Responsiveness, Smallest Detectable Change, and Minimal Important Change of the Turkish Pelvic Floor Distress Inventory‐20 Following Pelvic Floor Muscle Training

ABSTRACT

Background and Purpose

The Pelvic Floor Distress Inventory‐20 (PFDI‐20) is a widely recognized clinical tool designed to assess the severity of distress caused by pelvic floor symptoms. Within the context of conservative physiotherapy and pelvic floor rehabilitation, this study aimed to determine the responsiveness, smallest detectable change (SDC), and minimal important change (MIC) of the Turkish PFDI‐20 in women with pelvic floor dysfunction (PFD).

Methods

This multicenter, retrospective study consisted of 200 women (mean age, 46.92 ± 11.13 years) with PFD recruited across three clinical sites who had undergone a standardized 8‐week conservative pelvic floor muscle training intervention. The PFDI‐20, which includes three subscales (Pelvic Organ Prolapse Distress Inventory [POPDI‐6], Colorectal‐Anal Distress Inventory [CRADI‐8], and Urinary Distress Inventory [UDI‐6]), was administered at baseline and follow‐up clinical assessments. The instrument's responsiveness was assessed using the Wilcoxon signed‐rank test, standardized response means (SRMs), and effect sizes (ESs). The preliminary SDC estimates were calculated at the 95% confidence level. Receiver operating characteristic (ROC) analysis was used to determine preliminary MIC values, maximizing both sensitivity and specificity.

Results

For PFDI‐20 and UDI‐6, the responsiveness was excellent (ES: 1.24, SRM: 1.41, for PFDI; ES: 1.47, SRM: 1.53, for UDI‐6), moderate to good (ES: 0.78, SRM: 0.94), and moderate (ES: 0.55, SRM: 0.76) for POPDI‐6 and CRADI‐8, respectively. The preliminary SDC and MIC values were 55.60 and 20.83, 23.39 and 20.83, 17.71 and 9.37, 29.13 and 25.00, for PFDI‐20, POPDI‐6, CRADI‐8, and UDI‐6, respectively. The ROC analyses showed moderate to excellent discriminative ability, with an area under the curve of 0.85, 0.72, 0.72, and 0.85 for PFDI‐20, POPDI‐6, CRADI‐8, and UDI‐6, respectively.

Discussion

These findings support the responsiveness of the Turkish PFDI‐20 when applied in a conservative physiotherapy setting, providing preliminary estimates, specifically a SDC of 55.60 alongside a MIC of 20.83. However, derived from an active intervention cohort and imbalanced anchor groups, these values should be interpreted with caution as practical response thresholds within longitudinal clinical evaluations.

1. Introduction

Pelvic floor dysfunction (PFD) is a multifactorial condition encompassing key pathologies such as urinary and anal incontinence, bowel and bladder disorders, and pelvic organ prolapse. As outlined by the NICE (National Institute for Health and Care Excellence) guidelines, these conditions involve complex impairments of the pelvic floor muscles, connective tissues, and supporting structures (National Institute for and Health and Care Excellence (Great Britain) ²⁰¹⁰). Among these, urinary and anal incontinence and pelvic organ prolapse represent the core diagnostic entities linked to underlying structural deficiencies (National Institute for and Health and Care Excellence (Great Britain) ²⁰¹⁰).

Due to the complexity of PFD, subjective patient‐reported outcome measures (PROMs) are widely used to quantify symptom burden and track changes over time (Jeyaraman et al. 2025; van der Willik et al. ²⁰²¹). However, since statistical significance does not always equate to clinical relevance, a key challenge is determining whether an observed score change is genuinely meaningful to the patient (AbdulRaheem ²⁰²⁴). To address this gap, clinimetric indices such as responsiveness, smallest detectable change (SDC), and minimal important change (MIC) are essential (Terwee et al. 2007).

Responsiveness reflects an instrument's capacity to detect longitudinal changes over time (Terwee et al. 2021). To evaluate whether a change is both reliable and meaningful, the SDC, which exceeds measurement error (Terwee et al. 2007), must be considered alongside the MIC, denoting the smallest change perceived as beneficial by the patient (Terwee et al. 2021; Willigenburg and Poolman ²⁰²³). Therefore, assessing both indices is crucial to distinguish statistical significance from true clinical relevance.

The Pelvic Floor Distress Inventory (PFDI‐20) measures the distress caused by pelvic floor symptoms and has been validated in several languages and adapted into Turkish (M. D. Barber et al. ²⁰⁰¹; Teleman et al. ²⁰¹¹; De Tayrac et al. ²⁰⁰⁷; Due et al. ²⁰¹³; Yoshida et al. ²⁰¹³; Utomo et al. ²⁰¹⁴; Grigoriadis et al. ²⁰¹³; Sanchez‐Sanchez et al. ²⁰¹³; Toprak et al. ²⁰¹²). However, data regarding its longitudinal responsiveness, SDC, and MIC following conservative treatments like pelvic floor muscle training (PFMT) remain lacking in the Turkish population (de Arruda et al. ²⁰²¹; Kaplan et al. ²⁰¹²). Furthermore, while limited studies have reported partial MIC values, comprehensive SDC and MIC values for the Turkish PFDI‐20 and all its subscales have yet to be established (Utomo et al. 2014; M. Barber et al. ²⁰⁰⁵; Wiegersma et al. ²⁰¹⁷; Karjalainen et al. ²⁰²¹). Therefore, this study aimed to determine the responsiveness, SDC, and MIC of the Turkish PFDI‐20 and its subscales in women with PFD undergoing conservative PFMT intervention. We hypothesized that the Turkish PFDI‐20 and its subscales would demonstrate acceptable responsiveness and that preliminary SDC and MIC estimates could be established within this population.

2. Methods

2.1. Patient Selection

The design, methodology, and reporting of this multi‐center study were conducted in strict accordance with the updated Consensus‐based Standards for the Selection of Health Measurement Instruments (COSMIN) Reporting Guideline 2.0 to ensure clinical clarity, transparency, and reproducibility (Gagnier et al. 2025; de Arruda et al. ²⁰²⁵). This retrospective study received ethical clearance from the Xxxxx Xxxxx Xxxxxx Xxxxxxxxxx Health Research Ethics Committee (Approval Date: XX Xxx XXXX; Approval Number: XXXX), with the requirement for informed consent waived due to the study's design. A multi‐center retrospective design was purposefully employed to capture authentic ‘real‐world’ patient changes over time within routine clinical practice, rather than evaluating or establishing direct treatment efficacy in a controlled experimental setting. This approach minimizes the Hawthorne effect while ensuring high statistical power through a large, diverse patient cohort that reflects established clinical practice.

To ensure a representative study population and eliminate selection bias, patient selection was conducted using a rigorous consecutive sampling method. Because of the retrospective and multi‐center design, the sample size was determined by the total available clinical records meeting the inclusion criteria. The recruitment process involved a comprehensive electronic and physical archive screening of all consecutive medical records of female patients who had completed a standardized conservative treatment for pelvic floor complaints between June 2025 and February 2026 across the urology and pelvic health departments of three distinct public hospitals. To ensure reproducibility, a standardized screening protocol was applied across all three participating sites by the tracking investigators to identify eligible charts. Patients were included strictly based on predefined objective criteria: a clinical diagnosis of PFD and documented pelvic floor muscle weakness. Individuals with potential confounding factors were excluded to maintain a homogeneous cohort. Specifically, the exclusion criteria comprised concurrent neurological disorders, pregnancy, and malignancies, ensuring the statistical integrity of the retrospective analysis. Out of 204 screened medical records, 4 patients were excluded based on these predefined clinical exclusion criteria, resulting in a final analyzed cohort of 200 participants. This sample size (n = 200) fully satisfied the COSMIN recommendations for measurement properties, which prefer a minimum cohort of 100 participants to achieve robust statistical estimations and prevent overfitting in psychometric evaluations.

2.2. Intervention

According to the reviewed clinical records, all included patients had undergone a standardized 8‐week PFMT protocol administered by physiotherapists specializing in pelvic health. This intervention had been uniformly implemented across the three study sites in accordance with established international guidelines (Dumoulin et al. 2015), ensuring clinical consistency in the retrospective data. The training regimen consisted of two components: strength training, with 8–12 maximal contractions sustained for 1–2 seconds each; and endurance training with 15–20 submaximal contractions held for up to 10 seconds. A standardized 5–6 second rest period separated all contractions. Both strength and endurance exercises were performed as a single set within the training sequence. The daily volume was progressively increased: 3 sets per day for the initial 2 weeks, 4 sets per day for the subsequent 3 weeks, and 5 sets per day for the final 3 weeks. Although the protocol was conducted as a home‐based program, patients attended weekly clinical follow‐up sessions for supervised exercise monitoring and adherence assessment.

2.3. Outcome Measures

Baseline data acquisition extracted from the medical charts involved a comprehensive assessment of demographic characteristics, physical parameters, and urogynaecological clinical histories for all participants. Beyond these fundamental variables, two primary instruments, the Turkish PFDI‐20 and the Patient Global Impression of Improvement (PGI‐I) scale, were systematically utilized within the routine clinical workflow. The Turkish PFDI‐20 had been completed by patients at two distinct intervals: before the initiation of the 8‐week intervention (baseline) and immediately upon completion of the protocol (follow‐up). Following the 8‐week training period, participants' subjective perceptions regarding their global symptomatic change had been formally documented using the PGI‐I rating.

The distress associated with PFD was quantified using the Turkish PFDI‐20, a multidimensional self‐report instrument whose psychometric properties and cultural adaptation were previously validated by (Toprak et al. 2012). The Turkish PFDI‐20 is a comprehensive short‐form scale comprising 20 items organized into three symptom‐specific domains: the POPDI‐6, which assesses obstructive and prolapse‐related symptoms; the Colorectal‐Anal Distress Inventory‐8 (CRADI‐8), evaluating bowel and anorectal dysfunction; and the Urinary Distress Inventory‐6 (UDI‐6), focused on lower urinary tract symptoms. According to the official scoring manual, each item is rated based on the presence of the symptom: a “no” response receives 0 points, while a “yes” response is further rated by the severity of distress as not at all (1), somewhat (2), moderately (3), or quite a bit (4). In strict accordance with the scoring algorithm validated by (Toprak et al. 2012), the score for each subscale is calculated by obtaining the mean score of the answered items within that specific domain and multiplying it by 25. This mathematical transformation standardizes the subscale scores onto a scale ranging from 0 (best) to 100 (worst). The sum of these domain scores yields a global score ranging from 0 to 300, where higher numerical values indicate a more profound impact of symptoms and greater distress experienced by the patient, rather than the anatomical severity of the impairment itself.

The PGI‐I was utilized as a validated, generic PROM to evaluate the participants' perceived global symptomatic change recorded over the follow‐up period (Viktrup et al. 2012). The PGI‐I employs a 7‐point Likert‐type scale, requiring participants to assess their post‐treatment status relative to their baseline condition. Response options are categorized as “very much better”, “much better”, “a little better”, “no change”, “a little worse”, “much worse”, and “very much worse” (Viktrup et al. 2012). By integrating the participant's subjective evaluation, the PGI‐I serves as a critical anchor for capturing the clinical meaningfulness of the perceived changes from a holistic and patient‐centered perspective.

2.4. Statistical Analysis

Statistical analysis was performed using IBM SPSS Statistics Standard Concurrent User V‐26. The normality of the data distribution was assessed via the Shapiro‐Wilk test. Descriptive statistics were utilized to summarize the findings, with categorical variables expressed as frequencies (n) and percentages (%), and numerical variables reported as mean ± standard deviation (SD) based on their distribution. Crucially, as each subscale of the PFDI‐20 represents a distinct clinical domain, they were treated as independent primary variables. Consequently, no multiplicity adjustment was performed, a strategy consistent with established psychometric evaluation studies. Regarding missing data management, there were no missing data (0%) across the primary baseline and follow‐up assessments for the final analyzed cohort (n = 200), ensuring complete data integrity for all psychometric estimations in strict accordance with the COSMIN checklist.

2.4.1. Floor and Ceiling Effects

Floor and ceiling effects were assessed for the Turkish PFDI‐20 and POPDI‐6, CRADI‐8, and UDI‐6 subscales at both baseline and follow‐up time points. These effects were determined by calculating the percentage of participants who achieved the lowest (floor) and highest (ceiling) possible scores. Floor and ceiling effects were considered present if more than 15% of the respondents achieved the minimum or maximum possible scores, respectively (Terwee et al. 2007).

2.4.2. Responsiveness

To determine the responsiveness (i.e., sensitivity to change) of the Turkish PFDI‐20, a longitudinal comparative analysis between the two time points was performed. The Wilcoxon signed‐rank test was utilized to assess differences in paired scores, where a statistically significant p‐value indicated the instrument's capability to reflect longitudinal changes in symptom distress. This analysis was further augmented by calculating the effect size (ES) and the standardized response mean (SRM) to quantify the magnitude of change captured by the instrument (Cohen 1988). The ES was calculated by dividing the mean change by the baseline standard deviation and expressing the outcome in standard deviation units to interpret the magnitude of change relative to initial score variability. For instance, an ES of 0.30 signifies a mean change equivalent to approximately one‐third of the baseline standard deviation. In accordance with Cohen's widely accepted guidelines, the magnitude of the calculated ES was interpreted as small (0.2), moderate (0.5), or large (0.8) (Cohen 1988). Conversely, the SRM was utilized to reflect the magnitude of change relative to the variability of the changes themselves, rather than the initial population variability. Following established guidelines, SRM values were interpreted as follows: 0.2–0.5 as a small effect, 0.5–0.8 as a moderate effect, 0.8–1.0 as a good effect, and values greater than 1.0 as an excellent effect (Cohen 1988).

2.4.3. Smallest Detectable Change

Aligned with the COSMIN framework, the statistical equations for calculating measurement error and reliability ensure our SDC estimates match international psychometric standards (Gagnier et al. 2025; de Arruda et al. ²⁰²⁵). The clinical application of these SDC and MIC estimates is consistent with recent pelvic floor health evaluations that utilize these indicators to quantify sensitivity to change (Karshenas et al. 2026; Celenay et al. ²⁰²⁶). To determine the SDC, which distinguishes true score variance from measurement error at a 95% confidence level, a structured statistical approach was followed. The Standard Error of Measurement (SEM) was first calculated using the formula SEM = SD_pooled × √1–R. In this equation, the reliability coefficient (R) was defined as the Intraclass Correlation Coefficient (ICC_2,1), representing absolute agreement, derived from the test‐retest reliability analysis of the current study sample. Subsequently, the SDC was calculated as SDC₉₅ = SEM × 1.96 × √2. This methodology ensures that any observed change exceeds the boundaries of measurement noise, providing a preliminary estimate of the instrument's measurement error.

2.4.4. Minimal Important Change

The MIC was established using an anchor‐based approach, specifically employing Receiver Operating Characteristic (ROC) curve analysis. This methodology stratified participants into ‘improved' and ‘non‐improved’ cohorts based on their reported Patient Global Impression of Improvement (PGI‐I) scores. To dichotomize the 7‐point PGI‐I scale into a binary anchor state for the ROC analysis, participants who reported their post‐treatment status as “very much better”, “much better”, or “a little better” were classified into the “improved” cohort. Conversely, participants reporting “no change”, “a little worse”, “much worse”, or “very much worse” were categorized into the “non‐improved” cohort. To determine the optimal cut‐point for the MIC, the Youden index (J = sensitivity + specificity–1) was utilized, selecting the threshold score that maximized the mathematical balance between sensitivity and specificity, thereby minimizing the potential for misclassification between the groups (Turner et al. 2009). Furthermore, the Area Under the Curve (AUC) was calculated to assess the overall discriminative performance of the change scores. AUC values were interpreted according to established estimates: an AUC of 1.0 represents perfect discrimination, whereas 0.5 indicates performance no better than chance. Values ranging from 0.7 to 0.8 were considered fair, those from 0.8 to 0.9 were deemed good, and an AUC exceeding 0.9 was classified as excellent, in strict accordance with the diagnostic accuracy guidelines recommended (Çorbacıoğlu and Aksel 2023).

3. Results

A retrospective review of clinical records initially identified 204 patients with no missing data for potential inclusion in the study. Following the exclusion of 4 patients who met the clinical exclusion criteria, a final cohort of n = 200 participants with complete datasets (0% missing data) was included in the primary analysis. The detailed flow of participant recruitment, screening, and inclusion is visually represented in Figure 1. The participants' baseline physical, demographic, and clinical characteristics are presented in Table 1. Accordingly, the analysis of n = 200 participants revealed a mean age of 46.92 ± 11.13. The PGI‐I score indicated a high rate of perceived symptom change, with 93.0% of participants reporting positive global change at the follow‐up assessment.

FIGURE 1.

Flowchart of the participant.

TABLE 1.

Baseline physical, demographic, and clinical characteristics of participants.

Characteristics	n = 200
Age, Mean ± SD	46.92 ± 11.13
Body mass index, Mean ± SD	28.44 ± 5.54
Education n (%)
Literate	27 (13.5)
Primary school	74 (37.0)
Secondary school	37 (18.5)
High school	26 (13.0)
University	32 (16.0)
Postgraduate	4 (2.0)
Gravida, Mean ± SD	2.96 ± 1.89
Parity, Mean ± SD	2.23 ± 1.35
Menstrual status n (%)
Regular menstruation	91 (45.5)
Irregular menstruation	66 (33.0)
Natural menopause	37 (18.5)
Surgical menopause	6 (3.0)
Birth type, n (%)
Cesarean	33 (16.5)
Vaginal	123 (61.5)
Both	22 (11.0)
No birth	22 (11.0)
Pelvic floor dysfunction, n (%)
Urinary incontinence	199 (99.5)
Anal incontinence	183 (91.5)
Pelvic organ prolapse	124 (62.0)
Pelvic organ prolapse staging, n (%)
Cystocele
0	95 (47.5)
1	68 (34.0)
2	36 (18.0)
3	1 (0.5)
Rectocele
0	163 (81.5)
1	30 (15.0)
2	7 (3.5)
Uterine prolapse
0	168 (84.0)
1	22 (11.0)
2	9 (4.5)
3	1 (0.5)
Bladder diary, Mean ± SD
Daytime voiding frequency	11.16 ± 3.31
Nocturnal voiding frequency	1.28 ± 0.84
Voided volume	191.98 ± 87.03
Urinary incontinence episodes	1.58 ± 1.72
Patient global impression of improvement, n (%)
Very much worse	0 (0.0)
Much worse	0 (0.0)
A little worse	1 (0.5)
No change	13 (6.5)
A little better	52 (26.0)
Much better	77 (38.5)
Very much better	57 (28.5)
Patient global impression of improvement, n (%)
Improved	186 (93.0)
Not improved	14 (7.0)

Abbreviation: SD, Standard deviation.

3.1. Floor/Ceiling Effects

The floor and ceiling effects for the Turkish PFDI‐20 and its subscales at baseline and follow‐up assessments are detailed in Table 2. For the total score and all subscales (POPDI‐6, CRADI‐8, and UDI‐6), the observed floor and ceiling effects remained low, with the highest floor effect observed in the POPDI‐6 at follow‐up assessment (12.0%) and the highest ceiling effect in the UDI‐6 at baseline (3.5%). Since all observed values remained well below the widely accepted 15% threshold, the PFDI‐20 and its subscales demonstrate sufficient measurement breadth to capture score variance within this study population without significant data skewness at the extreme ends of the scale.

TABLE 2.

Floor and ceiling effects of the PFDI‐20 and its subscales at baseline and follow‐up.

n = 200	Baseline floor effect (%)	Follow‐up floor effect (%)	Baseline ceiling effect (%)	Follow‐up ceiling effect (%)
PFDI‐20	0.5%	0.5%	0.5%	0.5%
POPDI‐6	9.0%	12.0%	0.5%	0.5%
CRADI‐8	8.5%	10.0%	0.5%	0.5%
UDI‐6	0.5%	2.0%	3.5%	0.5%

Abbreviations: CRADI‐8, Colorectal Anal Distress Inventory‐8; PFDI‐20, Pelvic Floor Distress Inventory‐20; POPDI‐6, Pelvic Organ Prolapse Distress Inventory‐6; UDI‐6, Urinary Distress Inventory‐6.

3.2. Responsiveness

The responsiveness of the Turkish PFDI‐20 and its subscales reflecting score variations between baseline and follow‐up assessments is presented in Table 3. The magnitude of longitudinal score changes captured by the instrument was higher for the total PFDI‐20 and UDI‐6 than for the POPDI‐6 and CRADI‐8. For PFDI‐20 and UDI‐6, the instrument's sensitivity to detect change was good (ES: 1.24, SRM: 1.41, p < 0.001, for PFDI; ES: 1.47, SRM: 1.53, p < 0.001, for UDI‐6), while it was moderate to good (ES: 0.78, SRM: 0.94, p < 0.001) and moderate (ES: 0.55, SRM: 0.76, p < 0.001) for POPDI‐6 and CRADI‐8, respectively.

TABLE 3.

Mean change in scores and responsiveness of the Turkish PFDI‐20 and subscales.

n = 200	Baseline	Follow‐up	p value	Mean change in score	ES (95% CI)	SRM (95% CI)
PFDI‐20	117.26 ± 45.74	65.23 ± 37.49	< 0.001	52.02 ± 36.66	1.24 (1.08–1.40)	1.41 (1.23–1.59)
POPDI‐6	33.71 ± 21.23	18.96 ± 15.93	< 0.001	14.75 ± 15.53	0.78 (0.63–0.93)	0.94 (0.79–1.09)
CRADI‐8	24.64 ± 17.65	15.71 ± 14.15	< 0.001	8.93 ± 11.66	0.55 (0.41–0.69)	0.76 (0.62–0.90)
UDI‐6	58.90 ± 21.15	30.60 ± 16.86	< 0.001	28.30 ± 18.47	1.47 (1.30–1.64)	1.53 (1.35–1.71)

Abbreviations: CI, Confidence Interval; CRADI‐8, Colorectal Anal Distress Inventory‐8; ES, Effect size; PFDI‐20, Pelvic Floor Distress Inventory‐20; POPDI‐6, Pelvic Organ Prolapse Distress Inventory‐6; SRM, Standardized response mean; UDI‐6, Urinary Distress Inventory‐6.

3.3. Smallest Detectable Change and Minimal Important Change

The ICC, SEM, SDC, MIC, and AUC with sensitivity and specificity for the PFDI‐20 and its subscales are presented in Table 4. The ICC values (with 95% CI) were 0.78 (0.71–0.83), 0.67 (0.58–0.74), 0.64 (0.55–0.72), and 0.61 (0.51–0.69) for PFDI‐20, POPDI‐6, CRADI‐8, and UDI‐6, respectively. The SEM values were 20.05, 8.43, 6.39, and 10.51 for PFDI‐20, POPDI‐6, CRADI‐8, and UDI‐6, respectively. The SDC values were 55.60, 23.39, 17.71, and 29.13 for PFDI‐20, POPDI‐6, CRADI‐8, and UDI‐6, respectively (noting that these values, derived from an active intervention cohort, reflect both measurement error and longitudinal clinical changes). The MICs for identifying perceived clinical status changes were 20.83, 20.83, 9.37, and 25.00 for PFDI‐20, POPDI‐6, CRADI‐8, and UDI‐6, respectively. The ROC analyses showed moderate to good discriminative ability, with an area under the curve of 0.85 (sensitivity 84.0%, specificity 86.0%), 0.72 (sensitivity 30.0%, specificity 93.0%), 0.72 (sensitivity 38.0%, specificity 93.0%), and 0.85 (sensitivity 50.0%, specificity 93.0%) for PFDI‐20, POPDI‐6, CRADI‐8, and UDI‐6, respectively. Figure 2 provides the ROC curves of the PFDI‐20 and its subscales.

TABLE 4.

Preliminary estimates of ICC, SEM, SDC, and MIC for the Turkish PFDI‐20 and subscales derived from an active intervention cohort.

n = 200	ICC (95% CI)	SEM	SDC (95% CI)	MIC	AUC %95 CI (bootstrap) (Sensitivity, specificity)
PFDI‐20	0.78 (0.71–0.83)	20.05	55.60 (50.28–62.51)	20.83	0.85 (0.74–0.92) (0.84, 0.86)
POPDI‐6	0.67 (0.58–0.74)	8.43	23.39 (21.21–26.31)	20.83	0.72 (0.50–0.91) (0.30, 0.93)
CRADI‐8	0.64 (0.55–0.72)	6.39	17.71 (16.03–19.98)	9.37	0.72 (0.53–0.88) (0.38, 0.93)
UDI‐6	0.61 (0.51–0.69)	10.51	29.13 (26.30–32.88)	25.00	0.85 (0.71–0.93) (0.50, 0.93)

Note: Values for ICC, SEM, and SDC were derived from an active intervention cohort undergoing structured treatment rather than a clinically stable population. Therefore, these parameters should be interpreted as preliminary, intervention‐derived estimates rather than conventional measurement‐error thresholds.

Abbreviations: AUC, Area under the curve; CI, Confidence Interval; CRADI‐8, Colorectal Anal Distress Inventory‐8; ICC, Intraclass Correlation Coefficient; MIC, Minimal Important Change; PFDI‐20, Pelvic Floor Distress Inventory‐20; POPDI‐6, Pelvic Organ Prolapse Distress Inventory‐6; SDC, Smallest Detectable Change; SEM, Standard Error of Measurement; UDI‐6, Urinary Distress Inventory‐6.

FIGURE 2.

The receiver‐operating‐characteristic curves of the Pelvic Floor Distress Inventory‐20, Pelvic Organ Prolapse Distress Inventory‐6, Colorectal Anal Distress Inventory‐8, and Urinary Distress Inventory‐6.

4. Discussion

In this study, the Turkish PFDI‐20 demonstrated moderate to good responsiveness in capturing longitudinal score variations in women with PFD. The preliminary SDC values of the Turkish PFDI‐20 and its subscales (POPDI‐6, CRADI‐8, UDI‐6) were 55.60, 23.39, 17.71, and 29.13, respectively. Furthermore, the preliminary MIC values of the Turkish PFDI‐20 and its subscales (POPDI‐6, CRADI‐8, UDI‐6) were 20.83, 20.83, 9.37, and 25.00, respectively.

The current psychometric evaluation of the PFDI‐20 contributes to evidence supporting its utility as a reliable PROM to track longitudinal changes in PFD. In the present study, the Turkish PFDI‐20 and its subscales demonstrated moderate to good responsiveness between the assessment time points. This finding exhibits favorable external validity, aligning consistently with previous psychometric studies across multiple linguistic adaptations (e.g., Dutch, Spanish, Norwegian) (Utomo et al. 2014; Sánchez et al. ²⁰¹⁵; Teig et al. ²⁰¹⁷), and is comparable even to values observed in reconstructive surgery cohorts (Kaplan et al. 2012). Crucially, these responsiveness indices address a major gap highlighted in the global literature by (de Arruda et al. ²⁰²¹), who pointed out that high‐quality longitudinal responsiveness data for the PFDI‐20 remain remarkably scarce internationally (de Arruda et al. ²⁰²¹), thereby providing valuable evidence that prior global syntheses found lacking. A notable limitation observed in our findings is the lower responsiveness of the CRADI‐8 subscale compared to POPDI‐6 and UDI‐6 (Utomo et al. 2014; Kaplan et al. ²⁰¹²; M. Barber et al. ²⁰⁰⁵; Sánchez et al. ²⁰¹⁵). This recurring pattern points to an inherent psychometric constraint within the instrument itself, emphasizing that clinicians should interpret colorectal‐anal symptom changes with caution and underscoring the need for more sensitive measures in daily practice (National Institute for and Health and Care Excellence (Great Britain) ²⁰¹⁰).

Unlike prior validation efforts focused primarily on responsiveness, this study provides rigorously calculated SDC and MIC thresholds for the Turkish PFDI‐20 following a structured PFMT intervention. The calculated SDC values (55.60 for the total score and 29.13 for the UDI‐6) establish essential thresholds to distinguish true clinical recovery from measurement variability. These preliminary estimates are vital for clinicians and researchers to interpret score fluctuations with confidence, ensuring that treatment responses are evaluated with clear clinical relevance.

The MIC provides a vital clinical threshold by quantifying the magnitude of patient‐perceived improvement (Terwee et al. 2021). Addressing a global deficit highlighted by (de Arruda et al. ²⁰²¹) regarding under‐reported subscale metrics (de Arruda et al. ²⁰²¹), our calculated total PFDI‐20 MIC (20.83) aligns well with international literature, such as the Dutch adaptation (22.9) and conservative cohorts (13.5) (Utomo et al. 2014; Wiegersma et al. ²⁰¹⁷). Crucially, establishing subscale‐specific MICs (POPDI‐6: 20.83; CRADI‐8: 9.37; UDI‐6: 25.00) supports guideline‐driven practice emphasizing multi‐compartmental assessment (National Institute for and Health and Care Excellence (Great Britain) ²⁰¹⁰). These distinct thresholds show that meaningful change varies across pelvic domains, where a 25.00‐point change is needed for urinary symptoms (UDI‐6) compared to 9.37 points for colorectal‐anal distress (CRADI‐8). In conclusion, the SDC and MIC values established for the Turkish PFDI‐20 offer clinically meaningful preliminary estimates that account for both measurement error and true clinical recovery. These metrics provide a practical reference for routine clinical monitoring in Turkish‐speaking populations while warranting cautious interpretation in light of study limitations.

A notable finding is that the MIC values were generally lower than their respective SDC thresholds, indicating that patient‐perceived improvements can manifest within measurement error margins. Although an MIC exceeding the SDC is traditionally preferred for individual tracking, an MIC smaller than the SDC is a recognized phenomenon in pelvic floor literature, showing alignment with Dutch validation and conservative management cohorts (Utomo et al. 2014; Wiegersma et al. ²⁰¹⁷). While warranting cautious interpretation due to anchor distribution constraints, contextualizing this balance between subjective patient perception and statistical precision addresses the global reporting deficit highlighted by (de Arruda et al. ²⁰²¹), offering valuable guidance for clinical responsiveness in pelvic floor health.

To ensure international standardization and fulfill COSMIN criteria, the longitudinal design and predefined hypotheses of this study establish a rigorous framework for evaluating responsiveness. For the anchor‐based MIC methodology, aligning our analysis with an external patient‐reported criterion provides a clinically relevant reference point. Although COSMIN guidelines suggest optimal precision with evenly distributed cohorts, a limitation carefully considered regarding our distribution asymmetry, analyzing parameters such as the SDC offers crucial estimates for interpreting longitudinal score variations within active treatment timelines. Integrating these structural perspectives provides a transparent, methodologically sound foundation for cautiously interpreting the longitudinal utility of the questionnaire.

A primary strength of this study is its comprehensive methodological framework, establishing the first SDC and MIC values for the Turkish PFDI‐20 under conservative management. A multi‐center design across three sites enhances generalizability, while standardized PFMT protocols administered by specialist physiotherapists ensure evaluation under standardized clinical conditions. Furthermore, favorable AUC values, particularly for the total PFDI‐20 and UDI‐6, demonstrate strong discriminative power, providing clinicians with a reliable tool to track longitudinal symptom changes.

Despite its contributions, several limitations must be acknowledged. First, calculating reliability indices (ICC, SEM, SDC) from an active intervention cohort reflects both measurement error and true clinical change, requiring cautious interpretation per COSMIN guidelines. Second, the retrospective design introduces potential selection bias though mitigated by a large sample (n = 200) and stringent criteria. Third, while the PGI‐I is a validated anchor, its subjective nature and severe group asymmetry (93% improved vs. 7% non‐improved) can influence ROC precision, meaning MIC thresholds and AUC values should be applied cautiously in cohorts with lower response rates. Fourth, the absence of quantitative adherence logs leaves home‐based compliance as an unmeasured confounding variable. Fifth, the lack of diagnostic stratification in our heterogeneous cohort prevents subgroup analyses that might capture specific subscale variations (POPDI‐6 or CRADI‐8). Lastly, the exclusive focus on the PFDI‐20 without concurrent Pelvic Floor Impact Questionnaire‐7 (PFIQ‐7) evaluation limits broader differentiation between symptom changes and health‐related quality of life.

In light of these limitations, future research should prioritize prospective longitudinal designs to evaluate the long‐term stability of the SDC and MIC thresholds across diverse settings. Specifically, test‐retest reliability studies should be conducted on a stable cohort to establish a pure measurement error baseline independent of treatment effects. To address unquantified adherence, future trials should incorporate objective tracking tools, such as digital logs or smart sensor‐based devices, to capture precise treatment exposure. Furthermore, researchers should stratify participants by specific PFD phenotypes and diagnostic categories (isolating urinary/anal incontinence and prolapse stages) to determine whether psychometric estimates remain consistent. Integrating objective physiological parameters, such as urodynamics, anorectal manometry, or dynamic ultrasonography, would also offer a more comprehensive validation. Additionally, future trials should recruit balanced response spectrums or use advanced statistical adjustments to mitigate spectrum bias. Lastly, research should concurrently integrate the PFDI‐20 and the PFIQ‐7 to bridge symptom changes with health‐related quality of life.

5. Implications of Physiotherapy Practice

The Turkish PFDI‐20 and its subscales demonstrate favorable responsiveness, with promising indices and significant proportions of patients exceeding SDC and MIC thresholds. From a clinical perspective, these findings provide preliminary estimates for monitoring patient‐reported changes during conservative physiotherapy and rehabilitation. Specifically, because these indices stem from an active intervention cohort rather than a stable population, clinicians can appropriately and tentatively use the derived total SDC threshold of 55.60 as a cautious preliminary estimate accounting for both measurement variability and active intervention improvements. Furthermore, given potential instabilities from imbalanced anchor groups (93% improved vs. 7% non‐improved), observed changes meeting the MIC criterion of 20.83 should be viewed cautiously as preliminary estimates of meaningful patient‐perceptible outcomes rather than rigid clinical cutoff scores.

These subscale values underscore the need for domain‐specific evaluations of prolapse, colorectal, and urinary distress rather than reliance on global scores alone. When applying these findings in daily clinical routines, practitioners must carefully weigh individual patient presentations, treatment timelines, and the acknowledged methodological constraints. Ultimately, this balanced approach will empower clinicians to make measurement‐driven, rigorous assessments and enhance patient‐centered monitoring while avoiding overinterpretation of the longitudinal data.

Funding

The authors have nothing to report.

Ethics Statement

This retrospective study received ethical clearance from the Izmir Katip Celebi University Health Research Ethics Committee (Approval Date: 15 May 2025; Approval Number: 0274), with the requirement for informed consent waived due to the study's design.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.