Keywords
Quality and logistical aspects - Preparation - Quality management
10.1055/a-2125-0025
Real-world studies in many fields of medicine are becoming quite popular. They usually
represent manuscripts with obvious shortcomings, including retrospective character,
historical/not randomized comparison groups or lack of them, large amounts of missing
data or various definitions of important aspects of the study. But they also have
multiple advantages, including less “sterile” populations studied as compared to those
in well-defined typical randomized controlled trials (RCTs), which is seen in inclusion/exclusion
criteria that are much less strict and exclusive in real-world studies. Altogether,
clinical scenarios in real-world studies much more closely resemble what is going
on in existing health care systems. Therefore, some experts claim that real-world
studies may be regarded as the ones verifying recommendations and also presenting
data that can become hypothesis-generating.
In this recent issue of EIO, the paper by Jose Esteban and colleagues [1] on bowel preparation for colonoscopy represents a real-world data study with typical
shortcomings and advantages. The study was based on medical records from 10 centers
(8 in Spain and 2 in Portugal). Patients were adults and used 1L PEG plus ascorbic
acid. Adequate bowel prep was assessed using the standard Boston Bowel Preparation
Scale (BPPS) system. The main outcome measures were: adequate cleansing in the whole
large bowel, high-quality cleansing in the right colon, and frequency of adverse events
(AEs). Fortunately, all critical data defined as mandatory were available (sex, age,
indication, dosing regimen, BPPS and number of polyps in each segment) and those not-mandatory,
not surprisingly, were missing (e.g. body mass index) in roughly 50% to 90% of cases.
The interesting aspects of this study are multiple but I want to focus on two of them.
First, two main regimens were used in participating centers: overnight split-dose
regimen and same-day regimen (the morning of the colonoscopy). Interestingly, despite
current recommendations, split dosing was used infrequently (in only 32.8% of participants).
It appears that there is a notion that split dosing in participating countries may
lead to low adherence due for cultural and social reasons [2]. Specifically in Portugal, centers adopted the same-day regimen predominately due
to the fact that colonoscopies are frequently performed in the afternoon. In this
context, it has to be stressed that the overall adequate bowel preparation as well
as high quality in the right colon were statistically significantly higher for split
dosing than for same-day. And it confirms what we have known for a long time. Therefore,
it needs to be considered in real practice in those two countries. The authors claim
that the worse bowel prep with the same-day regimen was due to the interval between
dosing and the start of colonoscopy, which was longer than 5 hours. Whatever the reason,
it should corrected for the sake of Spanish/Portuguese patients.
Another aspect of this real-world study was a bit shocking. This was the lack of information
in the medical documentation about whether complete cecal intubation was achieved
– in as high a proportion of colonoscopies as 50%. Lack of this information may happen
occasionally, but not in such a high percentage. This is alarming, and in my view,
publication of this real-study and appropriate action by Spanish/Portuguese endoscopic
bodies to correct this finding may provide great benefits to society and patients.
It is also worth drawing attention to confounding aspects of the study, which are
unfortunately present in most of papers dealing with bowel preparation for colonoscopy.
First, scoring of the quality of bowel preparation is very subjective. This is usually
expressed by high interobserver and intraobserver variability observed in other studies.
A very interesting finding is that high-detector endoscopists (achieving high adenoma
detection rates [ADRs]) usually report lower scores for bowel preparation quality
[3]. This goes with the fact that perfectionist endoscopists (high detectors) are not
so easily satisfied with imperfect bowel prep. They want to have a perfect view because
they usually have higher expectations than the average endoscopist. That is why it
is quite difficult to prove that the better bowel prep, the higher ADR and polyp detection
rate. Such a finding, which is logical, was confirmed in only a small fraction of
studies. Second, authors of this study and frequently other authors of papers dealing
with the so-called “low-volume preparation” do not report the volumes of plain water
that was drunk by participants as supplementary hydration. That supplementary hydration
is recommended in Product Characteristics and in Instructions for Patients but then
disregarded in analyses. It is understandable that the expression “low-volume” applies
to volume of “unpleasant or special taste” liquid, but for the sake of critical comparisons,
such data should be available. But this was a real-world study, so lack of that data
is understandable.
In summary, I would like to thank the authors for their effort in performing this
study and providing data thanks to which the real-world situation in endoscopy may
hopefully improve. Let us call for more high-quality, real-world studies honestly
performed like this one.
