The influence of defaecating proctograms on clinical decision‐making in pelvic floor disorders

Defaecating proctogram (DP) studies have become an integral part of the evaluation of patients with pelvic floor disorders. However, their impact on treatment decision‐making remains unclear. The aim of this study was to assess the concordance of decision‐making by colorectal surgeons and the role of the DP in this process.

comprehensive symptom assessment with a detailed history and focused examination, and then often endoanal ultrasound, manometry, pudendal nerve latency testing and a defaecating proctogram (DP) [3].Treatments include dietary manipulation, medication, physiotherapy and surgery, often facilitated through multidisciplinary teams to integrate the complex factors that contribute to clinical decision-making [4].
Fluoroscopic DP is an investigation that provides dynamic imaging of the anorectal junction, pelvic floor descent, evacuation time and completeness, rectocele trapping of contrast, intra-anal intussusception, external prolapse and enterocele [5,6].DP facilitates classification into clinical groups: pelvic floor dyssynergia; rectocele dysfunction; intussusception and prolapse; enterocele.
However, the adoption of DP is variable owing to limited correlation with symptoms and variability in the interpretation of structural findings [3,[7][8][9].
Nevertheless, DP remains widely used in pelvic floor clinics, often influencing the extent of intervention, including decisionmaking around proceeding to surgery.This study aimed to assess the impact of DP on clinical decision-making and the concordance in treatment decisions between colorectal surgeons.

Ethics statement
Ethical approval of this study was obtained by Northern Regional Ethics Committee with institutional approval from the Research Review Committee of the Auckland District Health Board.

Study design
A prospective, clinician survey based on retrospective, systematically collected data was constructed.Changes in a surgeon's decision regarding treatment for pelvic floor disorders after the addition of the DP results to patients' diagnostic work-ups were quantified.

Defaecating proctogram
A DP is a fluoroscopic method for evaluating the dynamics of rectal emptying.It provides a 'near' physiological picture of evacuation compared with other imaging methods and physical examination [10].The technique itself involves instillation of 100-120 mL of barium paste into the rectum followed by fluoroscopic imaging of the rectum, anus and pelvic floor while defaecation is performed [5].The following features are extracted: sphincter function, intussusception into the anal canal, paradoxical puborectalis contraction, nonemptying rectocele and excessive pelvic floor descent (in centimetres).

Patient population
The study population consisted of patients who were referred between 2007 to 2010 to the Pelvic Floor Clinic (Auckland Hospital, Auckland, New Zealand) for further assessment of pelvic floor disorders.The medical history of patients within this cohort was attained from the Auckland District Health Board (ADHB) database and the treatment they received was noted.Only those patients who received one of three treatment options following their Pelvic Floor Clinic consultation were considered for this study.These treatments options were: physiotherapy only, anterior Delorme's procedure and anterior mesh rectopexy.Patients who received the aforementioned management but did not undergo a DP were excluded.

Clinical information and DP results
Clinic letters for the patients who met these criteria were obtained and two surveys were created, with each report representing a deidentified patient.The first survey included all relevant clinical information obtained at the Pelvic Floor Clinic consultation, including the history, examination, manometry, transanal ultrasound and electromyography but excluding the DP result.The second survey included precisely the same clinical information as the first but also included the DP result for each patient.The second survey was provided in a different order to mitigate the impact of the first review of each report on subsequent decisions.Parameters included in the DP report included sphincter length, level of the sphincter and anorectal angle at rest and on straining, the presence of prolapse, intussusception, rectocele, paradoxical puborectalis contraction and retention of barium paste.These DP results were interpreted and reported by a single radiologist (see the example in Appendix S1).

Assessment by surgeons
Four surgeons participated in this study.All had completed their fellowship in general surgery, had undergone further colorectal training and had an interest in the management of pelvic floor disorders.Each surgeon reviewed the clinic letters for the de-identified patients that were presented to them in two subsequent online surveys.Although the surgeons were aware the patients had received one of the three treatments, they were blinded to the actual treatment of individual patients.Based on the clinical information presented to them, each surgeon was required to choose the most appropriate way to manage each patient by choosing one out of the three treatment options: physiotherapy only, anterior Delorme's procedure or anterior mesh rectopexy.The two surgical options would be in addition to physiotherapy, and this physiotherapy would be tailored in a case-by-case manner at the discretion of the physiotherapist.Each time a decision was made they were also required to rank their confidence.Only when the first survey with treatment decisions without the DP had been completed and submitted for all patients was the second survey including the DP results sent to the surgeons.The same process with identical questions was used for both online surveys.The surgeons could not access their responses to the first survey when answering the second survey.

Statistical analysis
The frequency of change in treatment decision after viewing the DP was calculated for each surgeon.Interobserver agreement on treatment choice between the four surgeons for the entire cohort, before and after DP, was quantified by using the unadjusted Kappa coefficient, as follows: κ = 0-0.20,poor; κ = 0.21-0.40,fair; κ = 0.41-0.60,moderate; κ = 0.61-0.80,good; κ = 0.81-1.00,excellent.Further analysis comparing the proportion of cases pre-and post-DP for which there was the most and the least variety of treatment decisions was performed by allocating a consensus score to each case, as follows: 4 = all surgeons agreed; 3 = three out of four surgeons agreed; 2 = decisions were split between two treatments; 1 = decisions were spread between all three treatments.A confidence rating was indicated by the surgeons for each of their treatment decisions on a five-point Likert scale, as follows: 1 = not at all confident; 2 = not so confident; 3 = moderately confident; 4 = confident; 5 = very confident.The frequency of the treatment decision being in agreement with the actual treatment received by the patient was also calculated using the unadjusted Kappa coefficient.Features of the DP reports which influenced decision-making were explored through mixed effects binomial logistic regression.Multivariate data were adjusted for age, sex and DP findings with surgeon included as a random effect.Data are reported as κ or odds ratio (OR) and 95% confidence intervals (CIs).All statistical analyses were performed using IBM SPSS Statistics (v.19.0) and R version 4.2.0 (R Foundation for Statistical Computing) [11].

RE SULTS
A total of 106 patients (median age 60 years, range 27-83 years; 97 female) who met the inclusion criteria were identified, each having results for all appropriate investigations including the DP.All patients underwent physiotherapy, with 27 patients having no other treatment, 54 patients having an anterior Delorme's procedure and 25 patients having an anterior mesh rectopexy.
Treatment choice, after the addition of DP results in the second survey, changed in 219 (52%) of 424 interventional decisions (Table 1).
Agreement between the four surgeons on management choice was only fair but improved significantly after addition of the DP from κ = 0.26 (95% CI 0.19-0.30) to κ = 0.39 (95% CI 0.33-0.44).Before and after DP at least three surgeons agreed (variance score of ≥3) on the treatment in 65 (61%) and 78 (74%) (p = 0.078 chi square) cases, respectively (Table 2).Three of the four surgeons reported a statistically significant increase in confidence level after the addition of a DP (Table 3).
Comparison between the actual treatments patients received and surgeons' decisions showed an increase in agreement, from κ = 0.21 (95% CI 0.14-0.27) to κ = 0.28 (95% CI 0.21-0.35),when DP results were provided.A summation of all treatment decisions made before and after DP showed a significant decrease in cases where no operation was recommended and a significant increase in the number recommended for anterior mesh rectopexy (Figure 1).There were 16 patients whose decision changed from agreement by at least three surgeons for either physiotherapy or anterior Delorme's procedure and to agreement by at least three for rectopexy following the DP.In all of these the DP showed at least anterior prolapse into the anal canal (Oxford prolapse grade 3-5).
Other proctographic features were inconsistent in this group (seven had no posterior component and seven had a normal functioning sphincter on DP).Only one patient's treatment decision changed to physiotherapy alone from surgery, and on this there was abnormal contraction of the levator plate and no evidence of a rectocele.nence symptoms, 70% of these subjects had DP evidence of outlet obstruction (i.e.rectal intussusception or rectocele) [13].Thus, DP has the advantage of identifying abnormalities that would otherwise be missed, and this may contribute to increased numbers of patients being recommended for surgery, as was found in this study.

DP factors and their association with treatment decisions
However, the correlation between defaecography parameters and symptoms remains variable [15,16].
While others have explored the contributions of proctographic findings in predicting surgical outcomes [17], there are few data on how this diagnostic modality influences clinical decision-making.
Of DP features such as poor sphincter function, paradoxical puborectalis contraction, nonemptying rectocele and excessive pelvic floor descent, and intussusception into the anal canal, the latter was the most predictive of changes in surgical decision-making.
Rectal intussusception on DP was the variable most likely to predict a decision change to offer rectopexy surgery.While there was a trend towards performing more surgery as a result of DP findings, it is important to remember, that expert recommendations are to trial conservative management of bowel symptoms prior to surgery [18,19].Therefore, DP findings should be contextualized with modern-day best practice whereby surgery is reserved for patients whose symptoms are secondary to confirmed rectocele and refractory to conservative approaches [14].
Although DP appears to have a positive influence on the management of pelvic floor disorders it is important not to disregard the persistent variability that this study has shown.One of the explanations of this variability is the inherent subjectivity of management decisions.In a diverse group of surgeons, disagreement on the best way to manage such a complex range of disorders is inevitable.Analysis of all treatment decisions made before and after DP showed that two out of the four surgeons had a significant increase in the number of mesh rectopexy operations they recommended.
This may have been influenced by preferences to perform laparoscopic surgery.This study also indicates that DP results may lead to Example defaecating proctogram of a patient with a distal rectal reservoir where an anterior Delorme's procedure was preferred.
more patients undergoing surgery, particularly driven by the presence of rectal intussusception into the anal canal.
As symptoms are varied and there are many possible treatment options for patients with pelvic floor disorders, choosing the appropriate management is difficult.In this study, if the surgeons believed a surgical procedure was necessary (i.e.physiotherapy alone was not enough) they had a choice of either a perineal (anterior Delorme's procedure) or transabdominal (anterior mesh rectopexy) operation.The addition of a DP provided more information to the surgeon, tended to harmonize decision-making and offered a more objective metric to standardize clinical decision-making during pelvic floor assessment.
Outcomes of patients assessed at the Pelvic Floor Clinic were outside the scope of this study, which focuses on factors influencing clinical decision-making.However, it is worthwhile noting the ongoing debate and contention in the management of rectal prolapse, particularly in the absence of definitive, high-quality evidence [20][21][22].
PROSPER, the most influential randomized trial in the setting of rectal prolapse, did not demonstrate any differences in symptoms scores, quality of life or recurrences within 3 years between abdominal or perineal approaches, irrespective of whether a suture or resection approach, or Altemeier's or Delorme's procedure was used, respectively [23].A recent Cochrane systematic review found that there was no clear evidence to prefer Delorme's procedure over anterior rectopexy in the setting of rectal prolapse [21].Generally, external rectal prolapse management has trended towards increase uptake of laparoscopic abdominal approaches and a decline in perineal approaches [24,25], but Delorme's procedure remains commonly performed [25].
There are several shortcomings to this study.The context in which the surgeons were required to make a treatment decision did not fully represent the true clinical setting.However, these compromises were made in an attempt to standardize the clinical information available to each surgeon, thus addressing their decisions in relation to the DP rather than other clinical variables.
Surgeons were presented with clinical information via an online survey and were required to make an individual decision about the best management plan.In a normal situation, the clinician has the opportunity to explore the symptoms in more detail with the patient and gain a great deal of information from the clinical interaction.It was assumed that, given the same information (with and without DP), surgeons would arrive at the same treatment decision; to mitigate against variations in treatment decisions on the basis of a second review of the report four surgeons were employed, and interrater agreement was assessed.They are also able to make use of a more collaborative approach by conferring with radiologists, nurses and physiotherapists, who may all have input into a patient's treatment.Multidisciplinary meetings provide an opportunity for discussion of complex cases, of which there are many, and allow for a more holistic decision-making process.With regard to the DP report, the clinician was not able to actually view the films and relied entirely on the written report.Furthermore, although the range of treatment choices matched the actual treatment received by the patients it does not truly reflect clinical practice, where other treatment options may be available.There are also several newer therapeutic options being employed in modern pelvic floor clinics, including sutured rectopexy, colporrhaphy, rectal irrigation and sacral nerve stimulation, which were not incorporated in the current study.Sequential assessment also has an important role in decision-making, and although this was the situation for the original surgeon caring for the recruited patients it was not the case for those completing the study.In this study it was also assumed that the 'gold standard' comparison for treatment decisions was the actual treatment that patients received.This shortcoming is inevitable in the setting where the optimal treatment for these patients is still unknown [20].

CO N FLI C T O F I NTER E S T S TATEM ENT
None of the authors have conflicts of interest to declare.

7 TA B L E 2 TA B L E 3 2
No factors measured on DP were predictive of a change in decision from physiotherapy to operation.A change in decision to perform TA B L E 1 Frequency of change in treatment decision after defaecating proctogram for each surgeon.Consensus in treatment decisions between surgeons (DP, defaecating proctogram).Variance scores were allocated to each case as follows: 4 = all surgeons agreed; 3 = three out of four surgeons agreed; 2 = decisions were split between two treatments; 1 = decisions spread between all three treatments.Average confidence rating by surgeons for decisions made (DP, defaecating proctogram).Note: Five-point Likert scale used, as follows: 1 = not at all confident; 2 = not so confident; 3 = moderately confident; 4 = confident; 5 = very confident.ananterior Delorme's procedure was predicted by the absence of intussusception into the anal canal on DP (OR 0.40, 95% CI 0.18-0.91,p = 0.029).A change in decision to perform anterior mesh rectopexy was predicted by intussusception into the anal canal (OR 5.23, 95% CI 2.22-12.34,p < 0.001) and absence of nonemptying rectocele (OR 0.27, 95% CI 0.12-0.64,p = 0.003).An example of Oxford grade 5 rectal prolapse where it was decided to perform mesh rectopexy is shown in Figure2.An example DP demonstrating a distal rectal reservoir is shown in Figure3, where an anterior Delorme's procedure was preferred.When assessing clinical decisions made after seeing DP reports, intussusception into the anal canal was associated with lower odds of being considered for nonoperative management (OR 0.11, 95% CI 0.03-0.45,p = 0.002) and anterior Delorme's procedure (OR 0.15, 95% CI 0.06-0.33,p < 0.001), but significantly higher odds of offering anterior mesh rectopexy (OR 20.24, 95% CI 8.09-50.66,p < 0.001).DISCUSS IONIn this study the inclusion of the DP report with the information available to the clinician making a decision regarding the treatment of a pelvic floor disorder resulted in a decision to change treatment in over 50% of the patients.This resulted in greater interobserver agreement and an increase in the number of operations recommended, particularly anterior mesh rectopexy.F I G U R E 1 Summation of all treatment decisions showing a significant decrease in cases with no operation and an increase in the number of patients recommended for surgery after viewing of the defaecating proctogram (DP) results.Example defaecating proctogram of a patient with Oxford grade 5 rectal prolapse with evidence of prolapse into the anal canal where it was decided to perform an anterior mesh rectopexy.Of note, intussusception into the anal canal was the most influential factor on DP in driving the decision to change treatment.The addition of a DP to the diagnostic information increased the clinician's confidence in their decision and increased the concordance with the actual treatment the patient received.DP is recommended in the routine work-up of patients with pelvic floor disorders [12-14].The cause of faecal incontinence in patients may not be obvious on routine work-up, and DP can reveal occult pelvic floor disorders.For example, intermittent relaxation of the internal anal sphincter in rectal intussusception could escape detection on the day of anorectal physiology testing.In a retrospective study by Dench et al. studying patients with only faecal inconti-

Future work could explore
a wider evaluation of the interpretation of DPs in clinical practice, including surveying a larger cohort of surgeons and opinion leaders.Such efforts could aid in standardization of interpretation and assessment of pelvic floor disorders and contextualize the role that DP had in their work-up.Such future work should incorporate the impact of decisions made on the basis of DP on clinical outcomes, which could form the basis of future therapeutic guidelines.The DP appears to have an important role in the management of pelvic floor disorders.This study has shown that the investigation improves interclinician agreement and increases their confidence in treatment decisions.It also results in a greater number of surgical procedures being recommended.Despite this there remains a considerable difference of opinion about how to manage individual patients, even between surgeons operating within the same colorectal team at the same hospital.The persisting variability in treatment decisions highlights the need for well-designed comparative studies to clarify how different symptom complexes and anatomical abnormalities can be managed using individualized treatment regimens.AUTH O R CO NTR I B UTI O N S Chris Varghese: Methodology; software; formal analysis; validation; visualization; project administration; resources.Leanora Gomes: Conceptualization; methodology; data curation; investigation; formal analysis; visualization; writing -original draft.David Milne: Conceptualization; methodology; software; data curation; investigation; validation; supervision; visualization; project administration.ACK N OWLED G EM ENTS Open access publishing facilitated by The University of Auckland, as part of the Wiley -The University of Auckland agreement via the Council of Australian University Librarians.FU N D I N G I N FO R M ATI O N This work was supported by funding provided by the University of Auckland Summer Research Scholarship Fund and the Wallath Trust.