A pilot trial of movement-based pelvic floor physical therapy to address pelvic floor myofascial pain and lower urinary tract symptoms

Abstract

Introduction and Hypothesis:

Pelvic floor myofascial pain (PFMP) is associated with lower urinary tract symptoms (LUTS). The objective of this study was to test movement-based pelvic floor physical therapy (PT) for patients with PFMP and LUTS. We hypothesized that movement-based PT designed to target PFMP would result in significant improvement in both PFMP and concomitant LUTS.

Methods:

This pilot trial enrolled patients with moderate-to-severe PFMP on palpation who were referred to movement-based PT to diagnose and treat pelvic floor dysfunction in the context of body alignment and movement patterns. The primary outcome was change in Urogenital Distress Inventory (UDI) scores after PT. Factors associated with PT attendance were also measured. We aimed to enroll 55 participants to achieve 80% power to detect a difference in 11 points on the UDI scores with an alpha-level of 0.05, accounting for a 10% loss to follow-up. The sample size was increased to 65 due to a higher-than-expected loss-to-follow-up rate.

Results:

65 patients were enrolled and 62 analyzed. 38 (61.3%) attended PT, and 30 (48.4%) completed a follow-up PT Attendance (PTA) survey. Overall UDI score and irritative, obstructive, and stress subscales (p<0.0001) improved in participants that attended PT as well as mean myofascial examination scores at each site.

Conclusions:

Participants that attended movement-based PT demonstrated an improvement in LUTS. Future studies should extend our findings by: 1) confirming whether the myofascial pain-directed elements of PT improved LUTS, 2) investigate whether movement-based PT improves prolapse symptoms, and 3) including a non-PT control arm to rule out the possible influence of a placebo effect and behavioral modifications on LUTS and PFMP.

Brief summary:

In this pilot trial, participants with pelvic floor myofascial pain demonstrated an improvement in their LUTS after movement-based PT.

Introduction

Pelvic floor myofascial pain (PFMP) with palpation has been identified in the muscles of the internal hip [obturator internus (OI)] and pelvic floor [levator ani (LA)]. We have previously shown that PFMP is common among women presenting for evaluation of pelvic floor disorder symptoms[1] when a standardized pelvic floor muscle examination[2] is universally performed. PFMP severity appears to correlate with degree of bother from pelvic floor disorder symptoms, including lower urinary tract symptoms (LUTS)[1].

LUTS, including urinary urgency, frequency, and sensation of incomplete bladder emptying, affect up to 75% of women and impair quality of life [3]. Depending on the presenting symptom, the mainstay of treatment typically involves either medication or surgery. Medications to address overactive bladder symptoms, typically anticholinergic agents, are associated with a high rate of discontinuation due to low efficacy, bothersome side effects, and possible long-term negative effects on cognition [4]. Surgical procedures for refractory overactive bladder, including botulinum toxin injections or sacral neuromodulation, are invasive and may not completely address patients’ symptoms. Furthermore, when presented with various treatment options, women tend to prefer conservative options [5,6].

Pelvic floor muscle exercises are an effective conservative treatment option for women with overactive bladder and stress urinary incontinence but are typically focused on strengthening “weak” muscles in order to prevent urine leakage with an urge or during physical exercises [7]. In women with PFMP and LUTS, muscle impairments specific to pain, peripheral sensitization, reduced muscle length and impaired coordination may also be present [8,9]. Pelvic floor physical therapy (PFPT) for PFMP typically focuses on soft tissue mobilization to address peripheral sensitization and reduced muscle length. An alternative physical therapy (PT) approach is movement-based PT, which provides a framework for categorization and subsequent individualized interventions to address the presenting pelvic floor muscle impairments [10]. Movement-based PT can be used to address both the specific muscle impairments and the habitual movement patterns that contribute to PFMP and LUTS [9,10]. For women with both LUTS and PFMP, little data exist to guide the best PT approach.

This pilot study was designed to test movement-based PT as a treatment strategy for patients with concurrent PFMP and LUTS. Given the correlation between PFMP severity and pelvic floor disorder symptom bother [1], we hypothesized that movement-based PT designed to target PFMP would result in significant improvement in both PFMP and concomitant LUTS.

Materials and Methods

Study Overview and Procedures

This was a single arm pilot trial conducted at one tertiary care institution. Patients presenting with LUTS who were found to have moderate-to-severe PFMP on examination (described below) were eligible. Participants provided baseline information on LUTS and were referred for movement-based PT at our institution. After completing a course of PT, participants returned for follow-up examination, LUTS assessment, and completion of additional questionnaires. Examiners at the follow-up examination were not blinded to the initial examination scores. Mid-way through enrollment, participant attendance and subsequent follow-up was lower than anticipated; thus, we introduced a Physical Therapy Attendance (PTA) survey to better understand factors influencing attendance and follow-up.

Study Population and Eligibility

All new patients presenting to the Female Pelvic Medicine & Reconstructive Surgery clinic for evaluation of pelvic floor symptoms undergo an evidence-based, standardized pelvic floor myofascial screening examination to assess pain to palpation of the bilateral OI and LA muscles.² Adult female patients were eligible for inclusion if they were found to have PFMP on examination of ≥ 4/10 in severity at ≥ two sites and were prescribed and agreed to attend PT to address their myofascial pain. To ensure consistency among PT evaluation and treatment, only patients willing and able to attend PT at our institution were eligible for inclusion. Patients were excluded if they were younger than 18 or non-English-speaking.

Intervention

Participants enrolled in this study saw one of four women’s health physical therapists at our institution, all of whom were either board certified in women’s health PT or had completed formal education related to pelvic health PT. All physical therapists utilized a movement-based approach to diagnose and treat pelvic floor dysfunction in the context of body alignment and movement patterns [10,11]. To diagnose and subsequently direct intervention for the presenting movement impairments, each PT evaluated first the movement patterns of the spine, hip and pelvis, followed by a pelvic floor muscle examination [12,13]. In addition, self-reported functional activities associated with presenting symptoms were observed. The standard examination included analysis of trunk motions (forward bending, return from forward bending, side bending, and rotation), hip motions [single leg stance, Stork Test [13], mini squat (hip and knee bending)], hip rotation and flexion range of motion, as well as sitting, sit-to-stand, sleeping position, position for intercourse, and carrying/lifting objects. If thoracic spine movement impairments were noted, the effect of arm motion on the spine was included in the examination. To categorize the specific pelvic floor muscle movement impairments an analysis was done on the presence of pain to touch, length, strength, and motor performance during activation, relaxation, lowering, and coughing [9]. Interventions were determined based on the presenting movement-based diagnoses. Pelvic floor muscle-specific care was typically organized to first address pain, then impairments related to movement. Pain management strategies included reduction peripheral sensitization through local application of ice and manual therapy. In the presence of a non-relaxing pelvic muscle, muscle length issues were prioritized prior to addressing strength and coordination activities. Trunk, hip, and pelvic girdle impairments were addressed through movement pattern training for functional activities as well as specific motor performance exercises for the trunk and hip muscles. Functional activity training included practicing corrected bending, sit-to-stand, squatting, sleeping position, carrying and reaching tasks. PT visits were typically once per week for 6–8 weeks. The specific frequency and duration of PT visits was left to the discretion of the treating physical therapist.

Outcomes

The primary outcome was change in Urogenital Distress Inventory (UDI) score after PT. The UDI is a validated 19-item scale used to measure the presence and degree of LUTS bother [14,15]. It assesses the full range of LUTS and can be considered in 3 subscales: obstructive/discomfort, irritative, and stress. Scores range from 0–300; higher scores indicate more severe LUTS. Secondary outcomes included change in myofascial examination scores, change in LUTS Tool scores, and patient-reported improvement. The LUTS Tool is a validated measure used to assess frequency and bother of 18 LUTS defined by the International Continence Society [16]. Patient-reported improvement was measured by the Patient Global Impression of Improvement (PGI-I). Additional PT-related factors were measured including number of sessions attended, time since last appointment, diagnosis made by the physical therapist, and overall satisfaction with PT as a treatment modality as reported on the Short Assessment of Patient Satisfaction (SAPS) [17].

The PTA survey was developed using factors reported in the literature to be associated with attendance at PFPT [18,19]. It was administered as part of the follow-up questionnaires for all participants enrolled in the study. Participants that had completed their follow-up before PTA survey implementation or who had been lost to follow-up were contacted, consented by telephone, and completed the survey by telephone or email.

Statistical Analysis and Power Calculation

Baseline demographic and clinical characteristics were compared between participants that did and did not attend PT using student’s t-test and Chi-square test as appropriate. Changes in UDI and myofascial pain scores from baseline to follow-up visit were compared using Wilcoxon signed-rank test. LUTS Tool outcomes were dichotomized to “never/rarely” or “at least sometimes” and compared from baseline to follow-up by McNemar’s test. PGI-I scores were dichotomized to “improvement” (“a little better”, “much better”, “very much better”) or “no improvement” (“no change”, “a little worse”, “much worse”). Spearman’s rank correlation coefficient was used to estimate the correlation between changes in UDI and PFMP scores. Factors associated with attendance at PT were evaluated using student’s t-test and Chi-square test as appropriate. SAS version 9.4 (Cary, NC) was used for data analysis.

We originally powered our trial to detect a change of 11 points in UDI-scores before and after PT, the established minimally important difference for this measure [20]. Fifty participants were required to achieve 80% power with an alpha-level of 0.05. Accounting for a 10% loss to follow-up, we planned to enroll 55 participants. Midway through recruitment, it became apparent that our loss-to-follow-up rate was likely to exceed 10%, thus we increased our recruitment goal to 65 participants. This study was approved by the Institutional Review Board at Washington University in St. Louis (IRB #201609115).

Results

Sixty-five patients met inclusion criteria and were enrolled. Three patients who did not complete the initial UDI were excluded, leaving 62 participants for analysis. Of those, 38 (61.3%) attended PT. Follow-up UDI and myofascial examination scores were available for 35 participants (56.4%): 32 that attended PT and 3 that did not. Most participants were postmenopausal (69.4%, mean age 55.6 years), white (85.3%), and overweight (mean BMI 29.6 kg/m²; Table 1). 37.3% were not sexually active; among those who were, 22 (55%) reported dyspareunia. Baseline demographics were similar between those that did and did not attend PT except for race. Participants that attended PT were more likely to be white (p=0.007) and to incorporate behavioral modifications (p<0.0001) and vaginal icing (p<0.0001). Baseline UDI and myofascial examination scores did not differ between participants by PT attendance. The highest scores were observed in the irritative subscale for both groups (Table 1).

Table 1.

Demographic variables of included participants.

	Overall group (N=62)	Did not attend PFPT (N=24)	Attended any PFPT (N=38)	p-value
Age1	55.6 (16.1)	52.5 (15.2)	57.5 (16.5)	0.703
BMI1	29.6 (6.4)	31.7 (7.2)	28.3 (5.5)	0.120
Gravidity2	2 (2–3)	3 (2–4)	2 (1–3)	0.082
Parity2	2 (1–3)	2 (2–3)	2 (0–3)	0.501
Dyspareunia3	22 (37.3)	10 (45.5)	12 (32.4)	0.419
Not sexually active3	22 (37.3)	8 (36.4)	14 (37.8)	0.779
Postmenopausal3	43 (69.4)	15 (62.5)	28 (73.7)	0.352
Prolapse symptoms3	11 (18.6)	3 (13.6)	8 (21.6)	0.220
Race3				0.007
White	52 (85.3)	16 (69.6)	36 (94.7)
Black	9 (14.8)	7 (30.4)	2 (5.3)
Hysterectomy3	21 (33.9)	9 (42.9)	12 (57.1)	0.631
Right handed3	54 (90.0)	22 (91.7)	32 (84.2)	0.077
POP-Q Stage3				0.144
0/1	54 (88.5)	21 (87.5)	33 (89.2)
2	5 (8.2)	1 (4.2)	4 (10.8)
≥3	2 (3.3)	2 (8.33)	0 (0.0)
Did you perform vaginal icing? 3	26 (78.8)	2 (8.3)	24 (63.2)	<0.0001
Duration of vaginal icing? 3				0.0057
<1 week	3 (12.0)	1 (50.0)	2 (8.8)
1–2 weeks	5 (20.0)	1 (50.0)	4 (17.4)
2–3 weeks	1 (4.0)	0	1 (4.4)
3–4 weeks	5 (20.0)	0	5 (21.7)
5–6 weeks	2 (8.0)	0	2 (8.7)
>6 weeks	9 (36.0)	0	9 (39.1)
Incorporated behavioral modifications3	23 (37.1)	1 (4.2)	22 (57.9)	<0.0001
Modification of fluid intake	6 (9.7)	0	6 (15.8)
Weight loss	3 (4.8)	0	3 (7.9)
Elimination of bladder irritants	9 (14.5)	0	9 (23.7)
Timed voiding	19 (30.7)	1 (4.2)	18 (47.4)
Baseline UDI Scores2,4
UDI Irritative	72.9 (62.5–81.3)	72.9 (64.6–83.3)	72.9 (60.4–81.3)	0.500
UDI Stress	62.5 (62.5–75.0)	68.8 (62.5–78.1)	62.5 (56.3–75.0)	0.368
UDI Obstructive	63.6 (61.4–68.2)	64.8 (62.5–69.9)	63.6 (61.4–68.2)	0.327
UDI Total	204.2 (192.8–220.5)	208.7 (195.4–219.0)	202.8 (191.9–222.2)	0.370
Baseline Myofascial pain scores
Right SI Joint3	21 (33.9)	10 (41.7)	11 (29.0)	0.303
Left SI Joint3	18 (29.0)	8 (33.3)	10 (26.3)	0.553
Right ASIS3	36 (58.1)	13 (54.2)	23 (60.5)	0.621
Left ASIS3	36 (58.1)	14 (58.3)	22 (57.9)	0.973
Rectus insertion3	22 (35.5)	11 (45.8)	11 (29.0)	0.176
Right OI2	6 (4.5–8)	5.5 (3–9)	7 (5–8)	0.184
Right LA2	5 (4–7)	5 (4–8)	5 (4–7)	0.511
Left LA2	5 (3–6)	5 (4–5)	5 (3–6)	0.969
Left OI2	7 (4.5–8)	6.5 (4–9)	7 (5–8)	0.693
Chief Complaint to PT3,5
Pelvic pain			21 (33.9)
Urinary incontinence			20 (32.3)
Urinary urgency/frequency			14 (22.6)
Frequent UTIs			3 (4.8)
Prolapse			1 (1.6)
Fecal incontinence			1 (1.6)
Musculoskeletal/Neuromuscular Diagnosis3,5
Movement Pattern Coordination Impairment			34 (89.5)
Lumbar Extension			7 (18.4)
Impaired intraabdominal pressure regulation			7 (18.4)
Thoracic Flexion			6 (15.8)
Lumbar Rotation			5 (13.2)
Pelvic impairment			5 (13.2)
Thoracic Rotation			4 (10.5)
Hip adduction with medial rotation			4 (10.5)
Force Production Deficit			6 (15.8)
Shortened pelvic floor			4 (10.5)
Impaired endurance			3 (7.9)
Lumbar compression			2 (5.3)
Number of PFPT sessions attended2			5 (4–7)
Time since last PFPT session (weeks) 2			6.5 (4–10)

1 unreported race, 2 unreported handedness. Myofascial pain scores reported as yes (n, %) for right SI, left SI, right ASIS, left ASIS, rectus and median (IQR) for bilateral OI and LA.

¹Mean (SD);

²Median (IQR);

³n(%);

⁴p-values calculated with Kruskal-Wallis test for UDI, OI, and LA scores between patients who did and did not attend PFPT; chi-square for SI joint, ASIS, rectus insertion.

⁵Chief complaint to PT and musculoskeletal/neuromuscular diagnoses are not mutually exclusive, many patients had more than 1 chief complaint and diagnosis; data only available for participants that attended PT

Among the 38 participants that attended PT, median attendance was 5 sessions (IQR 4–7 sessions). Follow-up occurred a median of 6.5 weeks after their last PT session (IQR 4–10 weeks; Table 1) and 19 weeks after their initial visit (IQR 13–27 weeks). One participant was still attending PT at follow-up. One pelvic floor physical therapist saw 43.8% of participants, although all four physical therapists saw participants in this study. Pelvic pain and urinary incontinence were the most common chief complaints reported by participants at their initial PT visit (n=21, 33.9% and n=20, 32.3% respectively; Table 1). Urinary urgency/frequency was also frequently reported (n=14, 22.6%). Participants were given a range of movement diagnoses (Table 1). 80% of participants continued some or all of their home exercises after their last session.

Outcomes After PFPT

Among participants that attended PT, overall UDI score improved significantly by a mean of 125.2 points (p<0.0001; Table 2). Irritative, obstructive, and stress subscales also improved significantly (all p<0.0001), with the greatest improvement observed on the obstructive subscale (−49.3 points) and similar improvement in irritative (−36.4 points) and stress (−39.5 points) subscales. Three participants did not attend PT. These participants reported a non-significant improvement in overall UDI score (−82.9 points, p=0.50), irritative (−26.0 points, p=0.50) and obstructive subscales (−43.2 points, p=0.25), and worsening of symptoms on the stress subscale (mean +4.2 points, p=1.00). Small sample size precluded comparison of symptom changes between participants that did and did not attend PT.

Table 2.

Baseline and follow up Urogenital Distress Inventory (UDI) and myofascial pain scores.

	Baseline				Follow up				Attended PT
	Full sample (n=62)	Did not attend PT (n=24)	Attended PT (n=38)	p-value1	Full sample (n=35)	Did not attend PT (n=3)	Attended PT (n=32)	p-value1	Mean Change2 (n=32)	Pre vs Post p-value3 (n=32)
UDI Scores4
UDI Irritative	72.9 (62.5–81.3)	72.9 (64.6–83.3)	72.9 (60.4–81.3)	0.500	37.5 (12.5–54.2)	60.4 (54.2–66.7)	35.4 (12.5–50.0)	0.124	−36.4 (17.9)	<0.0001
UDI Stress	62.5 (62.5–75.0)	68.8 (62.5–78.1)	62.5 (56.3–75.0)	0.368	25.0 (0.0–62.5)	75.0 (75.0–100.0)	25.0 (0.0–50.0)	0.014	−39.5 (26.2)	<0.0001
UDI Obstructive	63.6 (61.4–68.2)	64.8 (62.5–69.9)	63.6 (61.4–68.2)	0.327	15.9 (0.0–29.5)	25.0 (25.0–36.4)	12.5 (0.0–26.1)	0.107	−49.3 (15.1)	<0.0001
UDI Total	204.2 (192.8–220.5)	208.7 (195.4–219.0)	202.8 (191.9–222.2)	0.370	76.3 (50.0–128.8)	160.4 (154.2–166.7)	71.2 (48.1–125.0)	0.034	−125.2 (41.2)	<0.0001
Myofascial pain scores5
Right SI Joint	21 (33.9)	10 (41.7)	11 (29.0)	0.303	13 (37.1)	0 (0.0)	13 (40.6)	0.378	0 (0)	0.484
Left SI Joint	18 (29.0)	8 (33.3)	10 (26.3)	0.553	12 (34.2)	0 (0.0)	12 (37.5)	0.424	0 (0)	0.337
Right ASIS	36 (58.1)	13 (54.2)	23 (60.5)	0.621	14 (40.0)	1 (33.3)	13 (40.6)	0.613	10 (31.3)	0.569
Left ASIS	36 (58.1)	14 (58.3)	22 (57.9)	0.973	14 (40.0)	1 (33.3)	13 (40.6)	0.613	9 (28.1)	0.426
Rectus insertion	22 (35.5)	11 (45.8)	11 (29.0)	0.176	13 (37.1)	1 (33.3)	12 (37.5)	0.765	0 (0)	0.574
Right OI	6 (4.5–8)	5.5 (3–9)	7 (5–8)	0.184	5.5 (3.5–9)	6.5 (6–7)	5 (3–10)	0.674	−1.0 (3.7)	0.145
Right LA	5 (4–7)	5 (4–8)	5 (4–7)	0.511	2 (0–5)	1.5 (0–3)	2 (0–5)	0.504	−1.5 (3.8)	0.047
Left LA	5 (3–6)	5 (4–5)	5 (3–6)	0.969	4 (0–5)	6.5 (5–8)	3 (0–5)	0.167	−0.5 (3.6)	0.395
Left OI	7 (4.5–8)	6.5 (4–9)	7 (5–8)	0.693	4 (1–7)	7 (5–9)	4 (1–5)	0.248	−2.1 (3.8)	0.014

SI, sacroiliac joint; ASIS, anterior superior iliac spine; OI, obturator internus; LA, levator ani.

¹p-values calculated with Kruskal-Wallis test for UDI, OI, and LA scores between patients who did and did not attend PFPT; chi-square for SI joint, ASIS, rectus insertion.

²Mean(SD), for sample that attended PFPT; external sites (Right SI joint, Left SI joint, Right ASIS, Left ASIS, Rectus Insertion) reported as n(%) with improved pain.

³p-values calculated with Wilcoxon signed rank test for change in UDI scores and change in OI and LA pain scores; chi-square for SI joint, ASIS, rectus insertion.

⁴Median (IQR).

⁵Myofascial pain scores for SI joints, ASIS, rectus insertion reported n(%) with pain; OI and LA scores reported median (IQR).

Among participants that attended PT, mean myofascial examination scores improved at each site, and improvement was statistically significant in the left OI (−2.1, p=0.014) and right LA (−1.5, p=0.047; Table 2). Change in myofascial pain scores and change in UDI score after PT were not significantly correlated (ρ= −0.053, p=0.809; Supplemental Table 1). Participants that attended PT reported significant improvement in urinary frequency, stress urinary incontinence, and unexplained leakage (Supplemental Table 2). Most participants (80.7%) reported improvement in their urinary symptoms on the PGI-I; 14 (45.2%) reported “much better” or “very much better” after PT. Most participants reported being satisfied (43.8%) or very satisfied (21.9%) in response to the question, “How satisfied were you with the effect of your physical therapy?” on the SAPS.

Physical Therapy Attendance (PTA) Survey Results

30 participants (48.4%) completed the PTA survey: 19/38 (50.0%) that attended PFPT and 11/24 (45.8%) that did not. Participants that attended PT were more likely to have checked with their insurance carrier regarding coverage (p=0.011), had higher educational achievement (p=0.022), and were less likely to have concerns regarding travel to and from PT (p=0.010). There was no difference in apprehension about PT, type of insurance, method of transportation, perceived health status and impression of utility of PT between those who did and did not attend PT (Table 3).

Table 3.

Factors associated with attendance at PFPT

	Overall (n=30)	Did not do PFPT (n=11)	Attended any PFPT (n=19)	p-value
Do you think PFPT will improve your symptoms?				0.839
Yes	19 (63.3)	7 (63.6)	12 (63.2)
No	4 (13.3)	1 (9.1)	3 (15.8)
Unsure	7 (23.3.)	3 (27.3)	4 (21.1)
Are you nervous about PFPT?				0.284
Yes	10 (33.3)	5 (45.5)	5 (26.3)
No	20 (66.7)	6 (54.6)	14 (73.7)
What is your primary insurance?				0.795
Private	17 (56.7)	6 (60.0)	11 (55.0)
Public	13 (43.3)	4 (40.0)	9 (45.0)
Did you check with your insurance carrier to determine if PFPT would be covered?				0.011
Yes	14 (45.2)	2 (16.7)	12 (63.2)
No	17 (54.8)	10 (83.3)	7 (36.8)
How do you travel to your appointments?				0.095
Walk/bicycle	1 (3.2)	0 (0)	1 (5.3)
Drive	25 (80.7)	8 (66.7)	17 (89.5)
Transported by family/friend	5 (16.1)	4 (33.3)	1 (5.3)
Does traveling to your PFPT appointments influence whether you think this treatment is an option for you?				0.010
Yes	12 (38.7)	8 (66.7)	4 (21.1)
No	18 (58.1)	3 (25.0)	15 (79.0)
Unsure	1 (3.2)	1 (8.3)	0 (0.0)
In general, would you say your health is:				0.128
Excellent	5 (16.7)	0 (0.0)	5 (26.3)
Very good	9 (30.0)	2 (18.2)	7 (36.8)
Good	10 (33.3)	5 (45.5)	5 (26.3)
Fair	4 (13.3)	3 (27.3)	1 (5.3)
Poor	2 (6.7)	1 (9.1)	1 (5.3)
What is the highest grade or year of school you completed?				0.022
Grade 12 or GED	4 (12.9)	4 (33.3)	0 (0.0)
College 1–3 years	11 (35.5)	4 (33.3)	7 (36.8)
College 4 years or more	16 (51.6)	4 (33.3)	12 (63.2)
What is your household income in dollars?				0.174
<$20,000	5 (16.7)	4 (36.4)	1 (5.3)
$20,000 – $40,000	2 (6.7)	1 (9.1)	1 (5.3)
$40,000 – $60,000	3 (10.0)	0 (0.0)	3 (15.8)
>$60,000	14 (46.7)	4 (36.4)	10 (52.6)

Discussion

Pelvic floor PT is an effective management strategy for patients with urinary incontinence [21,22], but regimens often involve primarily strengthening exercises [21]. In this single arm trial, we found evidence to suggest that movement-based PT designed to address underlying PFMP may also be effective for improving LUTS.

Alternative explanations for improvement in LUTS after movement-based PT

As our trial included only one arm, additional possible explanations for our findings must be considered, including a placebo effect or regression to the mean. The improvement in UDI score and UDI-stress subscale scores (−125.1 and −39.5 points respectively) following PT for PFMP greatly surpassed the published MID for these scores (−11 points for UDI and −8 points for UDI-stress) [20]. Interestingly, the mean difference in UDI score (but not UDI-stress subscale) for participants that did not attend PT also greatly surpassed the MID (mean difference −82.9 points), raising the possibility of a placebo effect. Other studies on LUTS interventions have shown a placebo effect of 9–34% [23], especially in symptoms of urinary frequency [24]. UDI score-improvement was greater overall and across all subscales for participants that attended PT, although follow-up UDI scores were only available for 3 participants that did not complete PT, precluding meaningful comparison of these groups. The greatest improvement in individual subscale scores for both participants that did and did not attend PT was on the obstructive subscale (−49.3 points and −43.2 points respectively). An MID for the obstructive scale of the UDI has not been published.

LUTS improvement in this study, particularly in the non-PT group, may also be explained by regression to the mean. Findings from the Nurse’s Health Study demonstrate that, over 10-years, 26–56% of women younger than 56 with moderate to severe urinary incontinence will have improvement in symptoms without intervention and another 3–11% will experience symptom resolution [25]. Heidler et al also noted a mean annual remission rate of 4.6% in symptoms of urgency and frequency [26]. Our findings may reflect the dynamic natural course of LUTS rather than a true effect of PT. Although both plausible explanations for our findings, we feel it is unlikely that either a placebo effect or regression to the mean explains the significant improvement noted in participants that attended PT given the degree of symptomatic improvement and the timeframe over which this improvement occurred.

Another factor that may explain some of our observed improvement in LUTS is concomitant behavioral modifications. Participants that attended PT were more likely to incorporate behavioral modifications to address LUTS including modifying fluid intake, weight loss, elimination of bladder irritants, and timed voiding [21]. This may indicate that certain patients are intrinsically more motivated to incorporate behavioral modifications to address their LUTS as adherence to PT also requires motivation. Furthermore, attendance at PT may serve to reinforce the importance of behavioral modifications as physical therapists at our institution also review behavioral interventions with patients who present with LUTS. Participants that attended PT were also more likely to incorporate vaginal icing for PFMP, a novel intervention with anecdotal evidence [27]. Results from a randomized controlled trial investigating the utility of this intervention at our institution are forthcoming. Thus, the improvement in LUTS demonstrated in this study may not be solely due to improvements in the extra-pelvic movement impairments.

Impact of movement-based PT on myofascial pain

Our findings demonstrating an improvement in LUTS after movement-based PT lend additional support to the proposed link between PFMP via palpation (i.e., not associated with chronic pelvic pain) and LUTS. Movement-based PT designed to address underlying PFMP resulted in an improvement in both LUTS and myofascial examination scores, although actual myofascial pain score improvement and the correlations between change in myofascial pain score and LUTS were not significant. This may be explained by the time between the last PT visit and follow-up examination, as well as a possible delayed influence of PFMP on LUTS. Participants returned a median of 19 weeks after their initial visit and 6.5 weeks after their last PT session. Although 80% of participants reported continuing some or all of their home exercises, the time from last PT session may explain the non-significant difference noted in myofascial pain scores despite attendance at PT. If patients returned for examination closer to their last PT session, the improvement in myofascial pain scores may have been greater. In addition, it is possible that PFMP on palpation may take time to contribute to LUTS. If so, the timing of follow-up examination may have enabled us to detect worsening myofascial pain on palpation before it contributed to a detrimental effect on LUTS. This would also explain the non-significant correlation between these outcomes. The examiners at the time of the follow up visit were not blinded to the original myofascial pain scores, however, follow up examinations occurred at a median of 19 weeks after the initial visit and pre-intervention myofascial pain scores were not reviewed prior to repeat examination, thus the risk of bias is likely low. Additionally, our study was powered to detect a difference in LUTS based on the UDI, and we may have been underpowered to detect a difference in myofascial pain scores after PT.

Factors influencing attendance at PT

Unfortunately, participation in PT as a treatment regimen is limited, even among participants enrolled in a trial designed to investigate outcomes after PT. In this prospective study, we specifically recruited patients who expressed a willingness and ability to attend PT at our institution, yet non-attendance was high similar to studies conducted outside of a trial protocol [28,29]. This high rate of non-attendance may be due to restricted access given notable sociodemographic differences between participants that did and did not attend PT in this study. Importantly, we did not provide compensation, transportation, or price-reduction for PT for participants in this study. While this may have impacted attendance and our ability to fully investigate our primary outcome, this is more consistent with “real world” experience as affordability is known to influence adherence to PT [18].

Some authors have shown that anxiety about PT and perceived lack of utility of PT are barriers to PT attendance, specifically for patients with high-tone pelvic floor dysfunction [18,29]. We did not find a difference in reported apprehension about or perceived utility of PT between participants that did and did not attend PT. This suggests that participants in our sample felt well-informed on the utility of and approach to PT, which emphasizes the importance of adequate counseling on treatment expectations.

Strengths and limitations

Although our final sample size was inadequate to draw definitive conclusions on the correlation between the change in myofascial pain scores and LUTS, we did have a sufficient sample size to detect an improvement in LUTS following movement-based PT and to identify factors associated with PT attendance. In addition, our ability to engage enrolled participants in follow-up surveys to measure factors associated with PT participation, even when they did not comply with the recommended treatment, allowed us to draw meaningful conclusions on factors associated with the barriers to care. Since the study was not specifically designed to investigate this question, we may not have captured all relevant factors associated with these barriers to care. Nevertheless, we were able to identify important differences between groups. Prior data has demonstrated a correlation between PFMP and prolapse symptoms, specifically pelvic pressure[1], but we were unable to analyze whether these symptoms improved after movement-based PT in this sample because prolapse symptoms were not measured at follow-up. All physical therapists involved in this study were specifically trained in both pelvic floor and movement pattern training. Though the specific treatment was not standardized across the study, overall methods and approach were consistent across participants. Overall, we believe this improves the validity of our findings.

Conclusions

LUTS and myofascial pain scores improved after movement-based PT for PFMP. Unfortunately, barriers to attendance at PT make this conservative management strategy unavailable for some patients. Follow-up studies should expand on these findings by: 1) confirming whether the myofascial pain-directed elements of PT improved LUTS, 2) investigating whether movement-based PT improves prolapse symptoms, and 3) including a non-PT control arm to rule out the possible influence of a placebo effect and behavioral modifications on LUTS and myofascial pain.