverse processes (L1-L2) were measured to assess side-bending asymmetries (SB_L and SB_R; Figure 2). Due to the depth of sacral landmarks from the skin, there was difficulty imaging and measuring an accurate distance from L5 to sacrum. For this reason, side-bending measurements were only made for L1 to L4. Rotational and side-bending asymmetry were calculated from these measurements using the following: R_A=|R_L - R_R| and SB_A=|SB_L - SB_R|.

Figure 2.

Grayscale ultrasonography was performed in the low back (L1-L5) region in all participants. Longitudinal static image of transverse processes and intertransverse ligaments were recorded twice by both operators. The anterior-posterior distance (yellow vertical line, 4.6 cm) between the skin and the transverse process, and the distance between adjacent transverse processes (green horizontal lines, 1.70 cm between L3 to L4, and 1.54 cm between L4 and L5), were measured using electronic calipers online and again offline. Sample measurements were taken at L₄ on the left side.

Figure 3.

Box-and-Whisker plots illustrate the difference in spine rotation asymmetry (R_A) before (purple colored box) and after (orange colored box) OMT (P=.01). A significant decrease in R_A after OMT indicates the improvement of asymmetry resulting from an effective OMT technique.

All measurements were repeated after the patient received OMT.

Osteopathic Manipulative Treatment

Following US assessment, participants received a single 15-minute session of OMT to address the previously diagnosed dysfunctions. The techniques used are shown with each participant's corresponding diagnosis in Table 1. The scope of this study did not include evaluating the efficacy of any particular management modality. For this reason, the osteopathic physicians (J.P., J.C, C.E., K.B.) were given instructions to use any modalities they believed would be most effective. OMT was performed with the main goal of improving lumbar asymmetry. Physicians prioritized the management of nonlumbar diagnosis to maximize the improvement of the lumbar dysfunction. Physicians used single or multiple OMT techniques based on the assessment to improve lumbar disorders. Techniques used in this study included: muscle energy technique, myofascial release (MR), balanced ligamentous tension, HVLA, Still technique, articulatory technique, facilitated positional release, and counterstain. The physicians were instructed to treat and reassess the participants.

Statistical Analysis

All variables, including the thickness of the soft tissue and the distance between transverse processes, were expressed by the mean and standard deviation (SD). Asymmetry was calculated for rotation by measuring the soft tissue thickness from the skin to transverse process bilaterally, and for side-bending by comparing the distance between adjacent transverse processes bilaterally. Calculations were made between the left and right sides of the same segment using the following equations, which we adapted from a previous study:¹⁸

Rotational asymmetry = | R_L−R_R |

Side-bending asymmetry = | SB_L−SB_R |

R_L is the rotational measurement of the left transverse process, and R_R is the rotational measurement of the right transverse process; SB_L is the measurement between the transverse processes on the left side, and SB_R is between the transverse processes on the right side. Both asymmetry assessments were measured pre- and posttreatment.

To assess any change in asymmetry after treatment, pretreatment asymmetry was compared to posttreatment asymmetry using the following equations, which were adapted from a previous study¹⁸:

Change in rotational asymmetry = | Pre-R_L−Pre-R_R |−| Post-R_L−Post-R_R |

Change in side-bending asymmetry = | PreSB_L−PreSB_R |− | PostSB_L−PostSB_R |

Pre-R_L is the rotational measurement on the left side before OMT and Pre-R_R is the rotational measurement on the right side before OMT; and Pre-SB_L is the intertransverse ligament length on the left before OMT and Pre-SB_R is the intertransverse ligament length on the right before OMT.

The absolute values allowed us to analyze the group together regardless of the direction of rotation. Without the absolute values, the right and left (positive and negative) asymmetries would negate each other and result in a population with no asymmetry. Additionally, greater absolute changes in participants with greater height and mass could introduce bias into our data. We did not have a mechanism to distinguish whether large changes in absolute value were due to the relative size of the participant's anatomy or whether the participants possessed a more significant dysfunction; this is an area for further research. To eliminate this bias, we also represented the changes as mean (%) improvement for the entire group (Table 2) and the subgroups (Table 3).

Statistical analysis included using an unpaired t test to examine the difference in soft tissue thickness between the sides with and without spinal asymmetry and using a 2-tailed paired t test to analyze the difference in soft tissue thickness before and after OMT. Intra- and interobserver reliability were tested using intraclass correlation coefficient (ICC). The correlation between US findings and osteopathic assessments was analyzed using Pearson correlation coefficient (PCC). A P value less than .05 was considered statistically significant. All statistical analysis was conducted by using SPSS software (SPSS version 25, IBM).