Investigation of Optimal Cue to Instruction for Pelvic Floor Muscle Contraction in Women Using Ultrasound Imaging

Abstract

Background Pelvic floor muscle (PFM) training is recommended as first line conservative management for stress urinary incontinence (SUI). The fundamental issue of how to optimally contract the PFM has not previously been investigated. An effective voluntary PFM contraction is known to positively influence the bladder neck and urethra which are urethrovesical (UV) structures associated with continence. The PFM may be globally or selectively contracted according to cue to instruction. The main research question was to investigate which cue to instruction for a PFM contraction has the potential to optimise position of UVSs following a brief period of practice in continent nulliparous pre-menopausal women (aiming to provide normative data) and parous menopausal women with previously unreported SUI.

Hypotheses Posterior or combined cues for instruction of PFM contraction are more influential in optimising UV position (UVP) during PFMC following brief practice than an anterior cue. Posterior or combined cues are equally influential in altering UVP.

Aims Preliminary aim was to investigate the reliability and suitability of 2-DRTUS and angle of urethral inclination (AUI) for imaging of selective contraction of the PFM and ease of reading images by a non diagnostic imaging researcher. Principal aim was to investigate if there is an optimal cue to instruction for a PFM contraction in two groups of women. Study 1: pre menopausal nulliparous continent women (to provide normative data) and Study 2: post menopausal parous stress incontinent women. Secondary aims were investigation of posture; ability to selectively contract the PFM contraction; and cue preference.

Method Study 1: Twenty women who were able to effectively and selectively contract were taught the following cues: anterior; posterior; anterior and posterior combined. Following 4 weeks of practice, perineal 2-D RTUS images of three PFMC for each cue were captured in supine and

standing twice (for repeatability analysis) five minutes apart. Two raters measured AUI. Data analysis was undertaken using a Customized General Linear Model (GLM) ANOVA with Bonferroni correction for interactions between all variables; subject, cue, posture and test. Seventeen data sets were available for analysis. Study 2: Methodology was based on Study 1. Twenty-one women were taught the study cues, followed the practice protocol and underwent data collection in the supine position. Twenty-one sets of data were available for analysis.

Results Reliability: ICC [1,3] for intra rater reliability was 0.957 [CI 95%: 0.946 to 0.967 p=0.000], inter rater reliability [2,1] 0.820 [CI 95%: 0.768 to 0.861] and for repeatability [1,3] 0.781 [CI 95%: 0.690 to 0.849 p=0.000] (continent) and 0.954 [CI 95%:0.931 to 0.971 p=0.000] (incontinent).

Principal results Study 1: anterior vs posterior cues (difference) 3.979˚ (CI 95%: [0.503 to 7.455 p=0.021]); anterior vs combined 3.777˚ (CI 95%: [-0.099 to 6.853 p= 0.059]) posterior vs combined cues -0.602˚ (CI 95%: [-2.874- 4.078 p=1.00]). Aggregated data from tests 1 and 2: anterior vs posterior 4.240° (CI 95%: [1.213 to 7.267 p=0.003]); anterior vs posterior 3.756° (95%CI: [0.729 to 6.783 p=0.009]); posterior vs combined-6.48° (95% CI: [-3.511 to 2.542 p=1.000]).

Principal results Study 2: anterior vs posterior 3.936˚ (95%CI: [0.863 to 7.008p=0.008]; 4.946˚ anterior vs combined (95%CI: [1.873 to 8.018 p=0.001]); posterior vs combined 1.010° (95%CI: -[2.062 to 4.082 p=1.000]). Aggregated analysis was anterior vs posterior 3.703˚ (95%CI: [1.639 to 5.761 p=0.000]); anterior vs combined 5.089˚ (95%CI: [3.0287 to 7.1503 p=0.000]) and posterior and combined 1.389° (95%CI: [-0.672 to 3.450 p=0.309]).

Secondary results: 2-D RTUS and the AUI were found to be suitable for investigating selective PFM contraction. Posture: supine vs standing (difference) 9.496˚ (p=0.000); (posture did not affect absolute AUI). Three continent (13%) and 2 incontinent (7%) subjects were unable to selectively contract the PFM. Cue preference in both studies was posterior or combined.

Conclusions AUI was significantly narrower/optimal when instruction for PFM contraction included a posterior cue, in both continent and stress incontinent women. This is proposed to be due to optimal recruitment of puborectalis. Puborectalis may be more important in urinary continence than widely recognized. This study has provided seminal information with respect to optimal cue to contraction for a PFM contraction and will change practice. Investigation of the potential impact of these findings clinically is required. It is proposed that further understanding will lead to standardisation of PFM instruction, ease of comparability between PFM research studies, and will clarify PFM instructions for the media and lay public.