To what extent are anorectal function tests comparable? A study comparing digital rectal examination, anal electromyography, 3-dimensional high-resolution anal manometry, and transperineal ultrasound

Anorectal function disorders like fecal incontinence and chronic constipation are very common. Generally, a conservative approach with lifestyle advices, fibers, laxative, and pelvic floor physical therapy will improve complaints in many patients. When unsuccessful, or the underlying cause seems unclear, these patients are referred to a specialist for further evaluation of anorectal function and possible therapy [1]. Besides digital rectal examination (DRE), a variety of tests are available to evaluate anorectal function. One may then objectively assess, e.g., low or high tone of the anal sphincter, paradoxical contraction, or inadequate relaxation of the pelvic floor.

Available tests are, for example, anorectal manometry (ARM), 3-dimensional high-resolution anorectal manometry (3D-HRAM), balloon expulsion test (BET), surface electromyography with or without an intra-anal probe (s-EMG), transperineal ultrasound defecography, and the classical defecography. Although some studies suggest that DRE alone is a useful tool to identify anorectal disorders [2, 3], others propose that anorectal function tests objectively evaluate anorectal function and might provide a predictive value for treatment results and influence management [4‐9]. Which anorectal function test is the most accurate is under debate.

The s-EMG with intra-vaginal or intra-anal electrode probes is commonly utilized by the pelvic floor physical therapist to confirm DRE and evaluate therapy [5, 10]. ARM is often considered the gold standard to measure anal pressures; however, lack of reproducibility mentioned in several studies makes the test questionable [11‐16]. Few studies compared ARM with anal s-EMG and showed limited concordance [17‐19]. A more recent study compared ARM with DRE to determine dyssynergia and concluded that there was a moderate agreement [20].

According to the ROME IV criteria, dyssynergia is established by two out of three anorectal function tests: first, abnormal anorectal evacuation pattern measured with ARM or s-EMG; second, abnormal BET; and third, impaired rectal evacuation diagnosed on imaging studies (e.g., defecography) [7]. Furthermore, examinations as DRE and transperineal ultrasound are not mentioned in this context and a clear gold standard for one of these tests is not suggested. One could wonder whether a restricted use of these additional tests is justified. Could we rely on DRE and use additional tests only in complex patients?

Another reason to perform anorectal function tests is an attempt to objectively measure the anal pressures. Since there is no gold standard, a reappraisal for DRE by experienced investigators seems worthwhile investigating.

The Proctos Clinic is a tertiary referral center for specialized proctological care with experienced surgeons, a pelvic floor physical therapist, and a fully equipped anorectal function laboratory. The aim of our study was to examine the correlation of the anal pressures between DRE, 3D-HRAM, and the s-EMG. DRE, 3D-HRAM, s-EMG, BET, and the transperineal ultrasound were compared to diagnose dyssynergia. Furthermore, we sought to assess the level of agreement between DRE performed by the surgeon and the pelvic floor physical therapist.

The present study provides an overview of the correlation between outcomes of frequently performed anorectal function tests and compares their ability to measure dyssynergia. Furthermore, this study measured the level of agreement between DRE performed by the surgeon and the pelvic floor physical therapist in a tertiary referral center.

Despite the surgeons and the pelvic floor physical therapist being experienced, performing several digital rectal examinations per day, the agreement of the anal tone between their DRE was not perfect. The assessed tone during rest, squeeze, and straining did not correlate in 22%, 22%, and 16%, respectively. To the best of our knowledge, no literature concerning the interrater agreement of DRE has been published. Interrater agreement has only been studied in vaginal digital assessment concerning the pelvic floor function and digital rectal examination in the context of prostate cancer [31‐33]. Overall, the agreement was substantial to almost perfect. The small differences in classification of DRE between the surgeon and pelvic floor physical therapist may be explained by differences in interpretation of the indifferent movement of the pelvic floor. Not a single examination was classified both as relaxation and paradoxical movement.

The correlation between the surgeons’ DRE, pelvic floor physical therapists’ DRE, and the 3D-HRAM in our study was moderate and somewhat better for squeeze tone/pressures than resting tone/pressures. Several studies compared DRE with ARM and showed an overall good agreement of pressures, however similar to our study, slightly better for squeeze pressures, but results are not consistent [9, 15, 34‐39]. For example, the study by Beatrice et al. showed that DRE correlates well, but not perfectly, with the ARM for resting pressures, (r = 0.71 (p < 0.001) [9]. However, Orkin et al. observed an excellent agreement between DRE and the ARM for resting pressures (r = 0.82) and for squeeze pressures (r = 0.81) [34]. In contrast, Soh et al. described a poor agreement between DRE and ARM for resting pressures with a k coefficient of 0.01 and a moderate agreement for squeeze pressure with a k coefficient of 0.42 [35]. Pinto et al. showed a moderate to strong agreement for resting pressure with a Gamma index of 0.7 and a strong correlation of the squeeze pressures with a Gamma of 0.96 [37]. All studies—including ours—report that the examinations were performed by experienced examiners but the results vary considerably. Nevertheless, ARM can be performed with a variety of types of equipment, techniques, and study protocols, making results less reproducible and thus difficult to compare [40, 41]. A recent study by Prichard et al. described even significantly different results during ARM between operators despite using similar instructions to patients [16]. Even a small difference in outcome could lead to a different interpretation. It must be noted that in contrast to most ARM studies we used the 3D probe.

DRE correlated better with 3D-HRAM in patients referred for fecal incontinence. With 54%, this was the largest group in this study. However, defining “normal” resting and squeeze pressures for ARM values is quite difficult. There is obviously an overlap since several studies showed different values for normal and abnormal resting and squeeze pressures for ARM [14, 23‐28]. To be accurate in comparing between groups, the pressures should be adjusted according to age, gender, and parous and nulliparous females. But these differences were small, and to make comparisons between tests manageable in this study, we did not differentiate.

The surgeons’ DRE and the pelvic floor physical therapist’s DRE were compared to the s-EMG and showed some discrepancies. The surgeon’s DRE and the pelvic floor physical therapist’s DRE were categorized as “low” whereas the s-EMG categorized “high” in three and four patients, respectively. However, one patient was categorized “high” with DRE and “low” with s-EMG. This can probably be explained by the fact that patients who can hardly control their external anal sphincter might overcompensate with their levator muscle. As we measured with s-EMG, the mean of the total electrical activity of the external anal sphincter including the levator muscle, the EMG activity might be higher than expected. When retrospectively assessing the 3D-HRAM, these patients showed indeed higher pressures of the posterior levator muscle on the 3D image in contrast to the sphincter and vice versa for the patient with a chronic anal fissure. Furthermore, high tone on the levator muscle with DRE might be turgor which is not measured with s-EMG. For this reason, comparing s-EMG with other tests might not be appropriate and should probably be used only to confirm physical examination and biofeedback registration.

The correlation between s-EMG and the 3D-HRAM was better for squeeze pressures and electric activity than resting pressures and electric activity with an agreement of 59% and 37%, respectively. A study from 1989 also showed limited concordance with a correlation coefficient of 0.55 (p < 0.001) between the maximum squeeze pressure with ARM and maximum contraction pattern with the EMG [17]. Regarding diagnosing dyssynergia while straining with s-EMG and 3D-HRAM, our results were not in line with the results by Chiarioni et al. [30]. In our study, s-EMG and ARM were concordant in 52% while Chiarioni et al. described an agreement of 88% for classifying patients’ dyssynergic or not dyssynergic. Both tests are used to test the anorectal function but are used for different purposes in clinical practice. The question that remains is how relevant small differences are in clinical practice.

The results of the six different function tests used to diagnose pelvic floor dyssynergia, namely, DRE by both the surgeon and the pelvic floor physical therapist, 3D-HRAM, s-EMG, BET, and transperineal ultrasound (with echo lucent gel), were to some extent comparable. Although most comparisons were statistically significant, the correlation remained low. Discrepancies with TPUS could be explained by the nonanatomical supine position of the test and the fact that the patient is not in private environment. Three patients who evacuated the gel—although not completely—but were not able to expel the balloon within 1 min were referred for PFPT because of fecal incontinence. It is very likely that these patients lost the gel by leaking, not because of the push effort. This makes these tests not suitable to compare.

Furthermore, the tests are performed in different postures; the balloon expulsion is performed in a private setting, in sitting position, whereas the other tests are performed by an examiner with the patient lying in the left lateral position. 3D-HRAM measures the anorectal pressures, s-EMG measures electrical activity, and TPUS is visually assessed by the doctor where evacuating echo lucent gel might support their findings. Some discrepancies cannot be explained except the snapshot nature of the tests. It is known that the diagnostic accuracy of ARM is limited for discriminating between healthy people and patients with functional constipation [42]. Unfortunately, previous studies with TPUS assessed its accuracy for detecting rectocele, intussusception, or enterocele or used a total pelvic floor ultrasound without echo lucent gel. No previous studies reported its accuracy to diagnose dyssynergia. However, based on our experience, the TPUS is a low-cost and easy tool for surgeons to perform. Surgeons are able to perform their own test in the outpatient clinic, and moreover, it has comparable results with the classical defecography [43] which makes it worth considering a relevant anorectal function test. The BET is a frequently used test for assessing defecatory dysfunction since it is a simple and low-cost procedure. Different protocols are used to perform the procedure, air filled or water-filled balloon, lying or seated position. Time values that are considered abnormal range from 1 till 5 min [28‐30, 44, 45]. In our study, a balloon expulsion time of more than 1 min was considered prolonged. This was categorized as dyssynergia by the 3D-HRAM in 32% of the cases. In contrast to older studies, more recent studies demonstrated poor agreement between BET and ARM [46, 47].

According to the ROME IV criteria, dyssynergic defecation is established by two out of three tests: (1) ARM or s-EMG, (2) balloon expulsion test, or (3) defecography. Remarkable is that the ARM or the s-EMG should be abnormal and that DRE and the transperineal ultrasound are not mentioned in this workup [7]. This might be confusing and suggests that none of the tests can be considered as golden standard. Furthermore, anorectal function tests provide additional workload and costs whereas DRE is widely available and dyssynergia is a widespread phenomenon. The ROME IV criteria are merely used to standardize patients in an attempt to objectivize dyssynergia. Also, Bordeianou et al. had their doubts about which test to assign the highest value, the s-EMG, BET, or ARM, prior to referral to the pelvic floor physical therapist with dyssynergia [48].

Undoubtedly, this study has a number of limitations which should be acknowledged. First, the surgeons and the pelvic floor physical therapist were unblinded to the patients’ medical history when performing the DRE which likely has influenced the results by information bias. Secondly, although all surgeons and the pelvic floor physical therapist were given instructions before the study started on how to perform a complete structured DRE and systematically describe the physical examination in the electronic health record, variety in performing and assessing DRE is insurmountable. The single observer for all 3D-HRAM results might be a lowness or a strength in this study. A considerable limitation of this study is that we were not able to use controlled normal s-EMG values since they have not yet been published. Furthermore, the results of the study would have had more relevance if there was a gold standard or known sensitivity of the tests. This issue is also reflected in the ROME IV criteria for dyssynergic defecation as mentioned above. Unfortunately, not all patients underwent all tests due to logistic problems in the outpatient clinic concerning the tests in the context of the study. Consequently, some patients did not undergo the BET or the TPUS. Lastly, there might have been interpretation bias by assessing straining movement of the pelvic floor. It is not known how “indifferent” movement of the pelvic floor is defined among the examiners; does this mean “no movement” or also “relaxation but not enough”? This probably resulted in different outcomes.

This study showed that squeeze pressures were more often similarly categorized than resting pressures in anorectal function tests. It further shows that the surgeons’ DRE and the pelvic floor physical therapist’s DRE more often similar assessed in comparison to anorectal functions tests as 3D-HRAM, s-EMG, or TPUS. Still, the correlation between all tests is quite disappointing and this raises questions regarding when to perform these tests in addition to DRE. Or does this mean that we can suffice with an expert’s DRE when referring to the pelvic floor physical therapist for dyssynergia? The pelvic floor physical therapist will evaluate therapy with his/her own DRE with or without s-EMG, not with ARM or transperineal ultrasound. Perhaps we should only perform anorectal function tests in patients who are refractory to conservative treatment like lifestyle and pelvic floor physical therapy or when more invasive procedures like surgery or Botox, e.g., are considered. Furthermore, these tests are valuable when evaluating new (surgical) therapies.

This study shows that DRE has a good correlation among experienced investigators. Since commonly performed anorectal function tests correlate poorly with DRE and with other anorectal function tests, DRE by an experienced investigator suffices in daily clinical practice. When conservative treatment fails, further investigation is warranted; however, these results should be interpreted with caution.