Routine versus postoperative CRP/CT guided selective diverting ileostomy after laparoscopic low anterior resection for rectal cancer: a comparative randomized controlled trial

Abstract

Background

Diverting ileostomy is routinely used to protect the anastomosis after laparoscopic low anterior resection (LAR) for rectal cancer, but it exposes most patients to stoma-related morbidity even though only a minority develop a true anastomotic leak. This trial evaluated whether a postoperative C-reactive protein (CRP) and computed tomography (CT) guided selective ileostomy strategy could reduce the rate of ileostomy formation compared with routine diversion.

Methods

In this RCT, 200 patients with stage II or III middle or lower third rectal cancer undergoing laparoscopic LAR were randomized 1:1 to a CRP/CT-guided selective ileostomy group (n = 100) or a routine diverting ileostomy group (n = 100). In the CRP/CT group, serum CRP was measured daily from postoperative day 1; a threshold of 114 mg/L on day 3 or a rise greater than 15 mg/L per day triggered confirmatory contrast-enhanced CT, with ileostomy formed only if findings were suspicious for leak. The primary outcome was the stoma-free status throughout the 12-month follow-up.

Results

Baseline demographic and tumor characteristics were similar between groups (all p > 0.05). Ileostomy formation was markedly lower in the CRP/CT group than the routine group (22.0% vs. 100.0%, p < 0.001). Anastomotic leak rates did not differ (0% vs. 1.0%, p = 1.000), nor did overall complication rates (31.0% vs. 32.0%, p = 0.879). Operative time (148.4 ± 14.0 vs. 166.6 ± 13.1 min, p < 0.001) and intraoperative blood loss (168.9 ± 28.2 vs. 197.3 ± 39.7 mL, p < 0.001) were significantly lower in the CRP/CT group. Stoma-related complications were far less frequent with the selective strategy (3.0% vs. 30.0%, p < 0.001). The CRP/CT group reported significantly better global health status, physical functioning, and emotional functioning scores (all p < 0.001).

Conclusion

A postoperative CRP/CT-guided selective ileostomy strategy reduces the need for diverting ileostomy after laparoscopic low anterior resection for rectal cancer while improving operative outcomes and quality of life; however, confirmed leakage was infrequent in both groups. This approach may serve as an alternative to routine diversion in appropriately monitored surgical units.

Trial registration

This trial was retrospectively registered at ClinicalTrials.gov (Identifier: NCT07722533) on July 17, 2026.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12893-026-04248-6.

Introduction

Total mesorectal excision (TME) with the double-stapling technique is the standard surgery for cancer of the middle and lower rectum. This technique, first described using the end-to-end anastomosis (EEA) stapler, allows a low anastomosis while preserving anal continence and avoiding permanent colostomy [1]. Modifications to double-stapled low pelvic anastomosis have made it the most widely used method for restoring bowel continuity after low anterior resection (LAR) [2]. TME, introduced by Heald, reduced local recurrence and supported sphincter-preserving surgery over abdominoperineal resection [3].

Anastomotic leakage (AL) remains the most feared complication of LAR. Reported incidence varies widely, influenced by tumor level, surgical technique, and patient risk factors [4]. AL after laparoscopic colorectal surgery is linked to poor blood supply, technical difficulty, and problems creating a tension-free join deep in the narrow pelvis [4, 5].

To reduce leak-related harm, creating a temporary diverting ileostomy remains the most common preventive method worldwide. Meta-analyses confirm that a defunctioning stoma lowers the rate of clinical leak and reoperation after LAR [6, 7]. However, this protection carries a real cost. Living with a stoma is linked to lower quality of life, more sexual and body-image problems, and higher anxiety and depression than in patients without a stoma [8, 9].

Because routine ileostomy is applied to nearly all patients while only a minority develop a true leak, researchers have sought ways to predict AL early, so a stoma is made only when needed. C-reactive protein (CRP) is the biomarker supported by the strongest evidence [10]. CRP rises quickly after surgery and falls once its trigger resolves; an elevated or rising CRP on postoperative day 3 to 5 predicts AL with a high negative predictive value, useful mainly for ruling out a leak [11–13]. Computed tomography (CT) has been added to strengthen CRP-based prediction, although a negative CT does not fully exclude a leak, and delayed reintervention after a false-negative scan worsens outcomes [14].

Despite this evidence, most guidelines, including those of the American Society of Colon and Rectal Surgeons, still recommend routine diverting ileostomy for anastomoses below the peritoneal reflection [15]. Diagnostic delay from imperfect biomarkers and imaging remains a recognized cause of delayed reintervention and worse outcomes after a missed leak [16]. Whether a selective strategy, using postoperative CRP trends confirmed by CT, can cut needless stoma formation without raising true leak-related harm after laparoscopic LAR remains unclear.

This randomized controlled trial compared an innovative CRP/CT-guided selective ileostomy strategy with routine diverting ileostomy after laparoscopic LAR for rectal cancer. The primary objective was to assess the reduction in ileostomy rate achieved through early detection.

Methods

Study design

This was a prospective, randomized, controlled, single-center trial reported following the CONSORT 2010 guideline for parallel-group randomized trials. Patients were assigned in a 1:1 ratio to either the CRP/CT-guided selective ileostomy group or the routine diverting ileostomy group.

Setting and participants

The trial was conducted in the colorectal unite, surgery department of Alexandria Main University Hospital between January 2023 and June 2025. Follow-up for the prespecified 12-month primary endpoint was completed in June 2026. The database was locked on June 18, 2026, and statistical analyses were undertaken between June 30, 2026, and July 22, 2026. Eligible patients had stage II or III, primary, resectable, operable middle or lower third rectal cancer, staged using the AJCC/TNM eighth-edition classification [17], completed neoadjuvant long-course chemoradiotherapy, and followed for at least 12 months. Patients with stage I disease, perforated or obstructed tumors, metastatic or recurrent disease, synchronous colonic resection, total neoadjuvant therapy, stapling problem, or a positive intraoperative air-leak test were excluded.

Sample size

The prespecified primary endpoint was stoma-free status throughout the 12-month follow-up. A previous study of highly selective diversion after low anterior resection reported that 76% of patients remained free from any stoma throughout one year [18]. Because all patients in the routine-diversion group were expected to undergo ileostomy according to protocol, the anticipated primary-endpoint proportions were 76% in the CRP/CT-guided group and 0% in the routine-diversion group. Using Fisher’s exact test with a two-sided significance level of 0.05, 80% power, and 1:1 allocation, the minimum required sample size was eight patients per group. However, because this design-dependent endpoint required only a small sample, enrollment was prospectively increased to 100 patients per group to improve the precision of treatment-effect estimates and permit a broader exploratory assessment of clinically important secondary outcomes, including anastomotic leakage, reoperation, stoma-related complications, and quality of life. This expanded sample was not calculated using a non-inferiority margin and was not considered sufficient to establish non-inferiority or equivalence for uncommon safety outcomes.

Randomization and allocation concealment

Eligible and consenting patients were randomly assigned in a 1:1 ratio to the CRP/CT-guided selective ileostomy group or the routine diverting ileostomy group. Randomization was computer-generated, and allocation was concealed using sequentially numbered, sealed, opaque envelopes opened only after the intraoperative air-leak test confirmed anastomotic integrity, to prevent selection bias linked to intraoperative findings. No stratification factors were applied beyond the study eligibility criteria.

Blinding

Given the nature of the intervention, patients and the operating surgeon could not be blinded to group assignment, since the surgical decision to fashion a stoma is directly observable at the time of surgery or during postoperative monitoring. To reduce assessment bias, outcome assessors recording postoperative complications, stoma-related morbidity, and quality-of-life questionnaires were kept separate from the treating surgical team wherever feasible. Laboratory staff measuring serum CRP were blinded to group allocation, and radiologists reporting CT scans were unaware of the clinical indication driving the request.

Preoperative workup and neoadjuvant therapy

Baseline assessment included digital rectal examination, rigid proctoscopy, colonoscopy with biopsy, routine laboratory tests, tumor markers, pelvic MRI, and staging CT of the chest, abdomen, and pelvis. All patients received neoadjuvant long-course chemoradiotherapy (45–50.4 Gy in 1.8–2 Gy fractions) with concurrent oral capecitabine, followed by restaging and surgery at 7 weeks [3]. All patients received a low-residue diet for 2 days before surgery and clear fluids on the day preceding surgery. Mechanical bowel preparation was performed with polyethylene glycol electrolyte lavage solution, administered as 4 L orally between 14:00 and 18:00 on the day before surgery. Oral antibiotic bowel preparation comprised neomycin 1 g plus metronidazole 1 g orally at 14:00, and 22:00 on the day before surgery. Intravenous Ceftriaxone 2 g plus metronidazole 500 mg was administered within 60 min before skin incision. A baseline serum CRP was drawn the day before surgery.

Surgical intervention

All patients in both study arms underwent laparoscopic low anterior resection with total mesorectal excision using endo-staplers, high ligation of the inferior mesenteric vessels, and routine splenic flexure mobilization. A colo-anal anastomosis was fashioned with a circular stapler, an intraoperative air-leak test confirmed anastomotic integrity in every patient; positive tests led to exclusion from the study, and followed by pelvic drain placement. At this point, randomization envelopes opened and the intraoperative and postoperative management diverged according to group allocation.

Routine diverting ileostomy group

A defunctioning diverting loop ileostomy was fashioned for every patient at the time of the primary operation, regardless of intraoperative findings or postoperative course, in line with conventional practice for anastomoses below the peritoneal reflection.

CRP/CT-guided selective ileostomy group

No stoma was created at the primary operation. Instead, a marking stitch was placed at the terminal ileum in all patients, in case a defunctioning diverting ileostomy was required. Patients entered a structured postoperative surveillance protocol. Serum CRP was measured daily from postoperative day 1. Because no single universally accepted alarm threshold exists, the lowest reported cutoff of 114 mg/L on postoperative day 3 was adopted for a conservative, safety-first approach [19]. Patients with CRP below this level and a continued daily decline were discharged on postoperative day 7 without a stoma. Patients with an elevated day-3 CRP or a subsequent rise of more than 15 mg/L per day underwent contrast-enhanced pelvi-abdominal CT. Suspicious CT findings, such as peri-anastomotic air or fluid collection, prompted ileostomy formation at that point. Unsuspicious findings led to continued daily CRP monitoring and repeat non-contrast CT every other day until the patient either stabilized and was discharged without a stoma, or an ileostomy became necessary due to persistently high CRP or evolving suspicious CT findings. Patients with findings suspicious for leakage underwent surgical reintervention for assessment and formation of a diverting loop ileostomy. The reintervention was performed by laparoscopy. The pelvis and anastomotic region were assessed for an anastomotic defect, enteric contamination, purulent fluid, and pelvic collection. No intraoperative air-leak testing, endoscopy, rectal contrast, or other assessment was performed due to fear of iatrogenic injury at the anastomosis line. Peritoneal lavage and pelvic drainage were undertaken when indicated. A diverting ileostomy was created according to the predefined protocol because the imaging and inflammatory findings were considered sufficiently concerning to justify protective diversion.

If the initial contrast-enhanced CT did not demonstrate findings considered sufficiently concerning to justify diversion, patients underwent continued daily clinical assessment and serial CRP monitoring. Repeat imaging every other day was considered when clinical concern persisted or increased, including worsening abdominal or pelvic symptoms, fever, tachycardia, peritoneal signs, persistent ileus, or a persistently rising CRP trajectory. When repeat imaging was clinically indicated, a low-dose protocol was used where feasible.

Follow-up

All patients underwent standardized clinical assessment during the index admission, including daily review for abdominal or pelvic pain, fever, tachycardia, ileus, peritonitis, purulent or enteric pelvic-drain output, and other features raising concern for anastomotic leakage. Patients were followed weekly for four weeks after discharge then at 3, 6, 9, and 12 months to assess for manifest anastomotic leak, pelvic sepsis, wound healing, bowel function, and ileostomy-related complications. Longer-term follow-up captured time to ileostomy reversal and any late oncological events, including local recurrence. A manifest anastomotic leak is an anastomotic leak that causes obvious symptoms and requires medical or surgical intervention. Anastomotic leak was defined according to the International Study Group of Rectal Cancer as a defect of the intestinal wall at the anastomotic site, including staple or suture lines of a neorectal reservoir, resulting in communication between the intra- and extraluminal compartments [20]. A pelvic abscess in proximity to the anastomosis was also classified as anastomotic leakage. Leak severity was graded according to its impact on management: grade A required no active therapeutic intervention, grade B required active therapeutic intervention without surgical reintervention, and grade C required surgical reintervention.

Patient-reported health-related quality of life was assessed four weeks after the index operation using the authorized Arabic version of the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Core 30, version 3.0 (EORTC QLQ-C30). The questionnaire was administered by an assessor independent of the treating surgical team. The prespecified QLQ-C30 domains reported in this study were global health status/quality of life, physical functioning, and emotional functioning. Scoring was performed according to the EORTC QLQ-C30 Scoring Manual. For each multi-item scale, the contributing item responses were averaged to obtain a raw score and then linearly transformed to a score ranging from 0 to 100. Higher global health status and functional-scale scores represent better quality of life and functioning [21]. Stoma-specific quality of life was assessed using the 20-item Stoma-QoL questionnaire developed by Prieto, Thorsen, and Juul. Each item has four response categories scored from 1 (“always”) to 4 (“not at all”), producing a raw sum from 20 to 80. The raw sum was converted using the validated Rasch conversion table to a final score from 0 (worst stoma-related quality of life) to 100 (best stoma-related quality of life). The questionnaire was administered only to patients with an ileostomy at the four-week assessment. Accordingly, 22 patients in the CRP/CT-guided group and 100 in the routine-diversion group were eligible and evaluable [22].

Outcome measures

The prespecified primary endpoint was stoma-free status throughout the 12-month follow-up, defined as avoidance of any stoma throughout the 12-month follow-up. Any patient who underwent stoma formation during this period was classified as not achieving the primary endpoint. Because diversion was mandatory in the routine-diversion group but selective in the CRP/CT-guided group, the primary endpoint quantified the ability of the selective pathway to avoid temporary stoma exposure.

Secondary outcome measures included: suspected anastomotic leak, defined as fever, tachycardia, gastrointestinal upset, ileus, CRP above the alert threshold or rising, or suspicious CT findings; intraoperative outcomes (operative time, blood loss); postoperative complications; total hospital stay and time to stoma reversal; stoma-related complications (high-output stoma, peristomal skin problems); and patient-reported quality of life, measured using validated questionnaires after 4 weeks from the surgery covering global health status, physical functioning, emotional functioning, and stoma-specific concerns. All outcomes were assessed by the same predefined criteria in both study arms to ensure comparability. Anastomotic leakage, pelvic sepsis, and reoperation were analyzed as exploratory safety outcomes. No non-inferiority margin was prespecified, and the study was not designed or powered to establish non-inferiority or equivalence for these outcomes.

Total operative time was measured from skin incision to final skin closure and included all components of the index procedure, including formation of a diverting loop ileostomy when performed. Estimated intraoperative blood loss represented the total blood loss recorded for the entire operation, including the stoma-construction phase. Because allocation occurred after completion of the colorectal anastomosis and confirmation of a negative intraoperative air-leak test, any between-group difference in these outcomes should be interpreted as reflecting the additional operative steps associated with routine ileostomy construction rather than differences in the laparoscopic low anterior resection, total mesorectal excision, or anastomotic technique.

Statistical analysis

Data were analyzed on an intention-to-treat basis using the IBM SPSS software. Continuous variables were expressed as mean plus or minus standard deviation and compared using the independent t-test; categorical variables were expressed as frequencies and percentages and compared using the chi-square or Fisher exact test as appropriate. Multivariate regression was used to adjust key outcomes for baseline demographic and tumor-related covariates. A p-value below 0.05 was considered statistically significant. Missing data were minimal and mainly related to patients lost to follow-up rather than missing baseline or outcome measurements. Twelve-month follow-up was required only for the prespecified primary endpoint of stoma-free status throughout the 12-month period. No secondary or quality-of-life outcome required 12-month assessment. Secondary outcomes were ascertained at defined earlier time points: intraoperative outcomes at surgery; postoperative and stoma-related complications, manifest anastomotic leak, and length of stay during the index admission; quality of life at four weeks; and time to stoma reversal at reversal. All 13 patients subsequently lost to follow-up (7 CRP–CT-guided, 6 routine-diversion) had already undergone diverting ileostomy before their loss and had therefore already met the primary endpoint criterion of stoma formation, so their primary endpoint was fully determined. Accordingly, no analyzed endpoint contained missing outcome data, and all analyses were based on 100 patients per group, with the exception of the stoma-specific quality-of-life analysis, which by design included only patients who had an ileostomy (22 CRP–CT-guided and 100 routine-diversion). No imputation was performed because no outcome data were missing.

Ethical considerations

The study protocol was reviewed and approved by the Faculty of Medicine, Alexandria University Ethics Committee prior to enrollment (Approval Code: 0305903; Date: December 15, 2022). This trial was prospectively designed and received ethics approval before enrollment commenced. However, it was registered retrospectively on ClinicalTrials.gov (NCT07722533) on July 17, 2026 after recruitment had been completed. Written informed consent was obtained from every participant after a full explanation of the study purpose, procedures, and the potential harms and benefits of each treatment arm. Participants were informed of their right to withdraw from the study at any stage without any effect on their ongoing medical care or their relationship with the treating team. Patient confidentiality was maintained throughout data collection and analysis, and all study procedures conformed to the ethical standards of the 1964 Declaration of Helsinki and its later amendments.

Results

Of 231 patients assessed for eligibility, 31 were excluded (18 did not meet the inclusion criteria and 13 declined to participate), leaving 200 patients who underwent randomization. Patients were allocated equally to the CRP/CT-guided selective ileostomy group (n = 100) and the routine diverting ileostomy group (n = 100). During the 12-month follow-up period, 7 patients in the CRP/CT-guided group and 6 patients in the routine ileostomy group were lost to follow-up. Despite these losses, all 100 patients originally allocated to each arm were retained in the final intention-to-treat analysis, giving a total of 100 patients per group for the primary and secondary outcome assessments (Fig. 1).

Fig. 1.

CONSORT flowchart showing the enrollment, exclusion reasons, allocation, follow up and final analysis groups

The two groups were well matched for age, sex, BMI, smoking, comorbidities, tumor distance from the anal verge, tumor length, T/N stage, tumor grade, histological type, and CEA level (all p > 0.05). Post-neoadjuvant restaging response, surgery interval, and lymph node yield were also similar between groups (Tables 1 and 2, and 3).

Table 1.

Comparison between the two studied groups according to demographics data

	CRP/CT (n = 100)	Diverting Ileostomy (n = 100)	Test of Sig.	p	Effect size (95% C.I)
Age (years)
Min. – Max.	49.0–71.0	49.0–71.0	t =  1.092	0.276	Cohen’s d=-0.154 (-0.432–0.123)
Mean ± SD.	58.80 ± 6.90	59.85 ± 6.69
Median (IQR)	56.50(53.0–65.50)	59.0 (54.0–65.50)
Sex
Male	61 (61.0%)	59 (59.0%)	χ2= 0.083	0.773	Phi = 0.020 (-0.116–0.148)
Female	39 (39.0%)	41 (41.0%)
BMI (kg/m2)
Min. – Max.	25.80–32.40	24.50–32.30	t =  1.804	0.073	Cohen’s d = 0.255 (-0.023–0.533)
Mean ± SD.	28.90 ± 1.68	28.45 ± 1.83
Median (IQR)	28.65(27.75–30.05)	28.30(27.25–29.70)
Smoking	27 (27.0%)	30 (30.0%)	χ2 = 0.221	0.638	Phi = 0.033 (-0.108–0.173)
Comorbidities	42 (42.0%)	39 (39.0%)	χ2 = 0.187	0.666	Phi = 0.031 (-0.167–0.103)
DM	20 (20.0%)	22 (22.0%)	χ2 = 0.121	0.728	Phi = 0.025 (-0.120–0.163)
HTN	21 (21.0%)	18 (18.0%)	χ2 = 0.287	0.592	Phi = 0.038 (-0.183–0.100)
IHD	7 (7.0%)	6 (6.0%)	χ2 = 0.082	0.774	Phi = 0.020 (-0.153–0.121)
CKD	3 (3.0%)	1 (1.0%)	FEp	0.621	Phi = 0.071 (-0.173–0.073)

U: Mann Whitney test, χ²: Chi square test, FE: Exact p value for Fisher Exact test

IQR Inter quartile range, SD Standard deviation, t Student t-test

p: p value for comparing between the studied groups

Table 2.

Comparison between the two studied groups according to baseline characteristics data

	CRP/CT (n = 100)	Diverting Ileostomy (n = 100)	Test of Sig.	p	Effect size (95% C.I)
Distance from Anal Verge (cm)
Min. – Max.	5.0–9.0	5.0–9.0	U =  4826.00	0.649	r = 0.032 (-0.107–0.170)
Mean ± SD.	5.97 ± 0.90	6.09 ± 1.07
Median (IQR)	6.0 (5.0–6.0)	6.0 (5.0–7.0)
Tumor Length (cm)
Min. – Max.	2.0–5.0	2.0–5.0	U =  4923.500	0.838	r = 0.014 (-0.125–0.153)
Mean ± SD.	3.25 ± 0.76	3.22 ± 0.72
Median (IQR)	3.0 (3.0–4.0)	3.0 (3.0–4.0)
T Stage (TNM)
T1-2	35 (35.0%)	37 (37.0%)	χ2= 0.195	0.978	Cramér’s V = 0.031 (0.033–0.220)
T3a	19 (19.0%)	18 (18.0%)
T3b	23 (23.0%)	21 (21.0%)
T3c	23 (23.0%)	24 (24.0%)
N Stage (TNM)
N0	16 (16.0%)	19 (19.0%)	χ2= 0.555	0.817MC	Cramér’s V = 0.049 (0.015–0.205)
N1	81 (81.0%)	79 (79.0%)
N2	3 (3.0%)	2 (2.0%)
Stage
IIA	16 (16.0%)	20 (20.0%)	χ2= 0.781	0.885 MC	Cramér’s V = 0.059 (0.036–0.233)
IIIA	35 (35.0%)	34 (34.0%)
IIIB	46 (46.0%)	44 (44.0%)
IIIC	3 (3.0%)	2 (2.0%)
Tumor to mesorectal fascia (mm)	(n = 65)	(n = 65)
Min. – Max.	3.0–17.0	3.0–17.0	U =  2011.500	0.635	r = 0.034 (-0.106–0.172)
Mean ± SD.	7.26 ± 4.26	7.49 ± 4.13
Median (IQR)	6.0 (3.0–10.0)	6.0 (4.0–11.0)
Number of mesorectal suspicious node
Min. – Max.	0.0–6.0	0.0–5.0	U =  4997.500	0.995	r = 0.000 (-0.138–0.139)
Mean ± SD.	1.80 ± 1.08	1.75 ± 1.05
Median (IQR)	2.0 (1.50–2.0)	2.0 (1.0–2.0)
CEA (baseline)
Min. – Max.	3.0–18.0	3.0–18.0	t =  1.257	0.210	Cohen’s d = 0.178 (-0.100–0.455)
Mean ± SD.	10.63 ± 3.89	9.94 ± 3.87
Median (IQR)	11.0 (7.50–13.0)	10.0 (7.0–13.0)

U: Mann Whitney test, χ²: Chi square test, MC: Monte Carlo Exact p-value

IQR Inter quartile range, SD Standard deviation, t Student t-test

p: p value for comparing between the studied groups

Table 3.

Comparison between the two studied groups according to preoperative characteristics and operative details

		CRP/CT (n = 100)	Diverting Ileostomy (n = 100)	χ2	p	Effect size (95% C.I)
Surgery Interval (week)
Min. – Max.		7.0–10.0	7.0–10.0	U =  4979.500	0.958	r = 0.004 (-0.135–0.142)
Mean ± SD.		8.03 ± 0.98	8.06 ± 1.06
Median (IQR)		8.0 (7.0–9.0)	8.0 (7.0–9.0)
Restaging	Response
	No Response	26 (26.0%)	28 (28.0%)	χ2= 0.219	0.896	Cramér’s V = 0.033 (0.017–0.194)
	Partial Response	68 (68.0%)	65 (65.0%)
	Complete Response	6 (6.0%)	7 (7.0%)
	T Stage (TNM)
	T0	6 (6.0%)	7 (7.0%)	χ2= 0.511	0.972	Cramér’s V = 0.051 (0.054–0.245)
	T1-2	65 (65.0%)	62 (62.0%)
	T3a	17 (17.0%)	16 (16.0%)
	T3b	8 (8.0%)	10 (10.0%)
	T3c	4 (4.0%)	5 (5.0%)
	N Stage (TNM)
	N0	35 (35.0%)	37 (37.0%)	χ2= 0.087	0.768	Phi = 0.021 (-0.164–0.115)
	N1	65 (65.0%)	63 (63.0%)
	Stage
	I	22 (22.0%)	26 (26.0%)	χ2= 2.359	0.670	Cramér’s V = 0.109 (0.066–0.288)
	IIA	7 (7.0%)	4 (4.0%)
	IIIA	43 (43.0%)	36 (36.0%)
	IIIB	22 (22.0%)	27 (27.0%)
	CR	6 (6.0%)	7 (7.0%)
Operative details	Operative Time (min)
	Min. – Max.	121.0–178.0	141.0–189.0	t =  9.491*	< 0.001*	Cohen’s d = 1.342 (1.034–1.648)
	Mean ± SD.	148.43 ± 14.03	166.64 ± 13.09
	Median (IQR)	147.0 (139.0–159.0)	167.0(158.0–176.5)
	Intraoperative Blood Loss (ml)
	Min. – Max.	130.0–250.0	140.0–250.0	U =  3108.500	< 0.001*	r = 0.329 (0.200–0.448)
	Mean ± SD.	168.9 ± 28.24	197.3 ± 39.67
	Median (IQR)	170.0(150.0–185.0)	185.0(160.0–240.0)

U: Mann Whitney test, χ²: Chi square test

IQR Inter quartile range, SD Standard deviation, t Student t-test

p: p value for comparing between the studied groups

*: Statistically significant at p ≤ 0.05 (All tests were two-tailed)

Operative time was significantly shorter in the CRP/CT group than the diverting ileostomy group (mean 148.4 ± 14.0 vs. 166.6 ± 13.1 min, p < 0.001). Intraoperative blood loss was also lower in the CRP/CT group (168.9 ± 28.2 vs. 197.3 ± 39.7 mL, p < 0.001) (Table 3).

The prespecified primary endpoint was avoidance of any stoma throughout the 12-month follow-up.

The primary endpoint, stoma-free status throughout the 12-month follow-up, was markedly increased with the CRP/CT-guided approach. In the CRP/CT-guided group, 78 of 100 patients did not undergo ileostomy during the study period, whereas 22 underwent selective ileostomy. Among CRP/CT patients, 26.0% met the CRP trigger threshold and 26.0% underwent confirmatory CT, of which 84.6% showed features suspicious for leak; ileostomy in this group was created at a median postoperative day 4 (IQR 4–5) (Table 4).

Table 4.

Comparison between the two studied groups according to pathology & postoperative daily monitoring assessment data

	CRP/CT (n = 100)	Diverting Ileostomy (n = 100)	Test of Sig.	p	Effect size (95% C.I)
Histological Type
Classic Adenocarcinoma	91 (91.0%)	93 (93.0%)	χ2= 1.052	0.799MC	Cramér’s V = 0.074 (0.010–0.188)
Mucinous Adenocarcinoma	8 (8.0%)	7 (7.0%)
Signet-Ring Cell Carcinoma	1 (1.0%)	0 (0.0%)
Tumour Grade
Poorly Differentiated	20 (20.0%)	18 (18.0%)	χ2= 0.132	0.936	Cramér’s V = 0.026 (0.019–0.196)
Moderately differentiated	73 (73.0%)	75 (75.0%)
Well-differentiated	7 (7.0%)	7 (7.0%)
Lymphovascular invasion	16 (16.0%)	18 (18.0%)	χ2 = 0.142	0.707	Phi = 0.027 (-0.109–0.161)
Perineural invasion	24 (24.0%)	22 (22.0%)	χ2 = 0.113	0.737	Phi = 0.024 (-0.161–0.124)
Number of LN Harvested
Min. – Max.	11.0–14.0	11.0–14.0	U =  3640.00	< 0.001*	r = 0.247 (0.112–0.373)
Mean ± SD.	12.46 ± 0.83	12.88 ± 0.95
Median (IQR)	12.0 (12.0–13.0)	13.0 (12.0–14.0)
Number of LN Positive
Min. – Max.	0.00–3.0	0.00–3.0	U =  4865.00	0.730	r = 0.024 (-0.115–0.163)
Mean ± SD.	1.38 ± 1.07	1.34 ± 1.01
Median (IQR)	2.0 (0.00–2.0)	1.0 (0.00–2.0)
CRP Trigger Met	26 (26.0%)	–	–	–	–
Serum CRP (mg/L)	8 (8.0%)	–	–	–	–
CRP Trajectory
Declining	43 (43.0%)	–	–	–	–
Stable	39 (39.0%)	–
Rising	18 (18.0%)	–
CT Performed	26 (26.0%)	–	–	–	–
Leak Suspicion	22 (84.6%)	–	–	–	–
Ileostomy	22 (22.0%)	100 (100.0%)	χ2= 127.869	< 0.001*	Phi = 0.800 (0.731–0.869)
POD of Ileostomy	(n = 22)
Min. – Max.	3.0–6.0	–	–	–	–
Mean ± SD.	4.41 ± 0.96	–
Median (IQR)	4.0 (4.0–5.0)	–

χ²: Chi square test, MC: Monte Carlo Exact p-value

IQR Inter quartile range, SD Standard deviation, U: Mann Whitney test

p: p value for comparing between the studied groups

*: Statistically significant at p ≤ 0.05 (All tests were two-tailed)

Overall complication rates were comparable between groups (31.0% CRP/CT vs. 32.0% diverting ileostomy, p = 0.879), including wound infection, ileus, chest complications, and urinary problems. Manifest anastomotic leak was rare and not different between groups (0% vs. 1.0%, p = 1.000). In contrast, stoma-related complications were far more frequent with diverting ileostomy (30.0% vs. 3.0%, p < 0.001), including high-output stoma (8.0% vs. 1.0%, p = 0.035) and peristomal skin complications (15.0% vs. 2.0%, p = 0.001). At ileostomy reversal, complications occurred only in the diverting ileostomy group (17.0% vs. 0%, p < 0.001), mainly wound infection (11.0%). Index hospital stay was longer in the CRP/CT group, reflecting the extra observation period needed for serial CRP monitoring (7.62 ± 1.08 vs. 3.97 ± 0.70 days, p < 0.001). Time to stoma reversal did not differ significantly between the small subgroup of CRP/CT patients who received an ileostomy and the diverting ileostomy group (7.95 ± 0.84 vs. 7.89 ± 1.03 weeks, p = 0.953) (Table 5).

Table 5.

Comparison between the two studied groups according to outcomes & complications

Outcomes & complications	CRP/CT (n = 100)	Diverting Ileostomy (n = 100)	Test of Sig.	p	Effect size (95% C.I)
Complications
No	69 (69.0%)	68 (68.0%)	χ2= 0.023	0.879	Phi = 0.011(-0.140–0.157)
Yes	31 (31.0%)	32 (32.0%)
Wound Infection	13 (13.0%)	12 (12.0%)	χ2 = 0.046	0.831	Phi = 0.015(-0.160–0.123)
Ileus	3 (3.0%)	7 (7.0%)	χ2 = 1.684	0.194	Phi = 0.092(-0.039–0.220)
Chest Complications	5 (5.0%)	4 (4.0%)	FEp	1.000	Phi = 0.24 (-0.159–0.122)
Urinary Tract Infection	6 (6.0%)	6 (6.0%)	χ2 = 0.000	1.000	Phi = 0.000(-0.137–0.142)
Urinary Retention	4 (4.0%)	3 (3.0%)	FEp	1.000	Phi = 0.027(-0.162–0.122)
Bowel obstruction	0 (0.0%)	1 (1.0%)	FEp	1.000	Phi = 0.071 (0.062–0.139)
Clavien-Dindo Grade	(n = 32)	(n = 40)
Min. – Max.	1.0–2.0	1.0–3.0	U =  543.00	0.192	r = 0.092 (-0.047–0.228)
Mean ± SD.	1.56 ± 0.50	1.73 ± 0.51
Median (IQR)	2.0 (1.0–2.0)	2.0 (1.0–2.0)
Ileostomy Complications
No	97 (97.0%)	70 (70.0%)	χ2= 26.456	< 0.001*	Phi = 0.364 (0.256–0.457)
Yes	3 (3.0%)	30 (30.0%)
High-Output Stoma	1 (1.0%)	8 (8.0%)	FEp	0.035*	Phi = 0.169 (0.041–0.254)
Peristomal Skin Complication	2 (2.0%)	15 (15.0%)	10.865*	0.001*	Phi = 0.233 (0.122–0.328)
Stoma Prolapse	0 (0.0%)	2 (2.0%)	FEp	0.497	Phi = 0.101 (0.065–0.162)
Stoma Retraction	0 (0.0%)	4 (4.0%)	FEp	0.121	Phi = 0.143 (0.069–0.212)
Stoma Necrosis	0 (0.0%)	1 (1.0%)	FEp	1.000	Phi = 0.071 (0.063–0.137)
Hospital Stay
Min. – Max.	7.0–11.0	3.0–5.0	U =  100.00	< 0.001*	r = 0.877 (0.840–0.905)
Mean ± SD.	7.62 ± 1.08	3.97 ± 0.70
Median (IQR)	7.0 (7.0–8.0)	4.0 (3.0–4.0)
Stoma Reversed	22 (100.0%)	100 (100.0%)	–	–	–
Interval to Reversal (week)	(n = 22)
Min. – Max.	6.0–9.0	5.0–9.0	U =  1091.500	0.953	r = 0.004 (-0.135–0.143)
Mean ± SD.	7.95 ± 0.84	7.89 ± 1.03
Median (IQR)	8.0 (7.0–9.0)	8.0 (7.0–9.0)
Reversal Complications
No	100 (100.0%)	83 (83.0%)	χ2= 18.579*	< 0.001*	Phi = 0.305 (0.230–0.375)
Yes	0 (0.0%)	17 (17.0%)
Bowel Obstruction	0 (0.0%)	1 (1.0%)	FEp	1.000	Phi = 0.071 (0.063–0.133)
Manifest Anastomotic Leak	0 (0.0%)	1 (1.0%)	FEp	1.000	Phi = 0.071 (0.062–0.138)
Wound infection	0 (0.0%)	11 (11.0%)	χ2 = 11.640	0.001*	Phi = 0.241 (0.163–0.310)
Ileus	0 (0.0%)	4 (4.0%)	FEp	0.121	Phi = 0.143 (0.072–0.212)
Local Recurrence	0 (0.0%)	0 (0.0%)	–	–	–

χ²: Chi square test, MC: Monte Carlo Exact p-value

IQR Inter quartile range, SD Standard deviation, U: Mann Whitney test

p: p value for comparing between the studied groups

*: Statistically significant at p ≤ 0.05 (All tests were two-tailed)

Patient-reported outcomes strongly favored the CRP/CT strategy. Global health status/QoL scores were markedly higher in the CRP/CT group (80.6 ± 22.1 vs. 41.1 ± 12.0, p < 0.001), as were physical functioning (69.3 ± 26.7 vs. 21.8 ± 7.5, p < 0.001) and emotional functioning scores (64.1 ± 22.7 vs. 29.6 ± 12.8, p < 0.001) (Table 6).

Table 6.

Comparison between the two studied groups according to patient-reported outcomes

Patient-reported outcomes	CRP/CT (n = 100)	Diverting Ileostomy (n = 100)	Test of Sig	p	Effect size (95% C.I)
Global health status / QoL
Min. – Max.	16.67–100.0	16.67–83.33	U =  1066.00	< 0.001*	r = 0.690 (0.610–0.757)
Mean ± SD.	80.58 ± 22.10	41.08 ± 12.04
Median (IQR)	91.67 (83.33–91.67)	41.67 (33.33–41.67)
Physical functioning
Min. – Max.	13.33–100.0	0.00–40.0	U =  1069.00	< 0.001*	r = 0.692 (0.612–0.758)
Mean ± SD.	69.33 ± 26.73	21.80 ± 7.51
Median (IQR)	80.0 (60.0–86.67)	26.67 (20.0–26.67)
Emotional functioning
Min. – Max.	0.00–100.0	0.00–58.33	U =  1226.00	< 0.001*	r = 0.660 (0.574–0.732)
Mean ± SD.	64.08 ± 22.74	29.58 ± 12.78
Median (IQR)	66.67 (58.33–83.33)	33.33 (25.0–33.33)
Stoma-QoL Score	(n = 22)	(n = 100)
Min. – Max.	10.0–21.67	6.67–21.67	t =  0.989	0.325	Cohen’s d = 0.233 (-0.230–0.695)
Mean ± SD.	15.68 ± 3.28	14.75 ± 4.14
Median (IQR)	15.0 (13.33–18.33)	15.0 (11.67–18.33)

U: Mann Whitney test

IQR Inter quartile range, SD Standard deviation, t: Student t-test

p: p value for comparing between the studied groups

*: Statistically significant at p ≤ 0.05 (All tests were two-tailed)

On multivariate regression, group assignment was the only independent predictor of both operative time (B = − 17.9, p < 0.001) and blood loss (B = − 30.5, p < 0.001), after adjusting for age, sex, BMI, comorbidities, tumor stage, grade, and nodal involvement. In addition, avoidance of diverting ileostomy remained an independent predictor of better global health status (B = 40.7), physical functioning (B = 49.2), and emotional functioning (B = 35.8), all p < 0.001, independent of age, stage, and tumor grade. Stoma-specific QoL scores did not differ between the small ileostomy subgroup of CRP/CT patients and the diverting ileostomy group (15.68 ± 3.28 vs. 14.75 ± 4.14, p = 0.325) (Table 7).

Table 7.

Univariate and multivariate linear regression analysis for the key outcomes of CRP/CT versus the reference diverting ileostomy (n = 200)

Key Outcome	Univariate		Multivariate
	p	B (95% CI)	p	B (95% CI)
Operative Time	< 0.001*	-18.210 (-21.993 – -14.427)	< 0.001*	-17.854 (-21.743 – -13.965)
Intraoperative Blood Loss	< 0.001*	-28.400 (-38.003 – -18.797)	< 0.001*	-30.490 (-40.564 – -20.415)
Global health status / QoL	< 0.001*	39.500 (34.538–44.462)	< 0.001*	40.712 (35.623–45.801)
Physical functioning	< 0.001*	47.532 (42.056–53.008)	< 0.001*	49.236 (43.625–54.847)
Emotional functioning	< 0.001*	34.502 (29.357–39.646)	< 0.001*	35.766 (30.418–41.114)

B Unstandardized Coefficients, C.I Confidence interval (Lower limit – Upper Limit)

*: Statistically significant at p ≤ 0.05 (All tests were two-tailed)

Full univariable and multivariable linear regression analyses of factors associated with clinical and patient-reported outcomes are presented in the Supplementary Tables.

Discussion

This randomized controlled trial shows that a CRP/CT-guided selective ileostomy strategy can markedly cut the rate of diverting ileostomy after laparoscopic low anterior resection for rectal cancer, from 100% under routine practice to only 22%. This finding supports a growing shift in colorectal surgery away from blanket protective stoma use and toward risk-based, individualized decision-making [23, 24]. Serial postoperative CRP was used as a prediction tool to identify patients at increased risk of anastomotic leakage and to trigger further evaluation; it was not considered diagnostic in isolation. Contrast-enhanced CT was then used as a targeted early detection tool in patients with an abnormal CRP trajectory, allowing identification of radiological features suggestive of leakage and facilitating timely intervention. Final leak classification was based on the overall clinical course, imaging, operative findings, and follow-up assessment.

Our results agree with earlier work showing that CRP is a reliable early biomarker for anastomotic leak, with a high negative predictive value that allows safe exclusion of leak in most patients [11, 12]. We used a conservative cutoff of 114 mg/L on postoperative day 3 [25], lower than many published thresholds of 140 to 190 mg/L [11, 12], to prioritize patient safety over further reduction in ileostomy rate. This cautious threshold likely explains why 26% of our CRP/CT group needed confirmatory CT, higher than some other selective protocols, but it also kept our leak-related morbidity low.

Adding CT to CRP monitoring, rather than relying on CRP alone, strengthened our diagnostic pathway. CT is useful mainly to confirm suspected leak once CRP crosses the alarm threshold, since a positive scan for peri-anastomotic air or fluid substantially raises diagnostic confidence [26]. However, CT is not perfect: sensitivity in coloanal anastomoses can be as low as 67%, meaning some early leaks may be missed on a first scan [26]. This is why our protocol repeated non-contrast CT every other day in patients with unsuspicious initial findings but persistently high CRP, rather than relying on a single negative scan to rule out leak, since a false-negative early CT has been linked to delayed reintervention and worse outcomes [14].

Similar dual-marker strategies combining CRP with imaging have been explored recently, and multiple studies confirm that a multimodal, individualized approach improves early detection of leak compared with either test alone [27, 28]. Comparable frameworks using pelvic and risk-factor-based selection of ileostomy have shown that a majority of patients can safely avoid stoma formation when leak risk is assessed systematically rather than applied universally [10, 23].

Reducing unnecessary ileostomies produced a clear quality-of-life benefit in our cohort, consistent with prior evidence that a stoma is linked to worse physical, emotional, and global health-related quality of life [8, 9]. Stoma-related complications, including high-output stoma and peristomal skin problems, were also far less frequent in the selective group. This matches published data showing that stoma reversal itself carries a real complication burden, further disadvantaging the routine strategy [29]. However, an important cost of the selective pathway was that 22 patients underwent a second operation for postoperative ileostomy formation after raised concern for suspicious leakage. These procedures should not be interpreted as confirmed leaks unless a defect, extraluminal communication, or an anastomosis-related pelvic abscess was demonstrated.

Our approach does have a trade-off: the CRP/CT group required a longer index hospital stay to allow serial monitoring, and 84.6% of CT scans performed in this group were suspicious for leak, so most confirmatory scans led to timely ileostomy rather than false alarm. This finding is consistent with reports that a well-designed CRP/CT algorithm can identify leak early enough to avoid a delayed, more dangerous reintervention while still limiting overall stoma exposure [14]. Selective use of ileostomy based on individual risk is now increasingly discussed in colorectal surgery articles and reviews, although routine diversion for very high-risk anastomoses below the peritoneal reflection remains supported by current society guidance [15]. Long-term bowel function after ileostomy avoidance also deserves attention, since a defunctioning stoma itself is a known independent predictor of worse bowel function scores after anterior resection [30].

The incidence of confirmed anastomotic leakage in this trial was unexpectedly low. Contemporary rectal-cancer series have generally reported rates of approximately 10%; for example, the multicenter RALAR study documented leakage in 10.2% of 5,398 patients undergoing anterior resection [31]. Nevertheless, the occurrence of only one confirmed leak among 200 patients requires cautious interpretation. Several features may have contributed to the low observed rate. The study included elective operations performed in a specialized colorectal unit using a standardized laparoscopic technique, routine splenic-flexure mobilization, intraoperative air-leak testing, and postoperative surveillance. Moreover, patients with perforated or obstructed tumors, metastatic or recurrent disease, synchronous colonic resection, or a positive intraoperative air-leak test were excluded. In particular, excluding patients with a positive air-leak test selected a cohort with technically intact anastomoses at the completion of surgery and may have reduced the observed postoperative leakage rate. These factors limit generalizability to an unselected middle- and low-rectal cancer population.

Recent evidence supports individualized diversion while underscoring the need to preserve anastomotic safety. The international PICARR survey demonstrated considerable variation in protective-ileostomy decision-making, although neoadjuvant chemoradiotherapy and an anastomotic height of ≤ 10 cm were each considered sufficient indications for diversion [32]. In a multicenter cohort of 306 rectal cancer patients with protective ileostomy, early and late dehydration–electrolyte imbalance occurred in 16.3% and 13.7%, respectively, and postoperative stoma-related complications independently increased the risk of early dehydration and related readmission [33]. Moreover, a recent SICCR survey found that 52.4% of respondents used protective stomas selectively, alongside widespread use of intraoperative leak testing and increasing adoption of perfusion assessment. Collectively, these observations support reducing avoidable stoma exposure through structured risk assessment, objective evaluation of anastomotic integrity, and close postoperative surveillance rather than routine diversion or reliance on a single diagnostic marker [34].

Limitations of our study include that the primary endpoint was inherently influenced by the allocated interventions because ileostomy was mandatory in the routine-diversion group and selective in the experimental group. Accordingly, the large difference in stoma-free status throughout follow-up should be interpreted as evidence that the protocol reduced exposure to temporary diversion, not as evidence of comparative safety. Furthermore, only one anastomotic leak occurred, and no non-inferiority margin was prespecified. The study was therefore underpowered to exclude clinically important differences in anastomotic leakage, pelvic sepsis, reoperation, or mortality. Moreover, The trial was registered on ClinicalTrials.gov only after recruitment had been completed and therefore did not satisfy prospective-registration requirements and its single-center design could restrict the applicability to institutions employing varying CRP test standards or CT techniques.The decision threshold for postoperative diversion was intentionally conservative and relied partly on CRP/CT abnormalities that are not specific for anastomotic leakage. Consequently, some patients may have undergone protective ileostomy for false-positive or indeterminate imaging findings. Longer-term functional outcomes, including low anterior resection syndrome, were not assessed and merit future study [30]. In addition, preoperative baseline health-related quality-of-life measurements were not obtained. Although randomization reduces the likelihood of systematic baseline differences between groups, the absence of baseline EORTC QLQ-C30 data precluded assessment of within-patient change after surgery. Therefore, the observed differences in global health status, physical functioning, and emotional functioning at four weeks should be interpreted as cross-sectional early postoperative comparisons rather than as definitive evidence of improvement from baseline.

Conclusions

A postoperative CRP/CT-guided selective ileostomy strategy reduced the need for diverting ileostomy after laparoscopic low anterior resection for rectal cancer from 100% with routine practice to only 22%; however, confirmed leakage was infrequent in both groups, and the small number of events precludes drawing conclusions about the comparative safety or non-inferiority of the two approaches. This selective approach also shortened operative time, lowered intraoperative blood loss, and produced far fewer stoma-related complications than routine diversion, while significantly improving patient-reported global health status, physical functioning, and emotional wellbeing. Although the CRP/CT strategy required a longer index hospital stay to allow serial monitoring, this trade-off appears justified by the substantial reduction in ileostomy-related morbidity and the meaningful gains in quality of life. Based on these findings, postoperative CRP/CT-guided selective ileostomy represents an alternative to routine diverting ileostomy after laparoscopic low anterior resection in appropriately monitored surgical units. Larger multicenter trials with longer-term functional and oncological follow-up are recommended to confirm these findings and to refine the optimal CRP threshold and imaging protocol before wider adoption.

Abbreviations

AJCC

American Joint Committee on Cancer

AL

Anastomotic Leak

ASCRS

American Society of Colon and Rectal Surgeons

BMI

Body Mass Index

CEA

Carcinoembryonic Antigen

CI

Confidence Interval

CKD

Chronic Kidney Disease

CONSORT

Consolidated Standards of Reporting Trials

CRP

C-Reactive Protein

CT

Computed Tomography

DM

Diabetes Mellitus

EEA

End-to-End Anastomosis

HTN

Hypertension

IHD

Ischemic Heart Disease

IQR

Interquartile Range

ITT

Intention-to-Treat

LAR

Low Anterior Resection

MRI

Magnetic Resonance Imaging

QoL

Quality of Life

RCT

Randomized Controlled Trial

SD

Standard Deviation

SPSS

Statistical Package for the Social Sciences

TME

Total Mesorectal Excision

TNM

Tumor, Node, Metastasis (staging system)

Authors’ contributions

AMM (Moaz) conceived and designed the study. AMM (Moaz), WE, and YZ contributed to patient recruitment, surgical management, and/or data acquisition. AB, BKH, and AMM (Mohieldein) contributed to data analysis, statistical interpretation, and critical review of the imaging and biomarker protocol. KEB drafted the initial manuscript and performed the statistical analysis. WE and YZ supervised the surgical intervention and clinical follow-up. All authors contributed to data interpretation, critically revised the manuscript for important intellectual content, and approved the final version for submission. All authors agree to be accountable for all aspects of the work.

Funding

No specific funding was received for this study.

Data availability

The data are available upon reasonable request from the corresponding author.

Declarations

Consent for publication

Not applicable.

Ethics approval and consent to participate

The study protocol was reviewed and approved by the Faculty of Medicine, Alexandria University Ethics Committee prior to enrollment (Approval Code: 0305903; Date: December 15, 2022). This trial was prospectively designed and received ethics approval before enrollment commenced. However, it was registered retrospectively on ClinicalTrials.gov (NCT07722533) on July 17, 2026 after recruitment had been completed. Written informed consent was obtained from every participant after a full explanation of the study purpose, procedures, and the potential harms and benefits of each treatment arm. Participants were informed of their right to withdraw from the study at any stage without any effect on their ongoing medical care or their relationship with the treating team. Patient confidentiality was maintained throughout data collection and analysis, and all study procedures conformed to the ethical standards of the 1964 Declaration of Helsinki and its later amendments.

Competing interests

The authors declare no competing interests.