Abstract
Aim
Quality of life (QoL) is a crucial and core outcome in assessing the effectiveness of treatments for cryptoglandular anal fistula. Despite its extensive impact, there is a lack of patient‐centred, disease‐specific QoL measurement instruments of adequate quality. The aim of this study is to develop a disease‐specific measurement instrument that can accurately measure QoL for patients with cryptoglandular anal fistula.
Method
Semi‐structured qualitative patient interviews and a systematic review of current instruments were used to generate items for the draft instrument. This underwent successive rounds of cognitive interviews to refine its wording and structure. Individual item and overall scale content validity were determined by asking experts to rate the relevance of each item and those deemed irrelevant were removed. The final instrument then underwent psychometric testing and test–retest analysis to determine its sensitivity and stability.
Results
A total of 148 patients were involved in item generation, scale development and psychometric testing. A 22‐item measurement instrument has been developed; it is scored on a scale of 0–100, where 0 indicates the worst QoL and 100 demonstrates perfect QoL. The scale demonstrates excellent internal consistency (Cronbach‐α = 0.927), strong content and construct validity [correlation with Perianal Disease Activity Index = −0.713, Hospital Anxiety and Depression Anxiety (−0.659) and Depression (−0.673) subscales and Short Form‐12 physical (0.609) and mental (0.589) component scales] and strong reliability and responsiveness.
Conclusion
We have developed a cryptoglandular Anal Fistula Quality of Life scale (AF‐QoL), a comprehensive, disease‐specific patient reported outcome measure assessing QoL in patients with cryptoglandular anal fistula.
Keywords: anal fistula, patient reported outcome measures, proctology, PROM, QoL, quality of life
What does this paper add to the literature?
We describe the development and psychometric testing of a patient‐centred, disease‐specific quality of life measurement instrument for cryptoglandular anal fistula. The scale demonstrates excellent sensitivity and stability, and is intended for evaluative purposes as a patient reported outcome measure in clinical practice and trials and studies of fistula treatment.
INTRODUCTION
The study of cryptoglandular anal fistula has been a consistent focus in colorectal research owing to the challenges involved in its assessment and management. An under‐reported element of the disease is its impact on patient QoL, which is known to be extensive and debilitating [1]. More recently, an international consensus of patients, researchers and clinicians experienced in managing anal fistula has identified QoL as a critical outcome that should be measured in all interventional studies of fistula treatment [2]. Despite this, only 14% of studies have measured QoL as an outcome [3], mostly through the use of generic measurement instruments, with the finding that QoL in patients with a fistula is lower than in the general population [4]. A reduction in QoL correlates with faecal incontinence [5], but studies have found varied results depending on severity. Patients with minor impairment have shown no reduction in QoL posttreatment [6] or following fistula healing; however, QoL is reduced for patients with severe deterioration in continence after surgery [7]. This complexity shows that continence cannot be used as a surrogate for QoL, as the benefits of successful surgery and symptom resolution may outweigh the impact of continence disturbance [7].
There is a clear need for disease‐specific instruments that demonstrate greater sensitivity to changes in QoL and can directly detect the impact of fistulas and their treatment [8]. Few disease‐specific instruments have been designed for cryptoglandular anal fistula. However, the lack of patient involvement in their development and a failure to adequately capture the extensive impact of disease have led to doubts regarding their content validity [9].
This study aimed to develop a disease‐specific, patient‐centred Anal Fistula Quality of Life (AF‐QoL) assessment tool for patients with cryptoglandular anal fistula.
METHOD
The AF‐QoL scale was developed in accordance with the COnsensus‐based Standards for the selection of health Measurement INstruments (COSMIN) standards for patient reported outcome measure (PROM) development studies [10]. Qualitative and quantitative methods were used and ethical approval was granted by West Midlands–Black Country Research Ethics Committee (REC reference 19/WM/0296) and Wales Research Ethics Committee 6 (REC reference 20/WA/0055), and also approved by the Health Research Authority and Health and Care Research Wales.
AF‐QoL study management team
AF‐QoL development was overseen by the study management team, consisting of colorectal surgeons experienced in assessing cryptoglandular anal fistula, qualitative researchers with experience in PROM development and patient representatives.
Definition of construct and conceptual framework
The aim of the AF‐QoL is the subjective assessment of the specific and direct impact of having a fistula or its treatment. The QoL definition guiding PROM development is ‘those aspects of self‐perceived well‐being that are related to or affected by the presence of disease or treatment’ [11]. Instrument design is based on Wilson and Cleary's model of health‐related QoL [12, 13], which describes causal links between biological function and physiological factors (e.g. the presence of a fistula) with symptoms such as pain and discharge, which directly impact functional status, ultimately affecting an individual's general health status and their global QoL [13, 14]. This model is particularly relevant to fistula patients due to the interrelated nature of the domains, its practical application and the broad view of QoL extending beyond physiological factors [1].
Target population and scope
AF‐QoL aims to measure QoL in patients with a cryptoglandular or idiopathic anorectal fistula, without restriction in disease complexity, chronicity or patient demographics. The scale is developed for evaluative purposes in clinical practice (including assessing the impact of treatment), clinical trials and studies of fistula treatment.
Participant selection
Patients aged 18–80 years with an idiopathic or cryptoglandular anorectal fistula, capacity to provide informed consent and understanding of written and spoken English were eligible. Patients with Crohn's anal fistula were excluded. A purposive sampling strategy was employed to ensure variation in disease complexity and duration. Participants were recruited from a tertiary centre in London, UK, specialized in treating complex anal fistula and internationally via social media. Different cohorts of patients were recruited for each stage of the study (see below).
Stage 1: Item generation
AF‐QoL items were generated using both inductive and deductive methods, including semi‐structured qualitative interviews and a systematic review of measurement instruments [9].
Qualitative interview methods have been described elsewhere [1]. Themes relating to the predefined QoL domain were extracted from interview analysis and formulated into draft items. These were supplemented with findings from the systematic review of disease‐specific QoL measurement instruments [9] and any additional items were included. The study management team ensured that all items were sufficiently unique and represented QoL and placed no restrictions on the number of items to be included. The readability of the draft instrument was assessed by assigning a Flesch Reading Ease score between 1 and 100, with 100 denoting the highest readability score and text scoring between 70 and 80 determined to be easy for the average adult to read. The Flesch–Kincaid grade level was also assessed to determine the approximate reading grade level for an item of text.
Stage 2: Scale development
Cognitive interviews
Participants were interviewed while completing the draft questionnaire and verbalized their responses. Verbal probing was used to identify difficulties in understanding questions or response options. The relevance, comprehensiveness and comprehensibility of the instrument [10] was assessed using specific prompts shown in supplementary information Table S1.
Cognitive interviews were conducted by clinician researchers, audio‐recorded and transcribed verbatim. Rounds of three to five interviews were conducted, and after each round the draft scale was revised according to the feedback received. Rounds continued until no further changes were suggested in three consecutive interviews, this version of AF‐QoL was assessed for content validity.
Independent experts in QoL and anal fistula were asked to assess each draft item and rank its relevance in assessing QoL on a four‐point Likert scale (1 = not at all relevant, 4 = extremely relevant). The item‐level content validity index (I‐CVI) of each item was calculated and those with an I‐CVI < 0.78 were discussed amongst the study management team for exclusion [15]. The scale‐level content validity index was calculated by averaging the I‐CVIs for the remaining items on the scale (S‐CVI/Ave). A scale with excellent content validity should have an S‐CVI/Ave of 0.90 or higher [15].
Stage 3: Item reduction and psychometric analysis
The draft AF‐QoL instrument was disseminated amongst a sample of the target population. Demographic information was collected, after which participants completed composite anchor questions (Table S2) consisting of a self‐reported score rating the severity of their fistula symptoms and QoL, alongside the draft AF‐QoL, Short Form‐12 (SF‐12) [16], Hospital Anxiety and Depression Scale (HADS) [17] and the patient reported component of the Perianal Disease Activity Index (PDAI) [18]. Participants were asked to complete these same tests 2–3 weeks later.
Item reduction
After completing AF‐QoL on the index occasion, item–total and item–item correlations (ITC and IIC) were conducted to eliminate those items that were too closely associated with the overall AF‐QoL score or with other items on the instrument to minimize scale redundancy. Stepwise regression was conducted to identify the best grouping of questions explaining most of the variance in the overall AF‐QoL score. The study management team discussed whether the outlying items could be excluded.
Psychometric analysis
Internal consistency was determined using the Cronbach‐α statistic, with an acceptable value being between 0.7 and 0.9 [19]. Construct validity was assessed by correlating AF‐QoL scores with other validated measures of mental well‐being, symptom severity and functional status. Pearson's correlation coefficient was determined for AF‐QoL and the HADS anxiety subscale (HADS‐A), HADS depression subscale (HADS‐D), the physical (PCS‐12) and mental (MCS‐12) component scores of SF‐12 and patient‐reported PDAI scores. An appropriate correlation coefficient is between 0.4 and 0.8 [20, 21, 22].
Principal component analysis (PCA) was used to identify AF‐QoL subscales, the method for which is outlined in Data S1.
The final instrument was assessed for floor and ceiling effects, where more than 15% of participants reported the lowest or highest possible scores, suggesting that the scale is unable to discriminate those with extreme responses [22].
Test–retest analysis
Participants completed the same series of questionnaires 2–3 weeks later. Those experiencing a change in their symptoms and QoL were identified by composite anchor questions summarized in Table S2.
Reliability
The Pearson correlation coefficient was calculated for the two sets of observations taken from those reporting stable symptoms (Table S2), with a value greater than 0.75 suggesting an acceptable correlation [20, 21, 23]. The intraclass correlation (ICC) was calculated using an analysis of variance and expressed as a number between −1 and 1, where −1 means both sets of results change inversely, 0 indicates no reliability and 1 suggests perfect reliability. An acceptable ICC is ≥0.70 [22].
Responsiveness
Patients reporting any change in their symptoms in the retest phase were included in responsiveness analyses described in Table S3. The effect size (ES) measures effectiveness in detecting change [19, 23], with small (0.2–0.49), moderate (0.5–0.79) and large (≥0.8) ESs described in the literature [24]. The responsiveness ratio (RR) accounts for nonspecific score changes in those who are clinically unchanged [23]. The standardized response mean (SRM) is the difference in means between test and retest for those reporting change in symptoms divided by the SD of change scores. A more sensitive instrument has a larger SRM [23, 25], and an SRM >0.8 indicates a large ES [26].
The retest sample was divided into those who reported no change, an improvement or worsening of symptoms. Mean changes between test and retest scores were calculated with corresponding confidence intervals (CIs) and compared using the paired t‐test. A p‐value < 0.05 was determined to be statistically significant.
Interpretability
Patients completed a self‐reported visual analogue score reporting how bothersome their fistula was on a scale of 1–10 (1 = not bothered at all, 10 = extremely bothersome). The minimally important change (MIC) is the mean change of AF‐QoL score for those who reported a change of one point and is the smallest difference in AF‐QoL score that would determine a change in management or demonstrate that an intervention had made a difference [27].
The final items retained for AF‐QoL were discussed amongst the study management group to ensure clinical relevance and comprehensiveness, readability and acceptability.
RESULTS
A total of 148 patients participated in the various stages of the study; their demographics are summarized in Table 1.
TABLE 1.
Participant demographics.
| Stage 1: qualitative interviews (N = 10), n (%) | Stage 2: cognitive interviews (N = 18), n (%) | Stage 3: psychometric testing (N = 120), n (%) | |
|---|---|---|---|
| Gender | |||
| Male | 5 (50) | 9 (50) | 46 (38) |
| Female | 5 (50) | 9 (50) | 74 (62) |
| Mean age (years) (SD) | 44.2 (1) | 47.2 (12.9) | 45 (11.8) |
| Duration of diagnosis (months) | |||
| 0–24 | 5 (50) | 4 (22) | 38 (32) |
| 25–48 | 1 (10) | 4 (22) | 33 (28) |
| >48 | 3 (30) | 6 (33) | 48 (40) |
| Unknown | 1 (10) | 4 (22) | 1 (1) |
| Number of previous procedures | |||
| 0–1 | 6 (60) | 2 (11) | 20 (17) |
| 2–4 | 4 (40) | 12 (67) | 48 (40) |
| >4 | 0 | 4 (22) | 52 (43) |
| Active fistula | |||
| Yes | 106 (88) | ||
| No | 2 (2) | ||
| Not sure | 12 (10) | ||
| Stoma | 7 (6) | ||
| Mean SF‐12 scores | |||
| Mean PCS‐12 (SD) | 40.9 (10.9) | ||
| Mean MCS‐12 (SD) | 40.1 (11) | ||
Abbreviations: MCS‐12, mental component score of SF‐12; PCS‐12, physical component score of SF‐12; SF‐12, Short Form‐12; SD, standard deviation.
Stage 1: Item generation
Interviews from 10 patients were used for item generation [1]. Extracted items related to physical symptoms, life impact and treatment. Items from preexisting measurement instruments were already included in the extracted themes.
The study management team formulated draft AF‐QoL items, response options, introductory text and instructions. The instrument recall period was agreed to be 4–6 weeks to obtain an average overview of fistula impact and avoid recall bias. The early draft of AF‐QoL to be used in cognitive interviewing consisted of 44 items, the Flesch reading ease score was 86.4 and the Flesch–Kincaid grade level was 2.4 (reading age 7–8 years).
Stage 2: Scale development
Cognitive interviews
A total of 18 interviews were conducted over four rounds. Revisions were related to wording and six additional questions were suggested. Verbal completion time ranged between 5 min in the final interviews to 45 min in earlier interviews, reflecting fewer changes suggested in final rounds of testing.
A total of seven experts assessed content validity for the draft AF‐QoL. The mean individual item rankings (I‐CVI) and exclusions are shown in Table S4. The study management team decided that 5 of 14 items with an I‐CVI of <0.78 should be retained as these were emphasized as being important determinants of QoL during patient interviews. Nine items were excluded (Table S4). The overall S‐CVI/Ave following exclusions was 0.88, and it was this 41‐item scale that underwent psychometric testing.
Stage 3: Psychometric testing
A total of 120 participants completed AF‐QoL and the validated scales and 78 participants completed AF‐QoL on a second occasion for retest analysis. Participants demonstrated lower than average QoL as determined by mean SF‐12 scores (average US scores = 50.0; Table 1).
Scoring AF‐QoL
The draft AF‐QoL contained 10 questions with a ‘not applicable’ answer option because they addressed conditional situations, such as the presence of a seton or being in a relationship. The study management team determined that the overall AF‐QoL score would be the average of the applicable answers. Each response was scored in percentiles, so that the overall AF‐QoL score would be transformed into a 0–100 scale, with a higher score indicating a better QoL.
Item reduction and psychometric analysis
Primary exclusions after ITC and IIC are summarized in Table 2. Despite strong ITC for being worried about needing a stoma bag, the study management group agreed to retain the item due to its emphasis in patient interviews. Similarly, the study management group agreed to keep items regarding self‐consciousness and embarrassment as they were felt to represent different concepts, particularly by patient representatives.
TABLE 2.
Comparison of 41‐, 34‐, 21‐ and 22‐item versions of the AF‐QoL scale.
| Question | Primary exclusions | 34‐item AF‐QoL | 21‐item AF‐QoL (after stepwise regression) | 22‐item AF‐QoL (after SMG discussion) | |
|---|---|---|---|---|---|
| 1 | How often have you experienced pain or discharge from your fistula? | ||||
| 2 | How severe has the pain from your fistula been? | ||||
| 3 | How often have you experienced discharge of fluid, pus or blood from your fistula? | Included due to importance of discharge | |||
| 4 | How often have you been worried about not being able to control the passage of wind or leaking stool from your back passage or fistula? | ||||
| 5 | How often have you been bothered by your fistula symptoms being random and unpredictable? | ||||
| 6 | How often have your fistula symptoms interfered with everyday activity such as walking, sitting, driving etc.? | ||||
| 7 | How many nights have your fistula symptoms made it difficult for you to sleep? | ||||
| 8 | Changing dressings or gauzes interfere with my planned social or work‐related activities | ||||
| 9 | My fistula stops me from, or interferes with playing sports or physical activities | ||||
| 10 | I avoid going out with my friends because of my fistula symptoms | ||||
| 11 | Because of my fistula, I worry about finding or needing the toilet when I am away from home | ||||
| 12 | Because of my fistula, I worry about finding somewhere where I can clean myself when I am away from home | Q11 IIC = 0.871 | |||
| 13 | I worry about the smell of the discharge from my fistula | ||||
| 14 | The practical demands of managing my fistula have an impact on my daily activities | ITC = 0.929 | |||
| 15 | How many days have you had to take time off from work or school/university because of your fistula? | ||||
| 16 | I find it difficult to commute to and from work and other places because of my fistula | ||||
| 17 | I have changed the way I work in order to manage my fistula symptoms (e.g. practical adjustments to sitting, standing more, changing working times or the tasks you do at work) | ||||
| 18 | I have needed frequent courses of antibiotics to manage infections associated with my fistula | ITC = 0.386 | |||
| 19 | I am worried about developing an abscess | ||||
| 20 | I am worried about needing further surgery for my fistula | ||||
| 21 | I worry about becoming incontinent as a result of my fistula or surgery for my fistula | ||||
| 22 | I am worried about needing a stoma bag | ITC = 0.388 | |||
| 23 | I am worried that I might need a seton forever | ||||
| 24 | I am physically more comfortable with a seton inserted, compared to having a fistula without a seton | ITC = −0.149 | |||
| 25 | I cannot accept my seton as a permanent part of my life | ||||
| 26 | I find it harder to manage dressings with a seton | ||||
| 27 | My seton is painful | ||||
| 28 | I find it difficult to keep my seton clean | ||||
| 29 | Having a seton makes me more anxious about my fistula | ||||
| 30 | My quality of life is worse with a seton, compared to having a fistula without a seton | ITC = 0.369 | |||
| 31 | How much of the time have you forgotten that you have a seton in | ||||
| 32 | There are members of my family and/ or friends from whom I feel isolated as a result of my fistula and its impact | ||||
| 33 | I depend on my partner or another person to help me manage my fistula (e.g. changing dressings, cleaning yourself) and other tasks (e.g. daily chores, driving to places) | ||||
| 34 | I worry about the impact that my fistula symptoms are having on my partner | ||||
| 35 | My desire for sex and intimacy has been reduced by my fistula | Q36 IIC = 0.859 | |||
| 36 | My fistula has stopped me from having sex or being intimate | ||||
| 37 | My fistula makes me self‐conscious about my body | Q38 IIC = 0.806 | |||
| 38 | I feel embarrassed by my fistula | Q37 IIC = 0.806 | |||
| 39 | I feel sad and/ or upset about my fistula, its symptoms and the impact they are having | ||||
| 40 | My fistula makes me worried and uncertain about the future | ITC = 0.82 | |||
| 41 | I feel guilty about the impact my fistula symptoms are having on my life | ||||
| Internal consistency (Cronbach's α) | 0.950 | 0.922 | 0.927 | ||
| Construct validity | |||||
| PDAI | −0.688 | −0.705 | −0.713 | ||
| HADS‐A | −0.680 | −0.661 | −0.659 | ||
| HADS‐D | −0.677 | −0.676 | −0.673 | ||
| SF‐12 PCS | 0.584 | 0.609 | 0.609 | ||
| SF‐12 MCS | 0.618 | 0.587 | 0.589 | ||
| AF‐QoL 34 | 0.987 | 0.988 | |||
Note: Shaded squares indicate excluded items.
Abbreviations: AF‐QoL, anal fistula quality of life scale; HADS‐A, Hospital Anxiety and Depression Scale Anxiety Subscale; HADS‐D, Hospital Anxiety and Depression Scale Depression Subscale; IIC, item–item correlation; ITC, item–total correlation; PDAI, Perianal Disease Activity Index; SF‐12 MCS, Short Form 12 mental component score; SF‐12 PCS, Short Form‐12 physical component score; SMG, study management group.
After primary exclusions, stepwise regression was performed on the 34‐item scale. This identified 21 items accounting for most of the variation in the AF‐QoL score. The study management team compared the 34‐ and 21‐item instruments, and it was determined that the item referencing the frequency of discharge should be retained. A comparison of the 41‐, 34‐, 21‐ and 22‐item instruments is shown in Table 2. All versions of AF‐QoL show acceptable internal consistency and construct validity, correlating well with PDAI, HADS‐A, HADS‐D and the component scores of SF‐12. Furthermore, the shortened versions correlated well with the 34‐item version (Table 2).
The study management team agreed that the 22‐item version of AF‐QoL covered all key domains of QoL extracted from patient interviews and included items holding the greatest clinical relevance. Therefore subsequent analyses were conducted on this version of AF‐QoL.
PCA identified six factors representing psychological impact, impact on everyday activities, unpredictability/disease control, seton, impact on going out and discomfort and discharge (Table S5). No items were excluded based on factor loading. There were no floor or ceiling effects, as confirmed in the distribution of respondents’ overall AF‐QoL scores, seen in Figure 1.
FIGURE 1.
Histogram of anal fistula quality of life scale (AF‐QoL) score distribution.
Reliability
In retest analysis, 35 of 78 patients (45%) reported stable symptoms. The Pearson correlation coefficient for scores at both time‐points was 0.929 and the ICC was 0.865, demonstrating excellent stability and reliability.
Responsiveness
Some 43 of 78 patients (55%) reported a change in symptoms. The ES, RR and SRM were all at acceptable levels (0.548, 0.512 and 1.124, respectively). Comparisons of mean test and retest scores were performed and are shown in Table 3. Statistically significant changes were seen in those reporting improved or worsening symptoms.
TABLE 3.
Comparison of mean test and retest scores in patients reporting symptom stability and a change in symptoms.
| n | Test mean (SD) | Retest mean (SD) | Change mean (95% CI) | p‐value | |
|---|---|---|---|---|---|
| Stable | 35 | 45.525 (19.289) | 42.919 (17.902) | 2.605 (−0.317 to 5.528) | 0.8 |
| Improved | 17 | 33.717 (10.548) | 39.037 (13.366) | 5.32 (0.142–10.4988) | 0.04 |
| Worse | 26 | 39.138 (20.021) | 32.503 (14.571) | −6.635 (−11.3121 to −0.958) | 0.007 |
Note: p < 0.05 is considered statistically significant.
Interpretability
The MIC was 7.107, meaning that a change of seven points in the AF‐QoL score would be deemed to be an improvement or deterioration. The final version of AF‐QoL can be seen in Data S1.
DISCUSSION
We present the development and validation of a patient‐centred, disease‐specific cryptoglandular Anal Fistula Quality of Life (AF‐QoL) measurement instrument, aimed at quantifying the impact of having a fistula on patient QoL. The instrument contains 22 items, the overall AF‐QoL score is an average of the applicable answers (see Data S1), resulting in a numerical value on a scale of 0–100, where a higher score indicates a better QoL. Due to the number of ‘Not applicable’ answers, we opted to use percentiles to make the scores easier to calculate, understand and interpret.
AF‐QoL has been developed for evaluative purposes. In clinical practice, the scale can identify changes in disease‐specific QoL occurring either as a result of disease progression or resolution or in response to treatment. This can guide clinical decision‐making by recognizing when changes in management are required or by highlighting those treatments that support symptom management, particularly in patients with chronic disease. In research studies and clinical trials, AF‐QoL can provide a more accurate assessment of changes in QoL in response to treatment, which can help to establish the superiority of some treatments over others, particularly in complex disease where fistula resolution is unlikely.
Existing QoL measurement tools [28, 29] have had limited patient involvement, raising concerns regarding their content validity [10]. The participation of fistula patients in shaping AF‐QoL is evident from the range of topics contained within the instrument. Prior instruments have included some of these themes; however, AF‐QoL recognizes additional factors such as the psychological impact of treatment, defined by themes of anxiety relating to after care, medication use and continence, and further identified the impact of having a seton on QoL [1].
Patients with a cryptoglandular anal fistula of any duration were eligible to participate in all stages of this study; however, a greater proportion of participants experienced a shorter duration of disease and fewer procedures in the initial item‐generation stage than in the later development and testing stages (Table 1). It could be argued that this may have influenced the nature of the themes contained within AF‐QoL, in that themes that were relevant to newly diagnosed patients would be more likely to be incorporated into earlier iterations of the instrument. However, participants undergoing cognitive interviews (Stage 2) were specifically prompted to report any themes that were not covered by the draft instrument, and statistical analysis performed in Stage 3 would have identified any redundant items, therefore the impact of variable disease duration between stages on the final instrument should be minimal.
One of the challenges of developing and utilizing an instrument measuring such a broad range of domains is maintaining a balance between comprehensiveness and length. Statistical analyses successively eliminated redundant items. The final 22‐item instrument strikes a balance between including items that are paramount to determining QoL with a fistula and being too cumbersome to complete. It also includes all the QoL aspects defined in the recent core outcome set [2].
Several themes contained within the instrument are conditional yet essential determinants of QoL. Methods for dealing with missing data where questions are not applicable include substituting answers with the overall mean score [19]. We interpreted that an unapplicable question would have no bearing on an individual's QoL, and so rather than substitute this with a value that could incorrectly skew the results, we removed these from the denominator from which the AF‐QoL score is calculated. This requires more time to compute the overall score, but is outweighed by the benefit of producing a more accurate reflection of QoL.
The strengths of the study include patient involvement in shaping the final instrument, expert clinicians assessing content validity of individual items and clinician researchers with expertise in PROM development directing methodology and ensuring a rigorous and robust approach according to COSMIN guidance. This multidisciplinary approach means that the instrument retains validity and domain specificity, reflects the extent of the clinical problem and also demonstrates excellent internal consistency and construct validity.
There are a few limitations to AF‐QoL development. The S‐CVI/Ave (0.88) was marginally lower than the accepted 0.9 threshold [15], owing to retention of items falling below the threshold for inclusion, which was done on the basis of patients' views. There may be tension between expert clinician content validity and patient perspective, in which we believe the patient must prevail. The final 22‐item version of AF‐QoL has a Cronbach‐α value of 0.927, which is above the accepted threshold for internal consistency (0.9). However, values for Cronbach‐α up to 0.95 are acceptable [30], and the study management group felt that further exclusions would risk losing clinically relevant items that would be notable by their omission.
It is recommended that five to ten participants per item on the draft scale undergoing psychometric testing are recruited to participate in test–retest analysis [31]. We were unable to reach this, reflecting the relative infrequency of anorectal fistula, therefore sensitivity and stability analysis may not have been as robust. However, test–retest validation is an ongoing process, and while our initial analysis supports the validity of the tool, further prospective studies, conducted specifically on patients undergoing treatment resulting in a change in symptoms, could provide additional data to support the sensitivity of the instrument. There is no recommended time period between repeated administrations for reliability analysis; however, a prolonged interval may reduce the number of stable participants [22]. One of the key findings in patient interviews was the unpredictability of symptoms, so it is difficult to determine the appropriate time period for analysis.
Despite this, the 22‐item AF‐QoL demonstrates reliability, construct validity and internal consistency. Future studies should focus on assessing its measurement properties alongside other QoL instruments, and in determining reference values for the scale, from which the severity of QoL impact can be deduced.
AUTHOR CONTRIBUTIONS
Nusrat Iqbal: Investigation; writing – original draft; methodology; writing – review and editing; project administration; validation. Laith Alrubaiy: Conceptualization; investigation; writing – review and editing; methodology; formal analysis. Ailsa Hart: Conceptualization; writing – review and editing; supervision. Rachel Siviter: Methodology; validation; writing – review and editing; investigation. Lynsay Wilson: Investigation; methodology; validation; writing – review and editing. Phil Tozer: Conceptualization; investigation; methodology; writing – review and editing; supervision.
FUNDING INFORMATION
None.
CONFLICT OF INTEREST STATEMENT
None declared.
ETHICS STATEMENT
Ethical approval was granted by West Midlands–Black Country Research Ethics Committee (REC reference 19/WM/0296) and Wales Research Ethics Committee 6 (REC reference 20/WA/0055) and was also approved by the Health Research Authority (HRA) and Health and Care Research Wales (HCRW).
Supporting information
TablesS1‐S5:
Data S1:
ACKNOWLEDGEMENTS
The authors would like to thank Jordan Fletcher for graphic design of the final AF‐QoL instrument.
Iqbal N, Alrubaiy L, Hart A, Siviter R, Wilson L, Tozer P. The development of a cryptoglandular anal fistula quality of life scale (AF‐QoL). Colorectal Dis. 2024;26:716–725. 10.1111/codi.16917
The data within this manuscript were presented as an oral presentation at the Association of Coloproctology of Great Britain and Ireland (ACPGBI) annual meeting, Manchester UK, July 2023, and as a poster at the European Society of Coloproctology (ESCP) annual meeting, Vilnius, September 2023.
DATA AVAILABILITY STATEMENT
The data supporting the findings of this study are available within the article and its supplementary materials.
REFERENCES
- 1. Iqbal N, Machielsen AJHM, Breukink SO, Woodcock R, Kane G, Alrubaiy L, et al. Living with cryptoglandular anal fistula: a qualitative investigation of the patient's experience through semi‐structured patient interviews. Qual Life Res. 2022;31:2505–2518. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Iqbal N, Machielsen AJHM, Kimman ML, Kane G, Woodcock R, Grossi U, et al. AFCOS: the development of a cryptoglandular anal fistula core outcome set. Ann Surg. 2022;277:e1045–e1050. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Machielsen AJHM, Iqbal N, Kimman ML, Sahnan K, Adegbola SO, Kane G, et al. Heterogeneity in outcome selection, definition and measurement in studies assessing the treatment of cryptoglandular anal fistula: findings from a systematic review. Tech Coloproctol. 2021;25:761–830. doi: 10.1007/s10151-021-02452-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Owen HA, Buchanan GN, Schizas A, Cohen R, Williams AB. Quality of life with anal fistula. Ann R Coll Surg. 2016;98(5):334–338. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Visscher AP, Schuur D, Roos R, Van der Mijnsbrugge GJH, Meijerink WJHJ, Felt‐Bersma RJF. Long‐term follow‐up after surgery for simple and complex cryptoglandular fistulas: fecal incontinence and impact on quality of life. Dis Colon Rectum. 2015;58(5):533–539. [DOI] [PubMed] [Google Scholar]
- 6. Jayarajah U, Wickramasinghe DP, Samarasekera DN. Anal incontinence and quality of life following operative treatment of simple cryptoglandular fistula‐in‐ano: a prospective study. BMC Res Notes. 2017;10(1):1–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Owen HA, Buchanan GN, Schizas A, Emmanuel A, Cohen R, Williams AB. Quality of life following fistulotomy—short term follow‐up. Color Dis. 2017;19(6):563–569. [DOI] [PubMed] [Google Scholar]
- 8. Patrick DL, Deyo RA. Generic and disease‐specific measures in assessing health status and quality of life. Med Care. 1989;27(3 Suppl):S217–S232. [DOI] [PubMed] [Google Scholar]
- 9. Iqbal N, Shah R, Alrubaiy L, Tozer P. Do patient‐reported quality‐of‐life (QoL) scales provide an adequate assessment of patients with Cryptoglandular anal fistulae? A systematic review of measurement instruments and their content validity. Clin Pract. 2022;12(4):628–639. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Terwee CB, Prinsen CAC, Chiarotto A, Westerman MJ, Patrick DL, Alonso J, et al. COSMIN methodology for evaluating the content validity of patient‐reported outcome measures: a Delphi study. Qual Life Res. 2018;27(5):1159–1170. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Ebrahim S. Clinical and public health perspectives and applications of health‐related quality of life measurement. Soc Sci Med. 1995;41(10):1383–1394. [DOI] [PubMed] [Google Scholar]
- 12. Wilson IB, Cleary PD. Linking clinical variables with health‐related quality of life: a conceptual model of patient outcomes. JAMA. 1995;273(1):59–65. [PubMed] [Google Scholar]
- 13. Ojelabi AO, Graham Y, Haighton C, Ling J. A systematic review of the application of Wilson and Cleary health‐related quality of life model in chronic diseases. Health Qual Life Outcomes. 2017;15(1):241. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Post MWM. Definitions of quality of life: what has happened and how to move on. Top Spinal Cord Inj Rehabil. 2014;20(3):167–180. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Polit D, Beck CT. The content validity index: are you sure you know what's being reported? Critique and recommendations. Res Nurs Health. 2006;29:489–497. [DOI] [PubMed] [Google Scholar]
- 16. Ware J, Kosinski M, Keller S. A 12‐item short‐form health survey: construction of scales and preliminary tests of reliability and validity. Med Care. 1996;34(3):220–233. [DOI] [PubMed] [Google Scholar]
- 17. Bjelland I, Dahl AA, Haug TT, Neckelmann D. The validity of the hospital anxiety and depression scale. An updated literature review. J Psychosom Res. 2002;52(2):69–77. [DOI] [PubMed] [Google Scholar]
- 18. Irvine E. Usual therapy improves perianal Crohn's disease as measured by a new disease activity index. J Clin Gastroenterol. 1995;20(1):27–32. [PubMed] [Google Scholar]
- 19. Streiner DL, Norman GR, Cairney J. Health measurement scales: a practical guide to their development and use. 5th ed. Oxford University Press; 2015. [Google Scholar]
- 20. Murphy KR, Davidshofer CO. Psychological testing: principles and applications. 2nd ed. Upper Saddle River, N.J., London: Prentice‐Hall, Inc; 1991. p. 514. [Google Scholar]
- 21. Marx RG, Bombardier C, Hogg‐Johnson S, Wright JG. Clinimetric and psychometric strategies for development of a health measurement scale. J Clin Epidemiol. 1999;52(2):105–111. [DOI] [PubMed] [Google Scholar]
- 22. Terwee CB, Bot SDM, de Boer MR, van der Windt DAWM, Knol DL, Dekker J, et al. Quality criteria were proposed for measurement properties of health status questionnaires. J Clin Epidemiol. 2007;60(1):34–42. [DOI] [PubMed] [Google Scholar]
- 23. Deyo RA, Diehr P, Patrick DL. Reproducibility and responsiveness of health status measures statistics and strategies for evaluation. Control Clin Trials. 1991;12(4 Suppl.):142–158. [DOI] [PubMed] [Google Scholar]
- 24. Cohen J. A power primer. Psychol Bull. 1992;112(1):155–159. [DOI] [PubMed] [Google Scholar]
- 25. Liang MH, Lew RA, Stucki G, Fortin PR, Daltroy L. Measuring clinically important changes with patient‐oriented questionnaires. Med Care. 2002;40(4):45–51. [DOI] [PubMed] [Google Scholar]
- 26. Cohen J. Statistical power analysis for the behavioural sciences. 2nd ed. New Jersey: Lawrence Erlbaum Associates; 1997. [Google Scholar]
- 27. Jaeschke R, Singer J, Guyatt GH. Measurement of health status. Ascertaining the minimal clinically important difference. Control Clin Trials. 1989;10(4):407–415. [DOI] [PubMed] [Google Scholar]
- 28. Tabassum A, Zahid A, Byrne CM, Young CJ. Development of perianal sepsis index (PASI) for assessing quality of life in chronic perianal sepsis. Eur Surg. 2018;50:270–277. [Google Scholar]
- 29. Ferrer‐Márquez M, Espínola‐Cortés N, Reina‐Duarte A, Granero‐Molina J, Fernández‐Sola C, Hernández‐Padilla JM. Design and psychometric evaluation of the quality of life in patients with anal fistula questionnaire. Dis Colon Rectum. 2017;60(10):1083–1091. [DOI] [PubMed] [Google Scholar]
- 30. Tavakol M, Dennick R. Making sense of Cronbach's alpha. Int J Med Educ. 2011;2:53–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Boateng GO, Neilands TB, Frongillo EA, Melgar‐Quiñonez HR, Young SL. Best practices for developing and validating scales for health, social, and behavioral research: a primer. Front Public Health. 2018;6(June):1–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
TablesS1‐S5:
Data S1:
Data Availability Statement
The data supporting the findings of this study are available within the article and its supplementary materials.